Well Logging and Geology
Content
Foreword
As my previous books got a very good reception by reviewers (G.B. Asquith from AAPG, H.M. Johnson from
Tulane University, A.V. Messineo from SPWLA, M. Verdier from IFP), I have been encouraged to work on a complete
revision and updating.
As a geologist by education, I started my career with a natural turn for geological applications of well logs.
I am more and more convinced of the fundamental link between logging data and geological parameters. I observed
ever since that many geologists do not realize sufficiently the true geological value of well logging data. Well logs
constitute a fantastic source of information on subsurface geology with a vertical resolution that can even reach 1
cm. Needless to say that logs are the eyes and tools of the subsurface geologist. It is the reason why "Well Logging
and Geology" is the first book of a series of three.
Moreover, too often petrophysicists and log analysts tend to restrict the applications of well logs to the determination of effective porosity, water saturation and permeability. The reservoir volume, and subsequently the hydrocarbon
volume in place, and the reservoir complexity are not correctly evaluated as the reservoir is not put back in its precise geological setting. The latter requires the precise determination of its original depositional environment, its diagenetic environment, the nature of its surrounding facies and finally a detailed and accurate description of the tectonic
structure that seismics cannot provide with sufficient resolution and accuracy. This can explain that the success rate
in field development reaches only 80% in favorable cases.
Unfortunately most of the time, production geologists do not have the logging data set which would help them
improving their reservoir description. Too often, oil company managers want to reduce costs on logging acquisition
as much as possible. The objective in itself is reasonable because reducing costs is always a good thing in itself and
in the bottom line companies want to make profit. The problem is that before selecting an asset reduction of 25% of
the logging cost will save only 2.5% of the well cost assuming the logging cost would represent statistically only
10% of the well cost! As a consequence, reduction of logging data, (most of the time natural gamma ray spectrometry and dipmeter or image data) will not allow the precise and accurate 3D description and evaluation of the reservoir. In addition, the error on volume estimation will be certainly much higher than the error on porosity measurement! In the end, the short term economy realized on logging can prove to be disastrous on a long term. An incomplete and imprecise knowledge of the local reservoir geology may result in much higher expenses due to wrong
location of production and injection wells! Do not consider the logs as a commodity but as a very valuable source of
information. And even if the logging measurements are affected by errors, remember, as pointed out by Philippe
Theys (Petrophysics, vol. 44, no 1, p.14, 2003),that "when you evaluate the magnitude of the errors that affect logging measurements, you find that the ones induced by the limitations of physics or technology are smaller than the
human errors". You would not rely on a diagnostic made by a doctor who would only examine your throat, take your
temperature and measure your blood pressure! Why would you trust a reservoir evaluation based on a reduced logging set?
Consequently, do not reduce too much the logging data set! But, at the same time exploit completely the logging
information. Do not leave the logs "sleeping" in their drawers! Analyze and interpret them in depth! Squeeze the
lemon!
The objective of this book is to demonstrate to geologists, petroleum engineers and log analysts what kind of
geological information can be extracted from logging data (a great deal actually!). For that purpose, I compare the
geological and well logging approaches, showing that one can extract from log data the information that generally
one asks to cores.
This "Well Logging and Geology" book will be completed by the book on "Well Logging - Data Acquisition and
Applications" and the book on "Well Logging - Reservoir Evaluation". These three books should demonstrate that
logs are a fundamental source of information for geologists and petrophysicists.
Oberto SERRA
Serralog 8 2003
i
TABLE OF CONTENTS
Foreword
Acknowledgements
Chapter 1
i
ii
GeneraIities
Introduction
Definitions
Rock properties
Rock physical properties
Rock chemical properties
Rock geological attributes
Sources of information
Outcropts
Information provided by subsurface data
Information provided by surface seismics
Information provided by drilling
Goal of well log interpretation
Interpretationmethodology
Log quality control
Tool calibrations
Repeat section
Depth match
Composite-Log
Segmentation in electrobeds
Well and Logging Types
Open holes
Vertical wells
Deviated wells
Horizontal to subhorizontal wells
Cased holes
Conclusion
Workstations
Formation Image Examiner
FracView fracture synergy log
DIAMAGE
GeoFrame Geology
Data preparation
Geology
Petrophysics
References and Bibliography
1
1
1
1
1
2
2
2
2
4
4
5
11
11
18
18
18
20
21
21
23
23
23
23
23
26
26
26
27
27
28
29
29
29
34
34
Chapter 2
Well Logging and Rock Composition
Introduction - Review of general geological concepts
The three major rock types
Igneous rocks
Sedimentary rocks
Detrital or clastic rocks
Chemical and biochemical rocks
Metamorphicrocks
Relative abundance of rocks
Rock classification
Rock composition
Elemental composition
Mineralogicalcomposition
Determination of the rock composition - Traditional method
Elemental and chemical composition
Mineralogicalcomposition
Determination of the rock composition - Well logging approach
Elemental composition
Natural gamma ray spectrometry
Induced gamma ray spectrometry
Determinationof the mineral composition
Exploitation of the elemental composition
Measurements involving interactions of photons
Measurements of the sound wave velocities
Cross-plot interpretation
Cross-plot definition
Cross-plot analysis
37
37
37
37
38
38
38
38
38
40
40
42
45
50
50
52
54
55
55
57
60
61
64
66
67
67
71
2003
Serralog i.:
Formations of very high resistivities
Identification of igneous rocks
Identification of evaporites
Identification of metamorphic rocks
Formations with variable resistivities
Identification of detrital rock type
Identification of biochemical rock type
Automatic lithology determination - LITHO-ROCKLASS
Introduction
Construction of the database
Logs used in LITHO-ROCKLASS
Quantitative interpretation
SERRALOG approach
Platform Express lithology display
Neural network approach
References and Bibliography
73
74
76
85
86
86
92
97
97
98
99
99
101
106
112
114
Chapter 3
Well Logging and Rock Texture
119
Introduction - Review of petrographic concepts
119
Definition
119
Textural components
120
Texture of detrital rocks
120
Grain size
120
Sorting
121
Grain shape
122
Packing
122
Orientation
123
Influence of grain properties
123
Influence of grain size
123
Influence of sorting
124
Influence of shape and roundness of grains
125
Influence of the orientation of grains
125
Influence of packing
125
Influence of the mineral composition of grains
126
Influence of other textural components
126
Texture of carbonate rocks
127
Determinationof textural parameters - Traditional approach
129
Determinationof textural parameters - Well logging approach
131
Detrital rock texture
133
Grain size
133
Sorting
137
Grain orientation
138
Arrangement or packing
139
Grain shape
140
Cement
140
Carbonate rock texture
141
Porosity
141
Porosity and pore size determination from NMR
145
Automatic extraction of textural informationfrom dipmeter 147
148
Quantitative analysis of image data - BorTex program
Influence of textural parameters on reservoir properties
154
References and Bibliography
155
Chapter 4
Well Logging and Rock Structure
Introduction - Review of general geological concepts
Definition
Importance of Sedimentary structures
Classification of sedimentary structures
Primary sedimentary structures
Stratum, bed, lamina description
Shape of sedimentation unit
Nature of bed boundaries
Groups of beds (“bedsets”)
Internal organization of beds
Secondary sedimentary structures
Disconformities
159
159
159
159
159
160
160
160
162
163
163
171
171
1
Intrusions- Dikes
Diapiric structures
Determination of rock structure - Traditional approach
Determination of rock structure - Well logging approach
Bed description
Bed shape and bedding planes
Bed thickness
Nature of bed boundaries
Internal organization of beds
Illustrationof structures by images
Bed boundaries
Internal organization
Graded bedding
Automatic analysis of dipmeter data - The SYNDIP program
The SYNDIP program
Descriptionof a SYNDIP display
Rules for interpretationof dipmeter and image data
Quantitative analysis of image data
Sedimentological applications of sedimentary structure detection
References and Bibliography
171
171
171
172
174
174
175
175
175
180
180
181
183
185
187
188
189
191
191
193
Chapter 5
Well Logging and Facies, Sequence and Environment
Introduction - Review of general geological concepts
Definitions
Importance of facies and sequence analysis
Traditional approach of facies analysis
Well logging approach
Historical
The Electrofaciesconcept
Electrofaciesanalysis from well logs
Manual identification of electrofacies
Automatic electrofacies identification - FACIOLOG
Automatic segmentation of logs
Clustering techniques
Final electrofacies determination
Principal component analysis
Translation of electrofacies to facies
Result presentation
The Electrosequence concept
Sequence analysis from well logs
Gradational transition - Elementary sequence
Abrupt contact
Sequences of electrofacies
Applications of facies and sequence analysis
Reconstitutionof the depositional environment
Interest of the reconstitution of the environment
Recognitionof the environments from logging data
Illustration of environments from logging data
Reconstruction of the geometry of the facies
Mapping of electrofacies
Constitution of a data base of electrofacies
Quantitative interpretation
Choice of a more judicious core sampling for analysis
References and Bibliography
197
197
197
200
201
201
203
206
207
208
211
212
214
216
217
221
223
223
224
225
226
227
227
228
231
231
232
234
235
236
236
236
236
Chapter 6
Well Logging and Diagenesis
Introduction - Review of geological concepts
Definition
The diagenetic environment
Factors affecting diagenesis
The fundamental processes
Physical processes
Chemical processes
Biochemicaland organic processes
Diagenetic changes
Rock components
Compaction
Cementation
Recrystallization
Transformation
Mineralogical replacement
Selective leaching
239
239
239
239
240
241
241
241
241
241
242
242
242
242
242
242
243
11
Authigenesis
Stylolitization
The diagenetic phases
Effects of diagenesis on the sediment properties
The need for diagenetic studies
Diagenetic Studies - Traditional methods
Well logging approach
Diagenesis in detrital sequences
Compaction
Pressure-controlledsolution
Cementation
Authigenesis
Nodules
Pyrite crystals
Epidiagenesis
Diagenesis in carbonate sequences
Crystallization
Neomorphism
Selective leaching
Stylolitization
Cherts and anhydritic nodules
Hard-ground
Diagenesis in volcaniclastic sequences
Diagenesis in shales
Diagenesis of organic material
References and Bibliography
~~
Chapter 7
Well Logging and Compaction
Introduction - Review of geological concepts
Definition
Compaction study
Sand compaction
Carbonate compaction
Shale compaction
Principal stages in shale compaction
Compaction mechanisms
The Hubbert 8 Rubey law
Compaction effects
Effects on shale porosity
Effects on shale density
Effects on chemistry of interstitial fluids
Effects on geothermal gradient
Effects on mineralogicaltransformation
The influence of time
Compaction of organic sediments (peats)
Compaction anomalies
Definition
Origin of compaction anomalies
Zones exhibiting undercompaction
Traditional geological approach
Well logging approach
Associated phenomena
Construction of the compaction profile
Manual plot
Construction of the profile at the well site
DENSON automatic program
Conduct of a study ..
Advantages of the DENSON program
Construction of the normal compaction trend
Detection of undercompacted zones by VSP
Study of sand compaction using well logs
Study of carbonate compaction using well logs
Study of shale compaction using well logs
Sonic velocity in shales
Resistivity of shales
Formation factor
Hydrogen index of shales
Capture cross-section of shales
Natural gamma ray spectroscopy log
Applications of compaction studies
Hydrodynamic applications
Detection of entry into an undercompactedformation
Evaluation of formation pressure
243
243
244
245
245
246
250
250
250
250
251
252
252
253
254
254
255
255
256
257
258
259
259
260
260
262
265
265
265
265
267
270
271
271
273
273
274
274
275
275
276
276
276
277
277
277
277
278
278
278
278
278
278
278
278
279
280
281
281
281
283
283
283
283
283
284
284
284
285
285
285
285
Serralog @ 2003
Evaluation of pressure gradients
Geological applications
Detection of unconformities
Reconstructionof maximum burial depths
Differential compaction
Correlations between wells
Sedimentologicalapplications
Applications to reservoir studies
Geophysical applications
Geochemical applications
Decompaction
Objectives
Procedures
Applications
References and Bibliography
287
289
289
289
290
290
290
292
292
293
293
293
294
294
294
Chapter 8
Well Logging and Tectonics
Introduction
Review of geological concepts
Definition
Concepts of stress
Mechanical behavior of rocks
Elastic behavior
Plastic behavior
Viscous behavior
Factors controlling rock behavior
Confining pressure
Temperature
Time
The actual behavior of rocks
Types of state of stress
Rock strength
General orientation of stresses
The results of stresses : strains
Continuous strain : fold
Tensile stress
Compressive stress
Descriptive elements of a fold
Fold shape
Fold type
Nomenclatureof folds
Discontinuousstrain : fractures and faults
Movements along a fault
Fault description
Fault classification
Determination of tectonic structures - Traditional approach
Determination of tectonic structures from seismic data
Determination of tectonic structures from gamma ray
Determination of tectonic structures from dipmeter or image data
Structural analysis of dip data
Statistical study of dip data
Dip and azimuth histograms
Polar plots
Azimuthal frequency plot
Density stereograms
Construction of density stereograms
Interpretationof these diagrams
Stereographic plotting techniques
Stereographic projection
Principle
Planes and lines on a stereogram
Polar plot
Cyclographic plot
Plotting a line
Construction of the WULFF stereonet
Construction of the SCHMIDT stereonet
Representationof structural elements
Plotting of dip data
Use of stereograms
Methods of Dip data interpretation
Visual arrow-plot analysis
“Quick-look technique : dip patterns
Determining structural dip
299
299
299
299
301
303
303
305
305
306
306
306
306
306
307
307
307
308
308
308
308
309
310
310
310
311
311
311
312
313
31 3
314
315
31 8
320
320
320
320
321
321
322
322
322
322
323
323
323
323
324
324
324
327
328
330
330
330
331
Serralog C 2003
Detecting faults
331
Detecting folds
332
Bengtson technique
332
Automatic and interactive programs
333
DipTrend program
333
Automatic structural dip determination
334
Analysis of borehole images
337
Structural dip versus regional or seismic dip
339
Reconstructionof the tectonic features of a well
339
Cross-section construction
340
Automatic cross-section construction
340
Hypotheses for cross-section construction
341
Similar folds versus parallel folds
341
Constructing cross-sections
342
Selecting a zone
342
Determiningthe tructural axis and translation Dlane 342
Determining the cross-section plane
344
Constructing the cross-section
345
Exploitation of dip data
345
Workstations
350
Other applications of stereographic projection
351
Find the line of intersection of two planes
351
Determine the angle between two planes
351
Determinationof true dip of a bed __.
351
Determination of the apparent dip ...
351
Dip removal
352
References and Bibliography
352
Chapter 9
Well Logging and Fractures
Introduction
Review of general concepts
Relation of fracture to stress orientation
Importance of fractures
Mechanical properties evaluation
Mechanical behavior of the reservoirs Stress computations
Fracture-Pressurecomputations
Dynamic elastic properties
Inherent strength computations
Detection of fractures on cores
Detection of fractures from well logs
Natural gamma radioactivity
Caliper
Temperature
Formation density
Photoelectric capture cross-section
Neutron hydrogen index
Sonic measurements
Effects on body waves
Arrival time
Attenuation of body waves
Normalized Differential Energies (NDE)
Amplitude spike analysis
Crisscross patterns
Effects on Stoneley wave
Resistivities
Dipmeter
Resistivity curves
Azimuth curve of pad 1
Caliper
Dips
Spontaneous potential
Ultrasonic image tools
Resistivity image tools
Fracture evaluation
Depths of fractured zones
Type of fracture
Fracture orientation
Fracture length
Fracture density
Fracture aperture
Fracture porosity from images
Fracture porosity from photoelectric index
355
355
356
356
358
358
358
358
359
359
360
360
360
360
361
362
363
363
363
363
364
364
365
365
366
367
368
369
370
371
372
372
372
373
374
377
377
377
378
378
378
379
379
380
III
Well Logging and Geology
IV
Fracture porosity from dual laterolog (DLL)
Determinationof fracture permeability
Fracture effects on reservoir interpretation
Tortuosity factor m
Saturation evaluation
Formation factor - Porosity relationship
Lithology determination
Core orientation
Hole slippage
Detection of borehole damage in deviated wells
Relationship between stress-induced features and
stress axis
Stress determination
Prediction of the fracture efficiency
RECAP
References and Bibliography
380
381
381
38 I
382
384
384
384
385
385
387
388
390
390
39 1
Chapter 10
Well Logging and Source Rock
Introduction
Traditional geological approach
Well logging approach
The Carbolog method
The uranium concentration
The A log R method
Methods based on cross-plots
Methods based on nuclear measurements
References and Bibliography
395
395
395
396
396
396
396
398
399
400
Chapter I 1
Well Logging and Sequence Stratigraphy
Introduction
Relative dating
Definition of parastratigraphic units
Stratigraphic phenomena
Break detection
Break definition
Break significance
Break classification
Breaks without major lithological change
Change in type of fluid
Textural change
Diagenesis
Tectonic accident
Unconformity
Breaks corresponding to major litholonical chanae
Diagenesis together witha mineraiogical change
Erosion
Unconformity
Ashes, lava flows or volcanic intrusions
Emersion
Transgression
Sequence stratigraphy from well logs
Nomenclature issues
Using log patterns
Simplistic use of log patterns
Surface hierarchy from well logs
Building a conceptual geological model
Methodology
Correlations from well logs
Principe of causality
Concepts of log correlation
Concept of similarity
Concept of rhythmicity
Concept of lateral variability
Intrinsic value of correlations. Concept of reliability
Stratigraphic value of log correlations
Search for time markers
Extension to seismic sections
References and Bibliography
403
403
403
404
404
404
404
404
404
404
405
405
405
405
405
406
406
407
407
408
408
408
409
409
412
413
415
415
415
416
418
418
418
419
419
419
420
420
421
422
-
Index Glossary
Index of referenced authors
Iv
423
425
Serralog 0 2003
ct
W
I-
n
a
I
0
1
GENERALITIES
Introduction
Humanity has constant need for resources provided by
the planet Earth: foods, water, raw materials, ores and
energy, either natural (sun, wind, hydraulic, tidal, geothermal, nuclear) or fossil (coal, petroleum, gas). An on-going
effort in research and exploration is therefore necessary
to discover, exploit and develop them. And, in this effort, it
is better to rely on research, technology and geology than
on hazard and luck!
Definitions
Geology is by definition "the study of the planet Earth.
It is concerned with the origin of the planet, the material
and morphology of the Earth, and its history and the processes that acted (and act) upon it to affect its historic and
present forms" (Bates & Jackson, 1980).
This study, even when it includes fundamental
research, is aimed at a better understanding of the
influences that have formed, transformed and modeled
our planet, of the laws ruling the formation of rocks, their
distribution, their transformation, their deformation, and of
the laws governing the accumulation of those raw materials and ores which are of economic interest. This scientific guidance is available to help us discover the low cost
minerals and energy resources that humanity needs so
much.
The geologist's work consists (Table 1-1):
- in observing and describing, as much as possible
completely and objectively, the rocks attributes and the
geological phenomena;
- in interpreting these observations by comparing them
with those made on recent series or phenomena (existing
models), those coming from the study of reconstructed
models or from laboratory experiments, observations
made on ancient formations which have been studied in
detail and are well understood (ancient models and application of theories of actualism or uniformitarianism as
developed by Hutton, 1788)
In the case of the study of sedimentary rocks, this
interpretation must lead to a reconstitution of the geographic and climatic frame and, consequently, to the understanding of conditions under which the sediments and
rocks are formed. From this reconstitution the geologist
will try:
- to predict those zones most favorable for the accu-
Serralog 0 2003
mulation of mineral resources,
- to specify the spacial extent of these resources,
- to evaluate their volume from the rock mineral resource content estimation.
Petrophysics is the "study of the physical properties of
reservoir rocks" (Bates & Jackson, 1980).
The petrophysicist's work, or log analyst's work
consists - in petroleum industry - in interpreting the physical properties of the reservoir rock crossed by a well, in
order to evaluate the volume of hydrocarbon both in place
and extractible inside a field. This evaluation is necessarily based on a reduced number of points of observation
and measurement (the drilled wells). In each well is measured the local reservoir thickness, and are evaluated its
local effective porosity, saturation, and permeability.
These local data must be extended from each well in
order to evaluate the final hydrocarbon volume.
Therefore, to be more accurate this volume evaluation
requires to put back the reservoir in its precise geological
setting. The latter requires the determination of its original
depositional environment, its diagenetic environment, the
nature of its surrounding facies and a detailed and accurate description of the tectonic structure. This study must
be achieved by a team composed of a geologist, a geophysicist, a petrophysicist and a reservoir engineer. This
team will base its study on the rock properties and the different sources of information.
Rock properties
The rock properties can be classified in three principal
groups.
Rock physical properties
The rock physical properties are numerous. They are
listed hereafter :
hardness, density, ductility, elasticity, compressibility,
magnetism, magnetic susceptibility, magnetic permeability, electrical conductivity, dielectric permittivity, natural
radioactivity, sound travel time, thermal conductivity,
power of gamma ray absorption, neutron interactions,
proton precession, porosity, capillarity, interfacial tension,
permeability, water saturation.
Most of these properties are measured in borehole by
well logging tools. The physical principles of these tools,
1
Table 1-1
Geologist’s work
-
1 Observation and description
of sedimentary bodies
Composltlon
elemental
mineralogical
Texture
-
-
-
2 Comparison with present
or well known old models
4 Prediction
Depositional
Environment
Palaeogeography
Location and
extent of
economical
objects
Determination
of mineral
resources
Volume
estimation
Lithology
Facies
paleocurrent
Thickness
Width
Length
Geometry
Relationship with
surrounding
sedimentary bodies
Evolution
-vertical
lateral
-
Diagenetic phenomena
Structural dip
Folds
Faults
Fractures
Stylolites
-
Sequences
Rhythms
Cycles
Geologic history
Diagenetic sequences
I
External structure
Stresses
their applications and, as well as, the interpretation of the
data they record, are explained by numerous papers and
books (cf. Volume 1).
Rock chemical properties
These properties are fundamentally linked to the minerals and materials which compose the rock. Rocks are
composed of minerals that are more or less stable.
The rock decomposition is due to weathering which is
a twofold process:
- mechanical weathering or rock fragmentation,
- chemical weathering by decay of unstable minerals,
each helping and reinforcing the other, the smaller the
pieces, the greater the surface area available for chemical attack, and the faster the pieces decay.
Some minerals are also easily dissolved under certain
conditions but can be regenerated as such by reverse
conditions: for instance salt, calcite.. .: the less soluble,
the more resistant. Quartz wins hands down.
Stability, as it affects resistance to weathering, is also
a relative property.
The mineral stability follows the reverse of the
Bowen’s reaction series (Fig. 1-1 next page). Fe-oxides
are the most stable. They are successively followed by Aloxides, Quartz, Clay minerals, Muscovite, K-feldspar
(orthoclase), Biotite, Na-feldspar (albite), Amphibole,
Pyroxene, Ca-feldspar (anorthite), and Olivine which is
2
-
3 Interpretation
~
Tectonic style
Structural traps
Location,
nature and
distribution
of permeability
barriers or of
permeability
pathways
the least stable.
Rock geological attributes
The geological attributes of sedimentary rocks include:
- the facies of each depositional unit. This facies is
defined by :
- the composition,
- the texture,
- the sedimentary features or structures,
- the colour,
- the fossils
of the depositional unit;
- the facies succession or sequence,
- the transformations undergone since the deposition
of the sediment, in other terms the observed diagenetic
phenomena,
- the deformations undergone under stress action.
Sources of information
As previously indicated the team in charge of the formation analysis will use several sources of information to
achieve the goals of their studies. These sources are
reviewed hereafter.
Outcrops (Cliffs, quarries, trenches, ditches, tunnels,
mines)
Serralog 0 2003
Generalities
I Chapter 1 I
3
ROCK
TYPE
MB)I
1II)c.
MOST
UNSTABLE
HIGH
FIRST TO
CRYSTALLIZE
[XMC
0
ANORTHrn
0
fca kldrp.f)
\
/
I
HIGH
LOW
R1
DUNITE
PERlDOTlTE
WEAT
PVROXENITE
GABBRO
n.s.0
DlORlTE
>
t
.o
Andrnsn.
B
OUARTZ DlORlTE
0.ClI.
ORANODIORITE
I I
LOU
TEMP.
LASTTO
CIYSTALLIZE
4
1
GRAWITE
UlsCbVrrE
Rhylrolll.
0
I
t
WRTZ
@
LOW
MOST
STlBLE
WEIS:
LOW
NtW
1: single tetrahedron. 2: single chain. 3: double chain.4: sheet. 5: 3-D lattice
Figure 1-1 - The Bowen's reaction series and its interpretation both for rock type formation and for weathering intensity and results
(adapted from Leet et al., 1982).
In certain types of research (metallic ores), natural or
man-made outcrops are still the essential source of geological information (Fig. 1-2). In others (petroleum, coal,
geothermal resources) outcrop information has been progressively replaced by drilling data or completed by surface geophysics (gravimetry, seismic surveying, magnetism, electric and telluric measurements), or borehole geophysics, the latter including Wireline Logging (WL) or
Logging While Drilling (LWD).
On land, the basin exploration starts generally by a
sedimentologic, stratigraphic and tectonic study of outcrops surrounding the basin in order to determine its
hydrocarbon potential, it means the co-existence of reservoir-rocks, source-rocks, seal-rocks and traps. The analysis of rock samples allows the verification of this co-existence. The samples are collected with the required size.
They are easily located in their environment. It is generally easy to laterally follow the beds from which they are
extracted, and determine their geometry and the one of
their surrounding beds. Aerial and satellite photographs
can help in this work. The samples can be replaced in the
vertical sequence of deposition. At the end, it is possible
to practice all the analyses necessary to describe, evaluate and date them. It is also quite easy to come back, resample and re-evaluate them.
In sea, the basin evaluation requires fundamentally
geophysical techniques completed by a few drillings.
In any case, more generally, the geological knowledge
of a basin, on land or offshore, would not be as complete
and precise if it did not included information provided by
specialties such as surface geophysics, geochemistry,
petrology, mineralogy, paleontology, sedimentology, or
Serralog 0 2003
I
Figure 1-2 - Observation and description of outcrop
(Mt Saint-Odile, Vosges, France, photo 0. Serra).
3
petrophysical measurements realized in a well, such as
well logs, borehole seismics and tests. This is because in
modern exploration, drilling and geophysical techniques
are more often used, not only in the petroleum industry,
but also in subsurface storage, or to discover coal, uranium or even some metallic ores, and in geothermal
resources. With deeper targets, outcrops, as a source of
information, are less frequently used because extrapolations established from them are less reliable. Moreover,
the geological complexity of targets increases (stratigraphic traps, fluid permeability barriers, size and depth of
structures, ...).
As a result, at the present time, most knowledge concerning geological sedimentary basins (especially deep
basins) comes from drilling and geophysics. As mentioned by de Golyer (1921), Lahee (1932) and after them
by llling (1946) : "if geology has contributed greatly to the
growth of the oil industry the debt is not one-sided one.
Geology owes a great deal to the oil industry in the expansion of its knowledge and the increased efficiency of its
methods".
Information provided by subsurface data
Information provided bv surface seismics
This branch of Earth's science is fundamental for many
reasons. It is the only one that allows continuous study of
formations in subsurface. It completes our perception of
these formations on the outcrops. Two and, more often,
three dimensional pictures of subsurface can be obtained
by today's surface geophysical techniques (Fig. 1-3).
They are extremely important tools for the exploration of
subsurface, since it gives direct information, not only on
the shape and arrangement of beds (Fig. I-4), but also on
their nature, their petrophysical properties and even
sometimes their fluid content (seismofacies, "bright spots"
Fig. 1-5).
Unfortunately, seismic vertical resolution is close to 20
meters (Islam, 2000), 10 meters at best at low depth. In
many cases, at the reservoir scale, it is not sufficient to
describe it correctly. In addition, the hypotheses drawn by
the interpretation of this information must be verified by
Figure 1-3 - Example of geological interpretation of surface seismic data in terms of depositional environment and tectonic setting
(courtesy of GECO).
4
Serralog 0 2003
Generalities
drilling. The translation of surface geophysical data into a
geological interpretation will be considerably easier and
more reliable if it is supported by well log measurements.
In other words, well log data provide the link between geophysics and geology (Fig. 1-6). The former has thus to be
correlated with the latter. Well logs are the only means for
providing an accurate transfer of time data to depth data.
I Chapter 1 I
5
a
b
Figure 1-4 - From this time slice representation of the seismic data, it is
easy to recognize a meandering channel system.
C
Figure 1-5 - On this seismic cross-section the gadoil and oil/water
contact are detected thanks to the changes in wave attributes
(amplitude, phase).
They allow for the transfer of amplitude and signal frequency data to sedimentological or economic data
(facies, porosity, fluid content, Fig. 1-7).
d
information provided bv drillinq
Two kinds of data are obtained during drilling
those linked to rock samples (full diameter cores,
sidewall cores, cuttings), and to fluid samples;
- those provided by physical measurements made in
drilled holes essentially by logging tools.
-
Core information
Standard cores, due to their size and if they are continuous, Constitute a sampling Of good quality similar to
the one collected on outcrops. They will allow all types of
analyses, providing information at microscopic scale
electron microscope
is
(grain and pore size) if scanning
used (Fig. 1-8).
Serralog 0 2003
Figure 1-6- a) Surface seismic section (minimum phase). 6) Acoustic
impedance vs porosity. c) Correlation between logs, formation impedances, stratigraphy and VSP (minimum phase). d) Acoustic impedanve vs porosity, the colour coding indicating the depth of each point,
shallow depth in blue/green, deeper depth as red/yellow
(adapted from Poster, 1988).
5
Figure 1-7a - The well log impedance model, extracted wavelet, and
cross-plot information have been combined to indicate the probable
effect on the surface seismic section of a porosity increase.
1500
Figure 1-7b - The seismic events now correlate with the main formation
boundaries with a higher resolution than the original data of Fig. 1-5a
(adapted from Poster, 1988).
However, the bed geometry will not be determined due
to the lack of dip and azimuth measurements of bed and
borehole, their size and the lack of lateral continuity.
Moreover, cores may also have drawbacks such as
partial recovery or none at all, sample damages linked to
release of the stress after their extraction, and the types
of processing they support during all the steps of analysis
which may modify significantly their petrophysical properties (cleaning, drying, ...).
Figure 1-8 - SEM photograph showing quartz grains and “books”of
kaolinite. It gives also an idea of the pore size, pore geometry and clay
distribution.
In some cases, there is considerable uncertainty over
the depth of a given cuttings sampling and furthermore it
can be difficult to restore the constituents and thickness of
the lithological column from cuttings alone. This is due to
mud swirling, caving, the loss of some constituents (siltysize grains, salts) during washing or by total lost-circulation events. The reduced dimension of this type of rock
samples does not generally allow a full analysis, and
makes an observation incomplete.
Therefore, geologists and, in particular, subsurface
geologists are often short of, or deprived of representative or good quality samples. Sidewall cores can partly
compensate for this shortcoming but, because of their
small size, observations or measurements may still be
inaccurate compared to those carried out on larger samples.
Consequently:
- Geologists and petrophysicists are unable to respond
to the fundamental question that one asks in hydrocarbon
exploration: does it exist economic hydrocarbon accumulations in a well?
- They cannot make certain studies yet necessary for
the basin exploration or field development pursue.
Well loaaina information
Cuttina information
Unfortunately, for economic and technical reasons,
coring can be a rare operation, particularly under certain
drilling conditions or in certain types of formations.
Therefore, often the only rock sample available consists
of the cuttings obtained during drilling, or sidewall cores
cut with the help of a rotary or of a bullet core barrel which
runs with a wireline.
6
Fortunately, borehole geophysical techniques allow
geologists to fill the gap represented by the lack of representative rock samples. Between those techniques well
logging occupies a special and very important place. This
is due to the quality and diversity of tools of acquisition
and methods of interpretation. Even, thanks to technical
progresses it is now possible to obtain physical measurements during drilling. The vertical resolution of these
measurements ranges from few meters (Vertical Seismic
Serralog 0 2003
Generalities
Profile) to about one centimeter (resistivity imaging
tool).
Well logs are of special interest in that:
-they provide the only source of data to give accurate
information on the depth and the apparent, and even real,
thickness of beds if a dipmeter or image data have been
recorded, thanks to the borehole deviation and dip of bed
boundaries they provide.
- They give a nearly continuous analysis of the formations (one sampling every 15 cm for traditional open hole
logs; this frequency can be increased to one sample
every 2.5 cm if required, and one sampling every 5 mm or
even 2.5 mm for dipmeter and image tools). In contrast,
information extracted from cores is more discontinuous,
and often scattered in depth. Even in the case of continuous coring, all analyses are not systematically made on
each plug selected on the cores (a plug each foot [30
cml).
-They generally analyze a volume of rock that is oftengreater than the one represented by a core or plug, and
consequently than a cutting. Consequently, they are more
representative of the mean properties of the rock, especially in heterogeneous rocks. But, at the same moment
they provide a very detail description of the formations if
images are recorded (resistive objects of 1 cm size and
conductive features as thin as 1 micron [lpm] are detected).
- They give a vision of the formations, at the scale of
meter for standard logs and centimeter for images, indeed
particular but nearly continuous and objective.
The information logs provide is
. quantitative and, consequently, it allows us to think
about geological objects represented by well logging
measurements by using the full the computer's capacity to
process the information;
. precise, even if, sometimes, errors are present as in
any measurement;
- objective and repetitive;
- permanent; whereas the cores are destroyed for analysis, preventing any further study, log data can be reinterpreted with new ideas, new techniques, or new parameters;
- obtained rapidly, even on the well site;
- economic. Coring and core analysis and description
are expensive and time consuming, and the desired information is obtained only several weeks later. Do not forget
that the wireline logging cost represents generally between 5 and 10% of the drilling cost and that logs provide
approximately 90% of the information extracted from a
well (Fig. 1-9). A reduce logging set represents certainly a
short-term economy, but can generate much more development cost on a long-term basis!
- The measurements made with logging tools are
strongly dependent on geological parameters (Table 1-2).
Consequently, the information they provide is of the
utmost interest for geologists.
In conclusion, well logs contain information that can
interest geologists and consequently can be used. Well
serralog Q 2003
1 Chapter 1 I
7
logs "photograph" the drilled formations and record real
facts that cannot be modified (Fig. 1-10). "They provide a
spectral picture, albeit particular and incomplete, which is
practically continuous and always permanent, objective
and quantified. It is easily understood that the "picture" will
be clearer when the number and the diversity of the log
measurements are greatest. One can say that logging
tools are to the subsurface rock description what the
eyes, geological instruments and products (hammer,
magnifying glass, compass, chloridric acid, alizarin...) are
to the surface outcrop. Thus, the logs can be considered
as the "signature" of the rocks since they depend on their
physical properties" (Serra, 1986). The physical characteristics of drilled formations that logging tools measure
result in fact on the one hand from the physical, chemical
and biological conditions - hence also geographical and
climatic - that existed in the environment of deposition of
the sediment and which determine the original facies and
characterize the environment, and, on the other hand,
from the evolution that these sediments have undergone
during their geological history to become sedimentary
rocks.
I
Figure 1-9 - Logging represents in average 10% of the well cost but if
provides approximately 90% of the information linked to the wellbore.
This is the reason why log data are so important. It is
no more possible to conceive any geological synthesis
and reservoir evaluation without the exploitation of well
log data. Doing so would correspond practically to base
our knowledge of the basins and reservoirs on the one
hand on outcrops, consequently on a small surface and
volume fraction of the formations composing them, on
surface geophysics but without calibration and scaling on
well logs, and, on the other hand, on cuttings, exceptionally on cores.
In addition, log interpretation should not be limited to
reservoir evaluation as too often practiced. The reservoir
must be put back in its geological setting (sedimentological, diagenetic and tectonic) to better evaluate its extent
and its volume. This is also the reason why geologists
must achieve well log analysis. You wouldn't think of a
radiologist analyzing medical images without knowing the
anatomy! Why would you entrust your log-interpretationto
a log-analyst who wouldn't know about geology? Only
geologists have sufficient geological knowledge to be able
to extract the required geological information. But, of cour7
se, they must also have a good understanding of the PhYsical parameters recorded by logging tools and the link
existing between the logging parameters and the geological attributes (Table 1-2).
(WL). In addition, logging techniques allow the sampling
of cores and fluids.
Table 1-2
Relative influence of geological attributes on well logging measurements
(adapted from Serra et Abbott, 1980).
Type of measurement
“Log interpretation consists, thus, of a data “translation” from log parameters to geological data. To do this we
need a good “dictionary” or an “interpreter” who knows
the two “languages“ welr (Serra, 1986).
Well logging data can be acquired either during the
drilling itself or after a period of drilling by tools gone down
at the end of a cable. The first one is called LWD (logging
while drilling), the second is known as Wireline Logging
8
Well seismic information
The Vertical Seismic Profile (VSP) obtained in a well
(Fig. 1-11) provides the best vision and exact depth of
each reflector, its signature and the effect over the underlying reflectors. It enables the transformation to be made
from reflectors to beds, providing a precise measurement
to go from depth to time and vice versa.
Senalog 0 2003
Generalities
I
Chaoter 1
I
9
Foreset laminations
Transport current
direction : N 200'
-
Drilling induced fracture
Foreset laminations
Transport current
direction : N 40'
-
Drilling induced
fractures
.
Foreset laminations
Transportcurrent
direction : N 120
-
Foreset laminations
Transport current
direction: N MI'
Ebsbnal contest
-
Figure 1-1Oa - Subdivision of a 2.2 meters bed in its depositional units from the borehole image analysis.
GR (API)
CAL (in)
SP (mV)
0
4
-80
150
14
20
2
10 ohm-m
MSFL
LLD
LLS
100
0
Pe (b/e) 10
45
NPHl (%)
1,95
Pb (g/cm3)
140
At (ps/ft)
-15
2,95
40
~~
Figure 1-lob - The 2.2 meters, corresponding to the image of Fig. 1-1 Oa, are indicated by a red strip on the standard logs.
Semlog 0 2003
9
Next Page
10
Well logging and Geology
seismic section and seismogram (Fig. 1-12).
0.9
I
.+
Figure 1-12 - Display of seismic section, seismogram and logs with
reflection coefficient and conversion of two way time in depth
(courtesy of Schlumberger).
Figure 1 - l l a - Example of VSP (courtesy of Schlumberger).
For example, it is possible to "see" the beds which
underlie formations with strong reflectance (anhydrite,
halite, compact dolomite or limestone), or with strong attenuation (undercompacted shales), or formations below
the bottom well depth. It also makes it possible to remove
the multiples that often complicate the interpretation of
seismic sections. VSP can also be used to analyze the
rock properties through the study of seismic waves (Direct
Signal Analysis method).
Wd1
Figure 1 - l l b - The VSP recorded in the well is inserted into the seismic
section allowing a better interpretation and the precise location of the
faults (courtesy of Schlumberger).
A graphic display of logs versus time instead of depth
can also be easily obtained and reproduced alongside
10
In fact, measurements of the density and the traveltime of acoustic sound derived from boreholes with well
logging tools, make it possible, after corrections, to determine the accurate acoustic impedance and reflection
coefficient for each boundary in formations. A reflectivity
log made from acoustic and density data provides a basic
document for the establishment of a theoretical seismic
section through a program, which, when correlated with
real seismic sections, permits its conversion into depth
and makes a composite log possible (Fig. 1-12).
Even if well logs constitute a fundamental source of
geological and petrophysical data, it is however important
to never under-estimate the problems that can occur
during their interpretation. These problems are linked to
the following conditions.
- Logging tools have different vertical resolution and
depth of investigation. Consequently, combination of their
data is not always valid as the different log responses do
not correspond necessarily to identical volume of rock.
- A perfect depth-matching of log data is sometimes
difficult to achieve if logging speed variations are not
recorded by an accelerometer tool added to the logging
string.
- Due to their lack of resolution, some of the recorded
data, especially those corresponding to the transition from
one bed to another, have no geological meaning. This is
especially crucial in thin beds. To avoid any misinterpretation the data corresponding to bed transition should be eliminated.
- The geological "bed" notion must be adapted to the
type of logging data and replace by the "electrobed"
concept. An electrobed is "the smallest unit that can be
recognized on logs from the surrounding units due to
significant changes in logging parameters". The minimal
thickness of an electrobed varies as a function of the logging type of measurement (Fig. 1-13). It depends on the
Serralog 0 2003
Previous Page
Chapter 1
intrinsic and qualitative vertical resolution. In order to
detect the thinnest units of deposition, imaging tools are
fundamental as they have the highest vertical resolution:
approximately 1 cm.
Sonic BHC
(ilyni
Figure 1-13 - Example of segmentation of log responses in electrobeds.
- What is generally called a bed may often be subdivided into several depositional units as illustrated by Figure
1-14.
Lithology
DePositional
units
Geological objects
11
- Most of the equations used for the quantitative evaluation of mineralogical and petrophysical properties are
em piricaI.
In fact, the properties of a bed essentially depend on
the own properties of each unit that composes it. Each
unit shows variable thickness and extent as a function of
its depositional environment. In other terms, a formation is
composed of a vertical and lateral succession of two
types of geological objects: volumes (beds, strata, layers)
delimited by surfaces. Too often, standard logs, due to
their lack of resolution, reflect imperfectly the average
petrophysical properties of a bed, not the real ones of
each depositional unit that composes it. Due to their very
high vertical resolution, images allow the detection of
each unit. It is also fundamental to quantify the information they provide (see further). An electrobed recognized
on a log with lower vertical resolution may in fact be composed of several smaller units, each of them having their
own petrophysical properties. The interpretation should
be based on this new concept.
As a conclusion we must keep in mind these problems
and proceed first to a log quality control. We must detect
the deficiencies when they occur, be able to make the
appropriate corrections, when they are possible, eventually reject the unreliable data, in order to correctly interpret the well logging information. But, in any case, extract
all the information included in logging data. Squeeze the
lemon to its ultimate drop! Accumulate real facts of observation in order to improve the accuracy and reliability of
your interpretation. You would not thrust medecine doctors who will base their diagnosis only on measurement of
your fever, your blood-pressure and examination of your
throat and ears. Why would you believe in an interpretation based on a reduced set of data?
Goals of well log interpretation
Figure 1-14 - The geological reality : generally several units compose
a bed and two types of geological objects : volume and surface must
be taken into account and their properties evaluated
(adapted from Blatt et al., 1980).
- Certain logging measurements are more sensitive to
borehole environmental effects (mud type, caving, temperature...) than others.
- A deep invasion by the mud filtrate may have a strong
effect on certain measurements and consequently obscure the original fluid in place.
- Very often the measurements are not corrected for
the influence linked to the apparent angle between the
measuring tool and the formation. This is very important in
deviated or horizontal wells.
Serralog 0 2003
Extracting geological information, and petrophysical as
well, related to bed thickness, lithology, composition, texture, structure (internal and external), diagenesis, stratigraphy and reservoir characteristics are the goals of subsurface data analysis. Characterizing these formation properties will assist geologists in determining the facies, the
genetic sequence, the depositional environment and,
consequently, the volume of the reservoir beds.
Geologists can also detect diagenetic effects and provide
a very detailed and accurate sequence stratigraphy analysis from those data.
To achieve these goals the analyst must use all the
available information provided by surface and subsurface
geological and geophysical data.
Interpretation methodology
An interpretation methodology is proposed in Table 13 (see next page). It integrates all available data in two
11
Well logging and Geology
Table 1-3
Interpretation methodology
(modified from Serra et a/., 1993).
Third step
steps for a single well. The third step corresponds to the
study of several wells of a field or a basin, each well
having been previously interpreted as indicated in the
two-step procedure.
Fundamentally, in a first step the subsurface geologist
extracts from well logs recorded in each well of a basin or
field the information he generally obtains, on surface, from
outcrop observation and rock analysis. For that purpose
he must apply to the well log analysis the same approach
12
he uses to observe and describe a core or an outcrop. He
must consider any well log object as a geological object.
His approach, similar to the one of the surface geologist, summarized in Table 1-1, consists of a first analytical
phase during which he accumulates the maximum data
and information coming from an objective, detailed, meticulous and complete observation and description of all
available well log data relative to each electrobed.
Serralog 0 2003
Generalities
He interprets this set of data in order to extract information relative to:
- lithology: elemental and mineralogical composition;
- texture: grain size, sorting, packing, cement type and
percentage, porosity, pore size, pore type and pore distribution;
- internal structure: sedimentary features, direction of
transport current and progradation, nature of bed boundaries, biological activity, energy in the depositional environment, figures of slump, slide, or avalanche;
- external structure or geometry: volume of each unit of
deposition: thickness, length, width, shape of limiting surfaces, dip and azimuth; determination of deformations
undergone since the sediment deposition: structural dip,
folds, faults, fractures; direction of stresses;
- facies and sequence: spatio-temporal relationship
between volumes (electrofacies) defining sequences of
deposition;
- depositional environment: the succession of facies
and their geometry allows the determination of the depositional environment and through that of the extent of
each unit of deposition;
- stratigraphy: relative datation, strata succession;
detection of reverse or recumbent series, erosion, emersion, transgression, unconformities;
- diagenesis: transformations undergone by sediments, dolomitization, compaction, cementation, dissolution; organic matter maturation;
- petrophysics: porosity (percentage, type, pore size,
pore distribution and geometry); permeability; anisotropy;
cation exchange capacity; irreducible water saturation,
wettability;
- economics: nature of mineral resources; extension;
volume; depth; proximity of consumption zones.
In a second phase, the geologist collects all the data of
each well he analyzed and with the help of the other members of the team he synthesizes the extracted observa-
Fundamenta I pre1iminary remark
Being conscious or not - it should be better to be
conscious! - logging tool measurements depend strongly
on the geological parameters which have controlled the
rock properties since its deposition.
Consequenfly:
- any log interpretation is by nature a geological interpretation;
- logs constitute a fundamental source of geological
and petrophysical information in subsurface.
The goals of this book are:
- to convince the reader that many types of geological
information can be extracted from logging data;
- to show how this information can be obtained from
the logging measurements.
In order to extract this information from logs one must
apply to well logs the same approach which a geologist
applies to outcrops and core examination.
Serralog 8 2003
I Chapter 1
13
tions and descriptions. Applying a continuous change of
scale of observation by taking into account the seismics,
logs, core and test data (Fig. 1-15), together they interpret them by reference to well known modern or ancient
models (experience) and to laws that control deposition,
transformation and deformation of sediments. In sedimentary rocks, this consists in the reconstruction of the
paleogeography, the diagenetic history and the tectonic
and stratigraphic setting of the formation and reservoir.
It is obvious that the accuracy of interpretation will be
improved by integrating the information provided by all
available data. Indeed, accumulation of observations and
real facts of different natures allows the interpreter to
avoid erroneous conclusions. Any interpretation that
would be based on a reduced set of information would
suffer of a lack of reliability. In the early stage of a field or
basin study, integration of log and image data calibrated
on core analysis, results of processing of well and surface seismic surveys, guided by information provided by
logs and dips, and test measurements are a must to ensure an accurate and reliable evaluation of the economic
potential of formations.
The accuracy will be, as well, improved by eliminating
all the artifacts generated by the lack of resolution of standard tools. This is the reason why the electrobed concept
is important and why the use of images is fundamental.
Workstations simplify interpretation as they allow the
user to easily process raw data, manipulate a variety of
data types, and display the results of processing and
interpretation at any scale (see further Fig. 1-16).
Interpretation of images is a multi-step process of observation, description, interpretation, and implication or
exploitation (Serra, 1989; Bourke et a/., 1989). The following set of guidelines for image interpretation is recommended:
For observation and description of sedimentary features, use images that have the same horizontal and vertical scales.
Interpretation of enhanced images should be done in
conjunction with static normalized images to check the
resistivity range of the observation. A calibrated microresistivity curve, displayed on the side, may replace the static images.
All images should be interpreted in context of other
open-hole logs, preferably recorded at a high sampling
rate, to relate events seen on the images to petrophysical
properties and vice versa.
The geological interpretation must use all the available data and especially, as much as possible, images or
dipmeter data. Due to their very high vertical resolution
(Fig. 1-17 and Table 1-4) images are fundamental to
determine the internal organization of any depositional
unit, especially reservoir, to better understand its complexity and/or its dynamic behavior. It is obvious that a
reduce set of logging data will not allow a reliable description of the formations.
13
14
Well logging and Geology
Figure 1-15 - From seismic data to thin section analysis, through standard logs and images. This change of scale allows a much better link between the different sources of information and, consequently,a better understanding of the attributes and properties of each depositional unit recognized with techniques of higher resolution
(courtesy of Baker Hughes).
Figurel-16 - Example of data reproduced on the screen of a workstation. One can see quantitative interpretation results, borehole wall images
indicating a highly laminated reservoic dip data and cross-plot of density vs neutron measurements. This plot indicates as well the laminated sandshale sequence of this reservoir. The shale percentage computed from the standard logs is more important than the one shown by the images.
Gas effect can also be observed (courtesy of Schlumberger).
14
Serralog 0 2003
Generalities
D
m
c
Y
0
E
I-
Figure 1-1 7 - Vertical resolution of logging tools.
Only bed boundaries, laminations and fractures that
are not oblique to the borehole axis will be common to
both images and core photographs (Fig. 1-18). In heterogeneous formations, features present in the core will not
occur on the borehole wall at the same depth and perfect
correspondence between core and image is unlikely.
Any geological interpretation of well logs must extract
the fundamental geological attributes of each depositional
unit as it is done from outcrops or cores (see next page,
Fig. 1-19).
These geological attributes are:
- Thickness.
- Elemental and mineralogical composition.
- Texture.
- Internal structure.
- Facies.
- Sequence or facies succession.
- Depositional environment determination.
- Diagenetic effects undergone by the sediment since
its deposition.
- Deformations undergone since the sediment deposition.
- Stratigraphic history.
Although the details of images are crucial for formation
description, analysis of the data can be very time consuming. A multi-step, systematic approach is recommended
to achieve optimum results in the least amount of time
(Harker eta/., 1988).
1. Display the images at a compressed scale (e.g.,
1:lo0 or 150) together with relevant companion logs (Fig.
1-16). This display is used first to perform depth matching
and correlate images with open-hole logs and arrow plots.
2. Zone the major depositional units using this composite log. A lithologic column obtained by processing openhole data may be preferred to raw data, because it directSerralog 8 2003
I Chapter 1 I
15
ly indicates the main lithofacies.
3. Identify major sedimentary discontinuities and tectonic features using an arrow plot. Use seismic information
and local knowledge to narrow the possibilities.
4. Define the most probable depositional environments
from all available information - lithofacies, thickness,
genetic sequences, nature of bed boundaries, etc.
5. Display the images at 1 5 or 1:lO both horizontal
and vertical scales. At these scales, image details can be
interpreted in terms of texture, sedimentary features, etc.,
within the interpreted environmental context. Local
enhancement may be necessary to highlight faint features.
6. If cores are available, describe them in detail.
Comparing sedimentary features between cores and images permits assessment of lateral variation of thin beds
and other features at the scale of the borehole. Images
also provide better visualization of minor faults and
slumps that are ambiguous or absent on dipmeter logs or
may be too large to be interpreted from core data alone.
7. Summarize all the information provided by the images and focus on the nature of bed boundaries, bedding
types, dips, fractures, stylolites and pore types and distributions.
U
Figure 1-18 - Depth-match of core and FMS images with the
DlAMAGE program. Observe the traces of fracture and
laminations. Their vertical extent is not the same as soon as their
apparent angle is not nil. This is due to the higher diameter of the
borehole than the core diameter
(courtesy of Schlumberger and ELF).
15
16
Well logging and Geology
GRAIN
SIZE
Figure 1-19 - Typical geological analysis that must be achieved from well logging data as the latter provide practically all the information necessaty
to make this geological description as done by Van Wagoner et al., 1990.
16
Serralog 0 2003
Generalities
I Chapter 1 I
17
-
Table 1-4
Vertical resolution of the principal logging tools compared to the geological objects.
,E!ec!W!.im?P. * . -. I
.Dipmeters
. .. . .. . .
Ultrasonic images
e
icrolog
Electromagnetic (GHz) W
Micro SFL
Litho-density - - - t---+,
Radioactivity- - - - Acoustic *-\
Azimuthal resistivitye-1
Laterolog I-*
Spherical log +e
Neutron - - - +e
Spectrometry
Array induction
Dual induction- - sp
t
I
1
1
1
1
I
0.05
0.0 1
o~ooo6.001
I
I
1
I
I
I
-
I
0.05
I
0.5
0. I
VSP I*
I
I
1
I
1
5
I
.+,
I
10
Scale (m)
c'
Y Crystals
8
3
cn 0
- - -Rock fragments - - - - - - - - - -
rd)
8
's"
Silts
w
I Sands I 1
1
Pebbles Cobbles1 Boulders
$
m
Shell debis _-------_-Nodules - - - - - - - - - - Sand lenses - - - - - - - - Crystals - - - - .r(
CI
09
....1
2
m
8
.rl
w
0
m
Fractures
od)*
g
d)
- -
-Small sedimentary features - - - - - - -Large sedimentary features
sedimentary bodies
Faults
. . . . Higher sensitivity than resolution
the purpose of interpretation.
Grade 1 - A unique interpretation of the image is
possible independent of other data. Trained analysts can
interpret the images from their prior experience with confiSerralog (i3 2003
--
- - - -Higher resolution techniques
contoured features with well-defined shapes. Other logs
may offer information to aid the interpretation.
Grade 2 - Ambiguous. A conceptual model or other
log data are necessary to interpret the images correctly
and completely.
17
18
Well logging and Geology
Grade 3 - Core calibration needed. Images are used
to complement core interpretation. After calibration to core
data, the images may be used to evaluate zones where
cores have not been recovered.
the analysis of borehole images and nuclear magnetic
resonance data for texture. As a reliable knowledge of the
lithology is fundamental, a particular care must be applied
to log interpretation to obtain this information.
To be used in any interpretation the information provided by image data needs sometimes to be quantified and
classified based on their origin (see Fig. 1-20 next page).
This is achieved, for instance, through the TexScan or
BorScan programs of Schlumberger. Its logic will be
explained later (Chapter 3).
The log study continues by determination of the sedimentary features, from borehole images or dipmeter data,
which will inform us about the facies and the depositional
environment through the analysis of the facies succession
or sequence.
Geological interpretation of well logs requires the
same type of work than generally done when studying
outcrops or cores (cf. Table 1-1). As logs strongly depend
on geological parameters, detailed and precise observation and description of events seen on logs and images,
comparison with previous studies, interpretation based on
experience, and prediction must be also the successive
steps of geologists working on subsurface data.
During their studies, workstations will help considerably the geologists in 'their geological interpretation as all
the data can be shown and studied at the same time as
they can be reproduced on the screen at the required
scale (Fig. 1-16).
As previously mentioned, and this must be well
understood, any log interpretation is by nature a geological interpretation as log responses depend strongly on
geological parameters. We insist again, it is important to
understand this fact. The goal of the geological interpretation of well logs is to convert the logging data into geological attributes. This requires a good knowledge of the
link between the log measurements and the geological
parameters. This link is shown in Table 1-2 (see above)
and will be developed and illustrated in the next chapters.
Thus, once more, extracting information related to
lithology, composition, texture, structure (sedimentary and
tectonic), thickness, geometry, diagenetic effects and
reservoir characteristics of each depositional unit crossed
by a well is the main goal of geological interpretation of
well logs, images and dip data.
Characterizing these formation properties assists geologists in the description of the facies, genetic sequence,
depositional environment, types of traps and stratigraphy.
In addition, reservoir engineers can determine the presence of permeability pathways or barriers and reservoir
flow properties, allowing them to better determine the
scheme of the field development and the calculation of
the net pay.
As we can understand, the study of any formation
starts with the determination of the lithology of each bed
which composes it, it means its mineral composition and
its texture. This first step requires the interpretation of the
log data with the help of cross-plots for composition, and
18
This study will be completed by the determination of
the transformations (diagenesis) and deformations (tectonics) undergone by the formations since their deposition.
Finally, formation evaluation and reservoir interpretation will take into account the geological data provided by
the logs to more accurately determine the petrophysical
properties of the reservoirs and their extent. In other
terms, the understanding of the reservoir properties and
behavior will be improved if the reservoir can be put back
in its geological setting (environmental, diagenetic and
tectonic).
Before covering these different steps, preliminary
works must be achieved.
Log Quality Control
Previously to any log interpretation it is very important
to check the quality of the logging data.
This starts by the control of the calibrations.
Tool calibrations
Each tool when it goes out from the factory must.be
calibrated. This is achieved by laboratory measurements
either in special holes like the pits in Houston, or in calibrating blocks existing in each service company. But, as
the tool-component properties can vary with time (age of
the sources or of the detectors, electronics influenced by
temperature...), it is of the utmost importance to check at
the shop and on the well itself the good functioning of
each tool and to recalibrate it.
Repeat section
After each run, a part of the well (generally the most
interesting section) is again recorded. This repeated
interval is called the repeat section. Comparison of the
two runs allows the detection of possible problems (Fig. 121 next pages). The repeat cannot be perfect, especially
for nuclear data. This is linked to the statistics but also to
the fact that the tool (litho-density for instance) does not
follow the same trajectory. In addition, it can occur that the
gamma ray of the repeat run shows higher values. This is
generally due to delayed gamma rays induced by neutron
Serralog 0 2003
Generalities
I Chapter 1 1
19
QUANTITATIVE BOREHOLE IMAGE ANALYSIS
OEOLOG~ICAL
OBJECTS DOWN TO A FEW MILLIMETERS SIZE
MATHEMAT~CAL
TOOLS ALLOW TO AUTOMATICALLY
A FAST ZONATION CAN BE EXTRACTEDFROM
BINARIZED IMAGES DERIVED WITH AN
ARE IDENTIFIED WITH BOREHOLE IMAGES
EXTRACT AND Q U A ~ F Y
THEM
Landscape view of the
conductivity image.
Landscape view of the
gradient image.
ADAPTAT WE BACKGROUND
FOREACH ZONE, GEOLOGICALOBJECTS ARE AUTOMATICALLY
RESISTIVITY FILTER
EXTRACTED AND THEIR ATTRIBUTES ARE COMPUTED
THIS METHOD INVOLVES A PRIMARY STEP OF INTERPRETATION AND THE IMAQE ANALYSIS
IS GUIDED B Y THE QEOLOQICAL KNOWLEDGE OF THE FORMAWONS
DEPOSITIONAL
SURFACE ATTRIBUTES:
- Resistivity contrast
- Dip end azimuth
- Planarity
- Proximity
LAYER
ATTRIBUTES:
- Medianconductivity
- Thickness
DIAOENETIC
HETEROGENEITY
TYPE AND DISTRIBUTION:
- Size, shape, area
- Frequency
- Connectedness
- Median conductivity
STRESS INDUCED
FRACTURE
ATTRIBUTES:
- Polarity (resistive,conductive) - Dip, azimuth
- Length
- Electrical aperture
- Density
- Planarity
THEINTEGRATIONW m i STANDARD LOGGING MEASUREMENTS, CORE AND FORMATIONTESTS PROVIDE
VOLUMES AND
A QUANTITATWE ClIARAClERUATlON OF THE INT€RNAL ORQANlzATloNOF RESEWTHEIR BOUNM#Q SURFACES TO BE?TER DEFENE THE FLOW PROPERTIESAND RESERVOIR MOMLS
Figure 1-20 - Image analysis and quantification
(from Delhomme & Motet, 1993).
Serrabg 8 2003
19
20
Well logging and Geology
E
P
Figure 1-21 - Comparison of main log (left) and repeat section (right) allowing the control of the log quality. Obsetve the small depth shift on
resistivity and neutron-density between the main log and its repeat section.
interactions with elements.
Another test of the data quality is to cross-plot the
repeat-section data versus the main-run data on the same
depth interval (Fig. 1-22).
Comparison of histograms of log data recorded during
the main run and the repeat run is also another way to
check the logging qualities (Fig. 1-23).
Depth match
As the tools are either run one after the other or combined (triple-combo or Platform Express) but superposed,
depth match of the log data are needed to put back at the
same referenced depth the different measurements corresponding to the same bed. As already mentionned,
tools can stick and suddently jump due to the cable elasticity. In the absence of accelerometer data, this step of
the preliminary works is sometimes difficult to achieve
correctly. Analysis of the tension curve, the caliper and the
Dr curve, allows sometimes the determination of the
depths where tool-sticks can occur. With accelerometer
data, this depth match is generally automatic as the variations of logging speed are recorded and the length between measurement points are well known (Fig. 1-24).
Figure 1-22 - Cross-plots of repeat section data versus main log data. GR of repeat run (GR-R) is very often higher than GR main run. Density
repeat (RHOB-R) is closer to the density of the main run but still with variations.
20
Serralog 0 2003
Generalities
---
I
Chapter 1
I
21
Figure 1-23 - Comparison of histograms of log data, on the same depth interval, corresponding to the main log (on left) and to the repeat section
(on right). Some differences can be observed even if the general aspect is preserved.
Mr--
Composite-Log
As soon as the depth match of logs is correctly achieved, the composite-log must be realized (Fig. 1-25 which
provides an example of composite-log including image
data). This document combines all the information recorded in a well and will help the analyser in his work as he
will have all the available data under his glance.
Segmentation in electrobeds
Figure 1-24 - Comparison of logs before (second track) and after (third
track on the right) corrections for running speed variations recorded by
the accelerometer included in the tool string
(courtesy of Schlurnberger).
Serralog 0 2003
The next step of the preliminary work consists to segment the interval to analyse in electrobeds. As already
mentionned, due to their lack of resolution, some data are
not representative of geological realities. The segmentation of logs in electrobeds allows the elimination of these
non-representative data and the correction of the log data
for error measurements and statistics. The logic of segmentation is explained (Fig. 1-26, next pages). Example
of log segmentation is shown in Figure 1-27 next pages.
Several log analysts think that clustering technique or
neural network approach eliminate automatically the nonrepresentative data. In certain conditions - thick beds and
high resolution tools -those data are partly included in the
cluster. In thin bed conditions, as the data corresponding
to the transition zone from one bed to the following can be
numerous and generate clusters which have no meaning.
In any case, the non-representative data affect the log
responses corresponding to each bed modifying the
representative values of each bed.
21
Next Page
22
Well logging and Geology
Gamma Ray
0
( W
LLd
LLS
MSFL
200
0 Pe(b/e) 10 -----20 Th (ppm) -20
-10 K(%)
10
20 U (ppm) 0 -----1.95 Density (g/cm3) - 2.95
45
Neutron (P.u.)
-15
200 Transit time (ps/ft)
0
---
--------
Hole
drift
0" 10"
Caliper
1.25 C 1-3
11.25 C2-4
(in.)
11.25
1.25
0"
- 0.5
.......&
.... 0.5.
1
FMS Images
ohm-m 1000
360"
II
-
Figure 1-25 Example of composite log combining standard logs with image data repmduced at the same verfical scale.
22
Semlog Q 2003
Previous Page
Generalities
I Chapter 1 I 23
which is illustrated by Figure 1-31.
-
Figure 1-28 Zone investigated by a
tool in a vertical well. In such a case,
the vdume investigated may be parallel to the bedding and may cornspond
to a unique bed if this bed is sufficiently thick compared to the vertical
resoltion of the tools
sy of Schlumbetyer).
L
-
Figure 1-29 Case of beds not perpendicular to the borehole axis. In that case, if
the beds are relatively thin compared to
the tool vertical resolution, the measurement is not representative of any bed
physical ptvpe@.
+
-
Figure 1-26 Explanation of the log-segmentationlogic taking into
account vertical resolution, statistics and measurement emrs.
Well and Logging Types
In petroleum industry several well types exist. The logging data must be associated to the well type in which
they have been recorded before any interpretation.
Open holes
In those types of wells, generally any type of logging
data can be recorded but one must know several parameters related to the well: the borehole size, the orientation and the reached depth. In addition, the apparent dip
between the borehole axis and the beds must be known
as well.
F
N
s
Fardetector
Neardetector
Swm
-
Figure 1-30 Case of a deviated well in the direction of the dipping. As
it can be imagined the volume analyzed by the near and the far detector of this litho-density tool does not cornspond to the same volume
of rock at the bed boundary.
Vertical wells
The spatial geometries between tool and beds are
illustrated by Figures 1-28 and 1-29. They correspond to
many old situations as, initially, the wells were fundamentally vertical and the logging acquisitionwas essentially by
wireline techniques.
Deviated wells
In that case the geometry is most of the time the one
illustrated by Figure 1-30.
Horizontal to subhorizontal wells
The recorded data are influenced by a different volume
serralog 82003
c.--
-
Figure 1-31 Case of a horizontal
well. As it can be observed, the zone
investigated has a symmetry perpendicular compared to the bedding
(courtesy of Schlumbetger).
J
Interpretationof the logging data, and more especially
of the resistivity data and the images of the borehole wall,
must take into account the relative angle between the
bedding and the tool axis (Fig. 1-32).
23
24
Well logging and Geology
Figure 1-27 - Comparison of standad log presentation (top) with segmented log presentation (bottom). This segmentation eliminates the non
representative data and generates a geological display of the interval in electrobeds closer to the bed notion.
24
Serralog 0 2003
Generalities
I Chapter 1 I
25
Table 1-5
Storage of the standard logging data corresponding to each electrobed.
4
n
When recorded in a deviated or horizontal well the
investigation of the borehole by any tool is different from
the one realized in a vertical well (Figs. 1-28 to 32). This
must be taken into account when analyzing the data.
Figure 1-32 - The volume investigated by a tool is not affected by the
same type of formation in a vertical, deviated and horizontal well. The
tools are analyzing in adifferent way the beds composing the formations. This generates an anisotropy and a different type of images.
This must be taken into account when analyzing logging data
(courtesy of Schlumberger).
Since more than 20 years, thanks to the technological progresses, the acquisition of logging data is now
possible during the drilling itself (Logging While Drilling or
LWD).
The development of deviated and horizontal wells has
also been possible taking into account the information
provided by Measurements While Drilling (MWD) and
sensors close to the rock bit (well trajectory, drilling effi
ciency and formation properties). Even when deviated or
horizontal, the wells start vertically, with a conventional
rotary bottomhole assembly.
Figure 1-33 - RAB image in a horizontal well
(courtesy of Schlumberger)
CDNtool PowapUku,
The drillstring and bit are rotated from the surface
either by a rotary table on the derrick floor or a motor in
the traveling block, called a topdrive or “locked” assembly.
The rotary assembly can be replaced with a steerable
motor, driven by mud flow. When mud is flowing, the
motor rotates the bit, but not the drillstring which slides
along. This type of drilling is called sliding mode. The
direction of the bit is measured and sent to surface by
MWD equipment. These measurements include azimuth
and inclination with respect to the vertical.
To the MWD measurements are added different LWD
measurements such as resistivity, density, neutron, photoelectric index Pe, sonic slowness, radioactivity and if
required Resistivity-At-the-Bit tool (Fig. 1-33) which can
provide images of the borehole. These measurements will
allow a better formation evaluation during the drilling itself
Serralog 0 2003
Figure 1-34 - On the top and the left, example of surface and
downhole components of LWD tools (IDEAL Integrated Drilling
Evaluation And Logging system of Schlumberger).
On the bottom right, sketch of the ADN tool of Schlumberger.
(courtesy of Schlumberger).
25
The recording of any wireline logging data in horizontal wells requires a special technique called tubing
conveyed system or Tough Logging Conditions (TLC* ) for
Schlumberger. This technique uses a drill pipe to push
logging tools and a side-entry sub for the entry of the
cable in the drill pipe. The side-entry sub precludes the
circulation of drilling mud to loosen or advance the drill
pipe.
The LWD measurements do not require any special
equipment other than the sensors inserted in the drill collars or the stabilizers. Another advantage is the acquisition of data during the drilling itself allowing a rapid reaction in order to maintain the drill bit inside the reservoir by
comparing top, right, bottom and left measurements of
density and photoelectric sensors (Fig. 1-35).
Based on this set of log data one can start an interpretation in order to completely describe these electrobeds in terms of composition, texture, structure, facies,
sequence, environment, diagenetic effects, tectonic deformations, petrophysical properties, etc.
This interpretation can be achieved “manually” or with
the help of specific softwares written for computers or
workstations (Table 1-6). In any case, the results of this
processing by any software must be controlled.
Table 1-6
Interpretation programs of Schlumberger.
4pplications
~~
Quality control
Lithology
I I
Wellsite
I
LOGOS
I
I
I
Reservoir
modeling
LQMS
FLIP*
Facies
FACI0LOG
DimQC
II
BorTex:
ScanTex
FasTex
SYNDIP
(SYNDIP)
(STRATIM
DUADIM)
FLIP*
I
GeoFrame*
Rockclass
RockCell
Litho QL
LITHO
LITHO
Platform FACIOLOG
Express ESTIME
Quicklool
Sedimentary
patterns
I
SediView
DipFan
analyzer
Stratigraphy
Figure 1-35 - On the left comparison of the sensors put into the drill
collar (bottom image) and on the stabilizer (top image). On the right:
ADN images acquired in a horizontal well (courtesy of Schlumberger).
FLIP*
DipTrend
FIL
DCA HDT
HDT AN
SHDT AN
FracView*
Cased holes
In such wells the presence of casing and cement between the tools and the formations reduce the possibility of
measurements. Nuclear measurements, such as natural
radioactivity, neutron interactions (thermal neutron decay
time, induced gamma ray spectrometry), and density can
be recorded. Resistivity measurements through casing is
also possible using a new tool, CHFR* (Cased Hole
Formation Resistivity), introduced by Schlumberger. In
any case, the interpretation of the recorded data is more
difficult as the nature and amount of cement is not precisely known. Images and dip data are not available.
StratLog II
StrucView
Cross-section
BorView
FasTex
Workstation programs
As workstations are now available in many offices the
next paragraph will describe the one developed by
Schlumberger.
The other service companies and the major petroleum
research companies have their own workstations based
fundamentally on the same principles.
Conclusion
Workstations
At the end of this preliminary work, it can be useful to
collect all the logging data corresponding to each electrobed into a table (Table 1-5 previous page). This part of the
work corresponds to the analytical step which, generally,
will not be argued.
The emergence of workstations in early 1980’s years
has considerably simplified the work of geoscientists. This
is due to several factors:
- import data from different origins in the same environment;
- set of softwares for processing and interpretation of
26
Serralog 0 2003
Generalities
raw and computed data, and display and storage of
results;
- interactivity;
- rapidity;
Workstations have relieved geoscientists of all the
repetitive, tedious, sometimes difficult, and time-consuming part of the analytical aspect of their work. They simplify interpretation as they allow users to easily manipulate any kind of data and display results of processing.
Integrating all well data (logs and images, cores, tests,
raw and computed data, well seismics) in the workstation
environment aids geologists, geophysicists, petrophysicists and reservoir engineers in their evaluation of formations. Geologists can improve the structural, sedimentological, diagenetic and stratigraphic analysis of logs calibrated on core data. Geophysicists can improve the processing and the interpretation of their seismic sections using
data provided by well logs. Petrophysicists can refine log
interpretation and minimize uncertainty taking into
account the geological information provided by the logs.
Reservoir engineers can better understand the dynamic
potential of a reservoir through recognition of petrophysical volume properties and surface transmissibility.
Chapter 1
I
27
could locally enhance part of an image to emphasize
details, make stereographic projections of selected dips
or have a three-dimensional view in any direction for better interpretation of the images. They could also select
thresholds for the images to make a vuggy porosity estimation or a sand bed count with cumulative bed thickness.
During the work session, users could make color
screen dumps of selected intervals. These dumps can be
updated and customized with TOUCH software. When the
interpretation is complete, a continuous color hard copy of
the images, plus the computed dips and the annotations,
can be made over the entire interval (Fig. 1-36).
A rapid description of the different processings available in the Schlumberger-GeoQuest workstations will allow
geoscientists to better know the type of information they
can extract from logs and images and what they can
achieve in terms of interpretation.
Formation Image Examiner
J
Figure 1-36 - Examples of functionalities existing in FLIP Oriented
images, curves, dips, three-dimensional views, azimuth frequency plots
and annotations can be generated and displayed on the screen or reproduced as a color screen dump
(courtesy of Schlumberger).
The Formation Log Imaging Product (FLIP) was the
first workstation introduced to interpret images. FLIP was
a fast, user-friendly, flexible tool used on a SUN workstation. This Formation Image Examiner displayed all the
information necessary for quality control and interpretation of Formation Microscanner*, FMI*, ARI* or acoustic
images. The display was easy to use and fast enough to
allow extensive browsing through the available information, allowing users to concentrate on the data rather than
on the program. Users could display the images at different horizontal and vertical scales, scroll up and down the
log, and view the borehole and its azimuth in three dimensions. They could use their expertise to detect, analyze
and interpret features on the images, and could isolate
individual features and analyze the interesting parts of the
images.
The interactive, user-friendly FracView program is
designed to derive quantitative information relating to
structural and stratigraphic information (Cheung & Heliot,
1990) as well as fracture dip, fracture orientation, fracture
density, fracture aperture and porosity (Luthi & Souhaite,
1990). Healed (Fig. 1-37) or open natural fractures (Fig.
1-38) can be recognized and separated from induced
fractures (Fig. 1-39) and bed boundaries (cf. Chapter 9).
Users had a set of built-in interpretation facilities at
their disposal. They could document their observations
directly by annotating the images or logs, but interactive
interpretationtools are available. Interactive dip computation allowed users to indicate directly on the image what
they recognized as bedding, laminations, erosional surfaces, stylolites, fractures or fault planes. In addition, they
Handpicking fractures may be time consuming and
tedious. This motivated the development of a batch-oriented product (SPOT) that automatically extracts conductive
fractures from a set of electrical images. FMI, Formation
Microscanner and ARI images can be processed.
Serralog 0 2003
FracView* fracture synergy log
Stereographic projections can be generated to improve the interpretation or to analyze the fracture network in
conjugated sets.
The image-derived quantitative information is displayed by a set of synthetic logs (Fig. 1-40).
27
28
Well logging and Geology
on a workstation. DIAMAGE is available on Sun workstations. It uses the X Windows SystemTM (MOTIF) interface.
Core photograph
Figure 1-37 - Conjugated cemented fractures in an eolian sandstone
(courtesy of Schlumberger).
0.50
Figure 1-40 - Example of quantitative logs generated by the Fracview
program. The density of fractures, fracture trace length, fracture aperture and fracture porosity are displayed as a function of depth. The
fracture dips are also reproduced and separated from the dips related
to bed boundaries (courtesy of Schlumberger).
Core photographs are taken by the AUTOCARTM
photography system developed by the Societe
Europeenne de Systemes Optiques (seso), which captures the image in colors of a cylindrical core's outer surface - without distortion - before any destructive intervention
occurs. The core is sprayed with water for a better color
contrast, and then is rotated on its axis (Fig. 1-41). A device synchronized with the rotation moves a film perpendicular to the core axis. The system controls photographic
focus, sprinkler startup, rotation, synchronization,
sequence and shutter action.
Figure 1-39 - Drilling induced
fractures associated with natural open fractures. The imgular features of the drilling induced fractures, which are not
continuous, generally parallel
to the borehole axis and
concentrated in a direction perpendicular to the natural fracture dip planes. The open conjugated shear fractures have
Figure 7-38 - open natural conjuga- dips equal to N270" and N73"r
with a strike approximately N-S
ted fractures in a granite
(courtesy of Schlumberger).
(courtesy of Schlumberger).
DIAMAGETM
D~AMAGEis a user-friendly software, written by ELFAquitaine and commercialized by Schlumberger, that
allows import, display and interpretation of core photographs and borehole wall images with other borehole data
28
Figure 1-41 - Photograph and sketch explaining the AUTOCAR system
(courtesy of Elf-Aquitaine).
The core photograph is scanned, and the image is loaSerralog 0 2003
Generalities
ded and displayed on the workstation using DIAMAGE
software (Fig. 1-42). Integration with all other well data
allows the following:
- precise and accurate depth match of the core with
borehole data
- automatic core orientation (depth and azimuth) by
selecting representative dips
- merging, splitting, flipping, rotating and shifting core
images for comparison with oriented borehole wall images
I Chapter 1 I
29
- statistical studies of fractures and borehole ovalization
- identification of drilling-induced fractures
- interactive dip computation and classification by fitting a sine wave or selecting points
- modifying image-color restitution using a variable
palette for enhancement of details
- interactive drafting of lithologic column
- detection on borehole images of fine details not visible on core photographs
- adding textual and graphics comments referenced in
depth
- simultaneous display of openhole logs and test
results (RFT* Repeat Formation Tester, MDT* Modular
Formation Dynamics Tester and DST [drill stem test])
GeoFrame* Geology
The GeoQuest GeoFrame integrated reservoir characterization system delivers the softwares users need to
develop fast and reliable, risk-reducing solutions that will
help geoscientists and engineers with prospect interpretation, well planning, reservoir behavior understanding and
field development management. GeoFrame comprises
applications specifically designed for project data management, seismics, geology, petrophysics, borehole geology, visualization and interpretation, and reservoir analysis,
launched from the GeoFrame Application Manager.
GeoFrame software applications have standardized features for software and data loading, as well as transfer for
moving back and forth between projects. This tight integration workflow process and ensures that changes to
one display are reflected in another. In addition to the processing softwares for dipmeters and imaging tools, described previously (cf. Volume 1 - Chapter 24), GeoFrame
offers a set of softwares for preparation of the standard
log data and for their interpretation that are described
now.
R
Figure 1-42 - Example of DIAMAGE display. On the left the FMS
image, in the center the oriented, depth matched core photograph.
On the right the dips computed from the images
(courtesy of Elf-Aquitaine).
- mirror imaging of core photographs to obtain a view
from inside for comparison with borehole wall images
- typology and identification of features and events
seen on borehole images
- precise location on images of plugs collected from
core if core photograph is taken after collecting the plugs
- detailed description of formations
- calibration of images for quantitative interpretation
- analysis of texture, sedimentary structure, facies and
sequence for sedimentology and stratigraphy applications
- three-dimensional analysis of sedimentary features
- classification of the hierarchy of surfaces (stratifications, bed boundaries, truncation, erosion, fractures, stylolites, etc.) and validation of permeability barriers
Serralog 0 2003
Data PreDaration
Data Preparation products - WellEdir and PrePlus* bring a powerful set of user-friendly tools for data editing
and environmental correction.
The WellEdit module comprises:
- general log curve editing,
- log depth match by stretch and squeeze depth shifting,
- core data and core image editing.
The PrePlus module applies appropriate environmental corrections to log data.
Geoloay
GeoFrame geological applications softwares include
StratLog /I*, Wellpix*, WellSkefch*, BorView* , BorTex
and DipFan.
29
30
Well logging and Geology
Stratlog I/ delivers all the tools needed to interpret
complex geology. It offers numerous easy-to-use features:
- single-well interpretation and display that incorporates any type of exploration and production data - geological, geophysical, petrophysical, engineering and production data into a single-well display to support your own
interpretation process;
- creation of structural and stratigraphic cross-sections;
- creation and manipulation of multi-horizon contour
maps.
-
The Wellpix module allows the user to rapidly pick
geologic markers and faults. Any marker picked in WellPix
is available in any GeoFrame application. To interpret
complex areas, WellPix offers a variety of features including fault gapping, flattening, independent scrolling and
log drag, stretch and squeeze.
Wellsketch brings powerful spreadsheet capabilities
and graphic editors to foster quick access, effortless editing and robust functionality.
The BorView program is the equivalent of the FLIP and
FracView programs written under LOGOS previously described. The BorView program offers a group of interactive
modules for interpreting Formation Microscanner and
FMI images. BorView incorporatesthe following modules:
- The lmageView interactive module supports the
display of high-resolution borehole images. Functions
include:
- interactive dip picking by selecting points or fitting a sine wave template directly on the images. The user
can immediately classify the computed dip as a bed
boundary, lamination, cross bedding, erosional surface,
stylolite, fracture or fault plane.
- histogram interaction and normalization
- fracture aperture analysis
- The Attributeview module computes a certain number of object attributes, such as dip, azimuth, planarity,
fracture density, fracture length, fracture aperture and
fracture porosity.
- Stereonetviewallows the user to make stereographic
projection (Wulff and Schmidt plots), histograms and
rosette diagrams, and all the manipulations on stereonet.
- StrucView allows the user to construct cross section
either from a dip file or through the automatic picking of
dips on the images. This module uses DipTrend logic.
The purpose of the BorTex program is twofold:
- extract and summarize the information about the
internal organization of the formation encountered by a
well, based on the analysis of dipmeter fast channel or
image microresistivity curves
- provide a zonation of the well into significantly different "morphofacies" using hierarchical clustering techniques, based on the information previously extracted and
summarized.
30
The information derived from the dipmeter curve or
image morphology is summarized by BorTex as logs, called BorTex "summary logs", so that it can be easily combined with traditional logs.
Intervals characterized by the same summary log
signatures define dipmeter curve- or image-derived
facies. Such facies are called "morphofacies" because
they are based on morphology, to distinguish them from,
e.g., "lithofacies" that are based on lithology.
BorTex thus combines in a single module new functionalities for electrical SCANner image TEXture analysis,
which is referred to as ScanTex,with the functionalities for
dipmeter FASt channel TEXture analysis offered by
FasTex of the DipFan program for dipmeter curve analysis (see below).
ScanTex inherits from some algorithms and concepts
originally developed for the SPOT software developed
under LOGOS. It is a geology-driven image analysis. Its
goal is a precise and complete extraction of the information provided by the images.
A common user interface gives access to both FasTex
and ScanTex algorithms. In wells where electrical imaging
tools such as FMI or Formation Microscanner tools have
been run, ScanTex algorithms is preferably used.
However, it is also possible to use the FasTex algorithms,
either to speed up the processing or for the sake of
consistency in a field study context. In such a case,
pseudo-dipmeter curves are first computed from
FMVFormation Microscanner data to replicate SHDT or
HDT fast channels.
For each pad or flap image, an image analysis technique reconstructsthe part of the image that extends from
right to left (i.e., the image-crossingcomponent). Then, a
reconstruction across the borehole, guided by the bedding dip information input to BorTex, is applied. This gives
the borehole-crossing component of the images.
Conductive and resistive heterogeneities are sorted into
spots and patches. Both spot and patches have boundaries that are located at maximum-contrast lines in the images but spots are local regions flagged by a local maximum and with normally a high contrast with respect to the
background. Patches are larger and more diffuse, showing only a medium contrast with respect to the background. In some instances, spots are small regions that
only show a medium contrast.
The results are displayed over the images, providing
boundaries of beds or layers, and contours of isolated
events. In addition, layer conductivity and high-resolution
textural curves are generated (Fig. 143).
A statistical study of the results is also achieved. This
allows a summary of the information and a data compression for comparison with other logs. Sliding statistics of
attributes (percentage, thickness, contrast, connectedness, size, aperture, etc.) are done. The results are
displayed as summary logs .
Based on these results, an automatic interval segmentation is achieved applying clustering techniques.
Serralog 8 2003
Generalities
814B.8
8741.2
8741.6
Figure 1-43 - Example of object attribute computation (a) original
image. (b) Vug area percentage and median size log (computed with a
0.5 ft [15.24 cm] sliding window). (c) Vug contours. (d) Vug connectedness log and boundaries of the vugs and conductivity paths (in white)
(from Delhomme, 1992).
DipFan (Dipmeter Facies Analysis) is a complete software package for geological interpretation. It is integrated
to GeoFrame. DipFan has been developed jointly by
Schlumberger and Agip SPA. Integration of dipmeter data
with other openhole logs allows exploitation of the sedimentological and stratigraphic information contain in the
complete logging suite. The interactivity of the chain permits a user-oriented interpretation of the well. The processing chain is built from six modules that can be run to
suit each particular case. Three primary modules, FasTex,
SediView and Sequence, are geologically oriented. They
can be used separately but overall results are enhanced
considerably by the three DimQC, StatPack and RockCell
that are optional utilities modules.
- DimQC performs basic quality control on the dipmeter input data. It represents the first step in the reliable
interpretation of dipmeter data. This module consists of
three parts.
1 - lndividual curve analysis. Individual curves, such as
calipers, tension or navigation (inclinometry), are tested
following user-defined thresholds. These tests allow the
detection of anomalies during the recording.
The dip data are tested in the same manner. All tests
are applied on either a level-by-level or on a window
basis.
The module can be used interactively or in batch mode
with default threshold values. The results of this quality
control can be displayed with doubtful data points indicated by changes in color.
2 - Fast channel qualification. In this part fast channel
data are controlled and any degraded data are crosschecked against those attached to the tension, caliper
and navigation channels. This process allows the degraded data to be classified as tool- or hole-related.
Serralog 0 2003
I
Chapter 1
31
3 - Severity curve computation. Two severity curves,
RawQC and TadQC, are computed in the final phase of
the DimQC module. The RawQC curve is computed from
the fast channel qualification results and the user can
either set the weight of each defect for the severity computation or use the default values. The TadQC severity
curve is then computed combining RawQC values with
the results of dip data and navigation curve checks.
The results are displayed as flagged zones and as
severity curves.
- StatPack is designed to provide the user with a number of basic data handling utilities including analysis techniques (stereoplots, frequency plots), statistical and
mathematical functions together with common graphic
presentations such as azimuth vector plots (Fig. 1-44).
This last plot displays the progression of dip azimuth with
depth and is useful in the identification of major breaks in
bedding.
well test
'
interval. 2900.21 - 3300.00m
Cut-off : Dip quality factor
Cut off (mirdmax) 12 00 / 20 00
Number of points 747
Depth wlor scale
R 2900-2940
2WO-3105
-_.-
__
3105-3173
3 3173-3275
3 3275-3300
Figure 1-44 - Example of azimuth vector plot
(courtesy of Schlurnberger).
For instance, FasTex uses the StatPack library in order
to run Cluster or principal component analysis, and results
are displayed under the form of dendrogram, cross-plot or
spider plot.
When running SediView the library provides statistics
on dips such as azimuth frequency plot, azimuth or dip
histograms, stereonet projection, Bengtson plot, dip vector plot and cross-plot of dip magnitude versus dip azimuth.
Other statistical analyses can be done such as minimum variance analysis or Markov analysis. Other utilities
(e.g. user-defined formulae) are also included.
- RockCell performs multi-well facies analyses based
on the assumption that the facies information is known
and contained in the well logs. One or more wells of a field
are selected as the "key wells" and are used to define the
set of facies estimators. The input data may be the open31
32
Well logging and Geology
hole logs, logs derived from dipmeter data, or images, or
indeed any other depth-indexed facies indicator. A model
is established which relates the estimators to each other
and to facies based on multi-dimensional histogram
techniques, neural network, or multidimensional ellipsoids.
Three phases form the RockCell module.
1 - Key well data analysis. This phase provides the
user with powerful working environment in which he can
display the results from an unlimited number of wells.
Data can be analyzed versus depth or in cross-plots.
Any user interaction in one view is visualized in the other.
Intervals representingvarious facies can be selected and
labeled on both the depth and the cross-plot views. Log
curves which represent a facies description from the key
wells can also be used.
2 - Model building. The model consists of the list of
estimators available in the different key wells and the
multi-dimensional histogram which relates the facies
information to the estimator value distribution. The log
curve axes of the multi-dimensional histogram are userdefined with each axis divided into equal bins. However,
the axes may also be functioned to suit specific needs.
3 Facies estimation. The goal of this final phase is to
assign the most likely facies to each level of the target
well. In practice, a number of facies will be found, and
these are retained in decreasing order of likelihood.
One or more facies models can be used for the interpretation. In this case, each model and the relevant set of
parameters are associated with a particular depth interval.
The results are displayed in a user-defined format
together with the estimators for quality control purposes.
In practice, only the first three facies estimated by the
model are displayed.
-
- FasTex analyses the character of the dipmeter microresistivity curves, or fast channels - electrical imaging tool
data sets may be input to the FasTex module since
pseudo-dipmeter curves that replicate the dipmeter fast
channels can be computed (Fig. 1-45).
As previously explained, these curves reflect the internal organization of the formations encountered by the
well. The module identifies intervals with similar curve
response and proposes a zonation of the formations into
"morphofacies": homogeneous, heterogeneous or layered.
In order to ensure the reliability of the answers, FasTex
is normally run after DimQC, this module having identified
bad hole or floating pad intervals.
The program performs a geologydriven analysis of
the dipmeter fast channels. FasTex algorithms basically
compare dipmeter fast channels with each other along the
apparent bedding planes. The borehole-crossing part is
called the background conductivity log. Conductive and
resistive anomalies (i.e., uncorrelated events) are extracted by detecting peaks and troughs that are present in the
fast channels but not in the background.
Summary logs are used to express the information
contained within the dipmeter fast channels. These curves reflect the layering and the degree of heterogeneityof
the formations. Attributes such as thickness, resistivity
contrast and polarity are used to describe raw curve
events.
Using the StatPack library, a clustering technique is
applied to the summary logs. The result is a segmentation
of the formations in different zones. The user can monitor
the aggregation process through various statistical plots
such as the dendrogram. The number of clusters, or
"morphofacies", is set via the input parameters and the
process is fine-tuned by using a vertical logic which takes
into account a minimum zone thickness.
-
Sediview derives accurate sedimentologicalinformation from dip and lithological results. The module helps
reservoir delineation by inferring geometrical information
from depositional features. The module also helps predict
locations and directions of permeability anisotropy.
SediView consists of four separate steps.
1 Data loading. Pre-computed dip data are loaded
from the GeoFrame database. Lithological information
can also be loaded from the same database, but may be
manually input or computed from application of simple
thresholds to the openhole logs.
2 - Data filtering. A variety of filtering techniques may
be applied to focus the interpretation on meaningful dips.
A typical filter would be the dip quality index, output by
DimQC. In addition, lithological zoning may be specified.
3 Structural dip evaluation. This phase evaluates and
removes the effects of post depositional deformationor tilting. An innovative and accurate approach, based on
Local Curvature Axis techniques, is used to derive the
structural dip which can then be removed in the standard
manner.
-
-
-
Figure 1-45 Image data (center) may replicate fast channels of HDT
(lefi) or SHDT (right) data in order to compare with dipmeter data
recorded in other wells (courtesy of Schlumberger).
32
The structural dip is traditionally determined from
zones where the dips are constant in magnitude and azimuth (green patterns) and that correspond to low energy
deposits (essentially shales or mark). The value of these
dips provide evidence of structural tilt. However, in thick
Serralog Q 2003
Generalities
sandstone intervals with cross-bedding or foresets the
determination of the structural dip is more difficult.
SediView makes use of the observation that the cylindrical or conical bed surface axes within each bedset are
often co-planar. The dip and orientation of this plane
reflects the effect of the post-depositional deformation.
The program for automatic structural dip determination in
such cases is based on the two following observations,
and one assumption (Beaudoin eta/., 1983). The first observation is that usually a few poles of consecutive dips
that are close in depth appear aligned on a great circle on
a Schmidt net (Fig. 1-46). This may be because of sedimentary structures, compaction or micro-structures.
-
Figure 1-46 Sedimentary features. (1 to 3) Their dips correspond to
orange sequences of small extent ffiing great circles that have cleady
different axis orientations.
(4) The axes of these sedimentary features fit a great circle as they
correspond to the same plane that charecterizesthe original depositional surface. This plane is the structural dip (SO)
(from Etchecopar & Bonnetain, 1992).
Whatever their origin, these "local" great circles can be
characterized through the orientation of their poles (Local
CurvatureAxis).
The second observation concerns the poles of the
local great circles. Most of the consecutive poles of local
great circles, also called local curvature axes, are themselves aligned on a great circle (cf. Fig. 1-46 (4)).
The basic assumption is that the plane that corresponds to the local great circle poles, or local curvature
axes, was nearly horizontalduring sedimentation and now
represents the structural dip. In other words, the axes of
microstructures or sedimentary structures, which are
Serrakg Q 2003
I Chapter 1 1
33
recognized as locally cylindrical, are assumed to have
been nearly horizontal during sedimentation.
The first pass of this module computes the local curvature axis of each bedset or laminaset. These axes are
plotted on the stereonet and analyzed interactivelyto define structural dip common to a group of beds. This value
is assigned to the corresponding interval. The technique
detects the subtle changes in structural dip that can occur
between depositional episodes.
4 - Sedimentary dip determination.After subtraction of
the structural dip for each depositional unit, a dip dispersion analysis is performed to organize the bedsets into
groups. The thickness and the azimuth of each depositional unit are then computed. These values are integrated by the interpreter with other data from images, logs
and cores as well as with general knowledge of the depositional environment in order to infer reservoir elongation
and transport current direction. Results are then
displayed.
- Sequence. The interactive Sequence module
consists of a number of independent interpretation submodules. The outputs can be combined with other DipFan
results to provide stratigraphic interpretation capability
with the maximum in user-friendliness.
Bed boundaries, defined either by the DipFan modules, or by stratigraphy and core analysis, are consolidated
and ranked into an organized geological sequence
display.
Three sub-modules form the Sequence processing
chain.
1 - Curve shape analysis. Sequence performs a semiautomatic analysis of the different logs for identificationof
bell, funnel and cylinder shape. The curve-shape analysis
is a two-stage process. Breakpoints are detected by
searching for levels at which the first derivative exceeds a
given threshold. Then, zones defined by two successive
breakpoints are classified in accordance with the slope of
the line that approximates the input log.
2 - Thickness analysis. The thickness analysis recognizes and describes thickening or thinning upward
trends in the analyzed sequence. Three algorithms have
been developed that enable the user to perform this analysis at different scales. A large scale, quick-look
approach over the whole sequence can be run using gravity analysis. For each of the facies, the average location
along the sequence is computed. This position, with
respect to the mid-point of the interval under study, indicates whether the facies is randomly distributed or shows
a preferred trend.
On a finer scale, regressionanalysis can be performed
within zones defined either automatically, or by the user.
Alternatively, parasequence stacking pattern analysis can
be inferred through interpretation of Fisher plots. The
Fisher plot is a simple graphic presentation of the vertical
trend of parasequence thickness. The resulting patterns
illustrate the deviation from the average throughout the
33
34
Well logging and Geology
stratigraphic interval.
-
3 Lithology analysis. This analysis statistically describes the lithology distribution within a sequence.
Petrophvsics
Included in the petrophysical product line are
PetroView Plus*, Rockclass* or Rockcell, Quickview,
ELANPlus* and MultiWell ResSum*.
Petroview Plus delivers accurate detailed petrophysical analyses quickly. This application guides you through
a deterministic petrophysical evaluation step-by-step.
PetroView Plus offers interactive parameters selection
from crossplots, histograms and log displays, and provides graphic presentations of input and output data.
Rockclass produces a zoned lithological display and
describes the gross mineralogy of a lithofacies, using
wireline log measurementsor computed mineral volumes
with a lithofacies database (cf. Chapter 2).
The lithologic display can be particularly helpful in correlating non-depth-indexed data such as seismic, or poorly-indexed well data such as cores, mud logs or cuttings.
The log derived lithological column becomes an especially powerful geologic tool when combined with dipmeter
correlations and high-resolution images.
The lithofacies database may be created interactively
by the user, or recalled from a collection of previously
saved files.
The Quickview petrophysical program provides quick
and easy analysis and all the necessary graphics and
tabular output to make accurate well decisions. This
module offers a standard methodology based on a well
log analysis model used since 15 years - for analyzing
openhole logs, providing both Archie and dual-water saturation solutions, effective porosity, shale volume, and
apparent grain density.
-
ELANPlus is an advanced petrophysical analysis
module that is derived from the well known ELAN software.
MultiWell ResSum is a module that computes reservoir thickness and associated propertiesfor multiple wells
and layers. To increase the precision of the interpretation,
true vertical thickness and true stratigraphic thickness calculations are available.
References and bibliography
ACKERT, D., BOETEL, M., MARSZALEK, T., CLAVIER, Ch., GOODE, P., THAMBYNAYAGAM, M., &
STAGG, T. (1988). Looking Sideways for Oil. The
Technical Review, 36, 1.
ALLEN, D. et a/. (1995). Modeling Logs for Horizontal
Well - Planning and Evaluation. Oilfield Review, 7, 4.
-
-
34
AULIA, K. et al. (2001). - Resistivity behind casing.
Oilfield Review, 13, 1, p. 2-25.
BABOUR, K., JOLI, F., LANDGREN, K., & PIAZZA,
J.L. (1987). Elements of Borehole Seismics illustrated
with three case studies. The Technical Review, 35, 2.
BATES, R.L. & JACKSON, J.A. (1980). - Glossary of
Geology. Amer. Geol. Institute, Falls Chuch, Virginia.
BEAUDOIN, B., PINOTEAU, B., DELHOMME, J.P.
(1983). - Geological analysis of logs and dipmeter data for
well-to-well correlation.
BEGUIN, P., BENIMELI, D, BOYD, A, DUBOURG, I.,
FERREIRA, A., McDOUGALL, A., ROUAULT, G. & van
der WAL, P. (2000). Recent progress on formation resistivity measurement through casing. SPWLA, 41th Ann.
Log. Symp. Trans., paper CC.
BENIMELI, D., LEVESQUE, C., ROUAULT, G.,
DUBOURG, I.,PEHLIVAN, H., McKEON, D, FAIVRE, 0, &
REBOLLEDO, K. (2002).- A new technique for faster
resistivity measurements in cased holes. Schlumberger.
BLATT, H. MIDDLETON, G., & MURRAY, R. (1972,
1980). Origin of sedimentary rocks. 1st and 2nd ed.
Prentice-Hall Inc. Englewood Cliffs, New Jersey.
BONNER, S. et a/. (1993). Measurements at the Bit:
A New Generation of MWD Tools. Oilfield Review, 5, 2/3.
BOURKE, L., DELFINER, P., TROUILLER, J.C., FETT,
T, GRACE, M., LUTHI, S. SERRA, O., & STANDEN, E.
(1989). - Using Formation Microscanner* Images. The
Technical Review, 37, 1.
CHEUNG, Ph., & HELIOT, D. (1990). - Workstationbased fracture evaluation using borehole images and
wireline logs. SPE Bull., no20573.
De GOLYER, E. (1921). - Debt of Geology to
Petroleum Industry. Bull. amer. Assoc. Petroleum Geol.,
5, p. 394-398.
DELHOMME, J.-P. (1992). A quantitative characterization of formation heterogeneities based on borehole
image analysis. SPWLA, 33d Ann. Log. Symp. Trans.,
paper T.
DELHOMME, J.-P., & MOTET, D. (1993). - Reservoir
description and characterization from quantitative bomhole image analysis. Schlumberger, Reservoir Characteriza.
tion, Africa and Mediterranean Region, no 5.
DESBRANDES, R. (1985). - Encyclopaedia of Well
Logging. Editions Technip, Paris.
DOVETON, J.H. (1994). - Geologic Log Analysis
Using Computer Methods. AAPG Computer Applications
in Geology, no.2.
DOVETON, J.H. (2002). The Geological application
of Wreline Logs: A Keynote Perspective. In: LOVELL, M.,
& PARKINSON, N. (ed.) (2002). - Geological Applications
of Well Logs. AAPG Methods in Exploration, 604 p.
ELLIS, D.V. (1987). - Well Logging for Earth Scientists.
Elsevier, Amsterdam.
HARKER; S.D., McGANN, G.J., BOURKE, L.T,. &
ADAMS, J.T. (1988). - Formation Microscanner Applications in Integrated Reservoir Description in
Claymore and Scapa Fields (North Sea). Geological
Applications of Wireline Logs, London, June 1-2, 1988.
Geological Society of London.
-
-
-
-
-
-
Serralog Q 2003
Generalities
HARKER, S.D., McGANN, G.J., BOURKE, L.T., &
ADAMS, J.T. (1988). - Methodology of Formation
MicroScanner lmage lnterpretation in Claymore and
Scapa Field (North Sea). Journ. Geol. Society of London.
HURST, A., LOVELL, M.A., & MORTON, A.C. (eds)
(1990). - Geological Applications of Wireline Logs.
Geological Society, Special Publication no 48, London.
HURST, A., GRIFFITHS, C.M., & WORTHINGTON,
P.F. (eds) (1992). - Geological Applications of Wireline
Logs /I. Geological Society, Special Publication no 65,
London.
HUTTON, J. (1788). - Theory of the Earth; or an investigation of the laws observable in the composition, dissolution, and restoration of land upon the globe. Royal
Society of Edinburgh. Transactions, v. 1, p. 209-304.
ILLING, V.C. (1946). - lnfluence of Oil on Geology. Oil
Weekly, September, p. 176.
ISLAM, M.R. (2000). - Emerging Technologies in
Subsurface Monitoring of Petroleum Reservoirs. 1st
Symposium on Well Log Analysis & Formation Evaluation
(WLA&FE), 29-31 Oct. 2000, Tripoli, Libya.
KHALIL, Mi, FERRARIS, P. & DAWOUD, A. (2001). Saturation monitoring with cased-hole formation resistivify. SPE Annual Technical Conf. and Exhib., New Orleans,
Louisiana, paper SPE 71716.
LAHEE, F.H. (1932). - Contribution of Petroleum
Geology to pure Geology in the Southern Mid-Continent
Area. Bull. amer. Assoc. Petroleum Geol., 43, p. 953-964.
LOVELL, M., & PARKINSON, N. (ed.) (2002). Geological Applications of Well Logs. AAPG Methods in
Exploration, 604 p.
LUTHI, S. (2001). Geological Well Logs. Their use in
reservoir modeling. Springer.
LUTHI, S., & SOUHAITE, Ph. (1990). - Fracture aperture from electrical borehole scans. Geophysics, 55, 7, p.
821-833.
LYNCH, E.J. (1962). - Formation Evaluation. Harper &
Row, New York.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed., Gulf Publishing Co., Houston.
POSTER, C. (1988). - Lateral Thinking in Seismics.
Schlumberger, Middle East Well Evaluation Review,
Special supplement: Borehole Seismics.
RIDER, M.H. (1986). - The Geological lnterpretation of
Well Logs. Blackie, Glasgow & London, Halsted Press,
New York.
Schlumberger (1991). - Log Quality Control
Reference Manual. M-090209.
SERRA, 0. (1984). - Fundamental of well-log interpretation. 7. The acquisition of logging data. Elsevier,
Amsterdam, Developments in Petroleum Science 15A.
SERRA, 0. (1986). - Fundamental of well-log interpretation. 2. The interpretation of logging data. Elsevier,
Amsterdam, Developments in Petroleum Science 158.
SERRA, 0. (1989). - Formation MicroScanner lmage
lnterpretation. Schlumberger, SMP-7028.
SERRA, 0. (1993). - Relations entre les donnees diagraphiques et les parametres geologiques. Les differents
domaines d'applications geologiques des diagraphies.
I ChaRter 1 I
35
Gkologues, no 99.
SERRA, 0. (1998). Role of well logging in the study
of sedimentary basins. In: Purser, B.H. (ed), Dynamics
and methods of study of sedimentary basins, Oxford and
IBH Publishing Co., New Delhi, India. Editions
Technip,Paris, p. 11-57.
SERRA, O., & ABBOTT, H. (1980). - The contribution
of logging data to sedimentology and stratigraphy. 55th
Ann. Fall Techn. Conf. SPE of AIME, paper SPE 9270.
SERRA, O., STOWE, I., & MOTET, D. (1993). - True
integrated interpretation. SPWLA - CWLS 34th Ann. Log.
Symp. Trans., paper 2.
THEYS, Ph. (1991). - Log data acquisition and qualify
control. Editions Technip, Paris.
TITTMAN, J. (1986).
Geophysical Well Logging.
Academic Press, Inc., Orlando, Florida.
Van WAGONER, J.C., MITCHUM, R.M., CAMPION,
K.M., & RAHMANIAN, V.D. (1990). - Siliciclastic
Sequence Stratigraphy in Well Logs, Cores, and
Outcrops. AAPG Methods in Exploration Series, no 7.
-
-
Remark : In all the chapters a * refers to Schlumberger's
mark.
-
-
Serralog Q 2003
35
WELL LOGGING AND ROCK COMPOSITION
Introduction
Review of general geological concepts
edly be useful to recall some basic concepts of geology,
petrography and mineralogy to non-geologists.
The three major rock types
The nature of a rock and its composition are the first
characteristics which the geologist attempts to determine,
and a knowledge of these enables him to name the rock.
This is also his first concern in the study of well logs.
He will therefore attempt to reconstruct the vertical
lithological profile from an analysis of the well logs. This
reconstruction is aimed at defining:
-the apparent thickness and the real thickness of each
electrobed, having an electrolithofacies and electrofacies,
or of each electrosequence,
- the type of rock and the elemental and mineralogical
composition of each electrobed.
In order to obtain the most reliable reconstructionpossible, the interpreter, or log analyst, must first of all procure a suite of logs which is as complete as possible, and
in addition, a good description of cores and cuttings from
the drilling process. Secondly, he must possess the basic
concepts which are necessary for a good understanding
of the problems involved in analyzing logs, and the implications that these will have in the choice of an interpretation model, be it manual or automatic.
Clearly, the interpretationof a volcanic or granitic rock
will proceed differently, and therefore will require a different model, from the interpretation of a sedimentary rock
such as an argillaceous sand made up of a mixture of
quartz, feldspar, kaolinite, illite and montmorillonite.
Moreover, taking account of the frequently limited availability of data and recorded logs, it is not always possible
to simultaneously determine the nature of the minerals
present and their proportions, especially in the case of
complex composition.
It is often necessary, therefore, to obtain additional
information from examinations of cuttings, core analysis
and local geological knowledge to establish the mineralogy model, and thereafter to restrict the investigation to
determining the percentages of each of the principal minerals assumed to be present in the rock.
Before proceeding with the problem of determining
rock type and composition from log data, it will undoubt-
Serralog 0 2003
Geologists divide rocks into three main categories
according to their mode of formation.
Igneous Rocks
These rocks arise from the solidification of a molten
mass known as magma. Depending on whether the solidification takes place at depth or on the surface, the rock
can have a coarse texture (slow crystallization), or a fine
or even glassy texture (rapid crystallization). The first type
is known as plutonic or intrusive igneous rock, and the
second as volcanic or extrusive igneous rock.
Plutonic rocks have the distinction of exhibiting practically no porosity, the crystals being tightly packed.
Alteration and fracturing of plutonic bodies usually favour
the development of some porosity and permeability,
thereby conferring some of the characteristic of a reservoir. Examples of hydrocarbon production from granitic
rocks exist in several countries: India, Egypt, Vietnam,
Libya.. .
Volcanic rocks, on the other hand, can be porous or
even very porous, but the pores are not always connected (pumice, vesicular basalt). They correspond to bubbles
separated by a thin lining of volcanic glass, which are
formed by violent depressurizationwhen fragments of viscous magma are projected into the atmosphere.
Plutonic rocks generally form the basement rock of
sedimentary sequences. Their upper zones are often
weathered, or broken down and may constitute a reservoir. Sometimes they can be found as intrusions, such as
dikes or sills, in the sedimentary formations (Fig. 2-1).
Volcanic rocks may be intercalated in the sedimentary sequence, and can therefore be encountered at any
level of the stratigraphic column. Dependingon their characteristics, they may constitute a reservoir rock.
37
38
Well logging and Geology
Figure 2-1 - Different types of intrusions of igneous rocks in sedimentary rocks (adapted from Schmidt & Shaw in Press & Sieve6 1982).
Sedimentary Rocks
These rocks arise from the consolidation of sediment
formed on the surface of the Earth or on the seabed by
the deposition of various materials, usually under the
action of gravity, acting on rock fragments or minerals of
any size transported by water, wind, or ice from their
source or parent-rock, but also by chemical precipitation
from solution or by secretion from living organisms. More
often, they are deposited in laminae, beds or strata. This
family of rocks is divided into several groups based on the
origin of the sediment.
Detrital or Clastic Rocks
These are formed from debris arising from the alteration and decomposition of pre-existing rocks, and may
be transported, often a considerable distance, by wind,
water or ice from the site of erosion and weathering to the
site of deposition.
Sediments which settle under the action of gravity at a
distance from their source are termed allochthonous or
exogenous. The particles are usually bound together by a
cement of chemical or biochemical origin formed subsequent to the deposition. It occupies part of the pore space.
This group is divided into several subgroups:
- Terrigenous rocks are formed from accumulations of rock debris from the alteration and erosion of
land-based outcrops.
- Pyroclasticrocks result from the accumulationof
fragments of solidified magma expelled into the atmosphere from volcanoes, deposited under the action of gravity, and then re-worked or altered to varying degrees by
the action of water.
- Bioclastic rocks result from the accumulation of
skeletons and other animal remains, typically carbonate
shells, and sometimes remains of vegetation.
which fall out of solution following changes of pressure,
temperature or concentration (evaporates), or in response
to chemical changes within the water, due to the activity
of organisms such as plankton, algae or bacteria (carbonates). Such rocks are termed autochthonous or endogenous because the site of deposition is usually the same as
the site of formation.
Frequently included in this group are the biochemical
rocks which result from the action of organisms such as
reef-building coral, or from an accumulation of organisms
having calcareous shells, or of siliceous organisms (giving
chert, radiolarites, diatomites), or from the transformation
of vegetable debris (humic or sapropelic) under the action
of anaerobic bacteria (giving peat, lignite, coal, hydrocarbons).
With the exception of purely chemical rocks of the
evaporate type, sedimentary rocks more often than not
exhibit connected intergranular, intraparticle and vuggy
porosity, which renders them potential reservoir rocks in
which fluids can accumulate (water, oil, gas). As a result,
they represent a major factor in the search for these substances.
Metamorphic Rocks
These result from the chemical, mineralogical, textural
and structural transformation of rocks under the action of
high temperatures, and frequently of high pressures. They
are divided into two groups depending on the type of
metamorphism by which they are created.
- Rocks associated with a general or regional metamorphism result from deep burial, and hence the simultaneous action of heat and pressure on pre-existing rocks,
and facilitate the modification of both texture and structure, and the formation of new minerals. This type of
metamorphism affects bodies of rock over large areas
and depths.
- Rocks associated with contact metamorphism are
produced by a mineralogical transformation of formations
in the vicinity of igneous intrusions, usually under the
influence of temperature alone.
The type of rock formed will depend on that of the original rock. Metamorphic rocks present any porosity or permeability other than that associated with the existence of
fractures. Regional metamorphic rocks sometimes form
the basement rock of sedimentary sequences.
Relative abundance of rocks
According to Clarke (1924), igneous rocks represent
95 % of the volume of the Earth's crust (lithosphere up to
a depth of 16 km), sedimentary rocks accounting for only
5 % (Table 2-1). If, however, one only considers the
exposed surfaces, sedimentary rocks account for 75 %,
while igneous rocks account for 25 % (Fig. 2-2).
Chemical and Biochemical Rocks
These are formed by the accumulation of precipitates
38
Serralog 0 2003
Composition
Table 2-1
Estimated volume of sedimentary rocks
(from Pettijohn, 1975).
kteference
Cubic kilometers
larke (1924)
oldschmidt (1933)
uenen (1941)
ickman (1954)
oklervaart (1955)
om 8 Adams (1966)
onov (1968)
3.7 x lo8
3.0 x lo8
13.0~10~
4.1 f 0.6 x 10'
6.3 x 10'
1 0 . 6 ~10'
I Chapter2 I 39
alone 90 % of the total.
As far as the sedimentary rocks are concerned,
Garrels 81MacKenzie (1971) have estimated that the proportions of clays, sands or sandstones, and carbonates
are 81 %, 11 % and 8 % respectively.
These values can be compared to those given by
Mead (1907), Clarke (1924), Holmes (1913), and
Wickman (1954) (cf.Table 2-4).
Table 2-3
Representationof the abundance of various types of volcanic rock in terms of the area they occupy
(from Daly, 1933).
9.0 x 10'
Exbusive Rocks
4.8 x 10'
According to the data of Wedepohl (1969, Table 2-2),
the granites, granodiorites, and quartz diorites represent
86 % of all plutonic rocks, the remainder being made up
of gabbros (13 %).
Pacific
'otal squan
Cordillera, Belt, square
miles
iquare miles
miles
Rhyolite
Dacite
2,145.7
82.1
Mica andesite
Hornblende andesite
Pyroxene andesite
(Ch@W)
3.0
21.6
3,966.0
Augite porphyrite
Basalt
Trachyte
Laiite
Phonoliie
Trachydolerite
eschenite
Nepherie basalt
(Texas)
Nephelite-meleliie
basalt (Texas)
Limburgite
Quah basalt
Totals
255.0
3,079.0
6.5
4.6
1.o
.....
.......
.......
.......
.......
130.0
5.5
0.3
02
2,146.7
82.1
3.0
21.6
3,966.0
255.0
3,209.0
0.3
1.2
.......
1.2
2.8
.......
2.8
2.5
8.0
9,584.0
.......
8.0
9,715.0
131.0
-
Figure 2-2 Histograms showing the relative abundance of igneous
and sedimentary rocks :(a) as a volumetric percentage of the Earth's
crust to a depth of 16 km, and (b) as a percentage of the surface exposure (fivm Pettijohn, 1949, based on data by Clarke, 1924, p. 34).
Table 2-4
Computed proportions of sedimentary rocks
(completed from Pettijohn, 1975).
I
Plutonic rocks
Granite and quartz monzonite
Granodiorite
Quartz diorite
Diorite
Gabbro
Others
TOM
Serrabg @ 2003
Percentage
I MtADm
A
-L
CI
(1907)a (1924)b (1913)~ (1954)d MACKENZIE
(1971)
Shale
82
80
70
83
81
Sandstone
12
15
16
8
11
Limestone
6
5
14
9
8
44
34
8
1
13
1
100
39
40
Well logging and Geology
Rock classification
In general terms, the classification of rocks is based on
the one hand on composition in terms of essential mineral percentage (e.9. quartz, feldspars, micas, amphiboles,
pyroxenes; calcite, dolomite; clays, etc.), and on the other
hand on textural properties such as the size of the crystals or particles and their arrangement.
Figs. 2-3 to 2-8, taken from Press & Siever (1982) and
Pettijohn (1975), give a summary of the classification of
rocks pertaining to the three fundamental families. Table
2-12 provides the composition, texture, parent rock, and
type of metamorphism associated with each of the metamorphic rocks.
Figure 2-4 - Classification of sedimentary rocks
(from Pettijohn, 1975, fig.2-3).
Metamorphic rocks are subdivided as a function of
temperature, pressure and foliation or granulation (Fig. 27). The composition of the rock which has been metamorphosed can also be taken into acount (marble from carbonates, quartzite from sandstones, gneiss from granite,
slate from shale, etc.).
Figure 2-3 - Classification of igneous rocks
(adapted from Press & Siever, 1982).
Sedimentary rocks are subdivided into exogenetic or
clastic rocks and endogenetic or chemical and biochemical rocks (Fig. 2-4)Pettijohn (1975). Clastic or detrital
(exogenetic) rocks are subdivided into cataclastic, pyroclastic, residues and epiclastic sedimentary rocks the latter including conglomerates, sandstones and mudstones
or shales as a function of the grain size (Fig. 2-5).
Figure 2-5 - Classification of detrital sedimentary rocks
(from Press & Siever, 1982, fig. 3-18a).
Rock composition
Sandstones themselves are subdivided into arenites
and wackes (Fig. 2-6 next page) as a function of the composition of the grains, their percentage and their size.
Chemical and biochemical rocks, called by Pettijohn
endogenetic rocks, are subdivided into non-evaporitic
rocks, which include limestones, dolostones, ironstones
and cherts, and evaporates which include gypsum, anhydrite, halite and many other salts such as polyhalite, carnallite, langbeinite, kainite, kieserite, sylvite, tachydrite,
bischofite, epsomite, trona and natron (Fig. 2-4 and Table
2-5).
40
The solid fraction of a rock is composed of minerals,
under the shape of grains or crystals, themselves com.posed d...moleGu198..m&-.I~ Qf.elementsQr-afoms(Fig.
2-8). The elements are listed in Mendeleiev’s periodic
table (Fig. 2-9).
Elements are fundamentally composed of a nucleus,
made of protons and neutrons, and electrons revolving
around it. Elements can combine to form molecules of
more or less complex composition.
Serralog 0 2003
Composition
I Chapter2 I
41
LlTNlC ARENITES
-
Figure 2-6 Classification of sandstones (adapted from Pettijohn, 1975, fig. 7-6).
Greenschist
facies
Zeolite
facies
Amphibolite
facies
Pyroxene
hornfels
facies
Temperature increase
Sanidinite
f8cies
t
Poorly foliated
Marble
Amphifflite
Quartzite
Foliated
Gneiss (coarse grained)
Schist (medium grained)
Sate (fine grained)
Tectonic granulation
Granulated
Mylonites
Cataclasites
-
Figure 2-7 Classification of metamorphic rocks. The textural gradations (bottom of diagram) do not necessarily correlate with the mineralogical
facies controlled by temperature and pressure (from Press & Sieve6 1982, fig. 3-19).
Serralog 82003
41
42
.
Well logging and Geology
Table 2-5
Classification of chemical or biochemical rocks (from Press & Siever, 1982, fig. 3-18b).
IrOnaOnOS
Evaporates
/Chefl
/Organic
Phosphates
I
Mineral
Calci
(aeonW
Dolomite
Ankerite
Magnesite
Coal
Oil
Gllr
Carbon +
hydroeen
Chemical
mporition
<1%
2.1%
2.3%
2.4%
4%
696
8%
28%
46%
Figure 2-10 - Relative abundance of elements in the earth's crust
(from Press & Siever, 1982). Compare with Table 2-6.
-
Figure 2-8 From element to outcrop through mineral and rock
(redrawn from Press & Siever, 1982).
Elemental covmposition
Table 2-6
Elemental composition of the Earth's crust
(igneous and sedimentary rocks)
recomputed from Clarke & Washington, 1924.
If 93 natural elements have been recognized, one has
to consider that only 8 are abundant as constituents of the
Earth's crust (Table 2-6 and Fig. 2-10). Elements are classified on the basis of electronic structure (Fig. 2-9).
The 8 most abundant elements represent more than
99% of the total mass of the Earth's crust. Also, as it can
be observed, oxygen is the most abundant component
both in weight percentage, atom percentage and in volume percentage. Oxygen is associated to a lot of other
elements to compose molecules and minerals. Table 2-7
lists the oxygen content of the most abundant minerals. In
average its weight percentage is close to 50%.
42
2.08
3.65
2.75
2.58
0.62
0.14
Atom, %
60.5
20.5
6.2
1.9
1.8
1.9
2.5
1.4
0.3
3.0
0.51
0.03
0.44
0.37
028
1.04
1.21
1.88
....
0.36
0.56
0.70
0.70
0.65
0.99
0.95
1.33
....
Serralog 0 2003
Composition
Table 2-7
Oxygen percentage of the most abundant minerals compbsing the Earth's crust.
Mineral
Quartz
Calcite
Domolite
Anhydrite
Orthose
Albite
Anorthite
Muscovite
Biotite
Glauconite
lllite
Kaolinite
Chlorite
Montmorillonite
43
them.
The chemical compositions, expressed as oxides, of
the more abundant rocks making up the Earth's crust are
presented in Tables 2-8 to 2-12 and Figure 2-11.
Oxygen (weight percentage)
53
48
52
47
46
48
48
50
55.7
52
53
By gain or loss of electrons, atoms become cations
(lost electrons) and anions (gained electrons). Cations
and anions bond to form molecules. Most of the crystal
space is occupied by anions that are generally larger than
cations, cations (smaller in size) fit into spaces between
Atome
1 Chapter 2 I
I
Figure 2- 11 - Oxide composition of some sandstones.
Less abundant but major geological importance
Metal
U:Solid at 25°C under 1 bar
Semi-conductor ~e : Gas at 25°C under 1 bar
Non metallic
Br: Liquid at 25°C under 1 bar
TC: Obtained by synthesis
Noble gases
Lanthanides & Actinides
Radioactive
The most abundant
Figure 2-9 - Periodic table of elements. The most abundant elements are indicated with a blue circle. The radioactive elements are indicated by a
red circle. The elements of lesser abundance in the earth's crust but of major geologic importance are indicated by an orange circle. They are constituents of common rock-foning minerals or of economic significance.
serralog 8 2003
43
Table 2-8
Chemical composition of the Earth's crust
(percentages by weight)
(from US Geol. SUN. paper 127, 1924).
Table 2-10
Chemical composition of the principal volcanic rocks
(from Daly, 1933).
Bfusive
Wde
Clarke and
Washington
59.07
15.22
3.10
3.71
0.11
3.45
5.10
3.71
3.11
1.30
1.03
0.30
72.m
0.33
13.49
1.45
0.88
0.08
0.38
1.P
3.38
4.46
1.47
0.08
Bfusive
Wte
6fuswe
jfusive Basal
Andeste
66.68
0.57
58.58
16.25
2.38
1.90
0.06
1.41
3.46
3.97
2.67
1.50
0.15
17.31
3.33
3.13
0.18
275
5.80
3.58
2.04
1.26
0.26
49.06
1.36
15.70
5.38
6.37
0.31
6.17
8.55
3.1 1
1.52
1.62
0.77
0.45
0.35
0.05
0.06
0.04
0.03
Table 2-9
Chemical composition of the principal plutonic rocks (from Daly, 1933).
FlutonicGranite
=2
Ti02
A403
Fe203
FeO
MlO
Me0
CaO
k20
60
YO
125
'
44
70.18
0.39
14.47
1.57
1.78
0.12
0.88
1.99
3.48
4.11
0.84
0.19
PkROnic
Gramdorite
65.01
0.57
15.94
1.74
2.65
0.07
1.91
4.42
3.70
2.75
1.04
0.20
Rutonic Quartz I Flutonic M
oioriie
61.59
I 56.77
0.66
0.84'
16.21
16.67
2.54
3.16
3.77
4.40
0.10
0.13
2.80
4.17
5.38
6.74
3.37
3.39
2.10
2.12
1.22
1.36
0.26
0.25
I
eI
I
I
Autonic
GaMro
48.24
0.97
17.88
3.16
5.95
0.13
7.51
10.99
2.55
0.89
1.45
0.28
Flutonic
Anorthosite
50.40
0.15
28.30
1.06
1.12
0.05
1.25
12.48
3.67
0.74
0.75
0.05
Flutonic lhnite
40.49
0.02
0.86
2.84
5.54
0.16
46.32
0.70
0.10
0.04
2.88
0.05
Serralog 0 2003
Composition
Table 2-11
Chemical composition of the principal sedimentary rocks
(from Clarke; Leith & Mead; eta/.).
Average
Shale
58.90
%O
A$%
15.63
4.07
2.48
2.47
3.15
1.32
3.28
3.72
F824
FeO
w
-0
4 0
60
YO
+
0.86
Ti2
0.17
2.87
1.48
p205
(32
MSce#aneoUS
lw.w
Average
sandstone
Average
lmtm
78.64
4.77
1.08
0.30
1.17
5.51
0.45
1.32
1.33
0.25
0.08
5.03
0.07
100.00
5.20
Table 2-13
Elemental composition of ocean waters.
Chbrine
sodim
Magnesium
0.81
0.54
7.92
42.74
0.05
0.33
I Chapter 2 1 45
I
Ppm
18980
10501
1272
Sulfur
884
Calcium
Rdassium
Others
400
380
118
Mineralogical composition
0.56
0.08
As previously mentionned rocks are essentially composed of minerals. Consequently, it is fundamental to
determine them to recognize the rock type.
0.04
41.70
0.05
Approxymately 3,300 minerals have been identified,
but the vast majority of them are rare and only found
either as trace minerals or occasionally, especially in
igneous and metamorphic rocks.
lw.w
Table 2-12
Chemical composition of the principal metamorphic rocks.
1
7293
12.67
blm
2.08
0.62
1.W
3.19
4.55
050
0.11
0.54
0.11
2
6133
1524
3.31
3.43
2.6l
4.79
329
2.05
2.38
3
5195
12.56
0.80
6.m
7.00
2.79
131)
2.67
0.14
1.a3
9940
1.02
0.43
9958
2.665
.....
.....
021
4
5153
1081
290
9.93
8.74
9.97
2.62
0.W
0.67
0.00
8.n
0.50
0.70
0.02
02%
99.96
0.38
9.
c
I
OM
024
0.57
0.71
lW.16
2.95%
5
16.45
132
4.89
5.04
3.17
2.62
2.14
0.68
0.03
0.54
Tl8Ce
6
6348
17.70
3.02
1.a
3.72
2.72
1.78
3.34
.0.87
0.23
0.52
0.58
0.48
100.38
2.m
0.45
100.81
2.680
63.04
7
8
on
8329
0.42)
1.or
5487
1. Granite gneiss, Argentina. M. Dittrich, analyst.
3. Chlorite schist, Lake Superior greenstone. S. Darling, analyst.
4. Hornblende schist, Lake Superior greenstone, near granite. F. Grout, analyst.
5. Biotitc schist, Lake Superior, from slate. F. Grout and S. Darling, analysis.
2.75
,
......
....
...........
......
) 4.40
......
...
......
100.35
......
......
10090
10
92M
421
1.w
427
3595
....
0.88
8.04
0.16
1.I6
0.96
......
43.49
2. Diorite gneiss, British Guiana. J. B. Harrison analyst.
124
0.16
4.68
27.13
TNC~
9
40.42
1Ad
0.18
1051
021
MW
Trace
1M
124
90.47
"
0.14
021
.....
100.68
11
4283
1.53
31.41
0.30
Mile
2337
b W
Trace
0.43
9987
In fact, and fortunately, the various rock types are
made up of a reduced number of minerals, and if one
considers the igneous rocks, not more than ten essential
minerals (or major components) are required for each
rock (Table 2-14).
6. Slate, Lake Superior, Knife Lake slate. D. Manuel, analyst.
7. Marble. Georgia. W. H. Emerson, analyst. Other marbles may be dolomitic.
8. Soapstone, steatite, South Africa. Van Riesen, analyst.
9. Serpentine, Massachusetts. George Steiger, analyst.
10. Quartzite, South Mountain, Pennsylvania. F. A. Genth, analyst.
Table 2-15 (next pages) lists the ninety most common
minerals which may be principal components, secondary
constituents or occasional inclusions in the composition of
a rock.
11. Garnet rock, Clifton, Arizona. George Steiger, analyst.
To be complete it is also important to mention the
average elemental composition of the ocean water (Table
2-13) as several rock types can be formed from those elements
According to Krynine (1948), "althoughover one hundred and sixty different minerals have so far been identified in sediments, less than twenty mineral species form
well over 99 per cent of the bulk of sedimentary rocks...
Table 2-14
Mineralogical composition of some plutonic rocks
(from Wedepohl, 1969).
Granite
Granodiorite
Quartz
diOrite
Diorite
Gabbro
Av. Igneous
rock(')
30
37
46
53
63
47
35
21
15
22
8
2
3
56
11
16
1
5
13
3
12
5
12
5
1
6-4
3
8
16
2
16
2
% In volume of:
Plagiodase
Quartz
Potassium feldspar
(including micro-perthite)
Amphbole
Biotite
Orthopyroxene
Clinopyroxene
Olie
Magnetite, ilmenite
Apatite
2
2
2
3
5
4
0.6
2-5
0.5
0.5
0.5
0.8
0.6
0.5- 1
n Clarke eta/., 1924.
They are known as "major" or "main" constituents".
Krynine adds: "it is rare indeed that more than five or six
minerals occur in sizeable amounts in any one rock ... In
addition, approximately twenty other minerals, known as
"accessory minerals", occur in very small amounts in sediments, although locally they may be of great importance".
These minerals are listed in Table 2-16, and the mineralogical compositions of the principal sedimentary rocks
are given in Table 2-17.
Reducing the number of fundamental minerals to 10,
still 95% of all the sedimentary rocks can be composed.
The important point is that sedimentary rocks are usually composed of a mixture of at most four minerals or
major contituents, that is, having a content of more than 5
%. The concept of "end-members", or major constituents,
introduced by Krynine (1948) and extended by Pettijohn
(1949) to describe the composition of a rock, is based on
this observation.
The extended form of the concept proposes that the
composition of any sedimentary rock in terms of minerals
can be represented by a point within a triangle or tetrahedron whose apexes correspond to the end-members in a
pure state (Fig. 2-12).
46
Table 2-16
The principal minerals found in sedimentary rocks
(from Krynine, 1948).
Major Constituents
>10%ofrock
I
< 10%ofrock
Accessory Minerals
I<l%ofrock
QUARTZ
Microdine
CLAY
MINERALS
(kaolin-bauxite)
FINEGRAINED
MICAS
(illite, sericite
muscovite)
DETRITAL CHERT
Sodic plagiodase
(albite-oligodase)
Coarse-grained
micas: muscovite
biotite chlorite
Hematite
Limonite
"IRON ORES":
MAGNETITE,
ilmenite,
DETRITAL
LEUCOXENE
STABLE GROUP:
ZIRCON TOURMALINE,
rutile
UNSTABLE GROUP:
APATITE, EPIDOTE
GARNET, HORNBLENDE
kyanite, sillimanite,
staurolite, titanite, zoisite
MICAS: frequently occu
as accessories rather thai
as major constituents
CALCITE
DOLOMITE
ANKERITE
ANATASE;
CHERT and opal
authigenic rutiie
"SECONDARY" QUARTZ
GYPSUM and anhydrite, and leucoxene
halie
Some hydromicas of the illite
sericite-chloriteseries
Phosphates and glauconite
Siderite and some iron ores
Serralog 0 2003
Composition
I Chapter2 I 47
Table 2-15
The 90 most abundant minerals.
Goethie
Hematite
Carbonate
group
Sulfide group
Sulfate
group
Chloride
group
Clay
group
Feldspar
group
Pyroxene
group
Amphbole
PUP
Mica
group
Olivine
group
Garnet
group
EDidote
group
Nepheline
Qroup
2eoIine
group
Zircon
,group
Phosphate
group
I
Alumina (bauxite)
Magnetite
lRAe
ICalcite, aragonite
Dolomite
Ankerite, magnesite
Siderite
Stroniianite
Trona, natron
Pynte
Marcasite
Galena
Chalcopyrite
Gypsum
Anhydrite
Epsomite
Barite, Celestine, Kieserite,
Langbeinite, Polyhalie
Halie
Syhrie, Bischofite, Carnallite, Tachydrite
Kaoiinite, dickite, halloysite
Smectite (montmorillonite), nontronite
beidellie. saponite
llliie
Chlorite (chlinochlore, thuringite)
Chamosite
Glauconite
Potassium Feldspars:orthoclase,
microline. anorthoclase
Alkali plagioclases: albie
otigocIase
andesite
Calcic plagioclases: labradorite
bytormite
anorthite
Enstatite, hypersthene,
diopside, augite
Hornblende,
antophyllie,
tremolie, actinolite,
glaucophane
Muscovite, biotite, lepidoliie, phlogopite
chlorite, pyrophillite
Oliiine. favalite. forsterite,
monticetiiie
Calcic garnets = grossularite, andradiie
IN0 calcic = almandine, pyrope, spessartite
End member
I
Figure 2-12 - Representation of rock composition by means of a
tetrahedron (from Krumbein, 1954).
This method of classification is based on the subdivision of a tetrahedron. Sectioning a triangle (or tetrahedron) into lines of equal percentage or equal ratio will
define areas (or volumes) corresponding to rocks with
well defined names (Fig. 2-13).
A
A
D
B
C
RATIO- CUT
A
A
PERCENTAGE CUT
I
I
IEDidote. zoisite
(aiinite '
(Nepheline
\Leucite
I Mordenite, heulandite, analcime, natrolite
Laumontite
Zircon
Thorite
Apatite
Monazite
D
A
A
A
B
C
A
D
A
RATIO-PERCENTAGE CUT
I
-
Figure 2-13 Division of a tetrahedron (or triangle) into lines of equal
ratio, equal percentage, or a combination of both
(from Krumbein, 1954).
Serralog Q 2003
47
Figures 2-14 and 2-15 illustrate the division of tetrahedra into different rock types.
Table 2-17
Sediment composition (from Krynine, 1948).
Mineral
Percentage
Quartz
31.5
Chalcedony
(chert)
9.0
7.5
Feldspars
Micas
chlorite
and
19.0
Clay minerals
7.5
Carbonates
20.0
3.0
3.0
Iron oxides
All others
Figure 2-14 - The fundamental tetrahedron for classifying sedimentav
rocks. (from Krumbein & Sloss, 1963).
Calcite
Pure limestone
95%
Shaly limestone
Figure 2-16 - Composition of detrital rocks as a function of grain size
(from Blatf et al., 1980).
65%
Marl
35%
calcardolomitic
claystone
the current. Coarse fragments settle rapidly under the
gravity action depending on the flow energy. Finest particles are maintained in suspension and carried far away,
settling down only when the energy decreases.
Claystone
Clay
Anhydrite
I
I
CONGLOMERATE
Calcareoanhydritic
dolostone
I
I
I
I
SHALE
Siltstone
Claystone
Dolomite
Rock F,.agments
Figure 2-15 - Example of subdivision of a tetrahedron for carbonate
classification.
Krynine (1948) evaluated the sediments composition
as consisting of the minerals and mineral families listed in
Table 2-17.
The composition of detrital rocks also depends on the
grain size as illustrated by Figures 2-16 and 2-17.
Pebbles and coarser particles will have generally the
composition of the parent rock, consequently will be composed of several minerals, while fine sands, silts and
clays will be composed of monocrystalline minerals.
This is due to the fact that, during weathering and
transport, rock fragments are broken down and abraded
reducing progressively the size of the particles carried by
40
I
SAND
...-.
Detrilsl carbonates
.,.._
---..-..
Feldspar
Figure 2- 17 - Other way to link the composition of particles to their
size.
Serralog 0 2003
Composition
It is also important to remember that every mineral
may have either a definite or a slightly variable chemical
composition (Table 2-18). This is due either to impurities
or substitution (cations having similar coordination numbers and similar ionic radii tend to substitute for each
other) or due to atoms or ions missing. By reference to
Table 2-19, which lists the ion radii of the most common
cations and anions, it is easy to understand what kind of
ionic substitution can occur in some minerals.
Table 2-18
Elemental composition of 3 orthoclases
(from Deer et a/., 1978).
=2
4 0 3
F%03
FeO
64.76
19.960.08
''
64.68
19.72
0.08
63.68
19.54
0.10
tr.
MO
tr.
tr.
CaO
,40
0.84
0.34
0.50
5.54
8.12
0.54
3.42
11.72
0.18
0.80
15.60
ro
KO
-
Minerals are classified as a function of their defining
anions (Table 2-20). Silicates are the most abundant as
they represent about 90% of the rock forming minerals.
They are classified as a function of their structure that is
based on the arrangement of silica tetrahedron: isolated,
ring, single and double chains, sheet and framework (Fig.
2-18 and Table 2-21 next page).
-Element Name 3ymbc
I
Sodium
I
Figure 2-18 - On the left, crystal structure (cubic) of halite: CI in blue,
Na in red. On the right, structure of the silica S O p quartz: silicon in
orange and oxygen in blue are set inside a qua& crystal form.
If very few minerals can be amorphous most of them
crystallize in systems with planes and folds of symmetry.
6 fundamental crystal systems exist (see further Fig. 2-20
next pages).
z
A
-
0
ion
Ion
barns
:harg
-
Radius E
+ I
0.01
C=2UA
Hydrogen
H
1
1.0080
0.33200
Helium
He
2
4.0026
>0.00700
Lithium
Li
3
6.9410
71.OOOOO
Beryllium
Be
4
9.0122
9.20000
+2
0.35
0.88769
5 10.8110
759.00000
+3
0.23
0.92498
0.99907
B
1.98413
0.99935
+ I
0.68
0.86443
Carbon
C
6 12.0112
0.00340
+4
0.16
Nitrogen
N
7 14.0067
0.00000
.5,-3
0.13.1.7
0.99952
Oxygen
0
8 16.0000
0.00027
-2
1.32
1.ooooo
Fluorine
F
9
18.9984
0.00980
-1
1.33
0.94745
Sodium
Na
I1 22.9898
0.53000
+I
0.97
0.95695
Magnesium
w
12 24.3120
0.06300
+2
0.66
0.98717
Aluminium
41
13 26.981 5
0.23000
+3
0.51
0.96362
Silicon
Si
14 28.0860
0.16000
+4
0.42
0.99694
Phosphorous
P
15 30.9738
0.01900
+5
0.35
0.96856
Sulphur
S
16 32.0640
0.52000
.2,+6
84,0.3
0.99800
Chlorine
CI
1;
33.2000C
-1
1.81
0.95902
35.453(
Potassium
K
I!
39.098:
2.10000
+ I
1.33
0.97191
Calcium
Ca
2(
40.080C
0.43000
+2
0.99
0.99800
Titanium
Ti
2;
47.900C
6.41000
+4
0.68
Vanadium
V
2
50.941;
5.06000 +4,+2
Chromium
Cr
2d
51.996(
3.10000
Manganese
Mn
2
54.938(
13.3000( *4,+i
Iron
Fe
21
55.847(
2.56000
Cobalt
co
2
58.933:
Ni
+3
1.63,0.81
0.91856
0.90299
0.63
0.92315
0.6,O.B
0.91012
+2,+? 1.74,O.P
0.93112
37.50000 +2,+3 1.72.0.6:
0.91629
2f
58.710C
4.54000
+2
Copper
U
:
25
63.5400
3.81000
+2,+'
Zinc
Zn
3(
65.370C
1.10000
Bromine
Br
35 79.9090
6.8000
-1
Nickel
Chloring
49
Table 2-19
Ion charge and ion radius of elements.
Boron
Orthoclase 2 Orthoclase 3
Orthoclase 1
I Chapter 2 1
+2
0.69
0.95384
3.72.098
0.91281
0.74
0.91785
1.96
0.87599
Strontium
Sr
38 87.620C
1.21000
+2
1.12
Zirconium
Zr
$0
91.220(
0.18200
+1
0.79
0.87700
Niobium
Nb
11
92.906C
0.15000
+5,+'
l.69,0.74
0.88261
Molybdenum
Mo
12
95.940C
2.6500
+6.+,
).62.0.7(
0.87555
Sylver
17 107.87Ol
63.8000
+2,+
1.89, 1.2
0.87142
Cadmium
Ag
Cd
18 112.4001 2450.000
+2
0.97
0.85409
Tin
Sn
50 118.6901
0.6300C
+4
0.71
0.84253
Barium
Ba
56 137.340
1.20000
+2
1.34
0.81549
Lanthanum
La
i7 138.9101
8.900
+3
1.02
0.82068
Cerium.
:e
58 140.120(
0.730
+3
1.03
Samarium
im
62 150.350( 5820.000
+3
0.96
0.86738
0.82786
0.82474
Europium
ZU
6:
+3,+2
.95, 1.0!
0.82917
Gadolinium
>d
61 157,2501 19000.000
+3
0.94
0.81399
Tantalum
Ta
7:
180.948(
22.000
+5
0.68
0.80686
Tungsten
W
74 183.850(
18.500
+6
0.62
0.80500
Lead
Pb
32 207.200
0.18800
+2
1.2
0.79151
Thorium
Th
9
232.038
7.40000
Uranium
U
92 238.050f
2.72000
151.9601 4100.000
+4
1.02
+4,+€ .94,0.80
0.77573
0.77294
Serralog 0 2003
49
50
Well logging and Geology
Determination of the rock composition
Traditional method
Table 2-20
Chemical classes of minerals.
aass
Defining anions
Exanples
Native elements
None : no charge ion
Cu,Au, Ag, diamond
Sulfides
Sz8 sirilar anions
write: FeS,
Oxides
0 2
Hemtite: F%O,
Hydroxides
(ow-
Brucite: Mg(OH),
Halies
Carbonates
sulfates
Ct, F, Br, t
Halite: Nacl
coy2
SO;2 8 sinilar anions
Calcite: C
a
m
,
Anhydrite: CaSO,
Phosphates
FO,J 8 sinilar anions
Apatite: Ca5F(F0,),
Silicates
SQ-2
Pyroxene: MgSii
Table 2-21
Structure of silicates.
As previously indicated, the rock composition can be
expressed in two different ways: either by elemental
analysis, or by mineralogical determination.
Elemental and Chemical Composition
This is provided by a chemical analysis of rock samples in the laboratory or by using X-ray diffraction (Fig. 219), alpha spectrometry for thorium (Th) and uranium (U),
or thermal neutron activation, the latter being the most
precise and accurate. The results can be expressed either
in terms of percentages of elements present or, more
often, in terms of oxides of these elements, (Tables 2-8 to
2-12), the oxygen being often strongly bonded to each of
the abundant elements.
N.B. The analysis must be repeated on several samples in order to
have a more precise evaluation of the element percentage of the rock
samples through statistical study.
.
-
Figure 2-19 SEM photograph of an anorthite coated by smectite and
below the spectrographic analysis by EDAX of this mineral.
50
Serralog 0 2003
Next Page
Composition
Isometk :3 equi-kngth
axes, all at right angles
c
u
r
Galena
Garnet
Diamond
Tetragonal:2 equal axes
and a thinl either longer
or shodec all at right
angles
Tetmgonal
Cassiterite
Wulferite
Barite
Olivine
Epidote
1
51
Magnetite
Pyrite
Halite
Fluorite
Zircon
Sulfur
Mamite
Epsomite
Celestite
Anhydrite
Monoclinc :3 unequal
axes, 2 at right angles,
a thinl perpendicular
to one but oblique to
others
Monoclinic
Chapter 2
Leucite
Rutile
Chalcopyrite
Ofthodmmbic: 3 unequal
axes, all at right angles
Olthorhombic
I
*psum
TMinic :3 unequal
axes, all at oblique
angles
Sphene
Hornblende
orthoclase
Anorthoclase
Oligoclase
Polyhalite
Triclinic-
Axinite
Rhodonite
Hexagonal: 3 equal
axes in same plane
intersecting at So',
a kwfth perpendicular
to other three
Calcite
Dolomite
Hematite
Corundum
Hegagonal & rhombohedral
Apate
QdlFiZ
-
Figure 2-20 The different basic crystal systems.
Serrabg Q 2003
51
Previous Page
Well logging and Geology
52
YAGYA
TVPE
MIOH
TEMP
FIRST TO
CRYSTALLIZE
ROCK
TVPE
DUNITE
PERIDOTITE
PVROXENITE
GABBRO
Basan
DlORlTE
cIn
Andwsitw
x
W A R T 2 DlORlTE
D.ClIW
ORANODlORlTE
GRANITE
WsEoYilE
*
LOW
TEMP
LAST TO
CRYSTALLIZE
*
Rhyolllw
0
I
t
WAATZ
@
LOW
YOST
SLOW
WEATt4ERING
STABLE
1: single tetrahedron. 2: single chain. 3: double chain. 4: sheet. 5: 3-D lattice.
Figure 2-21 - The Bowen's reaction series and its interpretation both for rock type formation and for weathering intensity and results
(adapted from Leet et al., 7982).
Mineralogical Composition
In spite of its usefulness - we shall see later the application of tools such as the natural or induced gamma ray
spectroscopy tools for this - the elemental composition of
a rock is not the best expression of rock composition,
either from a sedimentological or from a log analysis
standpoint. In fact, a rock is a mixture of minerals. These
are what gives its petrophysical characteristics, which are
generally those measured by the logging tools - density,
resistivity, sonic travel time, compressibility, etc. These
properties therefore depend on:
- the individual characteristics of each of the constituent minerals forming the rock,
-the relative percentages of each mineral,
- their distribution and bonding.
For this reason, it is preferable to express the composition of a rock in mineralogical terms.
One way is to convert oxide composition into mineral
composition. This is based, for igneous rocks, on the
ClPW (for Cross, Iddings, Pirsson and Washington) norm
or the Niggli's norm. The explanation of the rules for computation of the normative minerals is not the goal of this
booklet. One can say that the norms are based on the
Bowen's reaction series (Fig. 2-21) which admits that the
first to crystallize is olivine, or anorthite, then pyroxenes
and amphiboles or plagioclases poorer in Ca and so on.
The other way to determine the rock type and its min-
52
eral composition is by using different mineral properties color, density (Table 2-22), hardness (Fig. 2-22a), magnetism, radioactivity.. . -, techniques - staining (Fig. 222b), examination of thin sections under microscope
using cross nicols and optical properties of minerals (Figs.
2-23a to 2-23d), scanning electron microscope or SEM
(Figs. 2-23e to 2-23h) and X-ray or electron diffraction...and products - chloridric acid, alizarine, copper nitrate,
benzidine, safranine"y"... -.
Injection of colored fluid (epoxy), which polymerizes in
the pore space with time or temperature, allows the visualization of the pore system (Fig. 2-23c).
I
Figure 2-221, - Calcite crystals are stained in red by alizarine.
Sermlog 0 2003
Composition
I Chapter2 I
Table 2-22
Specific gravity (density) of the principal minerals.
Specific
I
Mineral
gravity
(9/cm3)
Coal
1.O-1.8
Lignite
1.15-1.3
Bituminous coal 1.14-1.40
1.32-1.7
Anthracite
1.47
Natron
1.60
Bischofite
1.592-1.602
Carnallie
1.66
Tachhydrite
1.68
Epsomite
2.07
Sulfur
2.1-2.2
Syhrite
1.9-2.3
Opal
Montmorillonite 2.0-2.1
2.16
Halite
2.13
Trona
2.15-2.24
Kainite
2.2-2.4
Glauconite
2.31-2.33
Gypsum
2.45-2.50
Leucite
2.4-2.63
Kaolinite
25 2 . 6
Kieserite
2.57-2.60
Anorthodase
2.55-2.58
Microcline
2.56-2.58
Orthoclase
2.55-2.65
Nepheline
2.59-2.64
Chalcedony
2.6-2.9
lllie
2.62-2.64
Albite
2.65
Quartz
2.65-2.90
Chlorite
2.71
Calcite
2.74-2.76
Anomie
2.76-3.1
Muscovite
I
Mineral
I Polyhalie
Dolbmite
Langbeinite
Pnkerite
Biotite
Anhydrite
Tourmaline
Apatite
Andalusite
Fluorite
Hornblende
Augite
Enstatite
Silimanite
Olivine
Charnosite
Goethiie
Garnet
Limonite
Pynhotite
Siderite
Sphalerite
Celestite
Chalcopyrite
Rutile
Barite
Zircon
Marcasite
pyrite
Magnetite
Hematite
Galena
Uraninite
Specific
gravy
Wcm 1
2.77-2.78
2.8-2.9
2.83
2.95-3.1
2.8-3.1
2.899-2.985
2.98-3.25
3.15-3.23
3.16-3.20
3.01-3.25
3.05-3.47
3.2-3.4
3.1-3.3
3.23-3.24
3.27-3.37
3.3-3.4
3.3-4.37
3.154.3
3.64.0
3.54.5
3.83-3.88
3.9-4.1
3.95-3.97
4.14.3
4.18-4.25
4.34.6
4.684.70
4.854.90
4.95-5.10
5.168-5.1 8
4.9-5.3
7.4-7.6
9 .O-9.7
-
Figure 2-23a Polish section of a granite.
Feldspar
-
Figure 2-22a Mohs scale of hardness.
Quartz
Figure 2-231, - Thin section of a granite.
Serralog 8 2003
53
Well logging and Geology
54
-
Figure 2-23f Another SEM photograph of "books" of kaolinite.
-
Figure 2-23c Thin section of a dolostone. Porosity is in blue.
m
-
Figure 2-23d Thin section of a compact sandstone. Quartz grains are
interlocked due to pressure solution.
-
Figure 2-239 SEM photograph of fibrous like crystals of illite.
I
Figure 2-23h - SEM photograph of chlorite crystals. On the right bottom comer one can see a quartz crystal overgrowth.
Determination of the rock composition
Well logging approach
-
Figure 2-23e SEM photograph of a "book" of kaolinite.
54
There are many different well logging measurements
which respond to the rock composition. Table 2-23 indicates the relationship that exists between the geological
attributes of rock composition and the different logging
parameters. As a result, an adequate logging suite will
Semlog 0 2003
Composition
I Chapter2 I 55
Table 2-23
Link between geological attributes of rock composition and well logging parameters.
Geological Parameters
Well Logging
Parameters
affected
i
Relevant
Logging Tools
Nalum
make the determination of rock composition much easier.
As rocks are made of beds composed of minerals
themselves composed of molecules and atoms (cf. Fig. 28), log responses are affected by those geological attributes. Two approaches are also possible in using well logs
for determining the rock composition.
Determination of elemental composition
Several logging techniques provide data related to the
elemental composition of the formations crossed by a
well.
Natural aamma rav sDectrometry
Natural gamma ray spectrometry (Fig. 2-24) allows the
measurement of the potassium (K), thorium (Th) and uranium (U) percentage.
Potassium is the seventh most abundant element of
the Earth's crust.
Potassium, thorium and uranium are relatively abundant in acidic igneous rocks, like in syenite, granite, granodiorite, diorite, but are rare in basic and ultra basic rocks
(Table 2-24).
Serralog 8 2003
J
-
Figure 2-24 Natural gamma ray spectrometry allowing the determination of K, Th and U percentage.
Using cross-plot Th vs U (Fig. 2-25) allows the determination of the igneous rock type. This determination
must be confirmed by using other log data (see further).
Potassium is associated to alumino-silicates such as
feldspars, micas, feldspathoids (leucite, nepheline), clay
minerals. Consequently, it is abundant in arkoses, subarkoses, graywackes, claystones. It exists as well in some
55
evaporites such as polyhalite, kainite, langbeinite, carnallite, sylvite (Table 2-25).
Table 2-24
Potassium, thorium and uranium percentage in igneous
rocks
(from U.S.G.S. in Adams & Gasparini, 1970).
I
I
Igneous rocks
Potassium
(%)
Granite G1
average
Granite 0 2
average
Precambrian
granite
Granite'
Svenite
4.28.14
4.55
3.25-3.90
3.74
24
2.74.7
1.95-2.77
2.21
24
-
. Rhyolite'
I
Trachyte'
Granodiorite
average
Diabase
(dolerite)
-..
57
I
I
I
I
I
1I
6.5-80
52
215-40
25.2
14-27
I
I
I
I
Uranium
@pm)
I
11-26
610-1800
1305
6 15
- --
a75
I
2.174.71
2.39
0.49-0.59
0.53
zE
3
I
Thorium
(ppm)
I
Kus U (ignww rocks)
99-125
110.6
1.3-4.6
2.4
I
I
I
1.34.42
4
1.32-3.4
1.99
3.24.6
3.6-6.9
1900-3000
2521
2.5-5
2-7
0.19-2.68
1.98
0.28-1
0.5
0
-
Figure 2-24a Th vs U (bottom lee) and K vs U cross-plot for the
determination of the igneous rock type
(based on data from U.S.G.S. in Adams & Gasparini, 1970).
T h w U (Spnitm)
P
3
Basic rocks
Andesite
average
I
0.46-0.58
0.5-0.6
1.16-2.68
1.39
0.61.7
2.16-3.55
2.39
Peridotite
Dunite
I
i
0-0.07(0.019)
0.02
Gabbro
Gabbro'
Basalt
average
Basalt.
I
3.85-27
2.7-3.9
5.8-10
6.81
2-7
5-10
6.96
I
I
0.018
0.0092
0.84-0.9
0.8-0.9
1.2-2.17
1.73
1-1.7
1.4-2.6
1.94
I
I
Uranium (ppm)
0.0048
0.0037
zE
a from Desbrandes(1982).
I
0
Uranium (ppm)
n
E
LL
Q.
u
-
F@um 2-24b Th vs U and K vs U cross-plot for the determination of
syenite (based on data from U.S.G.S. in A d a m & Gasparini, 1970).
Uranlum (ppm)
56
Thorium is fundamentally an indicator of the detrital
origin of the sediments as its solubility product is very low.
It is associated to some frequent stable minerals (zircon,
monazite) and to some clay minerals that adsorb thorium
ions on their platelets.
Serralog Q 2003
Composition
I Chapter2 I
57
In sedimentary rocks, uranium is essentially associated to uraninite which exists only in reducing environments. Consequently, it is often associated to source
rocks rich in organic matter (Fig. 2-25). However, it exists
also in heavy minerals and in phosphates where U
replaces Ca.
generate gamma rays of which the energy spectrum is
characteristic of the target nucleus (Fig. 2-27).
Figure 2-25 - Uranium and uranium potassium ratio vs percentage of
organic cabon (fmm Beers & Goodman, 1944 on left; from Supernaw
et al., 1978 on right).
I
Adams & Weaver (1958) proposed to use the thoriumuranium ratio as an indicator of the type of sedimentary
rocks (Fig. 2-26).
Table 2-25
Potassium bearing minerals.
Mineral name
FeMsPars
Microcline
orthoclase
Anorthoclase
Plagiodase
McaS
Muscovite
Bitite
Phlogopite
lllite
Glauconite
Fcldrpalhdds
Leucite
Nepheline
Kalsilite
Kaliophilie
olhcr clay minerals
Montrnorilonite
Chlorite
Kaoliite
s-
Polyhalite
Camalie
Kainite
Langbeinite
Glaserite
SyMte
I K content (weight %)
I
16 (ideal) to 10.9
14 (ideal) to 11.8
2.94 to 4.39
-
Figure 2-26 Distinction of the sedimentary rock type through the Th
content and the Th/U ratio (from Adams & Weaver, 1958).
Several logging tools allow the detection and evaluation of several fundamental elements.
Tools based on inelastic collisions of fast and high
energy neutrons (Fig. 2-28) provide information about the
carbon (C) and oxygen (0),and also the yield of some
other elements silicon (Si), iron (Fe), calcium (Ca), sulfur
(S), and magnesium (Mg) (Fig. 2-29).
O U
Inelastic collisions
offast neutrons
9.8 (ideal) 6.2 to 9.37
5.42 to 7.61 (average: 6.6)
5.38 to 9.9 (average: 8.5)
2.7 to 6.65 (average: 5)
T h e m neuron capture
17.9 (ideal)
3 to 10.14
21.29 to 23.51
24.7 (ideal)
0 to 1.32
0 to 0.35 (average: 0.1)
0 to 0.73 (average 0.35)
a,
.c.
E
p
=I
8
13.37
14.07
15.7
18.84
24.7
52.44
Induced aamma rav spectrometry
Figure 2-27 - Elements that can be determined from neutron interactions with nuclei and from natural gamma ray spectrometv.
Several types of interactions of neutrons with nuclei
can exist. They depend on the neutron energy. All of them
Serralog 0 2003
57
58
Well logging and Geology
Table 2-26
Different types of interactions of neutrons with elements with indication of the resulting reaction-product and energy of
the produced gamma rays (from Serra, 1993).
we
Inelastic collision
(,, nl)
e
'S0
Fast reaction
(n, a)
'to
Inelasticcollision
>
ip
w (r: 4,441
* ':c*
(n, n')
w (r: 3.68)
W
1:0*
(r: 6,131
I$
15h
*
15h
:pg
Thermal capture
(n,a)*
p-
:iNe
11%
Delayed thermal activation
7%
>
(n, 1 =$ yiMg'*
*
(n,)
Delayed thermal adivation
=3
38'
p-
3
11%
@,5
* @J*
*
23
100%
3
9.5'
=b
1%
3
23
Thermal capture
>
(n.
j
w (v:
:si
4,93,3,54,2,03
,...)
I ....)
Delayed thermal activation
3
a
*
(n.
3
gh%
b
p-
* $ 3 ~ * p- * ETi
8s
58
\
57,s
(r:
5.10,3.10.2.00)
(y 1,78)
(y 4,11,3.10,
...)
serrabg 82003
Composition
1 ChirpterZ I
59
ments composing the Earth's crust are detected and evaluated. Comparisons with core measurements show the
accuracy of this logging approach (Figs. 2-34to 2-37)and
its validity.
Therefore, it proves their interest for the direct determination of the content of certain elements, and, above
all, for a better analysis of composition in conjunction with
the other measurements, in particular hydrogen index and
photoelectric cross-section.
Figure 2-28 - Inelastic collision of fast and high energy neutron with
nucleus (courtesy of Schlumberger).
Figure 2-30 - Schematic of fast-neutron particle reaction
(courtesy of Schlumberger).
-
Figure 2-29 Inelastic spectra of elemental standards for the RST-A
tool (courtesy of Schlumberger).
Several other tools, based on fast-neutron particle
reaction (Fig. 2-30)or on thermal neutron capture and
activation (Fig. 2-31)allow the evaluation of the percentage of fundamental elements such as Si, Fe, Ca, s, Ti,
Gd, CI and H, and in some case Al and Mg (Fig. 2-32).
Elemental yields are processed from standard spectra for
the different tools, obtained using laboratory measurements. Table 26 lists certain types of reactions between
neutrons and elements and the nature of the result and
the energy of gamma rays emitted during these interactions.
Tools based on elastic collisions of neutrons (Fig. 233) and spectrometry of the induced gamma ray due to
thermal neutron absorption allow mesurement of Si, Fe,
Ca, S, Ti, Gd, CI and H (Fig. 2-32).
One can realize that, thanks to the spectrometry
measurement techniques, all the most abundant ele-
Serrabg 63 2003
-
Figure 2-37 Schematic of thermal neutron capture (first two steps)
and activation (second step on) (courtesy of Schlumberger).
59
60
Well logging and Geology
Neutron-based tools provide a breakdown either in
terms of light elements (H, C, 0 and Si) which have a
slowing effect (seen by the epithermal tools), or in terms
of thermal neutron absorbers (Gd, B, Li, CI, Fe) seen by
the thermal tools.
The litho-density tool provides a measure of the mean
atomic number of the rock components.
Determination of the mineral composition
This determination can be obtained from the elemental composition, assuming certain associations and rules,
or from other logging measurements.
-
Figure 2-35 Example of element percentage evaluation from the RST
tool and comparison with core measurements
(courtesy of Schlumbeger).
Figure 2-32 - Capture spectra of elemental standards for the RST-A
tool (courtesy of Schlumbeger).
1
Figure 2-33 - Schematic of the neutron scattering with a taget nucleus
in elastic collision.
-
Figure 2-36 Example of element percentage evaluation from the GLT
tool and comparison with core measurements. One can observe a
very good agreement between the measurements and the core data
(courtesy of Schlumbeger).
1
Figure 2-34 - Example of element percentage evaluation from the ECS
tool and comparison with core measurements
(courtesy of Schlumbeger).
80
Serralog 0 2003
Composition
I Chapter 2 I 61
Remark
It is highly recommended to never base the mineral
identification on a single cross-plot. Use all the reliable
information to avoid misinterpretation.
Silicon (wt %)
Figure 2-38 - Mineral identification combining Si and K.
-
Figure 2-37 Another example of elemental evaluation obtained with
the GLT tool. Observe the very good agreement between logging and
core measurements in this scientific well.
No point in this zone
EXDlOitatiOn of the elemental ComDoSition
For instance Si, Ca, Fe and S are generally only present in dry rocks. K may be in mud filtrate, if KCI has been
put in the mud, otherwise it is associated to feldspars,
micas and clays. CI is assumed to be present in water and
mud filtrate if evaporites minerals are assumed to be
absent. The sum of the weight percentage of all elements
in the dry rocks is equal to 100%. The model of conversion of element into minerals uses oxides and hydroxyl
content of clay minerals to predict the weight of unmeasured elements (0 and H).
Silicon (wt %)
-
Figure 2-39 Mineral identification combining Si and Ca.
100 = SiOg + CaC03 + CaSO4 + a OH(A1203 + FeqOg) + K20
Cross-plotsof elements can help selecting the mineral
model (Fig. 2-38 to 2-43). 2-plot technique (a third element on the 2 axis) can improve this selection.
Serralag @ 2003
Remark
On these plots several minerals are indicated by an
ellipse. These ellipses reflect the variations in elemental
composition of those minerals.
61
62
Well logging and Geology
No point in
this zone
Silicon (wt %)
Aluminum (wt %)
Figure 2-40a - Mineral identification combining Si and Fe.
-
Figure 2-42 Mineral identification combining A1 and K.
E
2
Y
8
K (Wgt Yo)
Figure 2-40b - Mineral identification combining Th and K.
A1 (wgt %)
-
Figure 2-43 Mineral identification combining Ca and Al.
s
No point in
this zone
r
Y
5
‘5
C
a
Silicon (wt %)
Remark
On these cross-plots only the principal minerals are
represented through their elemental composition. The
mineral composition of a rock can be evaluated by taking
into account the minerals, surrounding the point representing the elemental composition of the rock, common
on all cross-plots.
Another example of exploitation of such plots for rock
type determination is illustrated by Figures 2-44a and 244b.
Figure 2-41 - Mineral identification combining Si and A/.
62
Serralog (B 2003
Next Page
Composition
I
II ChaDter 2 II
63
/
/ -
Figure 2-44a - Igneous rock type identification through association of
Si and K. Each rock is represented by a point. In fact these points correspond to the mean values. Around them must be drawn an ellipse
with elongation following the amws.
I
x
16 18
14
12
10
-1
-
Figure 2-45 - In the leff track is reproduced the determinationof the
mineral composition of the formations crossed by this scientific well.
This determination is based on the elements reproduced on the other
tracks.
8 6 4 -
2 0-
sro,
.--.
I
Figum 2-44b - Determination of volcanic rocks using silica associated
to other oxides (from Dalx 1933).
Exploitation of the elemental composition in sedimentary rocks is illustrated by Figure 2-45. A model with 12
minerals has been used for quantitative computation by
the ELAN program of Schlumberger.
Another example 'of mineral percentage evaluation
from elemental analysis is provided by Figure 2-46.
Serralog Q 2003
In Figure 2-47, a quantitative interpretationof standard
logs and elemental analysis provided by natural and
induced gamma ray (spectrometry measurement) allows
the determinationof the mineral nature and percentage of
grains and crystals, the evaluation of the matrix material
(clay minerals) and the determination of the cement
nature in this detrital siliciclastic formation.
In Figure 2-49 are reproduced firstly the results of elemental analysis, secondly the results of quantitative interpretation mixing standard log data and elemental analysis. The mineral percentage evaluation, based on a 11mineral model, is displayed with the porosity computation.
All these examples show the interest of this elemental
analysis to better determine the lithology and the mineralogy of the formations crossed by a well.
63
Previous Page
Well logging and Geology
64
Other logging measurements can be used for mineral
or rock type identification.
Measurements involvina interactions of photons
gamma with electrons
The first type of such interaction corresponds to measurements linked to photons gamma colliding electrons. As
one knows, depending on the photon energy the photons
gamma either are deflected by their encounters and scatter with a reduced energy (Fig. 2-48), or transfer all their
energy to the electrons in the form of kinetic energy, the
electrons being ejected from their atoms (Fig. 2-50).
These interactions correspond to two types of measurements which are respectively: the electron density measurement, by Compton scattering, and the photoelectric
index measurement by absorption of photons gamma.
_I
Figure 2-46 - Mineral percentage determination from elemental composition (courtesy of Schlumberger).
The electron density is linked to the bulk density by the
following equation:
with :
pe = electron density
pb = bulk density (g/cm3)
Z = atomic number
A = atomic mass
N = Avogadro's number (6.02~1023).
-
J
-
Figure 2-48 Compton scattering of a photon gamma (or gamma ray)
with an electon. The gamma ray is deflected with a reduced energy
and the electron recoils
(courtesy of Schlumberger).
Figure 2-47 Example of a quantitative interpretation including elemental analysis coming from spectrometty measurement of natural and
induced gamma ray (courtesy of Schlumberger).
64
Serralog Q 2003
Composition I Chapter 2
65
.-
Figure 2-49 - Another example of formation evaluation integrating data provided by spectrometry of natural and induced gamma ray. The elemental
analysis results are on the left with comparison with core data. Mineral percentage and porosity are reproduced on the right
(courtesy of Schlumberger).
The attenuation of an incident gamma ray flux, Qo, by
photoelectric absorption alone can be written as:
where E is the gamma ray energy in keV and Z is the
atomic number.
where na is the number of atomes per cm3, x the attenuation depth and z the cross section for photoelectric
absorption, expressed in barndatom, z is given by the following relationship:
Serralog 0 2003
By analogy with Compton scattering, the counts N
seen by a detector at a distance L from a source radiating
No gamma ray are:
N = N0 e - n a T L
65
66
Well logging and Geology
their relative percentages in the zone of investigation on
the one hand, but also on the nature and percentages of
the fluids occupying the pore spaces on the other hand.
So, for a mixture of i atoms one can write:
where:
ne is the electronic density number
nei the electronic density number of the atom i
Vi the volumetric fraction of atom i
Pei the photoelectric index of atom i.
In general the volumetric photoelectric absorption
index is used. It is given by the following equation:
U = Pe pe
Figure 2-50 - Absorption of a photon gamma (or gamma fay) by an
electron that is ejected from its atom (courtesy of Schlumberger).
The photoelectric absorption index Pe is proportional to
the "average cross-section by electron", z /Z
The density measurement can be used for rock identification. Table 2-27 lists the mean densities for igneous
rocks. Refer as well to Tables 2-28and 2-29for the bulk
density of the most abundant minerals composing rocks.
Table 2-27
Mean densities of igneous rocks
(from Daly, 1933).
Pe- l Z
-K Z
K is a constant coefficient characteristic of the energy E:
Rock
Plutonic rocks
155
syenite
24
21
13
Diorite
11
Gabbro
27
40
3
15
8
12
Peridotite (fresh)
Dunite
pyroxenite
z and Ki are respectively the cross-section and the coefficient at energy Ei.
Finally, as the energy dependences upon z and K cancel out, Pe does not depend on the energy. It is well
approximated by :
Pe is expressed in barns/electron.
As we have seen, rocks are made up of a mixture of
minerals. Consequently, it is the physical properties of
these minerals, and ultimately the atomic properties of
their constituent elements which influence the log
responses. In fact, the responses are a function of the
characteristics of each mineral present in the rock and of
06
11
Norite
Diabase ( k h )
z=Zzi andK=ZKi
I
Numberof
sample3
Granite
Granodiorite
Quertz diorite
Practically the absorption is made over a range of
energies and :
I
Anorthosite
Meall
density
Range of
density
2.667
2.716
2.757
2.806
2.839
2.984
2.976
2.%5
3.234
3.277
3.211
2.734
2.5 16-2.809
2.668-2.785
2.630-2.899
2.680-2.960
2.721-2.W
2.720-3.020
2.850-3.120
2.804-3.110
3.152-3.276
3.204-3.314
3.100-3.318
2.640-2.920
2.370
2.450
2.474
2.550
2.772
2.330-2.413
2.435-2.467
2.400-2.573
2.520-2.580
2.704-2.851
Natural glasses
Rhyolite obsidian
Trachyte obsidian
Andesite glass
Leuoite tephrite glass
Basalt glass
15
3
3
2
-
The introduction of tools able to measure the slowness
of compressional and shear waves allows the determination of the rock elastic properties which can be exploited
to determine the rock type. For instance, a cross-plot, proposed by Pickett (1963),associating compressional and
shear slownesses (Fig. 2-51)allows the recognition of the
main sedimentary rock types.
serralog Q 2003
Composition
I Chapter 2 I
67
160
-
Figure 2-51 Cross-plot of compressional and shear slownesses for
identification of the lithology (from Pickett, 1963).
The velocity (or transit time) alone can be used for rock
identification as illustrated by Figure 2-52.Combined with
other log data, it allows the recognition of evaporite minerals (cf. Figs. 2-61 to 2-63).For instance, combination of
density and sonic travel time allows the determination of
many non-porous rock types such as igneous rocks (Fig.
2-53)and evaporites (cf. Fig. 2-61).Don't forget that non
porous rocks, igneous, metamorphic and evaporites, are
also characterized by a very high resistivity generally without change of value between micro- and macro-devices.
The logging parameters of the most abundant minerals are listed in Table 2-28.
At
Figure 2-53 - Cross-plot of density vs sonic travel time for igneous
rocks.
that compose them is better achieved from cross-plot
analysis.
Cross-plot interpretation
The lithology can be determined by a "manual" or
"visual" interpretation, which is fundamentally based on
cross-plot analysis hereafter explained.
Cross-plot definition
A cross-plot is a graphical representation of one
parameter, plotted on the X axis, versus another, plotted
on the Y axis, from a sample at a given depth. This represents a 2 dimensional space. Cross-plots are convenient
to evaluate any existing relationship between two parameters. Those parameters must correspond to the same
volume at the same depth.
Figure 2-52 - Compressional sound velocity of the major igneous rocks
(from Christensen, 7965).
The logging parameters of some accessory minerals
are listed in Table 2-29and the logging parameters of the
micas and clay minerals in Table 2-30.
Logging parameters of major igneous and metamorphic rocks are listed in Tables 2-31 and 2-32.
A third dimension can be added by plotting on an axis,
perpendicular to the plane made by the two first, a third
parameter. In that case the plot is named a Z-plot. One
can plot on the 2 axis either the frequency or the average
value of a third parameter. A fourth dimension can possibly be added, if a color graphic plotter or screen is available, by superposing a fourth parameter using a color
scale which is a function of this parameter value.
The determination of the rock type and of the minerals
Serralog 0 2003
67
Table 2-28
Logging parameters of the most abundant minerals.
08
Serralog Q 2003
Chapter 2
Composition
69
Table 2-29
Logging parameters of accessory minerals, coals and fluids.
a May contain up to 2 500 ppm of thorium
Table 2-30
Logging parameters of the principal phyllrte minerals.
Composition
Name
Group
Kaolinitc
Dickite
Hdloysite
Kao lioitc
Muacovdc
AI,[S,O,J(OH),
~
,
(A)
F
PA)-
7.2
i
0-06
AI,[Si,O,d(OH),4H2O
10 1
I
iSib5~~~0,,,1~0~~9'F,c2
KCfi,[,S&@JO,,(OH),
,
70
with 1.0 < x.< 1 5
-(K,Na,Ca),JJ(Fe
FC.M&~
-*
i
A1
Montmorillonite
I
Chlorltc
'
(MsFc,Al),
ThkinBtc
Clmochlorc
-Chamositc
I(~,A!)&.&OH)La
Rich U-I iron
Rich m magnesium
FeleA<d(MeFc,&4,
__
~
-4.30
.
(WC)
U
(trlrm3)
16-1.8
15-1.9
-1.5
4-5
4-5
3-3.5
13-16
13 5-16
12-12.7
GR
(API)
80-130
80-130
-100
' 15-25
20-37
2 6-35
7-12
15-23
130-235
-
-2 85
20-35
24-40
5 8-7.3
IOz18
18-26
155-210
2.0-2.2
50+
60+
1.6-2.3
3 5-5
11-15
- 140-200
_
-37
-5 2
-6.3
-17
-25
8-11
-2G35
--27-40
12-25
16-40
3 5 3 3 ' 2.6-2 8-6.0
L
(p.~)
40-46
42-50+
33-38
(g/crn3)
-2.60
-2.60
-2.08
Pe
-_
*s&L
(p.~)
-34
-35
(CU)
I
( 1 0 CaNa)o,(AI,M_&Fc>+
I(%W800,J(OH), nH,O
hton1tc
Bcidellilc
Nontronite
Chlorite
*me
- Pb
~
1- -
[B,.,,AI,,,4O,oI(OH)4
Smcctite
K
~ ~ 4 1 ~ . 0 i o l ~ 7 ~2 ~ ;~ -0.20
~
IllllC
Glaucon-itc
Sleet
1' 0.5-1.5
9.717 2
I
(2&l&LdOH)s
0-03)
'
-276
'
-3.1
-2 8
-3.19
*
45-50+
47-56+
'
so+
50-250
L
,
. *
.
14.
:
_-
- 4 -
-0 13
-
-
-
5Oi
44-50+
so+
18-c8-_ 5-13.
A -
25-32
8-10
-
2
For more detail refer to document ELEMENT, MINERAL, ROCK CATALOG (SERRA. 1990)
type, for instance carbonate formations must be separated from siliciclastic detrital formations, formations of difSerrabg 0 2003
ferent ages as well. This helps the analyst to avoid any
confusion,
09
70
Well logging and Geology
Table 2-31
Log responses of the principal igneous rocks ( completed from Desbrandes, 1982).
ROCb
PMonlc
Syenite
Granite
Diorite
Gabbro
Peridotite
Dunite
Vokanic
Rhyolite
Trachyte
Andesitc
Basalt
Th
PI
Pe
4%
c
@pm)
(g/Cm?
(we)
@.uJ
(C.U.)
3.59
2.86
3.93
4.47
3.66
3.32
6
3
8
9
9
17
I300
14-80
8.5
2.7-3.9
0.05
0.01
2500
1.3-6.9
2
0.8-0.9
0.01
0.01
2.7-4.5
2.7-6.14
1.1
0.46-0.6
6-15
9-25
1.9-2
2-7
2.5-5
2-7
0.8-2
1-1.7
2-4
5.7
1.7-2.9
0.6-1.7
2.6-2.8
25000
107-234 2.56-2.68
2.8-2.9
68
2.9-3.1
25-32
2.3-3.3
3.5
3.0-3.7
0.5
0.2
0.02
2.3-2.5
2.4-2.5
2.4-2.6
2.7-2.9
76-164
143-247
4 1-70
26-69
Vpl
vp4
(1o3dS)
(~dm/~)
6.7
6.1-6.5
6.6
6.7-7.1
3.9-8.1
7.8-8.3
6.6
5.8-6.4
6.4
6.6-7.0
3.6-7.2
7.5-8.2
12
21
3.2-7.0
5.36
12
5.4-6.2
5.2-6.0
Table 2-32
Log responses of the principal metamorphic rocks (from Desbrandes, 1982).
Pr
(g/cm3)
____-_--
Igneous
-
_-7-21
-
..
3-4
5-8
1
5-10
I_____.-___._._-.
I-I.
-
_
I
_
_
_
-
-____..I._
13-14
5.9-6.0
-..--I_-_--
Calcitic marble
Dolomitic marble
__-.
-
0.9
0.5
~
2.62-2.70
2.62-2.70
40-60
196-252
-
---
I
_
I
_
_
1.8
2.3
-
____-_
2.7-2.8
5-5.2
ll_l_-_-_-__-___
_.__
6.2-6.4
_
I
_
-2
-2
4.2
5.5
.I___
_ l
6.0-6.1
6.0-6.1
i ^ I ___
_
_
_
I
_
--_I
_
_I
_
_
I
I
-
Carbonates
.__l__.._.l_____...______I-
'
-
~
_
.
I
_._..I_-
I
_
_
3
c Vp 1 kbar Vp4 kbars
(c.u.) (103m/s)
(103m/s)
4-7
5.0-5.2
2.5-3 128-260 2.75-2.85
_
.._I_.-.
_
I
_
4.1
3
18-19
6.2-6.5
6.4-7.0
2.9-3.1
3-4 -172-216
___-___I---.---I_.
3.7
6-10 ~_.___.____.I__
17
5.2
3-3.5 108-156 2.75-2.8 - __-_
__
.I__
___^I
Micaschist
8-11
I-"--------2-5=Schist
.I-__.
..
..__Phyllites
5-9
_________._.I____.I
Sand8
_._I-_
..
Whitequartzite
2-4
l__l_._._.___._l
._
Red quartzite
37-41
- _-
-
2.65-2.90- 4-4.5
_I_-._---"
-____I.-._
I
_
_
Shales
4rr.n.
(p.u)
__ __-
3-5
Gneiss
Pe
(We)
___ 10-30
--2.8-2.9
.____I
10-30
A
4.5-4.1
,--..I-.I_
0_7
2.5 '-6-6.5
6.0-6.2
6.9-7.1
6.2-6.3
6.2-6.3
I
_
.-.____I___
-.
has meant that
I
- A minimum thickness for zones is needed to base the
interpretation on sufficient statistical data. This also
decreases to a reasonable level the number of cross-plots
to analyze,
- Never base your interpretation on one type of crossplot (i.e. density-neutron), as too often carried out,
- Use all the possibilities of the Z-plot technique. By
adding a third parameter one can validate the data or
determine the influence of typical minerals.
. For instance CALI (or Cal) on the Z-axis enables to
validate the data or, in other terms, not take into account
samples which are affected by cave or rough hole; allows
the recognition of porous and permeable zones (mudcake deposits).
. For instance GR, POTA (K) and THOR (Th) on the Zaxis enable to recognize the influence of radioactive minerals (i.e. clays, orthoclase, mica or zircon).
70
. For instance SP or SSP, if sufficient contrasts exist
between the mud and the formation water salinity, helps
to determine where are the reservoirs, even if they are
radioactive.
If you plot data corresponding to a systematic sampling (for instance samples each half a foot) do not forget
that some of the data do not represent any geological
reality due to the lack of vertical resolution of standard
logs. Eliminate data corresponding to an isolated value
(frequency 1 or 2). Those data correspond either to transition from one bed to another or to thin beds.
Consequently, it is highly recommended to adopt a segmentation of log data in electro-beds before any crossplot (Fig. 2-54).
Serralog Q 2003
Composition [Chapter 2
- __
I
71
J
Figure 2-54 - (a) On the left standard presentation of logs. (b) On the right presentation of logs after segmentation in electro-beds. This segmentation takes into account the vertical resolution of each log measurement, the statistics and the uncertainties on the measurements.
Cross-plot analvsis
Several cross-plots are essential for the determination
of the principal minerals composing a rock. The use of the
Z-plot technique is even preferred as it introduces a third
parameter allowing us to avoid misinterpretation. Never
try to determine the mineralogical composition from a single cross-plot or a reduce set of log data.
For lithological and mineralogical determination, the
most useful cross-plots are listed hereafter:
- Density versus (vs) neutron with GR, Th, K, Pe, SP,
eatt, Cal and frequency on the Z axis. This plot is interesting and popular because it gives a good evaluation of
the porosity if the rock is clean (without clay or shale),
even if one does not know perfectly the lithology.
- Pe vs K with Th, rma, SP, eatt, Cal on the Z axis.
- Pma vs Uma with Th, K, GR, eatt, SP, Cal on the Z
axis.
- Th vs K with SP, Pma, Pe, eatt, Cal on the Z axis.
- At vs PHlN with SP, Pe, Th, K, eatt and Cal on the Z
axis.
- AtC VS AtS.
These plots are reproduced hereafter with the position
of the main minerals. Some of them are represented by
an ellipse. This ellipse reflects the variations in elemental
composition of those minerals (Fig. 2-55 to 2-59).
Many other cross-plots can be used. Remember that
Z-plots are very useful for a more precise and accurate
determination of the rocks and their mineral constituents.
Semlog 8 2003
-
Figure 2-55 Density vs neutron porosity croos-plot. This plot is very
popular as it gives an idea of the porosity even if the lithology is not
known correctly (courtesy of Schlumberger).
This will be illustrated by the determination of rocks
without porosity for instance evaporites (Fig. 2-60 to 2-64)
or igneous rocks, which are characterized by the absence
of effective porosity (Figs. 2-65 to 2-68). The interpretation of cross-plots corresponding to porous rocks requires
sometimes more data as the fluid filling the pore space
influences the log responses.
71
1.1
110
0.8
-
Figure 2-58a M vs N cross-plot for mineral identification
(courtesy of Schlumberger).
40
Figure 2-56 - Sonic slowness vs neutron porosity cross-plot.
(courtesy of Schlumberger).
If geochemical logging data exist, cross-plots of Al vs Si,
Ca vs Si, Fe vs Si, K vs Si, Fe vs Ca, Al vs K, Al vs Fe will
be very useful for a more precise determination of the
mineral composition.
T3
B
Q
Q,
a
K (wgt Yo)
Figure 2-57 - Pe vs Potassium (K) cross-plot. Several minerals are represented by ellipses to reflect their variations in elemental composition.
72
Figure 2-58b - Matrix Identification (MID) plot for identification of the
minerals composing a rock
(courtesy of Schlumberger).
Serralog 0 2003
Next Page
Composition
I Chapter2 I
73
Table 2-33
Tests for identification of rocks from resistivity, gamma
ray spectrometry, photoelectric index and matrix density
measurements.
A
n
6
.
INPUT LOG DATA
0
N
-
Figure 2-59 pma vs Uma cross-plot for mineral identification.
From a logging point of view, rock types can be divided into two main categories:
- rocks without interconnected porosity, and therefore incapable of constituting a reservoir. These always
exhibit very high resistivity. Igneous and metamorphic
rocks and evaporites belong to this category.
- Rocks with interconnected porosity which can
constitute potential reservoirs. They can sometimes
exhibit high resistivity due either to low porosity, or to a
high content of non-conductive fluids (hydrocarbons, bitumen or fresh water), or they can exhibit variable conductivity because of a proportion of the pore space is occupied by a conductive fluid (water with variable concentrations of salts). Sandstones, limestones, dolostones
belong to this category. Shales belong also to this category even if the porosity is not effective. Table 2-33 summarizes the tests that can be done to identify the rocks.
The preliminary remarks suggest the use of resistivity
values to break up the formations in these two categories.
1 - Formations of verv hiah resistivity (> 600 ohm-m)
As previously indicated, these formations can correspond to one of the following rock types:
- igneous rocks, either plutonic or volcanic, which have
not been altered or fractured;
- metamorphic rocks which have not been fractured or
altered;
- sedimentary rock of zero porosity (chemical rocks of
the evaporite class: gypsum, anhydrite, halite, polyhalite,
sylvite.. .) or of very low porosity (quartzites, compact and
well cemented carbonates which are not fractured);
- rocks which are porous (reservoirs), but filled with
non-conductive fluids (hydrocarbons, fresh water, bitumen, asphalt).
The selection of one or the other among the above
possibilities requires complementary measurements.
Serrabg Q 2003
1-1 - Separation of Resistivitv Curves
Quite often in sedimentary rocks the resistivity curves
measured by tools with different depths of investigation
(macro-devices: deep and shallow laterologs, deep and
medium inductions, and micro-devices: microlaterolog
and microspherically-focussed logs) show a continuous
separation. This can be attributed to invasion, and hence
will indicate a porous and permeable rock.
1-2 - No seoaration on Resistivitv Curves
When very high resistivity intervals show practically no
separation between the various resistivity curves, one can
reasonably conclude that the rock has no porosity, or at
least no connected porosity. The corresponding interval
can correspond to a non-porous rock.
If the radioactivity is low, and the hydrogen index is
between -2 and 2 P.u., the probability that this rock is sedimentary, be it of chemical origin or strongly cemented or
re-crystallized by diagenesis, is very high, especially if the
surrounding formations are of variable resistivity. If not, it
is more likely to be a metamorphic rock such as marble or
quartzite.
73
Previous Page
74
Well logging and Geology
Weak radioactivity and a high neutron hydrogen index
(> 35 P.u.) indicate a sedimentary rock of chemical origin
which is rich in water molecules within the crystals (gypsum, bischofite, tachydrite, epsomite, kieserite, trona).
The techniques for determining mineralogical composition
will be described later.
Weak radioactivity and a hydrogen index in the range
of 10 to 25 p.u. may indicate a basic volcanic rock such
as andesite or basalt.
Cases of variable but significant radioactivity, and
hydrogen index varying with radioactivity where the beds
are intercalated with non-radioactive beds of zero or high
values of hydrogen index most probably correspond to
sedimentary evaporites with varying amounts of potassium salts. If, on the other hand, the beds are fairly homogeneous and thick, the indications are for an igneous rock
(plutonic or volcanic), or for a foliated metamorphic rock
such as slate or micaschist.
It is important to recognize them. As indicated, generally
this determination starts by the fact that plutonic rocks
never present porosity, and therefore exhibit very high
resistivity, both on micro- and macro-devices, if they are
not fractured or weathered (Fig. 2-60).
Identification of igneous rocks
Plutonic rocks can be encountered either as sills or
dikes or even lacoliths, intercalated in sedimentary rocks,
and sometimes as batholiths constituting basement rocks.
-
Figure 2-62 Cross-plot GR vs density for identification of igneous
rocks (from Desbrandes, 1982).
1
-
Figure 2-60a Typical log responses in granite. The resistivity drops on MSFL are due to fracture influence generating boreholebreakout. The
peaks on Pe and U are due to pyrite influence.
74
Serralog 8 2003
Composition
-
POTA MOR
I Chapter 2 1 75
/WTA-PW
-
Figure 2-606 Several cross-plots corresponding to the previous log intewal showing the log responses in granite.
Some data are affected by weathering and by fractures filled with pyrite (higher Pe values).
Volcanic rocks exist as ashes, lava flows or tuffs.
Sometimes, they can show high porosity, most of the time
not connected (pumice, vesicular basalt). In that case
they are very resistive (Fig.2- 61). If the porosity is connected, due to fracturing or weathering, the resistivity can
be low and the rock can constitute a reservoir.
The radioactivity of the igneous rocks depends on the
Serralog 6 2003
content of radioactive minerals such as potassium
feldspars, feldspathoids, micas or trace minerals such as
zircon, allanite, xenotime or monazite. The potassium,
thorium and uranium contents decrease from acidic to
ultra-basic rocks (Fig. 2-24a). As a result, the radioactivity also decreases from acidic rocks to basic rocks (Table
2-31 and Fig. 2-62).
75
The speed of sonic waves is usually high, above 6000
m/s for plutonic rocks and 5250 m/s for volcanic rocks
without connected porosity.
The density of igneous rocks increases from acidic to
ultra-basic rocks (Fig. 2-63). Works made by Henkel
(1976), based on 30,000 samples from Northern Sweden,
links the density with the silica content (Fig. 2-64), and
with the ratio (Ca + Mg)/(K + Na) (Fig. 2-65). The silica
content decreases from acidic to basic rocks. On the contrary, Ca and Mg increase from acidic to basic rocks and
simultaneously K and Na decrease.
The sonic velocity increases from acidic to ultra-basic
rocks (cf. Fig. 2-52).
The determination of the igneous rocks must be based
on the analysis of the cross-plots illustrated by Figures 252 and 2-53, and Figures 2-62 and 2-63, completed by
the cross-plots of Figures 2-44a, 2-64 and 2-65 if elemental analysis by spectrometry has been obtained.
h
Identification of evaporites
CNL Neutron porosity index (P.u.)(apparent limestone porosity)
With the exception of certain carbonates of chemical
Figure 2-63 - Density vs neutron porosity cross-plot for igneous rocks.
I
Figure 2-61 - Typical log responses in basalt.
origin, the evaporites are characterized above all by high
resistivities of about equal magnitude on all the resistivity
measurements. This is a natural consequence of the fact
that after compaction these rocks have no porosity, and
do not, therefore facilitate invasion, and also that none of
76
the constituent minerals are conductive.
Their radioactivity will depend on the salts present.
Only potassium salts are radioactive, the activity being a
function of the potassium content. When potassium salts
Serralog Q 2003
Composition
I Chapter2 1
77
and will depend partly on the types of potassium salts
present and partly on their volumetric percentages.
Si02
Determination of the tvpe of evaDoritic salt
If the thickness of the salt bed is greater than the vertical resolution of the measuring device, which is the usual
case, it can be easily identified by its log response.
One can assume that, because of the way in which
these formations originate - chemical precipitation from a
brine -the constituent minerals are usually pure, and have
mineralogical, chemical and physical characteristics
which are well-defined and constant. It is only to be
expected then, that their log characteristics are also welldefined and constant. Table 2-28 sets out the characteristics of the more abundant salts and minerals.
It is clear from this table that ambiguities in identifying
an evaporite are practically non-existent given an adequate logging suite and a bed thickness greater than the
vertical resolutions of the tools. In such a case, the nature
of an evaporite salt is defined by the log value attained by
each measurement.
Density (g/cm3)
0
WhlU
1.64 Tuhhydde
Figure 2-64 - Density vs Si02 cross-plot for igneous rocks
2.0
(from Henkel, 1976).
Ca + l2Mp
K + l.UNa
DEVIATION
CAUSEDBY
SERPENTINISATION
2.2
p
O^
.
Fl
2.4
s
XI
2.6
Q
2.8
3.0
0
10
20
30
40
50
K (wgt %)
-
Figure 2-66 Density vs potassium cross-plot for determination of
evaporite minerals.
Density (g/cm3)
Figure 2-65 - Density vs ratio (Ca + 1.2 Mg)/(K + 1.43 Na) cross-plot
for igneous rocks (from Henkel, 1976).
form laminations of varying thickness and frequency
among deposits of non-potassium salts such as gypsum,
halite and anhydrite, the overall activity will be variable
Semlog Q 2003
Thus, if At = 67 f 3 p / f t (from the Borehole
Compensated Sonic, BHC), or Pb = 2.04 f 0.05 g/cm3 and
Pe = 4.5 f 0.3 (from the litho-density measurement), the
immediate conclusion is for halite. However, cross-plots
combining two or three log measurements are recommended for a more certain diagnosis. Figs. 2-66 to 2-70
show such plots, with the positions of the pure salts
marked.
Finally, one can employ the electro-lithofacies method.
The method consists of comparing the electrolithofacies
for each interval of stable response (facies) with the theoretical electro-lithofacies for each evaporitic salt.
77
78
Well logging and Geology
Several real examples of log responses in evaporites
will now be presented. They will demonstrate the validity
of the logging approach for the lithology and mineralogy
determination.
9
Q
tn
3.
W
Figure 2-67 - Density vs sonic travel time cross-plot for identification of
evaporite minerals.
Neutron porosity
Figure 2-69 - Cross-plot of sonic transit time vs neutron porosity for
identification of evaporites
K (wgt %)
Figure 2-68 - Cross-plot of sonic travel time vs potassium for identification of evaporite minerals.
Figure 2-71 reproduces a reduced set of log data
recorded in an evaporitic formation of North Sea essentially composed of anhydrite, halite and polyhalite, the latter being characterized by radioactivity pics. The crossplot (Fig. 2-72) reproduces the real position of sampled
levels of Figure 2-71. This plot confirms the mineral mixture but allows the precise determination of the logging
Darameters of those minerals. Polyhalite has a sonic travel time close to 58 p / f t and a gamma ray close to 220
API. The sonic travel time for halite shoud be 68 ps/ft.
On the previous cross-plots the mineral positions correspond to their theoretical values.
78
9"
Neutron porosity
Figure 2-70 - Cross-plot of density vs neutron porosity for identification
of evaporites.
In Figure 2-73 are reproduced the log responses,
recorded by the Platform Express equipment, and the
lithology determination realized at the well site.
Serralog Q 2003
Composition
The next example corresponds to a similar mineral
association. Based on cross-plot interpretation (Fig. 2-75)
the authors (Ford et a/., 1974)suggest a small shift of the
logging parameters of the three minerals. For polyhalite
the density is equal to the theoretical value, but the SNP
neutron porosity should be increased by 2 P.u.. Its sonic
value is similar to the previous one (58ps/ft).
I
Chapter 2 I 79
1
The sonic value for halite is also close to 68 ps/ft.
A
-
Figure 2-71 Gamma ray, neutron, sonic responses in
evaporite series in North Sea (Serra, 1980).
Figure 2-73 - Example of log responses in evaporites. In the lithology
track is reproduced the rock type determination made by interpretation
of the data provided by the Platform Express equipment
(courtesy of Schlumberger).
Based on these new values, the log interpretation
computes mineral percentages very close to the one evaluated by the geologist (Fig. 2-74).
The next five cross-plots (Figs. 2-76 to 2-80)allow a
more precise determination of the real logging parameters
of other evaporitic minerals.
Figure 2-72 - Cross-plots of some log values (points I to 20) extracted
from log data reproduced in Figure 2-70 (from Serra, 7980).
Serralog Q 2003
79
1TWIEl.N.1
Geologist's evaluation
LTW1CI n
.
1
Log interpretation
SNP Porosity (P.u.)
Figure 2-74
- Comparison of log interpretation with geologist's interpretation (from Ford et al., 1974).
Figure 2-75 - Density vs SNP porosity (plot on the bottom left) and
sonic travel time vs SNP porosity cross-plot (above), corresponding to
the previous log data. They suggest a small shift of the theoretical
parameters of evaporite minerals (from Ford et al., 1974).
.80
Figure 2-76 - Cross-plot density vs neutron porosity corresponding to
real log data. Accumulation of points close to typical minerals and
trends between minerals suggest the variations in composition of this
evaporitic formation composed of anhydrite, halite, tachydrite, camallite
with a marked trend toward sylvite.
SNP porosity (P.u.)
80
They illustrate also the mineral mixtures (thin intercalated layers of different mineral composition in this evaporitic environment) and the trends toward the principal
minerals. These plots, based on a standard sampling (one
Serralog 0 2003
Composition
I Chapter2 I 81
value each half a foot), reflect as well the lack of vertical
resolution of the logging tools, many values corresponding
to artefacts linked to transition values from one bed to the
following one.
Pe
( bfej
Figure 2-79 - Pe vs density cross-plot corresponding to real log data
allowing the statistical determination of the mineral logging parameters.
Figure 2-77 - Potassium vs neutron porosity cross-plot corresponding
to the same real log data. Accumulation of points close to typical minerals and trends between minerals suggest the variations in composition
of this evaporitic formation. Trends from halite to sylvite and
camallite are obvious.
-
1
3
Figure 2-80 - Another example of real log data combining density and
potassium (from Dragoset, 1977).
fi
E
Y
cb
v
The next example (Fig. 2-81) illustrates the log
responses in Prairie Evaporite Formation rich in sylvite,
marked by a very high gamma ray response.
GR
Figure 2-78 - Density vs potassium cross-plot corresponding to the
same real log data. Accumulation of points close to typical minerals
and trends between minerals suggest the variations in composition of
this evaporitic formation. Trends toward sylvite and camallite are
obvious.
Serralog Q 2003
NEUTRON
DENSITY
SONIC
Figure 2-81 - Composite-log of a section of the Prairie Evaporite
Formation of Saskatchewan, Canada, illustrating the log responses in
evaporitic formation rich in sylvite confirmed by the cross-plot of Figure
2-82 (from Raymer & Biggs, 1963).
81
The corresponding density vs neutron-hydrogen index
cross-plot (Fig. 2-82) reflects the sylvite influence.
'1
1
\
Neutron H index (%)
--"
/
0
I
Figure 2-82 - Density vs neutron-hydrogen index cross-plot corresponding to the sampling (points 1 to 20) of the previous data
(from Raymer & Biggs, 1963).
Figure 2-86 (next page). The log data (gamma ray, neutron hydrogen index, density and sonic) were interpreted
by Schlumberger's Global program. Z-plots, shown in
Figure 2-85, were used to statistically define the real logging parameters of the five minerals present in this formation: halite, bischofite, carnallite, kieserite and sylvite.
Good agreement is observed between the results
obtained by solving a system of linear equations and
those obtained by core analysis (thin section and X-ray
fluorescence analysis). This example confirms once more
the validity of the logging approach.
Finally, the log characteristics of each mineral are not
always known exactly, or can vary as a result of small elemental composition variations, or even calibrations errors
or borehole or mud effects. For this reason, several precautions are necessary: proceed to a prior check of log
quality and calibration accuracy, exact definition of the
mineralogical model and verification of the log characteristics of each mineral. For all these reasons, the actual
values do not correspond exactly to the theoretical values.
Of course, this must be taken into account, and crossplots should be used to define them more precisely.
The following example, from Crain and Anderson
(1966) on the same Prairie Evaporite Formation, compares log interpretation results with core analysis results.
One can observe a relatively good agreement if one takes
into account the lack of resolution of logging data (Fig. 283).
Figure 2-83 - Log interpretation results and core data comparison on
the Prairie Evaporite Formation
(from Crain & Anderson, 1966).
The next example (see Fig. 2-84 next page) illustrates
the logging responses in an evaporitic formation rich in
tachydrite, carnallite and halite with some clay intercalations. The logging parameters of tachydrite are well determined: the sonic travel time # 91 ps/ft, the density # 1.66
g/cm3 and the CNL neutron index between 68 and 58 p.u.
and the gamma ray close to 5 API.
f
-?2
f
Neutron porosity (P.u.)
Figure 2-85 - Z-plots corresponding to the previous log example (16301775 m). Positions of carnallite, halite and bischofite are indicated on
each plot and confirmed by the Z values
(from Haile & Blunden, 1984).
Another example from the Zechstein formation of
Netherlands (Haile and Blundein, 1984) is illustrated by
82
Serralog 0 2003
Composition
I Chapter 2 I
83
GR
Figure 2-84 - Composite log of an evaporitic formation essentially composed of halite, carnallite and tachydrite with some intercalation of clays.
SGR (API)
Figure 2-86 - Another example of log responses in an evaporitic formation from Zechstein of Netherlands. Comparison of core description and
Global processing results. A good agreement is observed between the two methods (from Haile & Blunden, 1984).
Serralog 0 2003
a3
Next Page
J
Figure 2-86c - Another example of Global interpretation results with the illustration of the log measurements and their manual interpretation;
Compare the log responses with the theoretical values of the evaporitic rock types. All of the evaporitic rocks show a very high resistivity
(courtesy of Schlumberger).
a4
Serralog 0 2003
Previous Page
Composition
Identification of metamorphic rocks
As we have already observed, these rocks have no
interconnected porosity other than that arising later from
fracturing or alteration. As a result, the resistivity is usually high, except in the case of phyllites which are rich in
phyllitic minerals such as mica and chlorite, and because
of twinning, which provides a large specific area and a
high cation exchange capacity (CEC). The presence of
metallic sulfides may also lead to high conductivities.
If an adequate logging suite is available, it is usually
possible to determine the type of metamorphic rock present.
The quartzites usually exhibit moderate radioactivity
(40-60 API). Their neutron hydrogen index is about zero,
their density and photoelectric capture cross-sections
close to those of quartz (Pb = 2.66 g/cm3 and Pe = 1.81
b/e respectively), and the sonic travel time is between 50
and 55 ps/ft. The presence of feldspars, micas or heavy
minerals will "pull" the cross-plot points towards these
minerals proportionally to their content, and this results in
a dispersion of the points.
Marbles are generally characterized by weak radioactivity (10-30 API), and a neutron hydrogen index of almost
zero. The density and sonic velocity increase, and the
capture cross-section decreases with increasing levels of
dolomite in the marble (Pb 2.72 to 2.85 g/cm3; At : 50 to
42 ps/ft; Pe 5.08 to 3.14 b/e). The presence of impurities
such as silicates, iron or sulfur will, of course, slightly
modify these parameters.
Like the quartzites, the granulites show moderate
radioactivity (50-55 API), but their density is higher and
varies between 2.73 and 2.93 g/cm3 according to the
amount of pyroxene (hypersthene). The potassium, thorium and uranium contents distinguish these rocks according to the degree of metamorphism to which they have
been subjected. The amphibolites, being rich in thorium,
are noticeably more radioactive (70-80 API). They are
also denser (2.8 to 3.2 g/cm3), and the sonic travel time is
always below 47 ps/ft.
Gneiss has a composition varying between that of
granite and gabbro, depending on the origin. The log
responses are therefore similar. Their foliated texture
affects both the resistivity and the sonic travel time and
acoustic wave attenuation. The latter is the result of
anisotropy, which is itself a function of the mica content
(as shown by Christensen, 1965), and also of the angle
between the plane of foliation and the axis of the measuring device. Thus, the travel time and attenuation are less
when this angle is go", but increases as the angle
approaches zero. This has not been confirmed, however,
by a comparison between measurements and core observations.
Serralog 0 2003
1
Chapter2
1
85
Slates and schists (e.g. micaschists, chloritoschists),
also have a foliated texture, and should therefore exhibit
the same phenomenon. This may help to explain the difference observed by Sanyal eta/. (1980) on the sonic logs
recorded in two neighboring wells over two identical and
perfectly-correlated sequences of schists, micaschists
and quartzite, when there were no notable differences on
the other logs.
The identification of the various types of schists should
be possible by combining the log data, in particular the
density, photoelectric capture cross-section, and radioactivity, or better, the potassium content. Thus, the chloritoschists should show a low potassium content (less than
1 %) and quite a high value of Pe (between 3 and 5 b/e),
while a micaschist will have a potassium content close to
4 or 5 %, with a value of Pe varying between 2.5 and 5 b/e
depending on the type of mica present (muscovite or
biotite).
The theoretical positions of the principal metamorphic
rocks are shown in the cross-plots of Fig. 2-87 (see next
page).
Once the type of metamorphic rock has been identified, the choice of mineralogical model, and hence of the
response equations and parameters to be used, becomes
a simple matter.
Several scientific wells were drilled in Germany (KTB
project - which stands for Kontinentales Tiefbohrprogramm der Bundesrepublik Deutschaland or German
Continental Deep Drilling Program) through metamorphic
rocks. Practically all the logging measurements available
at this period (from 1987 to 1994) were recorded in these
wells. Nearly 400 logging runs were made in KTB-VB
(4000 m deep) and 266 runs in KTB-HB (9101 m deep).
One of the most interesting tool was the GLT
(Geochemical Logging Tool). This tool provided concentrations of silicon, calcium, iron, titanium, gadolinium, sulfur, aluminium, potassium, thorium and uranium.
Anothertool with a semiconductor detector - germanium was also used, which gave a higher sensitivity. Sodium,
magnesium, manganese, chromium and vanadium concentrations were the other data available. The uranium
was used as a graphite indicator as it tends to concentrate
at graphite accumulations. Several rock types were recognized and identified: different gneissic rocks, paragneiss from a sedimentary parent and orthogneiss from an
igneous parent, amphibolites composed mainly of
feldspars and amphiboles, marbles, metagabbro and
metabasites.
The set of data recorded during this project allows a
much better determination of the logging parameters of
these rocks. The reader can refer to the numerous scientific publications covering this study (Emmermann et a/.,
1993; Bram & Draxler, 1993; Grau et a/., 1993; Bram et
a/., 1995).
85
F
3
E
P
M
Neutron porosity (limestone)
Figure 2-87a - Example of cross-plots showing the positioning of certain metamorphic rocks (from Sanyal et al., 1980).
2 - Formations with variable resistivities
These formations generally correspond to reservoir
rocks, or to claystones or shales, or to coal and lignite.
As previously indicated, multi cross-plots and specially Z-plot technique will allow the recognition of the rock
type and the nature of the fundamental minerals composing this rock. This type of analysis will be illustrated by the
following examples.
Identification of detrital rock type
The following example corresponds to a siliciclastic
formation, a succession of sandstones and claystones
with few intercalations of calcareo-dolomitic sandstones.
The sandstones are well characterized by a typical separation of neutron and density curves (in yellow on Figure
2-88 next page), a relatively low thorium content and an
average potassium content close to 2% corresponding to
approximately 20% of potassium feldspar (microcline or
orthoclase). Claystones are marked by high thorium values.
86
N
Figure 2-87b - Example of cross-plots showing the positioning of certain metamorphic rocks (from Sanyal et al., 1980).
Multi cross-plot (fundamentally Z-plot) analysis (cf. Fig.
2-89 following pages) allows the recognition of the principal minerals composing the rocks: quartz and potassium
feldspar, the latter being characterized by the points along
the sandstone line on the neutron-density cross-plot
showing a relatively high potassium and a low thorium
content, and the marked trend towards the orthoclase
position on the Pe vs K and Th vs K cross-plots. The
presence of calcite and dolomite as cement is marked by
thin layers (resistive peaks on resistivity measurements,
density, neutron and sonic curves as well) and low thorium content and low gamma ray. The clay type corresponds to a mixed layer illite-montmorillonite (high thorium content, close to 25 ppm, and potassium content
between 2.5 and 3%).
Another example of dolostone and arkose beds is illustrated by Figures 2-90 and 2-91 (next pages). The orthoclase (or microcline) presence is marked by a high potassium and a low thorium content.
Serralog 0 2003
Composition
DEPTH
Track 1
Track 2
Track 3
I Chapter2 I
87
Track 1
Figure 2-88 - Typical log responses in a siliciclastic formation.
Serralog (0 2003
87
88
Well logging and Geology
_I
Figure 2-89a - Density vs neutron porosity Z-plots corresponding to the example of Figure 2-88. Observe the low thorium content close to the
sandstone line that increases progressively toward the right indicating the position of claystone which present also the highest potassium percentage indicating the main clay type :possibly a mixed layer illite-montmorillonite. The sandstones content between 7.9 and 2.5% of potassium
indicating a subarkose to arkose nature for these sandstones.
F
E
Figure 2-896 - Cross-plots associating potassium and Pe values for the mineral identification.
88
Serralog 0 2003
Composition
I Chapter 2 I
89
Figure 2-89c - Example of Z-plots allowing the determination of the principal minerals and rock parameters. Observe the trend towards the orfhoclase position and the general trend towards the top indicating the increase in clay
Figure 2-89d - Example of Z-plots associating neutron and sonic allowing the determination of rock parameters. Observe the position of shale indicated by high values (red dots) of thorium and potassium.
Serralog 0 2003
89
90
Well logging and Geology
DEPTH
-
Figure 2-90 Example of log response in arkose intercalated in dolostone. The arkose is characterized by high gamma ray value linked to a high
potassium content (reaching approximately 5%), a low thorium (between 2 and 6 ppm), a Pe value close to 2-2. I We, and, sometimes, a curve
separation between neutron and density (neutron on right) in a limestone compatible scale presentation. Dolostone is characterized by low gamma
r a thorium
~
and potassium, a Pe close to 3 and a neutron curve on the left of the density curve. In that example, do not interpret the gamma ray
curve evolution as a clay indicator as too offen done.
-
Figure 2-91a Density vs neutron Z-plots with hquencv, thorium, potassium and Pe values on the Z axis.
Serralog Q 2003
Next Page
Composition
I Chapter2 I
91
WTW
-
Figure 2-91b Pe vs potassium Z-plots with frequency, density, neutron and sonic travel time on the Z axis.
mtm
-
Figure 2-91c Thorium vs potassium Z-plots with frequency, density, neutron and Pe on the Z axis.
Serrabg (b 2003
91
Previous Page
Identification of biochemical rock type
This identification is based on neutron and density
curves recorded in limestone compatible scale.
Limestones are characterized by neutron and density
curves practically superposed in a limestone compatible
scale presentation. Dolostones are characterized by a
small left shift of the neutron curve compared to the density curve. At the same moment thorium and potassium
curves must show very low values. The same type of
cross-plot analysis will confirm the mineralogy. Figure 292 (see next pages) illustrates a typical log response in
carbonate rocks.
DEPTH
The corresponding cross-plots are reproduced in
Figures 2-93 (see next pages). One can observe two
main populations of points close to respectively the limestone and dolostone lines.
Track 1
Figure 2-92 - Typical log response in a carbonate formation with successively from the bottom to the top: limestone, shale layers, dolostone, limestone, shale layers and dolostone at the top. Observe that the limestones towards the top are oil bearing while the dolostones below are water
bearing. This can be interpreted in the following way: at the origin, dolostone beds were limestone beds. An anticline was created in which oil was
accumulated by migration. The same formation outcropped and received rain falls which, by gravity, circulated into the lower part. As less salty it
generated a dolomitization of the lower limestone beds by diagenesis, the upper limestone oil bearing beds being protected by the closure of the
anticline.
92
Serralog Q 2003
NPHI(WI
Figure 2-93a - Density vs neutron Z-plots. Observe the two populations of points, one close to the limestone line, the other close to the dolostone
line, confirmed by the respective Pe values. Both populations show low values in potassium and thorium.
Figure 2-936 - Z-plots density vs Pe with frequency, thorium, neutron and sonic travel time on the Z axis. The marked trends towards lower density
and Pe indicate the porosity increase for limestone and dolostone.
Sernlog Q 2003
93
94
Well logging and Geology
-
Figure 2-93c Pe vs potassium 2-plots with frequency, neutron, thorium and sonic travel time on the 2 axis. Observe the trends towards the day
characterized by high potassium and thorium. The clay mineral is essentially illite.
Figure 2-93d - Sonic vs neutron 2-plots with frequency, thorium, potassium and Pe on the Z axis. The two populations of limestone and dolostone
are well marked specially on the 2-plot with Pe.
94
Serralog Cf) 2003
Composition
Another example in carbonate is illustrated by Figure
2-94 and cross-plots of Figures 2-95. Their interpretation
indicates clearly the limestone, dolostone and anhydrite
layers with sometimes some quartz or feldspar influence.
One can also observe some peaks of Pe indicating probably the presence of glauconitic pellets. This interval may
I Ch
be interpreted as starting with intertidal limestone
deposits composed of grainstones (more porous) and
mudstones (less porous zones), superposed by supratidal
dolostone deposits with anhydrite layers or nodules. The
dolomitic intervals above 8155 ft are oil bearing.
DEPTH
-
Figure 2-94 Example of log responses in carbonate formation with intercalations of anhydrite. Anhydrite beds are well Characterized by high resistivity and density values and low neutron values (close to 0).
Serralog 8 2003
95
Next Page
96
Well logging and Geology
-
Figure 2-95a Density vs neutron Z-plots with fmquencg sonic travel time, gamma ray and Pe on the Z axis. On the last Z-plot limestone and
anhydrite are well characterized by high Pe values.
Figure 2 - 9 s - Sonic travel time vs neutron Z-plots with fmquencg densiiy, gamma ray and Pe on the Z axis. On the last Z-plot limestone and
anhydrite are well characterized by high Pe values.
96
Serralog Q 2003
Previous Page
Composition
Automatic lithology determination:
The LITHO-ROCKLASS Programs
Introduction
As it seems obvious, the recognition of the rock type
requires a multi-cross-plots and histograms analysis. This
type of analysis is time consuming and, in the end,
tedious.
For many geologists and log analysts, interpretation of
log data, in order to give an accurate and reliable lithologic description, is not an easy job. To help them, this work
was done by geologists, experts in log and cross-plot
interpretation. The rock classification based on log
responses, or in other words the building of the electrolithofacies database, was achieved by interpretation of
more than 3,500 cross-plots with their histograms. These
log data came from more than 25 wells, distributed all
around the world and having crossed all types of sedimentary rocks, and on which existed also core analyses.
The goals of these programs are numerous:
- give an automatic, complete and continuous lithologic description of the formations crossed by a well, by comparing, at each sample level, the set of standard log
responses (gamma ray, thorium, uranium, potassium,
density, neutron, photoelectric index, sonic travel time.. .)
recorded in a well to an electro-lithofaciesdata base, previously built, for its attribution;
- provide the first step of a complete facies analysis;
- generate data for:
- sequence analysis
determination of the major geological events
which will be used as time markers for
correlations
mapping purpose
1. lithofacies maps
2. iso-liths maps
3. iso-percentage maps
4. iso-ratio maps ...
generate a more appropriate mineralogical model for
quantitative formation evaluation.
I Chapter2 I 97
The LITHO program uses a pre-built electro-lithofacies database (Fig. 2-96).
The construction of this data base relies on the fact
that a gaussian distribution generally describes quite well
natural sciences (Fig. 2-97). Each rock type is characterized by a gaussian distribution of its elemental and mineral compositionand consequently of its log responses (Fig.
2-98). The combination of several distributions generates
an hyper-ellipsoid, or hyper-volume, in the n dimensional
space, n representing the number of different log measurements used for this lithologic determination.
tv
Mean value
I
Standard Mation
Figure 2-97
- Gaussian distribution describing natural sciences.
-
-
-
.E
cn
C
B
Pe (b/e)
-
Figure 2-98 Combination of log distributions (here density and Pe)
characterizing a typical rock. In two dimensions this generates an
ellipse.
-
Figure 2-96 Schematic representation of the location of the main rock
types on cross-plots.
Serralog (b 2003
ROCKLASS is the name of the new software written
for workstations. It allows the interactive building of an
electro-lithofacies database by direct interpretation of
CrOSS-plots.
97
98
Well logging and Geology
Fluid point
Construction of the database
Several approaches exist.
1. Interpretive approach
This one corresponds to a multi cross-plot and histogram analysis of actual logs. It is based on a good geological knowledge of their interpretation and requires, as
much as possible, core data for validation. The original
electro-lithofacies database of LITHO was realized this
way.
2. Empirical approach
This approach uses clustering techniques similar to
those used in FACIOLOG. It requires a zoning of log data,
after their depth matching, and the elimination of too thin
beds or of ramps that can cre.ate artificial electro-lithofacies without any geological validity. A previous segmentation of logs is recommended.
3. Interactive approach
This one can be used with the help of workstation
(GeoFrame). In that case the ellipses are built directly on
cross-plots reproduced on the screen. To avoid any problem, beds must be selected on the composite log reproduced on the right side of the screen. Then, the ellipses
must be aligned along the equi-lithology lines (Fig. 2-99).
As much as possible, for lithological purpose alone and
for decreasing the number of electro-lithofacies, try to
build ellipses including the total porosity range (Fig. 2100).
-
Figure 2-99 The long axis of the ellipse must be aligned along the
equi-lithology line.
98
-
Figure 2-100 When building the ellipses include the total porosity
range and for that purpose go from -5% and 55% with a coefficient of
correlation equal to -0.9 for RHOB vs PHIN. This will reduce the number of electro-lithofacies.
When building ellipses, try to generate ellipses of
approximately same size to avoid conflicts between big
and small ellipses (Fig.2-99). As all the space must be
covered by hyper-ellipsoids to avoid any interval without
lithologic attribution, some hyper-ellipsoids cover the
same hyper-volume. A point, representing a bed, may be
closer to the center of the small ellipsoid but will be attributed to the bigger one due to the fact that the logic of attribution is based on the ratio of the distance to the center
to the distance from the ellipse external limit to the center
(Fig. 2-99).
-
Figure 2-101 Construct, as much as possible, ellipses of same size to
avoid affectation of the bed to a wrong lithology.
Serralog Q 2003
Composition
Equilithology line
I Chapter2 I 99
breaks due to fractures,
- Ap: a high value may indicate problems on the
density measurement due to rough hole conditions
T2 from CMR to discriminate between non-porous
and porous and permeable rocks, etc.
Threshold values from any log can also be used to orient the search.
-
Coefficient of correlation P 0
Example of LITHO results is reproduced in Figure 2103. One can observe a general good agreement
between this lithological description from logs and the one
made from core analysis. Too thin layers are not well
detected due to the lack of resolution of the logging tools.
Coefficientof correlation = 0.6
LITHOFACIES
CORE DESCRIPTION
Coefficient of correlation = 0.8
SUB
definition
-
Figure 2-102 Select the appropriate coefficient of correlation
between logs, when it exists, to be sure that the ellipse will be comctly
aligned along the equi-lithology line.
Select the appropriate coefficient of correlation
between logs, when it exists. This is fundamental for
measurements sensitive to porosity like density, neutron
and sound slowness.
Lous used in LlTHO-ROCKLASS
All the logs are useful for a reliable and accurate lithologic attribution. But, for the construction of the database,
measurements sensitive fundamentally to lithology and
composition (elemental and mineral) are used. They are:
density
- hydrogen index or neutron-porosity
- photoelectric index
- compressional and shear travel time of sound
- total gamma ray
potassium amount
- thorium and uranium.
-
-
Other measurements can be introduced if they exist:
weight percentage of Si, Ca, Fe, S, Al from GLT,
GST or RST logs
- yields ratios from GST
eatt and tpl from EPT tool.
-
-
The other logs will be used in a preprocessing step to
better orient the search of the appropriate ellipsoid:
- resistivity: very high resistivity may suggest either
an hydrocarbon bearing reservoir or a
- non reservoir rock such as halite, anhydrite, gypsum...
- spontaneous potential: if the SP curve is valid, a
SP deflection may indicate a porous permeable reservoir;
- caliper: mud-cake indicates a reservoir, cave may
indicate a shale or an unconsolidated sand, or borehole
serrakg ca 2003
Figure 2- 103 - Example of LlTHO processing results. Observe the
good agreement between the lithology from logs and the lithological
description from core. Too thin layers are not seen by the logs.
Quantitative interpretation
The determination of the mineralogical composition of
formations is based on the solution of a system of equations.
A detailed and correct description of a reservoir often
requires more parameters than the number of available
measurements. Thus the model usually has to be simplified in order to obtain the fundamental parameters of
porosity, saturation, hydrocarbon type, clay content and
lithology. In so doing it is of interest to estimate the quality of the interpretation and thus the validity of the chosen
model. This can be done by using the GLOBAL or ELAN
programs developed by Schlumberger (Mayer & Sibbit,
1980). GLOBAL may be described as series of processes
using the response equations given by the tools and introducing a degree of uncertainty to the measurements and
to the zoned parameters used, together with pre-defined,
geological and local constraints, as well as calculating an
incoherence function.
The tool response model is expressed as a system of
equations as follows
(ai = fi(X)>
where the ai corresponds to the tools or inputs, 11 is the
99
Well logging and Geology
100
vector of the unknowns or outputs. Thus in the RTGLOB
program we get:
a = (RLLD~RLLS, RILD? RMSFL)
x = (Rtl Rxo, di)
the chances of obtaining correct results are increased by
reducing the number of unknowns per lithology type, as
well as by adapting the equations to the actual models,
eg. type of distribution of clay or porosity. We are then in
a favourable situation, having a model which is overdetermined and adapted to reality and not dependent on
chance.
UIIY
Uncertainty or measurement dispersion is caused by
several factors : the equipment itself (statistics, electronic
noise, etc.), corrections made by the surface recording
system and environmental corrections. This uncertainty is
defined for each tool by a spread which is designated Oi.
Uncertainty in the response function of the tool
fi is also determined and has a spread which is designated 7.
The constraints are expressed as inequalities on the
results and are independent of the measurements
gj (XI 0
’
The constraints may themselves allow a certain tolerance and their spread is designated Zj.
The incoherence funcion is expressed as:
A(a,g) =
- fi(X))2
(gj(X))2
z (ai
___________
+ z----i
Oi2
+ Ti2
j
zt
where gj(X) is the negative element of the function.
Once established, the incoherence function is minimized
by an iterative algorithm using its partial derivatives.
Iteration proceeds until convergence is achieved. Quality
control of the result is achieved by comparing the original
logs with the ones recomputed from GLOBAL or ELAN
results. The incoherence curve is also reproduced and
when it is higher than 1, then the model chosen is unable
to correctly represent the tool responses. A low incoherence is only representativewhen the system is overdetermined. This does not mean that the result is correct but
only that the model satisfies the measurements made.
Discussion
These programs are powerful mathematical tools
adaptable to all types of interpretation models, since data
from any new tool or exterior parameter may be introduced. One just needs to enter the response equations.
Thus, for a recording by a GLT tool one can enter the
measurements made by this tool and thereby get a much
sharper and more exact interpretation (Fig. 2-104).
Comparison with the measurements made on core samples verifies the quality of this interpretation. If the model
introducedcan be guided by evidence derived either from
the geology or the results of processing by a LITHO and
SYNDIP or by a FACIOLOG (Suau & Spurlin, 1982) then
100
-
Figure 2-104 Example of ELAN results (courtesy of Schlumberger).
Examination of the incoherence curve can help to
check the quality of interpretation even though it only represents the coherence between measurements and
model. If incoherence is higher locally this may indicate
that at this point the model is not wholly adequate. One
must then analyze a log or the logs in order to try to establish the reason for discrepancy. Thus, if the recomputed
Serralog 8 2003
Composition
1-
101
density and the Pe value are too low compared with the
measurements made by the tools, this may indicate the
presence of a heavy mineral not included in the model
(eg. pyrite or siderite). The interpretation of the facies and
the environmentwill be helpful in choosing between these
hypotheses. We can introduce an increasing percentage
of the new mineral and first check if this improves the
incoherenceor not. If it does then we stop when the incoherence begins to increase once more.
can appropriate action be taken in the event of a tool malfunction, either by repairing the tool itself, or by re-running
the log with another tool of the same type or by running
one which provides comparable information. This is why
the various quality control and calibration control procedures were reviewed in detail in the first volume on log
acquisition and it may be useful at this stage to refer back
to that section.
Program selection depends on the complexity of the
geological model. It is however, useful to interpret an
interval using several models: a mineralogical model to
determine the percentage of minerals and to give a more
precise idea of the rock composition (Fig. 2-105 next
page); a textural model (Fig. 2-106 next pages) giving an
idea of grain size (sand, silt and clay) to obtain an idea of
sorting and permeability.
The goal of this filtering is to eliminate from the interpretation all the non representativedata linked to bad hole
conditions, lack of resolution of standard logs, measurement and statistical errors. This is achieved through a
segmentation of log data in electrobeds and ramps (Fig.
2-108 next pages) taking into account the tool vertical resolution, the measurement errors provided by service companies, nuclear statistics, depth match problems. This filtering has also the advantage to produce a document
closer to the geological reality as it generates electrobeds
and natural electrosequences or ramps.
Another example of well log interpretation using the
ELAN program of Schlumberger is reproduced in Figure
2-107 (next pages). In this processing and presentation, a
10 mineral model was used and the computation results
displayed as framework minerals (quartz, orthoclase,
albite and volcaniclastic), matrix (clay minerals), cement
(calcite) and accessory mineral (rutile).
SERRA LOG Approach
Previous mathematical methods have limitations as
the number of log measurements control the number of
unknowns that can be computed. In addition, quite often
the mineralogical model introduced in the computation
process is imposed for an interval in which lithology
changes may exist.
The basic logic of the SERRA LOG approach is that
the rock types existing in the Earth's crust are not too
numerous, and are well classified and described in books
of petrography and petrology. This classification is fundamentally based on the principal mineral association that
compose them. These minerals have generally typical log
responses. Consequently, it seems logical to classify the
rocks as a function of the complete set of log responses.
In fact the SERRA LOG approach is a processing using
inversion technique.
- Firstly, the rocks are created by an association of
minerals with defined ranges of percentage.
- Secondly, each rock is converted in log responses as soon as one knows the logging attributes of each
mineral entering into the rock composition and the
response equations for each tool.
The program SQWIZLOG, developed by the SERRA
LOG Company, consists in several steps.
quality control.
The various controls of log quality must be performed
at the wellsite as soon as the logs are recorded. Only then
- Filterina of the loa data
- Depth Matchina and Composite Loq
This stage in the preparation of the data is very important. During recording the tools can be momentarilystuck.
The cable elasticity generates a jump of the tools after
their sticking. This event is not at the same depth for each
measurement provided by the run (cf. Chapter 1, paragraph Depth match).
- Litholoaical determination
For lithological determination of each electro-bed the
program consults a data base that includes:
approximately the 150 most abundant minerals
with all their geological and logging attributes taking into
account their variations in elemental composition,
approximately 650 rock types with all their logging
attributes taking into account their mineral and elemental
composition with their range of percentage variation,
- filtering based on typical log responses of rocks
before consultation of the data base.
A better knowledge of the lithology and mineral
composition allows a more accurate evaluation of the
petrophysical properties of each sedimentation unit composing a reservoir. For instance, a wrong evaluation of the
clay content, based generally on the Gamma Ray
response, can induce an error on porosity and on saturation computation.
-
-
- Constructionof the electro-lithofaciesdata base
It consists to convert in log responses the rocks
described in books of petrography. This requires first to
know either the elemental composition of the rock or its
mineral composition, secondly the tool's response equations for each element or mineral.
For instance, assuming a tetrahedral classification of
biochemical rocks (Fig. 2-109) and that we want to characterize a pure limestone (with more than 95% of calcite),
101
Well logging and Geology
102
otmc
-
Figure 2-705 Example of results obtained with the GLOBAL program using a mineralogical model including quartz, orthoclase, biotite and illite as
clay. Compare with results reprvduced in Figure 2-706
(courtesy of Schlumberger, in Well Evaluation Conference, India, 7983).
102
senrrlog Q 2003
Composition
I Chapter 2 I 103
-
Figure 2-106 Example of results obtained with the GLOBAL program using a tectural model including sand, silt and clay.
Compare with the previous results shown in Figure 2-105
(courtesy of Schlumberger, in Well Evaluation Conference, India, 1983).
serralog 0 2003
103
Well logging and Geology
104
Permeability
Saturation
Total fluid volume
Matrix
Framework
Cements
-
Figure 2-107 Example of results of computation using the Schlumberger’s ELAN program. A 10 mineral model, induding GLT data, was used.
The presentation displays permeability data in track 1 (lefi), then saturation, then fluid content and mineralogical composition in terms of matrix,
framework and cement
(courtesy of Schlumberger).
104
Serralog 8 2003
Next Page
Composition
DEPTH
Track 1
Track2
Track 3
I Chapter 2 I
105
Track 4
Figure 2-108 - Segmentation of log data in electro-beds and natural ramps. Compare with raw logs reproduced in Figure 2-92.
the impurities being chert, dolomite, anhydrite or different
clay types, we will obtain for each log measurement the
following range of variations (cf. Table 2-34).
These different rock types can be positionned on a
cross-plot associating density and photoelectric index
(Fig. 2-110).
Shaly llmatone
Mul
- Consultation of the data base for attribution of the
rock tvDe to each electro-bed
Anhydrlte
The set of log responses corresponding to each electro-bed is compared to the electro-lithofacies data base
for attribution. To speed up and avoid possible conflict,
several thresholdings are applied to logging parameters.
-
Figure 2-109 Classification of biochemical rocks using a tetrahedron.
Serralog 8 2003
105
Previous Page
106
Well logging and Geology
This can be related to porosity and mineral changes.
Most of the time, these rock types correspond to the
same rock class (sandstones, limestones or dolostones)
with the same mineral association but with different mineral percentages. In that case the probability for the rock
class is equal to 1.
-I
Figure 2-110 - Classification of biochemical rocks using a tetrahedron.
Location of each class on a Pe vs m a cross-plot.
These thresholding are applied to resistivity, thorium,
potassium, density, photoelectric index, neutron porosity,
sonic values.
This type of processing is illustrated by the following
figures and cross-plots (2-111 to 2-119 next pages) completed by borehole wall images on the same interval
(Figs. 2-120 and 2-121 next pages).
Table 2-34
Range of variations of well logging responses for a pure
limestone (more than 95% calcite) without porosity.
If the rock classes are different, the probability of the
selected rock class will be less than 1. This probability is
computed as a function of the distance of the point - representing the data set (correspondingto the electrobed) in
the n-dimensional space (n being the number of logs
involved in the processing) - to the line joining the fluid
point to the point corresponding to the mean value of the
rock type, this value being computed assuming no porosity.
A similar approach is used to compute the probability
of the rock types corresponding to the same rock class.
The two probabilities (from 0 to 1) are reproduced as a
function of the depth on the side of the results.
PLATFORM EXPRESS* Lithology display
Parameters
Density
Slowing down length
Diffusion length
Thermal porosity
Epithermal porosity
Comp. transit time
Shear transit time
Photoelectric index
Volume photo. index
Gamma ray
Hydrogen wgt %
Carbon wgt %
Oxygen wgt %
Sodium wgt %
Magnesium wgt %
Aluminum wgt %
Silicon wgt YO
Potassium wgt %
Calcium wgt %
Iron wgt %
Sulphur wgt %
Thorium ppm
Uranium ppm
Symbol
Pb
Ls
Ld
f th
f epi
Dtc
Dts
Pe
Ue
GR
H
C
0
Na
Mg
Al
Si
K
Ca
Fe
S
Th
U
Minimum value
Maximum valul
2.6328
16.6
10.6
0
- 0.1
46.9
86
4.88
12.84
0
0
11.17
47.9
0
0
0
0
0
37.3
0
0
0
0
2.71 75
24.3
13.5
0.7
0.6
54.2
103
5.08
13.83
15.4
0.02
12
48.7
0.1
0.7
0.77
1.49
0.4
40
0.14
1.28
1.25
0.5
- Computation of the rock-type Drobability
As each rock type is represented by an hypervolume,
it may happen that a given set of log data may correspond
to several rock types of the electro-lithofacies data base.
106
In the PLATFORM EXPRESS lithology display, an
approach based on log segmentation is applied. Five log
data are used : density, neutron and photoelectric indices,
gamma ray and spontaneous potential. The three first log
data are converted in two new data: pmaand Uma.
The two
other data (GR and SP) are used to compute the clay
volume V,,. In fact, only three data are introduced. The
color change depends on Vcl and the position of each
point on a pma-U,,,,cross-plot. The colors assigned to the
plot are derived from a color cube with corners of yellow,
green, cyan and white on one face, which represents zero
volume of clay, and corners of red, black, blue and
magenta on the opposite face, which represents 100%
volume of clay. The colors on each V,, slice of the cube
show a gradation across the slice, and all slices show a
gradation of colors from V,, = 0% slice to V,, = 100% slice
(Fig. 2-124a & 2-124b next pages). Anhydrite is reproduced by a special blue (violet) added to the face of the cube
with less than 5% of clay. Halite plots in the yellow-white
region.
This lithology determination is based on an interpretation of cross-plots and clay indicators. As obvious if log
responses are affected by environmental factors (caves,
barite in the mud, etc.), by the bed thickness (surrounding
bed influence), or by special minerals (feldspars, micas,
pyrite, etc.), the lithology display will not represent the
geological reality. Other clay indicators must be used (i.e.
thorium and potassium) and environmental corrections
applied.
Serralog Q 2003
Composition
Figure 2-111
I Chapter 2 I 107
- Typical log responses in a detrital deposit.
-
Figure 2-112 The previous log data have been segmented in electmbeds and ramps.
Serralog Q 2003
107
108
Well logging and Geology
Figure 2-113 - Cross-plot associating neutron and density with on the Z-axis thorium. Observe the trend towards dolomite indicating the porosity
decrease due to dolomitic cement (indicated by the arrow).
I
Figure 2-114 - Cross-plot associating neutron and sonic with thorium on the Z-axis.
Serralog 0 2003
Composition
Figure 2-1 15 - Cross-plot associating potassium and photoelectric
index Pe with thorium on the Z-axis. Observe the trend towards position of orthoclase.
I Chapter 2 1 109
I
3
-
Figure 2-116 Cross-plot associating potassium and thorium with Pe
on the Z-axis.
3
Figure 2-117 - Interpretation in tern of lithology of the previous set of log data reproduced on the left.
Serralog d 2003
110
Well logging and Geology
0
Figure 2-118 - Interpretation of the previous data in terms of mineral percentages with on the right side the porosities from sonic (Wyllie equation),
and density
1
-
Figure 2-119 Borehole images on the same interval than the previous example and the results of their analysis in terms of more resistive and conductive isolated events compared to the resistivity background. These events comspond to microconglomerates. These pebbles are essentially
composed of chert (opal) as confirmed by the Uma vs Rhoma cross-plot (Fig. 2-120). The microconglomerate nature is confirmed by the borehole
images of Figure 2-121 which comspond to the interval marked by a red strip.
110
Serralog Q 2003
Composition
I Chapter 2
11 1
m
Figure 2-120 - Cross-plot Uma (X axis) vs Rhoma (Y axis) on the same
interval than the other cross-plots, with thorium on the Z-axis.
Observe the position of the points essentially in the triangle quartzmicrocline-opal, and some points in the quadrangle quartz-microclinecalcite-dolomiteindicating levels cemented by calcite or dolomite. The
red points with high thorium values correspond to shale but their position is influenced by cave effects on density
Figure 2- 123 - Color coding for pma-Uma cross-plot as a function of
Vcl. These colors are used in the lithology display realized on
PLATFORM EXPRESS answer product (courtesy of Schlumberger).
Pad orientation
-
Figure 2-121 Borehole image of a 1.5 m interval showing the microconglomertae nature of the sandstone.
3
Figure 2-124 - Example of lithology display made by Platform Express
(courtesy of Schlumberger).
-
Figure 2-122 Color cube of primary colors represent lithology classes
(courtesy of Schlumberger).
Serralog Q 2003
Ill
112
Well logging and Geology
Neural network approach
This approach uses a Simulated Neural Network
(SNN) to identify electrolithofacies.
A neural network is a lattice of neurons that are interconnected by synapses. It is inspired by the organization
of the human brain which is a specific architecture of
neurons connected via synapses with axons and dendrites (Fig. 2-125).
In the hardware structure of a neural network, neurons
are replaced by transistors, axons and dendrites by wires,
and synapses by capacitors, resistors and inductors.
As stated by Baldwin et al. (1990) “the simultaneous
operation of highly interconnected neurons produces
mindlike synergetic characteristics and qualities that can
be exploited to perform sophisticated log analysis and
interpretation”.
Several SNN types exist.
- The Self-organizing Maps (SOM) neural networks
(Fig. 2-126), originally devised by Kohonen, consists of
two layers of processing units: an input layer (log data)
fully connected to a competitive output layer (artificial
lithofacies).
-
Figure 2-125 Typical neural network of a human brain.
Each neuron may be connected to 10,000 other
neurons. Each neuron has one extension, called axon,
which may have numerous ramifications,and one or more
axon endings. The axon conducts the information only in
one way. The axon ending is composed of a certain number of very small bags which contains molecules. These
molecules transmit the information through the synapse
slit. The axon ending is the location where an electric
charge, coming from a neuron through the axon, is transformed into a chemical signal thanks to the neurotransmitters. The synapse slit is the space where the chemical
signals are transmitted from a neuron to the other. The
neurotransmitters are substances which match the molecules to the receivers in the same way than only one key
corresponds to a bolt. The receivers are located on the
sensitive surface of the transfer zone. The dendrites
receive information from the other neurons.
Each neuron operates very simply and processes
information in simple fashion. A neuron receives the information from the other neurons and “decides” to transmit it
or not as a function of the electric charge coming from the
cellular surface. However, each neuron of the lattice operates simultaneously and in parallel.
A Simulated Neural Network is a software equivalent
of a hardware structure designed to perform some task.
112
_I
Figure 2-126 - The selected lithofacies (Winner) collects information from all input data (preceding neurons). The output of each preceding neuron (input data) is modulated by a corresponding weight (wi)
before affecting the activity of the neuron.
Frayssinet et al. (2002) uses this approach to represent lithofacies. The map is a discrete set of formal
neurons. Each neuron of the map is associated to a referent vector in the data space. A rectangular map of size
(13x7) was generated (Fig. 2-127).
_I
Figure 2-127 Topologic map established with 4 log data (pb, Pe, GR,
4d.L=limestone; M=marls, gS=glauconitic sandstone; Sha=shale;
Sha 1=other shales; Si=silt; Sil =other silts; cS=coarse grain sandstone;
S=strict sandstone; SL=sandy limestone; B=sandy breccia; Lig=lignite
(from Frayssinet et al., 2002).
-
Serralog 0 2003
Composition
The labelling of the cells (neurons) was done by reference to the core data. Grey level between two neighboring neurons represents the distance between them. Rock
classes are represented by the gathering of neurons .
- The SNN can use Self-Organizing-Activation (SOA)
principles devised by Kohonen (1984) to describe observations of how neurons in an optic nerve operate and
interact. This is based on the fact that log analysts make
visual observation of logs, cross-plots and images in the
initial stage of well log interpretation. Baldwin et al. (1989)
had successfully applied 2D SOA principles to interpretation requiring visual correlation of dipmeter and full waveform sonic curve traces. They extended from two to eight
dimensions the SOA rules. They generated 8D inhibitory
and excitatory hyperspheres (Fig. 2-128). Each log measurement can be subdivided in a reduced number of segments (8). One neuron is assigned to each segment.
When combining log measurements one generates n8
hypercubes, n being the number of different log data
taken into account. Each hypercube corresponds to an
artificial lithofacies. This type of process generates a lot of
hypercubes or artificial lithofacies and requires large
memory and long processing time to obtain a solution. In
addition, many hypercubes do not correspond to any real
lithofacies.
When the 8D hypercube completes its operation a set
of hyperclusters of active neurons results. Each hypercluster in n dimensions represents a lithofacies type that
the log analyst can determine by visual observation of the
raw n data.
Surface neurons
Internal neurons
Synapses
Figure 2- 128 - Self-organizing hypercube (only three dimensions are
shown) containing excitation and inhibition hyperspheres
(from Baldwin et al., 1990).
To be sure that each lithofacies cluster represents only
a single neuron in n dimensions, a final stage was introduced by Baldwin et al. This stage is known as OnCenter-Off-Surrounding (OCOS) competitive activation.
“OCOS principles operate such that individual neurons
within a group compete among themselves until the strongest neuron eventually suppresses all others within that
group. The operation of the hypercube and OCOS processing completes the identification portion of the interpretation task.” (Baldwin et al., 1989).
The neural network structure is completed by a
Serralog $3 2003
Chapter 2
.
113
Competitive-Activation Pattern Classification (CAPC)
paradigm (McClelland & Rumelhart, 1988).
To explain the logics of the SNN and CAPC network
the following lines are extracted from the paper written by
Baldwin et al. (1989).
Figure 2-128 “shows the S N N during hypercube training of the CAPC network... It illustrates how the S N N
connects the hypercube and CAPC networks and describes the internal architecture of the CAPC network”
(Baldwin et al., 1989)... The CAPC network consists of
pools of input log neurons and one pool of lithofacies output neurons. The number of input pools equals the number of input logs. The number of neurons per input log
pool equals the number of segments into which each log
measurement has been broken... The number of neurons
in the output pool equals the number of lithofacies that the
SOA hypercube discovered. Every neuron in each input
pool is connected to every neuron in the output pool.
Additionally, these connections are bidirectional. That is,
every connection from an input pool neuron to an output
pool neuron is reciprocated as is shown for the third
neuron of the Cxo pool (Fig. 2-129)...
0 InhibitedNeuron
C Excited Neuron
* Inhibitory Synapse
Excitatory Synapse
-
Figure 2-129 - Lithofacies pattern identification and recognition by a
SNN trained from inputs of an auto-associated SOA hypercube. Due to
the complexity of the structure only few connections are reproduced
(i.e. the eight neurons of the K pool). This pattern illustrates how the
SNN connects the hypercube and CAPC networks and describes the
’
internal architecture of the CAPC network
(from Baldwin el al., 1990).
Input pools, however, are not allowed to connect to
other input pools. All within-pool neurons are interconnected with Off-Surrounding (0s)architecture (illustrated for
the fourth neuron of the K curve pool in Figure 2-128)...
The CAPC paradigm functions in the following manner. When an input signal consisting of one value per log
113
curve per depth is received by the CAPC network, only
one or at most two neurons within each log pool become
active... Neurons activated by the input signal begin activating lithofacies neurons via synaptic connections to the
output pool. Because these connections are bidirectionnal, activated lithofacies neurons begin sending activating
signals back to the log pools. Within-pool neurons also
interact via 0s competition. Thus, within-pool neurons in
each pool compete with each other for domination.
Simultaneouslx neurons in log pools activate neurons in
other log pools indirectly through activation of the lithofacies neuron pool”.
- Neural network systems can be trained to perform
particular tasks based on well log data. The Back
Propagation Neural Network (BPNN) is done by presenting the system with a representativeset of examples describing the problems to solve. Input samples (set of log
data) are connected to output samples (lithofacies description provided by core analysis). BPNN is in fact a
supervised training method. It requires the knowledge of
the output to a given input. After training, the output of the
Neural Network is compared to the desired output. The
observed difference is used to adapt the weights. The
adaptation of the weights starts at the output neurons and
continues towards the input data. After training, the neural
network approach can be used to recognize data that are
similar to those of the examples shown during the training
phase.
- The SNN approach is also used to generate a missing logging curve from available logging data. For instance the neutron curve must be predicted from a combination of density, gamma ray and photoelectric index.
Frayssinet et a/. (2000) based their method on a MultiLayer Perceptron (MLP). This method “acts as a non
linear regression method for the prediction task and as a
probability density distribution approximation for the outlier rejection task.... The MLP is made of three kinds of
neuron’s layers (similar to the back-propagation neural
network): an input layer, one or several hidden layers and
an output layer”.
Results of this method is shown in Figure 2-131.
J
Figure 2-131 - Reconstruction of the neutron curve from three other log
data and comparison with the recorded neutron curve
(from Frayssinet et al., 2000).
Discussions
The SNN approach has distinct advantages:
- computer codes are very short,
- they do not use a special programming language,
- they do not require the services of special knowledge
engineers or experts.
The SNN approach is able to learn interpretation techniques performed by experts.
However, most of the time core data are needed to
determine the type of lithofacies that the analyst wants to
predict from the logging data.
Quite often the logging measurements are not corrected for their lack of resolution. This can generate misinterpretation as the non-representative data, corresponding
to the transition depth between successive beds, represent clusters that can be confused with lithofacies.
Some logs used for the lithofacies determination (for
instance resistivity, SP) represent more often the porosity
and the fluid content than the lithology or the mineralogy.
R eferences and bibliography
Gamma ray
Neutron
Density
Figure 2-130 - Schematic of the back-propagation neural mtwork
(from Rogers et al., 1992).
114
ADAMS, J.A.S. (1954). - Uranium and Thorium contents of volcanic rocks. In: Nuclear Geology. J. Wiley &
Sons, New York.
ADAMS, J.A.S. & GASPARINI, P. (1970). - Gammaray spectrometry of rocks. Elsevier Publ. Co.,Amsterdam.
Serralog 0 2003
Composition
ADAMS, J.A.S. & WEAVER, R. (1958). - Thorium to
Uranium ratio as indicator of Sedimentary Processes:
examples of concept of Geochemical facies. Bull. amer.
Assoc. Petroleum Geol., 42 , 2.
AUBERT, G., GUILLEMIN, C. & PIERROT, R. (1978).Precis de mineralogie. Masson, Paris.
BALDWIN, J.L., OTTE, D.N. & BATEMAN, R.M.
(1989). - Computer emulation of human mental processes: application of neural network simulators to problems in
well log pattern recognition. Conference on Artificial
Intelligence in Petroleum Exploration and Production,
Texas A&M University, May 17-19, pp. 145-175.
BALDWIN, J.L., OTTE, D.N. & WHEATLEY, C.L.
(1989). - Computer Emulation of Human Mental
Processes: Application of Neural Network Simulators to
Problems in Well Log Interpretation. SPE 64th Annual
Technical Conference and Exhibition, San Antonio, Texas,
paper SPE 19619, pp. 481-493.
BALDWIN, J.L., BATEMAN, R.M. & WHEATLEY, C.L.
(1990). - Application of a Neural Network to the Problem
of Mineral Identification from Well Logs. The Log Analyst,
31, 5, pp. 279-293. Originally presented at M.G. .S., 3rd
International Symposium on borehole geophysic for minerals, geotechnical, and groundwater applications, Las
Vegas, Nevada, Oct. 2-5, 1989, paper U.
BARIAND, P., CESBRON, F. & GEFFROY, J. (1978). Les Mineraux: leurs gisements, leurs associations.
Editions du B.R.G.M..
BARTH, T.F.W. (1952). - Theoretical Petrology. John
Wiley & Sons, Inc., New York.
BAYLY, B.(1976). - lntroduction a la petrologie.
Masson, Paris.
BEERS, R.F. & GOODMAN, C. (1944). -Distribution of
radioactivity in ancient sediments. Bull. Geol. SOC.
America, 55.
BERRY, L.G. & MASON, B. (1959). - Mineralogy
Freeman & Co., San Francisco.
BISHOP, C. (1995). - Neural Networks for Pattern
Recognition. Oxford University Press.
BLATT, H., MIDDLETON, G. & MURRAY, R. (1980). Origin of Sedimentary Rocks. 2nd ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
BOLVIKEN, E., STORVIK, G., NILSEN, D.E., SIRING,
E. & van der WEL, D. (1992). - Automated prediction of
sedimentary facies from wireline logs. In: Geological
Applications of Wireline Logs II, HURST, A., GRIFFITHS,
C.M. & WORTHINGTON, P.F. (eds), Geol. SOC.London,
Special Publication no 65, pp. 123-140.
BOYER, S., ANDRIEUX, P. & MICHAUT, B. (1993). Principes physiques de fonctionnement des outils de diagraphie. Geologues, 99.
BRAM, K. & DRAXLER, J.K. (1993). - KTB Report 931: Basic Research and Borehole Geophysics (Report 14).
Stuttgart, Germany.
BRAM, K. et a/. (1995). - The KTB BoreholeGermany’s Superdeep Telescope into the Earth’s Crust.
Oilfield Review, 7, 1, p. 4-22.
BURKE, J.A., CURTIS, M.R. & COX, J.P. (1967). -
1
Serralog 0 2003
I Chapter 2 I
115
Computer Processing of Log Data enables better
Production in Chaveroo Field. J. Petroleum Technol.
BURKE, J.A., SCHMIDT, A.W., & CAMPBELL, R.L. Jr.
(1969). - The litho-porosity crossplot :a method of determining rock characteristics from the computation of log
data. SPWLA, 10th Ann. Log. Symp. Trans.
BUTTGENBACH, H. (1953). - Les mineraux et les
roches. Dunod, Paris.
CAILLERE, S. & HENIN, S. (1963). - Mineralogie des
argiles. Masson, Paris.
CAROZZI, A.V. (Ed.) (1975). - Sedimentary Rocks.
Benchmark Papers in Geology, 1 5 , Dowden, Hutchinson
& Ross, Inc., Stroudsburg, Pennsylvania.
CHRISTENSEN, N.I. (1965). - Compressional wave
velocity in metamorphic rocks at pressure to 10 kbars. J.
Geophys. Research. Vol. 70, no. 24.
CLARK, S.P.Jr. ed. (1966). - Handbook of Physical
constants. The Geol. Society of America.
CLARKE, F.W. (1924). - The data of geochemistry.
Bull. U S . Geol. Surv., 770, 841 p.
CLARKE, F.W. & WASHINGTON, H.S. (1924). - The
composition of the Earth‘s Crust. U. S. Geol. Survey,
Profess. Paper 127.
CLAVIER, C. & RUST, D.H. (1976). - MID-Plot:A new
Lithology Technique. The Log Analyst, 17, 6, p. 16.
CRAIN, E.R. & ANDERSON, W.B. (1966). Quantitative log evaluation of the Prairie Evaporite
Formation in Saskatchewan. J. Can. Petrol. Techn., 5, 3,
p. 145-152.
DALY, R.A. (1933). - Igneous Rocks and the Depth of
the Earth. McGraw-Hill Book Co., New York.
DALY, R.A., MANGER, G.E., & CLARK, S.P. Jr.
(1966). - Density of rocks. In: Handbook of Physical
Constants.
DEER, W.A., HOWIE, R.A. & ZUSSMAN, J. (1962,
1976). - Rock-forming Minerals. Longman, London.
DEER, W.A., HOWIE, R.A. & ZUSSMAN, J. (1978). An introduction to the Rock-forming Minerals. Longman,
London.
DELFINER, P., PEYRET, O., & SERRA, 0. (1984). Automatic determination of lithology from well logs. 59th
ann. Techn. Conf. of SPE of AIME, Houston, paper SPE
13290.
DESBRANDES, R. (1985). - Encyclopedia of Well
Logging. Editions Technip, Paris.
DOVETON, J.H. (1994). - Geologic Log Analysis
Using Computer Methods. AAPG Computer Applications
in Geology, no 2.
DRAGOSET, M.A. (1977). - An Evaporite Analysis in
Southeastern New Mexico. Dresser Atlas internal Report.
DREYFUS, G. (1997). - Les reseaux de neuronas, une
technique operationnelle pour le traitement des donnees
industrielles, economiques et financieres. Mesures, no
699, Novembre.
DREYFUS, G. (1998). - Les reseaux de neurones.
Mecanique lndustrielle et Materiaux, no 51, Septembre.
DUNHAM, R.J. (1962). - Classification of Carbonate
Rocks according to Depositional Texture. Amer. assoc.
115
Petroleum Geol., Mem. 1, pp. 108-121.
EDMUNDSON, H. & RAYMER, L.L. (1979). Radioactive parameters for Common Minerals. SPWLA,
20th Ann. Log. Symp. Trans., paper 0 .
ELLIS, D., HOWARD, J., FLAUM, C., Mc KEON, D.,
SCOTT, H., SERRA, O., & SIMMONS, G. (1988). Mineral Logging Parameters: Nuclear and Acoustic. The
Thechnical Review, 36, 1, pp. 38-53.
EMMERMANN, R., LAUTERJUNG, J. & UMSONST,
T. (1993). - KTB Report 93-2: Contribution to the 6th Ann.
KTB Colloquium. Geoscientific Results, Stuttgart,
Germany.
FAUSETT, L. (1994). - Fundamentals of Neural
Networks. Prentice-Hall International Inc., New Jersey.
FELDER, B. & BOYELDIEU, C. (1979). - The
Lithodensity Log. SPWLA, 6th Europ. Symp. Trans., paper
0.
FISCHESSER, R. (1977). - Donnees des principales
especes minerales. Societe de I’lndustrie Minerale, SaintEtienne, France.
FOLK, R.L. (1959). - Practical Petrographic
Classification of Limestones. Bull. Amer. assoc.
Petroleum Geol., 43, 1, pp. 1-38.
FORD, W.E. (1955). - Dana’s Textbook of Mineralogy.
John Wiley & Sons, Inc., New York.
FORD, M.E., BAINS, A.J. & TARRON, R.D. (1974). Log analysis by linear programming. An application to the
exploration for salt cavity storage locations. SPWLA 3rd
Europ. Log. Symp. Trans., London, paper B.
FRAYSSINET, D., THIRIA, S., BADRAN, F. &
BRIQUEU, L. (2000). - Use of neural networks in log’s
data processing: prediction and rebuilding of lithologic
facies. Petrophysics meets Geophysics, Paris, 6-8 Nov.
GARRELS, R.M. & McKENSIE, F.T. (1971). - Evolution
of Sedimentary rocks. Norton, W. W. & Co, New York, 397
P.
GOTTLIB-ZEH, S. (2000). - Synthese des donnees
geologiques et petrophysiques acquises en forages
petroliers B /’aide de traitements statistiques et neuronaux. Applications aux depbts turbiditiques et aux
pla tes-formes mixtes. These, MontpelIier University, 265
P.
GOTTLIB-ZEH, S., BRIQUEU, L., & VEILLERETTE, A.
(1999). - lndexed Self-organizing Map: a new calibration
system for a geological interpretation of logs. In: Proc.
IAMG, pp. 183-188.
GOTTLIB-ZEH, S., BRIQUEU, L., WLODARCZYK, S.,
& MOSS, B. (2002). - Neural applications to Geological
and petrophysical data for improved flow-unit identification. SPWLA Europ. Symp. Transaction, paper T2C.
GRAU, J.A., SCHWEITZER, J.S., DRAXLER, J.K.,
GATTO, H. & LAUTERJUNG, J. (1993). - Elemental
Logging in the KTB Pilot Hole - 1. Nal based spectrometry. Nuclear Geophysics, 2, p. 19-73-187.
GRIM, R.E. (1968). - Clay Mineralogy. McGrawHill
Co., New York.
GROUT, F.F. (1932). - Petrography and Petrology.
McGraw-Hill Book Co., New York.
116
HAILE, P.M. & BLUNDEN, H.A. (1984). - Zechstein
Magnesium rich Evaporite Deposits of northern
Netherlands and their volumetric analysis by GLOBAL.
SAID-SPWLA, 9th International Log. Symp. Trans., Paris,
paper 37.
HASSAN, M. & HOSSIN, A. (1975). - Contribution a
/’etude des comportements du thorium et du potassium
dans les roches sedimentaires. Comptes rendus
Academie des Sciences, Paris, 280.
HEIER, K.S. & ADAMS, J.A.S. (1965). - Concentration
of radioactive elements in deep crustal materials.
Geochim. Cosmochim. Acta, 29, p. 53?62.
HENKEL, H. (1976). - Studies of density and magnetic properties of Rocks from Northern Sweden. Padeagh
214.
HERRON, M.M. (1986). - Mineralogy from
Geochemical Well Logging. Clays and Clays Minerals, 34,
2, pp. 204-213.
HOLMES, A. (1913). - The age of the earth. Harper &
Row, New York.
HOSSIN, A. (1969). - Interpretation des diagraphies en
zones carbonatees et evaporitiques. Bull. A. F. T. P., 193,
p. 47-69.
JUNG, J. (1958). - Precis de petrographie. Masson,
Paris.
KELLER, G.V., MURRAY, J.C. & TOWLE, G.H.
(1974).- Geophysical Logs from the Kilauea Geothermal
Research Drill Hole. SPWLA, 15th Ann. Log. Symp.
Trans., paper L.
KENNEDY, M.C. (2002). - Solutions to Some
Problems in the Analysis of Well Logs in Carbonate
Rocks. In: LOVELL, M., & PARKINSON, N. (ed.) (2002). Geological Applications of Well Logs. AAPG Methods in
Exploration, 604 p.
KERHERVE, J. (1977). - Notes prepared for an
advanced seminar on interpretation of well logs in volcanic rocks.
KHATCHIKIAN, A. (1982). - Log Evaluation of OilBearing lgneous Rocks. SPWLA, 23rd Ann. Log. Symp.
Trans., paper AA.
KLEIN, G. deVries (1980). - Sandstone Depositional
Models for Exploration for Fossils Fuels. 2nd ed. CEPCO
Div., Burgess Publishing Co., Champaign, Illinois.
KOHONEN, T. (1984). - Self-organization and associative memory. Springer-Verlag, New York, 255 p.
KOHONEN, T. (1992). - The Self-organizing Map.
Information Sciences, Springer-Verlag, New York., vol.
30, 312 p.
KRAUS, E.H., HUNT, W.F. & RAMSDELL, L.S. (1936).
- Mineralogy. An introduction to the study of Minerals and
Crystals. McGraw-Hill Book Co., New York.
KRUMBEIN, W.C. (1959). - The tetrahedron as a
facies mapping device. J. Sed. Petrol., 24 , 1, p. 3-19.
KRUMBEIN, W.C. & SLOSS, L.L. (1963). Stratigraphy and Sedimentation. 2nd ed. WH. Freeman &
Co, San Francisco.
KRYNINE, P.D. (1948). - The megascopic study and
field classification of sedimentary rocks. J. Geology, 56 ,
Serralog 0 2003
p. 130-165.
LARSEN, E.S. Jr. & PHAIR, G. (1954). - The distribution of uranium and thorium in igneous rocks. In: Nuclear
Geology. J. Wiley & Sons, New York.
LEET, L.Don, JUDSON, S. & KAUFFMAN, M.E. (1978,
1982). - Physical Geology. 5th ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
LEITH, C.K. & MEAD, W.J. (1951). - Metamorphic
rocks. Henry Holt & Co., New York.
LINK, P.K. (1982). - Basic Petroleum Geology. OGCl
Publications, TuIsa.
LOMBARD, A. (1956). - Geologie Sedimentaire. Les
series marines. Masson, Paris.
LOMBARD, A. (1972). - Series sedimentaires.
Genese-Evolution. Masson, Paris.
LOVELL, M.A., & PEZARD, P.A. (1990). - Electrical
properties of basalts from DSDP Hole 504B: a key to the
evaluation ofpore space morphology. In: Hurst, A., Lovell,
M.A., & Morton, A.C. (eds), Geological Applications of
Wireline Logs, Geological Society, Special Publication no
28, London.
LOVELL, M.A., HARVEY, P.K., JACKSON, P.D.,
BREWER, T.S., WILLIAMSON, G., & WILLIAMS, C.G.
(1998). - lnterpretation of core and log data - integration
orcalibration? In: Harvey, P.K., & Lovell, M.A. (eds), CoreLog Integration, Geol. SOC. London, Special publication
136, pp 39-52.
LOVELL, M., & PARKINSON, N. (ed.) (2002). Geological Applications of Well Logs. AAPG Methods in
Exploration, 604 p.
MASON, B. (1952). - Principles of Geochemistry.John
Wiley & Sons, Inc., New York
MATTHEWS, R.K. (1974). - Dynamic Stratigraphy.
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
MAYER, C. & SIBBIT, A. (1980). - GLOBAL, a new
Approach to Computer-processed Log Interpretation.
SPE of AIME, Fall Mtg., Dallas, SPE 9341.
MEAD, W.J. (1907). - Redistribution of elements in the
formation of sedimentary rocks. Jour. Geol., 15, pp. 238256.
MILLWARD, D., YOUNG, S.R.,
BEDDOESTEPHENS, B., PHILLIPS, E.R., & EVANS, C.J. (2002).
- Gamma-ray, Spectral Gamma-ray and Neutron-density
Logs for lnterpretation of Ordovician Volcanic Rocks,
West Cumbria, England. In: Lovell, M., & Parkinson, N.
(ed.) (2002). - Geological Applications of Well Logs.
AAPG Methods in Exploration, 604 p.
MOSS, B. (1997). - The partitioning of Petrophysical
Data: a Review. In: Lovell, M.A., & Harvey, P.K. (eds),
Developments in Petrophysics, Geol. SOC. London,
Special publication 122, pp 181-252.
PETTIJOHN, F.J. (1975). - Sedimentary Rocks. 3d ed.
Harper & Row, Publishers, New York.
PETTIJOHN, F.J., POTTER, P.E. & SIEVER, R.
(1972). - Sand and Sandstone. Springer, New York.
PICKETT, G.R. (1963). - Acoustic character logs and
their applications in formation evaluation. J. Petrol.
Technol., 15, 6.
Serralog 0 2003
PIRSSON, L.V. (1949). - Principles of Geochemistry.
John Wiley & Sons, Inc., New York.
PIRSSON, L.V & KNOPF, A. (1947). - Rocks and Rock
Minerals. 3rd ed. John Wiley & Sons, Inc., New York.
PRESS, F. (1966). - Seismic Velocities. In: Handbook
of Physical Constants.
PRESS, F. & SIEVER, R. (1978, 1982). - Earth. 2nd &
3rd ed. WH. Freeman & Co, San Francisco.
RAYMER, L.L. & BIGGS, W.P. (1963) . - Matrix characteristics defined by porosity computations. SPWLA, 4th
Ann. Log. Symp., Trans., paper 10.
RITCH, H.J. (1975). - A n open hole logging evaluation
in metamorphic rocks. SPWLA, Ann. Log. Symp. Trans.,
Paper V.
ROGERS, S.J., FANG, J.H., KARR, C.L. & STANLEY,
D.A. (1992). - Determination of Lithology from Well Logs
Using a Neural Network. Bull. Amer. Assoc. Petroleum
Geol., 76, 5, pp. 731-739.
Schlumberger Ltd (1974). - Log Interpretation. Volume
/I - Applications.
Schlumberger Technical Services, Inc. (1982). Essentials of NGS Interpretation.
SANYAL, S.K., WELLS, L.E. & BICKHAM, R.E.
(1980). - Geothermal Well Log Interpretation. State of the
Art. final Report. Los Alamos Scientific Laboratory, LA8211-MS.
SCHOLE, P.A. & SPEARING, D. (Ed.) (1982). Sandstone Depositional Environments. Amer. Assoc.
Petroleum Geol., Mem. 31.
SELLEY, R.C. (1978). - Ancient Sedimentary
,Environments.2d ed. Chapman & Hall, London.
SELLEY, R.C. (1976). - An lntroduction to
Sedimentology.Academic Press, London.
SERRA, 0. (1980). - Aspects diagraphiques des evaporites. Bull. Cent. Rech. Expl. Prod. Elf-Aquitaine, Pau,
4, 1, p. 411-431.
SERRA, 0. (1986). - fundamental of well log interpretation. 2 - lnterpretation of logging data. Elsevier,
Developments in Petroleum Sciences 15B.
SERRA, 0. (1990). - Element, Mineral, Rock Catalog.
Schlumberger.
SERRA, 0. (1993). - Relations entre les donnees diagraphiques et les parametres geologiques. Les differents
domaines d’applications geologiques des diagraphies.
Geologues, 99.
SERRA, O., DELFINER, P. & LEVERT, J.C. (1985).Lithology determination from well logs: case studies.
SPWLA, 26th Ann. Log. Symp. Trans., Paper WW.
SHAND, S.J. (1951). - Eruptive rocks. 4th ed. John
Wiley & Sons, Inc., New York.
SMITHSON, S.B. (1971). - Density of metamorphic
rocks. Geophysics, 36, 4.
STUNDNER, M. & LECHNER, J. (2000). - Delineation
of Lithofacies based on Well Logs - Application of
Stochastic and Neural Network Models for Analyzing and
Interpretation. 1st Symposium on Well Log Analysis &
Formation Evaluation, 29-31 Oct., Tripoli, Libya, paper
HI1.
117
SULEIMAN, O.E. & KELEMEN, Z. (2000). - The Role
of Principal Component Analysis (PCA) in the Delineation
of Lithofacies based on Well Logs. 1st Symposium on Well
Log Analysis & Formation Evaluation, 29-31 Oct., Tripoli,
Libya, paper H/2.
SUPERNAW, I.R., McCOY, A.D. & LINK,A.J. (1978). Method for in-situ evaluation of the source rock potential
of each formation. U.S. Patent 4,071,744, Jan. 31.
TUREKIAN, K.K. & WEDEPOHL, K.H. (1959). Distribution of the elements in some major units of the
earth’s crust. Geol. SOC.Am. Bull., 72, pp. 175-192.
TURNER, F.J. & VERHOOGEN, J. (1951). - Igneous
and Metamorphic Petrology. McGraw-Hill Book & Co.
Inc., New York.
WAHLSTROM, E.E. (1947). - Igneous Minerals and
Rocks. John Wiley & Sons, Inc., New York.
WEAST, R.C. ed. (1981-82). - Handbook of Chemistry
and Physics. 62nd ed. CRC Press Inc.
WEDEPOHL, K.H (1969). - Handbook of
Geochemistry.
WESTPHAL, H & BORNHOLDT, S. (1996). Lithofacies prediction from wireline logs with genetic algorithms and neural networks. E. Schweizerbart’sche
Verlagsbuchhandlung, pp. 465-474.
WEST, F.G., & LAUGHLIN, A.W. (1976). - Spectral
Gamma Logging in crystalline basement rocks. Geology,
VOI. 4, pp. 617-618.
WHITFIELD, J.M., ROGERS, J.J.W. &ADAMS, J.A.S.
(1959). - The relationship between the Petrology and the
Thorium and Uranium contents of some granitic rocks.
Geochim. & Cosmochim. Acta, 17.
WICKMAN, F.E. (1954). - The “total” amount of sediment and the composition of the “average igneous rocks”.
Geochim. Cosmochim. Acta, 5, pp. 97-110.
WILSON, J.L. (1975). - Carbonate Facies in Geologic
History. Springer, New York.
ZHOU, C.D., WU, X.L. & CHENG, J.A. (1993). Determining reservoir properties in reservoir studies using
a fuzzy neural network. SPE Annual Technical
Conference and Exhibition, Houston, Texas, paper SPE
26430, pp. 141-150.
118
Serralog 0 2003
WELL LOGGING AND ROCK TEXTURE
Int roduction
Review of petrographic concepts
Before explaining how to extract textural information
from well logs it is important to review some concepts
which will help in the understanding of the links between
well logging data and textural parameters.
Definition
Texture is " the general physical appearance or character of a rock, including the geometric aspects of, and
the mutual relations among, its component particles or
crystals; e.9. the size, shape and arrangement of the
constituent elements of a sedimentary rock, or the crystallinity, granularity and fabric of the constituent elements
of an igneous rock. The term is applied to the smaller
ROCK
COMPONENTS
COMPOSITION
Figure 3-1
Serralog 0 2003
(megascopic or microscopic) features as seen on a
smooth surface of a homogeneous rock or mineral aggregate" (Bates & Jackson, 1980).
So, texture covers the geometrical aspects of the
constituent components of rocks : grains or particles and
crystals: their size, shape, appearance, their arrangements and sorting; also the grain-grain, grain-matrix or
grain-cement boundings.
Texture plays a very important part in sedimentary
rocks, because the petrophysical properties of a rock,
hence its porosity and permeability, depend essentially on
texture.
According to the origin of the rocks, Krumbein & Sloss
(1963) distinguish:
- fragmental texture, more specific to detrital rocks;
- crystalline texture, more specific to chemical or eruptive rocks.
- The components of texture.
119
Textural components
Regardless of the type of texture, sedimentary rocks
are marked by the textural characteristics of four principal
components (Fig. 3-1):
grains, or particles, and crystals;
matrix, that corresponds to the finest-grained
material enclosing, or filling the interstices or space between the larger grains, particles or crystals of a sediment.
It is deposited at the same time than the other elements;
- cement, binding the grains and the matrix;
porosity, or void, generally filled by fluid.
These components may have different mineralogical
compositions. Figure 3-1 shows the constituent minerals
of the different textural components, depending on the
type of detrital rocks.
These components are not necessarily present all
together in the same rock. For example in medium-grained and very well-sorted sand there will be neither matrix
nor cement. A consolidated, medium-grained, well-sorted
sandstone has no matrix, but contains a cement that can
be siliceous, calcitic or dolomitic, sideritic, or halitic.
The grains and matrix are generally associated,
because they are most often deposited together (sand
grains in an argillaceous matrix, or pebbles in a sandy or
shaly matrix). The cement, being the result of chemical
precipitation in the pore space, is always post-depositional.
As shown by Krumbein & Sloss (1963), the study of
texture must be subdivided into three categories depending on the nature of sedimentary rocks:
- purely chemical rocks, such as halite, gypsum, anhydrite... The texture of these rocks is characterized only by
the crystalline system, the size and imbrication of crystals;
- partly chemical or biochemical and partly detrital
rocks, such as carbonates, for which the three textural
components play, in turn or simultaneously, a very important part;
- detrital rocks, for which the three components play a
very important part.
ties
- grain size and its variation that will control the sorting;
- shape (or sphericity) of grains (Fig. 3-2);
- roundness;
- surface texture;
- grain orientation or fabric;
- mineralogical composition. This is not, properly spea-
king, a textural parameter, but depends on:
- density, hence the rate of sedimentation of each
cornponent ,
- possibility of dissolution (solubility) or of alteration, consequently the ulterior formation of vugs
or cement,
- wettability of rock.
Grain size
This parameter is very important as it is used to subdivide detrital sedimentary rocks into conglomerates,
sandstones, siltstones and claystones (Fig. 3-2). It is also
a guide to the proximity of the source area: coarser sediments are generally closer to the source, the grain size
decreasing with the transport distance due to the breaking
and abrasion of the particles during the transport (Fig. 33). Also, grain size is related to the dynamic conditions of
transportation and deposition.
Combined with the sorting of grain size, it indicates the
competence and efficiency of the transporting agent. The
agents (water, wind, ice) and modes of transport (traction,
saltation, suspension) differ materially in their sorting and
transporting ability.
Detrital rocks will be discussed first.
Texture of detrital rocks
According to Krumbein & Sloss (1963), the final texture of a sediment is influenced by six fundamental proper-
MILES FROM SOURCE
Figure 3-3 - Variation of beach sand size and sorting with distance
from source (from Krumbein & Sloss, 1963).
J
Figure 3-2 - Common terms used to describe grains as a function of their size either in mm or the Wentwodh-Lane phi scale.
120
Serralog 0 2003
Sorting
It characterizes “the dynamic process by which sedimentary particles having some particular characteristics
(such as similarity of size, shape, or specific gravity) are
naturally selected and separated from associated but dissimilar particles by the agents of transportation” (Bates &
Jackson, 1980).
A
Figure 3-4 - Visual aspect of sorting
(adapted from Trask, 1932 and Folk, 1965; in Pettijohn et al., 1972)).
A sorting coefficient is sometimes computed as an indicator of grain size variation. It corresponds to the ratio of
the maximum grain diameter to the ‘minimum grain diameter. This ratio is equal to l if all the spheres or grains
have the same diameter; it is higher if the grain size
varies. The sorting is also expressed through the standards deviation of grain size distribution (Fig. 3-4).
J
Figure 3-5 - Histograms of grain size distribution corresponding to
different types of environments
(adapted from Lung, 1927; Pettijohn, 1975; Wentworth, 1931; Krumbein
& Griffiths, 1938; Krumbein & Tisdel, 1940; Rosenfeld).
Grain size distribution is well described by histograms
(Fig. 3 4 , or cumulative percentage curve (Fig. 3-6), that
reflects partly the depositional environment as illustrated
by Figures 3-5 to 3-7. Histograms present a factual picture of the grain size abundance in each grade size, however they cannot allow the determination of the average
particle size and other properties that cumulative curve
allows (Fig. 3-7).
A
_I
Figure 3-6 - Cumulative curve of grain size with explanation of percentiles 10 and 90, median and lrst and 3d quartiles
( h n Kmmbein 8 Slow 1963).
Serralog 0 2003
Figure 3-7 - Grain size and sorting as a function of the depositional
environment.
121
Well Logging and Geology
122
Grain shape
The geometric form of a particle involves several separate but related geometric concepts.
- Measurement of three shape factors describing its
geometry (Fig. 3-8);
- Measurement of the sphericity and roudness of the
particle (Fig. 3-9).
Figure 3- 10 - Geometrical description of roundness of two different particles (from Krumbein, 1940).
The three previous grain parameters evolve with grain
transport as illustrated by Figure 3-11.
J
Figure 3-8 - Measurement of the three axes of a particle and its classification according to its shape (from Krumbein, 1941 and Zingg, 1935).
Shape
Size
Sorting
ROUNDNESS
Length of transport increases in that direction
-
Figure 3-9 Classification of grain shape as a function of their
sphericity and roundness
(from Krumbein & Sloss, 1963, fig. 4-10).
Sphericity expresses a relation between the surface
area of a particle and its corresponding sphere. It can be
expressed as the cube root of cVab (Sneed &Folk, 1958).
Roundness is expressed as the ratio of : (average
radius of corners and edges)/(radius of maximum inscribed circle) (Wadell, 1932). Because this is difficult to
apply, it is more convenient to work with a two-dimensional figure (Fig. 3-10) (projected image of the particle). In
this case the roundness is defined as the average radius
of curvature of the corners of the particle image divided by
the radius of the maximum inscribed circle. It is expressed
as :
p = Z(q/R)/N
where ri are the individual radii of the corners, N the
number of corners, and R the radius of the maximum
inscribed circle (Fig. 3-10).
122
Figure 3-11 - Modification of the shape, size and sorting with the distance of transport.
Packing
According to Graton & Fraser (1933, simple geometrical packing of equal-sized spheres are made in six different manners (Fig. 3-12). They proved that porosity varies
according to packing from 47.64 % for the most "open"
arrangement to 25.95 % for the most compact or "closed".
Allen (1984) made a complete review of several types
of packing (ordered, random or haphazard) of particles of
different shapes (spheres, prolate and oblate spheroids).
He concluded that "regular particles ... may form packing
of all three kinds, whereas natural particles, which are
irregular, can only form packings of random or haphazard
kinds." In fact, regular particles can exist in nature. They
correspond to oolites.
Semlog 0 2003
Gravity
a)
Gravity + current
Figure 3-12 - The six possible geometric arrangements of equal size
spheres (from Graton & Fraser, 1935).
-
Figure 3-14 Orientation of flakes. a) Flakes deposited under gravity
action. b) Flakes deposited under action of gravity and cunent
(from Potter & Pettijohn, 1977, fig. 3-2).
Orientation
Influence of arain size
The orientation of particles is defined by reference to a
horizontal plane and to the direction of current.
The orientation of pebbles is generally well defined,
because their size makes observation relatively easy (Fig.
3-13). The measurement and quantification of the orientation of small sized grains (sand, silt...) is much more difficult. However, for non-spherical grains it is generally
observed that the orientation of grains is the same as the
orientation of their axis of maximum elongation and is
parallel to the direction of current (Fig. 3-14)..
-
Figure 3-13 Schematic representation of the pebble orientation in different environments (from Rukhin, 1958).
Porosity is theoretically independent of grain size. An
arrangement of spheres with uniform size, which present
the same organization, will have the same porosity,
regardless of size. This ideal situation, which corresponds
to a maximum sorting, rarely occurs in nature. It can
sometimes be observed in washed or winnowed sands,
and more frequently in oolitic sands. Figure 3-15, from
Dodge et a/. (1971), seems to confirm that, above a certain level of porosity reflecting the best sorting and the
absence of cement, the porosity is more or less independent of grain size.
P
8
is
n
These different textural parameters affect reservoir
characteristics.
Influence of grain properties
Porosity and permeability, which are the main petrophysical characteristics of reservoir, depend on several
factors.
Beard and Weyl (1973) showed that the primary porosity and permeability of a detrital sediment, which has just
been
depend On five
size?
shape, roundness, orientation and arrangement of grains.
serralog Q 2003
Mean grain size (@ unit)
Figure 3-15 Relationship beheen porosity and mean grain size for a//
samples coming from sandstones of Paluxy Formation, Texas
(from Dodge et al., 1971).
123
124
Well Logging and Geology
In fact, Lee (1919), Von Engelhardt (1960) (Fig. 3-16)
discussing ancient sedimentary rocks, and Rogers &
Head (1961) (Fig. 3-17), and Pryor (1973), on the subject
of contemporary sands, show that porosity decreases
slightly, when grain size increases. This evolution is probably due to a number of factors which have only an indirect connection to grain size. Finer sands have a tendency to be more angular and are likely to organize according
to a less dense arrangement. Thus, they present a higher
porosity than sands with coarser grains.
matrix invades the large pores and fills up the big canals.
The combined influence of grain size and sorting on
porosity and permeability has been studied by Beard &
Weyl (1973). It is illustrated by Fig. 3-20.
Mean grain size ($ unit)
Mean diameter (Fm)
Figure 3-16 - Relationship between porosity and mean diameter of
grains in Bentheimer sandstones (from von Engelhardt, 1960).
Figure 3-18a - Relationship between permeability and mean grain size
for all the samples coming from sandstones of Paluxy Formation,
Texas (from Dodge et at, 1971).
r
5b
z
Q
a
Porosity (%)
Figure 3-17 - Relationship between porosity and mean diameter of
sand grains for several different sorting coefficients
A:So=2.086;B:So= 1.625;C:So=1.279;D:So=l.128;E:So = 1.061
(from Rogers & Head, 1961).
Contrarily, as demonstrated by Dodge et a/., (1971)
(Fig. 3-18), in a well-calibrated sand, permeability increases when the size of the grain increases. This is easily
understood because the pore size and the canals
(throats) which connect the pores to one another are
governed by grain size: the smaller the grains, the smaller the pores and the section of the canals will be. Thus,
capillary attraction will be stronger and permeabilitywill be
less.
Influence of sortinq
As investigated by Rogers & Head (1961) porosity and
permeability increase when sorting increases (Fig. 3-19).
In fact, in a poorly-sorted sand, the small grains (matrix)
are set in interstices left by the coarser grains. Thus, the
124
Figure 3-18b - Relationship Figure 3-18c - Influence of grain size on
between Wrmeabil#Yand
rock properties; surface area, throat
grain size Even if the rela- radius, tortuos@, Archie’s m factoc pertionship is not perfect, as
meability and irreducible water
expected, when the grain
(courtesy of Schlumberger).
size increases the permeability increases
(from von Engelhad,
1960).
Serralog 0 2003
Texture
E
E
v
E'
.-
POAOSITY, %
Figure 3-19 - Relationship between porosity and sorting coefficient of
sands for different grain sizes. A :median diameter = 0.106 mm;
B := 0.151 mm; C := 0.213 mm; D = 0.335 mm
(from Rogers & Head, 1961).
E
cu
g1
kj
n1
Influence of shape and roundness of arains
It seems that shape and roundness affect intergranular porosity. Fraser (1935) came to the conclusion that
sediments composed of spherical grains have lower porosity than those formed by grains with less sphericity He
attributed this to the fact that in the first type the grains
tend to settle according to a denser arrangement than in
the second type. Less spherical grains can pack together
in a way that creates wider volumes between them.
I.
The influence of shape and roundness on permeability is not still well known, but we observe that permeability
follows the fluctuations of porosity in connection with
variations in shape and roundness.
Porosity (%)
Figure 3-20 - Relationship between porosity and permeability for different grain size and sorting (from Beard & Weyl, 1973).
Influence of the orientation of arains
On the whole, the orientation of non-phyllitic (nonshaly) grains has no influence on porosity. On the contrary, it has a very strong influence on permeability or more
precisely on anisotropy or the direction of highest permeability.
Thus, in channel sands, the direction of maximum permeability is parallel to the axis of the elongation of sand
bodies.
In a littoral sand bar, maximum permeability is perpendicular to the axis of elongation of sand bodies, but is
parallel to the dominant direction of currents (Fig. 3-21).
The orientation of phyllitic particles (shales) will be the
same as the orientation of their large sides that are parallel to the plane of stratification (Fig. 3-14).
Serralog 0 2003
_I
Figure 3-21 - Grain orientation and directions of maximum permeability
in channel and bar sand bodies, upper and lower respectively
(from Pwor, 1973, in Selley, 1976, fig. 15).
Influence of packing
Allen (1984) showed that porosity is decreased by a
"widening of the range of particle sizes present in mixture", but is increased "by an increase of particle angularity
and surface mundness, and by the inclf.Jsionof strongly
anisometricparticles".
It seems obvious that permeability must follow a combecause the section Of pores and
parable
capillaries in compact arrangement is smaller than in
others. Figure 3-22 shows the theoretical variations in
permeability for ideal geometric packings of 500 pm-diameter spheres.
125
Next Page
126
I
Well Logging and Geology
Porosity (%)
I
Figure 3-22 - Porosity vs permeability for some ideal geometric packings of 500 pm-diameter spheres
(courtesy of R. Nurmi).
Naturally, the most "open" arrangements do not, in
fact, ever occur in ancient formations; very soon under the
influence of compaction the grains are organized according to the most "close" arrangement. The influence of
compaction will be studied later as it also affects the form
of grains.
Influence of the mineraloaical comDosition of arains
Grains composed of heavy and denser minerals will be
deposited with the minerals of the same weight, i.e. with
less density, but with bigger size.
This situation will lead to a decreasing of sorting, therefore of porosity and permeability.
Grains composed of unstable or chemically immature
minerals (pyroxene, amphibole, mica, feldspar,...) will
influence porosity of the sediments in which they occur.
Their ulterior alteration will cause the formation of authigenic clay minerals (kaolinite, montmorillonite, illite, chlorite,...) which will surround the grains or invade the pore
space, thus, causing major reductions in porosity and permeability (Fig. 3-23). But the type and distribution of authigenic clay minerals must also be considered because
they affect the permeability in a different way (Fig. 3-24).
Their influence has been studied by Neasham (1977).
Discrefe-particule or pore-filling clay minerals (for
example kaolinite "books") are characterized by uneven
distribution in the pore space. Crystals can reach large
size (more than 10 pm). As crystals grow, pore space and
permeability decrease, although there is always a small
amount of microporosity between individual plates.
Pore-lining clay minerals, essentially illite, chlorite and
montmorillonite, coat the pore walls with a thin layer of flakes that are parallel or perpendicular to the pore wall (Fig.
3-23b), but the growth does not reach far into the pore
space. A large amount of microporosity can be present
between flakes, although these pores are less than
micron in diameter. This type of authigenic clay greatly
reduces permeability and also influences many of the rock
electrical properties because it can considerably increases the surface area.
Pore-bridging clays, fundamentally illite fibers (Fig. 323c), are connected across the pore space. This type causes major reductions in permeability, since bridging is
126
-
Figure 3-23 The three types and distributions of dispersed authigenic
clay minerals in pore space. (a): pore filling; (b): pore lining; (c): pore
bridging (fmm Neasham, 1977).
most easily attained across the throats, and also it reduces the size of the pores. Porosity is less affected because microporosity is preserved between the very fine
fibers.
From the previous remarks, it seems obvious that the
knowledge of the type of distribution (laminated, structural
or dispersed), and of the nature of the clay minerals is of
the utmost importance to predict the permeability range
and the existence and distribution of permeability barriers.
Particles composed of soluble minerals (calcite, dolomite,...) lead to the creation of a secondary porosity by
their ulterior dissolution and the elimination of the solution
by hydrodynamism. They can also obliterate porosity and
consequently permeability by solution and formation of
cement by precipitation of new crystals in the pore space.
The diagenetic influences will be analyzed later
Influence of other textural comDonents on reservoir
characteristics
The other textural components, matrix and cement
also have an important influence on the petrophysical
characteristics of detrital reservoirs.
When the percentage of matrix and/or cement increases, the porosity and permeability diminish, since the fine
particles, that make up the matrix, and the cement tend to
occupy the pore space between the coarser elements.
Serralog 0 2003
Previous Page
Texture
I Chapter 3 I 127
Texture of carbonate rocks
!
Discrete Particle
0
0
The texture of carbonate rocks depends on the relative percentage of three components (particles, matrix and
cement), and on the type of pore distribution (Fig. 3-25).
Carbonate rocks have different components described
in Table 3-2.
Table 3-2
The textural components of carbonate rocks.
Pore Lining
bhodula LL (- rxtndw)
grain8 of non carbonate mmcirb: ex : quam
grains, c h ~ anhydritr,
.
phphaes,...
Imnclac;g
frrpnwmt of morM carbonrte rocka which
originated within thr drpo*riwl basin.
S l u l ~ a frsgmmt
l
(- biodrotr)
0
0
Palticlea
or
6
grains
Lumpr
or
(wou.
Allocherm
1
1 w
Porosity, percent
-
Poloid8 00.t.d by ab8l laminae.
Qdm
Phorth8,
grains
P s r d i s form in caverns (cavepearls) or in
CdiiChO CNStS bOl9OMh W&lOlOd ZOnOS
The matrix is really an important component only in
conglomerates, whereas in sands it is present only in very
poorly sorted sand bodies.
(Vadoss pirditht)
Algal encrusrod grains (- oncolii)
oncoliihs am famed by primitin Muo-gmn
algae (schizophycre or Qanophycae)
growing on a grain a d mncting urbonru
mud to their stickv sudaw.
Cement is developed after deposition either by chemical interaction between unstable grains and formation
water, or by circulation in the pore space of solutions
under hydrodynamic forces.
1.
algal lumps
Nucleus (of quam or skeletal grain) with
secretion of calcite (or silica or hematite)
with conantrio or radial struchm. am
grown in susp.mion in an agitating medium
~
(high O I I O ~ Oenvironment)
Figure 3-24 Influences on porosity and permeability of the distribution
type of authigenic clay minerals. Data from Neasham's study of 14
sandstones (from Neasham, 1977).
These different influences are synthesized in Table 3-
I
Composite arains (- gnpestone)
P.(oidsbondedbymicrite
&&Q
:carbonate mud (maxi. c w l size :
10 IM)
-
my k enmmtomd in mallquantities within
a grain-rupportrdcarbonate
mykrorkmbntmnitformraca*rt.
mudrock tanned :
micrit.oruloiknitr
Otthochorni
(FOLK, 1968
Table 3-1
Influence of textural parameters on porosity and permeability of detrital rocks.
ddolniaolpwitr
Dokmite
ddorprritr
Po-
These different parameters are used to classify the
carbonate rocks (Folk's classification). But in this chapter
the Dunham's classification will be adopted (Fig. 3-25).
Table 3-3 from Choquette and Pray (1970)illustrates
the different types of porosity in carbonate rocks, and links
them with original phenomenon and with time of pore formation.
Senlog Q 2003
127
DEPOSITIONALTEXTURE RECOGNIZABLE
Table 3-4
Comparison between porosity in sandstones and limestones (from Choquette & Pray, 1970).
Sandstone
Amount ofprinury porosity
Commonly 2540%
Cartmuart
Commonly 40-700/,
in wdimrnlr
Amount of
Commonly
Commonly none or only small
ultimate pros- half or more of fraction of initial porosity; 5ity in rocks
initialporosity; ISTO common in raavoir
IS-30%
mon
kiyure 3-25 - The Dunham's classification of carbonate rocks based on
com- facia
depositional texture.
It is obvious that porosity and permeability depend on
texture as illustrated by Fig. 3-26 which shows their
dependence on the texture of the carbonate rocks.
But, in these rocks we have to distinguish the original
characteristics, i.e. those existing at the time of deposition, from present characteristics. In fact, the original texture may have been deeply modified as a result of diagenetic phenomena that are often precocious and more
important in these formations than in detrital series.
Consequently, these modifications in texture bring about a
change in the reservoir characteristics themselves.
This is why these characteristics arise more from diagenetic phenomena than from texture, and why the study
of porosity is so essential.
Table 3-3
Different types of porosity in carbonate rocks
(adapted from Choquette & Pray, 1970).
rype(r) of pri- Almost exdumry porosity d w l y interpanide
interpanick commonly prcdominates, but intrapankle
and other t y p a arc imporrant
Typa(r) of ulti- A l W udu- Widdy varied bouur af portmate porosity d v d y primary depositio~lmodifiutionr
interpanick
Sim of'pores
Diameter and Dumer and throat rimcanthroat rim
d
y show littk relation to
closely d 8 t 4 aalimcnury panMc Mac or
t o d m t n c . r y rortin#
panick rim
and m
i
n
g
Shape ofpomg S t r o y &pen- G m t l y varicd. nrya from
darc on par- a r w y dcpmd+nt "paitiw"
tick s h a m or "nqativc" af pnicks to
"nytivc"
of form canpkcdy i
w
panida
of *pa
of dcporlunul or
dLpncciccmponau
Variable. rangiq from fairly
Uniformity of C a n m o d y
dzc.rhpe,and fairly uniform uniform to c x t r m d y hetwo-
distribution
I n f l u e m of
diapcncrir
within homo- gemus,
even within body
rock type
of a*
p n a body
~
nude up
Minor: u w l l y
minor reductionofprimary
porosity. by
Major: a n auk oblilcncc
or c a n p k t d y modify porority;
cmwnution and rdtuion imporunt
compaalon
and cementation
TIME
TYPE
lnflucna of
fracturins
Generally not Of major imponancc in maexof nu* im- voir propertia if proms
poruncc in
rcauvoir propCrtiCS
128
Variable. mniquantiutive vtrual a c i m a t a ransc from e u y
to vinurlly impossible; inslrumcnt mearunmcntr of poroc
ity. permeability and capillary
prruurccommonly nscdcd
Visual evaluation of porosity
and permcability
Semiquantiutivc visual minutes mmonly rd.tivdy easy
Adequacy of
core analysis
for rcrcrvoir
evaluation
Core plugs of Core plugs commonly itude-
Permeabilityporosity inter.
reliltions
Relatively con- Grcacly varicd; commonly insistent; com- dependent of panide size and
monly drpcn- w i n s
dent pn panicle
size and wnins
I-in. &ma quPre;mnw hok cora(-3-in.
commonlyadr- &meter) n u y be inadequate
quate for "nu- for large pom
trix" porosity
Serralog 0 2003
Texture
I Chanter 3 I
129
determination of size of grains or pores as small as a
micron.
-
Fig. 3-28 Thin section of a compact sandstone.
-
-
Figure 3-26 Porosity permeability relationship as a function of
the carbonate texture and its modification by cementation, dissolution
and fracturation (fmm Nurmi, 1986).
Determination of textural parameters
Traditional approach
Geologists determine the textural parameters through
different methods summarized in Figure 3-27.
Grain diameter in millimeters(kgarithmic scale)
-
Figure 3-27 Grain-size analysis methods
(adapted fmm KNmbein 8, Sloss, 1963).
The most common method for the determinationof the
grain size is sieving. This method is essentially used to
determine the grain size distribution or the sorting.
If the grains are large enough, direct measurements
are possible. In that case the measurement is realized following three perpendicular directions, called intercepts
(Fig. 3-8):longest dimension (a), intermediate (b) and
shortest dimension (c).
For smaller particles, their size can be also estimated
from their settling velocity which varies as the square of
the particle diameter (Stokes’s law), but also as a function
of their density.
For very small grains, examination of thin sections
through microscope (Fig. 3-28) or scanning electron
microscope (Fig. 3-29)are used. The latter allows the
Serralog Q 2003
-
Fig. 3-29 SEM photograph showing the grain size, the sorting, the
presence of authigenic clays coating grains or filling partly the pore
space.
Staining of thin section by Alizarin Red S (Fig. 3-30)
allows the distinction between calcite and dolomite grains
or cement. This technique is fast, easy and cheap.
Through thin sections or SEM photographs many textural as well as mineralogical parameters can be determined (shape of grains, grain overgrowth, sorting and packing, orientation... Fig. 3-31).This study can be completed
by energy dispersive analysis (EDAX) or electron microprobe analysis.
By injection of colored fluid (epoxy) the pore system
can be visualized (Fig. 3-32)and analyzed.
Capillary pressure measurementscan also be used for
evaluation of pore size and consequently, grain size.
Cathodoluminescencetechnique is also used to obtain
129
130
Well Logging and Geology
information on the spatial distribution of trace elements in
clastic grains and cements (Fig. 3-33). The differences in
luminescence is a function of subtle differences in the
trace elements composition.
In carbonate rocks, particle size and nature, cement
mineralogy (calcite, dolomite...) and type (micritic or sparitic) can be well described through thin sections and SEM
photographs (Figs. 3-34 to 3-36).
-
Figure 3-32 The porosity of this mature sandstone is well shown by
the epoxy (in green).
J
Figure 3-30 - Staining of a thin section allowing distinction between
calcite (stained in red) and dolomite remaining
unaffected (photo 0.06 mm large).
Figure 3-31 - (a) On this thin section quartz overgrowth is well detected. The original quartz grain, well rounded, is emphasized by dark
dots or “dust rims”. The overgrowth is marked by the euhedral crystal
shape around this grain. Calcite cement is also present (orange crystals). Pore space is in blue.
(b) This SEM photograph shows very clearly the quartz overgrowth
characterized by the euhedral crystal shape.
-
PI
Figure 3-33 On the left, a transmitted light through a thin section
shows a sutured angular quartz arenite which can be interpreted as
due to compaction. On the right, the photo, taken with cathodoluminescence, shows quartz overgrowth of well rounded quartz grains (in red
or light blue).
9
Figure 3-34 a): Thin section showing intra- and inter-granular porosity
(in blue epoxy) in an oolitic limestone. Micrite cement is around oolites.
Some sparitic cement (in white) is also present. b): SEM photo of oolitic sand. c): SEM photo of chalk. d): Thin section showing intergranular
porosity (in black) in an oolitic limestone.
-
130
Senalog Q 2003
Texture
I Chapter 3 I
d
Figure 3-35 - Thin section showing a very high porosity with large pore
throats indicating a high permeability in this carbonate.
Determination of textural parameters
Well logging approach
The influence of textural components on log parameters is summarized in Table 3-5.
Thus, it clearly appears that some log measurements
contain within themselves textural information. But, it is
131
3
-
Figure 3-36 SEM photograph of a dolostone. Shape and size of
crystals and pore size are easily determined.
not always simple to determine the origin of influence
because several textural parameters may have similar
effects. However, if we reexamine the problem, according
to the type of rock, it is sometimes possible to perceive
the privileged influence of certain parameters.
In order to extract information on the textural parameters themselves, tools of high vertical resolution are nee-
Table 3-5
Relationship between textural parameters and well logging attributes (from Serra, 1984).
Textural Parameters
fsiie
Petrophysical reservoir
characteristics depending
on textural parameters
Well logging
parameters
affected
7
Relevant
Logging tools
IL. AIT, LL. ARI, SFL, ML, MLL, MSFL. FMSFMI
FDC, LOT
CNL
Sonic, BHTV
TDT, GST, GLT
EPT
LOT
FMMl
FMSIFMI, EPT, Sonic
EFT, MSFL
IL AlT, LL. ARI, SFL, ML MLL, MSFL, W F M I
Tortuosity
Sonic
EPT
Sizeof pores and throats
which ccnbds:
IL, AIT, LL, ARI, SFL. ML. MLL MSFL. FMS/FMl
LDT, CNL, Sonic,EPT, TDT, GST
Total penne&lity
Relative permeability
lneduable saturabbn
CMR
IL, AIT, LL. ARI, SFL, ML, MLL, MSFL
-Anisolrow
Horizontal pemeability
Veldical permeability
Weaabilily
A
F W M I , UBI
n. (%)in
b
1 :geologist's definition
Serralog Q 2003
131
132
Well Logging and Geology
ded as the geological textural objects (grains, particles,
crystals) have a small size.
With standard tools, quantitative well log information is
acquired along one generatrix of the borehole (microlog,
microlaterolog, microsphericallyfocused log, litho-density,
electromagnetic propagation tool) or volumetrically
around the borehole (spontaneous potential, resistivity,
conductivity, radioactivity, neutron, sonic). The recorded
measurements by the standard tools correspond to an
averaged measurementthat views the formation as apparently homogeneous. They are essentially affected by
grain and matrix composition but also by porosity and permeability which depend on the textural parameters. It is
the reason why it was possible to extract information on
grain size evolution with depth by looking at ramps on the
SP, gamma ray and other logs (Fig. 3-37), or information
on sorting by taking into account the porosity level : high
porosity in a sand indicates a well sorted sand without
cement (Fig. 3-38). But, due to their lack of vertical resolution, standard logs will not inform directly on textural
parameters. In fact, the effects of those parameters on
porosity and permeability fundamentally affect the tool
responses.
1
-
Figure 3-38 Change of sorting detected from porosity change.
With dipmeter tools, which analyze along 4, 6, or 8
generatrices (Figs. 3-39 & 3-40), or with density which
analyze in 4 sectors (Fig. 3-43d), and more especially
with imaging technology, which analyze formations following 12 sectors'(AR1or HALS, Fig. 3-43b), or along a lot
of generatrices (FMI Formation Micro Imager, Formation
Microscanner FMS, STAR and EM1 tools Figs. 3-41),
even during drilling (RAB tool, Fig. 3-43a), or all around
the borehole wall (UBI [Fig. 3-43~1,CBIL, CAST), it is possible to detect heterogeneities either radially out into the
formation or circumferentially around the wellbore. The
size of the electrodes (1 cm for SHDT or 0.5 cm for FMI)
or of the ultrasonic beam allows the detection of very
small objects.
J
Figure 3-39 - Schematic of the SHOT-B sonde
(courtesy of Schlumberger).
MQI
Figure 3-37 -Grain size evolution detected from gamma ray log
(from Sera 8 Sulpice, 1975).
I32
d
Figure 3-40 - Photographs of dipmeter tools.
Serralog @ 2003
Texture
C
I
Figure 3-41 - Photographs of imagery tools.
I Chapter 3 I 133
d
-
Figure 3-43 c): Schematic of the UBI (Ultrasonic Borehole Imager)
tool; d): Schematic of the ADN fool (courtesy of Schlumberger).
The determination of the textural parameters from
standard and high resolution logging tools will be illustrated now.
PAD
Detrital rock texture
Grain Size
There is no general universal relationship between the
grain size and standard log measurement. Nevertheless,
in a number of cases we often observe, regionally, a very
clear correlation between log and grain size.
Different authors (Sarma et a/., 1963; Alger, 1966) showed the existence of a correlation between resistivity, or
formation factor, and grain size (Fig. 3-44).
FLAP
-
Figure 3-42 Pad and Flap photo and description of the FMI tool
(courtesy of Schlumberger).
a
b
Figure 3-44 - Relationship between formation factor, resistivity and
mean grain diameter (from Sanna & Rao, 1963).
Remark
-
Figure 3-43 a): Schematic of the
tool. b): schematic of the ARI
tool (courtesy of Schlumberger).
Serrelog Q 2003
In the example of this figure, for the same value of R, the formation factor increases when the arain size increases. This seems to
contradict the commonly accepted relation. This situation is undoubtedly due to the fact that infresh water formations the importanceof surface conductance increases when the grain size decreases.
133
Well Logging and Geology
134
Other authors have observed relationship between a
well logging parameter and grain size.
In the example of Fig. 3-37it is obvious that a correlation exists between gamma-ray and grain size measured
on core samples. Gamma-ray increases when grain size
decreases, because radioactivity is linked with the finest
grains. These fine grains consist of clay minerals and silt
size heavy minerals (zircon for instance). Further analysis
shows that these minerals are fundamentally of detrital (or
allogenic) origin and are of structural or laminated type.
Indeed, it is hardly conceivable that authigenic clays occupying the pore space would show such evolution, because in this case the percentage in the rock does not significantly evolve with granulometry.
trometry of natural gamma radiation), Pb, or SP (Figs. 351 and Fig. 3-52,see next pages), at the intersection
point of the two other values on the cross-plot.
*E
Sb
8
D(PS
CORRELATIONS
RESISTIVITY
CURVES
The "bell or "funnel shapes of spontaneous potential, gamma-ray, or resistivity curves (Fig. 3-45) introduced by SHELL geologists around 1956,are another application of this relation between a curve and a textural parameter.
In these cases if we cannot precisely define the absolute grain size without preliminary calibration, we are
nevertheless always able to determine a relative size.
These shapes (Fig. 3-45),in fact, explain the normal graded-bedding ("fining upward") or reverse graded-bedding
("coarsening upward"). In other words we can conclude
that shale distribution is essentially of structural andlor
laminated type.
I'
wcMQIII.cToIu)(o
Figure 3-46 - Example of structural and partly laminated shales as they
can be recognized on dipmeter resistivity curves. The curve evolution
indicates fining-up sequences confirmed by core description.
Mpo
-
Figure 3-45 Relationship between SP curve shape and grain size or
shaliness (adapted from SHELL'S documents).
We can also convert these curves into shale percentage and into permeability if a preliminary calibration was
done with the help of core measurements (Fig. 3-48).
The relation between radioactivity and granulometry is
not always proved. It may happen that the silty levels are
more radioactive than shales (Fig. 3-49). In this case it is
necessary to have several log recordings, enabling the
grain size determination. The "cross-plot" combining two
logs, such as gamma ray and spontaneous potential
helps in this analysis, particularly if the technique of "Zplot" is used. This technique presents the plotting of the
value of a third parameter, such as GR (gamma ray) Th or
K (amount of thorium or potassium obtained from a spec134
-
Figure 3-47 This figure is an enlargement Of One SeqUenCe Of Figure
3-46 (between 834.5 and 837 ff).
Serralog @ 2003
Texture
I Chapter 3 I
135
Figure 3 4 8 - Schematic relationships between dipmeter resistivity on
one hand, and grain size, shaliness and permeability on the other
hand. These relationships are applicable to the case illustrated
by Figure 3-46.
If the thickness of each granulometric sequence is
very small the logging macrodevices cannot detect these
tools having a resolution respectively of 1 cm and 0.5 cm.
tant value of neutron and sonic’ at the same time.
-
Figure 3 4 9 Example of silts (green shading) being more radioactive
than shales. The real grain size evolution can still be detected on resistivity, SP and neutron curves ( h m Sera & Sulpice, 1975).
It is preferable, then, to use the resistivity curves of
dipmeter tools (HDT or SHDT) or image tools. As previously seen, their high vertical resolution allows the
detection of sequences of a few centimeters thickness
(Fig.
, - 3-46).
If a poorly-consolidated detrital reservoir contains
hydrocarbons* we can Observe the existence Of a good
correlation between grain size and irreducible water-saturation (Fig. 3-53). The latter depends, in fact, on permeability that depends, in turn, on grain size (cf. Fig. 3-18b).
Serralog 0 2003
K
Figure 3-52 - Examples of cross-plots of SSP vs thorium (a), and
potassium (b) contents. They define the Ths. and Ksh for shales. The
and potassium
of ucleanmsands suggest the presence of radioactive minerals such as feldspars, micas and zircon.
They show the existence of silts more radioactive than shales.
135
Next Page
136
Well Logging and Geology
Fig. 3-51 - Crossplots of pb vs @ with frequency (a), thorium content (b), potassium content (c), and thorium-potassium ratio (d), on the same
interval than Fig. 3-50a, and their interpretation.
Figure 3-55b illustrates the relationship that exists between grain size, its geometric mean and the permeability.
Lastly, the invasion diameter, as it can be deduced
from a microlog, may constitute a good indicator of grain
size evolution. The example of Fig. 3-54 shows an evolution of microlog resistivity curves which can be correlated
with changes in grain size (normal graded-bedding in this
case), - the porosity being the same in this interval -,
these changes affecting the diameter of invasion.
But, the best tools for determination of grain size and
grain shape are the dipmeter tools and more fundamentaly the image tools. Their very high vertical resolution
allows the detection of objects as small as 1 cm for dipmeters and 0.5 cm for image tools. This is illustrated by
Figures 3-55 to 3-57.
When pebbles are present and larger than the electrode size (1 cm for dipmeters, 0.5 for imagery tools) they
can be detected by dipmeter tools or by the imagery tools.
The pebbles are generally more resistive than the surrounding matrix in which they are embedded.
Consequently, each pebble is distinguished by a peak of
resistivitywhose shape varies with the size, the proportion
and the arrangement of the pebbles. This gives a heterogeneous aspect to the curves with a practically total
absence of correlations between them (Fig. 3-55a at the
bottom). When the pebbles touch each other (grain supported conglomerate), the peaks are very close. When the
136
pebbles are isolated in a sandy or shaly-sandy matrix
(mud supported conglomerate), the peaks are isolated
(Fig. 3-55b). The other open-hole logs may indicate a
detritic formation dominated by quartz and often feldspars
and micas, or by pebbles originating from igneous rocks.
The NGS tool may be very useful to determine the type of
radioactive minerals.
a
b
Fig. 3-53 - a): Relationship between irreducible water saturation, porosity and grain size. b): relationship between grain size (in 4 scale), its
geometric mean and permeability (from Krumbein & Monk, 1942).
Serralog 0 2003
Previous Page
Figure 3-54 - Observe the microlog resistivity curves. Their evolution reflects changes in invasion diameter, possibly in relation with grain size
decrease. Observe also the number of sequences which can be detected in this massive sandstone reservoir thanks to the microlog. The average
thickness of sequence (between I and 2 m) suggests a turbidite deposit.
Isolated resistive peaks showing variable thicknesses
from button to button (Fig. 3-5513) indicate irregular particle size and angular shape with rapid thickness change.
or to a reef or packstone if it is observed in a carbonate
deposit.
When grains are coarser than 5 mm, electrical images
allow their detection and the determination of their shape,
size and frequency.
For example, very resistive well rounded white spots,
often in contact with others (Fig. 3-56), correspond generally to grain-supported conglomerates if the other standard log data indicate detrital siliciclastic deposits.
When the white spots have an irregular shape and size
and are isolated in a more conductive background they
indicate a mud-supported conglomerate or a debris flow
(Fig. 3-57).
Very conductive peaks (black spots or cubes) correspond to very conductive crystals, generally of pyrite or
galena (Fig. 3-58).
Black rounded spots in carbonates correspond most
probably to molds of oolites which have been dissolved by
selected leaching of particles (oolites) composed of more
unstable minerals (aragonite or impure calcite) (Fig. 3Serralog 02003
59).
Sorting
One can approach this parameter with a study of porosity evolution. In fact in sandy formation, at a given depth,
we can reasonably assume that arrangement and packing
are the same for all grain sizes, the porosity generally
decreasing when sorting decreases.
Fig. 3-60 gives an example of the change in sorting.
Levels 9 and 10 present an average porosity close to 35
%. Taking into account their depth (7000 feet) this high
porosity is certainly due to a good sorting. The low radioactivity of these levels combined with a strong deflection
of spontaneous potential seems to indicate the presence
of radioactive minerals other than shales. Granulometry of
these levels must be fine to very fine. Level 11 shows a
lower porosity (25 %) with less radioactivity and an identical spontaneous potential. This drop in porosity, seems,
therefore to indicate a poorly sorted sand (coarse to fine
grained).
Again, the best way to analyze this phenomenon is to
make a cross-plot (type Z-plot), combining the hydrogen137
neutron index and the density with radioactivity, spontaneous potential values, or the amount of thorium or potassium. On such cross-plot (Fig. 3-61), from the point defining the maximum porosity for a given interval (corresponding to the best sorting), a drop in porosity along the
sandstone line (not less than 15%), represents a decrease in the level of sorting.
measurements and from "red" or "blue" patterns on arrow
plots.
Varve?
CORE DESCRIPTION
Glacial
outwash
deposit
Boulder
Figure 3-55b - Observe towards 144 m on pads I , 2 and 3 the thickness variations of the peaks. Their thickness evolves from 1 m on pad
2 to 0 m on pad 4. This indicates a block or boulder which does not
cross the well. Such blocks are more probably carried by ice and deposited when the ice melts
Pad orientation
Figure 3-55a - Grain-supported conglomerate at the bottom, evolving to
sand with pebbles, sands laminated and homogeneous, and shale at the
top. The core description confirms this interpretation of the HDT curves.
Grain orienfafion
A preferential orientation of grains must theoretically
be indicated by an anisotropy in resistivity reflecting an
anisotropy in the permeability from which it is derived.
But, the contrast between vertical and horizontal resistivities being small (around 1 3 , it seems tenuous to attribute any variation in the two horizontal axes to anisotropy
because other explanations can be responsible of this
variation (pad contact, mud-cake thickness...).
It does not Seem possible at the present time to
approach this phenomenon quantitatively, H
~ we
can approach it by means of relative deflections or, more
specifically, by directions of currents determined from dip
138
Figure 3-56 - FMS example of grain-supported conglomerate. Observe
the well rounded white spots which are locally in contact with similar
others.
They
to 4 or 5 cm.
~
~ correspond
~ to pebbles
~ with a ~diameter close
,
The dark spots correspond to pebbles pulled out from the formation by
rock bit. The background corresponds to a sandy formation.
Serralog
2003
Texture
I Chapter 3 I 139
U
Figure 3-59 - Thin section showing oomoldic porosity (left). On the right
observe the shape and the size of the black spots on the FMI images
(courtesy of Schlumberger).
A
Figure 3-57 - Example of mud-supported conglomerate confined by
core photographs reproduced on the side. A coarsening-up sequence
can be deduced from the increase in size of the resistive features from
XI27 to X118 m. This vertical grading could not be seen in core since
large cobbles look like beds. Observe on images the rapid changes in
thickness of the white spots and their shape. This intervalk can be
interpreted as corresponding to debris flow
(from Karker et al., 1988).
J
Fig. 3-60 -Observe the change in porosity from levels 9 and 10 and 11.
It corresponds to a change in sorting. The two lower sands are fine,
clean, well sorted, slightly radioactive; the upper sand is coarser and
badly sorted (from Serra 8 Sulpice, 1975).
4
Figure 3-58 - Example of pyrite crystals detected on FMS resistivity
curves (conductive troughs) and on image (very dark cubic spots).
Orientation of pebbles can be sometimes observed on
images (Fig. 3-62). This can be interpreted for the recognition of the transport current direction (cf. Fig. 3-1 3).
Arrangement or Packing
This parameter is not accessible, because we can reasonably admit that after a burying of several hundred metres, packing of sediment is such that arrangement of
grains becomes more compact (or closed).
Serralog (8 2003
J
Figure 3-61 - Sorting decrease, grain size evolution and type of shale
distribution can be predicted from cross-plot analysis.
Remark
This porosity drop could correspond to a cementation by precipitation of silica. Even if this hypothesis cannot be totally rejected, it is however unlikely if we consider the existing high porosity.
139
Well Logging and Geology
140
tool should enable determinationof the orientation of mica
flakes and consequently give an indication of the existence of horizontal or vertical permeability barriers in a micaceous sand.
As we have seen, electrical images allow the recognition of the object shape as soon as their size is higher
than 5 mm. For instance, well rounded pebbles, rock fragments or cubic crystals of pyrite can be recognized (cf.
Figs. 3-56 to 3-58). In chalk, chert nodules are well recognized on image thanks to their irregular shape (Fig. 363).
J
-
Figure 3-62 On the left, inclined pebbles (white spots) can be observed in this mud-supported conglomerate confirmed by the core photograph. On the right another FMS example of mud supported conglomerat (fmm Serra, 1989).
The study of porosity within a short interval, or even
more understandably at any given point, cannot reveal
information on packing. All porosity variations can be
explained, in a plausible manner, by a change in sorting
or by diagenetic effects (cementation, dissolution).
On the other hand the evolution of porosity with depth
on a long interval will explain the modification of packing
under the effect of compaction, and/or of diagenesis. This
aspect will be analysed with the study of compaction.
Grain shape
If the sandy formation is chemically immature and, therefore, rich in feldspars, mica etc .... shown by recordings
of litho-density and by natural gamma ray spectrometry
(relatively high amounts of potassium), we can deduce
the existence of a textural immaturity and consequently of
angular grains. On the other hand, if the sand appears to
be very clean, with a very low radioactivity and high porosity, we can suppose the existence of a chemically and
texturally mature sand, hence probably well-washed or
winnowed and well-sorted sand with round spherical
grains.
From Sen (1980, 1981), the grain shape has a strong
influence when the electromagnetic field is applied perpendicularly to the flakes (micas). Thus, sands rich in
mica flakes should show a higher dielectric constant than
the one expected from the measured values for quartz
and mica. The analysis of the data recorded by the EPT
140
d
-
Figure 3-63 Chert nodules (white spots) in a chalk. Observe the
inegular shape of the white spots (courtesy of Schlumberger).
Cement
The cement proportion is more or less important. As
well known, porosity will be lower due to cementation. In
case of cementation, it is, therefore, not always easy to
draw certain obvious conclusions concerning the grains.
However, for the same content of cement, a decrease in
porosity must be attributed to a diminution of sorting.
The type of cement can be determined from the mineralogical composition of the rock, obtained from log analysis. In the case of ambiguity (calcitic or dolomitic
cement), we have to remember that because cementation
is always a postdepositional phenomenon, the volume of
cement added to the porosity value cannot exceed the
maximum porosity value that existed at the beginning of
the cementation process.
Thus, if we compute from cross-plots a percentage of
calcite cement which is higher than usual, either a hypothesis of dolomitic cement must be accepted, or, if the
existence of calcitic cement is confirmed from other data
(Pe from lithodensity tool LDT, cutting analysis), the presence of detrital limestone particles (bioclasts, oolites...),
associated with grains of quartz or feldspars must be
considered.
Serralog Q 2003
Texture
I Chapter3 I 141
Carbonate rock texture
We know that in this type of formation the precocious
diagenetic effects have a tendency to completely modify
the initial texture. Therefore, in this particular case, it will,
exceptionally, be possible to obtain information on texture.
The dipmeter resistivity curves after correlation with
cores, may indicate some types of texture. Fig. 3-64
shows that calcirudites or boundstones, calcarenites or
grainstones and calcilutites or mudstones can be recognized from the shape of resistivity curves.
As shown by Fig. 3-65 an excellent agreement between the core description given in terms of constituent
percentages and one resistivity curve is obvious. It allows
a very detailed description of the formation in terms of
mudstone, wackestone, packstone, grainstone and even
boundstone. It is of considerable interest for the palaeogeographic reconstructionof a reefal environment, especially if natural gamma ray spectrometry (NGS tool) data,
the photoelectric index (Pe), and the density measurement (LDT tool) are available. The uranium content measured by the NGS tool will reflect reducing (uranium present), or oxidizing (uranium absent) conditions, which in
turn may help to locate the back- or the fore-reef. The Pe
will indicate the importance of dolomitization.
By comparison with core analysis an empirical relationship has been established between (a) curve and dip
parameters, and (b) facies in carbonates. It is reproduced
in Table 3-6.
A high porosity accompanied by shallow invasion, if it
is continuous along a certain interval, can correspond to
chalk. Of course, the other open-hole logs must indicate a
limestone composition.
Foreset laminae detected on images in carbonates
indicate grainstone texture as foreset beds correspond to
grains, consequently to oolites or shell fragments carried
by current of high energy (Fig. 3-66 see next pages).
Other carbonate textures can be detected from image
analysis as illustrated by Figures 3-67and 3-68.
Porosity
The porosity of detrital siliciclastic deposits results
from textural parameter influence. As it is obvious, a very
well sorted sand will present a higher porosity than a
badly sorted sand; a cemented sandstone will be less
porous than a sand. Consequently, as soon as the porosity is evaluated it is important to extract as much as possible information linked to textural parameters: sorting,
and cementation.
The existence of vugs is sometimes easier to determine. With the help of the electrical images, the vugs can be
even seen and their percentage evaluated (Fig. 3-69).
They correspond to conductive peaks on the resistivity
curves, or to dark irregular spots on the images. The other
open-hole logs will confirm the carbonate nature of the
rock.
Semlog (9 2003
I
Ern
I*)
TEXTURAL
CURVES
RESISTWITY
INTERPRETATION
a
-
82
84
1
-
figure 3-64 Exam. I of HOT response in a m f a l environment. Note
that the aspect of the dipmeter resistiwify curves enables separation of
boundstone at the base (characterizedby peaks of variable thicknesses and absence of correlations), grainstone at the center (marked by
very small curve activities) and mudstone (characterized by high resistivity without variations).
Table 3-6
Relationship between dipmeter signatures and
carbonate texture (from Theys et a/., 1983).
141
porous
coral branch
not porous
Figure 3-65 - Comparison of textural interpretation of HOT micro-resistivity curves with core description (from Schlumberger, Well Evaluation
Conference, India, 1983).
XX15
2R
XX17
Figure 3-68 - Image which can be interpreted as representing boundstone texture. Observe the vertical distribution of white spots which can
correspond to reefal branchs.
Analysis of the images allows in many cases to determine the type of porosity distribution as illustrated by
Figures 3-69 to 3-72.
The vugs can be connected to secondary porosity due
to the dissolution of calcite or dolomite crystals by water
circulation. We know that, in this case, the sonic tool does
not (( see >) this secondary porosity. Consequently, a comparison of porosity determined from the combination of
density and neutron tools (tools that (( see >> the total porosity connected or not), with porosity deduced from sonic
tool will indicate this vuggy secondary porosity (Fig. 373a). This secondary porosity index (SPI) is computed
and reproduced alongside the left track in computational
result display (Fig. 3-75). For this purpose we can also
use the M-N-plot, or the MID-plot techniques (Fig. 3-73b).
Figure 3-67 - Example of mudstone texture in which several stylolites
can be observed.
142
Serralog 0 2003
Texture
I Chapter3 I 143
North
Figure 3-66 - Example of images in grainstone. The foresets observed in this limestone cannot be explained if they are not associated to particles
carried by a current. These particles are oolites or shell fragments. The direction of the transport currents is determined from the azimuth of the
foreset dips. This can be exploited to deduce the direction of the elongation of the oolitic bars and the permeability anisotropy
Serralog 0 2003
143
Next Page
A
Figure 3-69 - On the left, FMI example of pore distribution which can
be interpreted as illustrating fenestral porosity and alga mound deposit.
On the right, photo of algal mat with bird’s-eye pores.
Figure 3-70 - On the left, FMI example of vuggy porosity (black spots)
with anhydrite nodules (white spots) in a dolostone. On the right photo
of similar rock type.
Figure 3-72 - Vugs (black spots on the left images) are isolated by
thresholding of the resistivity (right images). The percentage of the
surface occupied by the vugs can be computed ( 72% in this case)
(courtesy of Schlumberger).
I
I
Figure 3-71 - On the left-up, FMI image of vuggy porosity. The shape
of the vugs suggests an oomoldic porosity most of the time connected
to others and to background porosity. On the right, photo of a thin
section showing a similar oomoldic porosity (in pink).
The lower figures show a SEM photo of the same oomoldic porosity
and a vuggy porosity on the right.
In fact, the first arrival corresponds to the one, the trajectory of
which, has missed most of the vugs. At best, the secondary porosity
index (sp~)
represents a minimal
value of the secondary porosity,
It
would be more correct to refer to an index of heterogeneity of pore distribution, which, of course is dependent on the presence of secondary
porosity. A regular distribution of small vugs would generate a nil secondary porosity index which would mean that the sonic tool would see the
total porosity.
144
Fig. 375% - Detection of secondary porosity by comparison of porosities computed from sonic tool, with those deduced from the neutrondensity combination.
Serralog 0 2003
Previous Page
cate an interpretation method of the measurement and
extract an evolution of the micro-geometry and of the
cementation factor with depth in carbonate series.
Rasmus & Kenyon (1985) developed a technique for
estimating separately the amounts of intergranular and
oomoldic water. They used the results to predict the presence of oil and its flow rate in carbonates.
Primary Medium
Distributed
porosity
Jncluded Spherical Pores
0.8
Figure 3-73b - Cross-plot M vs N for mineral and secondaty porosity
determination (courtesy of Schlumberger).
Electrically
isolated pores Electrically
connected pores -
Similarly to the sonic travel time, the resistivity is also
affected by the distribution of the pores in the rocks. The
resistivity does not "see" the isolated or not connected
pores (Fig. 3-74). Neither is it influenced by the size of the
pores, but only by the network of the capillaries which
connect them.
Brie et a/., (1985) developed a method of analysis of
acoustic and electric measurements, based on KusterToksoz model for acoustic velocities, and the MaxwellGarnett model for resistivity, which determines the
amount of spherical (oomoldic) porosity in carbonate
rocks. In addition, they proposed a way of evaluating the
cementation factor level by level from the sonic indication
of spherical porosity (Fig. 3-75). This value can be used to
obtain a more accurate value of the water saturation in
such reservoirs.They also mentioned that the accuracy of
the model could be increased by combining it with other
texture sensitive measurements such as dielectric permittivity.
From recent works it seems that dielectric measurements made by the electromagnetic propagation tool
(EPT tool), combined with data extracted from microresistivity devices, can help to determine the microgeometry of
a carbonate rock.
Kenyon (1984), Kenyon & Baker (1984) demonstrated
the need to introduce a "bimodal model" taking into
account the micro-geometry of the rock (shape of pores
and grains) to explain the EPT tool response. They advoSerralog 0 2003
Fig. 3-74 - Schematic representation of the effect of vugs (supposed to
be more or less spherical or oomoldic) on the electric properties.
Isolated pores are in blue (from Brie et al., 1985).
In conclusion, one can emphasize the importance of
the knowledge of the texture for permeability estimation
and for a more accurate saturation computation, especially in carbonate reservoirs. This knowledge will also
help in the quantitative interpretation by favouring the
choice of the value of the rn factor of the Archie's equation.
Porosity and pore size determination from nuclear
magnetic resonance measurements
It is important to remember that the transverse relaxation time T, measured by a nuclear magnetic resonance
tool is lithology independant and is a function of the hydrogen nuclei content in the pore fluid. T, is proportional to
V/S (V = volume, S = surface) which in turn is proportional to pore size (Fig. 3-76). For instance, for a spherical
pore SN = 3/r, where r is the radius of the pore.
145
Figure 3-76 - The decaying signal amplitude (top) is the sum of all the
decaying T2 signals generated by hydrogen protons in the measurement volume. Separating out ranges of T2 values by a mathematical
inversion process produces the T2 distribution (bottom). The curve
represents the distribution of pore sizes, and the area under the curve
is the porosity (courtesy of Schlumberger).
Fig. 3-75 - Example of results of spherical porosity computation
(from Brie et al., 1985).
Interpretation of the pore size distribution and the logarithmic mean T2 are used to calculate parameters such as
permeability and free-fluid porosity.
In siliciclastic deposits, the threshold is generally put at
33 msec (Fig. 3-77). for irreducible water saturation. In
carbonate rocks the threshold for vuggy porosity is put at
100 msec (Fig. 3-78). Figure 3-79 illustrates the influence
of pore size on the relaxation time.
146
3
Figure 3-77 - The area under the T2 curves to the left of 33-msec cutoff corresponds to very small pores with irreducible water saturation.
From 33 to 210 msec, the raea under the curve represents producible
fluid and pore size increase. Above 210 msec, the area represents oil
(courtesy of Schlumberger).
Serralog 0 2003
shows the validity of this technique (Fig. 3-80).
Pore diameter, microns
Time, msec
Figure 3-80 - The T2 time is calibrated in pore diameter
(courtesy of Schlumberger).
Automatic extraction of textural information from
dipmeter curves
J
Figure 3-78 - In this carbonate, the area below 100 msec represents
very small pores with irreducible water saturation. The area between
100 and 750 msec represents largerpores and vugs. Above 750 msec
the area may correspond to oil (courtesy of Schlumberger).
As previously seen, dipmeters reflect the internal organization of formations. This is illustrated by Figures 3-82.
In order to extract and quantify this information a program
was written (Delhomme & Serra, 1984). Its name is SYNDIP. An example of its results is given in Figure 3-83.
J
Figure 3-79 - Grain surface relaxation is the most important process affecting TI and T2 relaxation times. When the pores are small the relaxation
is rapid as the probability of collisions with grains is higher. When the pores are large this probability is lower.
The T, curve can be converted into pore diameter (Fig.
3-80). The width can be interpreted as reflecting the sorting. A sharp peak on T, above 33 msec, may represent
a deposit with well sorted grains and pores of similar size.
The threshold value at 33 msec has been established
from experiments.A large number of water-saturated core
samples from two wells was analyzed for porosity determination. The cores were centrifuged under 100 psi pressure to simulate draining down to typical reservoir capillary pressures. The amount of fluid drained (Fig. 3-81)
equals the free-fluid porosity, which is converted into an
equivalent area on the T, distribution curve. The area shaded from the right - determines
values.
A
parison of T, distribution taken before (curve in green)
and after (curve spun sample in purple) centrifuging
Serralog 0 2003
J
Figure 3-81 - Comparison of porosity extracted by centrifuging core
samples of two wells and free fluidporosity. Observe the good correspondance specially above 6 p.u. (courtesy of Schlumberger).
147
and dark when it reads the mud). Such effects are easily
recognized, and the intervals can be zoned out on the
affected pads during the image analysis.
Depth
(m)
Calibrated
dipmeter
resistivity
1
Figure 3-82 - The resistivity curves of dipmeter tools reflect the internal
organization of formations. Cutves flat, without character, indicate an
apparent homogeneous formation. Curves showing high activity
(numerous deflections or breaks, FBR) without evident correlations
indicate a heterogeneous formation (pebbly, vuggy or reefal, bioturbations). Curves showing high activity with high density of correlations
(DCL) indicate a thinly laminated formation.
(from Delhomme 8 Serra, 1984).
But, even if the resistivity dipmeter curves can reflect
formations heterogeneities like pebbles or vugs, images
are the corner stone of the textural determination. They
inform us about the internal organization of each depositional unit reflecting partly the energy in the medium of
deposition. Their analysis may be visual but can be as
well realized by their processing using a software. The latter simulates what the eyes can achieve. It is why this processing is now explained in detail and is commercialized
under the name of BorTex (previously ScanTex or
TexScan, or SPOT) of which the goals and approach are
explained here below and in a paper written by J.P.
Delhomme.
BorTex Program
The gray- or color-level changes relate only to conductivity changes. Interpreting ambiguous images may require knowledge of the geological context or other log data
(Serra, 1989). Dark rounded areas may be vugs filled with
conductive drilling mud, clay clasts or pyrite crystals, or
even pebbles pulled out by the drilling bit. Dark linear features may indicate fractures filled with mud or another
conductive material such as clay or pyrite. Large-scale
image level variations may relate to changes in shaliness
or porosity. The same holds true for automated image
analysis - the image must be interpreted in context.
Although the current is strongly focused into the formation, electrical images can be adversely affected by
bad borehole conditions. In very rough holes, a blurred
image can result (white when the pad touches the wall
148
Figure 3-83 - Example of SYNDIP result display. In the left track are
reproduced in thin lines the minimum and maximum resistivities recorded by SHDT buttons, in heavy line the average value. The second
track displays the internal organization of the beds. The limiting curve
is the frequency of inflexion points. The three columns on the side, are
flags indicating: detection of non planar surfaces, non parallelism between consecutive bed boundaries, and correlated parallel planes. In
green strip are indicated the average dip determined from a minimum
of 5 consecutive dips with angular spherical dispersion lower than 5".
The next track shows the conductivity curve shaded with a color scale
to reinforce the sand/shale opposition in sandkhale series. The blue
strip width indicates the number of buttons having showed high resistivity. The last track displays thickness of conductive or resistive beds in
a two mirror logarithmic scales. The cumulated resistive thickness is
indicated on the left (courtesy of Schlumberger).
Serralog Q 2003
Texture
I
Chaoter3
Table 3-7
Determination of carbonate texture from well logs and images.
Standard t d response
lmage characters
texture
bhck spots (faU of chert nodules, mrcasite)
W close to 4 - 4.2 We
-ll___l
Variable porosity
Nbst of thetime visible laminations, foresets or crossbedding
Stylolites may be present if corqacted
-.
Faelatiiely hom~geneouspale yellow or w hte knages.
Few white spots on dynarric normalized knages (shell
fragments) .
Few dark spots or patches corresponding to isolated
-
-
"WS
___-
___--__
Styldies m y be present
(cf. Fg.3-67)
__
._
-
Irregular orange or while patches on enhanced images
Imagery tools provide a cylindrical view of the formation.
This view can be considered two-dimensional either by
unrolling the cylinder or separately visualizing the strips
covered by each pad. The measurement resolution and
spacing are small enough with respect to the size and
spacing of most formation heterogeneities that one can
expect to image these heterogeneities. The shapes and
spatial arrangements that emerge with the images complement traditional log analysis. The effect of heterogeneities on pad tool readings (e.g., density) and the lateral
extent of one-dimensional high-resolution logs can be
evaluated.
In siliciclastic rocks, images allow us to recognize
debris-flow deposits or grain-supported conglomerates
(cf. Fig. 3-56 and 3-57), the siliciclastic nature of the rock
being confirmed by standard open hole logs. The size of
the rock fragments or pebbles and the volume percentage
they represent (Fig. 3-84) are determined by an automatic analysis of the image through the computation of the
area they cover on the image and an assumption about
their shape (cubes, parallelepipeds, spheres or ellip
Serralog Q 2003
soids).
In carbonate rocks, the images allow us to classify
them in terms of grainstone (cf. Fig. 3-66), mudstone (cf.
Fig. 3-67) or boundstone (cf. Fig. 3-68), or in terms of
debris flow or breccia (Fig. 3-85). The reasoning to apply
for carbonate texture determination are summarized in
Table 3-7. It includes both standard logs and images.
The following text is a summary of a paper written by
J.P. Delhomme.
Zonation based on imaae texture
The computerized analysis of borehole images leads
to zonation of the formations by their internal structure
(Fig. 3-86). Five zone types can be identified from electrical images: massive, laminated, interwoven heterogeneous, heterogeneous with conductive inclusions, and
heterogeneous with resistive inclusions. This classification is frequently used in the literature (Nurmi et a/. 1990).
This zonation can be obtained from electrical images
and automated using image topological properties. More
149
specifically, one can refer to the feature that extends
across the image. In a full-coverage image, this imagecrossing part obviously describes a feature of the formation that crosses the well. When a pad tool records several image stripes, further sophistication is required to bridge the inter image gaps.
Assume that the borehole electrical image can first be
made binary with reference to some adaptive threshold.
This threshold should track the evolution of the median
conductivity versus depth (cf. Fig. 3-67).
To get a zonation, the next step is to compute sliding
parameters that characterize the resistive and conductive
components and capture their mutual spatial arrangement. Specifically, the median conductivity values of both
components, their area percentages before and after eliminating non-crossing regions, and the lateral variation of
the apparent thickness of the image-crossing regions
must be captured.
The final step is to apply a set of rules.
- A massive zone is an interval where the median
conductivity values of the two components - if both coexist
- are not significantly different.
- A laminated zone is an interval where the area
percentages of, for example, the conductive component
are not significantly different before and after elimination
of the non-crossing regions. This also includes intervals
where the apparent thickness of the image-crossing
regions does not show significant lateral variations.
- An interwoven heterogeneous zone is an interval where the area percentages of, for example, the
conductive component are not significantly different befo-
re and after elimination of the no-crossing regions, but
where the apparent thickness of the image-crossing
regions shows significant lateral variation.
- A heterogeneous zone with conductive inclusions within a resistive background is an interval where
the area percentage of the conductive component is significantly smaller after eliminating the non-crossing regions
than before.
- A heterogeneous zone with resistive inclusions
within a conductive background is an interval where the
area percentage of the conductive component is significantly greater after eliminating the non-crossing regions
than before.
More elaborate rules exist that avoid the binarization
stage and are directly applicable to gray-scale images,
but the basic principles stay the same as those explained
here.
Quantitative analvsis.
When the zonation is performed, further quantitative
information can be derived from images. For instance, in
an electrical image of a typical carbonate rock, one identifies local objects (e.g., vugs or molds, hairline cracks or
fractures, chert or anhydritic nodules). The host rock
containing the vugs or molds is sometimes called matrix
rock in logging terms. In this publication, it will be referred
to as background rock to avoid any confusion with the
geological term "matrix" (defined as the finest-grained
material [not cement] filling the spaces between coarser
material [grains]). This background may reveal layering,
large patches and interwoven patterns.
1
Figure 3-84a - Open-hole logs corresponding to the images reproduced in Figure 3-846.
150
Serralog 0 2003
Texture
1 Chapter 3 1 151
il
Figure 3-84b - Example of image analysis in order to evaluate the area covered by conductive (C) and resistive (R) objects. On the left, images
with identification of the pebbles (more resistive spots than the background). On the right, analysis of the entire interval. The red strip locates the
image reproduced on the left [from Delhomme 8 Motet, 1993).
One of the first goals of borehole electrical image analysis was to extract and to quantify voids such as vugs or
molds in carbonate rocks. Another goal was to study their
interconnections, which generally constitute only a small
fraction of the total pore volume, but exert a decisive
control on the permeability. This approach is similar to that
one proposed for core analysis (Lucia 1983). The portion
of the total pore volume that comprises vugs, cracks and
fractures is called secondary porosity. Hence, the name
given to the image processing prototype is SPOT, which
stands for Secondary Porosity Typing (Delhomme, 1992).
However, the technique is equally applicable to heterogeneous siliciclastic reservoirs such as sandstones with
nodules, clay chips (flasers), clay galls or clay lenses, and
to conglomerates in which the rock heterogeneities usually determine a high- or low-quality reservoir.
The image analysis is accomplished through a series
of image transforms (also called filters). The transformation of images into other images (called image processing) and the transformation of images into data (called
image analysis) are both used.the program uses mathematical morphology transforms that simplify image data
while preserving essential shape characteristics.
Figure 3-84c - Details of FMS images showing the microconglomeratic
nature of the sandstone. The microconglomerate is composed of small
irregular granules.
Serralog 0 2003
151
2689
background. This change can be detected by applying
derivation operators. Higher contrast lines in the original
image correspond to crest lines in the gradient image,
which is also considered a relief (Fig. 3-87).
Among all the crest lines, the watershed, (divide) lines
are the water-parting lines that separate two different altitude minima. In the gradient landscape, such divide lines
can clearly be tracked along the crests of truncated coneshaped relieves (volcanoes) (Fig. 3-88). The morphological dual of the watershed line network is the thalweg network, which corresponds to water-collecting lines of the
landscape metaphor.
The watershed extraction is a gray-scale morphology
transformation that, when combined with other gray-scale
morphology tools, can be used in many gray-scale segmentation problems. It is a nonparametric and global
algorithm.
Figure 3-85 - Example of breccia. The angular shape of the white
spots are fragments of rock. Th core photograph on the side comes
from a nearby well in the same formation (from Pilenko, 1988).
The individual geological features (conductives or
resistives) are first extracted from the images. Then, the
features are characterized with attributes related to size
(area, perimeter), shape and connectedness. Finally, the
information is analyzed and summarized through summary logs to compare and integrate the data with standard
(e.g., porosity) logs (cf. Fig. 3-84), well test data and production logging surveys. Obviously, the first stage is critical to this analysis.
The simplest way to segment the image is by establishing thresholds (cf. Fig. 3-72). The landscape metaphor,
often used for gray-scale images, consists of topographic
representationof the image. The numerical value (i.e., the
gray-level) at each pixel stands for the elevation or terrain
altitude at that point. If the objects to be extracted have a
distinctive gray-level range, the points with an altitude
lower (more resistive) or higher (more conductive) than a
given threshold can be cut off to separate objects from the
background. The threshold is critical because it dictates
the contour and size of all objects. When the objects have
different gray-levels, the thresholding method fails. The
threshold value that would be required to extract every
object will outline some objects much larger than expected.
A second approach identifies the change in gray-level
that occurs at the boundary between objects and the
152
J
Figure 3-86 - Example of zonation of a formation, 2.5 meters thick,
based on image.
To avoid over-segmentation related to low-contrast
boundaries, a strategy called marker-controlled segmentation has been used. Rather than removing irrelevant
contour elements, the gradient image is modified so that
the resulting divides correspond only to the desired
objects. This strategy makes it easier to use external
knowledge, such as geological knowledge about the features extracted from the images.
The automated image analysis is performed in three
stages:
marking and outlining objects, computing object attributes, and calculating attribute statistics and summarizing
the results as logs. These stages are outlined in the following sections.
Serralog 0 2003
I
Figure 3-87 - 3 0 landscape representation of conductive peaks.)( axis
= pad width, Y axis = depth, Z axis = conductivity or gradient
(from Delhomme, 1992).
The adaptive threshold is used to reject low-contrast
local extrema. It is set at the value of image plateaus
obtained from morphologically eliminating highs and lows
(Rivest eta/., 1991). Then, the objects are contoured by
tracking the major contrast lines (i.e., the major gradient
modulus divides) between markers.
The gradient image is modified using a morphological
reconstruction operation (Rivest et a/., 1991). This operation fills the lows in the gradient image between object
markers and background marker, leaving only the highest
gradient crest.
In the final step of contour detection, the standard
watershed extraction is performed on the morphologically
reconstructed gradient image to produce one close
contour per marker The quality of the segmentation
depends on the reliability of the markers, but the marking
procedure may vary with the specific segmentation problem. The parameter tuning, if needed, is restricted to the
marker generation step.
In summary, to extract any geological object (vugs,
nodules, lenses or fractures), six processing steps are
successively performed:
1. if needed, apply a pre-filter to remove noise
from the image
2. determine the object marker image
3. determine the background marker image
4. compute the gradient modulus image
5. reconstruct the gradient modulus image from
object and background markers
6. extract the gradient watershed to find the
objectlbackground boundaries.
I
Figure 3-88 - 3 0 landscape representation of maximum gradient. M can
be compared to a volcano (from Delhomme, 1992).
Markina and outlinina objects.
A marker, which is a connected set of pixels included
in the region, first is automatically extracted for each of
the different regions to be segmented. The object markers
must roughly indicate the location of the objects, and the
background marker must surround the objects. Object
and background markers cannot overlap. Different prefilters and object and background marking procedures are
specified for each class of geological objects. The choice
is based on a priori geological knowledge of the expected
attributes of the objects.
For conductive spots (e.g., vugs filled with conductive
mud), the object markers are the image local maxima
(summits) that are higher than an adaptive threshold. The
background marker is the image thalweg network. For
resistive spots (e.g., nodules or pebbles in conglomerates), the object markers are local minima (sinks) that are
lower than an adaptive threshold. The background marker
is the divide network.
Serralog 0 2003
Computina object attributes.
To characterize each object in the "cylindrical twodimensional" view at the borehole wall, object attributes,
(e.g., area percentage in the image, apparent size, shape,
contrast and orientation in the image) are computed for
each object.
Connection between objects is also a very important
attribute, and it raises a theoretical problem for gray-scale
images. In the landscape metaphor, the whole landscape
is a totally connected surface. Only by establishing thresholds will the intuitive notion of connectedness appear.
For instance, if the sea level rises, the emerged lands are
no longer contiguous, and islands and continents are isolated. In fact, watershed and thalweg lines summarize the
gray-scale connectedness. The altitude drops can
approach this, from summits to saddle points along the
watersheds or by the jumps from sinks to saddle points
along the thalwegs. This principle is used to define the
vug connectedness from electrical images. Conductive
and resistive paths characterize connectedness. The
conductive paths used to track inter-vug channels or fractures are defined as arcs of the image divide network for
which the minimum (saddle-point value) is higher than an
adaptive threshold. Conversely, the resistive paths are
arcs of the image thalweg network for which the maximum
(saddle-point value) is lower than an adaptive threshold.
153
154
Well Logging and Geology
Some attributes can be used directly to infer threedimensional characteristics, such as area percentages
that can be translated into volume. Other attributes, such
as orientation characteristics, can be converted to three
dimensions
using
geometrical
transforms.
Connectedness attributes are not directly translatable
from two to three dimensions, apart from convex or spherical models that do not apply to many geological objects.
For such attributes, the cylindrical view is used to predict
three-dimensional connectedness and anisotropy.
Calculatina attribute statistics and summarizina results
as Ioas.
analysis of the images, in conjunction with the other openhole logs, allows this determination (Fig. 3-92). Images
showing no change in color will suggest a homogeneous
formation. Dark or white spots in a homogeneous background will suggest either clay galls or granules, nodules,
pebbles as a function of their size and resistivity. Dark
laminae in a sandstone may correspond to laminated shales. Vertical evolution of a homogeneous background will
suggest either a fining-up or a coarsening-up sequence in
grain size which in turn will suggest depending on the
thickness of this variation either fluvial channel fill, turbidite deposits or progradation of deltaic deposits.
To display results at a compressed scale for zoning
and comparing with standard logs, sliding window filters
are applied to output summary logs such as percentage
area of vugs, median vug size and vug connectedness
index, which can be expressed as the cumulative apparent fracture length in borehole wall per unit length of the
well.
Influence of textural parameters on
petrophysical properties
It is obvious and well known that textural parameters
control the petrophysical rock properties. This is the reason why it is so important to determine as much as possible this parameters.
In siliciclastic deposits determination of the grain size
and the sorting will help to link the permeability to the
porosity (cf Fig. 3-18, 3-20, Fig. 3-89 and Fig. 3-90). The
permeability also depends on the pore radius (Fig. 3-91).
J
Figure 3-90 - Influence of grain-size (expressed in 4 scale) and its geometric mean on the permeability (from Krumbein & Monk, 1942).
Porosity (%)
1
Figure 3-89 - Relationship between permeability and porosity as a
function of the grain size and sorting (from Chilingar, 1964).
In carbonates, the determination of the textura will
allow the determination of the rn factor of the Archie’s
equation (Fig. 3-92), and the link between the permeability and the porosity as illustrated by Fig. 3-26.
It is also important to determine the type of clay and
their distribution into the formations, as they will affect the
porosity, the permeability and the resistivity as well. The
154
Figure 3-91 - Influence of pore radius and porosity on permeability
(from Gaida et a/., 1973).
Serralog 0 2003
Texture
Porosity (%)
00
Figure 3-92 - Relationship between the formation factor and the prosity
allowing the determination of the m factor of the Archie's equation
(from Nurmi et a/., in Middle East Well Evaluation Review, 1986).
Figure 3-93 - Effects of the clay type and of their distribution on the
petrophysical properties of detrital deposits.
References and Bibliography
ALGER, R.P. (1966). - lnterpretation of Electric Logs in
fresh Water Wells in Unconsolidated Formations.
SPWLA, 7fiAnn. Log. Symp. Trans., Tulsa.
ALLEN, J.R.L. (1984). - Sedimentary structures. Their
character and physical basis. Developments in
Sedimentology, 30, Elsevier, Amsterdam.
BALOSSINO, P., & ANXIONNAZ, H. (1994).- Deriving
textural and geometrical information from dipmeter data
Serralog 0 2003
I Chapter 3 I
155
as a help to define subsurface geological models. Agip
s.p.a. & Schlumberger paper.
BATES, R.L. & JACKSON, J.A. (1980). - Glossary of
Geology. Amer. Geol. Institute, Falls Chuch, Virginia.
BEARD, D.C., & WEYL, P.K. (1973). - lnfluence oftexture on porosity and permeability of unconsolidated sand.
Bull. Amer. Assoc. Petroleum Geol., 57, p. 349-369.
BLATT, H., MIDDLETON, G., & MURRAY, R. (1972,
1980). - Origin of Sedimentary Rocks. 1st and 2nd ed.
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
BOURKE, L., DELFINER, P., TROUILLER, J.C., FETT,
T., GRACE, M., LUTHI, S., SERRA, 0. & STANDEN, E.
(1989). - Using Formation Microscanner Images. The
Technical Review, vol. 37, n o l .
BRIE, A., JOHNSON, D.L., & NURMI, R.D. (1985).
Effect of spherical pores on sonic and resistivity measurements. SPWLA, 26th Ann. Log. Symp. Trans., paper W.
CAROZZI, A.V. (Ed.) (1975). - Sedimentary Rocks.
Benchmark Papers in Geology, 15, Dowden, Hutchinson,
& Ross, Inc., Stroudsburg, Pennsylvania.
CHILINGAR, G.V. (1969). - Deltaic and shallow marine deposits. L.M.J.U. van Straaten, ed., Elsevier,
Amsterdam.
CHOQUETTE, P.W., & PRAY, L.C. (1970). - Geological
nomenclature and classification of porosity in sedimentary carbonates. Bull. Amer. Assoc. Petroleum GeoL, 54, p.
207-250.
CLAVIER, C., & RUST, D.H. (1976). - MID-PLOT: a
new Lithology Technique. The Log Analyst, 17, 6.
DELHOMME, J.P. (1992). - A quantitative characterization of formation heterogeneities based on borehole
image analysis. SPWLA, 33rd Ann. Log. Symp. Trans.,
paper T
DELHOMME, J.P. & MOTET, D. (1993). - Reservoir
description and characterization from quantitative borehole image analysis. Reservoir Characterization Africa and
Mediterranean Region, no 5, Schlumberger Wireline &
Testing.
DELHOMME, J.P. & SERRA, 0. (1984). - Dipmeterderived logs for sedimentological analysis. 9th Colloque
International de Diagraphies, SAID-SPWLA, Paris, paper
50.
DELHOMME, J.P., PILENKO, T., CHERUVIER, E. &
CULL, R. (1986). - Reservoir Applications of Dipmeter
Logs. 6lstAnn Fall Techn. Conf. SPE of AIME, paper SPE
15485, and in J. Petrol. Techn., 29, 2, pp. 180-186.
DICKEY, P.A. (1979). - Petroleum Development
Geology. Petroleum Publishing Co., Tulsa.
DODGE, C.F., HOLLER, D.P., & MEYER, R.L. (1971).
- Reservoir heterogeneities of some Cretaceous sandstones. Bull. amer. Assoc. Petroleum Geol., 55, p. 18141828.
DUNHAM, R.J. (1962). - Classification of Carbonate
Rocks according to Depositional Texture. Amer. Assoc.
Petroleum Geol., Mem. 1, p. 108-121.
ENGELHARDT, W.V. von (1960). - Der porenraum der
sedimente. Springer, Berlin, 207 p.
FOLK, R.L. (1959). - Practical Petrographic
Classification of Limestones. Bull. Amer. Assoc.
155
Petroleum Geol., 43, 1, p. 1-38.
FOLK, R.L. (1962). - Spectral subdivision of Limestone
Types.Amer. Assoc. Petroleum Geol., Mem. 1, p. 62-84.
FRASER, H.J. (1935). - Experimental study of porosity and permeability of clastic sediments. J. Geol., 43, p.
9 10-10 10.
FRIEDMAN, G.M., & SANDERS, J.E. (1978). Principles of Sedimentology. John Wiley & Sons, New
York.
GAIDA, K.H., RUHL, W. & ZIMMERLE, W. (1973). Examination of the voids in sandstones with the aid of
scanning electron microscope. Erdoel & Gas, 89, 9, p.
336-343 (in german).
GRATON, L.C., & FRASER, H.J. (1935). - Systematic
packing of spheres with particular relation to porosity and
permeability. J. Geol., 43, p. 785-909.
GRABAU, A.W. (1903). - Paleozoic coral reefs. Bull.
Geol. SOC.Amer., 14, p.337-352.
HOBSON, G.D., & TIRATSOO, E.N. (1975). lntroduction to Petroleum Geology. Scientific Press Ltd,
Beaconsfield, England.
KENYON, W.E. (1984). - Texture effects on megahertz
dielectric properties of calcite rock samples. J. Appl.
Phys., 55, 8, p. 31533159.
KENYON, W.E., & BAKER, P.L. (1984). - EPT interpretation in Carbonates drilled with salty muds. 59th Ann.
Tech. Conf. SPE of RIME, paper SPE 13192.
KENYON, W.E., KLEINBERG, R., STRALEY, Ch.,
GUBELIN, G. & MORRISS, Ch. (1995). - Nuclear
Magnetic Resonance lmaging - Technology for the 21s'
Century. Oilfield Review, Autumn, vol. 7, no 3.
KRUMBEIN, W.C. (1941). - Measurement and geological significance of shape and roundness of sedimentary
particles. J. Sed. Petrology, vol. 11, pp. 64-72.
KRUMBEIN, W.C., & MONK, G.D. (1942). Permeability as a function of the size parameters of
unconsolidated sand. Petroleum Technol., AIME Techn.
Publ. 1492, 5, p.1-11.
KRUMBEIN, W.C., & SLOSS, L.L. (1963). Stratigraphy and Sedimentation. 2nd ed. W.H. Freeman, &
Co., San Francisco.
KUSTER, G.T., & TOKSOZ, M.N. (1974). - Velocity
and Attenuation of Seismic Waves in TwoPhase Media :
Part I, Theoretical Formulations; Part ll, Experimental
results. Geophysics, 39, 5.
LANDES, K.K. (1951). - Petroleum Geology. John
Wiley, & Sons, New York.
LEE, C.H. (1919). - Geology and groundwaters of the
western part of San Diego County, California. Wat.-Supply
Irrig. Pap. Wash., 446, 121 p.
LEET, L.Don, JUDSON, S., & KAUFFMAN, M.E.
(1978). - Physical Geology. 5th ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
LINK, P.K. (1982). - Basic Petroleum Geology. DGCl
Publications, Tulsa.
LOMBARD, A. (1956). - Geologie Sedimentaire. Les
series marines. Masson, Paris.
LOMBARD,A. (1972). - Series sedimentaires. Genese
- Evolution. Masson, Paris.
156
LONGMAN, M.W. (1980). - Carbonate diagenetic textures from nearsurface diagenetic environments. Bull.
Amer. Assoc. Petroleum Geol., 64, p. 461-487.
MAXWELL-GARNETT, J.C. (1904). - Transactions of
the Royal Society, 203, p. 385, London.
MIDDLETON, G.V. (1976). - Hydraulic interpretation of
sand size distributions. J. Geology, 84, p. 405-426.
NEASHAM, J.W. (1977). - The morphology of dispersed clays in sandstone reservoirs and its effect on sandstone shaliness, pore space and fluid flow properties. 52nd
Ann Fall Meet SPE of AIME, paper SPE 6858.
NURMI, R.D., & FRISINGER, M.R. (1983) .- Synergy
of Core petrophysical measurements, log data, and rock
examination in carbonate reservoir studies. 58th Ann. Fall
Mtg. and Techn. Conf, of SPE ofAIME, paper SPE 11969.
NURMI, R. (1986). Carbonate Close up. Middle East
Well Evaluation Review, no 1.
PETTIJOHN, F.J. (1930). - Imbricate arrangement of
pebbles in a pre-Cambrian conglomerate. Jour. GeoL, 38,
p. 568-573.
PETTIJOHN, F.J. (1975). - Sedimentary Rocks. 3rd ed.
Harped & Row, Publishers, New York.
PETTIJOHN, F.J., POTTER, P.E., & SIEVER, R.
(1972). - Sand and Sandstone. Springer, New York.
PICKETT, G.R. (1960). - The use of acoustic logs in
the evaluation of sandstone reservoirs. Geophysics, 25,
1, p. 250-274.
PICKETT, G.R. (1963). - Acoustic Character Logs and
their Applications in Formation Evaluation. J. Pet.
TechnoL, 15, 6.
POTTER, P.E., & PETTIJOHN, F.J. (1977). Paleocurrents and Basin Analysis. 2nd ed. Springer, New
York.
PRYOR, W.A. (1973). - Permeability-porosity patterns
and variations in some Holocene sand bodies. Bull. amer.
Assoc. Petroleum Geol., 57, p. 162-189.
RASMUS, J.C., & KENYON, W.E. (1985). - An improved petrophysical evaluation of oomoldic Lansing-Kansas
City Formation utilizing conductivity and dielectric log
measurements. SPWLA, 26th Ann. Log. Symp. Trans.,
paper V.
REINECK, H.E., & SINGH, I.B. (1975, 1980). Depositional Sedimentary Environments. 1st and 2nd ed.
Springer, New York.
RIVEST, J.F., DELHOMME, J.P., & BEUCHER, S.
(1991). - Marker-Controlled Picture SegmentationApplied
to Electrical Logging Images. Proc. SPlE 1451: Non-linear
Image Processing II, p, 179-190.
ROGERS, W.F., & HEAD, W.B. (1961). - Relationships
between porosity, median size, and sorting coefficients of
synthetic sands. J. Sediment. Petrol., 31, p. 467-470.
RUKHIN, L.B. (1958). - Grundzuge des Lithologie.
Akademie-Verlag, Berlin.
RUSSELL, W.L. (1951). - Principles of Petroleum
Geology. McGraw-Hill Book Co., New York.
SARMA, V.V.J., & RAO, V.B. (1963). - Variation of
Electrical Resistivity of River Sands, Calcite and Quartz
Powders with Water Content. Geophysics, April.
Schlumberger (1979). - Well Evaluation Conference,
Serralog 0 2003
Texture
Algeria.
Schlumberger Middle East S.A. (1981). - Well
Evaluation Conference, United Arab Emirates / Qatar.
Schlumberger Technical Services, Inc. (1982). Essentials of Natural Gamma ray Spectrometry
Interpretation.
Schlumberger Technical Services, Inc. (1983). - Well
Evaluation Conference, India.
SELLEY, R.C. (1970, 1978, 1985). - Ancient
Sedimentary Environments. I s t , 2nd and 3rd ed. Chapman
& Hall, London.
SELLEY, R.C. (1976). - An lntroduction to
Sedimentology.Academic Press, London.
SEN, P.N. (1980). - Dielectric and conductivity response of sedimentary rocks. SPE of AIME, paper 9379.
SEN, P.N. (1981). - Relation of certain geometrical features to the dielectric anomaly of rocks. Geophysics, 46,
p. 1714-1720.
SERRA, 0. (1972). - Diagraphies & Stratigraphie. In:
Mem. B.R.G.M., 77, p. 775-832.
SERRA, 0. (1973). - Interpretation Geologique des
diagraphies en Series Carbonatees. Bull. Centre Rech.
Pau -SNPA, 7, 1, p. 265-284.
SERRA, 0. (1974). - Interpretation geologique des
Series deltai'ques a partir des diagraphies differees. Rev.
A.F.T.P., 227, OCt., p. 9-17.
SERRA, 0. (1989). - Formation Microscanner lmage
Interpretation. Schlumberger Educational Services, SMP7028.
SERRA, O., & ABBOTT, H. (1980). - The Contribution
of Logging data to Sedimentology and Stratigraphy. 55th
Ann. Fall Techn. conf. SPE ofAIME, paper SPE 9270, and
in SPE J., Feb. 1982.
SERRA, O., & SULPICE, L. (1975). - Sedimentological
Analysis of shale-sand series from well logs. SPWLA, 16th
Ann. Log. Symp. Trans., paper W
SERRA, O., & SULPICE, L. (1975). - Apports des diagraphies differees aux etudes sedimentologiques des
series argilo-sableuses traversees en sondage. 9th Cong.
Intern. Sediment., Nice, theme 3, p. 86-95.
SNEED, E.D., & FOLK, R.L. (1958). - Pebbles in the
lower Colorado River, Texas - a study in particle morphogenesis. Jour. Geol., 66, p. 114-150.
THEYS, P., LUTHI, S., & SERRA, 0. (1983). - Use of
dipmeter in Carbonates for detailed sedimentology and
reservoir engineering studies.
TRASK, P.D. (1932). - Origin and environment of source sediments of petroleum. Gulf Publ. Co., Houston.
VATAN, A. (1954). - Petrographie sedimentaire. Ed.
Technip, Paris.
VISHER, G.S. (1965). - Use of vertical profile in environmental reconstruction. Bull. Amer. Assoc. Petroleum
Geol., 49, p. 41-61.
WADELL (1932). - Volume shape and roundness of
rock particles. Jour. Geol., 40, p. 443-451.
WADELL (1933). - Sphericity and roundness of quartz
particles. Jour. Geol., 41, p. 310-331.
Van WAGONER, J.C., MITCHUM, R.M., CAMPION,
K.M. & RAHMANIAN, V.D. (1990). - Siliciclastic Sequence
Serralog 0 2003
I ChaRter 3 1
157
Stratigraphy in Well Logs, Cores, and Outcrops. AAPG
Methods in Exploration Series, no7.
ZINGG, T (1935). - Beitrag zur Schofferanalyse:
Schweiz. Min. u. Pet. Mitt., Bd. 15, p. 39-140.
157
Introduction
Review of general geological concepts
Definition
By definition a sedimentary structure (internal), or feature, is "a structure formed either contemporaneously with
deposition: a primary sedimentary feature, or by sedimentary processes subsequent to deposition: a secondary sedimentary structure" (Bates & Jackson, 1980).
According to Pettijohn & Potter (1964) the structure
is an inherent property of a rock and a guide to its origin.
Whereas the texture deals with the grain to grain relations
in a rock, structure has to do with discontinuities and
major inhomogeneities. The structure is concerned with
the organization of the deposit - the way in which it is put
together - . Hence structures are the larger features that,
in general, are best studied in the outcrop rather than in
the hand specimen or thin section
'I.
They explain the local variations of the composition or
texture.
A sedimentary structure refers to megascopic morphological features. These features have been studied for
some time, as they are often visible to the naked eye.
They include the thickness and the shape of beds, their
internal organization, the nature of their surfaces, joints,
concretions, cleavages, and fossil content.
Primary sedimentary structures are generated by
either current velocity, impeding the action of the gravity,
and its evolution, (scour and erosional marks, ripple
marks, cross-bedding, wavy-bedding, graded bedding, or
biogenic activity (tracks, trails, burrows, rootlets,...), or
action of climatic or physical agents (mud-cracks, pits,
load casts, dikes, convolute bedding, slump structures).
"Some structures are texture dependent. Ripples
marks and cross-bedding for example, characterize only
those sediments which have a grain size in the sand
range" (Pettijohn et a/., 1964).
Importance of sedimentary structures
The primary sedimentary structures are particularly
important because they will reflect the hydrodynamic
Serralog 0 2003
conditions at the time of deposition (e.9, energy, type of
current,...). They constitute an important element of the
facies of a sedimentary unit and will lead to a better definition of the depositional environment.
As mentioned by Selley (1970), structures "unlike
lithology and fossils are undoubtedly generated in place
and can never have been brought in from outside".
Consequently, it is essential to detect them by analysing open-hole logs and more specifically dipmeter and
image tool data.
Classification of sedimentary structures
There are several different types of sedimentary structures, all of which can be characterized by the shape of
their surfaces and their internal organization.
The classification of structures can be based on the
time of their formation. They are defined as:
- predepositional, (formed before the deposition of a
bed). They correspond to features observed on the surface of a preceeding bed such erosion or impressions;
- syngenetic or primary or syndepositional,(contemporaneous with deposits). These structures contain information on physical, chemical or biological conditions, existing in a depositional environment during sedimentation.
They are subdivided into inorganic and organic structures, depending on their origin; inorganic structures are a
result of physical agents; organic structures are formed in
connection with an animal or plant organic activity (burrow; impressions; root traces);
- epigenetic or secondary or postdepositional, (formed
after sedimentation). These are often of chemical origin
and their occurrence reflects diagenetic phenomena; or of
physical origin resulting from tectonic deformation.
A second classification can also be suggested which is
based on the agents or processes, which have created
the sedimentary structures:
- physical such as action of gravity, influences of current or stress (ripple marks, tool marks, convolute-bedding, slumping, mud-cracks,...),
- chemical such as dissolution, concretions;
- biological such as burrows, tracks, trails, foot impressions, root traces,...
A third classification of sedimentary structures based
159
on their location can be suggested:
- external structures, that cover the size and the shape
of beds, thus the nature of their boundaries and the shapes of the lower and upper bedding planes;
- internal structures, relative to internal organization of
the bed : i.e. massive, laminated, graded-bedding, growth
structures (stromatolitic limestones,...).
The classifications proposed by Krumbein & Sloss
(1963), by Selley (1976), essentially combine the time
period and the agent, whereas the classification of
Pettijohn & Potter (1964), and Blatt et al., (1979) are
based more on the location of features. We have retained
the classification of Table 4-1, derived from Pettijohn &
Potter, because it adapts more closely to log analysis and
illustration.
above and below." This general term includes bed and
lamina (McKee and Weir, 1953).
A bed is "the smallest formal unit in the hierarchy of
lithostratigraphy units. In a stratified sequence of rocks it
is distinguishable from layers above and below ... A bed
commonly ranges in thickness from a centimeter to a few
meters" (Bates & Jackson, 1980). A single bed is a sedimentary unit formed under essentially constant physical
conditions and constant delivery of the same material
during deposition. It corresponds to a volume with a certain geometry defined by surfaces. It may be internally
layered, consisting of smaller units or laminae with their
own characteristics (Fig. 4-1).
Lithology
Depositional
Junits
Table 4-1
Classification of sedimentary structures
(from Pettijohn & Potter, 1964).
Geological objects
BEDDING EXTERNAL FORM
1. Beds equal or subequal in thickness: beds laterally uniform in thick
ness: beds continuous
2. Beds unequal in thickness: beds laterally uniform in thickness: bed!
:ontinuous
3. Beds unequal in thickness: beds laterally variable in thickness: bed!
:ontinuous
1. Beds unequal in thickness: beds laterally variable in thickness: bed!
jiscontinuous
BEDDING INTERNAL ORGANIZATION AND STRUCTURE
1. Massive (structureless)
2. Laminated (horizontally-laminated; cross-laminated)
3. Graded
4. Imbricated and other oriented internal fabrics
5. Growth structures (stromatolites, etc.)
BEDDING PLANE MARKINGS AND IRREGULARITIES
1. On base of bed
(a) Load structures (load casts)
(b) Current structures (scour marks and tool marks)
(c) Organic markings (ichnofossils)
2. Within the bed
(a) Parting lineation
(b) Organic markings .
3. On top of the bed
(a) Ripple marks
(b) Erosional marks (rill marks: current crescents)
(c) Pits and small impressions (bubble and rain prints)
(d) Mud cracks, mud-crack casts, ice-crystal casts, salt-crys.
tal casts
(e) Organic markings (ichnofossils)
Figure 4-1 - Subdivisons of a bed in sedimentation units with their own
characteristics (adapted from Blatt et al., 1980).
A lamina is "the thinnest recognizable unit layer of original deposition in a sediment or sedimentary rock, differing from other layers in color, composition or particle
size, ... (commonly 0.05-1.00 mm thick). It may be parallel or oblique to the general stratification.... Several laminae may constitute a bed." A lamina is produced by minor
fluctuations in fairly constant physical conditions. Rarely,
a lamina can be up to a few centimeters thick (Reineck
and Singh, 1975)
A sedimentation unit is characterized by:
- its shap- the n'ature of its bobndaries,
- its internal organization.
BEDDING DEFORMED By PENECONTEMPORANEOUS PROCES
SES
1. Founder and load structures (ball-and-pillow structures, load casts)
2. Convolute bedding
3. Slump structures (folds, faults, and breccias)
4. Injection structures (sandstone dikes, etc)
5. Organic structures (burrows, "churned" beds, etc)
Primary sedimentary structures
Shape of sedimentation unit
One must stress that a bed or a sedimentation unit, is
that thickness of sediment which was deposited under
essentially constant physical conditions" (Otto, 1938),
separated from other under- and over-lying beds by physically and visually more or less well-defined bedding planes, "made evident because of the unlike texture or com"
Stratum. bed. lamina descrbtion
A stratum is a "tabular or sheet like body or layer of
sedimentary rock, visually separable from other layers
160
Serralog Q 2003
position" (Pettijohn & Potter, 1964).
The parameters defining the shape of a sedimentation
unit are:
- thickness;
- bedding planes;
- lateral dimensions.
reduction conditions at the sea floor explaining the absence of the benthonic fauna and related bioturbation.
The analysis of the vertical evolution of bed thicknesses within the same lithology can indicate a change in
rhythm of sedimentation. These changes can be used as
time-markers.
Bed thickness
Bed thickness can vary from a few millimeters to several meters (Campbell, 1967). A bed can be massive or
internally finely stratified, formed by a succession of finer
units (laminations). A lamination results from minor fluctuations in the physical conditions, which prevailed in the
depositional environment. They are particularly expressed
by changes in grain size (sand to silt) and sometimes in
composition (quartz to clay mineral). The thickness of a
lamination is measured in millimetres and generally does
not exceed several centimetres (Fig. 4-2).
0
s
Figure 4-3 -Approach of log normality of thickness of turbidite sandstone beds (from Scoff, 1966).
Surfaces and beddina Dlanes
v)
v)
E
Y
.-0
c
c
Figure 4-2 - Strata, laminae and beds as a function of their thickness
and vertical resolution of the logging tools for comparison.
Knowledge of the thickness of beds is important
because it is sometimes related to the granulometry or to
the depositional mode. For example, in sandy turbidites
and volcanic ashes the thickness of the beds and the size
of the grains are related and decrease in the direction of
flow (Scheidegger & Potter, 1971). This is therefore a
means to distinguish proximal from distal deposits. In another way, Scott (1966), after Schwarzacher (1953), points
out that the thickness of individual beds pertaining to the
same kind of deposit are distributed following a straight
line on a probability-logarithmicgrid (Fig. 4-3).
According to Pettijohn (1975) the existence of laminations in marine environments is either the indication of
very fast deposits, below the zone of wave action, or of
Semlog 0 2003
The form and attitude of strata and structures are defined by the shape of their surfaces. Subdividing strata into
beds requires the identification of bedding surfaces.
These surfaces should be distinguishable by textural or
compositional changes from one bed or lamina to another.
Bedding surfaces have no thickness, but have areal
extents equivalent to the beds with which they are associated. Consequently, the geometry of a bed depends on
the relative disposition of its two boundaries.
A surface is a "two-dimensional boundary between
geologic features such as formations or structures." Its
shape can vary from a plane to a very irregular surface. A
perfect plane is defined by three non aligned points in the
space. Irregular surfaces can be cylindrical, conical,
undulating or warped. They can be shaped like a spoon or
very irregular like a stylolitic boundary. Consequently,
layers can have different shapes according to the nature
of their bounding surfaces, as shown in Figure 4-4
(Campbell, 1967). Bed forms are determined by stream
power and grain size.
As Campbell (1967) has correctly analysed, the surface of a bed represents a surface of non deposition, or corresponds either to an abrupt change in the condition of
sedimentation (variation in the energy of environment) or
to a surface of erosion. Usually the upper surface of a bed
constitutes the lower surface of the following upper layer.
Hence the characterization of beds depends on the recognition of their surfaces.
161
Next Page
J
; 4-4 ~
- ~i~~~~~
~
~
~ different shapes that can be acquired by
beds and laminae, and the corresponding descriptive terms
(from Campbell, 1967).
F
Figure 4-5 - Relationship between the thickness and the Width Of sedimentary bodies, giving an external morphology of these deposits
(from Krynine, 1948).
Figure 4-6 - Geomorphology and characteristics of bed-load, mixed-load, and suspended-load stream channel segments and their deposits
(from Galloway, 1977, and in Galloway & Hobday, 1983, fig. 4-13).
Lateral dimension of beds
Taking into account the bed composition and its thickness, the nature of the bedding planes, and the bed frequency and depositional environment, it is possible to
estimate its lateral extent as suggested by Krynine (1948)
(Fig. 4-5). For instance, if several beds show planar and
parallel boundaries over a thick interval, one can assume
a large extent of each bed composing the interval.
If the environment is known, as in a fluvial system (Fig.
4-6) a better link can be done between the bed thickness
and its lateral extent.
162
Nature of bed boundaries
The transition from one layer to another can either be
abrupt or gradual. In the first case the boundary is well
defined and agrees with the bedding planes.
The boundary is conformable if it corresponds to a
short break in sedimentation without modification of the
depositional sequence or without erosion, and the beds
remain parallel.
The boundary will be unconformable if it corresponds
to a break in sedimentation, followed by a change in the
sequence of deposition underlined either by an erosional
or by a lateritized surface (if continental), by a truncation
and possibly by a change in dip magnitude and azimuth
Serralog 0 2003
Previous Page
Structure
I Chapter4 I
163
(Fig. 4-7).
Fininp-up
sequence
Composite
bedset
Simple
bedset
ONLAP
TRUNCATION
WWNLAP
WWNLAP
DOWNLAP
TRUNCATION
WWNLAP
TRUNCATION
J
TRUNCATION
WWNLAP
WWNLAP
TRUNCATION
Figure 4-9 - On the right, different bedding due to the change in composition or
in texture :shape, size and orientation of
the grains. These types of arrangement
affect the pore geometry and
consequently the petrophysical properties
of each unit of sedimentation
(adapted from Griffiths, 1961).
On the top-left, illustration of size change
through thin section.
TRUNCATION
TRUNCATION
HORIZONSOF BURROWS, ROOTS. ETC
GRAVEL
MUDSTONE
SANDSTONE
ORGANIC DEBRIS
BEDDING SURFACE
Figure 4-7 - Typical bedding surfaces and how to recognize them
(adapted from Van Wagoner et al., 1990).
The surface of the bed boundary can be planar, ondulated or irregular with load casts (Fig. 4-8). The latter
structures are generated by the differential loading of a
waterlogged sand on an unconsolidated plastic mud.
They are common feature of turbidite deposits, yet they
also occur in deltaic and fluvial sediments.
change of charateristics.
The group of beds is called simple if it consists of two
or more superimposed beds, having the same general
characteristics (mineralogy, texture, internal structure)
and surrounded by other beds of a different nature.
A sequence of beds is called composite if it consists of
a group of layers, having different composition, texture,
and internal structure (Fig. 4-10).
1
Figure 4-8 - Example of undeformed bed (the two beds at the top), bed
with load casts on its lower surface due to deformation of the
mud/sand interface, and segmentation of bed in pseudo nodules or
pillows.
The bedding surface can be associated to changes in
composition, in grain shape, grain orientation, grain size
and cementation (Fig. 4-9).
In the case of a gradational transition between beds
the boundary is not clearly defined and thus not visible. It
then agrees with a sequence, which is either granulometric (normal or reverse), or mineralogic, or both (sand to
shale, or shale to silt to sand).
Groups of beds (“bedset‘;)
McKee & Weir (1953) have defined a group of beds as
a succession of layers or laminae, essentially conformable, separated from adjacent sedimentary units by a surface of erosion or by a break in deposition, or a sudden
Serralog 0 2003
J
Figure 4-10 - Schematic illustration of bedding terminology: lamina,
bed, simple bed set, Composite bed set and bedding type
(from Reineck 8, Singh, 1975).
Internal oraanization of beds
Several types of internal organization can be recogni163
zed. They depend on the type of transport, the stream
power or the energy in the environment and the water
depth.
Massive bedding
-
A bed can be apparently homogeneous, (i.e. without
any visible variation due to textural changes or sedimentary features). This corresponds to a constant condition of
sedimentation, without apparent stratification either due to
the absence of current ripples or even to a massive sudden sedimentation. It can also correspond to an intensive
bioturbation, which has completely destroyed all traces of
stratification, or to diagenesis especially in carbonates.
Genuine depositional massive bedding is often seen in
fine-grained, low energy environment deposits.
Direction of transport current
Laminated bedding
As previously mentionned, a single bed can be internally finely layered, made up of smaller units laminae. A
lamina is produced as a result of some minor fluctuations
in rather constant physical conditions. A bed shows therefore stratifications either parallel, oblique or cross-bedded. These laminations are control by the stream power
(Fig. 4-11).
Figure 4-12 -At the top :terms used in ripple description. L = ripple
length, H = ripple height, I1 = horizontal projection of stoss side, 12 =
horizontal projection of lee side. In the middle :dynamic mechanism
for ripple formation. At the bottom :flow pattern over a lee face of a ripple with velocity distribution, flow separation and the three major zones
on the lee side (from Reineck & Singh, 1975).
Ripples and other bedforms are present as undulations on a non cohesive surface. They are the results of
the interaction of moving fluids (water and air) - currents
or waves, wind - on a sediment surface. Interaction of
these fluids with a non cohesive material composed of
grains produces small ripples, megaripples, plane beds or
antidunes as a function of the flow regime (Fig. 4-13).
L
3”
P
E
e
j;
K
6
MEDIAN FALL DIAMETER
Figure 4-11 - Types of laminations as a function of the stream power
and the grain size (data from Guy et al., 1966, in Allen, 1968, and in
Reineck and Singh, 1975).
Ripples. foresets. plane beds and antidunes
A ripple is described in terms of its size and shape. It
is composed of a crest, or summit, and a trough (Fig. 412).
164
Figure 4-13 - Relationship between stream power, median fall diameter, bed form and sedimentary structure (after Simons et al., 1965).
In a low flow regime resistance to flow is large and
sediment transport is nihil or relatively small depending on
the coherence of the material. Water surface undulations
are out-of-phase with the bed undulations. The bedform is
either small ripples or megaripples or a combination of
Serralog 0 2003
both. Individual grains move up the back of the ripples
and avalanche down the lee face (Fig. 4-12).
In an upper flow regime resistance to flow is small and
sediment transport is large. Water surface undulations
and bed undulations are in-phase, and segregation of bed
material is negligible. The usual bed forms are plane beds
and antidunes. Individual grains roll almost continuously
downstream in sheets a few centimeter thick.
As well known, the mechanics of particle motion are :
suspension, bouncing or saltation, and rolling (Fig. 4-14).
LAMINAR FLOW
Low Reynolds numbers
R < 1for particles moving in a fluid
TURBULENT FLOW
High Reynolds numbers
R 5 1 for particles moving in a fluid
A
Figure 4-15 - Illustration of laminar and turbulent flows.
In laminar flow stream lines remain distrinct. Flow direction at every
point remain unchanged with time. Current velocity is inferior to a few
tenthes of mm/s depending on the width and the depth of a channel.
In turbulent flow the flow lines are confused and heterogeneously
mixed, generating eddies and vortices.
Figure 4-14 - Particle motions : A = suspension; B = bouncing or
saltation; C = rolling. (adapted from Selley, 1976).
The particle motion through a fluid is expressed by two
numbers : the Reynolds number, R,and the Froude number, F:
where V is the velocity of the particle, d is the diameter
of the particle, p is the density of the particle, p is the viscosity of the fluid, U is the average current velocity, g is
the acceleration due to gravity and D is the depth of the
channel.
For low Reynolds numbers, the fluid flow is laminar, it
means flow lines are parallel to the bounding surface. For
high Reynolds numbers the flow is turbulent, generating
eddies and vortices (Fig.4-15).
A Froude number of 1 separates two distinct types of
fluid flow regime in open channel: lower flow regime for
F < 1, upper flow regime for F > 1.
The stream power is the product of the fluid velocity by
the shear stress zo which is equal to p S wherey is the
specific weight of the fluid and sediment, D is the stream
depth and S is the slope of the energy gradient. The relationship between current velocity and water depth is
reproduced in Figure 4-16.
The critical flow velocity required for transportation,
deposition and erosion as a function of grain size is illustrated by Figure 4-17 based on works realized by
Serralog B 2003
Depth (cm)
Figure 4-16 - Relationship between current velocity and water depth, in
relation to bedforms indicated by the Froude number
(from Simons & Richardson, 1962).
165
100.0
0
3
Figure 4- 17 - Relationship between fluid-flow velocity and sedimentary
processes such as transportation, deposition and erosion as a function
of grain size (from Hjulstroms in Sundborg, 1956).
Hjulstroms (in Sundborg, 1956). For transportation, as
one might expect, the critical velocity increases with grain
size. For erosion, one must take into account the cohesion of the material. For cohesive beds, such as shales, a
higher velocity is required to start erosion. This is related
to their resistance to friction. This effect is responsible for
the preservation of delicate clay laminae, termed flasers,
in tidal deposits. At the opposite, non-cohesive beds as
silt or sand, the erosion starts for a lower velocity.
The different sorts of fluid-flow - traction currents, density current, suspension - generate different types of sediments with their own types of sedimentary structures.
Traction currents, for instance, generate ripples, crossbedding, foreset bedding, dunes, herringbone. Sediments
linked to density currents, which are a combination of
traction and suspension, are characterized by mixtures of
sands, silt and clay, and typically show graded bedding
with generally lack of cross-bedding. Table 4-2, from
Selley (1976), links the sediment types to the sedimentary processes which have generated them.
Table 4-2
Relationship between sediment types and sedimentary
processes (modified from Selley, 1976).
I Sedimentary processes I
Traction current
Density current
Environment
Type of deposits
ISubaerial
Subaqueous
IPredominantlycross-bedded sands
Subaerial
Subaqueous
INubes ardentes
IGraded sands, silts and clays
I
Velocity repartition from center to the edges
Velocity repartition from the bottom of a channel
1
Subaerial
Subaqueous
Depthofthachannel
ILoess
INepheloid clays
Particles are transported following different modes as
a function of their size and the stream power (Fig. 4-18).
One must also take into account the repartition of the
fluid velocity in a channel as a function of the proximity of
the channel edges, the channel depth and the type of
flow: laminar or turbulent (Fig. 4-19).
166
Figure 4-18- Modes of transport ofparticles as a function of their size
(from von Engelhardt, 1977).
Generally unstratified poorly sorted
Mass gravity transport
Suspension
I
Transport power mtgp (cm)
Bottom of tha channel
Figure 4-19 - Influence of the proximity of the edges and bottom of a
channel on the fluid velocity.
The particles carried in suspension in a turbulent flow
have a distribution which is represented by Figure 4-20.
Serralog 0 2003
Chapter 4
I 167
Height
Y/h
asymmetrical wave ripples. Foreset
laminae with off-shoots.
Height a b o v e
stream bed (m)
straight-crested small-current ripples
Cross-bedding planar.
Relative concentration
Figure 4-20 - Relative concentration of suspended particles in a turbulent flow as a function of the grain size. v: settling velocity; n,,.' number
of particles at height y; na :number of particles at height a < y; y:
height above the bottom; h: water depth; p f fluid density, 70: maximum
shear stress (at the top from von Engelhardt, 1977; at the bottom from
Hjulstroms, 1939).
Ripple marks result from the interaction of waves or
currents with a non cohesive sediment surface. They
generally trend at right angles or obliquely to the direction
of fluid flow. Their shape and size vary considerably as a
function of the current velocity. Ripples are asymmetrical
wave-like beds occurring in fine sands subjected to gentle traction currents. Migration of ripples generates crosslaminated deposits. Several types of ripples exist. They
are reproduced in Figure 4-21.
Ripples do not form in clay nor in coarse sand or gravel. They are restricted to coarse silt and sand with a grain
size of less than about 0.6 mm in diameter (Reineck &
Singh, 1975). Their length ranges between 4 and 60 cm
and their height varies from 0.3 and 6 cm depending on
the grain size (higher if coarser). They occur in dunes,
rivers and deltas.
Megaripples range between 60 cm and 30 m in length,
and 6 cm to 1.5 m in height. The grain size is more than
0.6 mm.
Foreset laminae result from deposition on the lee side
of a ripple by the progradation process. Foreset laminae
may be angular, tangential or concave as a function of
velocity, bed shear stress, depth ratio and particles in
suspension (Fig. 4-22). At low velocities sediment
transport is by bedload movement. The grains move
Serralog 0 2003
undulatory small ripples. Cross-beding
weakly festoon-shaped.
lingoid small ripples. Cross-bedding
strongly festoon-shaped.
h a t e megaripples.
Figure 4-21 - Block diagrams illustrating different types of ripples
(from Reineck & Singh, 1975).
along the bed and are deposited on the upper part of the
lee face from where they move downslope under gravitational force producing a planar slip face. With increasing
167
velocities, a greater proportion of grains is maintained in
suspension, carried beyond the lee face, and deposited in
the form of bottomset and toeset. Angular contact is replaced by tangential contact and the angle of repose of the
lee face is reduced.
Figure 4-22 - Factors controlling the shape of the foreset laminae
(modified from Jopling, 1965).
Cross-beddinq
It is probably the commonest type of bedding in clastic
deposits.
A cross-bed is a single sedimentation unit consisting of
foreset laminae inclined to the principal surface of sedimentation. It is separated from the adjacent units by a surface of erosion, non deposition, or abrupt change in character. Its thickness may vary from a few millimeters to
tens of meters.
Three principal types of cross-bedding exist (Fig. 423).
- Planar or tabular cross-bedding corresponds to
cross-beds whose bounding surfaces form more or less
planar surfaces (Fig. 4-24).
I
Figure 4-24 - Typical tabular planar cross-beds
(adapted from Allen, 1963).
- Wedge-shaped cross-bedding corresponds to crossbeds that thin out.
- Trough cross-bedding corresponds to cross-beds
having a trough shape and Whose bounding surfaces are
Curved (Fig. 4-25).
Set
(Tabular)
Set
Figure 4-25 - Typical trough cross-beds (adapted from Allen, 1963).
Coset
set
(C)
Trough
Cross-Bedding
Figure 4-23 - Terminology for cross-bedding
(adapted from McKee & Weir, 7953).
168
Migration of small-current or wave ripples can generate small-scale cross-bedding. Individual units vary from a
few millimeters to 5 cm in thickness. They are usually
trough shaped (Reineck and Singh, 1975).
Large-scale cross-bedding correspond to individual
units usually more than 5 cm thick. They can be planar or
trough-shaped and are produced by mega ripples, sand
dunes, long shore bars, etc.
Their shape and slope depend on the velocity, bed
Serralog 0-2003
Structure
shear stress, depth ratio and sediment type. Figure 4-22
illustrates the shape of the foreset laminae as a function
of these factors (Jopling, 1965).
The type of internal structure of the sedimentation
units allows their identification (Fig. 4-26), and the determination of their origins.
Chapter 4 ' 169
Cross bedding with flarera
simple
bifurcated
Flaser bedding
wavy
bifurcated
wavy
Wavy bedding
with thick Ienrer
connected
with fiat lenses
Lenticular bedding
with thick iensea
single
with flat lenses
Figure 4-26 - Recognition of the sedimentation units based on their
internal structure (from Potter & Pettijohn, 1977).
Scour-and-fill cross bedding and trough cross-bedding
result from erosion of a sediment surface by the current
flowing over it and refilling of the trough or depression with
sediment when the current velocity decreases. During the
filling process, the sediment transported down the steep
upcurrent slope. This produces inclined laminae that gradually flatten during the latter stages of scour fill.
Figure 4-27 - Schematic classification of flaser, wavy and lenticular
bedding (from Reineck & Singh, 1975).
Chevron cross-beddina or herrinabone cross-bedding
This type of cross-bedding relates to cross-bedding
that dips in different or opposite direction in alternating or
superimposed beds, forming a chevron or herringbone
pattern (Fig. 4-28). This type of cross-bedding is typical of
tidal environments.
Flaser bedding
During high tides, mud is commonly deposited across
the ripple crest and into the troughs. Flaser bedding
results when this mud is preserved in the troughs, but little or no mud is preserved on the crests. Wavy or bifurcated flasers result when there is some mud preserved
across the crests, and simple flasers result when there is
no mud preserved on the crests (Fig. 4-27).
Wavy bedding forms when alternating mud and sand
layers occur as continuous layers (Fig. 4-27). Their genesis requires conditions where the deposition and preservation of both sand and mud are possible. Lower current
energies are necessary to avoid erosion of the previous
deposits.
Lenticular bedding forms when the ripples or sand lenses are discontinuous and isolated in all directions (Fig. 427). It is produced where there is meager sand supply and
under conditions more favorable for the deposition and
preservation of mud than of sand.
These structures are generally observed in tidal environments.
When the current velocity increases, it starts to truncate or erode the previous deposits. The surfaces that are
generated can have different shapes as a function of the
flow type. Consequently, the shape and organization of
beds and laminae strongly depend on the current or wave
activity. Terminology has been proposed to describe this
organization (Reineck and Singh, 1975).
Serralog 0 2003
J
Figure 4-28 - Example of chevron or herringbone cross-bedding. They
can be recognized only in 3-13 (from Reineck & Singh, 1975).
Graded bedding
Sedimentation units characterized by a gradation in
grain size are termed graded beds.
Pettijohn (1957) described two main types of grading.
In the first type there are no fines in the lowest part of the
graded bed (Fig. 4-29a). In the second type the fines are
distributed throughout (Fig. 4-29b).
169
Well logging and Geology
Coorse iaikgrading
a
b
Figure 4-29 - Different types of graded beding
(adapted from Pettijohn, 1957).
Dzulynski and Walton (1965) recognized 8 types of
graded bedding. They are illustrated by Figure 4-30.
Graded beds commonly occur in thick sequences of
flysch type of sediments. Commonly, graded beds are
separated from each other by parallel-bedded clayey
layers, which represent the product of normal sedimentation interrupted by turbidity currents. Bouma’s sequence
is an example of graded bed (Fig. 4-31).
In fact, there are two types of vertical evolution of grain
size in a bed : namely normal (fining-upward) or reversed
(coarsening-upward)graded beddings (Fig. 4-32). Normal
graded bedding is generally underlined by an abrupt
contact at the base.
Figure 4-31
- The various types of graded bedding (from sellex
1976).
Figure 4-32 - Typical graded bedding in the Bouma’s sequence and its
interpretation in terms of flow regime.
-I
Figure 4-30 - The 8 different types of graded beds
[from Dzulynski & Walton (1965).
170
Bioaenic activitv
Other typical sedimentary features are linked with biolgenic activity, either animal or vegetable. Tracks, trails,
burrows, and rootlets are the most important biogenic features. These trace fossils are defined and named by the
characteristics of their makings, rather than the organisms
responsible for making them. In some cases, the origin of
trace fossils is unknown (Fig. 4-33).
Serralog 0 2003
Next Page
Deep marine
weight of overlying material, accumulation of moisture,
earthquakes or regional tilting. These secondary structures appear as small deformations within the deposit, such
as slides, slumps and convolute bedding (Figs. 4-35 and
4-36).
*
Mwsmentdlrechon
Figure 4-33 - Illustration of biogenic activity, based on Seilacher, 1964,
1967; Rodriguez 8 Gutschick, 1970; Heckel, 1972 (from Selley, 1976).
l m
Figure 4-35 - Typical slump structure (from Selley, 1976).
Action of climatic or phvsical aaents
Current
c
Mud cracks, load casts and pits are typical features
generated by climatic (shrinkage) or physical forces (Fig.
4-34).
Figure 4-36 - Sketch of convolute bedding (from Selley, 1976).
Disconformities
Certain surfaces are related to sea-level changes.
Disconformities (breaks in sedimentation caused by erosion or non deposition) can result from transgression,
regression, the formation of hardgrounds, levels of
condensation or other diagenetic effects.
1
Figure 4-34 - Sketch of mud cracks generated by shrinkage of clay, silt
or mud, generally in the course of dfying (from Selley, 1976).
Dish structures
Dish structures consist of small meniscus-shaped lenses (4 to 50 cm long and one to a few centimeters thick)
generally found in sandstone. They are created by water
escaping during elutriation of clay soon after deposition of
the sand. These structures are common in turbidite deposits.
Flame structures
Flame structures correspond to "wave- or flame-shaped plumes of mud that have been squeezed irregularly
upward into an overlying layer. It is probably formed by
load casting accompanied by horizontal slip or drag"
(Bates & Jackson, 1980).
Secondary sedimentary structures
When shear stress is applied to a sloping surface, the
mass movement that results can create secondary sedimentary structures. This stress can be caused by the
Serralog 0 2003
-
Intrusions Dikes
They correspond to igneous intrusions that cut across
the bedding of sedimentary rocks or across other rock
types (cf; Fig. 2-1).
Diapiric structures
Salt or clay uplifts, sometimes igneous intrusions,
generate anticlinal structures called domes. The overlying
sedimentary strata are ruptured by the squeezing-out of
plastic material.
Determination of rock structure
Traditional approach
As the structures are megascopic features, in the
range of millimeters to a few meters, they are better described and studied on outcrops or by core observations.
For a correct description and analysis a 3-D observation
is absolutely necessary.
Due to the lack of vertical resolution of surface seismics, generally only diapiric structures can be recognized.
In wells, thanks to the vertical resolution of dipmeters
and imagery tools, most of the structures will be detected
171
Previous Page
Determination of rock structure
Well logging approach
and analyzed.
The interest of these well logging measurements will
now be illustrated by examples. Outcrop or core photographs of similar structures will be presented at the same
time to demonstrate the validity of this approach..
The link beheen structural geological features and
well logging parameters is indicated in Table 4-3.
Table 4-3
Link between the structural geological features and the dipmeter and imaging data.
Logs to use for detecting the phenomenon
Shape : lateral variation in thickness
External form of bed
?
On base
(lower face)
Bedding surface
features
fLoad marks
origin
Erosion
Truncations
Organic
origin
Trails
On top
(upper face)
4
f
Massive or homogeneous
Hetero!geneous
Ripples
Horizontal
Laminations
Oblique (foresets)
Internal organization
of beds
Graded bedding
Growth
Physical
origin
Post depositional
deformation
I
Physico-chemical
origin
Organic
I72
Normal
Inverse
Load marks
Convolute bedding
Injections
Fractures
Falling
Slumps
Folds
Faults
1{
svlOlites
Bird's eyes
Dish structure
Burrows
Serralog 0 2003
Structure
As previously indicated, the vertical resolution of the
dipmeter tools and more especially of imaging tools
allows recognition of features as thin as a few millimeters
thick (Table 4-4).
I Chapter 4 I
173
Consequently, most of the sedimentary features will be
detected by these two kinds of tools.
Table 4-4
Vertical resolution of the well logging measurements.
.Electrical
. . . . . .'. .images
.... . . I
Dipmeters - - . . - . - .w--1
Acoustic images
hicrolog w
b
Electromagnetic (GH) t+
Micro SFL W
Litho - density - - - W
Radioactivite - - - - Acoustic 1-b
Azimuthal resistivity
b
-I
Laferolog +
Spherical log t-w
Neutron - - - t,
Spectrometry +
Array induction
Dual ihduction - - SP
i
I
I
I
I
0.0005
0.001
I
I
I
I
0.005
0.01
I
I
0.05
I
I
0.1
I
I
0.5
I
I
1
-
b
VSP b-I
1
I
5
*
I
I
10
Scale (m)
O m
0 3 Fractures
a > * -----------------
0.0
0
&
5
r;/l
- - - Small sedimentary features - - - - - - - - Large sedimentary features - - - - - Sedimentary bodies - - Faults
b
. . . . . Higher sensitivity than resolution
Serralog Q 2003
- - - - - - Higher resolution techniques
173
Bed description
Bed shaue and beddina planes
The shape of a bed is determined by the attitude of its
limiting surfaces. Most of the individual surfaces penetrated by a well can be approximated by planes (excluding
trough cross bedding, stylolites and load casts) because
the size of the borehole is small compared to their radius
of curvature. But, it happens that the surface is not really
planar even at the scale of the borehole diameter.
The attitude of a structural plane is its position in three
dimensions, expressed quantitatively by its strike and dip.
The strike of a surface is the direction of its intersection
with a horizontal plane. The dip of a surface is the angle
that it makes with a horizontal plane, measured in the vertical plane perpendicular to the strike.
When the surfaces are parallel over a representative
thickness one can suppose even, parallel and continuous
surfaces and a relatively wide extent of the beds (Fig. 437).
GEODIP presentation they correspond to the computation
of four planes by combination of the resistivity curves
three by three (1-2-3, 2-3-4, 3-4-1, 4-1-2). In the LOCDIP
or SYNDIP program presentation they are emphasized by
a wavy symbol.
Figure 4-38 - Example of even, non parallel and continuous boundaries. a :shale bed; b :sandstone beds; c :cross-bedding; d :unparallel
boundaries; e :parallel boundaries. Observe the erosion at the base of
bed b.
Examples of discontinuous wavy boundaries are given
in Figure 4-39.
-I
Figure 4-37 - Example of even, parallel and continuous sutface boundaries. Observe the very good consistency of dips. Two higher dips are
related to wrong correlations.
Surfaces, which may be flat, but non parallel, at the
scale of the borehole, are shown in Fig. 4-38. They will
probably remain non parallel at the scale of the layers.
The upper and lower boundaries of each bed can show
different dip magnitudes and sometimes different azimuths suggesting non parallel boundaries. Examples of
wavy, non planar surfaces are given by Fig. 4-38. In the
174
J
Figure 4-39 - Example of discontinuous wavy beds orlenses. The
core description on the side confirms the HDT interpretation.
Serralog 0 2003
Structure
Bed thickness
DIPS
I
Chapter 4
I
175
CORRELATIONS
Standard tools do not allow a precise determination of
the bed thickness due to their lack of vertical resolution
and to the fact that the borehole deviation and the dip and
azimuth of the bedding surfaces are not known.
This is not the case with dipmeter and image tools.
As soon as laminae are thicker than 1 cm for dipmeter
tools and 5 mm for imagery tools, these tools represent
the best and most accurate approach for determination of
the bed or lamina thickness. This is due to the fact that the
borehole drift and the dip of bed can be taken into account
for the thickness computation. This determination is not
possible from core as its orientation and the borehole
deviation are generally not known precisely.
The bed thickness computation is explained on Figure
4-40.
North
Figure 4-40 - Graph explaining the bed thickness computation.
Figures 4-41 to 4-44 illustrate different bed thicknesses as they can be observed on dipmeter curve displays
and images.
Nature of bed boundaries
The standard logs allow the detection of surfaces limiting beds of different lithologies or textures, consequently
the apparent thickness of each bed, but they do not allow
the determination of the real thickness of the nature of the
surface.
On the contrary, dipmeters may allow the determination of the nature of the surfaces limiting beds as illustrated by previous figures.
But, the best way to determine the nature of the surfaces and their origin is from image observation. Surfaces
can be classified based on several parameters such as
resistivity contrasts and values, proximity between surfaces, consistency of dip magnitude and azimuth, etc.
Figure 4-45 illustrates several types of bed boundaries as
they can be observed on a Formation Microscanner
image. One can imagine the interest to recognize these
features for facies definition and reconstruction of the
environment.
Serralog 0 2003
Figure 4-4 I - Thin beds well detected from the HDT resistivity curves.
One can observe typical thickening-up sequences of resistive beds.
This suggests a distal turbidite deposit.
Erosions, truncations, load casts, hardgrounds, etc.
will be illustrated during dipmeter or images examples
showing internal organization of beds.
Internal oraanization of beds
Curve homogeneity, especially on dipmeter or image
data, may reflect a homogeneous massive bed as illustrated by Figure 4-46.
Remark
A homogeneous bed must not be confused with an interval without
curve activity due to poor choice of EMEX current. In this case the resistivity curve is saturated in the very high or the very low resistivity region
(compare the logs in Fig. 4 4 7 recorded on same interval, but with two
different EMEX). The EMEX current is the current emitted by the whole
sonde. It focusses the current emitted by the button electrodes. It is
generally chosen to have the best contrast either in low or in high resistivity environments. In the SHDT and image resistivity tools it can be
automatically adjusted.
175
Figure 4-42 - Very thin beds very well detected from the SHDT resistivitv curves. The bed thickness varies from less from 1 cm to 5 cm.
and ‘‘fedpatterns which suggest
shows a lot Of
This
fan progradation with direction of transport towards NE, and draping of
the previous fan by the following one.
Figure 4-44 - Recognition of thin beds by various log measurements,
On the left microSFL, then FMs images,, on the right EPT measurement (eaff and tpl).
Figure 4-43 - Segmentation of one meter FMS images in beds
(from Trouiller et al., 1989).
Laminations are generally well detected by dipmeter
resistivity curves (if the EMEX has been well selected) as
illustrated by Figures 4-41 and 4-42. GEODIP and LOCDIP processings allow the detection of each bed boundary and, consequently, one can recognize parallel laminations, foreset or cross-bedding (Fig. 4-48) which is difficult
to do from results of processing using correlation interval.
176
1
Figure 4-46 - Example of homogeneous bed seen on HDT
Logic based on the curve activity (FBR for frequency of
breaks) and the density of correlations (DCL) (see Fig. 449)aHows the classifications Of the laminae (Table 4-5).
Serralog 0 2003
Structure
FMS
IMAGE
NATURE OF
BOUNDARlES
HOLE
DMFf
mm
Chapter 4
177
OOllllEWlCIMY
WARP. PLANAR
SHARP, CONVEX
SHARP, CONCAVE
WARP. CONVEX
SHARP. WAW,
SYMMETRICAL
SHARP. WAW,
ASYMMETRICAL
SHARP. W A W
WAW. GRADATtONAL
SHARP. IRREGULAR
Figure 4-48 - Example of trough cross-bedding, foreset bedding, erosional surfaces and sand lens detected on HDT curves.
Table 4-5
Logic applied to the dipmeter curves for determination of
the type of internal organization.
ABRUPT
IRREGULAR
ABRUPT
PLANAR
Figure 4-45 - Several types of bed boundaries can be observed on
borehole wall images.
I
--I
Figure 4-47 - Example of change of curve appearence due to change
of the EMEX current. According to the choice of the EMEX value, the
curves are saturated in low or high resistivities.
Serralog Q 2003
This logic, explained by Figure 4-49, has been used in
the SYNDIP processing of which a result is reproduced in
Figure 4-50 (see further for explanation of the SYNDIP
program).
I77
178
Well logging and Geology
Figure 4-49 - Typical curve activity allowing the recognition of the internal organization of beds. No curve activity (flat curves) is interpreted as
indicating massive and homogeneous bed. High frequency of breaks
without correlations is interpreted as indicating a heterogeneous interval. High frequency of breaks with high density of correlations is interpreted as indicating either laminations or cross-bedding if the resistivity
contrast is low, or as succession of beds of two lithology types if the
resistivity contrast is high.
But the best tools for recognition of the internal organization of beds are imagery tools as for the first time we
“see” the formations. This will be illustrated by the following figures.
J
Figure 4-50 - Example of results of SYNDlP processing of resistivity
curves.
Figure 4-51 - FMS image illustrating cross-bedding with truncations.
On the left, raw images in dynamic presentation; on the right, same
images with superposed correlations for dip computation.
As known, processing of the recorded data provides
images which look like core photographs, with the following characteristics:
- high vertical resolution (0.2 in. [5mm])
178
- coverage of 80 % of the borehole wall in a 8 in. diameter (for FMI tool of Schlumberger)
- complete coverage of the borehole wall for ultrasonic
images
- very large dynamic range for electrical images - from
less than 0.1 to more than 10,000 ohm-m
- high sensitivity, allowing detection of very thin (fractures)
or small events (vugs, pyrite crystals) for electrical images
- high sampling rate - one sample each 0.1 in. [2.5mm]
for electrical images
- low sensitivity to heavy mud, borehole ovalization and
rugosity
- image enhancement allows detection of very minute
variations of conductivity.
Serralog Q 2003
Next Page
Structure
I Chapter4 1
179
J
Figure 4-53 - Other example of planar cross-bedding (foresets) with a
layer showing contorted bedding, with shale drape and shale protuberance towards the top of the image (courtesy of Halliburton).
Figure 4-52 - Example of cross-bedding showing different dips and azimuths indicating different transport current directions. The depth scale
is in meters. Several thin shale layers (more or less horizontal black
brown features) are intercalated. Observe small more or less vertical
features at 180" difference in azimuth. They correspond to small
drilling induced fractures (courtesy of Schlumberger).
Recommendations
Image interpretation is a multistep process of observation, description, interpretation by reference to previous
examples and experience, and implication or exploitation.
To be correctly interpreted images must have the
same horizontal and vertical scale. A 1/1Oe scale is generally the best scale for a good observation.
It is often necessary to use enhanced images but their
interpretation must be done with the static normalized
Serralog 0 2003
images as a reference to check the resistivity range of the
observations.
Images must be interpreted with the help of the other
openhole logs, preferably recorded at a high sampling
rate to relate events seen on the images to petrophysical
properties and vice versa.
When comparing images and core photographs, only
bed boundaries, laminations and fractures that are not
oblique to the borehole axis will be common to the two
sources of data (Fig. 4-54).
When cores exist one can orient them with the DIAMAG€ program written by ELF engineers and commercialized by Schlumberger. It is also possible to better locate the core losses. Another application is the precise calibration of resistivity with core porosity and permeability as
179
Previous Page
the plugs can be perfectly located on the images (see Fig.
4-54). In heterogeneous formations, what is seen in the
core cannot be on the borehole wall and perfect correspondence between core and image is unlikely.
Illustration of structures by images
Bed boundaries
Bed boundaries are marked by abrupt and relatively
high resistivity contrasts that cross all the images. These
lines are easily correlated from pad to pad and are visible
on static images. Careful examination of the shape of the
lines on each pad determines the nature of the bed boundary (cf. Fig. 4-45 and Fig. 4-56). These lines correspond
generally to surfaces or boundaries separating two beds
of different lithologies or porosity. Their length defines the
bed thickness and geometry.
When several consecutive bed boundaries, planar and
parallel over a relatively long interval, delineate consecutive conductive and resistive beds, they generally characterize parallel bedding. In sand-shale series, parallel bedding indicates a low energy environment, below the wave
or tide influence that was undisturbed by bioturbation.
1
Figure 4-54 - Depth match of image with photograph of the external
core surface using the DIAMAGE program of ELF. Observe the fracture on the image and its equivalent on the core. Its vertical extent is not
the same. This is due to the difference in diameter of the core and the
borehole as explained by Figure 4-55.
A
Figure 4-55 - Difference in size between core and borehole wall image
for an inclined surface.
180
X570
Figure 4-56 - Typical erosional surface marked by a sharp irregular
contrast with load cast.
Variations in thickness along the images are also observed. These features correspond to wavy bedding characteristic of a higher energy level than that reflected by
parallel bedding (Fig. 4-57). Laterally it can evolve either
towards parallel or lenticular bedding. The latter is characterized by lenses of more conductive (dark) or more
resistive (white) material (Fig. 4-58). Lithology of the lenses is not easy to determine. When dark lenses correspond to shales they indicate a higher energy than
wavy bedding; these lenses do not constitute permeability barriers. When white lenses correspond to sands, they
indicate a lower energy depositional environment.
Load casts are characterized by “ a slight bulge, a deep
or shallow rounded sack, a knobby excrescence or a
highly irregular protuberance”. They are produced “by the
exaggeration of the depression as a result of unequal
settling and compaction of the overlying material and by
the partial sinking of the such material into the depression
as during the onset of deposition of a turbidite on unconsolidated mud‘ (Bates & Jackson, 1980).
Serralog 0 2003
Structure
I Chapter 4 I
181
may be abrupt, tangential or concave. These features correspond to foreset bedding or cross-bedding. Jopling
(1965) gives a list of useful indices of foreset laminae
which may serve as a qualitative guide in determining the
current strength and the mean grain size.
Figure 4-57 - Non parallel wavy bedding as seen on Formation
Microscanner images (courtesy of Schlumberger).
Pad orientation
1
Figure 4-59 - On the lefi, example showing different types of bed boundary, very thin laminations in siltstone (in red) and in sand (in yellow),
with some thin shale laminae (flasers ?). A shale bed at the bottom and
in the center. Observe vertical red lines which can correspond to fractures (courtesy of Halliburton).
On the right, foreset and truncation in eolian sand with the core photograph on the side. One can recognize on image thin layers more
cemented which are not detected on core (courtesy of Schlumberger).
Figure 4-58 - Isolated sand lens in a shaly formation as seen on
Formation Microscanner images.
Internal oraanization
Features that are parallel, planar and oblique to bed
boundaries may be observed on a relatively thick interval
on all the pads. They are sometimes interrupted by other
parallel features inclined in different directions. Resistivity
contrasts, caused by changes in porosity (possibly related
to diagenetic effects) and/or grain size (alternating coarser laminae and thin fine-grained laminae), are often observed (Fig. 4-59). The shape of the contact with the base
Serralog 0 2003
The angle of foreset dips is about 30" at low velocity
and is less than 20" at high velocity. The foreset slope
becomes steeper when the sand grains are coarse and
angular, sorting is poorer and clay is absent (McKee,
1957).
With increasing velocity, the character of contact between foreset and bottomset changes from angular to tangential and even to a sigmoidal shape.
Two populations of laminae dipping abruptly and practically in opposite directions correspond to herringbone
cross-bedding.
Observation of foresets in carbonates indicates that
carbonates have a grainstone texture (Fig. 4-60 and 4-61)
since foresets can only be generated with progradation of
clastic material.
The dip azimuth of the foresets gives the direction of
181
the transport current. This also corresponds to the direction of the progradation of the body (Fig. 4-60).
Foresets introduce permeability anisotropy and, sometimes, permeability barriers when they are cemented or
truncated by other features. This must be taken into
account when defining the block size and the position of
injection wells during a secondary recovery phase.
-I
Figure 4-62 - Small scale cross-bedding organized in cosets, truncations and overturned feature (at the top) are well seen on these FMS
images. On the side a 30 view of the overturned feature is reproduced
(from Serra, 1989).
Figure 4-60 - Foresets observed in a carbonate. They indicate a
grainstone texture of this formation.
Fgure 4-61 - Photograph of the same formation seen on outcrop. One
can see in the Massangis quarw, France the foreset laminae in this
carbonate of the Dogger formation.
Very fine, hairline features with curved surfaces generate with other similar surfaces a festoon shape. They correspond to trough cross-bedding. Depending on their
size, they may not necessarily cross the borehole and are
not seen on all pads (Fig. 4-62).
182
d
Figure 4-63 - Succession of planar cross-bedding with shale drapes,
contorted bedding and non planar cross-beds with shale intrusion characterizing the upper section of a fluvial channel sand
(courtesy of Halliburton).
Serralog 0 2003
Structure
Chapter 4
183
Contorted and overturned structures can be well recognized on images as illustrated by Figures 4-62 and 4-63.
Flasers and water escaped features are also easily
observed on images without need of core (Fig. 4-64).
A
Figure 4-64 - Flasers, bifurcated flasers (dark “chips” features), and
water escape features are well observed on this FMI image over this
sand deposited in a intertidal environment with a marked increase in
water depth from the bottom to the top (courtesy of Schlumberger).
Deformation of layers or laminae are sometimes observed as previous illustrated (cf. Fig. 4-62 and 4-63). The
deformations can be over a longer interval. They correspond generally to slump structures (Figs. 4-65 and 466) or to microfolds (Fig. 4-67 next page).
1
Figure 4-66 - 3 example of slump structures. At the top, coming from a
2-pad FMS tool (Serra, 7989). At the bottom, image coming from a FMI
tool (courtesy of Schlumberger).
Irregular features either more conductive or resistive,
especially those vertical or oblique intersecting laminae or
bed boundaries, can sometimes be observed. These images frequently correspond to bioturbation by animal or
plant activity (Figs. 4-68 and 4-69). They indicate either an
oxidized environment with strong biologic activity or emersion when they correspond to root traces.
In some cases typical features correspond to fossils as
illustrated by Figure 4-70.
Graded beddinq
Ll
Figure 4-65 - Slump structure seen on FMS-2 images (internal figure)
confirmed by the Core Photograph (external View) (from Serf8 7989).
Serralog 0 2003
Progressive vertical evolution of the color (or gray) correspond to graded bedding. If they have sufficient vertical
evolution, the other logs such as gamma ray, thorium,
density, neutron and sonic should show similar trends.
183
Next Page
o
90
~0210360
Yettical burrow
Boturbated
zone
Root traces
Figure 4-69 - These FMI images show in a sandstone-mudstone
sequence vertical and horizontal burrows and zone of bioturbation
superposed to a period of emersion confirmed by the root traces seen
at the bottom of these images (courtesy of Schlumberger).
Core photograph
Static FMS image
Enhanced FMS image
832.4
Figure 4-67 - Microfold with steeply dipping limb
(courtsey of Schlumberger).
832.4
Core photograph
Enhanced FMS image
Burrow
Pyrite
-
Figure 4-68 The white vertical feature on the leff of the FMS image
corresponds to a vertical burrow. Other white dots correspond to
oblique or horizontal burrows. The dark feature with sharp and linear
edges corresponds to a pyrite crystal. The core photograph on the left
collected at the same depth shows similar features confirming the
interpretation (from Serra, 1989).
Figure 4-70 - The crescent shape of the white even observed below
the arrow on the enhanced FMS image (central track) suggests an
oyster shell. On the core photograph at the same depth a similar oyster is observed (from Serra, 1989).
When these vertical trends are observed over a long
interval they correspond either to a fining-up sequence in
a fluvial deposit (Fig. 4-71), or to a coarsening-up sequence due to progradation in a deltaic environment (Fig. 472).
When these trends are observed on short interval they
correspond to turbidite deposits. In such case one can
also observe thickening-up sequences (Fig. 4-73 nextpage). In such case the other logs cannot detect such trends
due to their lack of vertical resolution.
Other features such as conglomerates, grain or mudSerralog Q 2003
Previous Page
Structure
supported, and breccias can be recognized on images
Chapter4
j
185
(cf. Figs 3-56, 3-57, 3-62, 3-63, 3-84c and 3-85).
0
DEPTH
Figure 4-71 - Example of openhole logs recorded with a high sampling rate indicating a fining-up sequence with an abrupt erosional contact at X76
m and X53.5 m. The images on the right, reproduced at the same vertical scale, show the same trends. White yellow color corresponds to sandstone, dark brown color to shale, intermediate color to silt (from Serra, 1989).
140
200
260
320
20
80
140
In the same way, anhydritic or chert nodules can be
recognized. They are characterized by very resistive
peaks on few resistivity curves, or more or less irregular
white spots on images.
The determination of the mineralogic nature of the very
resistive features requires other log data to recognize the
nature of the deposit: detritic for conglomerates, carbonate for chert or anhydrite nodules the latter being frequent
in dolomitic deposits (Fig. 4-74 next page).
Automatic analysis of dipmeter data.
The SYNDIP program
Figure 4-72 - Typical coarsening-up sequence seen on FMS images
(from Serra, 1989).
Serralog 0 2003
As demonstrated in the previous chapter and in the
first part of this one, the interest of the dipmeter data to
extract information on the texture and the sedimentary
structure of rocks is obvious. This information is very
185
186
Well logging and Geology
PAD ORIENTATION
PAD ORIENTATION
Figure 4-73 - Typical grain size evolutions detected by FMS images.
They correspond to color trend between white which corresponds to
sand and black which correspond to shale. Observe also the thickening-up sequence. These two types of observation characterize
turbidite deposits (Sera, 1989).
important for a better and more accurate definition of the
electrofacies, consequently of the facies and the sedimentary environnements, and cannot be ignored.
But this information is qualitative by nature: character
of the dipmeter resistivity curves, type of the bed boundaries, dip evolution with depth...
Consequently, its utilization for the automatic determination of electrofacies is only possible if the data, extracted from the dipmeters and their processing by GEODIP
or LOCDIP programs, are quantified. On the other hand,
to be used those data must be assigned to the electrobed
or electrosequence.As known, fast channel dipmeter data
are sampled each 5 mm for HDT tool, or 2.5 mm for SHDT
or for imagery tools instead of each half-foot (15 cm) or
exceptionally each 1.2 inch for wireline tools.
To achieve these goals, dipmeter resistivity curves and
186
Figure 4-74 - Anhydrite growth (white spots) in a sandy dolomite.
Central images correspond to an isolation of the white spots. Right curves are the results in density of spots (in blue) and the area they cover
(in violet) coming from a quantitative analysis of the central images
(from Delhomme & Motet, 1993).
dip computation results were initially described by a series
of parameters :
- the variability or activity of the curves (VAR), which
reflects the homogeneity (very low variance), or the
heterogeneity (high variance) of the formation,
- the frequency of events (peaks or troughs) per curve
(FRE) over a given interval (6 in.) as recognized by the
GEODIP program (cf. Fig. 4-49),
- the average event thickness per curve (ALT) (P9, parameter of pattern vector computed during the GEODIP processing),
- the balance of positive to negative excursions of the
resistivity curves or, in other words, the ratio of the average thickness of the peaks over the average thickness of
Serralog 0 2003
the troughs in the interval (BAL). If these peaks and
troughs correspond to sand and shale beds, this parameter is used to compute a sand-shale ratio (cf. Fig. 4-50),
-the density of correlations found by the GEODIP or LOCDIP program (DEN), the frequency of events (FRE) can
be high and the density of correlation (DEN) can be low if
these events are not similar and consequently not easily
correlated (case of conglomerates, recifal boundstones,
zones of bioturbation),
- the sharpness of the curve events (averaging from the
four resistivity curves) (SHA),
- the average resistivity of the interval (SRES).
These parameters were computed for each 6-in. interval corresponding to 30 samples. They represented an
attempt to convert the information included in the curve
shape to a quantitative form useable in the definition of
electrofacies similar to a log. It is for that reason that the
name of "synthetic logs" was given to these computed
curves.
Since some of the HDT or SHDT-derived logs can
reflect same phenomena (i.e. FRE, DEN, ALT, VAR), and
consequently can be highly correlated (Fig. 4-75), a certain amount of redundancy is present. Some of these
synthetic logs exhibit a high degree of correlation with
openhole logs such as gamma ray, or spontaneous potential (Fig. 4-76). Comparison of the two shows that positive
excursions of the new curve indicates heterogeneity
(laminations, poor grain sorting, etc.) and the negative
side indicates homogeneity (well-sorted sands, etc.). This
curve can then be used in the processing as a textural
and structural indicator, instead of FRE, DEN, VAR ....
The SYNDIP Proaram
SYNDIP is a program developed by Schlumberger
(Delhomme & Serra, 1984), to replace the previous procedure. It generates dipmeter-derived (or synthetic) logs
from HDT or SHDT raw data and computation results. The
so-called synthetic logs are based on the features and
likeness of the microresistivity curves, and on the evolu-
J
Figure 4-75 - Correlations between some of the synthetic logs derived from dipmeter data (from Serra, 1989).
Serraiog 0 2003
187
tion versus depth of dips and planarity. SYNDIP can be
focused, according to different criteria (dip quality and planarity, size of curve events ...).
In normal use, SYNDIP synthetic logs are output with
a half-foot sampling rate; this rate was chosen in order to
be consistent with the sampling rate of most open-hole
logs. However, for detailed studies, another rate (e.9.
1.2", that is the usual sampling rate for EPT tool measurements) can be selected.
The frequency of inflexion points, correlated or not, on
a single dipmeter resistivity curve is first computed as an
indicator of curve activity. The sharpness of this synthetic
log is ensured by the tact that no window is used to compute frequency; rather, it is based on the thickness bet
ween consecutive inflexion points (ATBR).
maintained.
An idea of the bed internal organization is obtained by
combining curve activity with density of correlations. High
activity with no correlation will reflect a coarse heterogeneous bed organisation (conglomerates, flaser bedding,
reefs, vuggy limestones, bioturbation ...). Low or no activity with no correlations will reflect either a homogeneous
bed or a very finely heterogeneous bed. High activity with
high density of correlations will correspond to thinly larninated formations (cf. Table 4-5).
As for bed boundaries, a non-planarity flag is output
when the GEODIP or LOCDIP planarity coefficient is
below a certain threshold. This indicates non planar and
possibly erosive surface.
The parallelism between consecutive bed boundaries
can be appreciated by comparing the corresponding dipping planes; the angle between a layer's top and bottom
is computed and a non-parallelism flag is output if the
angle is higher than a certain value (e.9. 10'). Dip dispersion is also computed in a given sliding window (typically
3 m or 10 ft wide); it is an angular standard deviation computed on the unit hemisphere.
Intervals with low dip dispersion, thin beds, small
angles between bed tops and bottoms, and with a lithology that corresponds to low energy deposits (shaly-silty or
shaly-marly laminations) are those intervals where the
structural dip can be picked.
The nature of contacts and transitions is also identified. The high vertical resolution of the dipmeter and its
sampling rate allow separation of abrupt changes from
gradational ones. This last type corresponds to conductivity ramps, which generally reflect grain size or lithology
evolution (sequences). SYNDIP ramp analysis outputs
small and large-scale ramps.
Lastly, in SYNDIP, dipmeter data are corrected for
EMEX (i.e. for the value of the total current sent out into
the formation), rescaled and averaged in conductivity over
the output sampling interval since levels appear in parallel.
Description of a SYNDIP Display
20-
2
0-
Figure 4-76 - Correlations between HOT derived synthetic curve (DEN)
with SP curve (from Serra, 1989).
A bed, or a lamina, can be defined as the depth interval between two consecutive correlations found by GEODIP or LOCDIP programs. Correlation lines link upper or
lower inflexion points of similar events on several dipmeter curves. Layer thicknesses are thus computed between
consecutive correlation links (ATCL). When correlation
links are closer than the output sampling interval, an average thickness is simply computed, but the resolution is
188
An example of the SYNDIP graphic output is given in
Fig. 4-50. The first track reproduces, on a logarithmic
scale, 3 calibrated resistivity curves of dipmeter : the minimum and maximum values recorded by the 4 or 8 buttons
in thin lines, and the average value computed from these
4 or 8 resistivity measurements as heavy lines. This presentation enables detection of homogeneous or heterogeneous beds.
The second track displays the internal organization of
the beds : small dots represent the homogeneous or massive beds, bubble-like shading correspond to heterogeneous beds, and dark gray with horizontal lines the laminated beds. The limiting curve is the frequency of inflexion
points, an indicator of the total curve activity (I/ATBR),
and the frequency of correlations (I/ATCL).
On the side, in three columns, are flags indicating
detection of non planar surfaces, non parallelism between
Serralog 0 2003
Structure
consecutive bed boundaries, and correlated parallel planes. The following column indicates intervals containing a
minimum of five dips with angular spherical dispersion
lower then 5",with the computed average dip. This represents the structural dip if the interval corresponds to a low
energy environment (shale, silt or marl).
Sometimes, selected dip results are represented in a
third track. The dip, which is taken as representative of a
1.5 ft wide interval, is the one which minimizes the spherical distance to the (n/2)th closest dip result among the n
ones found in this interval. The solid curve to the right is
the dip spherical standard deviation, and the dashed line
is the angle between current layer's top and bottom.
The next track shows a conductivity curve, shaded
with a continuous grey or colour scale to reinforce the
sand/shale opposition in sandlshale series (cf. Fig. 4-50).
Colours are selected by using the histogram of dipmeter
resistivity measurements. Sometimes, two scales of
conductivity ramps are displayed on the left side.
The last track displays thicknesses of conductive or
resistive layers in two mirror logarithmic interpolation between the 8 resistivity curves, after equalization, following
the dips computed by the LOCDIP program.
Geologists must be convinced, if necessary, that the
measurements realized by dipmeters contain much more
than the dip information. The resistivity curves reflect in
detail the internal organization of the formations crossed
by the well. This information must be exploited in order to
have a return on investment. Oil companies paid to aquire this information. Consequently, why not exploit it?
"Squeeze the lemon until its last drop!"
Complementary remarks on features seen by dipmeters and imagery tools
The above mentioned examples demonstrate that dipmeter and imagery tools provide a mass of information,
covering the texture as well as the sedimentary structure
of rocks, the direction of transport and the thickness of the
beds. The advantages of their integration in all subsurface sedimentological studies can therefore be understood.
It is obvious that HDT; SHDT and imagery tools do not see
all sedimentary features: they detect only those presenting a minimum resistivity contrast. Thus, the figures visible to the naked eye due to a color change are not detected unless this change is accompanied by a resistivity
variation. The features appearing on the surface of a bed
without repercusion on a vertical section (tool marks, rain
imprints, ...), cannot be recognized because the tool only
analyses a cylindrical section.
In general, one can admit that all events detected by
dipmeter or imagery tools (corresponding to resistivity
variations) inevitably explain a change of geological parameters (mineralogy, texture, sedimentary structure, fluids
...) with the condition, of course, that the pad is properly
applied to the borehole wall and the tool is working correctly.
Serralog 0 2003
I Chapter 4 I
189
Rules for interpretation of dipmeter and imaae data
The interpretation of a dipmeter arrow-plot should
never be implemented without the integration of all other
available data, consequently the open-hole logs.
- The first step of the interpretation must be the compilation of a composite log combining at 11200 scale all available logs and the GEODIP or LOCDIP plot or images
after depth matching. The result of a processing describing the lithology or mineralogy can also help considerably if it is reproduced alongside.
- The second step consists of the observation and description of the following aspects, by refering if necessary
to the GEODIP or LOCDIP plot at 1/40 scale, or to the
image data at 1/10 or 1/5 scale. At this stage it is necessary to separate dips corresponding to tectonic structure
from sedimentary dips generated by current (Fig. 4-76).
- The investigated interval is subdivided into zones with
approximately constant characteristics.
- The mineralogical composition of each zone, and of
each event in the zone, is defined as precisely and carefully as possible. To achieve that, the data of other logs,
especially litho-density, neutron, sonic, natural gammaray spectrometry ..., or a LITHO display, must be used.
- The nature of the contact (abrupt or gradational, planar
or warped, conformable or not) is observed.
- The type of layer succession is described simple or composite, with or without parallel boundaries, continuous or
discontinuous.
-The thickness of each bed type and its evolution with the
depth must be noted.
- The existence of current features (flaser, wavy, or lenticular bedding), revealed by the thickness of the events
and the dip variations (magnitude and azimuth), or by the
images, must be extracted and analyzed.
- Internal structure of a bed including the curve shape
(massive and homogeneous, heterogeneous, resistivity
ramp revealing a graded bedding), and dip evolutions
(oblique bedding, cross-bedding, foreset,...) must be
added. The amplitude of dip variation is also analysed.
The absence of these variations indicate either a low
energy, or, on the contrary, a very high energy. The choice between the two hypotheses is derived from the mineralogical nature or the vertical position of this phenomenon in the granulometric sequence, and from the dips.
The important variations of the dip magnitude indicate
changes of energy in the environment.
- Finally, the sequential evolutions, the rhythms or cycles,
the evolution in thickness of each bed and sequence as
they are revealed by the dipmeter resistivity curves or the
images, are studied.
- The third step corresponds to the direct interpretation. It
involves the translation of the observed features into geologically meaningful interpretations. For instance a "blue
pattern" (increasing dip magnitude upward) will be interpreted as a foreset.
- The fourth step constitutes the deductive interpretation
189
190
Well logging and Geology
plldsrad w s
Local great
circle axes
Figure 4-77 - To restitute correctly the sedimentary dips, reproduced on the right, it is necessary to substract the structural dip. The left figure
reproduces the dips determined over a 6 meters interval. It is difficult to separate the structural dip from the sedimentary dips. Based on the determination of the local great circle axes (cf. Fig. 4-77 below), the structural dip can be determined and substracted (central figure)
(from Etchecopar & Bonnetain, 1992).
Figure 4-78 - Trough cross-bedding surfaces fit great CifC/eS With different axes. Those axes fit a great circle which allows the determination
of the structural dip (from Etchecopar & Bonnetain, 1992).
190
in terms of a depositional environment. Analysis of azimuth frequency plots established on selected intervals will
help to define the uni-, bi- or polymodal nature of the dips.
To achieve this, we have to integrate data obtained
from the observation of cuttings and cores, such as the
presence of glauconite, lignite fragments, phosphate,
shells; heavy minerals; grain size, granulometric sorting,
nature of cement, shale type, ...
One proceeds by the elimination of the environmental
hypotheses which do not fit with the observed features.
The final selected solution among the remaining hypotheses, is the one which fits the best with the geological knowledge we have of the formation, the basin and the main
tectonic features.
It is suggested to summarize all the observations in a
table (Table 4-6) or in several columns put alongside the
composite log at 1/200 scale, one column for lithology,
one for sedimentary features, one for sequential evolution, one for the direct interpretation and complementary
one for the finalinterpretation in terms of facies,
subenvironments and environment.
Serralog 0 2003
Structure
Quantitative analysis of image data
As done for dipmeter data, images must be analyzed
quantitatively in order to extract and classify the information they contain (Fig. 4-79 next page).
As previously seen in Chapter 3, heterogeneities,
represented by features more or less resistive than the
background, are analyzed thanks to a processing similar
to the BorTex program developped by Schlumberger.
Their frequency, their shape and their size can be determined.
Surfaces can be classified as a function of several criteria:
- amplitude of the resistivity contrast on each side of
the surface. Low amplitude will suggest the persistency of
the same lithology; high contrast will suggest an important
change in rock characteristics either due to lithology change or to porosity change.
- proximity of succesive surfaces. A very short depth
1
Chapter 4
~
191
ne bedding. Change of dip and azimuth with hogh resistivity contrast will probably suggest an unconformity.
Sedimentological applications of
sedimentary structure detection
The main application is the determination of the depositional environment.
Without minimizing the sedimentological interest of the
CLUSTER, MSD or BORDIP type programs - see the
works and papers of Gilreath et a/. (1964, 1969, 1971),
Campbell, (1968), Goetz et a/. (1977), Selley, (1979),
Serra (1985) - it should be underlined that these processings do not exploit the very detailed analysis of the dipmeter that is now possible with thelast dipmeter tool (e.g.
SHDT) or imagery tool and improved processing techniques. The curves were generally not shown and the precise events from which the dips are computed are still
unknown. These dips are determined with the help of a
Table 4-6
Summary of the internal structure information.
I
interval between consecutive surface with low reistivity
contrast will suggest laminations.
- planarity of the surfaces. Irregular non planar surface
will probably correspond to an erosional surface r to a presence of load cast.
- dip and azimuth consistency. Practically no change in
dip and azimuth combined with low resistivity contrast and
high proximity will correspond to parallel laminae or foresets from which the transport current direction can be
determined. Change in dip and azimuth with low resistivity contrast can correspond to cross-bedding or herringboSerralog 0 2003
correlogram established from a correlation of events in a
given interval, without distinction and selection of their origin. The obtained dip is consequently an average dip for
an interval which can cover several types of surfaces:
sedimentary units, each of them possibly having sedimentary features with different dips, fractures, and isolated events. For these reasons it is highly preferable to use
a GEODIP or LOCDIP presentation for sedimentological
interpretation of dipmeter data.
As previously stated the primary sedimentary structu191
Q U A N T I T A T I V E BOREHOLE I M A G E ANALYSIS
mw
QEOLOWCM OBJECTSDOWN TO A
UILUYIETERL) s m ARE IDENT~REDwmi BOREHOLE IMAGES
MATHEMATICAL
TOOLS ALLOW TO AUTOMATICALLY Ucriuct AND QUANTIFY THEM
A FAST ZONATION CAN BE EXTRACTED FROM
BlNARlZED IMAGES DERIVED WITH AN
ADAPTATWE BACKGROUND
RESISTIVITY FILTER
Landscape view of the
conductivity image.
Landscape view of the
gradient image.
FOR EACH ZONE, GEOLOGICAL OBJECTS ARE AUTOMATICALLY
EXTRACTED AN0 THEIR ATTRl9UTESARE COMPUTED
THIS METHOD INVOLVES A PRIMARY STEP OF INTERPRETATIONAND THE MACBE ANALYSIS
IS QUIDED BY THE QEOLOQlCAL KNOWLEDOE OF THE FORMAWOMS
DEPOSlTlONAL
SURFACE
ATTRIBUTES:
- Resistivitycontrast
- Dip and azimuth
- Planarity
- Proximity
LAYER
ATTRIBUTES:
-Medianconducthrity
h w
-T
- _
DlAQENETlC
HETEROGENEITY
TYPE AND DISTRIBUTION:
- Size, shape, area
- Frequency
- Connectedsless
- Median#Mducthllty
STRESS INDUCED
FRACTURE
ATTRIBUTES:
- Polarity ( r e s i s t i V 9 , w ~ )- Dlp, pzlmuth
- Lenom
- Density
- Electrlcalaperbwe
- Planarity
Figure 4-79 - Quantitative borehole image analysis and classification of the geological objects as a function of their origin
(from Delhomme & Motet, 1993).
192
Serralog 0 2003
Structure
res -which, can be detected on dipmeter logs or better on
images - are particularly important because they reflect
the hydro- (or aero-) dynamic conditions prevailing in the
environment at the time of deposition.
Another important application is in petrophysical interpretation. As soon as cross-bedding or foresets are recognized, the grain size can be evaluated (cf. Fig. 4-11) and
through the relationship between porosity and permeability as a function of the grain size (cf. Fig. 3-89 Chapter 3)
the permeability estimated. In addition, the permeability
anisotropy can be evaluated (Fig. 4-80).
Figure 4-80 - Permeability anisotropy as a function of the internal organization of the laminae.
Another petrophysical application of dipmeter and
imagery tools is for interpretation of thin bedded reservoirs. The lack of resolution of the standard logs does not
allow their recognition. Only dipmeter and imagery tools
have a vertical resolution allowing to detect succession of
two different types of depositional units with different
petrophysical properties and thicknesses (cf. Figs. 4-41,
4-42, 4-44, 4-52 and 4-53).
References and Bibliography
ALLEN, J.R.L. (1963). - The classification of crossstratified units, with notes on their origin. Sedimentology, 2, p.
93-114.
ALLEN, J.R.L. (1968). - Current Ripples. North
Holland, Amsterdam.
ALLEN, J.R.L. (1970). - Physical Processes of
Sedimentation. Elsevier, New York.
BATES, R.L. & JACKSON J.A., ed. (1980) - Glossary
of Geology. American Geol. Institute, Falls Church,
Virginia.
BLATT, H., MIDDLETON, G. & MURRAY, R. (1972,
1980). - Origin of Sedimentary Rocks. 1st and 2nd ed.
Serralog 0 2003
I
Chapter 4
I
193
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
BOYELDIEU, C. & JEFFREYS, P. (1988). FormationMicroScanner: New developments. SPWLA,
11th Europ. Eval. Symposium Trans., Oslo, paper G.
BOURKE, L.T. (1989). - Recognizing artifact images of
the Formation Microscanner. SPWLA, 30th Ann. Log.
Symposium Trans., Denver, paper WW.
BOURKE, L.T., DELFINER, P. FETT, T, GRACE, M.,
LUTHI, S. , SERRA, 0. & STANDEN, E. (1989). - Using
Formation Microscanner Images. The Technical Review,
37, 1, p. 16-40.
CAMPBELL, C.V. (1967). - Lamina, Laminaset, Bed
and Bedset. Sedimentology, 8, p. 7-26.
CAMPBELL, R.L. (1968). - Stratigraphic applications
of dipmeter data in Mid-Continent. Bull. Amer. Assoc.
Petroleum GeoL, 52, 9, p. 1700-1719.
CAROZZI, A.V. (Ed.) (1975). - sedimentary Rocks.
Benchmark Papers in Geology, 15, Dowden, Hutchinson
& Ross, Inc., Stroudsburg, Pennsylvania.
Chambre Syndicale de la Recherche et de la
Production du Petrole et du Gaz naturel (1966). - Essai de
nomenclature & caracterisation des principales structures
sedimentaires. Ed. Technip, Paris.
Chambre Syndicale de la Recherche et de la
Production du Petrole et du Gaz naturel (1974). Methodes modernes de geologie de terrain. 1 - Principes
d’analyses sedimentologiques. Ed. Technip, Paris.
COLLINSON J.D. & THOMPSON, D.B. (1982). Sedimentary Structures. George Allen & Unwin PubL Ltd.,
London.
CROSBY, E.J. (1972). - Classification of Sedimentary
Environments. In : “Recognition of Ancient Sedimentary
Environments”, edited by RIGBY J.K. & HAMBLIN, WK.,
SEPM, special publication 16.
DELFINER, P., PEYRET, 0. & SERRA, 0. (1984). Automatic determination of Lithology from Well Logs. 59th
Ann. Techn. Conf. SPE of AIME, Houston, Texas: paper
no SPE 13290.
DELHOMME, J.-P. (1992). - A quantitative characterization of formations heterogeneities based on borehole
image analysis.
DELHOMME, J.P., & MOTET, D. (1993). - Reservoir
Description and Characterization from Quantitative
Borehole lmage Analysis. Schlumberger, Africa and
Mediterranean Region, Reservoir Characterization, 5; p.
5-19.
DELHOMME, J.-P. & SERRA, 0. (1984). - Dipmeterderived Logs for SedimentologicalAnalysis. SPWLA, 9th
Europ. Intern. Format. Eval. Trans., paper 50.
DUNHAM, R.J. (1962). - Classification of Carbonate
Rocks according to Depositional Texture. Amer. Assoc.
Petroleum GeoL, Mem. 1, p. 108-121.
DZULINKI, S . & WALTON, E.K. (1965). - Sedimentary
features of flysch and greywackes. Elsevier, Amsterdam.
EKSTROM, M.P., CHEN, M.Y., ROSSI, D.J., LOCKE,
S. & ARON, J. (1986). - High Resolution Microelectrical
Borehole Wall Imaging. SPWLA, 10th Europ. Symp.
Trans., Aberdeen, paper K.
EKSTROM, M.P., DAHAN, C.A., CHEN, M.Y., LLOYD,
193
194
Well logging and Geology
P.M. & ROSSI, D.J. (1986). - Formation lmaging with
Microelectrical Scanning Arrays. SPWLA, 27th Ann. Log.
Symp. Trans., Houston, paper BB.
von ENGELHARDT, W., & GAIDA, K.H. (1963). Concentration changes of pore solutions during the compaction of clay sediments. J. Sediment. Petrology, 33, p.
919-930.
ETCHECOPARA.& BONNETAIN, J.-L. (1992) - Cross
Sections from Dipmeter Data. Arner. Assoc. Petrol. Geol.
Bull., 76, 5, p. 622-637.
FONS, L. Sr. (1969). - Geological application of well
logs. SPWLA, 10th Ann. Log. Symp. Trans.
FRIEDMAN, G.M. & SANDERS, J.E. (1978). Principles of Sedimentology. John Wiley & Sons, New
York.
GALLOWAY, W.E. (1977). - Catahoula Formation of
the Texas Coastal plain: depositional systems, composition, structural develoment, ground-water flow history, and
uranium distribution. Bur. Econ. Geol. Univ. Texas, Austin,
Rept. Invest. no 87.
GARY, M., McAFEE, R.Jr. & WOLF, C.L. (1972). Glossary of Geology. Amer. Geol. Institute, Washington,
D. C.
GILREATH, J.A. & MARICELLI, J.J. (1964). - Detailed
Stratigraphic Control through dip Computations. Bull.
Amer. Assoc. Petroleum Geo/., 48, 12, p. 1902-1910.
GILREATH, J.A., HEALY, J.S. & YELVERTON, J.N.
(1969). - Depositional Environments Defined by Dipmeter
Interpretation Gulf Coast Assoc. GeoL SOC.Trans., 19, p.
101-111.
GILREATH, J.A. & STEPHENS, R.W. (1971). Distributary Front Deposits lnterpreted from Dipmeter
Patterns. Gulf Coast Assoc. Geol. SOC.Trans., 21, p. 233243.
GILREATH, J.A. & STEPHENS, R.W. (1975). Interpretation of Log Responses in a Deltaic Environment.
Amer. Assoc. Petroleum Geol. Marine Geology
Workshop, Dallas, Texas.
GOETZ, J.I., PRINS, W.J. & LOGAR, J.F. (1977). Reservoir Delineation by Wireline Techniques.Paper presented at 6th Ann. Conv. Indonesia Petroleum Assoc.,
Jakarta, May 1977.
GOODALL, I., WESTLAKE, I., BLOCKVAGLE, G.,
MUNDAL, I., & MULCAHY, M. (2002). - Resistivity lmage
Data, Voring Basin, Offshore Norway. In: LOVELL, M., &
PARKINSON, N. (ed.) (2002). - Geological Applications of
Well Logs. AAPG Methods in Exploration, 604 p.
GUY, H.H., SIMONS, D.B., & RICHARDSON, E.V.
(1966). - Summary of alluvial channel data from flume
experiments 7956-67. U.S. Geol. Surv. Profess. Papers
462-1, 96 p.
HACKBARTH, C.J. & TEPPER, B.J. (1988). Examination of BHTV, FMS, and SHDT images in very
thinly bedded sands and shales. 63rd SPE Ann. Techn.
Conf. & Exhib., Houston, paper SPE 18118.
HARKER, S.D., McGANN, G.J., BOURKE, L.T. &
ADAMS J.T. (1988). - Formation Microscanner Applications in Integrated Reservoir Description in
Claymore and Scapa Fields (North Sea). Geological
194
Applications of Wireline Logs, London, June 1-2, 1988,
Geological Society of London (in press).
HARKER, S.D., McGANN, G.J., BOURKE, L.T. &
ADAMS J.T. (1988). - Methodology of Formation
Microscanner lmage Interpretation in Claymore and
Scapa Fields (North Sea). Journal of the Geological
Society of London (in press).
HECKEL, P.H. (1972). - Recognition of ancient shallow
marine environments. In: Rigby & Hamlin (eds):
Recognition of ancient sedimentary environments, Spec.
Publs. S.E.P.M., 16, p. 226-286.
HEPP, V. & DUMESTRE, A.C. (1975). - CLUSTER - A
method for selecting the most probable dip results from
dipmeter survey. SPE of AIME, 50th Ann. Fall Mtg.,
Dallas, Paper SPE 5543.
HOBSON, G.D. & TIRATSOO, E.N. (1975). lntroduction to Petroleum Geology. Scientific Press Ltd,
Beaconsfield, England.
HJULSTROM, F. (1939). - Transportationof detritus by
moving water. In: Trask (ed): Recent marine sediments.,
Amer. Ass. Petrol. Geol.
HUTIN, R. (1986). - Formation lmaging with Micro
Electrical Scanning Arrays. A New Generation of
Stratigraphic High Resolution Dipmeter Tool. SPWLA,
10th European Formation Evaluation Symposium
Transactions, Aberdeen, Scotland, April 22-25, 1986,
paper L.
INGRAM, R.L. (1954). - Terminology for the thickness
of stratification and parting units in sedimentary rocks.
Bull. Geol. SOC.Amer., 65, p. 937-938.
JOPLING, A.V. (1965). - Hydraulic factors Controlling
the Shape of Laminae in Laboratory Deltas. Journal of
Sedimentary Petrology, 35, 4 (December 1965), p. 777791.
KRUMBEIN, W.C. & SLOSS, L.L. (1963). -Stratigraphy
and Sedimentation. 2nd ed. W HFree. man & Co., San
Francisco.
KRYNINE, P.D. (1948). - The megascopic study and
field classification of sedimentary rocks. Jour. Geol., 56, p.
130-165.
KUENEN, Ph. H. (1953). - Significant features of graded bedding. Bull. Amer. Assoc. Petroleum Geol., 37, p.
1042-1066.
LANDES, K.K. (1951). - Petroleum Geology. John
Wiley & Sons, New York.
LEET, L.Don, JUDSON, S. & KAUFFMAN, M.E.
(1978). -Physical Geology. 5th ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
LINK, P.K. (1982). - Basic Petroleum Geology. OGCI
Publications, Tulsa.
LLOYD, P.M. (1986). - The formation Microscanner :
A New Generation High Resolution Logging Technique.
LLOYD, P.M., DAHAN, C.A. & HUTIN, R. (1986).
Formation lmaging from Microelectrical Scanning Arrays :
A New Generation of Stratigraphic High Resolution
Dipmeter Logging Tool. SPWLA, 10th Europ. Symp.
Trans., Aberdeen.
LLOYD, P.M., DAHAN, C., & LUTHI, S.M. & BANAVAR, J.R. (1988). - Application of Borehole lmages to
Serralog 0 2003
Structure
Three-Dimensional Geometric Modeling of Eolian
Sandstone Reservoirs, Permian Rotliegende, North Sea.
AAPG Bulletin, 72, 9, p. 1074-1089.
LOMBARD, A. (1956). - Geologie Sedimentaire. Les
series marines. Masson, Paris.
LOMBARD, A. (1972). - Series sedimentaires. Genese
-Evolution. Masson, Paris.
LOVELL, M., & PARKINSON, N. (ed.) (2002). Geological Applications of Well Logs. AAPG Methods in
Exploration, 604 p.
MATTHEWS, R.K. (1974). - Dynamic Stratigraphy. An
introduction to Sedimentation and Stratigraphy. PrenticeHall Inc., Englewood Cliffs, New Jersey.
McGOWEN, J.H. & GARNER, L.E. (1970). Physiographic features and stratification types of coarsegrained point bars : Modern and ancient examples.
Sedimentology, 14, p. 77-111.
McKEE, E.D. (1957). - flume experiments on the production of stratification and cross-stratification. J . sediment. Petrol., 27, p. 129-134.
McKEE, E.D. (1957). - Primary Structures in Some
Recent Sediments. AAPG Bulletin, 41 , 8, p.1704-1747.
McKEE, E.D. (1966). - Structures of dunes at White
Sands National Monument, New Mexico (and comparison
with structures of dunes from other selected areas).
Sedimentology, 7, p. 1-69.
McKEE, E.D. & WEIR, G.W.(1953). - Terminology for
stratification and cross-stratification in sedimentary rocks.
Bull. geol. SOC.Amer., 64, 4, p. 381-390.
MIDDLETON, G.V. (ed.) (1965). - Primary sedimentary structures and their hydrodynamic interpretation.
SEPM, Spec. Pub. 12.
MIDDLETON, G.V. (1976). - Hydraulic interpretation of
sand size distributions. J. Geology, 84, p. 405-426.
NURMI, R.D. (1984). - Geological evaluation of stratigraphic high resolution dipmeter data. SPWLA, 25th Ann.
Log. Symp. Trans., New Orleans.
NURMI, R., CHARRARA, M. & WATERHOUSE, M.
(1988). - Heterogeneities in Carbonate Reservoirs:
Detection and Analysis Using Borehole Electrical
Imagery. Geological Applications of Wireline Logs,
London, June 1-2, 1988, Geological Society of London.
OTTO, G.H. (1938). - The Sedimentation Unit and its
Use in Field Sampling. Journal of Geology, 46, 4, p. 569582.
PAYENBERG, Tobias H.D. (2002). - Paleocurrents
from FMS and Scribe-oriented Core - A comparison. In:
Lovell, M., & Parkinson, N. (ed.) (2002). - Geological
Applications of Well Logs. AAPG Methods in Exploration,
604 p.
PAYRE, X. & SERRA, 0. (1979). - A case study Turbidites recognized through dipmeter. SPWLA, 6th
Europ. Log. Syrnp. Trans., London, paper K.
PERRIN, G. (1975). - Comparaison entre des structures sedimentaires a l'affleurement et les pendagemefries
de sondages. Bull. Centre Rech. Pau, SNPA, 9, p. 147181.
PETTIJOHN, F.J. (1930). - Imbricate arrangement of
pebbles in a pre-Cambrian conglomerate. Jour. GeoL, 38,
Serralog 0 2003
Chapter 4
I
195
p. 568-573.
PETTIJOHN, F.J. (1975). - Sedimentary Rocks. 3rd
ed. Harper & Row, Publishers, New York.
PETTIJOHN, F.J. & POTTER, P.E. (1964). -Atlas and
Glossary of Primary Sedimentary Structures. Springer,
New York.
PETTIJOHN, F.J., POTTER, P.E. & SIEVER, R.
(1972). -Sand and Sandstone. Springer, New York.
PEZARD, P.A. & LUTHI, S.M. (1988). - Borehole
Electrical Images in the Basement of the Cajon Pass
Scientific Drillhole, California; f facture Identification and
Tectonic Implications. Geophysical Research Letters 15
(August supplement, 1988), p. 1017-1020.
PILENKO, T. (1986). - Formation Microscanner
Applications in Italy. Milano, Italy. Schlumberger Italiana,
S.p.A., 1988.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Houston.
PLUMB, R.A. & LUTHI, S.M. (1986). - Application of
Borehole Images to Geologic Modeling of an Eolian
Reservoir. 61st SPE Annual Technical Conference and
Exhibition, New Orleans, paper SPE 15487.
POTTER, P.E. & PETTIJOHN, F.J. (1977). Paleocurrents and Basin Analysis. 2nd ed. Springer, New
York.
PRESS, F. & SIEVER, R. (1982). - Earth. 3rd ed. W H.
Freeman & Co, San Francisco.
READING, H.G. (Ed.) (1978). - Sedimentary
Environments and Facies. Blackwell Scientific
Publications, Oxford.
REINECK, H.E. & SINGH, I.B. (1975, 1980). Depositional Sedimentary Environments. 1st and 2nd ed.
Springer-Verlag, New York.
REINECK, H.E. & WUNDERLICH, F. (1968). Classification and Origin of Flaser and Lenticular
Bedding. Sedimentology, 11, 1-2, p. 99-104.
RUSSELL, W.L. (1951). - Principles of Petroleum
Geology. McGraw-Hill Book Co., New York.
SCHEIDEGGER, A.E. & POTTER, P.E. (1971). Downcurrent decline of grain size and thickness of single
turbidite beds : a semi-quantitative analysis.
Sedimeniology, 17, p. 41-49.
Schlumberger Ltd (1970). - fundamentals of Dipmeter
Interpretation.
Schlumberger Ltd (1974). - Log Interpretation. Volume
I1 - Applications.
Schlumberger ( I 979). - Well Evaluation Conference.
Algeria.
Schlumberger Ltd (1981). - Dipmeter Interpretation.
Volume 1 - fundamentals.
Schlumberger Technical Services, Inc. (1986). Carbonate Close up. Middle East Well Evaluation Review
1, p. 22-35.
Schlumberger Technical Services, Inc. (1987). Surprising Sandstones. Middle East Well Evaluation
Review, 2, p. 6-19.
Schlumberger Technical Services, Inc. (1988). Sandstones - The Inside Story. Middle East Well
Evaluation Review, 5 p. 32-47.
195
Schlumberger Technical Services, Inc. (1989). - Peak
Viewing for Thin Beds. Middle East Well Evaluation
Review, 6 p. 46-55.
Schlumberger Technical Services, Inc. (1982). Essentials of Natural Gamma ray Spectrometry
Interpretation.
Schlumberger Technical Services, Inc. (1983). - Well
Evaluation Conference. India.
Schlumberger WTA Marketing Services (1992). - FMl*
Fullbore Formation Microlmager. SMP 9210.
SCHOLLE, P.A. & SPEARING, D. (Ed.) (1982). Sandstone Depositional Environments. Amer. Assoc.
Petroleum GeoL, Mem. 31.
SCHWARZACHER, W. (1953). - Cross-bedding and
grain size in the Lower Cretaceous sands of East Anglia.
Geol. Mag., 90, p. 322-330.
SCOTT, K.M. ( I 966). - Sedimentology and dispersal
pattern of a Cretaceous flysch sequence, Patagonian
Andes, southern Chile. Bull. Amer. Assoc. Petroleum
GeoL, 50, p. 72-107.
SEILACHER, A. (1964). - Biogenic Sedimentary
Structures. In: lmbrie & Newell (eds): Approaches to
Paleoecology. John Wiley & Sons, New York.
SELLEY, R.C. (1970, 1978, 1985). - Ancient
Sedimentary Environments. Is t, 2nd and 3rd ed.
Chapman & Hall, London.
SELLEY, R.C. (1976). - An lntroduction to
Sedimentology.Academic Press, London.
SERRA, 0. (1985, 1989). - Sedimentary Environments
from Wireline Logs. 1st & 2d ed. Schlumberger, M081030, SMP-7008.
SERRA, 0. (1989). - Formation Microscanner lmage
Interpretation. Schlumberger Educational Services, SMP7028.
SERRA, 0. & ABBOTT, H. (1980). - The Contribution
of Logging data to Sedimentology and Stratigraphy. 55th
Ann. Fall Techn. conf. SPE of AIME, paper SPE 9270, and
in SPE J., Feb. 1982.
SIMONS, D.B., RICHARDSON, E.V., & NORDIN, C.F.
(1965). - Sedimentary structures generated by flow in
alluvial channels. In: Middleton (ed): Primary Sedimentary
Structures and their Hydrodynamic interpretation. Spec.
Publs. S.E.P.M., 12.
SPWLA Reprint Volume (1990). - Boehole Imaging.
STEINMETZ, R. (1967). - Depositional history, primary sedimentary structures, cross-bed dips, and grain size
of an Arkansas river point bar at Wekiwa, Oklahoma. Rep.
F67-G-3 (In: REINECK & SINGH, 1975).
THEYS, P., LUTHI, S. & SERRA, 0. (1983). - Use of
dipmeter in Carbonates for detailed sedimentology and
reservoir engineering studies.
TROUILLER, J.-C., DELHOMME, J.-P., CARLIN, S. &
ANXIONNAZ, H. (1989). - Thin-Bed Reservoir Analysis
from Borehole Electrical Images. 64th SPE Ann. Techn.
Conf. & Exhib., paper SPE 19578.
VINCENT, P., GARTNER, J. & ATTALI, G. (1979). GEODlP - An approach to detailed dip determination
using correlation by pattern recognition. J. Petroleum
Technol., Feb. 1979, p. 232-240.
196
van WAGONER, J.C., MITCHUM, R.M., CAMPION,
K.M., & RAHMANIAN, V.D. (1990). - Siliciclastic
Sequence Stratigraphy in Well logs, Cores, and Outcrops.
A.A.P.G. Methods in Exploration Series, no 7.
WALKER, R.G. (Ed.) (1979, 1984). - Facies Models.
1st and 2nd ed. Geoscience Canada, reprint series 1,
published by Geol. Assoc. Canada.
ZEMANEK, J. GLEN, E.E., NORTON, L.J. & CALDWELL, R.L. (1970). - Formation inspection with the borehole televiewer. Geophysics, 35, p. 252-269.
Serralog Q 2003
WELL LOGGING, FACIES, SEQUENCE
AND ENVIRONMENT
Introduction
Review of general geological concepts
Definitions
Since its introduction by Gressly (1838), the term
facies has been used in many different ways and these
uses have been the centre of considerable debate.
Usages and definitions have been reviewed by Moore
(1949), Weller (1958), Teichert (1958), Krumbein & Sloss
(1963), and more recently by Selley (1970), Reading
(1978) Middleton (1978) and Walker (1984).
Without reopening the debate, the general, almost
identical, definitions proposed by the Glossary of Geology
(Bates & Jackson, 1980) and by several geologists are listed here below.
- "The aspect, appearance, and characteristics of a
rock unit, usually reflecting the conditions of its origin;
esp. as differentiating the unit from adjacent or associated
units" (Bates & Jackson, 1980).
- Haug (1907) : "the sum of the lithologic and palaeontologic characteristics of a [sedimentary] deposit at a
given place".
- Moore (1949) : "any areal/y restricted part of a designated stratigraphic unit which exhibits characters significantly different from those of other parts of the unit'.
- Selley (1970) : "a mass of sedimentary rock which
can be defined and distinguished from others by its geometry, lithology, sedimentary structures, palaeocurrent
pattern, and fossils".
It is obvious from such definitions that a facies has a
necessarily limited extension, both stratigraphic and geographic, even if it can be found at different levels in the
same stratigraphic unit.
Geologists have noticed that a facies observed in a
stratigraphic unit can show similar features and characteristics to those described in other units of different ages or
coming from other regions of the earth. This is related to
the fact that such facies, with the same aspects, were
deposited under identical physico-chemical conditions.
Consequently, the term facies can be used with an
extended meaning for designating sedimentary rocks with
the same aspect, fauna and flora, even if they are different in age, reflecting similar physico-chemical conditions
Serralog Q 2003
in their environment of deposition.
In the following, the term facies will cover the more
general meaning given above. But, it will always be descriptive, without any genetic or environmental connotation. It will correspond to the general aspect of a sedimentary rock as it results from the sum of lithological,
structural and organic characteristics which can be detected in the field, and which distinguish this rock from other
surrounding rocks.
These present characteristics, on one hand, are the
results of the physical, chemical and biological conditions
under which the sediment was deposited, and, on the
other hand, are derived from its evolution under diagenetic influences since the time of its deposition.
It is from its characteristics and the context in which it
is found - vertical and lateral sequential evolutions, timespace relationship with neighbouring facies, regional tectonic control in the period of deposition - that one will be
able to determine its origin, its depositional environment
and its geological history.
Selley (1970) states "that a facies has five defining
parameters, viz. geometry, lithology, palaeontology, sedimentary structures and palaeocurrent pattern".
Generally, a facies is surrounded by other facies which
are related to it. This means that in a given environment
the facies are not randomly distributed, but constitute a
predictable association or sequence.
As pointed out by Selley (1970) "there are a finite number of sedimentary facies which reoccur in time and space
in the geological record' as "there are on the earth's surface today a finite number of sedimentary environments".
As pointed out by Middleton (1978) "it is understood
that (facies) will ultimately be given an environmental
interpretation". "However, many, if not most, facies defined in the field have ambiguous interpretation... The key
to interpretation is to analyse all of the facies communally in context. The sequence in which they occur thus
contributes as much information as the facies themselves" (Walker, 1984).
As also stated by Selley (1976) the idea of facies analysis "can be usefully extended to consider not just a vertical sequence, but a whole body of rock; what has been
termed a genetic increment of strata (Busch, 1971)".
197
"A genetic increment of strata (Fig. 5-1) is a mass of
sedimentary rock in which the facies or subfacies are
genetically related to one anothef (Selley, 1976). An
example of genetic increment would be one Bouma's
sequence.
"A genetic sequence of strata includes more than
one increment of the same genetic type" (Selley, 1976).
An example of genetic sequence would be several successive Bouma's sequences.
The general meaning of a sequence is a "succession
of geologic events, processes, or rocks, arranged in chronologic order to show their relative position and age with
respect to geologic history as a whole" (Bates and
Jackson, 1980).
Nomarine coal suMacfes
(swamp subenvironment)
Carbonaceous
sand fecies
(deltaic environment)
Carbonaceous sand subfacies
(delta front subenvironment)
Marine shale subfacies
(otrshmsubenvironment)
Figure 5-1 - Illustration of genetic increment and genetic sequence
(adapted from Selley, 1985).
Lombard (1956) has introduced the concept of lithological sequence that he defines as "a series of two lithological units, at least, forming a natural succession, without
any other important break except for the joints of stratification. The thickness of the bed is not considered'. He
distinguishes three orders of sequence:
- thin microscopic sequences (i.e. varves);
- medium macroscopic sequences (i.e. cyclothem);
- large megascopic sequences (i.e. stage, system).
Other concepts must be added. A granulometric
sequence corresponds to a grain size evolution without
change in mineralogy (i.e. coarse, medium, fine, very fine
sands). It can be fining upward, or coarsening upward.
A facies-sequence corresponds to a series of facies
which gradually merge into each other. The sequence
may be bounded at top and bottom by a sharp or erosive
junction, or by a hiatus in deposition. An example is the
Bouma's sequence.
Following the order of succession of the facies A, B
and C, or terms of the sequence, we have:
- a rhythm (Fig. 5-2), which corresponds to ABC, ABC,
AB, ...; such succession characterizes a rhythmic sedimentation and the results are rhythmites (e.9. cyclothems,
turbidites, varves);
198
- a cycle (Fig. 5-2), which corresponds to the succession of two sequences with opposite evolution ABCBA;
such succession characterizes a cyclic sedimentation.
The first type of succession is more frequent than the
second.
A lateral evolution or association of related facies
deposited at the same time, in different places in the
same environment but forming a continuum, creates a
lateral sequence; a succession of superposed terms in
relation to time corresponds to a vertical sequence (Fig.
5-3).
A
Figure 5-2 - Illustration of rhythm and cycle. In X, the three genetic
, and 111) show symmetric cyclic motifs: DCD, DCD,
increments (II1
DCD. In Z,the same three genetic increments are superposed with
asymmetric rhythmic motifs: AB, AB, AB (from Selley, 1976).
But, as new terms have been introduced by Vail,
Mitchum, Posamentier and others, (most of the time
without taking too much into account previous acceptations, definitions and works made essentially by european
geologists but also by north american geologists), it
seems important to clarify the sequence analysis concept
by going back to the basics and review the definition of
the terms now used in geology and sequence stratigraphy
(cf. Table 5-1).
Seauence
- "An unconformity-bounded stratal unit' (Sloss, 1948).
- "A relatively conformable succession of genetically
related beds or bedsets bounded at its top and base by
unconformities and their correlative conformities"
(Mitchum, 1977; Vail et al., 1977). "It is composed of a
succession of systems tracts and is interpreted to be
deposited between eustatic-fall inflection points" (ibid.).
Svstem tract
"A linkage of contemporaneous depositional systems "
(Brown and Fisher, 1977).
Paraseauence set (cf. Fig. 5-3)
"A succession of genetically related parasequences
forming a distinctive stacking pattern and commonly
bounded by major marine-flooding surfaces and their cor
relative surfaces" (Van Wagoner et al., 1990).
Serralog 0 2003
I
I
Facies, Sequence and environment Chapitre 5 199
Table 5-1
Stratal unit hierarchy
(modified from Van Wagoner et al., 1990).
Stratal Units
Thickness range
Area range
Time range
(feet)
(Sq. miles)
(Yeam)
Paraseauence
"A relatively conformable succession of beds or bedsets bounded by marine-flooding surfaces and their correlative surfaces" (Van Wagoner, 1985).
Bed set
"A relatively conformable succession of genetically
related beds bounded by surfaces (called bed set surfaces) of erosion, non-deposition, or their correlative conformities" (Van Wagoner et a/., 1990).
For Reineck and Singh (1975) "a simple bed set
consists of two or more superimposed beds characterized
by similar composition, texture and internal structure... A
composite bed set denotes a group of beds and bedsets
differing in composition, texture, or internal structure, but
associated genetically, representing a common type of
depositional sequence".
Tool Resolution
Use
surfaces) of erosion, non-deposition or their correlative
conformities" (Van Wagoner et al., 1990).
Lamina
"The smallest megascopic layet" (Van Wagoner et al.,
1990). For Payne (1942), it corresponds to a stratum less
than 1 cm in thickness. For Campbell (1967), the thickness of a lamina is lower than 3 cm.
But, as previously seen, sequence is also defined in a
more general meaning.
In accordance with this more general definition one
can recognize and adopt the other stratal units introduced
by Busch (1971) and Selley (1976) which are the genetic
increment and the genetic sequence or elementary
sequence.
Other types of sequences must be introduced as well.
Litholoaic sequence or Iithoseauence
"A relatively conformable succession of genetically
related laminae or lamina-sets bounded by surfaces (called bedding surfaces) of erosion, non-deposition, or their
correlative conformities" (Van Wagoner et al., 1990). From
the Glossary of Geology, 1980, it is "the smallest formal
unit in the hierarchy of lithostratigraphic units".
Lamina set
"A succession of, at least, two lithologic units (genetically related) forming a continuous succession without
interruption other than the stratification surfaces"
(Lombard, 1956).
Facies tract ("faciesbezirk" de Walther)
"Conformable vertical sequence of genetically related
facies" (Walther, 1893, 1894) generated by a lateral
sequence of environments (in Selley, 1970).
"A relatively conformable succession of genetically
related laminae bounded by surfaces (called lamina set
Serralog 0 2003
199
Facies-seauence
I
I
“A series of facies which pass gradually from one into
the other. The sequence may be bounded at top and bottom by a sharp or erosive junction, or by a hiatus in deposition indicated by a rootlet bed, reworking or early diagenesis. A sequence may occur only once, or it may be
repeated (i.e. cyclic)“ (Reading, 1978).
Virtual series
“A succession of lithofacies which are superposed
without major discontinuities“ (Lombard, 1956). It would
correspond to the logical succession of lithofacies if the
sedimentary process could evolve as suggested by the
previous deposits and was not interrupted due to external
causes.
Textural seauence
A succession of, at least, two types of texture occurring
in a same lithologic unit (i.e.: bounstone, grainstone,
packstone, wackestone and mudstone textures in limestone; grain size evolution: coarsening-up or fining-up
sequences in sandstones; change in sorting in a sand
body, etc).
Seauence of structures
A succession of, at least, two types of sedimentary features occurring in a same lithologic unit (i.e.: massive,
laminated, cross-bedded, and laminated sands composing units A, B, C and D of the Bouma’s sequence).
Seauence of thickness
A succession of beds of same lithology, with decreasing or increasing thickness upward (i.e.: thickening-up or
thinning-up sequences observed in turbidity deposits)
(Mutti & Ghibaudo, 1972; Ricci-Lucchi, 1975 in Walker,
1984).
J
Figure 5-3 - Progressive development of genetic increment and genetic
sequence with parasequence boundaries. Stage one corresponds to a
progradation of parasequence A during a time when the rate of deposition exceeds the rate of water depth increase. Stage two corresponds
to a rapid water depth increase flooding the top of the parasequence A
creating a surface of non deposition with respect to siliciclastic sediment. Thin carbonates, glauconites, organic-rich marls, or volcanic
ashes may be depoited on the surface. Stage three corresponds to a
progradation of the parasequence B during a time when the rate of
deposition exceeds the rate of water depth increase. Bedsets in parasequence B downlap onto the boundary of parasequence A. There is
an abrupt deepening of facies across the boundary of parasequence A
(from Van Wagoneret al., 1990).
Walther introduced the term facies tract (faciesbezirk)
for a conformable vertical sequence of genetically related
facies and observed that the various deposits of the
same facies area and similarly, the sum of rocks of different facies areas, were formed beside each other in
space, but in a crustal profile we see them lying on top of
each other... it is a basic statement of far-reaching significance that only those facies and facies areas can be
superimposed, primarily, that can be observed beside
each other at the present time” (translation from Blatt et
a/., 1972). This principle is at the origin of the Walther’s
law which can be stated as “a conformable vertical
sequence of facies was generated by a lateral sequence
of environments” (Selley, 1976) (Fig. 5-4).
“
Environments A
Importance of facies and sequence analysis
The study of facies and their arrangement or association in lateral and vertical sequences is the only way to
establish the depositional environment, and thus to
reconstruct the palaeogeography at the time of deposition. The physical, chemical and biological conditions
existing in an environment, which define it, can only be
determined by its imprints on the sediments. Among these
imprints, the primary sedimentary features are more
important because they have been formed in-situ. The
sequences will reflect the modifications in the conditions
both in space and in time.
Figure 5-4 - Cross-section illustrating the Walther’s law. Environments
A, B and C deposited during the same time period the lateral sequence
of facies a’, b’and c‘respectively in the increment of sedimentation
designated I. The increments I, I1 and 111 together constitute a genetic
sequence of strata. The same facies a’, b’ and c’ are superposed in
position X (adapted from Selley, 1976).
As pointed out by Walker (1984) “careful application of
the Walther’s law, therefore, suggests that in a vertical
200
Serralog c.c3 2003
I
I
Facies, Sequence and Environment Chapter 5 201
sequence, a gradational transition from one facies to another implies that the two facies represent environments
that once were adjacent laterally. The dangers of applying
the law in a gross way to stratigraphic sequences with
cyclic repetitions of facies have been emphasized by
Middleton (1973)".
De Raaf et al; (1965) and Reading (1978) indicated the
importance to clearly define gradational facies boundaries
in vertical section as opposed to sharp or erosive boundaries. In case of sharp or erosional boundaries between
two facies there is no way of knowing whether the two vertically adjacent facies represent environments that once
were laterally adjacent. Sharp breaks between two different facies may signify fundamental changes in depositional environment and the beginnings of new cycles of sedimentation.
Based on these remarks, as one can imagine, it is very
important to determine the type of boundaries between
facies.
Walker (1976) made a comparison between the results
of facies sequence analysis and facies models (established on the basis of modern environments). This comparison allows, by analogy, the determination of the depositional environment by application of uniformitarianism or
actualism, theories introduced by Hutton (1785) and Lyell
(1830): "the present is the key to the past'.
Van Wagoner et a/. (1990) indicate the necessity of
facies determination and facies succession for the determination of the depositional environment and for sequence stratigraphy applications (Fig. 5-5).
Traditional approach of facies analysis
The determination of the facies, sequence and environment is based on the observation and description of
the traditional geological objects: outcrops and cores. The
results and synthesis of this study is perfectly illustrated
by Van Wagoner et al. (1990) as previously seen (cf. Fig.
5-5).
Many books have been written on the facies models
(Walker, ed., 1984) and sedimentary environments in
which the traditional approach for determination of facies,
sequences and environments is abundantly described
and illustrated (Krumbein and Sloss, 1951, 1963; Selley,
1970, 1978, 1985 and 1976; Schwarzacher, 1975;
Reineck and Singh, 1975; Reading, 1978; Friedman and
Sanders, 1978; Scholle and Spearing, ed. 1982; Scholle
et al., ed. 1983; Galloway and Hobday, 1983; Davis, ed.
1978, 1985; Selley, 1988...). Please refer to these books,
if needed!
Well logging approach
As demonstrated in the previous chapters, practically
all the facies attributes (Table 5-2) can be extracted from
well logging data, as soon as a certain methodology is
applied.
Serralog 0 2003
Table 5-2
Sedimentary rock characteristics. Comparison of the
geological approach and the well logging approach.
Geologicalapproach
Fades
Well logging approach
Electrofacies
I
Definition
I
Definition
Set of geobgiial attributes that charactefees ]Set of w ell lcg attributesthat characterizesan
a bed. abws its differentiationfromthe
electrobed and abws its differenfiationfromthe
surrounding beds and reflects the conditions surrounding electrobeds
(physical, chenicaland biological and so
clirretic) of its orun
Attributes
WOlcgY
- Conposition
- ninerakgical
Attributes
BecboWhofacies
Wo-density. hydrogenindex, natural and induced
radioacWty, sound sbw ness. resistivity data...
- elemental
- Texture
Natural and inducedradicWty ...
(grain size. sorting. pore size. pore type and
distribution. p cking... )
S.P.. reskW!. porosw took. borehole images.
dipmeter curve data and nuclear magnetic
resonance
hternalsbuctures: energy, transport current
direction, environmnt
Borehole images. dipmter data
FossiS : age. environmnt
Up data, correlabns between w e b
-Width
I
-apparent
- real
I
I
Standard bgs
Upmeter data, borehole m
e
s
It appears clearly that practically all the parameters or
attributes which determine a facies can be extracted from
well logs, if we except color and fossils. Consequently, it
seems quite reasonable to admit that a valuable facies
recognition can therefore be achieved with log data. We
call this facies determination electrofacies to indicate its
origin, and its definition is : " the set of well log responses
which allows differentiation of an electrobed from its surrounding electrobeds".
It is obvious that :
- a complete set of open hole logs, possibly calibrated
on core data, allows the determination of the lithology, and
even the principal minerals composing the rocks, with a
good accuracy (cf. Chapter 2);
-as previously explained and illustrated, electrical images provide information about texture, sedimentary features, the nature of surfaces, organic activities and thickness, as well as diagenetic activities. In addition, they
identify significant geological events or markers with limited vertical extent such as hardgrounds, concretions,
vugs, roots, burrows, clast lag, debris, breccias and
conglomerates (cf. Chapter 3 and 4).
Consequently, most of the attributes that characterize
a facies can be extracted from logs and, so, can be used
to define it. In addition, interpretation of the image data by
a program similar to BorTex provides a useful description
201
GRAIN
SIZE
TROUGH CROSS B€DS
TROUGH CROSS BEDS - SMALL SCALE
SlGMOlOAL CROSS BEDS
CURRENT - R
W SEEMW(0
CONTORTEO BEDS
PLANAR BEDS
HUMMOCLV BEDS
WAVE
wrnr~omffi
c u v CUSTS
AWNOANT
COMMON
RARE
ABSfNT
-
HST = HIGHSTAND SVSTEMS TRACT
TST TRANSGRESSIVE SVSTEMS TRACT
LST - LOWSTANO SYSTEMS TRACT
CHURNED BV BURROWS
Figure 5-5 - Typical type of formation description which includes grain-size data, facies characteristics, depositional environment and sequence
stratigraphy. All these steps can be achieved from well logging data, including obviously borehole image data, with a high vertical resolution
(from Van Wagoner, et al., 1990).
202
Serralog 0 2003
Facies. Seauence and Environment khaoter 51 203
of the internal organization and the thickness of each unit
of deposition composing the formation. This information
completes the lithology determination using the standard
logs.
However, as the lithology and sometimes the texture
observed in the present day result from two factors:
- the original dynamic and biological conditions
prevailing at the time of deposition
- the subsequent diagenetic effects on the sediment since its deposition,
the original facies cannot be accurately determined by
relying only on lithology and textural parameters from
porosity levels or SP/gamma ray vertical evolution. When
sedimentary features are identified in images, however,
the original facies can be determined because these features reflect the dynamic and biological conditions prevailing during the deposition. In the same way, conglomerates, roots, hard-ground or burrows signify certain conditions which cannot be extracted from the standard logs.
All this information is critical for accurate facies and
sequence recognition.
Most of the time diagenetic effects do not significantly
modify the internal structure or sedimentary features,
except in some carbonates. Many of the common diagenetic effects can be identified on images (cf. Chapter 6).
For instance, cemented nodules or layers, dissolution features, and stylolites are easily recognized. Diagenetic
sequences can also be determined from log data (cf.
Chapter 6). In any case, these diagenetic features characterize diagenetic environments which are usually linked to the depositional environments.
From the previous remarks it seems evident that
image data should be recorded at least once in a field.
Whenever possible, they should be calibrated on core
description and analysis. Finally, they should be run in
other wells or replaced by dipmeter data if typical curve
responses are demonstrated to reflect features seen on
the images.
As soon as the facies of each unit composing a formation is determined, the vertical facies succession, combined with the surface type (hardground, erosion, truncation) or features (roots, burrows) observed on images,
and completed by information on facies thickness, enables the determination of genetic increment units and
genetic sequences (see next topic).
The overall depositional environment can then be derived from these data. In addition, breaks of higher order
and wider extent can be detected, which is important for
sequence-stratigraphy analysis.
zing the SP curve, the type of contact (abrupt or progressive) between sands and shales, and the character of the
curve (smooth or serrated; concave, rectilinear or convex)
one can establish the classifications shown in Fig. 5-6.
UPPER BED CONTACT
Figure 5-6 - Classification of SP curve shapes as a function of the
curve vertical evolution (adapted from SHELL-PECTEN).
Pirson (1970, 1977) associates a facies and a depositional environment to each shape, and he interprets the
curvature of the curve as an indicator of the speed of
transgression or regression processes (Fig. 5-7).
Fig. 5-7 - Classification of SP curve shapes in terms of sedimentary
patterns (courtesy of Pirson, 1970, and Gulf Publishing Co., fig. 2-1).
Several geologists (Visher, 1965; Serra & Sulpice,
1975; Coleman & Pryor, 1980; Galloway & Hobday, 1983)
have used log patterns for several purposes:
- vertical grain size evolution,
- determination of vertical sequence,
- recognition and mapping of log facies
- interpretation of depositional environment.
Historical
It seems that the idea of using wireline logs as sedimentological tools, first came in 1956-1957, from engineers working for the SHELL-PECTEN Company in
U.S.A. Studying the Mississippi delta, they stated that the
spontaneous potential curve (SP) presented characteristic shapes. Each of these shapes corresponds to a facies
or facies succession of a particular sand body. By analySerralog 8 2003
Figures 5-8 and 5-9 (see next pages) summarizes
some of the log patterns and their interpretation.
Several geologists have used this very rapid and synthetic method of log pattern analysis to construct facies
maps (see Fig. 5-10 next pages).
203
1
3
2
4
Figure 5-8 - Examples of interpretation of log patterns (SP curve), grain size and sedimentary structures.
I - Case of a low-sinuosity braided channel. Sequence A dominated by migration of a gravely longitudinal bar. Sequence B records deposition of
successive transverse bar cross-bed sets upon a braid channel fill.
2 - Case of a laterally accreting (A) and symmetrically-filling channel segments (6) of an anastomosed channel system.
3 - Case of a meanderbelt sand body produced by a high-sinuosity channel. Sequence A illustrates a complete fining-upward sequence typical of
the mid- or downstream point bar. Section B illustrates the truncated vertical sequence commonly found in the upstream end of the bar.
4 - Case of a chute-modified point bar. Sequence A corresponds to upstream portions of the point bar capped by chute-channel deposits.
Sequence B corresponds to downstream, the channel and lower point bar deposits are capped by chute-bar sediments
(adapted from Galloway & Hobday, 1983).
204
Senalog 'B 2003
Next Page
Facies, Sequence and Environment (Chapter 5! 205
1
3
2
4
Figure 5-9 - Examples of interpretation of log patterns (GRcurve), grain size and sedimentary structures.
1 - Case of an upward-coarsening parasequence formed in a beach environment on a sandy, wave- or fluvial-dominated shoreline.
2 - Case of a similar upward sequence but formed in a deltaic environment on a sandy, wave- or fluvial-dominated shoreline.
3 - Case of stacked upward-coarsening parasequences interpreted as formed in a beach environment on a sandy, wave- or fluvial-dominated shoreline where the rate of deposition equals the rate of accomodation.
4 - Case of two upward-fining parasequence interpreted as formed in a tidal flat to subtidal environment on a muddy, tide-dominated shoreline.
FS = Foreshore. USF = Upper shoreface. LSF = Lower shoreface. D.LSF = Distal lower shoreface. SH = Shelf. OSMB = Outer steam-mouth bar.
DF = Delta front. PRO D = Pro delta. CP = Coastal plain. SBT = Subtidal. INT = Intertidal. SRT = Supratidal.
(adapted from Van Wagoner et at., 1990)
Serralog Q 2003
205
Previous Page
206
Well logging and Geology
Knowing the fundamental reasons for the choice of the
SP or GR curve (other logs often being unavailable, resistivity curves being heavily affected by the presence of
hydrocarbons, large number of wells to be studied, etc.)
and its real possibilities in detrital sand-shale series (in a
majority of cases it reflects shaliness and grain size evolution), we have to acknowledge that the use of such
curve alone is often insufficient to clearly determine a
facies and a depositional environment. It may lead to a
misreading of the events because of "parasite effects" on
SP deflection (influence of R,, invasion, contrast of RR
/, ,
compact zones, thickness of beds, etc.). Finally, this curve
becomes unusable if the contrast of R,/R,
is insufficient.
The GR response may be affected by the presence of
feldspars, micas, zircon, uranium salts, etc., which may
be confused with clay content or grain size evolution. This
is why the shape of a single curve may only exceptionally define the facies and the depositional environment, particularly if we intend to use this method to study other
types of sediments such as carbonates and evaporates.
covering their chemical and mineralogical composition,
their texture and their structure. The higher the number of
well logs used, the richer the spectrum will be, and the
better defined the rock characteristics. Hence there will be
less risk of ambiguity and error in their interpretation.
Figure 5-11 - Facies determination from GR curve
(from Serra & Sulpice, 1975).
Moreover, dipmeter data processed by GEODIP program for HDT tool, or LOCDIP program for SHDT tool, or
image data allow, in many cases, the detection of sedimentary structures and the determination of the palaeocurrent pattern and the direction of transport.
Finally, the geometry is defined by both analysis of the
true thickness of the beds (only dipmeter or image data
provide this information), and by the lateral extent of beds.
This can only be defined by correlations between several
wells and by the drawing of isopach maps based on thickness data (Fig. 5-12 and 5-13).
G
H
Fig. 5-10 - Distribution of facies based on the shape of the resistivity
curve (from Lennon, 1976).
The electrofacies concept
As we have seen in previous chapters, every well logging measurement gives, more or less, some information
about the mineralogical composition, the texture, and the
sedimentary structures, even if this information is sometimes implicit. In other words, each well log gives a particular spectral picture of the rock properties.
In certain cases one or two spectral pictures, therefore one or two logs as we have previously noted, are sufficient for the determination of rock characteristics (for
instance, the use of the shape of the SP curve in the
sand-shale series of the Gulf Coast, or the use of GR
curve in Algeria : cf. Fig. 5-11). But, it is preferable to use
all the available log data for interpretation. Their number,
their diversity and their complementarities allow, in fact,
the establishment of a spectrum of rock characteristics
206
Figure 5-12 - Log-correlation between wells in the South Glenrock
Oilfield, Wyoming. Bar, beach and channel fill are recognized
(from Curry & C u q , 1972).
Most of the parameters defining the facies, or at least
the lithofacies in Moore's definition (1949), are directly
accessible from logs. The latter, therefore, produce a picture of the present facies. This picture is certainly particular, incomplete, sometimes confused, but always permanent and objective. If the set of well logging measurements is diversified and rich enough for a better covering
Serralog 0 2003
Facies, Sequence and Environment 'Chapter -5
il
of geological parameters, the picture will be sufficiently
precise. In other words the spectrum will be sufficiently
rich and detailed to permit a new representation of the
facies by means of log data.
J
Figure 5-13 - lsopach map of the lower Muddy in the South Glenrock
Oilfield, Wyoming, showing two buried streeam channels, based on
Figure 5- 12
(from Curry & Curry, 1972).
This parafacies was named by Serra (1970) electrofacies, the definition of which is:
"the set of log responses which characterizes an electrobed and permits it to be distinguished from the others"
(Serra, 1972; also in Schlumberger Well Evaluation
Conference, Algeria, 1979).
All log responses (electric, nuclear, acoustic, dipmeter,
images, etc.), that indicate the quantitative (log values
and dip data) as well as qualitative aspects (curve characteristics, textural and structural data) represent, therefore, the component elements of the electrofacies.
Electrofacies constitutes more than one element of a
facies. It is, in fact, its equivalent since it includes in itself
the parameters which define the facies.
But we have to realize that there is a parameter which
has never been taken into consideration by sedimentologists in their definitions of facies. This parameter is the
fluid that occupies the porous space of the rocks. If in surface outcrops it is neglected because it is absent or
without significance, it is always present in the subsurface and it influences the response of most logging tools.
Therefore, we cannot eliminate it and, in fact, it enters into
the definition of electrofacies. Consequently, several electrofacies, depending on the nature of fluids present in the
rocks (gas, oil, fresh or salty water), may correspond to
the same geological facies. This situation may be considered, at first sight, as an important disadvantage of the
electrofacies concept and of its utilization. In fact, it is not
important because the purpose of the electrofacies analysis is, first of all, to describe the formations as they are
"seen" by logging tools. After all, we can possibly utilize
logging tools, and their data, less sensitive to fluids or to
Serralog 0 2003
207
i
"
Y
l"lil(_)
porosity (i.e. natural or induced gamma ray spectrometry,
photoelectric index), or we can correct the log response
for porosity and fluid influence. For all that, in the absence of hydrodynamism, the fluid may be an important factor for the recognition of a depositional environment: fresh
water in fluvial or lacustrine environment, briny water in
swamp or marsh, salty water in sands deposited in marine environments...
The objection has often been raised that the electrofacies is only an equivalent of the lithofacies because it
does not contain palaeontological information. Without
being argumentative, it is necessary to make the followings remarks.
1) If the fossils are utilized as indicators of the depositional environment, we have to remember that in many
cases the fauna and flora are almost nonexistent, especially in sandstone deposits, and, consequently, the facies
recognition is realized without this information;
- the fauna and flora are not always good indicators of
the environment (ubiquitous species, mixtures, allochtone
species, etc.);
- in many cases the presence of macrofossils (animal
or vegetal) is shown by their influence on log responses,
particularly on dipmeters and images. Moreover, it is
controlled by the physico-chemical conditions existing at
the time of deposition, which also determine the other
parameters, especially the sedimentary structures;
- other elements of facies (mineralogy, texture, sedimentary features, palaeoccurents and geometry) are
often sufficient for a precise definition of the facies, and
also for specifying the depositional environment, especially if we use the additional information on sequential
evolution.
2) If the fossils are used to define the geological age,
we may note that this information is implicitly included in
the subsurface data through the depth data and the position in relation to markers. Besides, at this stage, the logs
allow a more precise appraisal of the lapse of time than
that defined by fossils.
Hence, we have good reasons to assimilate the electrofacies into the facies. Moreover, the reconstitution of
the time-space repartition of the different facies, and the
definition of their mutual relations is the final goal of facies
and sequence analysis, regardless of the methods used
to achieve this goal (traditional method by examination of
rocks, or from log data).
Depositional environments will therefore be defined by
these analyses, and thus we may forecast more accurately the continuity of a reservoir, the presence, nature and
distribution of permeability barriers, and the location of
mineral resources in exploitable economic quantities.
/J
Electrofacies analysis from well logs
The goal of this analysis is to describe objectively the
formations penetrated during drilling, through their well
logging responses, and to recognize all the different fundamental electrofacies present, in such a way that ultimately it will be possible to study their association in vertical
207
sequences, and, consequently, deduce the lateral evolution by applying Walther's law. In other words, the first
goal is to reconstruct the electrofacies models which will
help to define the depositional environments.
An electrofacies analysis can be carried out manually
or automatically. In both cases the basic approach is
essentially the same.
DEPTH
Track 1
Track 2
Manual identification of electrofacies
Originally a sedimentological study from logs involved
examining the shapes of various curves for indications of
the type of sedimentation and the depositional environment. As previously seen, a ramp or a gradient on a log
could indicate an upward-fining or coarsening of grains
(cf. Figs.5-6, 5-8 and 5-9). Classification of electrofacies
by the shape of spontaneous potential or GR response is
well known and has been used for many years. But, as
drawbacks of SP or GR curves may occur (Fig. 5-14 and
5-15), it is highly recommended to use all the available
data.
Figure 5-14 - Based on the GR curve evolution the interpretation
would be coarsening-up sequence followed by fining-up sequences.
DEPTH
Figure 5-15 - The complete set of log allows the recognition of dolostone invaded episodically by arkosic deposits as perfectly illustrated by the
potassium content. The shaly layers are indicated by peak of thorium with a value above 8 ppm.
Feldspathic dolostone
Dolostone
One starts from a contrived document, the compositelog. This document gathers together and depth matches
all the log data recorded in a well, including dipmeter data
and dip computation results obtained by GEODIP- Or
LOCDIP-processing types and borehole image data. It
can be obtained on a workstation (Fig. 5-16).
208
Arkose
One divides the studying interval into electrobeds or
electrosequences. For this purpose the amplitude of the
variations on macrodevice curves is analyzed. According
to its importance and its shape, it is subjectively decided
- either it corresponds to an electrobed boundary
- or it indicates a "noise" that is inherent either to the
Serralog Q 2003
Facies, Sequence and Environment IChapter 51 209
Figure 5-16 - Composite-log combining results of a quantitative interpretation of standard logs, borehole images and dip data with cross-plot of
density vs neutron measurements (courtesy of Schlumberger).
measurement (statistical variations of nuclear measurements), or to hole conditions (borehole wall rugosity, presence of caves, etc.), or to minor changes in the geological parameters;
- or it reflects a gradual evolution (ramp) with minor
variations.
In the following step one determines the electrolithofacies for each electrobed presenting a thickness greater
than the average vertical resolution of most macrodevices
(about 2 to 3 feet, or 60 to 90 cm).
As the synthesis of different measurements, corresponding to the same electrobed (especially when it
concerns a large amount of data) is not easy, a representation using rosettes, spider's webs (Fig. 5-17a), or histograms has been proposed to visualize the electrofacies
(Serra, in Schlumberger Well Evaluation Conference,
Algeria, 1979). As histograms cannot easily be obtained
manually, and a spider's web varies in shape according
the number of involved measurements (hence the branches), a ladder diagram presentation (Serra, in
Schlumberger Well Evaluation Conference,Algeria, 1979)
seems to be more usefull because the absence of one log
Serralog Q 2003
data will not modify the general shape of the figure (Fig.
5-17b). In these representations, each branch of the spider's web, or each bar of the ladder, represents a scaled
log axis with is range of variation. One plots on these
basic documents the minimal, maximal and median
values of each log on the interval corresponding to the
electrobed .
When the points on each axis are joined, a characteristic shape is formed. For each electrofacies, there will be
an allowed band on each axis; hence, an allowed area is
created corresponding to one electrofacies.
Figure 5-18 shows the shapes of two electrofacies.
Two shapes need to differ in only one axis to establish the
difference between two electrofacies. Essentially, the
comparison of shapes allows an analyst to break down a
logged interval into some 10 to 15 electrofacies (Fig. 519). This last figure shows the progressive change in shapes from one electrofacies to the next. The interval covered is shown in Figure 5-20. The shallowest depth is at the
top left corner, and figures are arranged in columns with
depth increasing downward. Correlation with facies defined from core analysis is also made.
209
Next Page
210
Well logging and Geology
A correlation between electrofacies from different wells
can also be achieved by this technique (Fig. 5-21).
Figure 5-1 7 - After logs have been zoned and electrofacies established, the log parameters are plotted on rosette or spider's web diagrams (left image), or on superposed bars of a ladder (right image).
(from Serra, in Schlumberger Well Evaluation Conference, Algeria,
1979).
MSFL
MSFL
Figure 5-18 - Spider's web diagrams for (a) a limestone and (6) a
sandstone (from S e r a & Abbott, 1980).
this is a long and very tedious operation. Therefore, the
following empirical method may be preferable. One draws
the electrofacies representation of the thin bed and indicate, for each measurement, by an arrow the direction
toward which the correction should displace the representative point. By comparison with preliminary defined
electrofacies, and by using dipmeter or image resistivity
curves, it is possible to estimate the closest electrofacies
(Fig. 5-22 & 5-23).
In the case of ramps or electrosequences, one defines
the electrofacies of the starting and stopping depths or of
the surrounding electrobeds.
This manual analysis is often long and sometimes
tedious. Certain traditional geologists may consider it as
uninteresting. In this connection we remind these geologists that this method is used by some oil companies in
the world, perhaps not in the above mentionned way, but
at least in spirit. For example this method allowed the
ELF-Aquitaine group to study more than 1000 wells and
to have, thus, a synthetic idea of the facies and environments for each well. They were able to draw facies maps.
It was done at low cost and in a very short time.
Otherwise, it would have been impossible to obtain this
information, because cores were rarely cut and the core
analyses were often unavailable (exchange wells). This
method enabled a more accurate covering of the seismic
profiles and a more reliable interpretation in terms of seismofacies. This gave the best possible knowledge of the
basins and consequently a more judicious and justified
choice of prospects.
J
-
Figure 5-19 Comparison of 20 electrofacies (from Serra & Abbott, 1980).
Determination of the electrofacies in thin beds requires a
preliminary step : the correction of the different measurements for the influence of the surrounding beds. For this
purpose environmental
correction charts are used. But,
210
But this analysis is subjective, because the results
may partly depend on the analyst performing the study.
For this
specialists dreamed for an automatic process using computers. The computer processed method
Serralog Q 2003
Previous Page
~~~~~
Facies, Sequence and Environment [ C h a p t a 211
(Serra & Abbott, 1980) described hereafter, was developed by Schlumberger and commercialized under the mark
of FACIOLOG. Its description is destined to explain the
different steps of data processing. Other approaches
could be imagined, and will doubles be developed. They
will certainly be inspired by the same general philosophy.
Automatic Electrofacies Identification : FACIOLOG
Essentially the same steps are taken as when applying
GtOMP
CURVES
Figure 5-20 - Correlation of electrofacies wifh core facies (from Serra 8, Abboff, 1980).
Well B Fades 2
Well A Facies 2
SHALE
Figure 5-21 - Correlation of electrofacies (from Sera & Abboff, 1980).
Serralog 0 2003
the manual method. We simply try to translate the
approach of the analyst into mathematical functions or
statistical processing.
The set of n log responses that characterizes an electrobed, or level of reading, may be considered as defining
the coordinates of electrofacies (here represented by a
point) in a n-dimensional space. Since the same causes
produce the same effects, we may think that another bed
with same geological facies and containing the same
fluids will have the same electrofacies. Consequently, its
representative point in n-dimensional space has to be
very close to the previous point. Hence, an electrofacies
must correspond to a cluster, or to a cloud of points very
close to each another in this space. Contrarily, the distinct
electrofacies must correspond to different and separated
clusters. They can possibly overlap in one or several
dimensions of the n-space.
If we start with "raw", unzoned data, we will observe a
certain dispersion of points corresponding to one electrofacies, and it will be more dispersed with a higher number
of log measurements. This dispersion is related to the
"noise" due to the tool, to hole conditions and even to
weak variations of the geological parameters. A preliminary zoning and an analysis of the principal components
211
THEK
LIGNITIC
BED
Figure 5-22 - Composite-log associating dipmeter (HDT) data.
GR Pb -
'h4b-
At -
L,
RL c
Figure 5-23 - (a) :Electrofacies of the coal bed at 1241-1243 m. (b) :electrofacies of the thin bed at 1238-1238.5 m which corresppnds to a coal as
indicated by dipmeter resistivity curves.
will, on the one hand, decrease this dispersion, and on the
other hand, reduce the dimensionality of the space (cf.
Table 5-3). Only after this step, can an automatic clustering be carried out.
We can, in certain cases (the absence of ramps and of
thin beds), immediately start the analysis of the principal
components and carry out a clustering. The results of the
processing leads to an automatic zoning (Fig. 5-24).
The use of automatic zoning concentrates the cluster
by attempting to eliminate measurement errors. If an ndimensional histogram is created and the frequency of
each cell analyzed, we obtain results as in Figure 5-25a.
The results on the same interval after zoning are shown in
212
Figure 5-25b. Obviously, the data distribution shown in
Figure 5-25b is easier to handle. Figures 5-26a and 5-26b
show the corresponding frequency plots on two selected
axes of the interval.
Automatic seamentation of loas
The clustering technique does not allow the detection
of natural ramps (or electrosequences) on curves. It is the
reason why another approach has been developped for
an automatic segmentation of log data into electrobeds
and electrosequences (ramps).
Serralog 0 2003
Facies, Sequence and Environment
a
b
Figure 5-25 - Frequency plots of the log data corresponding to Table
5-3. (a) :unzoned cross-plot; (b) :zoned cross-plot of density vs neutron (from Serra & Abbott, 1980).
Table 5-3
Number of cells with their number of elements.
Variable
I
Mnirmm
11
Maximum
I
Steps
I
Figure 5-24 - Example of automatic zoning of logs obtained by clustering techniques and comparison with raw logs (from Serra, in
Schlumberger Well Evaluation Conference, India, 1983).
An electrobed can be defined as a succession of levels
with contiguous reading, according to the sampling rate
used (6 or 1.2 inches). These levels present essentially
the same values. This means that their response variations do not exceed certain limits (allowed variations).
These variations correspond to the tool error in the measurement, and to minor acceptable changes in geological
parameters (Fig. 5-26).
These changes may be expressed in terms of measurement error or may follow a more complex law.
An electrosequence can be defined as a succession of
contiguous readings. The level having an order n, shows
a value higher than that of level with order n - 1, but lower
than that of level with order n + 1, or vice-versa. This evolution must continue in a depth interval thicker than the
vertical resolution of the measuring tool, to be considered
as representative of a natural electrosequence.
Otherwise, it corresponds to an artificial ramp, due to the
lack of resolution of measuring tool (Fig. 5-26).
Serralog Q 2003
213
Allowed variations on the measurement
including statistics and mlnor geoiogical changes
This symbol wmsponds to a sample
(measurement )
A thin electrobed Inside a thicker one
tConsaquently,wrrectlons must be applied
to roach the roal value of this thin electrob4
-Mean value of the bed
:samp#ng rate:
oach half a foot In this case.
or each Inch In case of high resolution recording
ty of points). The task of the clustering program is then to
distinguish each cloud. For this purpose, we determine,
for each level or reading depth, the distance dk of the k'
closest neighbour. It corresponds to the radius of the
smallest circle with its center at a known level and containing k neighbouring points. If any other level, situated
below or above a certain depth window (BAND) does not
have a lower dk value, and if among these neighbouring
k none of them has a lower dk value, a level will be taken
as a cell, or as a local mode (Fig. 5-29). Each cell obtained in this manner corresponds to an elementary electrofacies.
Natural ramp'
or
electrosequence
0.G
c
Mean value of the bed.
A thldc electrobed
Real thickness
of the middle
electrobed
Artifk4ai ramp
A thick electrobed
Real thickness
of the lower
electrobed
~Verticairemlutlon of the measurement
,expressedin numbor of samples
Allowed variationsbn the measurement
nciuding statistics and minor geological changes
Loo i
Figure 5-26 - Principle of the automatic segmentation of logs into electrobeds and electrosequences.
Hence, it is necessary to define for each tool, the vertical resolution and the amplitude of the accepted variations, expressed as a percentage. We can state that the
effect is minor if we do not exceed a certain value.
Generally, we carry out the zoning from active logs, on
which we rely to determine the limits and the type of evolution. We impose these limits to other so-called passive
logs. The type of evolution between these limits defines,
for passive logs, either a bed or a sequence. Figure 5-27
(next page) gives an example of automatic log segmentation obtained with the help of three active logs gamma ray,
neutron and density.
Clusterina techniaues
The method described in the previous section is
essentially a way of trying to divide n-dimensional log
space (n corresponds to the number of different log data
taken into account) into definable volumes corresponding
to each electrofacies (Fig. 5-28).
Many of the mathematical techniques known as clustering are adaptable to such problems. In more than two
dimensions, we can think of the electrofacies points falling
into clusters or clouds. (A cluster may be described as a
continuous region of the n-dimensional space containing
a relatively high density of points, separated from other
such regions by regions containing a relatively low densi214
Figure 5-28 - Three-dimensional case of three electrofacies on a threelog axis (from Serra & Abbott, 1980).
The shape of the clusters is difficult to define, as one
cannot assume, automatically, that the distribution within
each facies, for each log, is normal. A cloud is dispersed
in any direction by the effects of log errors, as described,
or by changes in the facies itself. Consider, for example,
the grain size variation within a sandstone. This will
disperse the sandstone points, leading to a "cloud" with a
concentration of points at one end and a tail. In the case
of a gradual change from one facies to another, the actuation is even more complicated since the clouds are not
separated. In most cases at facies interfaces the differing
vertical resolution of the logs introduces some scatter in
the response. For these reasons statistical techniques are
needed to establish criteria to separate clusters.
In deciding what kind of clustering technique to use,
several things are important
- the number of levels to be examined at any time;
- the number of variables (logs) to be considered;
- the size and distribution of each cluster;
- the occurrence of electrofacies with only limited
representation;
- the effects of gradational changes from one cluster to
another.
It is desirable to introduce information provided by a
dipmeter or by borehole image to really determine the
Serralog 0 2003
Facies, Sequence and Environment IChapter 51 215
Figure 5-27 - Example of log-segmentation results (lower figure) compared to the composite-log of raw data before segmentation (upper figure).
Serralog (9 2003
215
electrofacies. In that case this information must be quantified. This is achieved, for dipmeter, through the dipmeter
derived synthetic logs previously described and obtained,
for instance, from the SYNDIP program for dipmeter and
through BorTex program results for image data.
For data where the significance of changes over the
range of a log is heavily dependent on the log value, some
nonlinear-scale changes are necessary. For example,
resistivity values that vary from 0.1 to 10,000 are transformed to logarithm base 10. This makes the processing
easier and allows the use of a metric standard.
points for a partitioning algorithm, working not on the histogram but on the original data.
Single-link clustering methods on the original data
have not been used because of the amount of data and
the effects of chaining. However, clustering directly on
constant values within zones after zoning gives a reduced
data set and works well. This is, in effect, a method of
segmentation followed by clustering.
Figure 5-30 - Schemes explaining the clustering of points info local
modes and terminal modes.
X
Figure 5-29 - Determination of local modes or elementary electrofacies.
Final electrofacies determination
The previous described method defines a certain number of local modes (or elementary electrofacies) that is
much smaller than the original number of levels. A list of
the local modes classified following the order of creation
is also produced. It indicates the depth of the most representative value, the dk value, and the link with other
modes. The coordinates of the principal components are
also given.
Even if they describe logging reality better and more
objectively (compare zoned logs with raw logs of Fig. 524), these local modes are often too numerous to be easily correlated to geological facies. The analysis of their proximity to each other can group some elementary modes to
reduce the number of final electrofacies to a value close
to the number of geological facies (Fig. 5-30). One
method is to search within the multidimensional histogram
for cells of the locally highest frequency (modes) and then
to construct a dendrogram of similarities between the
modes (Fig. 5-31).
The dendrogram is a graphic representation of the distance, in n-dimensional space, between each mode and
its closest neighbour. Values along the x-axis of Fig. 5-31
correspond to these distances. They give an idea of the
degree Of
between modes and can be used to
justify the grouping of modes into the final electrofacies.
The analyst then chooses a reduced list of modes as seed
216
15 10
Figure 5-37- Example of dendrogram and its use for defining the terminal modes.
Serralog Q 2003
Facies, Sequence and Environment /Chapter 51 217
Principal CornDonent Analvsis
In fact, the clustering methods are applied on the principal component logs, which are derived from a Principal
Component Analysis (PCA).
As we noted previously, each log is influenced, in different degrees, by the geological parameters of the rocks.
In combining several logs we have, automatically, a certain redundancy of geological information. This is highly
useful for mutual verification of the data quality .The goal
of the Principal Component Analysis is to study the correlations between log data, to reduce the number of n variables to a lower number rn, by eliminating the insignificant
components.
The Principal Component Analysis, or PCA, is a statistical study of the logging data over a given interval. From
the search of correlations between logging parameters
(Tables 5-4 and 5 4 , PCA replaces n measured parameters (or n curves or logs as IL, SP, FDC, CNL, GR, etc.)
with n other uncorrelated parameters (or n PC logs). In
fact, this results in a change of coordinate axes (Fig. 532). The readings of n logs at a given depth can be considered as coordinates of a point (corresponding to the
depth) in a n-dimensional space.
the largest variability. The second (or PC 2) is the second
largest but in a perpendicular direction. etc. There is no
correlation between PC 1, ... , PC n, so their use eliminates the redundancy between original logs and permits isolation of the elementary effects. The amount of original
information that each PC axis contains decreases from
PC 1 to PC n (Table 5-6). When the n of them are taken,
the total amount of information contained in the original
set of logs is restored. PC axes of high order contain little
information, which in some cases can be considered
noise. In this case, we can eliminate them in a further
treatment. This corresponds to a filtering and reduces the
number of parameters to be taken into account (it reduces
the dimensionality of the cloud from n to rn, rn being lower
than n).
Table 5-4
Statistical analysis of logging data.
GR
hRI
M E
DT
VAR
F?=E
BAL
SW
ALT
I
60,0135 21,4679 85,8293 24,9209 110,7502
0,3139
0,1726
0,4866
0,3261
0,0739
2,3779
0,456 2,1101
2,5661
0,0931
72.564
74.362 146.926
100.7467 18.3361
1,97711
1,0761 6.19611 0,10981 6,30591
1.503
8.25
3,784
0,25
8,5
9,6154 89,5028
47,0103 14,2632 79,8874
2,9793
1,217
0,2996
8,4376
8,7372
6,662
3,0032
21,918
2,8018 24,7199
2
15
1,6
1
I
0.35
4%
0.35
06
0,35
Figure 5-32 - Principal Component Analysis (PCA) corresponds to a
change of axes.
0,35
Table 5-5
Correlation matrix between wireline logs.
But, if variations observed on a PC of high order correspond to events that appear on one log and not on the
others, these events cannot be considered as noise (i.e.
organic material only detected by the uranium content,
measured by the NGT tool).
Table 5-6
Evaluation of the principal axes and rank by amount of
original information carried.
1
VAR. FRE. BAL, SHA, ALT refer to synthetic logs coming from the SYNDIP program applied to HDT data processed by GEODIP program.
Over a given interval, all measurements or levels (or
sets of log data) define a cloud in n-dimensional space.
This n-dimensional cloud can be described by a set of
axes. PCA defines the principal axes of inertia. If each
data point has the same weight, the first axis of inertia (or
PC 1) will be aligned according to the direction of maximum length. In other words, the first axis is the one having
Serralog 0 2003
AXE
1
_w_ _ _
m
;1I
pelce"tage[
--
cumtatad
.- -___
I
percentage
64,6171
64,6171
22.341I
86.9581
b.l s e rcb
]
I
11 0 , 6 6 8 8 W l I
21 0.2312 MII
31
l
0,5081 W
4,9091
91,8671
-7,457
I
41
51
0,3224WI
0.2992~1
0,1008 D+O
3,1151
2.891I
0,974
94,9821
97.8731
98,847
-0,208
-0.3871
-2,965
0,786
0,294
99,634
99.928
-3,605
3,223
6
7
8
I
0,8140D01
0.3045D01
91 0.7478 DO21
0.0721
1001
-1,003
-2.5521
14.321
217
The distribution of the information carried by a log between the PC logs, can be represented by histograms
(Fig. 5-33). At the same time correlation between logs and
PC 4s provided (Table 5-7). This can help to understand
what type of information is essentially represented by a
given PC log.
Active
inertia = 0.00866
Active
Inertia = 0.00546
PC
PC
As previously mentionned PC logs do not carry additional information. It is just another way to present the
essential information contained in the original logs. This
technique makes clustering somewhat easier, with little or
no loss of information. Hence, after the principal component analysis of the standard logs, a reduced set of principal component logs is chosen. PC logs can be drawn as
logs (Fig. 5-34).
The definition of the axis is based on the cloud distribution. The latter will depend on the weight given to each
data. Hence, if we want to give priority to mineralogy, we
can put more weight on those logs that are more sensitive to composition (NGT, LDT, GST, ACT tools). This situation will amplify the variations of parameters measured by
these tools, and will diminish the weight given to those
logs particularly sensitive to porosity and saturation (FDC,
CNL, LL, IL ... ). This will naturally compress the variations
of parameters obtained from the last mentioned tools.
The results of either of these approaches is an affectation of each log sampling level to one electrofacies.
Figure 5-20 shows the results obtained by the partitioning
technique. The electrofacies is coded arbitrarily as a number from 0 to 20. Evidently, we expect the n’ electrofacies
to reappear whenever the core description indicates a
repeat of the lithofacies type. The maps from electrofacies
to core lithofacies can be made easily by means of the
representation of Figure 5-20. For example, the no7 electrofacies is a clean sand, whereas no 5 is a limestone.
Electrofacies 12 repeats several times. It corresponds to
the shale intervals.
Figure 5-33 - Histograms illustrating the contribution of the principal
components to the reconstruction of the logs.
Table 5-7
Correlations between wireline logs and principal
components.
1450
Figure 5-34 - Example of PC logs.
Once the principal axes of inertia have been calculated
from a set of “active” logs, it is also possible to project
other logs or curves (i.e., “passive” logs) onto these axes.
Thus, one can determine which part of the passive logs
can be explained by the information contained in the set
of active ones and how much the passive logs are related
to each PC log.
218
The correspondence with core data is shown.The
“ladder” presentation for some of the electrofacies (no 5,
6, 7, 8 and 12) also is shown. Aglance at the “ladder” corresponding to facies no 12 shows a wide variation of several logs. Furthermore, the GEODIP results shown on the
right indicate different response types within electrofacies
Serralog 0 2003
Next Page
1
Facies, Sequence and Environment Chapter 51 219
no 12. It can be subdivided by establishing subfacies in
the shale zone.
Figure 5-35 illustrates the result of direct clustering on
the zone values. In this case, the criteria used in the automatic zoning program have been relaxed to give longer
and, hence, fewer zones than in Figure 5-20. The dendrogram shows the zone "families" grouped in terms of their
similarities and distinguished by use of a cutoff level.
Each of these families is assigned an electrofacies number, as shown in the appropriate column. We can note
immediately that this method gives four different shale
types and three shaly sand types. With a more detailed
definition of the core, we probably would see differences
in the composition of each of these shales or shaly sands.
The main layers of marl, sand, and limestone are seen in
Figure 5-20.
- It is not recommended to try to obtain, at any price,
the same number of electrofacies as well recognized in
the core by facies analysis made by a geologist. Even if
the logging tools, and especially image tools, "substitute
the geologist's eyes", it is impossible for them to "see" the
formations as a geologist would see them. The tools do
not react to the same parameters as the eyes and the
brain of geologist do. To define a facies, he will often "filter" the available information, and rely on a feeling that is
based on some of the information : composition, colour,
certain elements of texture, and more often sedimentary
features and faunistic associations. He pays little attention
to the precise proportion of each mineral or element,
because these are studied on chosen samples, and they
may vary appreciably from one point to another. He rarely estimates the porosity and never includes the fluids.
Figure 5-35 - Direct clustering on zone values (from Serra & Abbott, 1980).
An example of electrofacies analysis in carbonates is
given by Figure 5-36. A final selection of 9 terminal electrofacies was made. The superposition of the squared
logs obtained from these 9 terminal modes on the actual
logs (Fig. 5-37) gives a good quality-control indicator of
the selection of terminal electrofacies. The 9 selected
electrofacies allow an adequate description of the main
geological facies over this interval. At the same time, the
representative mean values of the parameters characterizing each final electrofacies are reproduced under the
form of a listing (Table 5-8 next pages).
At this stage it is necessary to make the following
remarks.
Semlog Q 2003
The majority of tools are sensitive to minor variations in
composition, and the latter is not easily detected in a core
except through a very detailed and expensive analysis.
- The purpose of electrofacies analysis is to describe
the formations from the tool responses. If we wish to be
precise and objective it is necessary that the retained
electrofacies reflect, first of all, log parameters. This verification is carried out by superposing the average values,
defined for each electrofacies, on the raw data (Fig. 5-38).
- It is important that the geologist gets to know this new
concept and learns to exploit the results of this new
method of formation analysis. It is necessary that he
benefits from its advantages; objectivity, rapidity, quantifi219
Previous Page
A = carbonaceous shale; B = nodular bedded shale; C = dolomitic mudstone; D = packstone to wackestone with abundant rounded algae;
E = packstone to wackestonerich in foraminifers; F = packstone to wackestonerich in large foraminifers; G = bioclastic mudstone to wackestone;
H = packstone to wackestone rich in bioclasts; I = upper shale, calcareous.
Figure 5-36 - Open-hole logs, synthetic logs, GEODIP and FACIOLOG results and GLOBAL evaluation of a well
(from Schlurnberger Well Evaluation Conference, India, 1983).
cation, accuracy, and finesse of analysis. However, while
maintaining the detail made possible by analysis of the
electrofacies, the geologist can cluster several electrofacies under the same shading, representing one lithofacies.
- As it was previously mentionned he can give more
weight to those log data particularly sensitive to lithological parameters, or to sedimentary features.
220
- If the beds are very thin, compared to the vertical
resolution of the tools, or if the series are fundamentally
composed of thin sequences (i.e. distal turbidites, tempestites), the majority of open-hole logging tools will not
recognise the proper electrofacies of each bed or each
term of the sequence. The computer processing will thus
define an electrofacies corresponding to an average
response that combines the influence of two or more
Serralog CJ2003
Facies, Sequence and Environment IChapter 51 221
beds, and therefore facies. This electrofacies will not give
an accurate idea of the real facies. In such a case, it is
necessary to use tools with a high vertical resolution (dipmeters, EPT, image tools, NMR tools), and a combination
of the results obtained by processing of the data with, for
instance, the LITHO and SYNDIP programs for dipmeter
tools or with BorTex for image tools. This allows conversion of the dipmeter or image resistivity curves in terms of
lithology and structure (Fig. 5-38), by analyzing dipmeter
or image resistivity histograms (Figs. 5-39 and 5-40)and
the nature of the events on the curves.
Translation of electrofacies to facies
An electrofacies is an abstract mathematical concept
that needs to be "translated" into geological words to be
easily understood and used by geologist. Even if the two
approaches are very similar, they differ in language : a set
of words and qualitative adjectives for the description of
the facies by a geologist, a set of log parameters for characterizing an electrofacies by a log analyst. To pass from
one language to the other it is necessary to "bring together the two vocabularies". The most accurate method is
to put alongside each other, at the same scale, the two
Table 5-8
Listing of the mean values of the parameters characterizing each terminal mode or electrofacies.
I
I
I
I
I
I
11
-3.4071 6.8871 33,7851 24.328)
21 -3.7041 1.5881 48.1121 18.5191
31 -2.1621 16.821I 45.7921 27.5261
2.4691 84.1751
2.5421 75.9251
2.461I 87,631I
2.7711
2.3081
1.7281
5.0671 41.6481
3.751 63.2891
3.231 50.6981
3.7591
3.6271
2.7221
5.2281
5.831
6.6781
41
-0.9921 11.2581
2.4911 87.9281
2.4181
4.3931
4.1071
4.9221
5
-1,156
0,933
1,747
2,308 93,075
2,423 101,593
2,429 100,284
1,616
2,627
3,581
4,194
2,338
3,825
4,936
6,802
5,831
5,523
6
7
29,801
7,02
2,649
67.671 27.2811
41,152
70,53
80,276
30,77
37,164
39,906
__
1
-
Figure 5-37 - Terminal modes or electrofacies :averaged squared logs
superposed on actual logs
(from Schlumberger Well Evaluation Conference, India, 1983).
Serralog 0 2003
4,094
54.531
43,311
37,255
5,5 41,979
types of analysis, and to attach to each electrofacies the
descriptive terms used by the geologist to describe facies.
One must attach to the set of log values which characterize an electrofacies (pb, I$,., Pe, At, GR, Th, K ...), the
set of words and adjectives describing the composition,
the texture and the structure of the corresponding bed
(e.g. feldspathic well sorted sandstone with dolomitic
cement, cross-bedded, having a porosity between 18 and
23 %, water saturated).
Sometimes the logs recognize an electrofacies that a
first core analysis has not differentiated. In this case, if it
is not linked to a fluid change, it is better to review and reanalyze the core in a more elaborate manner, and try to
explain the origin of this electrofacies (i.e. presence of
heavy radioactive minerals, of uranium linked to phosphates or organic matter ...).
If the interval was not cored, we will try to translate the
electrofacies into facies using interpretation techniques.
To achieve this objective the lithology must first be defined, and the features, extracted from the dipmeter or
image analysis, must be converted into qualifying adjectives.
The conversion of well log responses to lithology is
obtained by cross-plot analysis or by using an automatic
program based on the construction of a lithofacies data
base which itself is done by combining multi cross-plot
analysis and conversion of the composition of rock lithofacies into log responses. This last program is LITHO or
SQWIZLOG and has been described in Chapter 2.
The conversion of the dipmeter data in textural and
structural information is explained in Chapters 3 and 4
and can be achieved with the SYNDlP or BorTex programs described in the same chapters.
221
SvNTHETK=
LlTHOFAClES
DESCRIPTION
THICKNESSES
STRATK3RAPHK:
COLUMN
Z
SUBDEFlNlTlON
3
=Ill
n
Figure 5-38 - Example of the lithological interpretation of dipmeter resistivity curves by a combination of LlTHO and SYNDIP results.
-
Zoning has been defined
Figure 5-39 - Example of dipmeter resistivity histogram and its conversion in lithology.
222
Figure 5-40 - Example of resistivity histogram of image data.
Serralog 0 2003
Facies, Sequence and Environment‘Chapter 51 223
Result presentation
The presentation of the results of the interpretation by
the FACIOLOG program must include the original logs,
the dipmeter results provided by a processing of the HDT
or SHDT data using GEODIP or LOCDIP programs, and if
possible the lithological column obtained from LITHO or
SQWIZLOG as well as a display of the SYNDIP results.
Hence, we have the means of verifying the quality of the
electrofacies analysis. We may also add the results of a
quantitative interpretation provided by a program such as
ELAN. It is also possible to superpose on the raw logs the
zoning derived from the electrofacies analysis.
The electrosequence concept
We sometimes observe, on certain recordings, progressive evolutions of measured parameters (resistivity,
gamma ray, spontaneous potential, etc.) in relation to
depth (Fig. 5-41). These evolutions, having the shape of
ramps, were named electroseguences by Serra (1970).
The proposed definition of an electrosequence is :
“a depth interval thicker than the vertical resolution of
the measuring tool, presenting a progressive and continuous evolution between two extreme values of measured parameter, tracing a ramp”.
This variation may reflect:
- a progressive change in mineralogical composition
with depth, generally observed on Pe, density and radioactivity measurements: percentage evolution of shale in a
sand or in a limestone; enrichment of a limestone in dolomite, or of a sand in radioactive heavy minerals (Fig. 5-
Figure 5-42 - Example of ramps corresponding to grain size evolution
detected on well logs. Observe, between 210 and 182 ft, the progressive evolution of GR, thorium (Th), uranium (U)and density curves. They
suggest a progressive enrichment of the sand with heavy radioactive
minerals, thorium and uranium-bearing, which can be correlated with a
decrease in grain size, heavy minerals being more abundant in the
silty fraction than in the sandy. Observe the neutron curve (IHCNL)): it
does not show significant porosity variations in the same interval. The
potassium curve (K) is practically zero, indicating the absence of clay
minerals
U
Figure 5-41 - Electrosequences (ramps) clarly seen on well logs.
Serralog Q 2003
223
-the evolution of a textural parameter: grain size change reflecting fining or coarsening upward sequences; sorting decrease, etc.;
- a simultaneous variation in mineralogical composition
and in texture (conglomerate to sand to shale);
- a saturation evolution in the transition zone between
oil- and water-bearing reservoirs that appears particularly
on resistivity curves.
This electrosequence is not necessarily observed on
all curves and, moreover, if it has a small vertical extent,
it will be detectable only by microdevices (dipmeters,
borehole imagery tools, microlog Fig. 5-43).
medium to fine
:graMsadStone
n
wor wave
laminated
i
Imedium
%Ya%
%&
' toe
grad
medium grirned
s;ndslone m a l l y
graded, clean
Figure 5-44 - Thin microsequences or ramps within one macrosequence seen on HDT dipmeter resistivity curves
(from Sera &Abbott, 1980).
ters (hydrogen-index and density for instance) shows a
continuous form (Fig. 5-45), even if each separate curve
(Fig. 5-46) does not clearly show a smooth evolution.
COAL
Figure 5-43 - Example of dipmeter resistivity evolution suggesting thin
fining upward sequences contradicting the evolution suggested by the
gamma ray log (from Serra, in Schlumberger Well Evaluation
Conference, Algeria, 1979).
It corresponds, most often, to a first order sequence
(thin or microscopic according to the definition of Lombard
1956), and sometimes to a second order (medium or
macroscopic), if detailed study of the curves shows very
fine variations in the general trend (Fig. 5-44).
The "bell" and "funnel" shapes of the SP curve correspond to electrosequences (cf. Fig. 5-6), and not to
facies or environment.
It is possible to extend the electrosequence concept to
all depth intervals in which the cross-plot of two parame-
Figure 5-45a - Plot of one progradational sequence showing the crossplot trend (from Rider 8 Laurier, 1979).
The "boomerang" shape_lSometimes observed on density vs neutron cross-plots and well known by log analysts, is another example of an electrosequence of this
type (Fig. 5-47).
Sequence analysis from well logs
It consists of analyzing, on the one hand, the type of
transition from one electrofacies to another (gradational or
abrupt contact), and on the other hand, the arrangement
of electrofacies in vertical sequences, at different scales
(elementary, meso-and mega-sequences).
1
Facies, Sequence and Environment Chapter 51 225
ANALVIIm
correspond to a transition from water-bearing to hydrocarbon-bearing
reservoir, and thus may indicate a progressive change in saturation. A
look at other logs less or not sensitive to fluids (particularly gamma ray)
distinguishes this case from that of a sedimentary sequence.
OP ONE
PIOORAOATtONAL IEOUENCI
ON COC.CNL
caoem
PLOT
FREcuENcTPL0-l
-
Interval 4940 4680 R
Figure 5-456 - Interpretation of Figure 5-45a cross-plot in terms of
facies sequence (from Rider & Laurier, 1979).
Figure 5-47 - The %oomerang”shape observed on neutron vs density
and sonic vs density cross-plots reflects a grain size evolution in a
sand-shale deposit.
LEGEND
Figure 5-46 - Log response of an idealized prograding sedimentary
sequence from shale to sand, shale and coal
(from Rider & Laurier, 7979).
Gradational transition - Elementary seauence
A gradational transition corresponds to a ramp or an
elementary electrosequence, simultaneously detectable
on one, two or several curves : most often gamma ray,
spontaneous potential, resistivity, but sometimes sonic,
density, photoelectric index and neutron.
It may happen that, following the vertical evolution of
thicknesses of each elementary sequence, the macrodevices give a hypothesis for the sequential evolution that
does not agree with that given by the dipmeter- or imageresistivity-curveevolution (Fig. 5-43). In this case, only the
data derived from dipmeter or image are valid.
Rem
ark
It is necessary to remember that a ramp on resistivity curves may
Serrabg 8 2003
When the sedimentary sequence is not very thick (few
centimeters or decimeters), it is not detectable on macrodevices. In this case the electrosequences often appear
more clearly on dipmeter or image resistivity curves (Fig.
5-48 and 5-49).
Remark
Once more, we must be careful in the interpretation of curve evolution in the case of deviated wells. In fact, some sequences are purely
artificial and simply due to the influence of surrounding layers which,
because of the apparent low dip between the axis of the hole and the
beds, are situated behind the bed immediately in front of the measuring
devices (Fig. 5-50).
In a terrigenous detrital series, the elementary electrosequences (cf. Fig. 5-8) indicate evolution both in grain
size and in lithology : normal graded bedding (fining
upward) or reverse graded bedding (coarsening upward).
In a carbonate series, an elementary electrosequence
more often indicates a change in lithology (enrichment in
shale or in dolomite), a diagenetic influence, or, sometimes, a textural evolution (transition from grainstone or
packstone to wackstone or mudstone Fig. 5-51).
225
z
P
w F
DIPS
aP
88
8
n
CORRELATIONS
RESISTIVITY
CURVES
shale
sand
shale
sand
shale
Flning up
wquencs
sand
shale
sand
shale
sand
shale
sand
shale
sand
Figure 5-48 - Example of gradational evolutions or ramps very easily
seen on dipmeter resistivity curves (HOT tool) corresponding to electrosequences. Observe as well the thickness evolution of these electrosequences (between 1 foot and 3 feet). This reflects a thickening
upward sequence of fining upward sequences: With such thicknesses
these sequences cannot be correctly detected by other standard tools.
Figure 5-49 - Example of grain size (fining up) and thickness (thickening and thining up) sequences well seen on FMS images
(from Serra, 1989).
Hole
Axis
A b r u ~ tcontact
It corresponds to an abrupt and significant change of
reading observed at the same depth on one, and more
generally, on several logs simultaneously. Its interpretation needs the study of the relationship between electrofacies. From that analysis one should be able either to
determine sequences of electrofacies, or the significance
of this change : transition from one term of the sequence
to the following one, or break in the sedimentological
sequence, or fault, unconformity ... For instance, if the
abrupt change corresponds to the transition from anhydrite to halite, it only reflects a natural evolution inside the
evaporitic sequence from a less to a more resticted environment.
226
Figure 5-50 - Scheme explaining the different evolutions observed on
dipmeter resistivity curves, related to the apparent low angle between
the axis of the hole and the beds.
Inversely, the abrupt change from a shale to a sand, or
from a sand to a limestone, may indicate a fundamental
Serralog 6 2003
Next Page
Facies, Sequence and Environment Chapter 5 / 227
change in depositional sequence. It marks the beginning
of'a new cycle of sedimentation, after a break or an erosion even if it does not correspond on the dipmeter or
image to a non planar surface (wavy symbol on LOCDIP
display, 4 dip computation on GEODIP display).
Mudstone
But, undoubtedly the best and most accurate method
will be the analysis of their arrangement by probability
methods. Once the electrofacies have been defined, a
study of their vertical organisation can be made following
a method proposed by de Raaf ef a/. (1965). Applying a
procedure suggested by Selley (1970), the succession of
electrofacies has been represented in a diagram resembling a spider's web. An example of what is now referred
to as a "Facies Relationship Diagram" (FRO) is given in
Fig. 5-52. In this diagram, the probability of transition from
one facies to another, as observed in the well, is shown by
numbered arrows.
Observed facies relationship
-Sharp
..__
+Gradational
Grainstone
Figure 5-52 - Example of a Facies Relationship Diagram, showing the
observed number of sharp and gradational transitions between facies
(from Walker, 1979).
SS = scoured surface; A = poorly defined trough cross-bedding;
B = well defined cross-bedding; C = large planar tabular cross-bedding; D = small-scale planar tabular cross-bedding; E = isolated
scours; F = trough cross laminated fine sandstones and shales;
G = low angle stratification.
Boundstone
Figure 5-51 - Example of textural evolution in a limestone. Observe the
general trend on the averaged resistivity curve on the leftside.
Finally, an abrupt change may be linked to a fault or to
an unconformity that brings into contact two rocks having
different properties. The interpretation of the dipmeter and
image data generally allows recognition of these phenomena, and hence determination of the origin of this abrupt
change.
Seauences of electrofacies
The electrofacies, as well as the facies, do not superpose each another randomly, except if there are tectonic
accidents, unconformities, erosion, or periods of non
deposition.
As we have seen previously some cross-plots indicate
directly these sequences of electrofacies. Rider & Laurier
(1979) used this technique to characterize deltaic sequences. They give the log responses for an ideal sedimentary cycle (Fig. 5-46) that consists of a progradation from
shale to sand, with occasional coal deposits. Their
method uses the location of each facies on crossplots
(Fig. 5-45).
Serralog Q 2003
In order to construct a FRD, a computer program tabulates a difference matrix (observed minus random transition probabilities) as shown in Table 5-9. The observed
transition probability is the number of observed transitions
in the well from one electrofacies to another converted to
probabilities. The random transition probability is obtained
on the assumption that all facies transitions are random,
and depends only on the absolute abundance of the
various electrofacies. The difference between the two probabilities gives the difference matrix. It is to be noted that
some values are high-positive (transitions much more
common than if electrofacies were random) and some are
high-negative (transition much less common than random). The last row gives the total number of occurrences
of a particular facies (e.g. facies H is observed only once).
Remark
Walker (1984) mentionned that this method is statistically incorrect.
He suggests using a more complex methods of Markov chain analysis.
The interpretation of the FRD suggests a cycle of
deposition as represented by Fig. 5-53.
Applications of facies and sequence
analysis
Facies and sequence analysis is a fundamental step of
the geological study of a formation. Consequently, it is
normal that its applications cover a certain number of
fields related to sedimentology, and the quantitative inter227
Previous Page
pretation of reservoirs.
Table 5-9
Observed minus random transition probabilities
for well of Fig. 5-36
(from Serra, in Schlumberger Well Evaluation
Conference, India, 1983).
FACES
1
2
3
I
I
1 1 2 1 3 1 4 1
I 18 I 39 I -20 I
1-8
-19 -19
1-1 1-1 I
I 9 I
**L
1-1
-
I
I
5 1 6 1 7
-2 I -18 I -15
-20 -17
85
0 I 3 1-9
I
I
I
I
I
I
I
why we cannot be solely interested in the electrofacies of
sedimentary bodies. In fact, when we analyze the shape
of the SP curve we make this kind of interpretation unconsciously, the "bell" and "funnel" shapes corresponding to
electrosequences as previously noted. But, still we cannot
be conclusive, because such sequential evolution can
belong to several environments. The choice between the
several hypotheses will be done by taking into account
other information related to the thickness of each electrofacies and electrosequence, and its evolution with depth
(Figs. 5-54 and 5-55) and in space through well to well
correlations (Fig. 5-56).
Alluvian fan
(Steel eteL, 1977)
River dominated
delta
(Miall, 1979)
5,o
Wave dominated
delta
(Miall, 1979)
mud
silt
sand
conglomerate
coal
trough
crossbedding
herringbone
crossbedding
Iowan I
crossading
crossbed8ng
hummoc
Figure 5-53 - Spider's web diagram for well of Fig. 5-37 representing
the facies relationship as obtained from data of Table 5-7, and its FRD
interpretation (from Sera, in Schlumberger Well Evaluation
Conference, India, 1983).
.
Reconstitution of the depositional environment
As it was pointed out by Middleton (1978), the final
goal of facies and sequence analysis is the reconstitution
of the depositional environment. But, Walker (1979) stated "many, if not most, facies defined in the field have
ambiguous interpretations - a cross-bedded sandstone
facies, for example, could be formed in a meandering or
braided river, a tidal channel, an offshore area dominated
by alongshore currents, or on an open shelf dominated by
tidal currents". In such cases, "the key to interpretation is
to analyze all of the facies communally, in context. The
sequence in which they occur thus contributes as much
information as the facies themselves". Finally, the
sequence will usually allow recognition of the depositional
environment.
In the same way, we can only reconstruct the depositional environment, from logs, by replacing the electrofacies both in vertical and lateral electrosequences. This is
228
RftS
roots
shell debris
bioturbation
Barrier island
(Davies etel., 1971)
4
4
Prograding storry
dominated shoreline
Submanne fan
(Hamblin 8 Walker, 1979)
(Walker, 1979)
-
Figure 5-54 Lithology, grain size and thickness evolution allow the
determination of the depositional environment (in Miall, 1984).
Serralog @ 2003
Facies, Sequence and Environment IChapter 51 229
Gravel
Sand
Silt
Clay
Meandering
system
Braided
system
Figure 5-55 - Grain size evolution, type of bed boundary, thickness and
frequence of occurrence must be taken into account to determine the
depositional environment (adapted from Sellev, 1976).
An environment is a general term used by geomorphologists or oceanographers to characterize physiographic or morphologic units (mountain ranges, desert, deltas, continental shelves, abyssal plains ... A sedimentary
depositional environment is a geographically restricted
part of the earth's surface, which can be easily distinguisWELL
hed from its adjacent areas by the complex of physical,
chemical and biological conditions, influences or forces
under which a sediment accumulates. This complex largely characterizes the environment and determines the
properties of the sediments deposited within it (Krumbein
& Sloss, 1963; Selley, 1970; Reineck & Singh, 1975; Blatt
et al., 1980).
The physical conditions which act on and control an
environment are numerous. They include
- the climate: the weather and its components
. temperature variations (diurnal, nocturnal, seasonal);
. importance and frequency of rainfalls, snowfalls;
. humidity;
. wind regime (dominant direction, velocity, and their
variations);
. all these factors acting on the vegetal cover;
- the altitude and the topographic profile nature, size,
shape and slope of the mountains or of the receiving
basin, the energy of the flow, the water depth, which will
control the hydraulic regime; in marine environments : the
bathymetry, the amplitude of the tides, the waves, the current system of the water mass, the wind regime, the
Coriolis forces.
The chemical conditions which operate within an environment include the geochemistry of the rocks at the surface, and of the waters (river, lake, sea, ocean) : salinity
(nature and percentage of salts in solution), pH, Eh, gas
in solution.
The biological conditions comprise both fauna and
flora, terrestrial or aquatic, and bacteria present in the
environment.
These three conditions or factors are not independent
but, to the contrary, are strongly linked. Any change in one
of them has immediate repercussions on the others.
Following the medium (air, ice, water) and the relative
importance of each condition, factor and process, an environment can be depositional, erosional or non depositional (equilibrium). As a broad generalization, subaerial
environments are essentially erosional, while subaqueous environments are mostly depositional.
WELL
WELL
Figure 5-56 - Correlations between facies and sequences represented by postulated core and well-log (SPor GR) responses following the
Walther's law (from Van Wagoneret al., 1990).
Serralog 8 2003
229
In the study of ancient deposits, because we analyze
sedimentary rocks, we will always refer to depositional
environments. However, it must be kept in mind that
depositional sequences can be interrupted by periods of
non deposition or even erosion, which of course have no
associated sediments. It will be important to detect and
localize them, both in time and space, because they will
enable a better definition of the environment and the geological history of a sedimentary basin. They can also be
used for correlation purposes.
The determination of the depositional environment will
only be possible through the description of the imprints, or
responses, that the physical, chemical, biological and
geomorphological conditions characterizing the environment left in the deposit.
Those imprints define a facies which is, as previously
seen, the "sum of the physical, chemical and biological
characteristics which differentiate a sedimentary body
from another".
As pointed out by Krumbein & Sloss (1963), "the study
of any sedimentary environment includes considerations
of the four basic environmental elements" listed in Figure
5-57. To each element are attached several factors. "The
relative influence of environmental elements varies with
the nature of the environment. Similarly, the factors in any
element vary in their importance within different environments".
The material factors include the medium (air, fresh or
salty water, ice), the pH, the Eh, the temperature, the dissolved salts (Ca++, Na+, C03-- , SO4--, Cl-), and gases
(CO, O,, SH,), and the solids. "Each of the material factors may have some effect on sediments being deposited,
but the importance of the effect is controlled, in part, by
factors of other elements". For instance in a high energy
environment (zone of breaking waves), the solids (sand
and gravel) are dominant, salts and gases have very little
influence. On the contrary, in a low-energy environment
(quiet water), they become dominant in controlling organisms, and hence play a part in the deposition of carbonate.
"Boundary and energy factors include water depth,
distance from shore, topography of the bottom, and, in a
general way, the geography of the depositional area". The
energy controls the quantity of particles in suspension,
and the size of the settling grains. It also controls the current type (laminar or turbulent) and consequently the
nature of sedimentary features.
The energy and material factors control, in turn, the
biological factors. "In clear warm seas, well oxygenated
and mildly alkaline, organisms may thrive and produce
abundant carbonate sediment" (Krumbein & Sloss, 1963).
On the contrary, turbid medium, or reducing conditions,
may restrict living conditions for some organisms.
The area variations of the elements and factors within
an environment, called the environmental pattern by
Krumbein & Sloss, (1963), induce progressive evolution
of the nature and properties of the sediments creating a
lateral sequence of facies.
But, the accumulation of sediments in an environment
can modify its elements and factors sufficiently to create
new conditions and consequently new properties. "This
sort of interlocked relation between process and response is called feedback". These continuous modifications at
any given geographical point contribute to create a vertical sequence of facies.
Consequently, an environment will be characterized by
a typical sequence of facies both in space and time.
Following the Walther's law' these sequences are similar.
So, the knowledge of the vertical sequence, which can
be obtained from wireline logs, helps to predict the type of
ENVlRONMENTAL FACTORS
RESULTING FEATURES
- Size of the deposits
Geometry
Geometry of the environment
- Properties of the sediments
- composition
- texture
Energy of the environment
- sedimentary structure
- color
- fossils
-I
Lithology FACIE!
Paleocurrents
Paleontology
Matrials of the environment
3iological elements of the environment
- Lateral variations in the
sedimment properties
+
Lateral
sequence
V
Feedback on the environmental factors: prog
Figure 5-57 - A generalized sedimentary environment process-response model (from Krumbein & Sloss, 1963).
230
Serralog Q 2003
Facies, Sequence and E n v i r o n s 231
lateral sequence and the depositional environment.
By considering all the possible combinations of the
physical, chemical and biological processes or all elements and factors which characterize an environment,
one could guess that there should be an infinite number
of environments. In fact, due to the strong interdependence of all these processes and factors, a finite number of
environments has been recognized.This is also related to
the fact that there are a finite number of physiographic
types. It is for this reason that, even if two environments
or morphologic units are never totally identical, the number of major environments is reduced if one considers the
dominant factors. They are listed in Table 5-1 0. Of course,
especially when we study modern deposits, several subor even sub-sub-environments can be added to take into
account small variations. But, for the ancient sediments
the separation between them is not an easy task. It is for
that reason that we will limit the illustration of their recognition by wireline logs to the major environments.
Table 5-10
A classification of depositional sedimentary environment
(from Selley, 1970).
Fanglomerate
Fluviatile
Continental
Lacustrine
Eolian
Lobate (deltaic)
Shoreline
Linear (barrier)
Marine
I
Braided
Meandering
Terrigenous
Mixed carbonat
terrigenous
Shelf
Turbidite Carbonate
This table tabulates only those environments which have generated large volumes
of ancient sediments.
Interest of the reconstitution of the depositional
environment
The reconstitution of the depositional environment is
not a pure academic objective. It is a necessity in the
search for mineral natural resources such as hydrocarbons, coal, phosphates, potassium salts, uranium...
because these resources are, most commonly, strongly
associated to specific depositional environments.
In petroleum exploration, one of the essential objectives is the evaluation of the hydrocarbon potential of a
basin. This requests the determination of the quality, the
thickness and lateral extension (the volume) of the different facies which represent source rocks in which will be
generated hydrocarbons, reservoir rocks in which hydrocarbons will accumulate by migration from the previous
rocks, and cap rocks which will constitute impermeable
Serralog CJ2003
traps avoiding any dismigration of the hydrocarbons from
reservoirs.
In reservoir study, only the reconstitution of the depositional environment gives a correct idea of the lateral
evolution of the facies and consequently of the petrophysical properties of the reservoirs, and to enable prediction
of the existence, nature, importance and distribution of
permeability barriers. This information is of the utmost
importance to evaluate the production potential of a field
and to locate extraction and injection wells for a better oil
recovery.
Recoanition of the maior depositional environments
from loaaina data
In the study of ancient deposits, the traditional
approach of geologists consists of analyzing rock samples by defining their facies through the mineralogical,
textural and structural characteristics of the rocks.
The knowledge of the lithology and the mineralogy is
important, but it is necessary to control if diagenesis has
modified the original rock type.
The grain size and the textural maturity can help to
recognize the depositional environment as illustrated by
Figure 3-7 of Chapter 3.
As previously seen, sedimentary features constitute a
fundamental information for the recognition of the facies
and the environment.
But, the facies knowledge is not sufficient to identify an
environment. As pointed out by Walker (1979), “a crossbedded sandstone facies, for example, could be formed in
a meandering or braided river, a tidal channel, an offshore area dominated by alongshore currents, or on an open
shelf dominated by tidal currents.” Additional information
related to the thickness of each facies, or sequence of
facies, and to its evolution with time, and consequently
with depth, will often allow discrimination between two or
three possible environments (Figs. 5-54 and 5-55).
In a multi-well study the facies and the environments
will be more accurately recognized using correlation techniques (Fig. 5-12), and facies mapping (isoliths or isopachs, Fig. 5-13), the geometry of sedimentary bodies
being an important parameter for facies and environment
recognition. For example, the meandering channel fill system is easily recognized from the isopach map of the
lower Muddy in the South Glenrock Oilfield, Wyoming
(Fig. 5-13 from Curry & Curry, 1972).
In fact, the depositional environment will be better and
more accurately defined by integrating all the information
on facies and sequences of facies. “The key to interpretation is to analyze all of the facies communall~in context,.
The sequence in which they occur thus contributes as
much information as the facies themselves” (Walker,
1984).
These sequences can be described, as suggested by
Visher (1965), by a vertical profile (Fig. 5-58), or, in other
terms introduced by Walker (1976), by a facies model. As
illustrated by Figure 5-59 from Walker, the facies model is
obtained by “distilling and concentrating the important
231
~
features that' several similar environments "have in common".
As Walker explained "a facies model could be defined
as a general summary of a specific sedimentary environment, written in terms that make the summary useable in
at least four different ways".
For Walker a facies model must "act as
- a norm, for purpose of comparison;
- a framework and guide for future observations;
- a predictor in new geological situations;
- a basis for hydrodynamic interpretation of the environment or system that it represents".
In a subsurface study involving well logs, the equivalent of the facies model will be the concept of an electrofacies model. As seen in previous chapters, wireline logs
enable determination of all the components of facies
(mineralogy, texture, sedimentary features, geometry).
One can also define the organization of facies in sequences. Many authors (Fisher, Visher, Pirson, Coleman,
Galloway ...) have used the SP-resistivity curve shapes to
determine facies and environments. An illustration is
given by the small sketches of Figures 5-60 and 5-61 from
Fisher (1969) which quickly determine the different facies
and sub-environments of two delta systems.
But, as previously mentioned, SP and resistivity curves
are not always sufficient to correctly determine the facies.
The other logs, especially the dipmeter or images, bring
information which help to be more accurate and precise.
It is the reason why I suggest they be introduced, when
available, in any sedimentological study.
Local examples
A
B
C
D
E
distilled away
d
differ from norm
I1
A
@-
Model as framework
\'
@ Model as norm
@
54
, I
'y
Pure essence of
environmentalsummary
7G d f f
\
Model as basis for
hydrodynamic interpretation
Faciesmodel
\
-=-
0
Model as
c
.... ... a
pydictor
\L
[Evidence in local area X +
z:zE
= predictions in area
XI
~~~
Figure 5-59 - Distillation of a general facies model from various local
examples, and its use as a norm, framework for observation. predictor, and basis for interpretation of new example
(adapted from Walker, 1984).
SP
Resistlvity
SP ResbUvitySP Resistivity
SP
Resistivity
Figure 5-58 - Determination of facies from the grain-size evolution, the
sedimentary features and the geometry. Case of fluvial deposits
(from Visher, 1965).
Illustration of depositional environments from l o w
data
In the following pages, the main geological
ristics of several major environments, on a single vertical
profile, are described, through the geological faciesmodel concept.
232
Figure 5-60 - Examples of interpretation of vertical SP and resistivity
profiles in high-constructive lobate delta systems, Gulf Coast Basin
(from Fisher, 1969).
Serralog Q 2003
Facies, Sequence and Environment )Chapter 51 233
SP
SP
Resistivity
Resistivity
SP
Resistivity
SP
Resistivity
SP
Resistivity
SP
Resistivity SP Reslsthrity
A - Fluvial channel fades
-
E Chand -diannel mouthbar
facies (progradational)
stacked ~ 8 sbader
~ l
hies
D - Sbandplain-shelf
fades
E - Sbandpiainlagoonmarsh facies
F Lagoon-marshf!ad basin facies
~
G - shelf facies
Figure 5-61 - Example of interpretation of SP and resistivity profiles in high-destructive, wave dominated delta system, Delta Coast Basin
(from Fisher, 1969).
The geological parameters are then "translated" into
well-log responses, both openhole logs and dipmeter or
image data, the validity of which is in turn controlled and
illustrated by examples. In each case the most important
logging features which allow their recognition are indicated. As suggested by Walker (1976), such an electrofacies model will act as a norm to which any actual example will be compared, and as a guide for future observations and enrichment of the model.
Gamma ray
Calipar
(API)
Figures 5-62 and 5-63 illustrate other examples of log
interpretation in deltaic environment.
The type of environment corresponding to these
examples is illustrated by the sketch of Figure 5-64.
Neutron (%)
Dipmeter arrow plot
Figure 5-62 - Other example of interpretation of log profiles in terms of facies (from Serra & Sulpice, 1975).
Serralog Q 2003
233
DEPTH
Track 1
Track 2
Track 3
Track 4
fed
Figure 5-63 - Example of log response in a deltaic environment. The bottom sand corresponds to a mouth bar capped by a transgressive level
(resistive and compact level rich in shell debris). The second sand corresponds to a tidal channel deposit; the third sand to a beach or barrier bar
deposit; the top sand to a tidal channel deposit. The resistive peaks in the third sand corresponds to layers rich in shell debris.
Reconstruction of the geometry of the facies
After an electrofacies and electrosequence analysis
for each well of a field, or possibly of a licence or even a
basin, we can, through the correlations of electrofacies,
234
jointly with chronostratigraphic correlations between
wells, reconstitute the space-time distribution of the different electrofacies. The application of mapping techniques
will define the geometry of each facies or group of facies
(see Fig. 5-10 and 5-13).
Serralog Q 2003
Facies. Seauence and Environment ,ChaDter 51 235
Figure 5-64 - Sketch of environment corresponding to Figs. 5-62 and
5-63 (based on Weber 8, Daukoru, 1975).
Figure 5-66 - lsopach map of total sands in the Bisti Oilfield, New
Mexico (from Sabins, 1972).
The correlations between wells (Fig. 5-65) and the isopach map of total sands (Fig. 5-66) in the Bisti Oilfield
show clearly the barrier bar nature of the reservoir (from
Sabins, 1972).
application by the realization of isopach or percentage
maps of one typical electrofacies or group of electrofacies. One can also generate maps of sand-shale ratios
computed from FACIOLOG, LITHO or SYNDIP. For
Figure 5-65 - Log correlations between wells in the Bisti Oilfield, New Mexico (from Sabins, 1972).
Mapping of electrofacies
As seen previously, SP and resistivity curves can be
used for mapping purposes, to define the spatial repartition of typical facies. This helps to better define the environment (see Fig. 5-1 0 and 5-67). We can generalise this
Serralog Q 2003
instance, the ratio of the cumulative thickness of resistivity peaks to the cumulative thickness of conductive
troughs computed in a given interval, leads to an accurate sand-shale ratio map if the peaks correspond to sand
beds and the troughs to shale beds. These maps describe objectively, precisely and quantitatively, the lateral evolution of the facies.
235
Choice of a more judicious core sampling for
analysis
I
Figure 5-67
- Reconstruction of the depositional environment through interpretation
of the SP shapes.
This aspect is often neglected. In bringing together the
cores and the results of electrofacies analysis, or at least
the open-hole logs and more precisely the dipmeters and
borehole images, we are able to select in a more reasonable manner the sampling intervals and rates for detailed laboratory analysis. This way to sample allows both a
saving on expense and time, and a clarification of certain
log responses not clearly understood and interpreted.
Thus, if we refer to the example of Fig. 5-48, if the
sampling and the analysis were concentrated on one, two
or even three electrosequences with a sampling interval
of 5 cm, we should be able to calibrate the dipmeter resistivity curves (Fig. 5-68) in terms of grain size, shaliness
and permeability.
Constitution of a data base of electrofacies
By combining the results of electrofacies analysis processed on several wells of a field or a basin, we can constitute an electrofacies data base which can be used as
reference for the study of any new wells in the field or the
basin. Each set of log data will be cross checked with this
data base for attribution. If not attributed a new electrofacies must be introduced into the data base for its completeness.
This application is very important. It will later enable
definition of the depositional environment directly from
well logs using programs involving "expert systems".
Quantitative Interpretation
The electrofacies analysis leading to a reconstitution
of facies and depositional environment, we are able to
attribute to.each electrofacies, or group of electrofacies,
the most probable mineralogical model for quantitative
interpretation (i.e. quartz, potassium feldspar, plagioclase,
kaolinite as principal minerals entering into the composition of a sand). This limits the number of unknowns for
each electrofacies and optimises the interpretation by a
GLOBAL or ELAN type program as suggested by Suau et
a/. (1982).
Moreover, the texture and the sedimentary structure
indicated by dipmeters or borehole images enable definition of the types of distribution of shales and, consequently, a better choice of the response equations (for the
computation of shaliness and saturation), and the parameters a, m and n that relate porosity to the formation factor and saturation (textural and structural models).
One can also better specify the constraints and determine for each electrofacies the functions that permit an
approach to permeability. Finally, one can reduce the
computation time and thus carry out a higher number of
tests by working with local modes or electrofacies instead
of level by level.
236
Figure 5-68 - Calibration of the dipmeter resistivity curve in terms of
grain size, shaliness and permeability.
We should also be able to precise the facies and the
depositional environment (turbidite). In this way, all uncored intervals, and even all the wells of the same field,
representing the same characteristics, would be easily,
rapidly and economically interpreted.
References and Bibliography
BATES, R.L. & JACKSON, J.A. (1980). - Glossary of
Geology. Amer. Geol. Institute, Falls Chuch, Virginia.
BLATT, H., MIDDLETON, G., & MURRAY, R. (1972,
1980). - Origin of Sedimentary Rocks. 1st and 2nd ed.
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
BOUMA, A.H., & BROUWER, A. (eds) (1964). Turbidites. Elsevier, Amsterdam.
BUSCH, D.A. (1971). - Genetic Units in delta prospecting. Bull. Am. Ass. Petrol. Geol., 55, p. 1132-1154.
COLEMAN, J.M., & PRYOR, D.B. (1980). - Deltaic
sand bodies. Am. Ass. Petrol. Geol., Contin. Ed. Course
Note Ser. 15.
CURRY, W.H., & CURRY, W.H. Ill (1972). - South
Glenrock Oilfield, Wyoming : Prediscovery, Thinking and
Post-Discovery Description. In Stratigraphic Oil and Gas
Serralog Q 2003
Facies, Sequence and Environment /Chapter 51 237
-
Fields Classification, Exploration Methods, and Case
Histories. Arner. Assoc. Petroleum Geol., Memoir 16, p.
415-427.
DAVIES, D.K., ETHRIDGE, F.G. &BERG, R.R. (1971).
- Recognition of barrier environments. Bull. Amer. Assoc.
Petroleum Geol., 55, p. 3-26.
DELFINER, P., PEYRET, O., & SERRA, 0. (1984). Automatic determination of Lithology from Well Logs. 59th
Ann. Techn. Conf. SPE of AIME, Houston, Texas; paper no
SPE 13290.
DELHOMME, J.P., & SERRA, 0. (1984). - Dipmeterderived Logs for SedimentologicalAnalysis. SPWLA, 9th
Europ. Intern. Format. Eval. Trans., paper 50.
FISCHER, A.G. (1964). - The Lofer cyclothems of the
Alpine Triassic. In: Symposium on Cyclic Sedimentation.
(ed. by MERRIAM, D.F), p. 107-149. Bull. geol. Surv.
Kansas, 169.
FISHER, W.L. (1969). - Facies characterization of Gulf
Coast basin delta systems, with some Holocene analogues. Gulf Coast Assoc. Geol. SOC.Trans., 19, p. 239261.
FISHER, W.L., & McGOWEN, J.H. (1969). Depositional systems in Wilcox Group (Eocene) of Texas
and their Relation to Occurrence of Oil and Gas. Bull.
Amer. Assoc. Petroleum Geol., 53, p. 30-54.
FISHER, W.L., & BROWN, L.F. Jr. (1972). - Clastic
depositional systems -A genetic approach to facies analysis. University of Texas, Austin, Texas, Bureau of Econ.
Geol.
FISHER, W.L., BROWN, L.F., SCOTT, A.J. &
McGOWEN, J.H. (1969). - Delta systems in the exploration for oil and gas. - A research colloquium. University of
Texas, Austin, Texas Bureau of Econ. Geol.
FONS, L. Sr. (1969). - Geological application of well
logs. SPWLA, 10th Ann. Log. Symp. Trans.
FRIEDMAN, G.M., & SANDERS, J.E. (1978). Principles of Sedimentology. John Wiley & Sons, New
York.
GALLOWAY, W.E., & HOBDAY, D.K. (1983). Terrigenous Clastic Depositional Systems. Springer, New
York.
GILREATH, J.A., & STEPHENS, R.W. (1971). Distributary Front Deposits lnterpreted from Dipmeter
Patterns. Gulf Coast Assoc. Geol. SOC.Trans., 21, p. 233243.
GILREATH, J.A., & STEPHENS, R.W. (1975). lnterpretation of Log Responses in a Deltaic Environment.
Arner. Assoc. Petroleum Geol., Marine Geology
Workshop, Dallas, Texas.
GOETZ, J.I., PRINS, W.J., & LOGAR, J.F. (1977). Reservoir Delineation by Wireline Techniques. Paper presented at 6th Ann. Conv. Indonesia Petroleum Assoc.,
Jakarta, May 1977.
GRESSLY, A (1838). - Observations geologiques sur le
Jura Soleurois. Neue Denkschr. allg. Schweiz. Ges. ges.
Naturw., 2, p. 1-112.
GRIFFITHS, C.M. (1990).- The language of rocks: an
example of the use of suntactic analysis in the interpretation of sedimentary environments from wireline logs. In:
Serralog 0 2003
HURST, A., LOVELL, M.A., & MORTON, A.C. (eds),
Geological Applications of Wireline Logs, Geological
Society, Special Publication no 48, pp. 77-94.
HAUG, E. (1907). - Traite de geologie. Armand Colin,
Paris.
KRUMBEIN, W.C. & SLOSS, L.L. (1963). Stratigraphy and Sedimentation. 2nd ed. W.H. Freeman &
Co., San Francisco.
LENNON, R.B. (1976). - Geological factors in steamsoak projects on the west side of the San Joaquin Basin.
J. Petroleum Technol., July, p. 741-748.
LOMBARD, A. (1956). - Geologie Sedimentaire. Les
series marines. Masson, Paris.
LOMBARD, A. (1972). - Series sedimentaires. Genese
- Evolution. Masson, Paris.
LYELL, C. (1933). - Principles of Geology. John
Murray Ltd., London.
MIALL, A.D. (1977). - A review of the braided-river
depositional environment. Earth Sci. Rev., 13, p.1-62.
MIALL, A.D. (ed.) (1978). - Fluvial sedimentology.
Canad. SOC.Petrol. Geol., Mem. 5.
MIALL, A.D. (1979). - Deltas. In : Facies Models.
(Walker, R.G. ed.). Geoscience Canada, Reprint series 1.
MIALL, A.D. (1984). - Principles of Sedimentary Basin
Analysis. Springer, New York.
MIDDLETON, G.V. (1973). - Johannes Walther's Law
of the correlation of facies. Geol. SOC.America Bull., 84,
p. 979-988.
MIDDLETON, G.V. (1978). - Facies. In : FAIRBRIDGE,
R.W, & BOURGEOIS, J., eds. Encyclopedia of sedimentology. Stroudsburg, Pa., Dowden, Hutchinson & Ross, p.
323-325.
MOORE, R.C. (1949). - Meaning of facies. Geol. SOC.
Amer., Mem. 39, p. 1-34.
NORWOOD, E.M. Jr., & HOLLAND, D.S. (1974). Lithofacies mapping. A descriptive tool for ancient delta
systems of the Louisiana outer continental shelf. Trans.
Gulf Coast Assoc. Geol. SOC.,24.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Houston.
RAAF, J.F.M. de (1968). - Turbidites et associations
sedimentaires apparentees. I and /I. Kon. Ned. Akad.
Wetensch., Amsterdam, Verh., series B, 71, p. 1-23.
RAAF, J.F.M. de, READING, H.G., & WALKER, R.G.
(1965). - Cyclic sedimentation in the Lower Westphalian
of north Devon, England. Sedimentology, 4, p. 1-52.
READING, H.G. (ed.) (1978). - Sedimentary
Environments and Facies. Blackwell Scientific
Publications, Oxford.
RIDER, M.H. (1990).- Gamma-ray log shape used as
a facies indicator: critical analysis of an oversimplified
methodology. In: HURST, A., LOVELL, M.A., & MORTON,
A.C. (eds), Geological Applications of Wireline Logs,
Geological Society, Special Publication no 48, pp. 27-37.
RIDER, M.H., & LAURIER, D. (1979). - Sedimentology
using a computer treatment of well logs. SPWLA, 6th
Europ. Syrnp. Trans., paper J.
SABINS, F.F. (1972). - Comparison of Bisti and
Herseshoe Canyon stratigraphic traps, SanJuan Basin,
237
New Mexiso. In stratigraphic oil and gas field Classification, Exploration methods, and case histories.
Amer. Assoc. Petroleum Geol., Memoir 16, p. 610-622.
Schlumberger (1979). - Well Evaluation Conference.
Algeria.
Schlumberger Technical Services, Inc. (1983). - Well
Evaluation Conference. India.
Schlumberger Middle East S.A. (1984). - Well
Evaluation Conference. Egypt.
SELLEY, R.C. (1970). - Studies of sequence in sediments using a simple mathematical device. J. geol. SOC.
London, Quart. Jour., 125, p. 557-581.
SELLEY, R.C. (1970, 1978). - Ancient Sedimentary
Environments. 1st and 2nd ed. Chapman & Hall, London.
SELLEY, R.C. (1976). - An lntroduction to
Sedimentology. Academic Press, London.
SELLEY, R.C. (1978). - Concepts and methods of
subsurface facies analysis. Amer. Assoc. Petroleum
Geol., Continuing Education Course Note Series 9.
SELLEY, R.C. (1979). - Dipmeter and log motifs in
North Sea submarine-fan sands. Bull. Amer. Assoc.
Petroleum Geol., 63, 6, p. 905-917.
SERRA, 0. (1972). - Diagraphies & Stratigraphie. In:
Mem. B. R. G. M., 77, p. 775-832.
SERRA, 0. (1973). - Interpretation Geologique des
diagraphies en Series Carbonatees. Bull. Centre Rech.
Pau - SNPA, 7 , 1, p. 265-284.
SERRA, 0. (1974). - Interpretation geologique des
Series deltai'ques a partir des diagraphies differees. Rev.
AFTP, 227, Oct., p. 9-17.
SERRA, 0. (1977). - Methode rapide d'analyse faciologique par diagraphies differees. SPWLA, 5th Europ.
Symp. Trans., Paris, paper 9.
SERRA, 0. (1984). - Fundamentals of Well-Log
Interpretation. Volume 1 : The Acquisition of Logging
Data. Developments in Petroleum Science, 15A, 440 p.,
Elsevier, Amsterdam.
SERRA, 0. (1986). - fundamentals of Well-Log
Interpretation. Volume 2 : The lnterpretation of Logging
Data. Developments in Petroleum Science, 15B, 684 p.,
Elsevier, Amsterdam.
SERRA, O., & ABBOTT, H. (1980). - The Contribution
of Logging data to Sedimentology and Stratigraphy. 55th
Ann. Fall Techn. conf. SPE ofAIME, paper SPE 9270, and
in SPE J., Feb. 1982.
SERRA, O., MOTET, D., & VASQUEZ, J. (1993). Detailed Facies, Genetic Sequences and Depositional
Environment Analysis from Well Logs and Electrical
Images. A.A.P.G. Annual Convention, Poster session,
New Orleans, (Abstracts).
SERRA, O., & SULPICE, L. (1975). - Sedimentological
Analysis of shale-sand series from well logs. SPWLA, 16th
Ann. Log. Symp. Trans., paper W
SERRA, O., & SULPICE, L. (1975). - Apports des diagraphies differees aux etudes sedimentologiques des
series argilo-sableuses traversees en sondage. 9th Cong.
Intern. Sediment, Nice, theme 3, p. 86-95.
SHROCK, R.R. (1948). - Sequence in Layered Rocks.
McGraw-Hill Book Co., Inc., New York.
238
STEEL, R.J., MAEHLE, S., NILSEN, H, ROE, S.L., &
SPINNANGR, A. (1977). - Coarsening upward cycles in
the alluvium of Hornelen Basin (Devonian), Norway.
Sedimentary response to tectonic events. Bull. Amer.
Assoc. Petroleum Geol., 88, p. 1124-1134.
TEICHERT, C. (1958). - Concepts of facies. Bull. arner.
Assoc. Petroleum Geol., 42, p. 2718-2744.
VINCENT, P., GARTNER, J. & ATTALI, G. (1979). GEODIP - An approach to detailed dip determination
using correlation by pattern recognition. J. Petroleum
Technol., Feb. 1979, p. 232-240.
VISHER, G.S. (1965). - Use of vertical profile in environmental reconstruction. Bull. Amer. Assoc. Petroleum
Geol., 49, p. 41-61.
WALKER, R.G. (1975). - Generalized facies models
for resedimented conglomerates of turbidite association.
Bull. geol. SOC.Amer., 86, p. 737-748.
WALKER, R.G. (1976). - Facies Models. Turbidites
and associated coarse clastic deposits. Geoscience
Canada, 3, p. 25-36.
WALKER, R.G. (Ed.) (1979, 1984). - Facies Models.
1st and 2nd ed. Geoscience Canada, reprint series 1,
published by Geo% Assoc. Canada.
WALKER, R.G. & MUTTI, E. (1973). - Turbidite facies
and facies association. SEPM, Pacific Section, Short
Course, Anaheim.
WALTHER, J. (1893-4). - Einleitung in die Geologie als
historische Wisssenschaft. Fischer, Jena.
WEBER, K.J., & DAUKORU, E. (1975). - Petroleum
geology of the Niger delta. Trans. 9th World Petrol.
Congres, Tokyo, 2, p. 209-221.
WIDDICOMBE, R.E. & NOON, P. (1984). - Mulfiwell
FAClOLOG evaluation, Hartzog Draw field,Powder River
Basin, Wyoming. SPWLA, 25th Ann. Log. Symp. Trans.,
New Orleans.
WILLIAMS, P.F. & RUST, B.R. (1969). - The sedimentology of a braided river. J. sediment. Petrol., 39, 2, p.
649-679.
YAPAUDJIAN, L. (1972). - Une approche actualiste en
geologie sedimentaire (quelques donnees d'interpretation
des sequences de plateforme). In: Mem. BRGM, 77,
p.715-744.
Serralog Q 2003
WELL LOGGING AND DIAGENESIS
Introduction
Review of geological concepts
Definition
Diagenesis is defined as “all the chemical, physical,
and biologic changes undergone by a sediment after its
initial deposition, and during and after its lithification,
exclusive of surficial alteration (weathering) and metamorphism... It embraces those processes (such as compaction, cementation, reworking, authigenesis, replacement, crystallization, leaching, hydration, bacterial action,
and formation of concretions) that occur under conditions
of pressure (up to 7 kb) and temperature (maximum range
of 700°Cto 300°C) that are normal to the surficial or outer
part of the Earth’s crust; and it may include changes
occurring after lithification under the same conditions of
temperature and pressure.” (Bates & Jackson, 1980).
Several studies have been made of this phenomenon,
and the associated publications are listed in the bibliography. We now extract the some key points from these
publications.
‘Asdeposition continues, the lamina of sediment passes from the interface to successively lower positions and
enters a realm of greatur pressure, higher temperature,
and of changed chemical and biological conditions. These
new conditions promote the consolidation or lithification of
the sediment into a sedimentary rock” (ibid).
The diagenetic environment is characterized by the
original composition of the sediment and the original fluid
on which will act the temperature and pressure, therefore
the burial depth, and the possible circulation of fluids by
hydrodynamic or hydrothermic phenomena. Non-marine
environments are different from their marine analogues
due to the differences in fluids.
The original fluids occupying the pore space of the
rock at the moment of deposition are characterized by
their nature (water, air or gas), and possibly by their
content of salt and dissolved gases. The composition of
the fluids will be modified very rapidly by burial.
Interactions with the minerals and the action of bacteria
lead to a change in pH (hydrogen ion concentartion) and
in Eh (oxidation reduction potential) (Figs. 6-1 and 6-2).
The Diagenetic Environment
A “diagenetic environment is the environment of postdepositional change. It extends an indefinite distance
downward from the depositional interface” (Krumbein &
Sloss, 1963). In other words it may be defined as a certain volume of rock in which a series of transformations
takes place. “The nature of the depositional environment
and the rapidity of the postdepositional changes depend
upon the medium of deposition and the kind of sediment
being deposited‘ (ibid).
“The depositional interface represents an important
boundary condition that separates two different physiochemical regions. As a simple illustration, clay and silt
settle through sea water as a group of particles in aliquid
medium. When the Aparticles come to rest on the bottom,
they form a solid matrix with water-saturated pores. The
water has the same composition as the medium above,
but marked changes occur once it is sealed from free circulation by confinement in the pores” (ibid).
Serralog 0 2003
Figure 6-1- Diagram of changes in characteristics of marine bottom
mud below the depositional interface. Eh is in units of 0. I
(redrawn from data from Zobell; 1949,in Krumbein & Sloss, 1963).
239
Zone of intense dissolqtfi
Zone of Ininor dissolution
Zone of dolomitizatiau
Eh
Fresh water phreatic zone
PH
Zone of precipitation
Formation of meni=uS
and pendant calcite
cements
Formation of caliche crust
Intense solution near soil zone
Minor solution
Figure 6-2 - Diagram showing the limits on values of pH and Eh found
in natural environments, in particular in syndiagenetic, anadiagenetic
and epidiagenetic zones (based on Becking & Carrels)
(from Fairbridge, 1967).
Thereafter the diagenetic environment changes with
depth, but not necessarily in a uniform way, because
again grain size and composition, or the formation of new
(authigenetic) minerals, may create a permeability barrier,
preventing any further circulation of fluid. This is why the
diagenetic environment depends mainly on the original
facies, and consequently on the initial environment (Fig.
6-3).
a Zone
of active water
circulation and
significant precipitation
of rnanne cements
0 phdatic
Sta nant marine
zone with
Miqtization
and intr r a n k ?
little or no cementation cementzon
Factors Affecting Diagenesis
According to Wolf & Chilingarian (1976), there are
twelve factors which control the diagenetic process.
- geographic (climate --> humidity --> rainfall --> type
of terrestrial weathering --> surface water chemistry);
- geotectonism (rate of erosion and accumulation,
coastal morphology, emergence and subsidence, whether
eugeosynclinal or miogeosynclinal);
- geomorphologic position (basinal versus lagoonal
sediments, current velocity -. grain size, sorting, flushing
of sediments);
- geochemical factors in a regional sense (supersaline
versus marine water, volcanic fluids and gases);
- rate of sediment accumulation (halmyrolysis --> ion
transfer --> preservation of organic matter --> biochemical
zonation);
- initial composition of the sediments (aragonite versus
high-Mg and low-Mg calcite and isotopes and trace-elements content);
- grain size (content of organic matter --> number of
240
Zone of solution
near water table
0 Stagnant zone
8 Zone of active water
grain neomorphism
CtrculatiOn. Rapid
to calcite put little
neomorphism and
cementation by
cementation
equant calcite
Figure 6-3 - Schematic cross-sections showing diagenetic textures in
carbonates which have been developed in diagenetic environments
close to the surface (from Longman, 1980).
bacteria --> rates of diffusion);
- purity of the sediment (percentage of clay and organic matter --> basic exchange of clays altering interstitial
fluids);
- accessibility of rock framework to surface (cavity systems permit replacements);
- interstitial fluids and gases (composition, rate of flow,
and exchange of ions);
Serralog 0 2003
Diagenesis
- physico-chemical conditions (pH, Eh), partial pressures and C02 content);
- previous diagenetic history of the sediment (previous
expulsion of trace elements will determine subsequent
diagenesis).
These factors influence the local environments which,
in turn, influence the micro-environments. The factors
interact and overlap : the climate influences the geomorphology which in turn controls the grain size and hence
the type of bacteria, rate of diagenesis, pH and oxidation
reduction potentials, and finally the type of replacement
which takes place.
1 Chapter6 I
241
- Biochemical and Oraanic Processes
All organisms, and particularly bacteria, can modify the
pH and the oxidationheduction potential, Eh. The modification is particularly rapid within a few centimeters or tens
of centimeters of the surface (Zobell 1942), and produces
major changes within the sediments, such as reduction of
sulphates and conversion of organic matter (cf. Fig. 6-2).
In addition, there may be a reduction of grain size, mixing
of grains (bioturbation), corrosion of grains and creation of
gas bubbles.
Diagenetic Changes
The Fundamental Processes
There are three types of process involved.
- Phvsical Processes
The mechanical constraints of burial initially result in a
rearrangement of the grains of the sediment which tends
to produce compaction, and at a later stage may cause
fracturing.
Dessication will result in contraction, fissuring, penecontemporaneous internal deformation and internal
mechanical sedimentation.
Pressure and temperature will both increase with
depth and will modify the chemical equilibrium. They will
also cause rocks which are normally elastic to become
plastic or even viscous.
The combined effects of temperature and pressure will
also affect the solubility of the minerals, which may result
in either increased dissolution or increased precipitation
of salts in the pore space.
- Chemical Processes
The interactions between the solids and the fluids will
tend to establish a chemical equilibrium. This may result
in dissolution, alteration, oxidation or reduction. It may
also cause the growth of crystals (Fig. 6-4), or their replacement with new types.
The main factors controlling diagenesis are mineralogical composition, grain size, fluid content and organic
matter on the one hand, and temperature, pressure and
ambient chemical conditions on the other. As a result,
there are many different types of diagenetic changes.
Krumbein (1942) has listed some thirty of these. They
may be classified as summarized in Figure 6-5.
Lithification =d
-shale
includes
Sand
Sandstone
individual ~ ~ ~ , __+
d , , Conglomerate
limertone and
,
Processes sands, oozes
below
dolostone
Compaction
primarily
of muds
50-60% ofwater
Finely detailed
Structures
Recrystallization
of unstable
{I
minerals
ta
Dissolution of
more insoluble
minerals
Precipitation
of new minerals
or additions to
existing ones
New cryst@faces of
quartz preciptated on
rounded surface.,
New lcaolinite crystals
grown in pore space
1
_I
Figure 6-4 - On the left, SEM photograph showing quartz overgrowth.
On the right, thin section illustrating the quartz overgrowth.
Serralog 0 2003
Figure 6-5 - Changes in composition and texture resulting from diagenetic processes - Most of the changes tend to transform a soft and friable sediment into a hard, consolidated sedimentary rock
(from Press 8, Sieve6 1978).
241
- Rock components
According to Krumbein (1942), the raw materials of
diagenesis consist of organic or inorganic sediments and
of interstitial fluids. The sediments may be either allochthonous or autochthonous and will have been deposited
on the interface between the pre-existing formations and
the new depositional environment. The fluids will represent about 45 % of the volume of sands, 50 to 65 % of the
volume of silts, 80 to 90 % of the volume of argillaceous
muds, and 98 % of the volume of colloids at the moment
of deposition. Materials which are subsequently introduced or formed are included in the classification of "raw
materials".
The various minerals, materials and fluids making up
the sediment constitute a complex chemical mixture
which is generally not in an equilibrium state, and the
resulting interactions lead to a transformation of the sediment.
Both the size and composition of the minerals have a
bearing on the process. In fact, the percentage of certain
minerals depends on grain size (Fig. 6-6). Thus, phyllitic
minerals and organic matter are more abundant in the fine
than in the coarse fraction. Besides, some minerals are
more unstable or more soluble than others (aragonite,
pyroxenes, amphiboles, feldspars, micas), and will be
easily transformed or dissolved.
The size of the grains or crystals also controls both the
area of solid-fluid contact, and the permeability and
consequently the possibility of the fluids circulating. Grain
sorting will also control the location of the interactions.
The fluids will react according to their nature (water,
air, gas, ...) and their composition (salt and dissolved gas
content). We introduce the concept of diagenetic environment to describe how the fluids may evolve with time as a
result of hydrodynamic phenomena, or under the influence of temperature and pressure.
original void spaces, in other words, a reduction in the
initial porosity.
The amount of compaction depends on the initial
porosity and on the size, shape and sorting of the grains.
It also depends on the rate of sedimentation and time
period. Compaction is particularly marked in detrital sediments.
Compaction will be studied in detail in the following
chapter.
Cementation
This is one of the most common diagenetic phenomena. It is the deposition of minerals within the pore space.
The minerals may be derived from the sediment itself by
leaching and redeposition. They may also be derived from
salts dissolved in interstitial or circulating water, and may
be a single type or a mixture of several types.
Cementation may occur quickly or over a long period.
The most common cements are calcite, dolomite, silica,
clay minerals and less frequently anhydrite, halite, pyrite,
siderite or haematite (Fig. 6-7). Cementation results in a
reduction of porosity, and the quantity of cement cannot
exceed the initial porosity (in detrital deposits).
Figure 6-7 - On the left :thin section showing quartz cement (quartz
overgrowth) and calcite cement (in red). The pore space is in blue.
On the right :thin section showing early meniscus calcite cement
between rounded grains.
Recrystallization
This is the process by which the crystalline texture of
the rock changes as a result of the finer crystals dissolving and then growing again in different forms. The phenomenon is more common in chemical or biochemical
rocks than in detrital rocks and increases with burial.
Scale
Transformation
Figure 6-6 - Relationship between grain size and composition of the
detrital fraction in clastic silicate rocks (from Blatt et al., 7980).
Compaction
This is a mechanical rearrangement of grains under
the weight of sediments above during the burial process.
The result is a reduction in volume at the expense of the
242
This is the replacement of a mineral by its polymorph,
the commonest example being the transformation of aragonite to calcite.
Mineraloaical Redacement
In this case a new mineral replaces a previously existing one (Fig. 6-8). Included in this category are the pheSerralog 0 2003
is
nomena of dolomitization, phosphatization, pyritization
and silicification, along with the transformation of gypsum
into anhydrite and montmorillonite into illite. Dapples
(1962) gives a list of diagenetic reactions divided into
three stages (Table 6-1).
a) MUDSTONE
RHOMBOHEDRON
c) COMPOSITE CALCITE
RHOMBOHEDRON
(direct replacement origin)
d) PARTIALLY-LEACHED
DEDOLOMITE
RHOMBOHEDRON
e) RHOMBOHEDRALPORE
Redoxomorphic stage
(reversible reactions)
Locomorphic stage
Phyllomorphic stage
(replacement reactions) (unidirectional reactions)
Fe+2
aragonite by calcite
Fe+3
Hematite + calate
siderite
e
+ siderite + ferrodolomite
(cementation origin)
Figure 6-8 - Diagrams showing the diagenetic history of a limestone
(mudstone) (from Evamy, 1967).
carbonates by quartz
or chert
Hematite + clay minerals + quartz or chert by
silica +hlorite + greenalite carbonates
+ stilpnomelane (?)
Hematite +chlorite
chamosite (?)
Hematite + iliite
glauconite
+
+
"Bauxite" + silica
kaolinite
+
"clay minerals"-.chlorite
feldspars by carbonates
"clay minerals" + Fe*2 +
biotite
"clay minerals" +chert
chlorite
+
quartz, chert, clay minerals "clay minerals" +quartz +
by carbonates
sericite
opal by chert or quartz
silica solution
chert
kaolinite + glauconite or
illite + Mg+2+chlorite + K'
glauconite + feldspar
Diaspore or boehmite +
silica +clay minerals
illite or glauconite +
chlorite + muscovite
Iliaspore(?) + silica + K C
(glauconite)
kaolinite + illite +
glauconite + calcite + Mg+2
+ muscovite + biotite +
feldspar + dolomite +
chlorite
Bauxite + hematite +
silica (minor) =clay
minerals + pyrite
illite or glauconite +
calcite + Mg+2 +micas
+ feldspar + dolomite
f) RHOMBOHEDRALPORE
partially filled by calcite
g) COMPOSITE CALCITE
RHOMBOHEDRON
calcite by dolomite
"clay minerals"+ muscovite
or biotite
montmorillonite +chlorite
+ dolomite
+ clay minerals
druse
243
3
Table 6-1
Characteristic diagenetic reactions
(from Dapples, 1972).
Hematite + calcite + Mg +2
b) DOLOMITE
Chapter6
+
illite
Kaolinite + K +
Biotite
glauconite
Feldspar + clay minerals
+ chert
Glass G montmorillonite +
cheri
+
feldspar + sericite
plagioclase +chlorite + chert
Selective Leaching
Leaching is quite a widespread phenomenon, and is
often associated with other phenomena such as compaction (the solubility increases at the points of contact between grains with increasing pressure), recrystallization,
cementation and replacement. Selective leaching, however, must be considered separately. It affects only certain
constituents of the rock which have increased solubility
under certain conditions of temperature, pressure and
concentration of salt or dissolved gas in the interstitial
fluids. Of particular importance is the leaching associated
with acidic waters of meteoric origin which are rich in C02.
They can create vugs and even caverns in the rock which
can make a significant contribution to its porosity (Fig. 69).
Authiaenesis
This phenomenon results in the appearance of new
minerals, either by direct introduction or, more frequently,
by alteration of pre-existing minerals. It is very similar to
the process of cementation in terms of the end effect. An
example would be the formation of kaolinite by the alteration of feldspars.
Serralog 0 2003
Figure 6-9 - On the right :thin section showing oomoldic secondary
porosity. On the left : SEM photograph showing a vug.
Stvlolitization
This is a pressure-controlled solution phenomenon
which gives rise to stylolites (Fig. 6-10).
243
a
b
Figure 6-10 - a) Schematic of a stylolite. b) Schematic of a horizontal
stylolite with tension gash. The maximum principal strees axis is vertical. c) Photograph of a stylolitic surface in 30.d) Photograph of a stylolite seen on a plane surface.
The diagenetic phases
The numerous researchers in the field of diagenesis
have identified several phases in the process. Table 6-2,
drawn from a synthesis of the work of several authors by
Dunoyer de Segonzac (1968), attempts to unify the
various terminologies which have been applied to these
phases. We can confine our interests to the following
three main phases.
The first, which occurs early in the diagenetic process
is syndiagenesis (Bissell, 1959). This is characterized by
the presence of substantial amounts of interstitial water,
which is expelled very slowly, and by extreme variations
of pH and Eh. As Dunoyer de Segonzac pointed out, this
only affects a few tens of centimeters of sediment. It is
during this phase that the sediments are transformed into
rocks which are coherent and solid (lithification).
The second phase occurs late in the diagenetic cycle,
and is known as anadiagenesis (Fairbridge, 1967, Fig. 611). It is a phase of compaction and maturation which is
characterized by the expulsion and (usually) upward
migration of interstitial water and other fluids, such as oil
or gas, and by conditions of reduction. It takes effect over
several hundreds or thousands of meters of sediment
(Dunoyer de Segonzac), and generally results in a substantial reduction of porosity.
The third and final phase is epidiagenesis (Fairbridge,
1967). It is characterized by the modification of interstitial
waters as a result of the penetration and downward
migration of meteoric water, and by a reversion to conditions of oxidation. The phenomena of dissolution and of
the formation of "hard-ground'' mainly appear during this
phase. The zone affected is usually not very thick and
close to the surface.
These three phases more or less correspond to the
three periods and zones of diagenesis defined by
Choquette & Pray, 1970 (Figure 6-12)
Table 6-2
Various stages of diagenesis according to different authors (extracted from Dunoyer de Segonzac, 1968).
Stages
Authors
Van Hise
(1904)
Twenhofel
(1926,3930)
Krumbein
(1942,1947)
K. & Sloss
(1963)
Deposition
Deposition
of lithogenesis
Particles immobilized
in a sediment with a
Tectonic phenomena
place the sediments
V under conditions of
decomposition and
leaching or exposed
244
I
depositional
interface
katamophism weathering
(delithification)
Pettijohn
(1949,57)
I
Packham
Crook
(1960)
halmymlysis
I
weathering
Dapples
(1959)
Dapples
(1962)
I
I
halmymlysis
weathering
Williams
Turner
Gilbert
(1954)
weathering
Wolf
Conolly
(1965)
predepositional
I
I
halmymlysis initial stage
,
=depositional
1
Taylor
(1964)
preburial
syngenetic
Redoxo-
weathering
weathering
weathering
Serralog 0 2003
Next Page
Dianenesis
I
ChaDter 6
I
245
- The eogeneticperiod is the time interval between the
end of deposition and the start of effective burial. The
eogenetic zone is the interval of depth at or near the surface of sedimentation where the chemistry of the interstitial water is largely controlled by the surface environment
before any effective burial takes place (less than a few
tens of metres).
- The mesogenetic period is the post-depositional period from the start of burial to the start of the diagenetic
processes associated with erosion. The mesogenetic
zone is the depth interval within which the diagenetic processes associated with burial take place.
- The telogenefic period refers to the interval during
which rocks which have previously been buried are subjected to the diagenetic processes associated with subaquatic or subaerial erosion. The telogenetic zone is the
name applied to the corresponding depth interval.
Several diagenetic cycles incorporating some or all of
the above-mentioned phases may occur, each one leaving its marks and possibly obliterating those of preceding
cycles.
J
Figure 6-12- Schematic representation of the main areas and burial
zones in which porosity is created or modified
(from Choquette & Pray, 1970).
1
Figure 6-11- Idealized profile across a continental margin showing the
marine sedimentation sites and the three phases of diagenesis :(I)
diffusion during syndiagenesis; (2) movement of ascending liquid
during anadiagenesis; (3) movement of descending fluid during epidiagenesis (from Fairbridge, 1967).
Effects of diagenesis on the sediment properties
The sediment properties which were originally deposited will change as a result of physical, chemical, mineralogical, textural and structural changes associated with
the various diagenetic phenomena. The principal effects
are summarised in Table 6-3; among them are :
- Size and shape of particles : the grains may increase
in size and form crystals by growth or re-crystallization.
Alternatively, the grain size may decrease because of leaching.
- Mineralogical composition : transformation, replacement, cementation or authigenesis are the only means by
which the composition of the original sediment may change.
- Porosity: the majority of diagenetic processes, apart
from dissolution and occasionally replacement, involve a
reduction of porosity which may be quite substantial. This
reduction is referred to as poronecrosis. Conversely, the
Serralog 0 2003
phenomena of dolomitization and particularly of leaching
increase the porosity, sometimes quite visibly, by the formation of vugs and caverns. This effect is referred to as
porogenesis. However, the way in which the porosity
develops will depend on the type of sediment (detrital or
chemical : Table 6-4) and on the texture (Table 6-5).
- Permeability:precipitation of cement, recrystallization
and compaction will usually result in a substantial reduction of permeability along with a loss of porosity.
Dissolution and mineralogical replacement will only
increase the permeability if the newly-created pores are
connected to the existing pore system, and if the interconnecting channels are themselves enlarged.
- Sedimentary structure : the original sedimentary features can be obscured or deformed by certain of the diagenetic processes: recrystallization, compaction, the
action of organisms, leaching or expansion.
The need for diagenetic studies
Because of the ways in which diagenetic phenomena
modify the original sediment, it is important to be able to
evaluate the extent of these changes in order to construct
a reasonable model of the depositional environment and
to successfully identify the original facies.
In addition, by recognizing the various stages and
cycles of diagenesis, we are in a better position to reconstruct the geological history of a rock, and hence of the
sedimentary basin in which the rock has been deposited.
Diagenetic studies are generally performed on rock
samples by examination of thin sections and images
245
Previous Page
obtained by scanning electron microscopes.
Table 6-3
Relative effects of different diagenetic processes on the
properties of sediments (from Krumbein, 1942).
Property
Cornpaction
CemenRecrystation tallitation
Size
-
Shape and
roundness
-
Surfacetexture
-
Particleonentabon
7
-
Mineral
composition
-
ac
Porosity
Permeability
Color
x
X
x
xx
xx
m
xxx
X
xx
Replace- Differen- Authigenesis
rnent
tia!
solution
xxx
-
xx
xx
-
xx
xx
xx
X
XXX
xx
7
X
-
7
xxx
xx
xx
X
!a
X
x
xx
-
Aspect
xxx
Curhonute
mary porosity
in sediments
a%
Commonly 40-70%
Amount of
ultimate porosity in rocks
Commonly
half or more of
initial porosity;
I5-30% com-
Commonly none or only snuill
fraetton of initial porosity: 515% common in reservoir
facies
Typds) Of ptimary porosity
Almost exclu-
xx
xx
xx
Legend : x = small to moderate effect; x x = moderate to large effect; x x x = property most strongly affected by a given process; means a negligible effect; and
? indicates an unknown effect.
mon
Diagenetic studies are fundamentally based on :
tnterparticle commonly predominates. but intrapartidc
and other types are imponant
Almort exclusively primary
interparticle
Widely varied b u r of postdepositional modifications
S i z a of pores
Diameter and
throat sires
clovly related
to sedimentary
particle s i n
and sorting
Diameter and throat dzes cbmmonly show little relation to
sedimentary panicle uzc or
wrting
Shape of pores
Strong depcn&nce on partick s h a m
“negative” of
particles
Greatly varied. ranges from
strongly dependent “positive”
or “negative” of particks to
form completely independent
of s h a p a of depositional or
dmgenetic components
Uniformity of
size, shape. and
distribution
Commonly
fairly uniform
within homogeneour body
Variable. ranging from fairly
- analysis of rock-sample reaction to hydrochloric acid
(HCI);
- observations of polished surfaces using staining
techniques;
- analysis of thin sections;
- analysis of SEM photographs and Energy Dispersive
Analysis (EDAX);
- electron microprobe analysis;
- cathodoluminescence techniques.
Through these analyses the different phases of the
diagenetic history of a rock sample can be reconstructed
as previously illustrated by Figure 6-8 and by the following
figures.
Figure 6-13 (see next page) illustrates some of the diagenetic modifications that can occur into a sediment. As
previously indicated these modifications can occur either
in vadose (Fig. 6-14) or phreatic zones (Fig. 6-15) or induced by burial. Figure 6-16 indicates the influence of Mg
and Na on the morphology of carbonate crystals (Folk,
1974).
Staining techniques allow the distinction between carbonate minerals, aragonite from calcite by Feigel’s solution, dolomite from calcite by Alizarine Red-S in a 0.2%
HCI solution (Fig. 6-17), Mg calcite by use of Clayton yellow stain, and Fe calcite by use of a potassium ferricyanide stain in a weak HCI solution.
Figures 6-18 and 6-19 illustrate several diagenetic
phenomena that can be recognized in sedimentary rocks
by observation of thin sections.
Figures 6-4 (page 3) illustrate overgrowth of quartz
grain. Cathodoluminescence technique, allowing the
detection of secondary phase of cementation, is illustrated by Figure 6-20.
dvely interparticle
Typdr) of ultimate porosity
Diagenetic studies
Traditional methods
246
Sandstone
Commonly 25-
Amavnt of pri-
X
xx
X
xx
Figure 6-21 illustrates stylolitization seen on thin section. Secondary porosity development by selected leaching is illustrated by Figure 6-22 and geopetal fabric by
Figure 6-23.
Table 6-4
Comparison of porosities in sandstones
and carbonates (from Choquette & Pray, 1970).
uniform to extremely hetero-
geneous. even within body
made up of single rock type
Influence of
diagenesis
Minor; usually
minor reductionofprimary
porosity by
compaction
and cementatior
M a p r ; can create, obliterate.
or completely modify porosity;
cementation and solution important
Influence of
fracturing
Generally not
of m a p r imponance in
reservoir prop-
Of major importancc in m e r voir properties if present
erties
Visual evaluation of porosity
and perme
ability
Semiquantitative visual a t i -
mates commonly retatively eaay
Variable; rcmiquantiiative visual a t i m a t a ranse fromesay
to virtually impouible; inurument mcssuremcnts of masicy. permeability and capillary
pnwrrecommonly necded
Adequacy of
-re analysia
for nrcrvoir
evaluation
Core plugs of
I-in. diameter
commonly a&
quste for “matrix” porosity
Core plugs commonly inadequate;evenwhole cora(-3-in.
diameter) m a y k inadequate
for large p o r a
Permcabilityporosity interrelations
Relatively consiatent; commonly dcpendentpn panicle
dzeand wrting
Greatly varied; commonly in-
dependent of particle sire and
sorting
The traditional methods are the only ones that allow
the determination of the successive diagenetic phases
and periods which have modified the original properties of
Serralog 6 2003
Diagenesis
1
Chapter6
I
247
Table 6-5
Different types of porosity evolution during diagenesis (courtesy of Gulf Oil of Canada Ltd).
in matrix
dolostones
I
I
Preserved
sediment
porosity
No essentiel
changes in
sediment porosity
CompLction
pressure &
solution
I
Recrystallisation
& replacement
of matrix
I
I
I
leaching of
leaching of
leaching of
Arenite and rudite size carbonate particles
Fracture porosity can develop
in rocks
independent of composition
or texture
- with or without lime mud matrix
bioconstructed carbonate framework with
or without fine matrix
I
I
Cementation
I
I
Organic
cementation
I
Mineral
precipitation
I
Advanced.
dolomitization
-I
Porosity
Porosity
reduced or
eliminated
Porosity
eliminated
in matrix
mainly in
interfossil bridged
& intrafossil types
Porosity
reduced or
eliminated
I
Porosity reduces
reduced
or eliminated
mainly in
matrix
MUD-SUPPORTED SEDIMENTS
3
g
W
t
3f a
Intercrystalline
micropores & small
mesopores in
sucrose dolostone
Vugs in sucrose
or coarser
crystalline
dolostone
I
3
8
Dolomitization
2
n
I
-Iw
.v,
n ->
! &
I
Contemporaneous
solution of sand
or larger size
calcitic particles
in lime mud during
dolomitization
Vugs in dolostone
I
Leach ing of
incompletely or
non dolomitized
sand or larger
size carbonate
particles
Vuggy & moldic
porosity
in limestones
Leaching of more
soluble carbonate
particles
I
I
I
I
d
Serralog 0 2003
247
e
Figure 6-13 - The upper five figures illustrate schematically and successively from left to right: the original sediment, its cementation, its partial
dissolution, its dolomitization and its recrystallization (courtesy of B. Purser). The four lower figures illustrate how these situations can be detected
by observation of thin sections through microscope. The last figure explains the link between mold and vug.
A
Figure 6-14 - Diagrams illustrating the different types of cement generated in the vadose zone (courtesy of 6.Purser).
Figure 6-16 - Diagenetic realms and carbonate crystal morphology.
Three main realms exist: 1) Na+ and Mg++ both high as in marine
cement (beachrock, subtidal, reef cement...). 2) Na+ high and Mg++
low, as in the subsurface zone, where Mg++ has been removed
through trapping by clays and dolomite (mixing zone between connate
and fresh meteoric Waters). 3) Na+ and Mg++ both low, in the meteoric
zone (from Folk, 1974)
-l
Figure 6-15 - Type of cementation as a function of the depositional
environment (courtesy of B. Purser).
the sediment since its deposition. In addition, they also
allow a very precise sequence stratigraphy determination
that cannot be dtected by others techniques.
I
Figure 6-17 - Staining of a thin section. Calcite cement is in red.
240
Serralog 0 2003
Diagenesis
ChaDter6
1
249
I
Figure 6-18 - On the left, illustration of meniscus cement between each
oolitic grain. On the right, illustration of microstalactitic cement
developped in the vadose zone. Porosity is in black.
w
Figure 6-22 - Thin section illustrating oomoldic secondary porosity
(in purple).
I
Figure 6-19 - On this thin section one can recognize two phases of
cementation. The first one (in white) is composed of sparry calcite
crystals, the second is composed of ankeritic dolomite which has filled
the pore space. This later deposit is associated with an unconformity.
Fiigure 6-23 Illustration of
geopetal sediment or vadose
silt (micritic
mud) filling the
lower part of a
gastropod. The
upper part is
filled by sparry
calcite.
The calcite stability is a function of both salinity and
Mg/Ca ratio in water as illustrated by Figure 6-24.
3
Figure 6-20 - Cathodoluminescence technique (right image) allows the
detection of quartz overgrowth which was not detected through thin
section analysis alone (left image).
Y
Figure 6-21 - On the left, theoretical diagram of a stylolite. On the
right, a sty/olite is well detected on this thin section.
Figure 6-24 - Calcite and dolomite stability as a function of the
salinity and the Mg/Ca ratio (from Folk & Land, 1974)..
Serralog 0 2003
249
Well logging approach
If a detailed analysis of rock samples can provide a
reconstruction of the diagenetic changes which have
taken place and hence the stages of diagenesis, it is not
then necessary to make use of well logs to analyze the
diagenetic history of the rock. Well logs analyze the rock
in its present state, and hence both the properties which
were determined by the original depositional environment,
and those which were the result of diagenesis. At best,
thus, they will only give a picture of the final state of the
rock which is the end result of a chain of diagenetic processes.
Having said that, it is only necessary, during the detailed analysis of the logs, to bear this in mind and to keep
an eye out for signs which may indicate certain diagenetic phenomena. Clearly, from what has already been said,
this approach must be achieved taking into account the
type of rock studied.
Diagenesis in detrital sequences
The type and extent of diagenetic changes depend largely on the maturity of the rock, that is, on its content of
stable minerals and on its textural properties (Fig. 6-25).
Rounded
Mature
Figure 6-26 - Diagram showing changes in the relative mineralogical
maturity of sandstones (from Weller, 1960, in :Chilingarian & Wolf,
Compaction of Coarse-grained Sediments, Developments in
Sedimentology, Elsevier, Amsterdam).
Table 6-6
Some reactions involving clay minerals during
the diagenesis of sandstones
(from Pettijohn e t a / , 1972).
Feldspar
Components added to (+)
or subtracted to (-)
- (K+, S O 2 )
Kaolinite
+ (K+, S O 2 )
Mite
Kaolinite
+ K+, - H2O
Muscovite
Precursor
Clay mineral formed
Kaolinite
Montmorillonite
+ K+, - (Si02, H 2 0 , Na+, Ca++, M g ++,Fe++..) Mite
Montrnorillonite
+ (Fe++, Mg++),- ( 3 0 2 , H 2 0 , Na+, Ca++) Chlorite
Volcanic glass
- (Na+, K+, Ca++)
+ (Fe++,Fe+++), - (
K', A1203)
lllite
Porous space
+ H20,
+ (A1203, Si02, H 2 0 )
Montmorillonite
Glauconite
Kaolinite
Pressure-controlled solution
Figure 6-25 - Diagram illustrating Folk's textural maturity concept (from
Weller, 1960, in :Chilingarian & Wolf, Compaction of Coarse-grained
Sediments, Developments in Sedimentology, Elsevier, Amsterdam),
For example, the constituent minerals of a quartz-arenite will react differently to those of an arkose or a graywacke to burial and to the subsequent diagenetic processes (Figure 6-26 and Table 6-6). Consequently, the recognition of the mineral composition of a rock will inform on
the possible development of cement linked to unstable
mineral alteration.
Compaction
This is one of the major diagenetic phenomena in
detrital sequences, regardless of composition. It is generally accompanied by secondary effects. In view of its
importance, the following chapter is devoted entirely to
the study of compaction.
250
The underlying concept of this phenomenon is that the
high pressures developed between the points of contact
of the grains result in increased solubility. This leads to
preferential leaching and this is further encouraged by the
presence of a film of clay. Fuchtbauer (1967) suggested
that pressure-controlled solution appears early, and is the
reason for loss of porosity in sands. Sippel (1968), however, demonstrated by cathodoluminescence technique
that detrital grains are not involved, and that authigenic
crystal growths appear in sands at concavo-convex or
suture contacts (cf. Fig. 6-20).
Regardless of its importance, pressure-controlled
solution is not the only cause of cementation.Alteration of
feldspars and other silicates by meteoric water produces
dissolved silica which may be re-precipitated in the sand,
or may contribute to the growth of quartz crystals. The
water itself may be supersaturated in dissolved silica.
This phenomenon is not distinguishable from silica
cementation by well logs, because the only detectable
result of both phenomena is a reduction in porosity.
Serralog 0 2003
Diagenesis
Cementation
When analysing logs from the diagenetic point of view,
it is necessary to establish the type of cement involved
and the degree of cementation. Sands which are composed of quartz grains (orthoquartzite) are recognised by
their very low radioactivity, due to the almost complete
absence of potassium and generally low content of thorium and uranium, the exact amount depending on the
presence of heavy, stable minerals such as zircon or
monazite. These sands have a cement which may be
composed of quartz, calcite, haematite or, less frequently,
of authigenic clay minerals such as kaolinite or dickite.
Arkoses and graywackes, which have a relatively high
potassium content and hence moderate to strong radioactivity, usually have a cement of calcite or haematite, or
of authigenic clay such as kaolinite, chlorite, montmorillonite or illite. Siliceous cement is rare because alkaline and
alkaline-earth cations from the alteration of feldspars and
from mafic minerals or volcanic grains combine with silica
to form authigenic clays or zeolites.
One of the most reliable means of evaluating the type
of cement is to use the litho-density (LDT)-neutron combination and the corresponding crossplot (Figure 6-27).
Orthoquartzites with a siliceous cement are clustered
around the point Q (quartz). At the same time, the density vs. neutron-hydrogen index plot reveals the loss of
porosity with reference to the general compaction gradient of sands, and from that, the quantity of cement
(according to Pettijohn, 1963 and Pettijohn et a/., 1972).
Before concluding that the cement is siliceous, we must
be sure that the decrease in porosity is not due to poor
sorting.
Chapter 6
1 251
distance of the point from Q to the length of the line Q-C.
In the same way, the points for a dolomitic cement fall
around a line joining Q and D (dolomite). If the cement is
a mixture of calcite and dolomite, the points fall between
the lines Q-C and Q-D (Fig. 6-28).
Figure 6-28a - Crossplot RHO6 vs PHIN with PEF on the Z-axis. The
very porous sand (30% porosity) shows a decreasing porosity with
displacement of the points towards the dolomite pole (arrow). A development of calcite would be marked by an increase in Pe value.
n
6
.
Figure 6-28b - Crossplot RHOMA vs UMA with THOR on the Z-axis.
One can observe trends towards opal and microcline with low thorium
values and trends towards dolomite and clay, but not towards calcite.
tn
Y
E
Q
Uma ( b / c m 3 )
Figure 6-27 (pmaja vs (Uma)a crossplot for determining the mineralogy of rocks and the cement type.
In the case of a calcite cement, the points are grouped
around a line joining the points Q (quartz) and C (calcite).
The percentage of cement is the-, given by the ratio of the
Serralog 0 2003
Certain sands have a halitic cement. Thus, in theory,
the points should fall around a line joining Q and the halite point H. However, if the well has been drilled with a mud
which is not salt-saturated, the halitic cement will dissolve, and the logging tools will not detect it (Figure 6-29).
Halitic cements are usually found near salt formations.
The Pe, which is less sensitive to porosity, may show
values which are consistently higher than those of quartz
(1.8 b/e), and this distinguishes this case from a gas sand.
The Schlumberger TDT tool is useful in this-situation
because of its sensitivity to chlorine (Figure 6-30).
Pyrite or haematite cements will displace the points
towards high values of ,U
,
and pma. If the quantities are
high enough and the cement is more or less in continuous
25 1
contact, the resistivity will be very low since pyrite and
haematite are both conductive.
For arkoses or graywackes, the points fall in the area
AI-Q-An-C-F for a calcareous cement. These rocks may
be recognised by plotting the potassium percentage on
the Z-axis.
CPI WITHOUT TDT
CPI WITH TDT
-a
I
v)
s
-J
Figure 6-29 - pb vs @J crossplot showing an example of a sandstone
having a halitic cement (courtesy of Schlumberger).
Authiaenesis
The formation of authigenic minerals in terrigenous
detrital sequences will depend on the textural and chemical maturity of the rock, on the types of fluid and hence on
the hydrodynamic conditions and on the compaction
(simultaneous action of temperature and pressure).
The formation of authigenic minerals in orthoquartzites
is usually limited to the precipitation of "books" of kaolinite (Figure 6-31). These may be recognised on the same
crossplot. The Pe value of kaolinite is close to that of
quartz (1.8 b/e), but its content of hydrogen ions gives it a
value of (P,,,~)~ and (U,,,a)a which allow them to be distinguished (cf. Fig. 6-27). Thus, the points are grouped
around the Q-K line.
It is difficult to distinguish authigenic kaolinite from kaolinite of detrital origin. However, it is reasonable to assume that authigenic kaolinite is purer, and that, as a result,
it contains practically no potassium, and even less thorium and that it therefore has very low radioactivity.
In immature sequences (arkoses and graywackes),
authigenic minerals are common. The most common are
the clay minerals (illite, montmorillonite, chlorite, etc.),
less frequently potassium or plagioclase feldspars. It is
not usually possible to distinguish between these authigenic minerals and their detrital counterparts from well logs.
The presence of authigenic minerals will have a considerable effect on permeability, but this will vary according to
how the clays are distributed. The distribution in its turn
often depends on the type of clay mineral (Figure 6-32
and Table 6-7).
252
Figure 6-30 - Sandstone with a halitic cement. Comparison between
two CPls run with and without the TDT log (courtesy of Schlumberger).
Nodules
The analysis of the microresistivity curves from dipmeter logs, combined with the dips, allows identification of
resistive events which only appear on a few curves
(Figure 6-33), more rarely on all 4 or 8 curves. In this last
case the thickness of the resistive events varies from one
Serralog 0 2003
Diagenesis
curve to another, and the dips computed at the bottom
and at the top of the event will be different (Figure 6-35).
b - Pore lining by grain coating
-
c Pore bridging by fiber crystals
Schematic
SEM microphotographs
Chapter 6
253
When these events appear suddenly in a homogeneous
sandstone, they correspond either to cemented balls, formed by dissolution of shell fragments and reprecipitation
of the released calcite in the surrounding pore space by
radiating diffusion, or to isolated pebbles in a sandy
matrix, or to anhydritic nodules. The choice between
these hypotheses can be made by checking the nature of
the previous deposit or the type of the environment. If a
grain size evolution can be detected from the logs, or if
evidence of grain supported conglomeratesexists (cf. Fig.
3-56), or if we are in an alluvial fan, a channel bed load, a
mass flow deposit, or a glacial environment, the pebble
hypothesis is the most likely. The cemented balls hypothesis should be prefered in a deltaic or mid fan turbidite
environment because it is impossible to explain how one
pebble could be transported and deposited in a region
where the energy was such that it generated a homogeneous sand (Fig. 6-36).
Anhydritic nodules is the most likely hypothesis in the
interdune sabkha deposits of an aeolian environment.
They will be also characterized by a very high resistivity
(> 500 ohm-m).
Cemented or anhydrite nodules are well detected on
images of the borehole wall. They correspond to isolated
white spots in a more conductive background (Fig. 6-37).
Pyrite crystals
Pyrite crystals, formed by diagenesis by reduction of
sulfates, are easily detected by the Formation
Microscanner tool. They are characterized by very
conductive peaks (Figure 6-38) on the resistivity curves,
Table 6-7
Characteristics of authigenic clays
(from Wilson & Pittman, 1977).
Figure 6-31 - Distributionmode of authigenic clays
(from Neasham, 1977).
Relationship
to sand size
detrital grains
Thickness of coating
or long dimension of
aggregates (microns)
pore filling
2 - 2500 (generally 2-20)
vermicule
pore filling
10 - 2500 (generally 20-200
sheet
pore filling
Morphology of
individual flakes
Form of aggregates
Kaolinite
pseudohexagonal
staked plates (books)
Dickite
pseudohexagonal
pseudohexagonal
Clay mineral typt
Chlorite
pseudohexagonal
surled equidimensional with rounded
2dges
squidimensional
Nith angular or
lobate edges
'an-shaped fibrous
iundles
lllite
rregular with
?longatespines
Smectite
lot recognizable
Mixed-layer
smectite-illite
subsequant with
stubby spines
Serralog 0 2003
plates ( 2-dimensional
card house)
honeycomb
pore lining
pore lining
0.1 - 1
Special features
Flakes notched or embayed
(twinned?)
Flakes notched or embayed
(twinned?)
Flakes notched or embayed
(twinned?)
2-10
2-10
rosette or fan
pore lining &
pore filling
4-150
(generally 4 - 20)
cabbage head
pore lining &
pore filling
8 - 40
pore lining
0.1 - 10
bridging between sand grains
pore lining
2-12
bridging between sand grains
2-12
bridging between sand grains
sheet
wrinkled sheet or
honeycomb
imbricated sheet
to ragged honeycomb
pore lining
253
or by dark black spots on the image display, these spots
reflecting sometimes the shape of the crystals.
4a
Figure 6-34 - Resistive events appearing on a few curves.
They correspond to cemented balls or lenses.
Figure 6-32 - On the left: influences on porosity and permeability of the
distribution type of authigenic clay minerals. Data from Neasham’s
study of 14 sandstones (from Neasham, 1977).
On the right: effect of type and distribution of authigenic clay onpermeability. The porosity is unchanged, in spite of a changein permeability of 4 orders of magnitude in the extreme cases
(from Stadler, 1973; in Blatt et al., 1980).
LOCDIP results
Figure 6-35 - One cemented ball crossed by the well. Observe the
change in thickness from pad to pad. The core photograph confirms
the presence of the ball.
P d3
PPd 1
Figure 6-33 - Effects of detrital and authigenic clays on petrophysical
characteristics of reservoir.
EPidiaaenesis
As we have already seen, this corresponds to the
development of secondary porosity by leaching, and is
detected by comparing the porosity measurement of the
sonic tool with that obtained from the density-neutron
combination. This occurs in detrital sediments in which
were present carbonate fragments (bioclasts, oolites...).
Diagenesis in carbonate sequences
The diagenetic phenomena affecting carbonates are
appreciably different from those affecting detrital sequences. Thus, compaction is of minor importance because
other diagenetic phenomena may have already affected
the rock and made it more consolidated even before the
completion of the burial process.
254
Figure 6-36 - Calcite nodules (white or yellow spots) in a deltaic (left)
and turbiditic (right) sandstone reservoir. The volume of these nonporous nodules , calculated from the left FMS image, corresponds to
12% of the total sandstone. Thus, the porosity of the sandstone itself,
will be higher than the porosity value determined from the neutron-density measurement. The permeability estimation must be evaluated from
the porosity of the sandstone, not from the bulk volume porosity.
(courtesy of Schlumberger).
Serralog 0 2003
Next Page
Diagenesis
I Chapter6 1
255
The points will plot very close to the calcite point, C, on a
(P,.,,~)~vs (U,.,,Ja crossplot (Figure 6-39), except in the
case of aragonite, for which the points fall towards the
bottom of the crossplot.
Figure 6-37 - Anhydrite growth in a sandy dolostone and evaluation of
their density and the area covered by these anhydrite crystals
(from Delhomme & Motet, 7993).
LlTHO
SYNDIP
DESCRIPTION
TRACES
IMAGE
msisnvmINCREASE
Figure 6-39 - Several Z-plots (thorium on the Z axis) for mineral and
cement determination.
~
the superposition of points on the limesUpper left: pb vs I P Observe
tone and dolostone lines respectively with the change in porosity indicating development of cement of same type except at the left corner
where a dolomitic cement appears in the limestone.
Upper right: Pb vs Pe. Observe the grouping of points along the limestone and dolostone lines.
Lower left: Pe vs potassium. The points evolve from the dolomite position to the calcite position with few points moving to the right with a
thorium increase indicating the influence of clay mineral (illite).
Lower right: (n,,,J, vs (Umala. The points are grouped around the calcite and dolomite position with some influence of clay moving the
points towards a lower position.
Neomorphism
Figure 6-38 - Pyrite crystals in a shale. They are characterized on the
Formation Microscanner curves by very conductive peaks, and on the
images by very black spots showing the shape of the crystals
(courtesy of Schlumberger).
Crystallization
This is the phenomenon by which primary pore space
is filled with crystals of calcite (sparite) or aragonite by
precipitation or drusy growth. This is accompanied by a
loss of porosity which is easily detected from the porosity
logs, assuming the type of mineralogy does not change.
Serralog 0 2003
It corresponds to a "transformation between one mineral and itself or a polymorph whether the new crystals are
larger or smaller or simply differ in shape from the previous ones, or represent a new mineral species" (Bates &
Jackson, 1980).
A first type of transformation corresponds to the change of aragonite to calcite. It is accompanied by a reduction of density (2.94 g/cm3 for aragonite against 2.71 for
calcite) and a loss of porosity. In theory it should be very
easy to detect from the logs. This transformation always
occurs early in the cycle, and the majority of ancient
sequences have been subjected to it.
Dolomitization (replacement of calcite with dolomite)
belongs to the second type of transformation. It brings
about a 13% reduction in volume, and a corresponding
increase in porosity if other phenomena do not interfere.
The log matrix density increases to a value of 2.87 g/cm3.
It is this parameter above all which allows this diagenetic
phenomenon to be detected. Crossplots of density-neu255
Previous Page
tron, M vs N, or (p),
vs (U,),
provide the best means
of detection (Fig. 6-40). (p,.,,,), and (Urn,), vary with the
intensity of dolomitization, which may itself vary from one
point to another.
(SPl)rel= (+N-D - +s) 1+N-D
Figure 6-41 - Crossplot M vs N showing how points are shifted as a
result of secondary porosity (courtesy of Schlumberger).
A
Figure 6-40 - Other example of 2-plots (GR on the Z axis), density vs
on the upper right,
neutron on the upper left and (p,,,,), vs
sonic transit time vs neutron on the lower left and porosity from
neutron-density vs sonic transit time on the lower right, showing the
loss of porosity of limestone and dolostone and the trend towards
anhydrite. Limestones are slightly radioactive due to glauconitic pellets.
Recrystallization cannot be detected from the logs
unless accompanied by a change in mineralogy or porosity.
Dedolomitization is the inverse transformation of dolomite into calcite, and so leads to a reduction in porosity
and a decrease in density. This process cannot be distinguished from cementation by calcite using only the logs.
Selective Leaching
The method for detecting this phenomenon with well
logs is well known. It involves comparing the porosity derived from the density-neutron combination with that obtained from the sonic. The sonic tool detects the arrival of
the fastest acoustic wave which is the one corresponding
to the most direct path, that is, the one which avoids the
vugs and caverns which slow down the waves.
The crossplot of Figure 6-41 illustrates the method.
The degree of leaching is measured by the secondary
porosity index, SPI, given by the following relation:
SPI = +N-D - +s
or by the relative secondary porosity index, given by
the following relation:
256
The development of secondary porosity has an effect
on the m parameter of the Archie’s equation, especially if
the pores are not connected. This requires a continuous
computation of the m exponent. This is achieved taking
into account the measurement of the transit time of an
electromagnetic wave. In that case one can write the following equation:
tpl
= tpmf.($ND.Sxo+ ($ND(l-Sxo)tphc
+
(l-@ND)tpma
in which:
tp,= transit time measured by the electromagnetic propagation tool,
tpmf= transit time in mud filtrate, generally close to 25
or 30 nanos/m,
+ND = porosity computed from the neutron and density
tphc = transit time in hydrocarbon (3.3nanodm in gas,
4.7-5.2 in oil),
tpma= transit time in rock solid (matrix), close to 8.7 for
dolostone and between 9.1 and 10.2 in limestone.
S,, = water saturation in the invaded zone.
The later parameter is the only unknown. It can be
extracted from the previous equation and inserted in the
Archie’s equation of saturation in the invaded zone:
s,
=@7iFG
in which:
Rmf= resistivity of the mud filtrate,
R,, = resistivity measured by a microresistivity device,
a = a parameter generally assumed to be equal to 1.
The only unknown from this equation is the m exponent which can be computed continuously and reproduced and compared to the secondary porosity computed
from the sonic and neutron-density combination (Fig. 642).
Serralog 0 2003
Chapter6
Variable m
i
257
Constant m
Figure 6-43 - The saturation computation using a constant m value
indicates gas in the lower part of the reservoir. Using a variable m this
lower part is water bearing. This was confirmed by the test of the two
zones, dry gas was produced in zone A, water in zone B
(Schlumberger, Middle East Well Evaluation Review, 2, 1987).
Figure 6-42 - Results of a quantitative interpretation
which displays the secondary porosity and the value of the m exponent
of the Archie's equation. Interconnecting vuggy porosity is identified by
the divergence of secondary porosity computed from sonic measurement and the Archie's m exponent curves. Above the two curves run
together, indicating isolated vugs.
(Schlumberger, Middle East Well Evaluation Review, 10, 1991).
This variation of the rn exponent has a big influence on
the saturation computation in the virgin zone as illustrated
by Figure 6-43.
Selective leaching and vugs can also be recognized
on dipmeters and micro-resistivity image tools. Vugs correspond to conductive peaks with irregular shapes, which
are not easily correlated even from button to button and
all the more so from pad to pad (Fig. 6-44). On the microresistivity images vugs are represented by dark gray
spots (Fig. 6-45) of which the shape, the frequency and
the connectivity can be analyzed and estimated(Figs. 646).
Stylolitization
This is a pressure-controlled solution phenomenon.
Stylolites are characterized by very fine joints of irregular,
sawtooth form, of variable height and interpenetrating
(Fig. 6-47). These joints often contain a concentration of
insoluble materials (clays, iron oxides, organic matter,
coal, lignite fragments). There is a loss of porosity due to
reprecipitation of dissolved calcite in the surrounding pore
space.
Serralog 0 2003
Pad 1
Pad 2
Pad 3
Pad 4
Figure 6-44 - Typical dipmeter resistivity curves in a vuggy carbonate.
Vugs are characterized by very thin and n a m w conductive spikes.
Stylolitization can be detected through well logs by a
sudden large drop in porosity and a simultaneous increase in radioactivity (Figure 6-48) often due to enrichment in
uranium. This phenomenon is most frequently encountered in carbonate sequences, but may also be found in
compacted sequences of terrigenous clastics.
In favorable cases (sufficient thickness, or resistivity
contrast between stylolites and surrounding beds), stylolites are easily recognized on the micro-resistivity images
as illustrated by the examples of the Figure 6-27. By stu257
dying them it should be possible to determine their frequency, and the axis of maximum stress.
Orlentatlon
SIZE ( m d ) (lin)
Figure 6-466 - Other example of analysis of the mold population on
iamges in size, area and density (from Herron et al., 1994).
Figure 6-45 - On the left: example of Formation Micro Imager (FMI)
images showing oomoldic econdary porosity in limestone.
On the right: extraction of the vugs by thresholding
of the microresistivity FMS curves
(courtesy of Schlumberger).
Original FMS Curves of vug density Boundaries
size and conne&ty
ofthe vugs
image
c%2%
zone
I
Connectiilty
of the W S
Figure 6-47 - Examples of stylolites easily recognized on
microresistivity images (courtesy of Schlumberger).
Figure 6-46a - Example of image analysis showing the extraction
of conductive events, analysis of their connectivity and estimation of
their density and area
(Delhomme, 1992).
Cherts and anhvdritic nodules
When very resistive events appear in a homogeneous
and porous limestone (chalk) characterized by flat curves
without variations, chert nodules seem to be the most likely hypothesis. When these events are observed in a
heterogeneous dolomitic, or better, anhydritic limestone,
they may correspond to anhydritic nodules generated in
the supratidal subenvironment. The micro-resistivity
image tools enable recognition of their shape and their
fre&ency. They correspond to very resistive peaks (R
500 ohm-m), and are seen as white spots with irregular
shapes on the images (Figure 6-49).
’
258
Figure 6-48 - Example of stylolites detected by gamma ray and uranium measured on core lfrom Hassan et al.. 1976).
Serralog 0 2003
Diagenesis
Chapter6
I 259
1
Figure 6-49 - Example of anhydritic nodules detected on the
2-pads Formation Microscanner images, confirmed by core photograph (courtesy of Schlumberger).
Hard-Ground
-I
Figure 6-51 - Oysters fixed on hard-ground in the Dogger formation
(Massangis quarv, Paris Basin, France).
It corresponds to “ a zone at the sea bottom, usually a
few cm thick, the sediment of which is lithified to form a
hardened surface, often encrusted, discolored, case-hardened, bored, and solution-ridden” (Bates & Jackson,
1980). It is often rich in oxides of iron and manganese.
(Figs. 6-50 and 6-51). It indicates a pause in sedimentation. It is frequently overlaid with a fine intraformational
conglomerate, rich in phosphate debris and glauconite in
a marly matrix. It appears on the logs as a very narrow
peak of very high density and resistivity which is best
seen by the micro-devices (MLL, ML and HDT, SHDT or
images), with a radioactive peak just above (Fig. 6-52).
Figure 6-50 - A “hard-ground horizon in a chalk
(from Selley, 1976).
Diagenesis in volcaniclastic sequences
The basic diagenetic reactions observed in volcaniclastic sequences are the transformation of glass into
clays (montrnorillonite) or zeolites. In addition, plagioclases may be altered to give clays, and the zeolites themselves may be transformed into other zeolites or into albite. The reactions which give rise to these diagenetic minerals are hydration, carbonization and dehydration.
In the absence of in-depth studies of logs in sequences of this type, it is not possible to indicate methods of
detection.
Serralog 0 2003
Figure 6-52 - Example of hard-ground detection using well logs.
Obsefve the thin resistivity peak at 1740 m, and the small radioactive
peak just above it. It is the equivalent of the hardground illustrated by
Figure 6-51 (from Serra, 1973).
259
Diagenesis in shales
As previously indicated clay minerals can evolve into
other minerals due to interaction between minerals and
fluids by additon or substraction of ions and influence of
temperature and pressure (Table 6-6). The detection of
this type of diagenesis requires the determination of the
depositional environment in order to evaluate the original
clay mineral type. Continental clayey deposits are characterized by irregular log responses due to the depositional process. On the contrary, marine clayey deposits
show generally a homogeneous log response.
ce of organic matter uranium is irreversibly adsorbed from
uranyl solutions in the presence of bacteria and humic
fractions. Organic matter and acidic pH are favourable to
convert U022+into the insoluble quadrivalent ion U 0 2 . In
acidic pH environments, humic and fulvic acids, ethers,
alcohol, aldehydes, favour the precipitation of uranium by
the reduction of U6+ to U4+,forming urano-organic complexes or chelates. Clays also encourage the formation of
schoepite by hydrolysis of the uranyl ions (Fig. 6-54).
Diagenesis of organic material
Organic material present in sediments evolves as a
function of the conditions of oxido-reduction into the depositional environment. Oxidation conditions transform the
organic material into C02. In reducing conditions, the
organic material is transformed and can become a source
rock which can generate biogenic methane, oil, bitumen
or thermal gas as a function of its maturation under temperature and pressure influences (Fig. 6-53).
mDRocARBoN
OENaUTlOtl INTENSITY
J
Figure 6-54 - Diagramatic sketch showing possible association and
time of emplacement of uranium with common constituents of marine
black shales. Uranium is represented by black squares
(from Swanson, 1960).
Several authors, Beers & Goodman, (1944) (Fig. 6-
55),Russel (1945), Swanson (1960) (Fig. 6-56), Hassan
(1973), Supernaw et a/. (1 978) (Fig. 6-57) have observed
a strong correlation between uranium and organic material.
Figure 6-53 - Organic material maturation and formation of biogenic
methane, oil and dry gas (from Sellex 1988).
"Transition zone"
The evaluation of the potential of source rock can be
achieved through the well log analysis.
L
Carboh (%)
Reducing conditions generate transformation of sulphates into sulphides (i.e. pyrite). They are also favorable
to preservation of uranium which will precipitate or accumulate in sediments under certains conditions. In presen260
0
Figure 6-55 - Relation between uranium and organic carbon in sedimentary rocks (from Beers & Goodman, 1944).
Serralog 0 2003
Figure 6-56 - On the left: diagram showing possible relation of uranium
content to total organic matter as controlled by the proporfion of humic
and sapropelic material making up the organic matter. On the right: oil
yield of a marine black shale as a function of the total organic matter
(from Swanson, 1960).
So, after calibration with core data, it is possible to
evaluate the organic carbon content of a source rock from
its uranium content and from that its hydrocarbon potential if the nature of the organic material is known (Fig. 658).
Organic carbon (%)
Figure 6-59 (next page) illustrates a formation very rich
in uranium shown by the spectrometry of the natural
radioactivity. It may correspond to a potential source rock
which could correspond to a mixed layer illite-montmorillonite taking into account the potassium and thorium
content.
Figure 6-57 - Relation between uranium (expressed as a
uranium/potassium ratio) and organic carbon
(from Supernaw et al., 1978).
Figure 6-58 - Sketch showing theoretical distribution of humic and sapropelic materials in a shallow sea in which black muds are accumulating, and
the estimated uranium content and oil yield of the resulting black shale. Increase of total organic matter seaward due chiefly to seaward decrease
in amount of detrital sediment; proportional increase in sapropelic matter seaward due to decrease of humic /and-plant debris and predominance of
planktonic matter (from Swanson, 1960).
Other attempts to evaluate the source-rock potential
include the measurement of the carbon-oxygen ratio
(C/O) by spectrometry of the gamma rays induced by
inelastic collisions of fast neutrons (Fig. 6-60 next page).
This measurement requires either a stationary run or
several runs with very low recording speed.
Serralog 0 2003
261
NATURAL GAMMA RAY SPECTROSCOPY DATA
Figure 6-59 - Example of fromation very rich in uranium as shown by
the natural gamma ray spectrometry data. It may correspond to a
potential source rock (courtesy of Schlumberger).
A
5
5
0
Total Organic Carbon (%)
Figure 6-60 - Evaluation of the organic carbon content through the
measurement of the carbon oxygen ratio (from Herron et al., 1988).
262
References and Bibliography
BATES, R.L., & JACKSON, J.A., eds. (1980). Glossary of Geology. 2nd ed., Amer. Geol. Institute, Falls
Church, Virginia, U.S.A.
BATHURST, R.G.C. (1976). - Carbonate Sediments
and their Diagenesis. 2nd ed. Developments in
Sedimentology, 12, Elsevier, Amsterdam.
BEERS, R.F., & GOODMAN, C. (1944). - Distribution
of radioactivity in ancient sediments. Bull. Geol. SOC.
America, 55.
BISSEL, H.J. (1959). - Silica in sediments of the Upper
Paleozoic of the Cordilleran area. In IRELAND, H.A., ed.,
Silica in Sediments - A symposium. Tulsa, SEPM special
publication 7, p. 150-185.
BLATT, H. (1979). - Diagenetic processes in
Sandstones. SEPM, Special Paper 26.
BLATT, H., MIDDLETON, G., & MURRAY, R. (1980). Origin of Sedimentary Rocks. 2nd ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
CAROZZI, A.V. (Ed.) (1975). - Sedimentary Rocks.
Benchmark Papers in Geology, 15, Dowden, Hutchinson
& Ross, Inc., Stroudsburg, Pennsylvania.
CHILINGARIAN, G.V., & WOLF, K.H. (eds) (1975). Compaction
of
Coarse-Grained
Sediments.
Developments in Sedimentology, 18A & 188, Elsevier,
Amsterdam.
CHOQUETTE, P.W., & PRAY, L.C. (1970). - Geological
nomenclature and classification of porosity in sedimentary carbonates. Bull. amer. Assoc. Petroleum Geol., 54, p.
207-250.
CURTIS, D.M. (1976). - Sedimentary Processes
Diagenesis. SEPM, Reprint series number 1.
DAPPLES, E.C. (1967). - Silica as an Agent in
Diagenesis. In: LARSEN, G. & CHILINGAR, CV., eds. :
Diagenesis in Sediments, p. 91-125, Elsevier,
Amsterdam.
DAPPLES, E.C. (1972). - Some concepts of
Cementation and Lithification of Sandstones. Bull. amer.
Assoc. Petroleum Geol., 56, p. 3-25.
DELHOMME, J.P. (1992). - A quantitative characterization of formation heterogeneities based on borehole
image analysis.
DICKEY, P.A. (1979). - Petroleum Development
Geology. Petroleum Publishing Co., Tulsa.
DUNHAM, R.J. (1962). - Classification of Carbonate
Rocks according to Depositional Texture. Amer. Assoc.
Petroleum Geol., Mem., 1, p. 108-121.
DUNOYER de SEGONZAC, G.D. (1968). - The birth
and development of the concept of diagenesis (18661966). Earth-Science Reviews, 4, 3, p. 153-201.
EVAMY, B.D. (1967). - Dedolomitization and the development of rhombohedra/ pores in limestones. J. sediment. Petrol., 37, 4, p. 1204-1215.
FAIRBRIDGE, R.W. (1967). - Phases of Diagenesis
and Authigenesis. In : LARSEN, G. & CHILINGAR, CV.,
eds. : Diagenesis in Sediments, p. 91-125, Elsevier,
Amsterdam.
FERTL, W.H. (1976). - Abnormal Formation
Serralog 0 2003
Diagenesis
Pressures. Developments in Petroleum Science, 2,
Elsevier, Amsterdam.
FOLK, R.L. (1959). - Practical Petrographic
Classification of Limestones. Bull. amer. Assoc.
Petroleum Geol., 43, p. 1-38.
FOLK, R.L. (1962). - Spectral subdivision of Limestone
Types. Amer. Assoc. Petroleum Geol., Mem. 1, p. 62-84.
FOLK, R.L., & LAND, L.S. (1974). - MgKa ratio and
salinity: two controls over crystallization of dolomite. Bull.
Am. Assoc. Petrol. Geol., 59, p. 60-68.
FRIEDMAN, G.M. (1964). - Early diagenesis and lithification in carbonate sediments. J. sediment. Petrol, 34, p.
777-813.
FRIEDMAN, G.M., & ALI, S.A. (1981). - Diagenesis of
Carbonate Rocks : Cement-Porosity Relationships.
SEPM Reprint series 10.
FRIEDMAN, G.M., & SANDERS, J.E. (1978). Principles of Sedimentology. John Wiley & Sons, New
York.
FUCHTBAUER, H. (1967). - lnfluence of different
types of diagenesis on sandstone porosity. Proc. 7th Wld
Petrol. Cong., Mexico, p. 353-369.
CARRELS, R.M., & MACKENSIE, F.T. (1971). Evolution of Sedimentary rocks. Norton, W W & Co., New
York.
GARY, M., McAFEE, R.Jr., & WOLF, C.L. (1972). Glossary of Geology. Amer. Geol. Institute, Washington,
D. C.
GRIM, R.E. (1958). - Concept ofdiagenesis in argillaceous sediments. Bull. amer. Assoc. Petroleum Geol., 42,
p. 247-253.
HASSAN, M. (1973). - The use of radioelements in
diagenetic studies of shale and carbonates. Int. Symp.
Petrography of Organic Material in Sediments. Centre
Nat. Rech. Scientifique, Paris, September.
HERRON, S., Le TENDRE, L., & DUFOUR, M.
(1988)- Source rock evaluation using geochemical information from wireline logs and cores. Bull. amer. Assoc.
Petroleum Geol., 72, 8.
HERRON, S., PETRICOLA, M., SAPRU, A., DEVRAJAN, N.R., & IYER, R.S. (1994). - Through the reef barrier. Schlumberger, Middle East Well Evaluation Review,
15.
HOBSON, G.D., & TIRATSOO, E.N. (1975). lntroduction to Petroleum Geology. Scientific Press Ltd,
Beaconsfield, England.
KRUMBEIN, W.C. (1942). - Physical and Chemical
Changes in Sediments after Deposition. J. sediment.
Petrol., 12, 3, p. 111-117. Also in S.E.P.M. Reprint Series
number 1, Sedimentary Processes, p. 13-19.
KRUMBEIN, W.C., & SLOSS, L.L. (1963). Stratigraphy and Sedimentation. 2nd ed. W.H. Freeman &
Co., San Francisco.
KRYNINE, P.D. (1948). - The megascopic study and
field classification of sedimentary rocks. J. Geology, 56, p.
130-165.
LANDES, K.K. (1951). - Petroleum Geology. John
Wiley & Sons, New York.
LARSEN, G., & CHILINGAR, G.V. (1967). Serralog Q 2003
I
Chapter6
I
263
Diagenesis in Sediments. Developments in sedimentology, 8, Elsevier, Amsterdam, 551 p.
LEET, L. Don, JUDSON, S., & KAUFFMAN, M.E.
(1978). - Physical Geology. 5th ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
LINK, P.K. (1982). - Basic Petroleum Geology. OGCl
Publications, Tulsa.
LOMBARD, A. (1956). - Geologie Sedimentaire. Les
series marines. Masson, Paris.
LOMBARD, A. (1972). - Series sedimentaires. Genese
- Evolution. Masson, Paris.
LONGMAN (1980). - Carbonate diagenetic textures
from near-surface diagenetic environments. Bull. amer.
Assoc. Petroleum Geol., 64, p. 461-487.
NEASHAM, J.W. (1977). - The morphology of dispersed clays in sandstone reservoirs and its effect on sandstone shaliness, pore space and fluid flow properties. 52nd
Ann. Fall Meet., SPE of AIMS, paper SPE 6858.
PETTIJOHN, F.J. (1975). - Sedimentary Rocks. 3d ed.
Harper & Row, Publishers, New York.
PETTIJOHN, F.J., POTTER, P.E., & SIEVER, R.
(1972). - Sand and Sandstone. Springer, New York.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Book Division, Houston.
POWERS, M.C. (1967). - Fluid release mechanisms in
compacting marine mudrocks and their importance in oil
exploration. Bull. amer. Assoc. Petroleum Geol., 51, p.
1240-1253.
PRAY, L.C., & MURRAY, R.C. (eds) (1965). Dolomitization and Limestone Diagenesis. A symposium.
SEPM, Special Publication 13.
PRESS, F., & SIEVER, R. (1978). - Earth. 2nd ed. W.H.
Freeman & Co, San Francisco.
RIEKE, H.H. Ill, & CHILINGARIAN, G.V. (1974). Compaction of Argillaceous Sediments. Developments in
Sedimentology, 16, Elsevier, Amsterdam.
RUSSELL, W.L. (1945). - Relation of Radioactivity,
Organic content, and sedimentation. Am. Assoc. Pet.
Geol. Bull., 29, 10.
RUSSELL, W.L. (1951). - Principles of Petroleum
Geology. McGraw-Hill Book Co, New York.
SELLEY, R.C. (1976). - An lntroduction to
Sedimentology. Academic Press, London.
SERRA, 0. (1973). - Interpretation geologique des diagraphies differees en series carbonatees. Bull. Centre
Rech. Pau-SNPA, 7, 1, p. 265-284.
SIPPEL, R.F. (1968). - Sandstone petrology, evidence
from luminescence petrography. J. sediment. Petrol., 38,
p. 530-554.
SUPERNAW, I.R., McCoy, A.D., & LINK,A.J. (1978). Method for in-situ evaluation of the source rock potential
of each formation. U.S. patent 4,071,744, Jan. 31.
SWANSON, V.E. (1960). - Oil yield and Uranium
Content of black shales. Geol. Survey, Prof. paper 356-A.
TAYLOR, J.M. (1950). - Pore space reduction in
sandstones. Bull. amer. Assoc. Petroleum Geol., 34, P.
70 1-716.
TRASK, P.D. (1932). - Origin and environment of source sediments of petroleum. Houston.
263
WELLER, J.M. (1960). - Stratigraphic principles and
practice. Harped New York.
WILSON, J.L. (1975). - Carbonate Facies in Geologic
History. Springer, New York.
WOLF, K.L., & CHILINGARIAN, G.V. (1976). Diagenesis of sandstones and compaction. In :
CHILINGARIAN, G. V. & WOLF, K H. (eds): Compaction
of Coarse-grained Sediments, II; Developments in
Sedimentology, 18b, Elsevier, Amsterdam.
ZOBELL, C.E. (1942). - Changes produced by
Microorganisms in Sediments affer Deposition. J. sediment. Petrol., 12, 3, p. 127-136. Also in S.E.P.M. Reprint
Series no 1, Sedimentary Processes, p. 20-29.
264
Serralog Q 2003
K
W
I-
n
a
I
0
WELL LOGGING AND COMPACTION
Introduction
Review of geological concepts
Definition
Compaction is the “reduction in bulk volume or thickness of, or the pore space within, a body of fine-grained
sediments in response to the increasing weight of overlying material that is continually being deposited or to the
pressures resulting from earth movements within the
crust” (Bates & Jackson, 1980).
Generally, compaction is the result of the mechanical
rearrangement of grains. The result is a reduction in volume at the expense of the original void spaces, in other
words, a reduction in the initial porosity.
The principal stress is therefore vertical, and directed
downwards (Fig. 7-1). However, further compression may
occur because the ensuing forces may result from tectonic movements in various directions. This kind of compression only occurs after compaction.
Compaction study
Compaction is particularly marked in detrital or particle-composed sediments. The amount of compaction
depends on the initial porosity and on the size, shape and
sorting of the grains. It also depends on the rate of sedimentation and period of time.
As pointed out by Fertl (1976), the compaction is related to several parameters which are:
(T :
the stress on the system,
the density of the formation,
p:
the porosity of the formation,
@
the permeability of the formation,
k:
D:
the burial depth,
the time since the starting of the burial,
t:
c:
the compressibility relationship,
v:
the velocity parameter for solids and
interstitial fluids in the system,
V:
the volume relationships.
The stress caused by burial is a function of the force
per unit area, and is represented by (T. Its unit is the pascal (= newtonlmz) and its dimension is kg/m/s*.
(T
= force/surface
(7-11
The force (whose unit is the newton) generated by the
sediments is equal to the product of their mass, M, and
the acceleration due to gravity, g. Therefore, eq. (7-1)
becomes:
C
(T
= (M x g) / surface
(7-2)
The mass of the sediments is equal to the product of
their average density (usually expressed in glcm3) and
their volume (surface x height). Thus:
(T
Figure 7-1 - Classification of compaction stresses : (C) polyaxial stresses (pz > py > px) (A) hydrostatic stresses (px = py = pJ; (B) triaxial
uniaxial stresses (the four faces which are
stresses (px = py < pJ; (0)
parallel to stress pz are stationary); (E) biaxial stresses (px = py and
the faces parallel to these two stresses are maintained stationary)
(from Sawabini et al., 1974).
Serraiog Q 2003
= (&
g surface x h) / surface
(7-3)
Or simplifying, we get the total overburden pressure’
(7-4a)
(7-4b)
265
where:
0 = the overburden pressure in pascal2,
-Pb = total average density of the sediment in g/cm3,
g = acceleration due to gravity (m/s2),
h = height (or thickness or depth) in meters.
The hydrostatic component pp is also known as the
interstitial fluid pressure, or pore pressure, and is equal to
the hydrostatic pressure (i.e. the product of average fluid
density and the height of the fluid column), if the fluid is
pure water and the compaction is normal.
0-
For porous formations, P b can be expressed as:
2-
(7-5)
4-
where:
ir, = average density of the fluid in g/cmJ,
pma= average density of the matrix in glcm3, (matrix as
understood by the log analyst, i.e. the total non-fluid part
of the formation),
;b = average porosity of the formations.
Overburden pressure increases with depth, and, as a
first approximation, would be assumed to be uniformly
proportional to depth (Fig. 7-2). However, this is not the
case, since the density of rocks increases with depth.
Assuming a uniform increase, we define an overburden
pressure gradient which is assumed to be equal to 0.231
kg/cmn/m (= 1 psi/ft) for an average density of rocks equal
to 2.31 gkm3 (Fig. 7-3).
For porous formations, by using eqs.(7-2) and (7-3),
the overburden pressure can be broken down into two
components, hydrostatic, pp (pressure of fluid column)
and lithostatic, pe (pressure of sediment column or intergranular pressure). This is represented schematically in
Figure 7-4. Thus,
N.B. In the case of non-porous rocks, pp = 0 and
(J
= pe.
6-
z
8-
5
10-
0
X
v
'
5
o.
1214-
16 18
1
20
0.7
08
a9
I
I
19
to5
Overburden gradient (psilft)
Figure 7-3 - Variation of geostatic pressure gradient with depth.
(1) theoretical; (2) in Texas and Louisiana; (3) in California; (4) in the
North Sea (from Fertl, 7976).
h
(kglcm2 )
1Mx)
.pP
Zoo0
Figure 7-4 - Schematic representation of geostatic pressure and
its components.
3m
Figure 7-2 - Variation of geostatic pressure with depth.
This pressure may exceed the hydrostatic pressure if
the interstitial fluid is subjected to an excess of pressure
due, for example, to tectonic stresses or compaction (in
the case of an undercompacted formation). The hydrostatic pressure gradient is defined as the ratio pp/h, and is
equal to 0.1 kg/cmZ/m for pure water. The ratio of fluid
pressure, pp, to overburden pressure, 0,represented by
A,is minimal in a hydrostatic environment or in the case of
normal compaction.
The pressure is sometimes expressed in psi or kg/cm2.
266
Serralog 0 2003
Compaction
h=p,/o
h = 0.435 psilft (0,l kglcmzlm)
(7-7)
In an undercompacted environment, the fluid supports
some or all of the lithostatic pressure. pp increases while
pe decreases. pp tends towards the geostatic pressure o
and 31, tends towards a value of 1 psi/ft (Fig. 7-5).
Pressure (psi)
I
Chapter7
I
267
These will occupy the volume allowed by the arrangement of coarse grains, and will continue to do so until the
point at which the stress begins to crush the grains.
The composition of the sand also has a part to play.
The porosity of a "clean" sand, that is without shale or
mica, decreases less rapidly with depth than a shaly
sand. This is because, under stress, mica or shale grains,
which are less resistant and more plastic, lose their shape
and crumble, thereby invading the porous space.
Furthermore, the presence of such minerals as amphiboles, pyroxenes, plagioclases, feldspars, etc., encourages
the development of other diagenetic phenomena because
of their chemical instability. Such phenomena will quickly
take over from compaction.
According to measurements carried out by various
researchers (Maxwell & Verral, 1954; Borg & Maxwell,
1956; Maxwell, 1960), well-sorted quartz sands can have
very high porosities (25 to 30 %) at ambient temperatures,
for pressures corresponding to a depth of 13,000 m
(40,000 ft).
Pressure (kg/cm2)
Figure 7-5 - The concept of compaction in subsurface.
Sand compaction
The first stage of the consolidation is a mechanical
rearrangement of the grains 3. During this they roll or slide
over each other easily and rapidly depending on their
shape and sorting, because of the vertical stress exerted
by overlying sediments at the time of burial. This produces
a tighter or more compact arrangement and hence a
reduction of porosity, leading to an increase in the density. Furthermore, as shown by Taylor (1950), this rearrangement will cause the number of lower and total contacts
for each grain to increase (Fig. 7-6).
It can be demonstrated that for a stack of spheres of
the same size the number of lower contacts will increase
from 1 to 3 or 4 (Fig. 7-7). Taylor also established that the
type of contact changes with burial (Fig. 7-6 and 7-8).
From having a tangential shape (a point), the contact
becomes elongated (a ridge), then concavo-convex and
finally sutured (a surface). The change in porosity due to
compaction also depends on sorting. Initially, a poorly sorted sand is less porous than one which is well sorted.
However, the reduction in porosity with depth is less rapid
for a poorly sorted sand. Once the coarse grains have
been rearranged with tangential or elongated contacts by
the displacement of small grains, they will carry most of
the load and thus protect the smaller grains from further
stress.
See Allen 8 Chilingarian, 1975 for more details.
Serralog 0 2003
sdoo
Figure 7-6 - Effect of depth of burial on the type of grain-to-grain
contact in the Jurassic and Cretaceous sandstones in two Wyoming
wells (from Taylor, 1950).
12
4
Figure 7-7 - The six stacking schemes for spherical grains of the same
size which control the porosity and number of contacts
(adapted from Graton & Fraser, 1935).
According to Maxwell (1964), porosity decreases more
rapidly with depth when the temperature gradient increases (Fig. 7-9). For low temperature gradients (7°C per
1,000 ft) and young formations (Miocene), Maxwell recorded, in Louisiana and Texas, a porosity of 20 % at a depth
267
of about 6,000 m. It can be concluded that a high temperature gradient encourages the early development of
other diagenetic phenomena, particularly cementation, at
which time compaction becomes a secondary effect.
No
contact
Point
contact
cult to determine the diagenetic phenomena responsible
for the reduction in porosity on the basis of individual samples. For this, a detailed analysis should be carried out on
thin sections by cathodoluminescence or a scanning electron microscope. It is clear that beyond a certain temperature and pressure which depends on the composition
and texture of the rock, other phenomena begin to take
over from mechanical compaction. For very clean, mature and well-sorted sands, (i.e. free of clay), and for a low
temperature gradient, these secondary effects appear
later, hence at greater depths.
Line
contact
Surface
contact
Figure 7-8 - Variation of the number of lower contacts and of the porosity with depth of burial (adapted from Taylor, 1950).
h
2
s
t!
v
5
Q
5
0
8
8
Porosity (%)
Figure 7-9 - Variation of porosity with depth using various temperature
gradients and ages (from Maxwell, 1964).
Another factor in the reduction of porosity appears to
be the passage of time, which also encourages other diagenetic phenomena. Based on the results of 17,367 porosity measurements made on subsurface cores, Atwater &
Miller, 1965 (in Blatt, 1979) established that porosity
decreases linearly and continuously at depths below 350
m (Fig. 7-10). McCulloh (1967), on the other hand, defined a change of porosity with depth which is not linear.
This study was based on 4,000 porosity measurements
from cores (Fig. 7-11). The contradictory results of these
two researchers seems to be due to their analysis of porosity changes on a statistical basis only, rather than studying the origin of porosity reduction.
In fact, these curves include all the diagenetic effects
associated with burial and not just compaction. It is diffi268
Figure 7-10 - Relationship between depth of burial and porosity, based
on measurements on 17,367 core samples from sandstones of the late
Tertiary in Louisiana. The points represent the average values over
1000-foot intervals (from Atwater & Miller, 1965, unpublished).
The analysis of the effects of burial on sands should be
based on changes in maximum porosity which should be
related to the porosity of "clean", pure, stable quartz
sands (i.e. well-sorted, with no unstable minerals). These
sands will undoubtedly be less influenced by diagenetic
effects other than compaction. Furthermore, considering
the stress exerted on the grains, the porosity seems to
decrease in steps. Each step corresponds, for a given
temperature gradient, to a certain pressure linked with a
resistance to normal or shearing stresses, to collapsing,
or to a change of mechanical behavior. Figure 7-12, taken
from Roberts (1969), shows that there is no modification
of the void ratio4 and hence porosity, until a certain pressure, known as the point of collapse, is reached. In the
same way, the variation in maximum porosity with depth
is seen to be stepped, as measured by Maxwell (1964)
and plotted in Fig. 7-14, other diagenetic effects being
minimal. The following can be concluded:
- The study of compaction in sands must be based on
the maximum porosity.
Serralog 0 2003
Compaction
- The reduction of porosity due to burial does not
decrease proportionally, but in steps.
- Compaction is often accompanied by secondary diagenetic phenomena especially in immature sands.
1
Chapter7
1
269
APPLIED PRESSURE. P S I
Total porosity (%)
APPLIED PRESSURE. K6/CM-'
Figure 7-12 - Relationship between void ratio and applied pressure (1)
in a Rhode Island sand; (2) in a Plum Island sand; (3) in quartz; (4) in
feldspar (from Roberts, 1969).
a0
n
u)
L
0
z
Y
->
m
C
cu
Probable maximum
average porosity of
reservoir sandstones
5
Q
d
&
.-3;;
rn
C
Maximum probable
porosity for most
sedimentary rocks
a
Portion of solid grains (G)
Probable upper limit of
porosity for virtually all
sedimentary rocks
Figure 7-13 - Relationship between porosity, void ratio and percentage
of solids (from Robertson, 1967).
Range of porosities of most
sediments and sedimentary
rocks
h
",
5
P
Figure 7-11 - Relationships between the total porosity of sedimentary
rocks and depths of burial, based on laboratory measurements on over
4000 core samples frdm the basins of Los Angeles and Venture, and
other locations in the US and Italy (from McCulloh, 1967). The curve
through the crossed points is from the Niger Delta.
The void ratio, e, is defined as the ratio of pore volume to solid
volume
e=V+Vasor e = $/(I - $)
(7-8)
Robertson (1967) defined the percentage of solids G as the ratio of
the volume of solids to the total volume (Fig. 7-13)
G=Va/Vb,OUG=l - $
(7-9)
This gives
e = $lG
(7-10)
Serralog 0 2003
0"
Miocene sandstones
Frio (Oligocene)sandstones, Acadia, Allen
Figure 7-14 - Variation of maximum porosity with depth (from Maxwell,
1964). The variation can equally well be plotted in steps.
- At a given depth, a porosity which is below the com269
270
Well logging and Geology
paction line indicates either a decrease in sorting, or the
development of other diagenetic effects (cementation,
authigenesis, etc.), connected with a combined effect of
pressure and temperature and chemical instability of the
minerals present.
- Levels of higher porosity indicate undercompaction
and suggest the presence of zones of high pressure.
Such zones are detectable by an increase in mud pit
levels during drilling, in which case mud weight should be
increased. Higher porosity may also indicate secondary
porosity due to epidiagenesis if the reservoir pressure is
hydrostatic.
Carbonate compaction
One can consider, as Coogan & Manus (1975) did,
that the compaction of carbonates depends on three principal factors - which are also those which act in the compaction of other sediments.
- Inherited factors which are related to the original
composition of the carbonate (particle mineralogy), and to
the texture (grain and crystal size and shape, sorting, packing), consequently to the depositional environment.
- inhibitory factors which inhibit compaction and are
related to physically and biologically induced chemical
changes during preburial lithification or alteration such as
synsedimentary or early cementation and dolomitization.
- Dynamic factors which are related to the depositional, diagenetic and tectonic environments, and overburden pressure, subsurface temperature, duration of burial
stress, pore pressure and pore fluids.
Completely contrary to quartz and shales, the carbonate minerals (fundamentally calcite and dolomite), are
very soluble, their solubility depending greatly on their
purity, the pH and Eh conditions, and the temperature and
pressure.
The initial (depositional) porosity of carbonates is often
very high (between 40 and 80 %). This high porosity
favors fluid circulation if it is associated with a high permeability, which depends on the size of the crystals and
the throats between the pores. Consequently, the interstitial waters are not always in chemical equilibrium with the
surrounding rock. This can create ionic exchanges between the solutions and the minerals composing the rock.
Furthermore, a high dispersion of the size of the particles
and crystals leads to different amounts of dissolving and
recrystallization. All these factors result in early diagenesis which leads either to lithification of the rock by cementation and/or recrystallization, or to a weakening of the
rock due to dissolution. In the first case compaction by
burial will have little effect. In the second, it will become
important and will generate other diagenetic phenomena :
pressure solution, fracturing, stylolitization, brecciation,
etc.
So, carbonate rocks react to burial in different ways
depending on the original facies, and the importance and
type of diagenetic phenomena supported since the depo270
sition.
Other factors must be taken into account
- the rate of sedimentation and subsidence;
- the geothermal regional gradient (i.e proximity of a
volcanic activity, etc.), the temperature effect on the sohbility of calcium carbonate being important;
- the maximum effective stress or overburden pressure;
- length of time of burial;
-the fluid pressure, which, if it is high, will decrease the
effect of compaction;
- the formation fluids, the presence of oil or gas can
considerably decrease the ionic exchanges and consequently the diagenetic phenomena. Oil-wet particle surfaces could substantially reduce the potential for solution of
carbonate minerals and the redeposition of calcite as
cement in the pore space.
In the case of chalks, the initial porosity is very high
(between 70 and 80 %). The size (1 to 20 pm) of the organisms which compose them (coccoliths and foraminifers,
Figure 7-15), and the good sorting, give this type of rock
poor permeability, limiting fluid movement. Furthermore,
the magnesium content is often close to zero, giving them
a high degree of diagenetic stability. The decrease of
porosity with depth is, in its initial phase, fundamentally
linked to compaction. The porosity evolution with depth is
close to that found for sands (Fig. 7-16). But, the compaction is sometimes delayed either when the rate of
sedimentation is high, or when the hydrocarbons have
early occupied the pore space at an early stage, or when
the two phenomena are combined as in the North Sea
example (Fig. 7-16).
In the case of reefs (bioherms), the framework itself of
the recifal body is the reason that these deposits can support an important burial without a significant loss of porosity. Other diagenetic effects (cementation, dolomitization)
can, however, modify their behavior.
_I
Figure 7-15 - SEM photogrzaph of chalk showing the particle size.
In the case of carbonate sands (bioclastic, oolites) the
compaction will depend on the importance and the nature
of the other diagenetic phenomena which themselves will
depend on the depositional and diagenetic environments.
Serralog 0 2003
Compaction
From a general point of view, taking into account all
phenomena, a recent study on carbonates from the south
of Florida, would seem to indicate that the loss of porosity of limestones is more rapid than that of dolomites (Fig.
7-17).
ChaDter7
I
271
Shale compaction
Shales are well-suited for studying compaction because of their high initial porosity and the lower importance of
other diagenetic phenomena. So, a great deal of work has
been done on them, resulting in numerous publicationss.
Of particular interest are those of Rieke & Chilingarian
(1974), and Fertl (1976). The following paragraphs summarize the basic points necessary to understand compaction.
PrinciDal Staaes in Shale ComDaction
Like all other sediments, shales expel their interstitial
water when buried. Due to their significant permeability,
related to their high porosity, there is a considerable period during which shales can potentially expel this water.
North'Sea Coccolith 'Chalk
~
1
I L I
I
I
I
50
10
30
I
20
I
I
10
Porosiw, Dercent
-
Figure 7- 16 Porosity loss with depth in chalk. Gulf Coast Chalk lost
more porosity than normal. This is related to a high geothermal gradient. On the other hand, North Sea Chalk was deposited so quickly
that much more original porosity was preselved (courtesy of R. Nurmi).
Several models of compaction have been proposed.
- For Athy (1930) compaction is simply the process of
expelling interstitial fluids, resulting in a porosity decrease. However, after the deposition and burial of a sediment, its pore volume may be modified by:
- deformation of the grains;
- cementation;
- dissolution;
- recrystallization;
- grain compression.
- For Hedberg (1936) compaction has three stages
(Figs. 7-18 and 7-19).
Stages
Figure 7-18 - Hedberg's compaction model (from Hedberg, 1936).
Porosity, percent
-
Figure 7-17 Comparison of limestone and dolomite porosity evolution
with depth in a south Florida formation. Above 1500 m limestones are
generally more porous, below they are less porous than dolomites
(courtesy of R. Nurmi).
Serralog 0 2003
- Mechanical rearrangement and loss of water
from the shaly mass in an interval between 0 and 800 psi.
For a small pressure variation, the reduction in porosity is
high.
- Mechanical deformation of the particles together
with further expulsion of adsorbed water between 800 and
6,000 psi (0 e 35 %). Some recrystallization of shale particles may occur.
-
A detailed list of references is given for readers interested by the
subject.
271
- Recrystallization with porosities below 10 %.
The reduction of pore volume is slow and only happens
with a large increase in pressure. The larger crystals may
enlarge at the expense of the smaller ones.
- Weller's model (1959), which is very similar to that of
Hedberg, gives the main stages of compaction as follows:
- expulsion of interstitial water until the moment
when the grains come into contact with each other; porosity is reduced from 85 to 45 %;
- grain rearrangement;
- soft shaly minerals are squeezed between the
grains of the more resistant minerals;
- deformation, tending to eliminate all porosity.
1. The expulsion of interstitial water until the grains
come into contact with each other. The porosity in shales
falls from 70-85 % to around 45 %, a reduction which is
attained rapidly over a depth of a few tens of metres.
2. Mechanical rearrangement of grains and continued
expulsion of fluid. Porosity falls to about 25 %, a reduction
which is slower and occurs over several hundred metres.
3. Mechanical deformation of the particles and expulsion of the adsorbed water. Soft minerals sink into the
interstices between harder minerals. The porosity falls
from 35 to 10 %. This reduction is even slower and may
occur over several thousands of metres.
4. The development of major diagenetic phenomena
caused by modification of the physicalchemical equilibrium by the combined action of pressure and temperature, concentration of water and the presence of other
fluids, e.g. gas and oil. This development includes:
- recrystallization,
- cementation,
- dissolution.
ILLlTEAND KAOLlNlTECOMPACTION HSTMlY
Depth of burial (ft)
Figure 7-20 - Compaction history of various types of marine-deposited
shales and its probable relationship to the formation of hydrocarbons
(from Powers, 1967).
Figure 7-19 - Relationship between depth of burial and grain proportion
or porosity, showing Hedberg's compaction stages (from
Baldwin, 1971). (A) Oklahoma (Athy, 1930); ( 0 ) Venezuela (Dallmus,
1958); (DG) Gulf Coast (Dickinson, 1953); (ER1) Santa Barbara Basin
(Emery & Rittenberg, 1952); (ER2) California Coast (Emery &
Rittenberg, 1952); (ER3) Los Angeles Basin (Emery & Rittenberg,
1952); (G) Lake Mead (Gould, 1960); (H) Venezuela (Hedberg, 1936);
(J) Joides, Well 1 (Beall & Fisher, 1969); (K) 15 cores from the abyssal
plane (Kerrnabon et al., 1969) (KH) Venezuela (Kidwell & Hunt,1958);
(M) curves 3, 4, 7, 8, and 10 (Meade, 1966); (RK) New Scotland
(Richards & Kelley, 1962); (SH) compaction test by Skempton
(Hamilton, 1959); (SJ) composite curve by Skeels (Johnson, 1950);
compaction test on blue marine shales (Terzaghi,1925); (W) compaction test by Warner (Beall & Fisher, 1969);(B) average curve by
Baldwin (Baldwin, 1971).
Density = 1.32
Recent burial
Pore water
Interlayer water
Swelling day solids
Non-swelling clay solids
Non-day solids
- Powers (1967) considers that the mineralogical type
of the shale causes compaction to have different effects,
as shown in Figure 7-20.
-Finally, Burst (1969) proposed a compaction model
based on three stages of dehydration (Fig. 7-21).
From these, the stages of compaction can be summarized as follows:
272
Figure 7-21 - Global composition of a marine shale during its dehydration (from Burst, 1969).
Serralog 0 2003
Compaction
ComDaction Mechanisms
Note: The following pages will consider vertical compaction along a
single axis caused by the weight of overlying sediments on shales.
The mechanism of compaction was described by
Terzaghi & Peck (1948) and later by Hottman & Johnson
(1965). To explain this mechanism they created a model
like the one in Figure 7-22, consisting of perforated metal
plates, separated by metal springs and water, all within a
cylinder. The springs simulate the grain-to-grain contacts
between the shaly particles and the plates simulate these
particles. Manometers record the fluid pressure.
Overpressure
STAGEA
Hydrostalic pressure
STAGE B
At stage A of the experiment the valves are closed and
no fluid can be expelled. When pressure is exerted (o),
the springs are not subjected to any pressure (p, = 0), and
the pressure applied is totally counterbalanced by the
equal and opposite pressure of the water (p,) occupying
the volume Vi, and pp = o.
A practical way of recording this pressure is to note the
ratio of fluid pressure, pp and the pressure exerted o
(7-7)
At stage A, h = 1, and the system is overcompressed.
At stage B, the valve is partially open, the plates are
moved downwards (compaction), the volume V1 decreases and the springs transmit part of the applied pressure.
0 = Pp
+
Pe
(7-11)
and h = I .
At stage C, the valve is completely open and enough
water is expelled to allow the springs to reach compaction
equilibrium. From this moment the system is in equilibrium
and no more water is expelled. The water occupies a
volume V, smaller than V, and Vi (compaction). Now, the
applied stress is supported by the springs and the water
which is at the hydrostatic pressure. In such a case, h
approximately equals 0.433.
This model simulates well what actually happens:
- o represents the overburden pressure exerted by the
overlying sediments;
Serralog 0 2003
Chapter7
I 273
- pp is the interstitial fluid pressure at depth Z (formation or hydrostatic pressure);
- p, is the compaction pressure to which the shaly
matrix is subjected at depth Z, or the lithostatic pressure
to which it is subjected, if the proposals of Chapman
(1972) are accepted.
Thus, if the pore fluids can escape freely during subsidence, either towards the surface or by continuous drainage, the volume of original water (porosity) decreases
proportionally with Z, pe is maximum and pp tends to be
hydrostatic. In these circumstances, compaction is said to
be normal. However, if, given the burial conditions, fluids
can only escape with difficulty, the following conditions will
be found:
- slight reduction in the volume of original water, therefore a small porosity decrease with depth;
- p, will be abnormally low;
- pp will be close to the overburden pressure.
Here the compaction is defined as abnormal or undercompaction can be assumed.
The Hubbert & Rubev Law
STAGE C
Figure 7-22 - Schematic representation of shale compaction
(from Terzaghi & Peck, 1948).
h=pp/o
~
It has been established by these two authors that the
effective stress p, exerted on porous shale (or on the
springs mentioned in the experiment) depends on the
degree of shale compaction, and that p, increases proportionally with compaction. A practical guide to the degree of shale compaction is its porosity @,defined as the
ratio of pore volume to the total volume. From this it can
be deduced that, for a given shale, there is for each porosity value 4 a certain maximum value of effective compression stress p, which the shale can support without
further compaction. This is expressed by:
@sh =
(7-12)
with:
Pe = 0 Pp
(7-6b)
where:
@shO = shale porosity at zero burial depth (Z = O),
@sh = shale porosity at burial depth Z,
pe = compaction pressure exerted on the solid matrix
at depth Z,
k = constant,
(T = total pressure exerted on a porous shaly element
at depth Z (i.e. overburden pressure),
pp = fluid pressure.
This equation shows that shale porosity at a given
depth is a function of fluid pressure. If this pressure is
abnormally high the shale porosity will also be abnormally high, as in the case of overpressured shale. In hydrostatic conditions:
o = pbw.g.Z
(7-13)
and
Pp = Pw4.Z
(7-14)
273
where:
pbw = average density of the overlying sediments,
impregnated with water of a density pw in g/cm3.
pW= average density of the fluids in overlying formations in g/cmJ,
g = acceleration due to gravity,
Z = burial depth in meters.
Substituting values for (T and pp in eq. (7-14)
$sh =
(7-15)
with:
(7-16)
c = k (Bbw - Pw)4
c = compaction factor, with dimension L-1 (inverse of
length).
Eq. (7-16) shows that shale porosity varies exponentially with burial depth. So, plotting @sh logarithmically
against Z on an arithmetic scale, eq. (7-16) is shown by a
straight line, characterising normal compaction.
Using measurements from Athy (1930) on shales and
mudstone of the Paleozoic of North Oklahoma, Hubbert &
Rubey established the following coefficients:
These values are close to those formulated by Masse
(1971):
c = 1,14.10-3( m-' )
(7-18)
Os-,O = 48 %
which were established from measurements on 113 shale
samples in 18 wells drilled in 8 basins (Paris, Aquitaine,
Illizi, Gassi-Touil, Gabon and Majunga). They were made
at different burial depths varying between 800 and 5,500
m and ranging from the Gothlandian to the Upper
Cretaceous (Fig. 7-23). The differing values of c can be
explained by the fact that Masse's line is an average,
which integrates data from different basins with different
ages of rocks, while that of Hubbert & Rubey applies only
to the Permian formations of North Oklahoma.
Compaction Effects
Effects on shale porosity
As already seen, the mechanical effect of compaction
is a reduction in volume and therefore in porosity as well
as an increase in density. The connate water is expelled
towards less compact, more permeable areas of drainage.
The highest variations in density and porosity should
occur between 300 and 800 m and the gradient of specific mass of the shales should be between 0.02 and 0.05
g/cm3/100 m.
From Figure 7-23 it appears that during deposition
(mud with almost 85 % water, buried at 800 m, @sh = 20
%) porosity should decrease by 65 %, corresponding to
an expulsion of around 650 litres of water per m3 of sediments.
The change in shale porosity with depth has been studied by various researchers, and their findings are shown
in Figure 7-24. The curves reveal a wide range of differences, possibly due to the very varied composition of the
shale, different amounts of quartz, mica and calcite,
various shaly minerals or a variety of organic matter, and
various geological ages. The analysis of the graph shows
that, in many cases, curves can be more or less superposed by a simple translation. This seems to indicate a more
complex geological history, such as the effects of tectonic
stress, ages, or of a significant erosion of the overlying
sediments, bringing deeply buried rocks closer to the surface. However, this conclusion cannot be reached in the
absence of detailed analyses.
N
Porosity (%)
Porosity (%)
Figure 7-23 - Variation of shale porosity with depth
(from Masse, 1971).
274
Figure 7-24 - Relationship between porosity and maximum depth of
burial of shales and shaley sediments (from various authors:
(1) Proshlyakhov (1980); (2) Meade (1988); (3) Athy (1930); (4) Hosoi
(1983); (5) Hedberg (1936); (8) Dickinson (1953): (7) Magara (1968):
(8) Weller (1959): (9) Ham (1966): (70) Foster & Whalen (1986).
In: Rieke & Chilingarian (1974).
Serralog 0 2003
Next Page
Compaction p h a p t e r 7
Systematic studies (Chilingar & Knight, 1960) and
(Chilingar et a/., 1963) have been made of void ratio in
relation to pressure for pure clays (montmorillonite, illite,
kaolinite, dickite and halloysite), and the results are
shown in Figure 7-25. They indicate a different behavior
for each type of clay mineral.
I 275
0
l0OQ
2000
n 3000
E
W
01
100
10000
Pressure (kglcm’l
5
8 4coo
n
Figure 7-25 - Variations in the porosity of various types of shale with
pressure (from Chilingar & Knight, 1980).
5000
The curves do not seem to follow an exponential law,
but can be taken as such, this having been validated by
an analysis of log parameters over a considerable interval
of depth.
Effects on shale density
The general density of a rock depends on the percentage per unit volume of each component, and its respective density. This is given by:
where:
Pbsh = density of the shale recorded on a density log,
pw = density of the liquid in the shale pores,
pma= density of the shaly matrix or clay minerals,
@ = shale porosity.
Since the density of shale matrix varies between 2.05
(montmorillonite) and 3 (ferriferous chlorite), and that of
the liquid is close to 1, any porosity reduction is indicated
by an increase in total density and vice versa.
If density is proportional to porosity, in normal compaction the density of shale varies exponentially with
burial depth. So, plotting the density of the shales on a
logarithmic scale against depth on a linear scale, the line
given by P b against depth will be straight (at least where
the shale porosity varies exponentially with depth). This is
shown in Fig. 7-26. This is achieved at the wellsite by
measuring the density of shale cuttings and recording
how this density varies with depth. Similarly, as we will
see, the compaction of shales can be studied by following
the changes in density obtained from the density log as a
function of depth.
Serralog 0 2003
6000
5
Density, (g/cni3)
‘igure 7-26 - Variations of shale bulk densities with depth in sedimentary basins. 2 = mudstone of Po Valley Basin (after Storer, 1959);
3 = average Gulf Coast shale densities; values derived from geophysical data (after Dickinson, 1953);
4 = average Gulf Coast shale densities; values derived from density
logs and formation samples (after Eaton, 1969);
5 = Motatan-I - Maracaibo Basin, Venezuela (after Dallmus, 1958);
6 = Gorgeteg no?- Hungary; calculated wet density values
(after Skeels, 1943);
7 = Pennsylvanian and Permian dry shale density values, Oklahoma
and Texas, Athy’s adjusted curve (after Dallmus, 1958);
8 = Las Ollas-I - eastern Venezuela (after Dallmus, 1958).
Effects on the chemistry of interstitial fluids
The following facts are based on various observations:
- the salinity of interstitial waters in normally compacted shales increases with depth,
- the water salinity in undercompacted formations is
generally lower at a given depth, unless there are salt formations nearby.
The work of Engelhardt & Gaida (1963) (Figs. 7-27
and 7-28) shows that the concentration of salt in interstitial water in shales decreases initially with compaction,
but starts to increase once a certain degree of compaction
has been achieved. This phenomenon is explained by the
clay membrane effect. Once a certain degree of compaction has been reached, the clays allow water to pass, but
retain ions selectively (see, for example, the work of
Hitchon, 1964).
275
Previous Page
Porosity
sphere compactedzone
..._
lrovlerrn line8
-Rwmal Rux lines
Qmund surface
"Void ratio"
Figure 7-29- Distribution of thermal flux and temperature profile in the
case of a thermal conductor (from Lewis 8 Rose, 1870).
Figure 7-27- Variation in the concentration of NaCl in the interstitial
water of pores of montmorillonite (from Engelhardt & Gaida, 1983).
Ground surface
Lens insulator (high porosity)
- - _Isotherm lines
Thermal flux lines
Temperature
gradient
-
Temperature
Temperature gradient +
Q)
s
Figure 7-30-Distribution of thermal flux and temperature profile in the
case of a thermal insulator (from Lewis & Rose, 1970).
"Void ratio"
Figure 7-28- Enlarged section of the preceding figure.
Effect on geothermal gradient
The flow of heat from the depths of the earth crosses
the layers of formations within a basin and eventually dissipates at the surface at a rate which depends on climatic
conditions. For a given flow rate, the geothermal gradient
is a function of the thermal conductivities of the intervening formations (Lewis & Rose, 1970).
Any localised bodies of high conductivity, such as a
salt dome, cause a concentration of thermal flux (Fig. 729a) and the geothermal gradient decreases (Fig.7-29b),
while the opposite occurs with bodies of low conductivity
(Fig. 7-30).
In reservoirs which are thick and permeable, convection currents can be established, and this implies good
thermal conductivity. Thus the hydrodynamic renewal of
formation waters dissipates the thermal flux and tends to
diminish the geothermal gradient.
Conversely, the overlying non permeable formations
will appear as insulators, resulting in an increase in thermal gradient. Consequently, in sequences in which the
sandkhale ratio is low (probably indicating undercompacted shales), there will be few thermal conductors and the
geothermal gradient will be high.
276
Effects on mineralogical transformations
Chemical equilibria are affected by the effects of temperature and pressure associated with burial. Elements
and compounds will be subjected to solution, precipitation, cementation and mineralogical changes. We will only
list the well-known transformations
- carbonates : aragonite -+ calcite -. dolomite
- sulfates : gypsum -+ anhydrite
- clays: montmorillonite K illite, montmorillonite -. chlorite, kaolinite -. chlorite, kaolinite -+ illite.
The influence of time
The experiments of Terzaghi & Peck indicate little
influence due to time, the equilibrium position being reached almost immediately due to the high permeability of
the disks (Fig. 7-22). In the case of shales, however, the
permeability can often be very low and the loss of water
can be much slower, hence the time taken to reach equilibrium is longer. The influence of the time factor is illustrated by the diagram in Figure 7-31.
The pressure profile as a function of time will depend
on the rate of sedimentation and type of substratum, as
illustrated in Figure 7-32. It is possible that, on a geological time scale, a similar influence could exist, which would
explain the spread observed in the densities of shales
according to their ages (Dallmus, Fig.7-33).
On the basis of these observations, then, it can be
concluded that the trend of normal compaction will vary
from one basin to another according to the characteristics
of each basin, that is:
- the type of substratum under the shales,
- the rates of sedimentation and subsidence, and
hence the tectonic framework,
Serralog 0 2003
e n n u -
_-1_
Compaction E Chapter7
277
undercompaction (excess porosity) or to formations which
have been deeply buried being raised by extensive erosion, or to the consequences of tectonic movements.
0
1
SP*U
3 - Pliocene to Lower Oligocene
Sand
Figure 7-31 - Schematic representation of the compaction of a shale
and of the influence of time (from Katz & Ibrahim, 1971).
- Recent to Miocene
2 - Pliocene to Lower Miocene
WXK)
4 - Eocene
5 - Readjusted curve of Paleozoic
Figure 7-33 - Normal compaction curves for shales of different ages
(from Dallmus, 1955, in Weeks, L.G.(ed), Habitat of Oil).
Separation line for flow at t 2
Separation line for flow at t,
Permeable substratum
,Impervioussubstratum
Figure 7-32 - Diagrams showing the influence of time and of the type
of substratum on the compaction of shales.
- the composition of the sediments, their porosity and
permeability.
These factors will determine the thickness of sediment
which the water expelled from the shales by compaction
will have to cross to reach porous and permeable rocks.
They also determine the time taken for this migration to
occur, and hence the degree of compaction at a given
instant.
The time taken in fact depends on the permeability
which is a function of the type and grain size of the
sediments. Thus the trend of normal compaction must be
established for each basin and for each period of sedimentation. Every deviation from this relationship between
porosity and depth will correspond either to a state of
Serralog Q 2003
Compaction of organic sediments (peats)
From Weller (1959) the compaction processes of
peats are close to those of shales: water expulsion, followed by a physico-chemical transformation of the sediments. The variation of thickness can reach ratios of 30 to
1 (Renault, 1899; Rukhin, 1953; mentioned by Ryer &
Langer, 1980).
Compaction anomalies
Definition
A compaction anomaly or undercompaction is the
state of a sediment which has been unable to expel its
interstitial water during burial.
Oriain of compaction anomalies
Compaction anomalies are to be found in sequences
which lack porous and permeable escape routes for the
interstitial water of the sediments which should be expelled during compaction. The water remains trapped and
takes on part of the lithostatic overburden. The fluid pressure pp therefore increases, while the sediment itself is
undercompacted.
277
Zones Exhibitina Undercompaction
The phenomenon of undercompaction is found in
basins of rapid detrital sedimentation, poor in permeable
deposits or rich in clays, such as outer deltaic deposits. It
is also observed in basins of mixed detrital and evaporitic
sediments. The map in Figure 7-34 from Fertl (1972)
shows the regions throughout the world in which the phenomenon of undercompaction, or more generally of overpressured formations6 are found.
. increased sonic travel time, At,
. reduced density, Pb,
. reduced resistivity, R,and formation factor, F,
. increased neutron-hydrogen index, QN,
. increased capture cross-section, X
- formation pressure exceeding hydrostatic pressure,
- reduced formation water salinity,
- increased temperature,
- increased geothermal gradient.
If these parameters are plotted against depth, then,
undercompacted formations will exhibit anomalies as
illustrated in Fig. 7-35.
IHN X
-1
figure 7-34 - Countries in which overpressured formations have been
encountered (from Fertl, 1872).
These phenomena are encountered from Cambrian to
Pleistocene, and from depths of a few hundred meters to
about 6,000 m.
Traditional geological approach
Study of compaction is fundamentally based on measurement of the shale cuttings parameters (i.e. density),
collected during the drilling. Drilling parameters are used
to predict penetration into undercompacted or over-pressured formations in order to adjust mud parameters and
avoid blowout (Fertl, 1976).
Well logging approach
Compaction is one of the major diagenetic phenomena in detrital sequences, regardless of composition. It is
generally accompanied by secondary effects which can
be well studied using well logging data.
Associated Phenomena
Undercompacted formations will exhibit the following
features:
- increased porosity, and corresponding effects on
those parameters which depend on it:
~____
Overpressure is not always associated with undercompaction,but
may also be due to artesian pressure, osmotic phenomena, excessive
overburden or tectonic constraints.
278
Figure 7-35 - Schematic responses of the various logs on entry into an
undercompactedshale (completed from Fed/& Timko, 1977).
From the foregoing it is clear that the study of compaction can be undertaken using well logs, in particular
those affected by porosity variations. In broad terms, it
involves examining the variation in log parameters with
depth. The logs of prime concern are the resistivity, density and sonic travel time, and the formations of most
concern are fundamentally shales, but also to some
extent silts and sands, and very occasionally carbonates
(chalk in particular).
Construction of the comDaction Drofile
Manual plot
This method consists of plotting the resistivity, density,
sonic travel time and neutron-hydrogen index of each
pure, homogeneous shale formation on a logarithmic
scale against depth on a linear scale.
Construction of the profile at the wellsite
The wellsite computers present in the truck allows the
realization of a composite log which can be used to select
the shale zones and thereby to obtain the log parameters
of these zones (Fig. 7-36). A compaction profile can also
be generated that plots the sonic travel time on a logarithmic scale (Fig. 7-37).
DENSON automatic program
The DENSON program, written in 1969 by ELF
(Chiarelli et a/., 1973), automatically provides a shale or
even a sand compaction profile on a plotter using log
measurements which have been previously digitized or
recorded on magnetic tape.
Serralog 0 2003
Compaction
The following parameters are plotted against depth, as
shown in Fig. 7-38 in the left-hand track, on linear scales:
- the shale content, or the total natural gamma ray
curve corrected for mud effects,
- optionally, the neutron curve; in the right-hand track,
on logarithmic scales
- the sonic travel time and/or the bulk density of the
shale intervals,
- their resistivity,
- the salinity computed in clean zones from the SP or
in the shale zones by the Pickett's method (1960).
SP
SONIC
I
Chapter7
I 279
- by comparing the At and P b readings with those predicted for a normally-compacted shale at the same depth
based on the trend of normal compaction established for
a given basin or region. Only points having At values
greater than or equal to that of the normal trend, and Pb
below or equal to the normal are plotted (method 2).
Conductivity
Figure 7-37- Compaction profile created automatically at the wellsite
(courtesy of Schlumberger).
Figure 7-39 (next page) shows the profiles obtained by
both methods in the same well.
The logic of the program is illustrated by the simplified
flow-chart in Figure 7-40. Data are entered in the program
from magnetic tape which has either been recorded
directly or which is derived from digitizing optical logs.
Conduct of a study
Figure 7-36- Selection of shale zones and of shale parameters on a
composite log made at the wellsite, and construction of the trend of
normal compaction for detection entry into undercompactedzones
(around 2300 m) (courtesy of Schlumberger).
At the beginning of a regional study, the first method is
applied to several wells in order to establish the trend of
normal compaction. It then becomes possible to apply the
second method to individual wells7.
The shale levels chosen for plotting are selected
according to two criteria:
- by determining shale content using gamma ray , thorium and potassium data, and/or SP, retaining only those
zones showing a shale content (Vs,,) above 80 % (method
11,
In practice, the trend is displaced slightly upward, enabling normally-compacted shales, or values displaced by statistical or lithological
variations which would otherwise have been undetectable, to appear.
Serralog 0 2003
279
Particular care must be exercised in the definition of
the mean radioactive response in the shales, and the
intervals chosen must be representative of a pure shale,
and not of other radioactive formations. If necessary, the
neutron log can be used. It is often necessary to adjust
the response level several times in the same well to take
account of variations in hole diameter and mud type, as
well as changes in shale type. The use of LITHO results
(Delfiner et a/., 1984) can help to select shale intervals
more accurately.
might not be so apparent on a manual plot which would
naturally have a lower density of points.
Figure 7-39 - comparison of compaction profiles obtained by the two
methods applied to the same well logging data set
(Chiarelli et al., 1973).
J
Figure 7-38 - Example of a Compaction profile obtained with ELFAquitaine's DENSON program. At and Pb are used to detect entry into
undercompacted zones (marked by an asterisk).
The main advantages are:
- the speed of obtaining a profile, and the accuracy of
the plot,
- the possibility of correcting responses for borehole
and mud effects by introducing correction charts into the
program,
-the easy detection of levels which are badly caved or
very radioactive thanks to the symbols used,
-the possibility of using only the levels which are most
representative of a pure shale, marked by the symbol X,
and defined from the neutron log, in the computation of
the trend of normal compaction,
- the high density of points (one every six inches),
which allows poor quality points to be eliminated, and
which clearly shows certain geological phenomena which
280
Figure 7-40 - Logic of the DENSON program
(from Chiarelli et al., 1973).
Serralog 0 2003
Compaction
Construction of the normal comDaction trend
Surface seismic
and overlaid VSP trace
Chapter 7
281
Acoustic impedance
For a single well
In the case of a shale, the trend is constructed by
connecting a line through the points corresponding, on the
one hand, to a pure, homogeneous shale, and on the
other, to the hydrostatic pressures of the beds, determined either from pressure measurements in surrounding
reservoirs, or from the density of the mud, used for drilling
the formations, which must be the closest to 1 gkm3 (Fig.
7-41).
In the case of constructing the trend for a sand, "clean"
sands which are at hydrostatic pressure are chosen.
However, several different trends can usually be constructed, each one corresponding to a certain degree of
sorting, since this is what controls the initial porosity.
For several wells
The points for each well satisfying the above-mentioned conditions, that is falling on the normal compaction
line, are plotted together on a semi-logarithmic scale. A
regression line is then estabished for all of these points,
and the equation of this line is used in the second method.
m
11o
msonic transit time (pdft)
Figure 7-42 - Overpressure prediction from a VSP recording
(from Schlumberger, Well Evaluation Conference, Nigeria, 1985).
Figure 7-4 1 - Detection of over pressure using the sonic
(from Schlumberger, Well Evaluation Conference, Algeria, 7979).
Detection of undercompacted zones bv VSP
As previously mentioned, the VSP recording allows us
to see below the bottom of a well. An adapted interpretation of the recorded data, using inversion techniques to
model the boundaries below total depth, enables the
determination of the top of undercompacted or overpressured zones which correspond to a rupture in acoustic
impedance gradient.
Serralog 0 2003
This is possible by the fact that VSP allows the accurate separation between downgoing and upgoing wave
trains.
A layered model below the recording point, with up to
fifty boundaries for a length of trace of about 1 second, is
built such that the incident downgoing wavelet produces
an upgoing synthetic trace as close as possible to the real
one. The choice of these boundaries is done by stepwise
regression. This process insures that we deal mainly with
important contrasts and that we stay relatively immune to
noise. The successive iterative process keeps the initial
position of the boundaries, but computes the contrast
across these boundaries to obtain a synthetic trace which
nearly coincides with the real upgoing trace. The absolute value of the acoustic impedance is known at the recording point from the density and sonic logs, so the series
of impedance contrasts will start from that reference
value. When the change between two successive iterations is less than a pre-determined level the solution is
output (Fig. 7-42). Geological knowledge (presence of
fault, dips) can be introduced at some point of the trace
with a constraint related to the degree of uncertainty of
this knowledge (Fig. 7-43). An important break in the
acoustic impedance gradient can suggest the entry in an
undercompacted or overpressured formation.
Study of sand compaction using well logs
Since burial results in reduced porosity, it is easy to
study this from logs, given the wide range of measurements they provide and the great number of wells drilled
281
282
Well logging and Geology
throughout the world in which density and neutron-hydrogen logs have been recorded. Well logs also provide other
opportunities for analysis, when combining natural radioactivity data (total or selective) with the Pe index measured by the LDT tool, giving an accurate picture of the composition of the sand, without long and costly laboratory
analyses.
radioactive minerals, such as feldspars, mica or shale.
Figure 7-44 shows, as a function of depth, porosity
measurements for clean sands (radioactivity below a certain limit) determined from density logs recorded by ELF
in 26 wells in Nigeria and corresponding to 1875 measurements carried out at depths of between 2,000 and
14,000 ft.
Figure 7-43 - Overpressure prediction from VSP
The upper figure corresponds to a slightly dipping transition to massive
shales. The lower figure corresponds to a case of highly dipping sharp
transition through a fault surface
(from Schlumberger, Well Evaluation Conference, Nigeria, 1985).
Any tool sensitive to porosity (density, sonic transit
time or neutron-hydrogen index) may be used for this purpose. Either the raw values, or those corrected for the
"parasite" effects of the borehole influence, can be plotted
as a function of depth. With density or sonic logs, the
porosity value derived from the raw measurements (Pb or
At) is used, choosing those of quartz as the standard
values of density and sonic travel time (pma and Atma)of
the matrix, as understood by log analysts. In this case, the
calculation is made automatically from the following equations :
2000
$s = (Atf - At) / (A& - Attma)
(7-21)
An upper limit for radioactivity can also be set, excluding from the calculation
and crossplot
high levels
which may correspond to chemically and texturally h m a ture reservoirs. These may include chemically unstable
282
Figure 7-44 - Plot of porosity measurements, based on logs run by
ELF-Aquitaine in 26 wells in Nigeria, as a function of depth. There is a
clear general trend, while the variation in maximum porosity appears to
progress in steps.
Serralog 0 2003
_ _ _ _ _ ~
Compaction F h a p t e r 7 1 283
The porosity is plotted logarithmically. The average
change, given by the middle line, and corresponding to
the line of regression seems to be exponential, and represents the change of porosity with burial, all diagenetic
phenomena being considered. The same graph shows
the change in maximum porosity as well as minimum, the
maximum being defined in clear steps as previously mentioned.
Study of carbonate compaction using well logs
Thus, if the sonic travel time in shales is plotted logarithmically against depth on a linear scale, the points
representing the shales should fall on a straight line. This
is verified by the experimental results shown in Figures. 746 and 7-47.
The sonic, then, provides a means of studying the
compaction of shales as a function of depth. Because it is
largely unaffected by caving, the borehole compensated
sonic measurement provides a better measurement than
the density.
The carbonate compaction can be studied through the
porosity evolution with depth. But, one must remember
that the porosity changes can be related to other phenomena than compaction: cementation, dissolution, pressure-solution (see further).
Figure 7-45 gives an example of compaction study in
a chalk from North Sea.
Figure 7-46 - Variation of At., with depth for Oligocene and Neocene
shales of the Gulf Coast (from Hoffman & Johnson, 1965).
Figure 7-45 - Example of compaction profile in chalk. Observe the two
types of trends. In water bearing formations (lower resistivity zones
with low radioactivity) the average porosity is much lower than in oil
bearing zones (high resistivity).
Study of shale compaction using well logs
Sonic velocitv in shales
It has been widely accepted following the work of
Wyllie et a/. (1956)B that the sonic travel time is more or
less linearly dependent on porosity, and that therefore the
travel time in shales is an exponential function of depth in
the case of normal compaction:
Atsh
= @A&+ ( I -
(7-22)
where:
AtShis the sonic travel time in shales in pseclft,
Atf is the sound travel time in fluids,
Atrnathe sound travel time in the shale matrix, @ is the
porosity of the shales.
In spite of the errors pointed out by Raymer et a/. in the Wyllie
equation in certain conditions, we will continue to use it because it has
been more or less verified for porosities in the range of 10 to 45 %.
Serralog 0 2003
Resistivitv of shales
The resistivity of a rock depends on:
- the conductivity of the constituent minerals (metallic
sulfurs, haematite and graphite are conductive, for example),
- the porosity and the salinities and saturations of the
pore fluids,
- the temperature.
All other factors being equal, decreased porosity will
give increased resistivity. The salinity, however, often
increases with depth and may reduce or even reverse the
trend due to porosity.
Formation factor
Foster & Whalen (1985) proposed a shale formation
factor based on the ratio of the shale resistivity to the
resistivity of the water in the surrounding sands to overcome this difficulty
(7-23)
This formation factor then increases with cornpaction
and depth (Fig. 7-48). This method requires a knowledge
of R,, which is not always possible due to inaccuracies in
determining Rwfrom the SP, or indeed because the SP is
not available. It also assumes, controversially, that the
water in the shales is of the same type as that in the
283
Next Page
284
Well logging and Geology
sands. This method is little used.
I
I
10
I
I
1
b
I
I
.
1
20 3040 6080100
2
Figure 7-48 - Example of variations of Fst, as a function of depth in the
Guff Coast (from Fertl, 1976).
Hvdroaen index of shales
The presence of hydrogen is mainly due to molecules
of water or hydrocarbons. The shales contain molecular
water on the one hand, but also water or hydrocarbons
associated with porosity. It is possible, then, at least theoretically, to observe the reduction in porosity with depth by
studying the behavior of the hydrogen index of the shales.
However, most neutron tools saturate rapidly at porosities
above 50 to 55 %, and in addition they are adversely
affected by caves, which are common in shales. The compensated neutron tools allow this parameter to be used in
the study of compaction because of their increased range
and corrections for the effects of caves. Also, since these
tools are sensitive to gas, which is often present in shales,
the shales can appear to be more compacted than they
really are. For all these reasons, this method is little used.
CaDture cross-section of shales
Timko & Fertl (1970) have suggested this measurement for the study of compaction. The dependence of this
parameter on porosity suggests its usefulness for observing the effects of compaction on porosity, but it must be
remembered that the measurement is also very sensitive
to chlorine, and hence to salinity. This generally varies in
a different way to porosity, and its effect can mask the
effect of porosity.
Figure 7-47 - Trend of normal compaction of shales of a Nigeria well
over an interval of 14000 feet.
284
Natural aamma ray sDectroscoDv loq
Clays are transformed mineralogically under the cornbined effects of temperature and pressure, for example
Serralog 0 2003
Previous Page
Compaction
montmorillonite and kaolinite are changed into illite. This
results in a change in the potassium content and in the
thorium to potassium ratio. The latter should decrease
with depth, being around 10 for montmorillonite and 20 for
kaolinite, but only 4 for illite (Hassan & Hossin, 1975). This
method is not yet widely supported, partly due to a lack of
measurements, but also because this effect is slow to
appear and because illite can appear detritally at any
depth and because the ratio can be affected by other
radioactive minerals such as feldspars, micas or zircon.
1
~~
Chapter7
1
285
marks the entry into sequences where the chances of finding continuous reservoirs is considerably reduced.
Moreover, the problems, and hence the cost of drilling
increase considerably as soon as undercompacted (overpressured) zones are penetrated. It follows that isobath
maps of the upper limit of undercompaction provides useful information about the thickness of the zones which are
most attractive in the search for oil.
Applications of compaction studies
Hydrodynamic applications
Detection of entrv into an undercompacted formation
Note: Obviously, with the exception of VSP interpretation below the
bottom of the well, this analysis can only be applied during the drilling
through LWD measurement, or retrospectively since the wireline logs
are not available until the formation has been drilled.
Entry into an undercompacted formation will be marked by a shift of points relative to the trend of normal compaction already established. The displacements depend
partially on the logs used (see the diagrams of Fig. 7-35),
and partially on the degree of undercompaction (compare
the five profiles in Fig. 7-49).
J
Figure 7-50- lsobath map of the entry into undercompactedformations
(from Chiarelli et al., 1973).
The detection of the entry into undercompacted zones
using the logs allows certain interpretations related to the
existence of high-pressure formations, detected by well
site measurements and wrongly attributed to undercompaction, to be corrected (Fig. 7-51).
Evaluation of formation Dressures
Top of undrrumpldlon
Figure 7-49- Identification of undercompactedzones. Studying the
profiles enables the point of entry into the undercompactedshales to
be pinpointed, and gives an idea of the degree of undercompaction
(from Chiarelli et al., 1973).
The boundary at which the undercompaction appears
can be located very precisely. lsobath maps showing the
entry into undercompacted formations can be constructed
by applying this method to several wells in a region (Figs.
7-49 and 7-50). Such maps can be useful in designing
drilling programmes (casing points, selection of casing
and mud types) in future wells to be drilled in the region.
The appearance of undercompacted zones usually
Serralog 0 2003
Theoretically, the study of the degree of undercompaction of the shale should help in the evaluation of the
pressures of the surrounding reservoirs.
Let A be a point in the undercompacted zone at a
depth ZA (Fig. 7-52). The shale here suffers the same
compaction as the point E, at depth ZE.
According to eq. (7-6),
in A we have:
GA = (Pp)A -k @e)A
(7-24)
in E we have:
(7-25)
OE = (Pp)E
(Pe)E
Since the shales at A and E have suffered the same
degree of compaction, ( P ~=)(p,JE9.
~
+
In practice we assume that for equal porosities, the pressure supported by the matrix is the same. This is the basis of the model of
Terzaghi & Peck used here.
285
R
At E the shales are normally-compacted, and the
hydrostatic pressure ( P ~is) written:
~
where pp is in kg/cm3, Z, is in meters and pw is the
density of the fluid in g/cm3. We also havelo
<TE = PbwE
(zE/lo)
(7-27)
where :
PbwE is the global mean density of the sediments in the
interval surface-Z,.
Likewise, we can write:
OA
=
PbwA
(7-28)
(zA/l O)
If the type of shale does not vary, we can take PbwA as
equal to &,WE as a first approximation, since the sediment
porosity is the same at Z, and ZA. We can then write:
OA
Figure 7-51 - Example of a pressure anomaly due to gas in a zone of
normal compaction (from S e r a et al., 1975).
= PbwE (zA/lo)
(7-29)
) ~ their values in eq. (7-24)
Replacing bAand ( P ~with
gives:
or again
Nota: If pw and Pbw are not known, the following values can be used
pw = I,O g/crn3 et Pbw = 2,31 g/crn3.
ZE
The real case, illustrated in Fig. 7-53, shows that such
an evaluation is practicable.
In any case, in practice a very good approximation of
the trend of normal compaction tied to the shaly facies
studied is necessary in order to be exact. This poses a
serious difficulty.
lo
More precisely, we have on the one hand
(7-27b)
-1
and on the other
(7-28b)
or again
21
.A=&(
DEPTH
Figure 7-52- Estimating formation pressure from the compaction profile
[theoretical example).
286
cJZ)dZ+
~'~JZ)dZ)
(7-28~)
ZE
which gives the following expression by replacing OA and ( p e ) ~by their value
(7-31b)
Serralog 0 2003
~-~
Compaction
~
~~
Chapter 7
287
Certain authors (Hottman & Johnson, 1965; Ham,
1966), propose charts (Fig. 7-55 cf. next page) or nomograms (Fig. 7-56), which allow an empirical determination
of the pressure from the comparison of the shale resistivity (Rs&,bs, or the sonic travel time (Afsh),,bs, observed in
undercompacted shale intervals with the resistivity (Rsh)",
or the sonic travel time
defined at the same depth
for normally-compacted shales. These charts or nomograms are only valid within the region for which they have
been established.
Evaluation of Dressure aradients in massive. undercompacted shales
As soon as we are capable of estimating the interstitial
water pressure within a massive shale, the most probable
direction and value of the pressure gradient can be determined. Figure 7-57 represents a profile in an undercompacted zone where the points corresponding to the same
facies are distributed about a line which is practically vertical. This signifies that the interstitial water is subjected to
a geostatic gradient directed from bottom to top. This phenomenon is also detected on the profile in Figure 7-54.
It goes without saying that such a considerable gradient can influence the migration and hence the distribution of hydrocarbons. Such a distribulion of pressure reduces the value of any reservoir whose source rock is situated on top of it. This seems to be the case for the well
represented by Fig. 7-58, which, having penetrated a
sequence very rich in hydrocarbon shows, then entered a
water-bearing reservoir.
Figure 7-53 - Evaluation of formation pressure. The computed pressure is 470 kdcm' at 2925 m, while the pressure measured
during a test on the reservoir above is 455.1 kg/cm' at 2800 m
(from Chiarelli et al., 1973).
The three plots of Figure 7-54 (see next page) show,
nonetheless, that it is always possible to evaluate the
amplitude of the anomalies, which represents, in spite of
all, considerable progress.
Serralog 0 2003
Figure 7-58 - Compaction profile showing a steep pressure gradient,
but also showing variations in shale composition
(from Chiarelli et al., 1973).
287
288
Well logging and Geology
slight undercompaction
average undercompaction
high undercompaction
Figure 7-54 - Indication of the degree of undercompaction by a shift in the trend of normal compaction (from Chiarelli et al., 1973).
Figure. 7-55- Charts relating the fluid pressure gradient to R,. and At,,,. The fluid pressure is equal to p,r - FPG . ZAN where F f G is the pressure
gradient of the fluid from the chart (from Hottman & Johnson, 1985).
X
c.
al
P
Shale resistivity (ohm-m)
Acwsli impedance (gem Ip%)
Figure 7-56 - Nomograms for estimating formation pressure from resistivity or sonic travel time, established for the Gulf Coast
(from Hottman & Johnson, 1965).
288
Serralog 0 2003
Compaction
Figure 7-57 - Evaluation of pressure gradients. The alignment of the
shale points along a vertical indicates that the interstitial fluid is subjected to a geostatic pressure gradient (from Chiarelli et al., 1973).
Geological Applications
Detection of unconformities
Unconformities sometimes show up clearly on compaction profiles. In fact, they are marked, generally, by a
sudden shift in the trend of normal compaction (Fig. 7-59).
This shift undoubtedly indicates several phenomena:
- change of shale type,
-the strong influence of time elapsed since deposition,
- probable erosion.
[ Chapter7 I 289
I
I
Reconstruction of maximum burial deDths
If the log responses of normally-compacted shales of
the same geological period from different wells are plotted
together, it is frequently observed that the points from the
various well do not fall exactly on the same line, but rather
on a set of parallel lines, one for each well (Fig. 7-60).
This generally indicates a structural uplift of the whole
interval, but by different amounts for the different wells, so
long as we accept that compaction is an irreversible phenomenon.
Using the trend of well (1) for which the actual depth is
the greatest for a given value of At, the maximum apparent depth reached by each of the different shales (A, B,
C or D) in the interval can be determined. Thus, the shales C of well (4), currently between 2100 and 2250 m have
been buried to at least 3100-3250 m, according to the
trend for well (1). They have therefore been uplifted again
by about-I000 m.
This application thus provides a relatively precise
palaeogeographic reconstruction in basins which have
been subjected to deep erosion. The reader is referred to
the article by Lang (1978) which provides a good example.
These studies can also have geochemical applications
by providing an indication of the degree of diagenetic
change and the maturation of organic matter in relation to
maximum depths of sediment burial (Fig. 7-61). Through
this type of analysis, potential gas or oil provinces can be
evaluated.
However, these methods can only be applied to basins
which have not suffered major tectonic constraints.
S
F
Figure 7-59 - Three examples of unconformities detected using compaction profiles
(S = Serra, 1972; C = Chiarelli et al., 1973; F =Feel, 1978).
C
Serralog 0 2003
289
Porosity %
oil-filled sandstones
o water-filled sandstones
\ present depth
maximum depth of burial
/ maximum depth of burial
before oil accumulation
-
Figure 7-60 - Determination of the maximum depth of burial. (a) method of determination; (b) application example (from Philipp et al., 1963).
Correlations between wells
As the recent work of Lang (1984) pointed out, these
compaction profiles (Fig. 7-63) can be useful for establishing correlations between wells. They have the advantage of presenting the log data in a compressed form (due
to the scale used), thus emphasizing the major phenomena such as erosion, condensation, stable periods, etc.
Erosion
Deposition
Differential
compaction
Figure 7-61 - Maturation of the organic matter as a function of the
depth of burial and the temperature.
Differential comDaction
Since the loss of porosity with burial differs between
sands and shales, the change in volume will not be the
same. This leads to changes in the shapes of bodies as
illustrated by the diagrams in Fig. 7-62.
290
Figure 7-62 - Schematic showing the effects of differential compaction
on the shapes of sedimentary bodies.
Sedimentoloaical amlications
The strong influence of lithology on sonic velocity and
resistivity presented certain difficulties, and indeed cauSerralog 0 2003
Compaction
I Chapter 7 I 291
f 0 P STRATIORAPWC SECTION
WfPALtO DATA
Figure 7-63 - Correlations based on compaction profiles (from Lang, 1984).
sed some errors in the first interpretations of compaction
profiles. The measured variations are, in fact, more frequently related to changes in the type of lithology rather
than porosity variations. The variations can be quite
abrupt, as the example in Figure 7-58 of three different
shale facies illustrates. The slowest facies in particular
corresponds to a sediment rich in organic matter, which
explains its acoustic properties. Equally, the variations
can be progressive, due to a gradual enrichment in silts,
carbonates, etc., as in Fig. 7-64, rather than a change
from a state of normal compaction to one of undercompaction.
votl
v.h
Figure 7-64b - Other example of sequential variations from shale to
carbonate detected from compaction profiles
(from Chiarelli et al., 1973).
Figure 7-64a - Detection of sequential variations from compaction profiles (from Chiarelli et al., 1973).
Serralog 0 2003
Furthermore, the variation can sometimes be so gradual and extend over such a long interval (Fig. 7-65) as to
render it practically undetectable by a simple examination
of the logs.hi^ can lead to
of interpretationboth in
determining the trend of normal compaction and in
291
(7-32)
Assuming pma= 2.65 gkm3 and pf = 1 glcm3, the density is converted to porosity, and the following relationship
is obtained:
(7-33)
if Z is expressed in feet, and
(7-34)
with Z in meters.
Note : This statistical law is only valid for normally-compacted
sands (Fig.7-66) which have not suffered diagenetic effects (other than
those associated with compaction), and only in the interval 2000 - 14000
feet in Nigeria.
J
Figure 7-65- Sequential variations over 1700 m clearly shown by the
compaction profile (from Serra et al., 1975).
detecting undercompacted zones. Compaction trends will
therefore be used as a sequential tool providing a first
sketch or compensating for the deficiencies of the gamma
ray, and in certain cases correcting the trend it provides.
Application to Reservoir Studies
Variation of porositv with deDth
Several authors (i.e. Teodorovich & Chernov, 1968)
have proposed formula to express the variation of porosity and permeability with depth. For instance, for sandstones we have:
and
Z being the depth in meters.
A study of the density variations of sands using the
DENSON program in twenty-six wells, corresponding to
1875 density-depth pairs distributed between 2000 and
14000 feet, yielded the following relationship:
292
J
Figure 7-66- Laws of variation of density, sonic travel time and
porosity of sands and shales as a function of depth.
Geophysical Applications
A knowledge of the variations in density and sonic
velocity with depth of both shales and sands or silts can
be of major interest to the geophysicist. It can result in a
better interpretation of seismic profiles in the transformation of isochrones to isobaths, in the computation of
reflection coefficients and in the estimation of sand-shale
percentages. This is how data from numerous wells in
Nigeria were used by the DENSON program to provide
the curves in Fig. 7-66, which are represented by the following equations:
( pb) sd=1.95e
5.2 ~x10-5i3m)
(7-32)
Serralog 0 2003
Compaction
(7-36)
(7-37)
(At)
1.87x104Z(m)
=200e-
(7-38)
Sh
Note : The above relationships have been derived statistically from
measurements in a depth interval of 2000 - 14000 feet, and have only
been verified within this interval.
Chapter7
, 293
The second point concerns the formation of hydrocarbons in undercompacted sequences. Is the expulsion of
hydrocarbons from undercompacted source rocks always
sufficient for the creation of accumulations of commercial
interest? To what extent can a lack of expulsion inhibit the
normal functioning of a source rock? What exactly is the
effect of overpressure on the maturation process of organic matter? Such are the questions to which it would be
interesting to have answers; but these answers - some of
which are only now becoming clear -, will only have credibility if they are based on a great number of observations.
This is why the use of well logs must provide the basis of
the answers.
Decompaction
Geochemical Applications
This subject has been touched upon in the preceding
paragraphs in the discussions of detailed studies of shaley sediments, detection of zones rich in organic matter as
well as the reconstruction of maximum burial depths.
These are all questions which are of some interest to the
geochemist.
However, certain other points have not yet been mentioned. The first concerns the detection of gas zones in
massive shales.
Figure 7-67 speaks for itself. It shows without any
ambiguity the strong effect of gaseous hydrocarbons on
the resistivity. This property, which is a nuisance in the
study of shale compaction using only resistivity logs,
becomes of interest in qualitative studies of the distribution of gas shows within shales when other logs are available.
Obiectives
A decompaction procedure is necessary if a reconstruction of the arrangement of the different sedimentary
bodies as it should be at the time of deposition is required.
In that case it is important to apply to each body an appropriate decompaction coefficient. For that, the compaction
rate supported by each individual sedimentary body must
be taken into account. This compaction rate takes into
account the mineralogical composition, the texture, the
thickness, the environment of deposition and the maximal
depth of burial of each body. Figure 7-68 shows the chart
for estimating the compaction rate of shales as a function
of depth of burial and thickness. The law of decompaction
for a given lithology is generally deduced from the compaction law experimentally established for the same lithology from porosity measurements on rock samples or
from log data.
+
Figure 7-67 - Detection of gas zones in a massive shale
(from Chiarelli et al., 1973).
Serralog o 2003
Shale thkkness In m
Figure 7-68 - Chad giving the coefficient of decompaction, z, for pure
shale as a function of depth and thickness.
293
Procedures
They are defined by Brown, (1975) and illustrated by
Figure 7-69. They consist, firstly of correcting the apparent thickness of each sedimentary body, fundamentally
shales and peats, for overburden compaction; secondly,
of fitting the geometry to a subjacent palaeosurface or
assumed reference datum.
I
U
n
Figure 7-69 - Decompaction procedures (from Brown, 1975).
ADDlications
This decompaction technique allows reconstruction of
the palaeotopography at the end of deposition. It helps to
establish both chronostratigraphicand facies correlations.
References and Bibliography
ALLEN, D.R., & CHILINGARIAN, G.V. (1975). Mechanics of Sand Compaction. In : Compaction of
Coarse-Grained Sediments, I, (edited by Chilingarian,
G.V, & Wolf, K.H.), Developments in Sedimentology, 18A,
Elsevier, Amsterdam.
ASQUITH, G.B. (1982). - Basic Well Log Analysis for
Geologists. Amer. Assoc. Petroleum GeoL, Methods in
Exploration Series.
ATHY, L.F. (1930). - Density, porosity and compaction
ofsedimentary rocks. Bull. Amer. Assoc. Petroleum Geol.,
14, p. 1-24.
ATWATER, G.I., & MILLER, E.E. (1965). - The effect of
decrease in porosity with depth on future developemnt of
oil and gas reserves in South Louisiana. Amer. Assoc.
Petroleum Geol., Progr. Ann. Meet., New Orleans, p. 48.
BALDWIN, B. (1971). - Ways of deciphering compacted sediments. J. sediment. Petrol., 41, p. 293-301.
BATES, R.L. & JACKSON, J.A. (1980). - Glossary of
Geology. Amer. Geol. Institute, Falls Church, Virginia.
BLATT, H. (1979). - Diagenetic processes in sandstones. in: Aspects of Diagenesis. (edited by Scholle, P.A., &
Schluger, P.R.), Spec. Publs SEPM, 26, p. 141-157.
294
BLATT, H., MIDDLETON, G., & MURRAY, R. (1972,
1980). - Origin of Sedimentary Rocks. 1st and 2nd ed.
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
BOATMANN, W.A. (1967). - Measuring and using
shale density to aid in drilling wells in high pressured
areas. J. Petr. Technology, Nov., 19, 11, p. 1423-1430.
BOND, L.O.,ALGER, R.P., & SCHMIDT, A.W. (1971).
-Well log Applications in Coal Mining and Rock
Mechanics. Trans. SME, 250.
BOLT, D.B. (1972). - Detection of overpressured formations during drilling operations in worldwide environments. Ann. Sw. Petr., Short Course Ass. Mtg Proc.,
Lubbock, Texas, 19-20-21 May.
BOREL, W.J., & LEQIS, R.L. (1969). - Ways to detect
abnormal formation pressures. Part I geopressure prediction. Part 11: geopressure detection whil edrilling. Part 111 :
surface shale resistivity. Part IV : geopressure detection
procedures. Petrol. Engineer, July to Nov.
BORG, L.Y., & MAXWELL, J.C. (1956). - Interpretation
of fabrics of experimentally deformed sands. Amer. J. Sci.,
254, p. 71-81.
BROWN, L.F.Jr. (1975). - Role of sediment compaction in determining geometry and distribution of fluvial and
deltaic sandstones. In : Compaction of Coarse-grained
Sediments, I, (edited by Chilingarian, G.V., & Wolf, K.H.),
Developments in Sedimentology, 18A, Elsevier,
Amsterdam.
BURST, J.F. (1969). - Diagenesis of Gulf Coast clayey
sediments and its possible relation to petroleum migration. Bull. Amer. Assoc. Petroleum Geol., 52, 3.
CHAPMAN, R.E. (1972). - Clays with abnormal interstitial fluid pressures. Bull. Amer. Assoc. Petroleum Geol.,
56, 4, p. 7910-795.
CHIARELLI, A., SERRA, O., GRAS, C., MASSE, P., &
TISON, J. (1973). - Etude Automatique de la sous-compaction des argiles par diagraphies differees. Rev. Inst.
franp. PeGtrole, 28, 1, p. 19-36.
CHILINGAR, G.V., & KNIGHT, L. (1960). Relationship between pressure and moisture content of
kaolinite, illite and montmorillonite clays. Bull. Amer.
Assoc. Petroleum Geol., 44, p. 101-1016.
CHILINGAR, G.V., RIEKE, Ill, H.H., & ROBERTSON,
J.O. Jr. (1963). - Relationship between high overburden
pressure and moisture content of halloysite and dickite
clays. Bull. Geol. SOC.Amer., 74, p. 1041-1048.
CHILINGARIAN, G.V., & WOLF, K.H. (Eds) (1975). Compaction
of
Coarse-Grained
Sediments.
Developments in Sedimentology, 18A & 18B , Elsevier,
Amsterdam.
COOGAN, A.H., & MANUS, R.W. (1975). Compaction and Diagenesis of Carbonate Sands. In :
Compaction of Coarse-grained Sediments, 1, (edited by
Chilingarian, G.V, & Wolf, K.H.), Developments in
Sedimentology, 18A, Elsevier, Amsterdam.
COSTLEY, R.D. (1964). - Abnormal pressure determination. LCCOS, 2, 217, 14 Jan.
CURTIS, D.M. (1976). - Sedimentary Processes
Diagenesis. SEPM, Reprint series 1.
DALMUSS, K.F. (1958). - Mechanics of basin evoluSerralog 0 2003
I 295
~
Compaction
tion and its relation to the habitat of oil in the basin. In:
Habitat of Oil (edited by Weeks, L.G.), Amer Assoc.
Petroleum Geot, Mem. 36, p. 2071-2124.
DELFINER, P., PEYRET, O., & SERRA, 0. (1984). Automatic determination of Lithology from Well Logs. 59th
Ann. Techn. Conf. SPE of AIME, Houston, Texas; paper
SPE 13290.
DELHOMME, J.P., & SERRA, 0. (1984). - Dipmeterderived Logs for Sedimentological Analysis. SPWLA, 9th
Europ. Intern. Format. Eval. Trans., paper 50.
DICKINSON, G. (1951). - Geological aspects of
abnormal reservoir pressures in the Gulf Coast region of
Louisiana, U.S.A. Proc. 3th Wld Petrol. Cong., section 1,
p. 1-16.
DICKINSON, G. (1951). - Geological aspects of abnormal reservoir pressures in Gulf Coast Louisiana. Bull.
Amer. Assoc. Petroleum Geol., 37, p. 410-432.
DICKINSON, G. (1953). - Reservoir pressures in Gulf
Coast Louisiana. Bul. Amer. Assoc. Petroleum Geol., 37,
2, p. 410-432.
EATON, B.A. (1972). - The effect of overburden stress
on geopressure prediction from well logs. J. Petrol.
Technol., 24, p. 929-930.
ENGELHARDT, W.V. von, & GAIDA, K.H. (1963). Concentration Changes of Pore Solution During
Compaction of Clay Sediments. J. sedim. Petrology, 33,
4.
FAIT, I. (1957). - Effect of overburden and reservoir
pressure on electric logging formation factor. Bull. Amer.
Assoc. Petroleum Geol., 41, 11, p. 2456-2466.
FERTL, W.H. (1972). - World wide occurrence of
abnormal formation pressure. Part 1. SPE of AIME
Abnormal Subsurface Pressure Symp., Baton Rouge, La.,
15-16 mat 1972, prepint, paper SPE 3844, p. 29-42.
FERTL, W.H. (1976). - Abnormal Formation
Pressures. Developments in Petroleum Science, 2,
Elsevier, Amsterdam.
FERTL, W.H., & TIMKO, D.J. (1970). - Occurrence and
significance of abnormal pressure formations. Oil and
Gas J., 5 Jan. 1970, 68, 1.
FERTL, W.H., & TIMKO, D.J. (1970). - How abnormal
pressure detection techniques are applied. Oil and Gas J.,
12 Jan. 1970, 68, 2.
FERTL, W.H., & TIMKO, D.J. (1970). - Association of
salinity variations and geopressures in soft and hard
rocks. SPWLA, IlthAnn. Log. Symp. Trans., paper C.
FERTL, W.H., & TIMKO, D.J. (1970). - Occurrence of
cemented reservoir roof rock and geopressure cap rock
and its implication in Petroleum Geology and
Geohydrology. SPE of AIME Fall meeting, Houston, oct
1970, paper SPE 3085.
FERTL, W.H., & TIMKO, D.J. (1971). - Parameters for
identification of overpressured formations. Conf. on
drilling and rock mechanics, Austin, Texas, 5-6 Jan.,
paper SPE 3223.
FERTL, W.H., & TIMKO, D.J. (1971). - Utilisation des
diagraphies en recherche, forage et production d’hydrocarbures pour Etude des problemes de pressions geostatiques. Rev. Inst. franp. Petrole, 26, 9, p. 687-714.
Serralog 0 2003
Chapter7
_
FERTL, W.H., & TIMKO, D.J. (1971). - Salinity plots
evaluate subsurface formations. World Oil, 173, 1, p. 93100.
FERTL, W.H., & TIMKO, D.J. (1972). - How downhole
temperatures, pressures affect drilling. World Oil, Jun.Sept., part. 1 to 4.
FINCH, W.C. (1969). - Abnormal pressure in the
Antelope field, North Dakota. J. Petrol. Technol., 21, p.
821-826.
FONS, L. Sr. (1969). - Geological application of well
logs. SPWLA, 10thAnn. Log. Symp. Trans.
FOSTER, J.B. & WHALEN, H.E. (1965). - Estimation
of formation pressures from electrical surveys, offshore
Louisiana. 40th Ann. Fall Meet. SPE of AIME, and in J.
Petrol. Technol., Feb. 1966,18, p. 165-171.
FOWLER, W.A. (1968). - Pressures, hydrocarbon
accumulation and salinities, Chocolate Bayou fields,
Brazoria County, Texas. 43rd Ann. fall Meet. SPE of AIME,
Houston, paper SPE 2226, and in J. Petrol. Technol., 22,
p. 411-423.
GARY, M., McAFEE, R.Jr., & WOLF, C.L. (1972). Glossary of Geology. Amer. Geol. Institute, Washington,
D.C.
GRATON, L.C., & FRASER, H.J. (1935). - Systematic
packing of spheres with particular relation to porosity and
permeability. J. Geol., 43, p. 785-909.
GRETENER, P.E. (1969). - fluid pressure in porous
media, its importance in geology. A review. Bull. Canad.
Petroleum Geol., 17, p. 255-295.
HAM, H.H. (1966). - A method of estimating formation
pressure from Gulf Coast well logs. Gulf Coast Assoc.
geol. SOC.,16, p. 185-197.
HASSAN, M., & HOSSIN, A. (1975). - Contribution a
Etude des comportements du thorium et du potassium
dans les roches sedimentaires. C. R. Acad. Sci. (Paris), D
280, p. 533-535.
HEDBERG, H.D. (1936). - Gravitational compaction of
clays and shales. Amer. J. Sci., 5th series, 31, 184, p. 241287.
HILL, G.A., COLBURN, W.A., & KNIGHT, J.W. (1961).
- Reducing oil-finding costs by use of hydrodynamic evaluation. In: Economics of Petroleum Exploration,
Development and Property Evaluation. Prentice-Hall,
Englewood Cliffs, New Jersey.
HITCHON, B. (1964). - Formation fluids. In: Geological
History of Western Canada.Alberta SOC.Petroleum Geot,
Calgary, p. 201-217.
HOSOI, H. (1963). - First migration of petroleum in
Akita and Yamagata Prefectures. Jap. Assoc. Miner.,
Petrol., Econ. Geol., 49 , 2, and 3, p. 43-55 and 101-114.
HOTTMANN, C.E., & JOHNSON, R.K. (1965). Estimation of formation pressures from log derived shale
properties. J. Petrol. TechnoL,I 7, 6, p. 718-722.
Houston Geol. SOC. (1971). - Abnormal subsurface
pressure. A study group report 7969-7977. Houston
Geo% SOC.,p. 1-92.
HUBBERT, M.K., & RUBEY, W.W. (1959). - Role of
fluid pressure in mechanics of overthrust faulting. Bull.
geol. SOC.Amer., 70, 2, p. 115-166.
295
_
JONES, P.H. (1969). - Hydrodynamics of geopressure
in the northern Gul fof Mexico Basin. J. Petrol. Technol.,
21, p. 803-810.
KATZ, D.L. & IBRAHIM, M.A. (1971). - Threshold
displacement pressure considerations for caprocks of
abnormal-pressure reservoirs. 5th Conf. on Drilling and
Rock Mechanics, Austin, Texas, SPE 3222.
KERR, P.E., & BARRINGTON, J. (1961). - Clays of
deep shale zone, Caillo ulsland, La. Bull. Amer. Assoc.
Petroleum Geol., 45, p. 1698-1712.
KRUMBEIN, W.C. (1942). - Physical and Chemical
Changes in Sediments after Deposition. J. sediment.
Petrol., 12, p. 111-117.
LANG, W.H.Jr. (1978). - The Determination of Prior
Depth of Burial (Uplift and Erosion) Using Interval Transit
Time. SPWLA, 19thAnn. Log. Symp. Trans., paper B.
LANG, W.H.Jr. (1984). - Conductivity and interval
transit time as correlation tools. The Log Analyst, 25, 3, p.
21-33.
LEWIS, C.R., & ROSE, S.C. (1970). - A theory relating
high temperatures and overpressures. J. Petrol. Technol.,
22, p. 11-16.
MAGARA, K. (1968). - Compaction and migration of
fluids in Miocene mudstone, Nagaoka Plain, Japan. Bull.
Amer. Assoc. Petroleum Geol., 52, 12, p. 2466-2501.
MAGARA, K. (1969). - Porosity-permeability relationship of shale. J. Canad. Well Log. SOC.,2, 1, p. 48-74.
MAGARA, K. (1971). - Permeability considerations in
generation of abnormal pressures. SOC.Petrol. Eng. J.,
11, 3, p. 236-242, and paper SPE 28916.
MARTIN, G.B. (1972). - Abnormal high pressure and
environment of deposition. 3rd Symp. on Abnormal
Subsurface Pore Pressure, SPE of AIME, Baton Rouge,
La., preprint paper SPE 3846.
MASSE, P. (1971). - Relations entre les caracteristiques des couches d’argilites et leur profondeur maximale d’enfouissement. C.R. Acad. Sc. Paris, 272, 11, p.
1481-1484.
MATTHEWS, W.R. (1971). - How well logs indicate
whatpressures to expect. Oil and Gas J., 13 Dec., 69, 50.
MATTHEWS, W.R., McCLENDON, R.T., & SOUCEK,
C.R. (1972). - How to predict formation pressures of cretaceous-jurassicage sediments, Mississippi. 3rd Symp. on
Abnormal Subsurface Pore Pressure, SPE of AIME,
Baton Rouge, La.; preprint paper SPE 3893
MAXWELL, J.C. (1960). - Experiments on compaction
and cementation of sand. In : Rock Deformation (edited
by Griggs, D. T, & Handin, J.), Geol. SOC.Amer., Mem. 79,
p. 105-132.
MAXWELL, J.C. (1964). - lnfluence of depth, temperature, and geologic age on porosity of quartzose sandstone. Bull. Amer. Assoc. Petroleum Geot, 48, p. 698-709.
MAXWELL, J.C., & VERRALL, P. (1954). - Low porosity may limit oil in deep sands. World Oil, 138, (5) : 106113; (6) 102-104.
McCULLOH, T.H. (1963). - Facteurs contrdlant les
variations de densite dans les roches sedimentaires
Oligocenes. Abstr. G.S.A. and Ass. SOC.Joint Mtg, New
York.
296
McCULLOH, T.H. (1967). - Mass properties of sedimentary rocks and gravimetric effects of petroleum and
natural-gas reservoirs. U S . Geol. Surv. Prof. Paper,
528A, 50 p.
McGREGOR, J.R. (1964). - How abnormal pressures
behave in coastal and offshore Louisiana. Oil and Gas J.,
82, 37, p. 128-129.
McGREGOR, J.R. (1965). - Quantitative determination
of reservoir pressures from conductivity log. Bull. Amer.
Assoc. Petroleum Geol., 49, 9, p. 1502-1511.
MEADE, R.H. (1966). - Factors influencing the early
stages of compaction of clays and sands -review. J. sediment. Petrol., 35, p. 1085-1101.
MYERS, R.L., & Van SICLEN, D.C. (1964). - Dynamic
phenomena of sediment compaction in Matagorda
County, Texas. Gulf Coast Assoc. Geol. SOC.Trans., 14,
p. 241-252.
OTVOS, E.G. (1970). - High pressure shales and their
depositional facies, southern Louisiana Cenozoic. J. sediment. Petrol., 40, 1.
PHILIPP, W., DRONG, H.J., FUCHTBAUER, H., HADDENHORST, H.J., & JANKOWSKY, W. (1963). -The history of migration in the Gifhorn trough (NW-Germany).
Proc. 6th Wld Petrol. Cong., Frankfurt, section 1, pap. 19,
p. 457-481.
PICKETT, G.R. (1960). - The use of acoustic logs in
the evaluation of sandstone reservoirs. Geophysics, 25,
1, p. 250-274.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Houston.
POWERS, M.C. (1967). - Fluid-release mechanisms in
compacting marine mudrocks and their importance in oil
exploration. Bull. Amer. Assoc. Petroleum GeoL, 51, p.
1240-1253.
PROSHLYAKOV, B.K. (1960). - Reservoirproperties of
rocks as a function of their depth and lithology. Geol. Neftii
Gaz, 1960 (12), p. 2429.
RATHBUN, P.C. (1968). - Abnormal pressures and
conductivity anomaly. Northern Green River Basin,
Wyoming. 43rd ann. fall Mtg. SPE of AIME, Houston,
Texas.
RAYMER, L.L., HUNT, E.R., & GARDNER, J.S.
(1980). - An improved sonic transit time-toporosity transform. SPWLA, 21stAnn. Log. Symp. Trans.
REHM, B. (1972). - World wide occurrence of abnormal pressures. Part 2. 3rd symp. on Abnormal Subsurface
Pore Pressure, SPE of AIME, Baton Rouge, La., preprint
paper SPE 3845.
RIDER, M.H. (1986). - The Geological lnterpretation of
Well Logs. Blackie Halsted Press, Glasgow.
RIEKE, H.H. (1970). - Compaction of argillaceous
sediments (20-5OO;OOO PSI). Southern Calif. Univ. Ph.D
thesis, 720 p.; Abstr. no 72-21, 698, Diss. Abstr. Int. Sect.
B., 33, 2, p. 755-8-756-B.
RIEKE, H.H. Ill, & CHILINGARIAN, G.V. (1974). Compaction of Argillaceous Sediments. Developments in
Sedimentology, 16, Elsevier, Amsterdam.
ROBERTS, J.E. (1969). - Sand compression as a factor in oil-field subsidence. In : Land Subsidence, 1.A.S.H.Serralog Q 2003
Compaction
~~
Unesco Publ., 89, AIHS, 2, p. 368-376.
ROBERTSON, E.C. (1967). - Laboratory consolidation
of carbonate sediments. In: Marine Geotechnique (edited
by Richards, A.F.), Univ. of Illinois Press, Urbana, Illinois,
326 p.
RYER, T.A., & LANGER, A.W. (1980). - Thickness
change involved in the peat-to-coal transformation for a
bituminous coal of Cretaceous age in central Utah. J.
sediment. Petrol., 50, 3, p. 987-992.
SAHAY, B. (1972). - Abnormal subsurface pressures,
their origin and methods employed for prediction in India.
3rd symp. on Abnormal Subsurface Pore Pressure, SPE of
AIME Baton Rouge, La., preprint paper SPE 39100.
SAWABINI, C.T., CHILINGAR, G.V., & ALLEN, D.R.
(1974). - Compressibility of unconsolidated, arkosic oil
sands. SOC.Petrol. Eng. J., 14, p. 132-138.
Schlumberger (1979). - Well Evaluation Conference.
Algeria.
Schlumberger Ltd (1981). - Data Processing Services
Catalogue.
Schlumberger (1985). - Well Evaluation Conference.
Nigeria.
SERRA, 0. (1972). - Diagraphies et Stratigraphie. In:
Mem. B.R.G.M., 77, p. 775-832.
SERRA, 0. (1974). - Interpretation geologique des
Series deltaiques a partir des diagraphies differees. Rev.
A.F.T.P., 227, OCt., p. 9-17.
SERRA, 0. (1984). - Fundamentals of Well-Log
Interpretation. Volume I : The Acquisition of Logging
Data. Developments in Petroleum Science, 15A, 440 p.,
Elsevier, Amsterdam.
SERRA, 0. & SERRA, L. (2000). - Diagraphies.
Acquisition & Applications. Editions Serralog, Herouville,
France.
SERRA, O., CHIARELLI, A., & GRAS, C. (1975). Automatic study of clay compaction : the DENSON program. The Log Analyst, Jul.-Aug., 16, 4, p. 3-12.
SMITH, J.E. (1971). - Shale compaction. 46th Ann. fall
Mtg SPE of AIME, New Orleans, preprint paper SPE
3633.
SMITH, J.E. (1971). - The dynamics of shale compaction and evolution of pore fluidpressures. J. Int Ass. Math.
Geol., 3, 3, p. 239-263.
SMITH, J.E., & THOMAS, H.G. (1970). - Origins of
abnormal fluid pressures. 20th ann. AAPG Guff Coast
Sect. and SEPM Gulf Coast Sect. Conv., Shreveport,
paper 1970. Abstr., Bull. Amer. Assoc. Petroleum Geol,
54, 9, p. 1791.
SPURLOCK, W.T. (1965). - The relationship between
compaction of sedimentary rocks and high formation fluid
pressure. Tenneco Oil Co.
TAYLOR, J.M. (1950). - Pore space reduction in
sandstones. Bull. amer. Assoc. Petroleum Geol., 34, p.
701-716.
TERZAGHI, K., & PECK, R.B. (1948). - Soil mechanics in engineering practice. 2nd ed. John Wiley & Sons,
Inc., New York.
TIMKO, D.J. (1965). - Practical applications of log calculated pressures as an aid to solving drilling problems.
Serralog 0 2003
1
Chapter7
I
297
~~
The Log Analyst, Sept. -Oct., p. 37-54.
TIMKO, D.J., & FERTL, W.H. (1970). - Hydrocarbon
accumulation and geopressure relationship and prediction of well economics with log-calculated geopressure.
SPE of AIME, 45th ann. fall Mtg., Houston, Texas; et in J.
Petroleum Technol., 23, p. 923-933, 1971.
WALLACE, W.E. (1965). - Abnormal subsurface pressures measured from conductivity and resistivity logs. The
Log Analyst, 5, 4, and Oil and Gas J., July, 63, 27.
WEEKS, L.G. (ed) (1958). - Habitat of oil. Amer. Assoc.
Petroleum Geol,
WELLER, J.M. (1959). - Compaction of sediments.
Bull. Amer. Assoc. Petroleum Geol., 43, 2, p. 273-310.
WILSON, M.D., & PITTMAN, E.D. (1977). - Authigenic
clays in sandstones: recognition and influence on reservoir properties and palaeoenvironment analysis. J. sediment Petrol., 47, 1, p. 3-31.
WYLLIE, M.R.J., GREGORY,A.R., & GARDNER, L.W.
(1956). - Elastic wave velocities in heterogeneous and
porous media. Geophysics, 21, 1.
297
K
W
F
L
a
I
0
18
WELL LOGGING AND TECTONICS
Introduction
After formation, hydrocarbons migrate under the
action of compaction, from the source rock towards their
ultimate reservoirs. This mechanism is called primary
migration. Once in the reservoir rock they continue to
move, a process called secondary migration, settling out
above the denser water occupying the porous spaces of
the rock. The rate at which this process takes place
depends on hydrodynamic gradient within the reservoir
which is controled by the permeability. As the hydrocarbons rise, displacing the formation water, they may
encounter permeability barriers that constitute traps,
under which they accumulate and form hydrocarbon-bearing strata.
Traps are usually classified into three categories (Fig.
8-1 next page):
- structural traps;
- stratigraphic traps;
- mixed or diverse traps.
Among them, structural traps play the major role in
accumulating hydrocarbons. According to Halbouty
(1976), 78 % of the 306 major hydrocarbon fields in the
United States (more than 14 million tons of oil or gas equivalent, for each one) are related to the existence of structural traps. According to Perrodon (1980), this type of trap
represents 89 % of the 266 biggest fields in the world
(more than 70 million tons of oil or equivalent of gas).
These traps are formed by the action of forces that deform
the rocks. Rocks with very low permeability or cap rocks
serve as the closure of these traps.
Stratigraphic traps, representing 10 % of the biggest
fields in the world, correspond to the permeability barriers,
formed by lateral variations of facies, stratigraphic pinchouts of the rocks, or unconformities.
The third category, mixed or diverse traps, are linked
to diagenesis or related to differential pressure.
The drilling of exploration, appraisal and development
wells and the analysis of well logs recorded in them, in
particular dipmeter or image data, leads to the establishment of correlations and the study of pressure gradients
throughout the reservoir. The data may be also used to
complement high resolution seismic. All of these methods
Serralog 0 2003
allow the picture of the subsurface structure to be refined
to a high level of precision.
Analysis of the correlations between 3 wells, X, Y and
Z (Fig. 8-2) results in the definition of the structural form
represented in Figure 8-2a. In this case the information
obtained from dipmeter logs modified, appreciably, the
final structural features, particularly the position of the
anticline crest (Fig. 8-2b).
Certain structures in the North Sea (such as Piper, Fig.
8-3), originally interpreted as anticlines or as pinching out
layers under an unconformity, appeared as more a complex series of faulted blocks as a result of the study of dipmeter data.
It is necessary to emphasize that the interpretation of
dipmeter or image data cannot be carried out without knowledge of lithology, environment and the tectonic features
of the basins in which the drilling is carried out. Moreover,
it seems that knowledge of some of the essential
concepts concerning the deformation mechanisms, the
mechanical behavior of the rocks, and the relation between the types of deformation and the sedimentary
basins is necessary. These concepts allow a better
understanding of the problems confronted. Some of these
concepts will be reviewed briefly in Chapter 9. Others are
reviewed hereafter prior to discussing the actual interpretation of dipmeter or image data.
Review of geological concepts
Definition
Structural geology is "the branch of geology that deals
with the form, arrangement, and internal structure of the
rocks, and especially with the description, representation,
and analysis of structures, chiefly on a moderate to small
scale" (Bates & Jackson, 1980).
Tectonics is "a branch of geology dealing with the
broad architecture of the outer part of the Earth, that is,
the regional assembling of structural or deformationalfeatures, a study of their mutual relations, origin, and historical evolution. It is closely related to structural geology, ...,
but tectonics generally deals with larger features" (Bates
& Jackson, 1980).
299
Figure 8-1 - The various types of trap (adapted from Penn Well).
w
WELLX
Figure 8-2 - (a) Structure deduced from the correlations between wells. (b) Structural cross-section after introducing dipmeter data
(from Schlumberger, Well Evaluation Conference, Venezuela, 1980).
PH
A
Figure 8-3 - (a) Map showing the position of the cross-section across the Piper field. (b) Cross-section across the Piper Field, North Sea.
300
Serralog Q 2003
Tectonics
A structure, in the tectonic sense of the term, is "the
general disposition, attitude, arrangement, or relative
positions of the rock masses of a region or area" (ibid.).
A fold is a continuous deformation, a "curve or bend of
a planar structure such as rock strata, bedding planes"
(ibid.).
A fault is "a fracture or a zone of fractures along which
there has been displacement of the sides relative to one
another parallel to the fracture" (ibid.). This is a discontinuous deformation that acts at the occasion of surfaces of
weakness, irrespective of any deformation of the formations on either side of the fault.
Concepts of stress1
Every element of a rock is subject to a series of forces.
These forces are of two types.
- The first type corresponds to the forces that are
applied to the whole body of the rock. These are called
body forces, and are proportional to the mass of the substances, e.g. gravity, centrifugal forces, magnetic forces.
They are measured in force unit per unit volume (dimension : mLT-2)'.
- The second type are known as surface forces. They
act on the surface of a body and, because of this, are
measured in force units per unit of surface area (dimension: mLT-2/L2 = mL-IT-2). "ln a solid, the force per unit
area, acting on any surface within it", is termed stress
(Bates & Jackson, 1980). Stress is equivalent to a pressure, in which the SI unit is pascal. Taking into consideration all the elements of a rock or bed (Fig. 8-4), the surface forces acting on any imaginary surface are represented by:
- the weight of the above sediments, or the geostatic
pressure, S, and the reaction of the material below;
the fluid pressure pp; if the fluid is in equilibrium (no
movement) the fluid pressure is equal to the hydrostatic
pressure;
- the tectonic forces, T.
1
ChaDter8
I
301
pe
Figure 8-4 - Surface forces acting on a body.
Torsion is "the state of stress produced by two force
couples of opposite moment acting in different but parallel planes about a common axis" (Fig. 8-5d).
Tension
Distortion
-
One must distinguish between the external forces that
act on a body, and the resulting internal actions and reactions that constitute the stress. If the forces acting on a
body are equal on all sides, the body is in equilibrium. The
all-sided pressure is called the confining pressure, C.
In many cases the forces acting on a body are not
equal in all sides. This will cause deformation. If the external forces tend to pull a body apart, the body is said to be
under tension. If it is subjected to external forces that tend
to compress it, it is said to be under compression. If two
equal forces act in opposite directions in the same plane,
but not along the same line, we have a couple, and the
body is said to be under distortion (Fig. 11-4).
This part of the chapter has been reviewed originally by Prof. F.
Proust and Dr. A. Etchecopar, I want to extend my thanks to their unvaluable contributionwhich helped me to issue a text without errors or mistakes.
m - mass, L - Length, T - Time.
*
Serralog 0 2003
Torsion
with variation in rotation
with regards with a same axis
figure 8-5 The four types of forces which can act on a body.
Let us take A as a point in a rock (Fig. 8-6), and C as a
small plane surface, defined by thgntegection of a plane
P passing throughA+A pressure, p = AF/hc will act on C.
We can break the p down into two components: ( 0 ) normal to X, called the normal stress, and (T),parallel to C,
called shear stress.
+
Generally, the pressure p as well as o and T, vary in
magnitude and direction depending on the orientation of
the surface on which they are applied. The set of all the
pressures exerted on point A on all planes that pass
through this point is called the state of stress.
301
the principal stress axes. On these three mutually perpendicular axes, the three principal stresses are as follows (Fig. 8-8):
- greatest or maximum principal stress, o1
- intermediate principal stress, 02;
- least or minimum principal stress, 03;
with 01>02>03.
Figure 8-6 - The resolution of pressure, p, acting on a point, A, of a
surface, 2, into two components: 0,normal to Z,called the normal
stress, and 7,parallel to Z,called shear stress.
The state of stress at any point may be described in
terms of nine stress components of which only six are
independent if the body is in equilibrium. The stresses on
each face of a cube (Fig. 8-7) can be resolved into three
parts, one normal stress, and a shearing stress which
itself can be resolved into two components parallel to the
direction of two of the coordinates.
A
B
D
C
Figure 8-8 - Principal stress axes and the stress ellipsoid.
When the normal stresses are equal no shearing
stresses exist in the material. This state of stress is known
as hydrostatic stress. When they are different, shearing
stresses appear. The geometric representation of the
state of stress at a point is known as the stress ellipsoid
(Fig. 8-8). One can demonstrate that six planes of maximum shearing stresses exist associated in pairs each pair
countaining one of the principal axis, and forming between them an angle of 90” (Fig. 8-9).
Figure 8-9 - Planes of maximum shearing stress.
Figure 8-7 - The stress components acting on the faces of a cube
(from Ramsay, 1967).
There is no direct way to measure the stresses in a
body, but they may be calculated if the external forces are
known.
But it is possible to calculate all the stresses at any
point of the body if the applied stresses at this point on
three mutually perpendicular planes are known. It is also
possible to demonstrate that at each point A, there exist
three orthogonal planes, called principal planes of stress,
for which the shear stress z = 0, and therefore the stress
is perpendicular to them. They constitute symmetry planes for the state of stress.
The three normal vectors to these planes are called
302
4
Figure 8-10 - Planes of maximum shearing stress (S1and SZ, and planes of rupture (Fland FZ, forming an angle 0, close to 30’ to the
maximum principal stress.
Serralog 0 2003
Tectonics
The greatest shearing stress always occurs on the planes which contain o2 axis (2 is maximum the stress difference, o1 - o3being maximum), and make an angle of 45"
to the principal stresses o1and o3irrespective of the signs
or values of the principal stresses (ruptures and slippages
are produced more or less along these planes, Figs. 8-8
and 8-10). In fact, fractures form an angle 8 less than 45"
and close to 30" with the principal axis. By reference to
Coulomb's work, this can be related to the concept of
internal friction which suggests that, at failure, the relationship between the magnitude of shear stress 121 and
normal stress o is :
where T~ is the cohesive strength (sometimes expressed as c for cohesive);
p being the coefficient of internal friction of the material which is related to the angle of internal friction @ by :
1
Chanter 8
I
303
tangents defines the angle of internal friction @,for each
state of stress.
Strain is the deformation caused by stress. This deformation may correspond to a change in volume which is
called dilation Or ComPresSion. It may also ~ ~ uin l at
change in shape: distofiion.
01
@ being related to 8 by the following equation:
The relation between stress and rupture may be determined graphically by the Mohr stress circle (Fig. 8-11)
which is a graphic representation of the state of stress.
(a)
(b)
(C)
Figure 8-12 - Marble cylinders deformed in a laboratory by compression. (a) :undeformed; (b) :20% strain, 270 atm. confining pressure;
(c) :20% strain,445 atm. confining pressure. 01 indicates the direction
of the maximum principal stress (adapted from Press 8, Sieve6 1978).
Mechanical behavior of rocks
Every stress field imposes a strain field, but the resulting deformation also depends on the nature and the
mechanical behavior of the deformed medium.
There are three principle mechanical behaviors.
Figure 8-11 - Mohr stress circle.
To determine the cohesive strength and angle of internal friction, a series of experiments with different values of
the confining pressure must be run on cylinders submitted
to compression tests (Fig. 8-12), and the results reproduced as a Mohr stress circle (Fig. 8-13). The lines drawn
tangent to the successive circles define the Mohr stress
envelope. Their intersection with the vertical axis define
,
corthe cohesive, or shear, strength of the rock T ~ which
responds to the "inherent strength of a material when normal stress across the prospective surface of failure is
zero" (Bates & Jackson, 1980). The slope of each of these
Serralog 0 2003
- Elastic behavior
This behavior is characterized by a possible return to
the initial state. Deformation appears immediately after
the force is applied and strain does not build up. The
deformation obeys Hooke's law, which states that strain is
proportional to stress. The solid regains its dimensions
and its shape when the stress is removed (Fig. 8-14a).
However, this return to the initial shape is not necessarily
immediate, and may indeed take some time. An elastic
solid stands up until a certain limiting stress, called the
elastic limit. If this is exceeded, the solid does not return
to its original shape. When the stress exceeds the elastic
limit, the deformation is plastic. It means that the solid
only partially returns to its original shape (Fig. 8-14e).
When the stress increases, at a certain value the solid
303
304
Well Logging and Geology
fractures. We reach the rupture point The relation existing
between stress and strain is expressed by a stress-strain
diagram (Fig. 8-15).
E
f
(d
Norm01strou in kilotun
Figure 8-15 - Stress-strain diagrams for different rock behaviors.
A :elastic; B :elasto-plastic; C :elasto-plastic with strength hardening;
D :actual elasto-plastic (from Billings, 1972).
tensile strength = 0
tensile strength at low stress,
ductility at high stress
b
5
Figure 8-13 - (a) :Mohr stress envelope (adapted from Billings, 1972).
(b) :Different types of Mohr stress envelopes in relation with the rock
type: (A) :wet clay; (6) :dry sand; (C) :rock materials
(adapted from Ramsay, 1967).
Visco-elastic behaviour
10
Figure 8-16 - Differential stress (0, - od versus strain diagrams explaining the transition from brittle to ductile behavior when the confining
pressure increases (od
Elasto-plastic behaviour
Figure 8-14 - Ideal mechanical models and stress-strain or strain-time curves for material behaviour
(adapted from Ramsay, 1967; Billings, 1972; and Hatcher, 1990).
304
Serralog 0 2003
Chanter 8
The resistance of a material to elastic deformation is
defined as the stress-strain ratio. This ratio is the Young's
modulus E:
E = CJ,/E~
fO
(8-4)
Figure 8-16a - Young's modulus.
with
CJ, = uniaxial compressional stress
E~ = strain. E~ is equal to the ratio of the change in
length, AI, to the original length, lo.
I 305
- Plastic behavior
As previously explained, deformation is permanent
only above a certain threshold. Before this point is reached the substance behaves elastically (Fig. 8-14e).
Plastic deformations result from processes such as intergranular movements, dislocation gIide (intragranular
movements), and recrystallization (including diffusion).
The rheologic model is a mass moving with friction.
Movement will only take place above a certain value of
traction (Fig. 8-14c).
- Viscous behavior
Rigidity measures the
resistance to change in
shape:
(8-6)
G=z/y
F
Figure 8-16b - Shear modulus.
where G is the rigidity or shear modulus, z (= F/A) the
shear stress, and y (= u/l) the shear strain.
G may also be expressed in another way, in terms of
Young's modulus and Poisson's ratio:
G = E / [2(1 + v)]
In viscous material deformation appears immediately
and the strain is unrecoverable (Fig. 8-14b). The shear
modulus is zero in fluids as they do not support a shearing force. Also there is no shear wave propagation and
the Young's modulus has no meaning because stretching
implies shearing.
(8-7)
Fr3
where v is the Poisson's
rat;o equal to the ratio of Figure 8-16c - Poisson's ratio.
transverse strain to axial strain in elastic deformation by
uniaxial stress. In an elastic solid, v is a constant. For isotropic materials, v is between 0 and 0.5.
Viscosity, q, is the property that has a substance to
offer internal resistance to flow. It is equal to the ratio of
the shearing stress, z, to the rate of shear strain, y, per
unit of time, or dy/dt. The rate of shear strain, y, is measured by the change in angle y~ per unit of time t (Fig. 818):
Y=tgW
(8-10)
W
where Ad (= w) is the change
in diameter (equal to E ~ ) .
The bulk modulus or incompressibility K is given by :
K = C J /~ E = Ah I (AVNO) (8-9)
P
where Ah is the change in
hydrostatic pressure, p, and AV Figure 8-16d- '"Ik
the change in volume compared to the original volume Vo.
The rheologic model of an elastic body is a perfect
spring without mass (Fig. 8-14a).
Stress
b
7gure 8- 17 - In a viscous
material its strain is a
function of time and the
rapidity of its strain is a
function of its viscosity.
Deformation speed
Serralog 0 2003
Figure 8-18 - The rate of shear
strain y is measured by the
angular shear variation yJ (from
Billings, 1972).
The viscosity unit is called poise. Viscosity is very high
for rocks but decreases when temperature increases
(Table 8-1). Viscosity is an important property in geological processes. It determines, for example, the flow of
magma or lava during intrusive or volcanic activity, and
the velocity of displacement in plate tectonics.
Water at 100°C and one atmosphere
Water at 30°C and one atmosphere
Water at 0°C and one atmosphere
Corn syrup, room temperature and pressure
Roofing tar, ready to apply
Lava, Mt. Vesuvius, 1400°C
Lava, Mt. Vesuvius, 11 00°C
Rock salt, near surface
Rocks in general
Mantle of earth
0.00284
0.00801
0.01792
7 x 102
3 x 107
2.58 x lo2
2.83 x 104
1017
1017 to 1022
1023
The rheologic model for viscous behavior is a damper,
a perforated piston moving without friction in a fluid (Fig.
8-14b).
305
Factors controlling rock behavior
In addition to their inherent properties (mineralogy, texture, structure), the mechanical behavior of rocks is
controlled by several factors such as confining pressure,
temperature and time.
jected to very short duration stresses, becoming plastic if
these stresses are applied over a long time. This effect is
observed in creep experiments, where a small load
applied for a sufficiently long time produces a strain that
may continue and eventually cause rupture. The same
stress in instantaneous tests would not cause any measurable strain. Figure 8-21 illustrates an ideal creep curve.
Confinina pressure
The strength of a rock increases with the confining
pressure. Figure 8-19 illustrates the effect of confining
pressure on the breaking strength of several standard
rocks. At low confining pressure, all the rocks deform only
a few percent before fracturing. Under a high confinin
pressure, we observe a different behavior for the rocks.
I
25
I
I
I
Shale.
W.ln
~
Shondng in pwernt
Figure 8-20 - Effect of temperature on deformation of marble
(from Griggs, 1939).
tA
Figure 8-19 - Effect of confining pressure on the ductility of several
common rocks (after Donath, 1970).
When fractures appear at less than 3-5 % plastic
deformation, the rocks are said to be brittle. When rocks
are able to sustain, under a given set of conditions, 5-10
% plastic deformation before fracturing, they are ductile.
Ductility is "a measure of the degree to which a rock exhibits ductile behavior under given conditions, commonly
expressed by the strain at which fracture commences"
(Bates & Jackson, 1980). As a consequence, when the
confining pressure increases a brittle rock becomes ductile (i.e. limestone).
TemDerature
The elastic limit decreases when the temperature
increases. Moreover, less stress is necessary to produce
a given strain when the temperature increases (Fig. 8-20).
Time
Time plays a very important part in the behavior of the
rocks. Rocks may exhibit elastic behavior if they are sub306
S = A + B log t + Ct
+D
I
Tim.
-.
-rgure 8-21 - Ideal creep cuwe. A :instantaneous deformation. B :primary creep. C :secondary creep. D :tertiary creep
(from Billings, 1972).
The actual behavior of rocks
In nature, rocks have a complex behavior of all three
types of response visco-elasto-plastic. One of these components may dominate according to physical conditions
(temperature and pressure) and the way the stress is
applied.
At low temperature the elastic deformation of the crystal of quartz shows an almost perfect reversibility.
Rocks which show a good reversibility and admit the
greatest elastic deformation are:
- quartzite, plutonic rocks;
- slates.
Such rocks are brittle.
Semlog 0 2003
Next Page
Tectonics
Some other rocks are more or less ductile, or show an
elastoplastic behavior. Few rocks, such as halite and
undercompacted shales, may have a plastic to viscous
behavior.
According to the previous factors, it is possible to
determine the different kinds of strain following the depth:
- an upper zone, where most of the rocks have an
elastic (brittle) behavior and consequently will fracture
under a certain value of stress;
- an intermediate or middle zone, where the rocks
have an elastoplastic to elastoviscous behavior (ductile);
- a deep zone, where rocks will show a plastic behavior. This zone is characterized by the appearance of
schistosity, and then of foliation. It corresponds to anchimetamorphism and to metamorphism. This type of rock
has no interest in oil exploration, since porosity and permeability disappear.
Tvpes of state of stress
These are three types of state of stress:
- tension or traction: stretches the material and may
increase its volume;
- compressional: leads to a decrease in the volume of
the material;
- pure shear stress: produces a change in shape, but
not in volume (Fig. 8-22).
1
Chapter 8
I 307
Table 8-2
Compressive, tensile, and shearing strengths of
some rocks in kg/cm*
(from Billings, 1942).
I
Rock
Compressive
Sandstone
Limestone
Granite
Diorite
Gabbro
Basalt
Felsite
Marble
Slate
100 to 1400
1 000 to 2 800
1 000 to 2 600
1 000 to 1 900
2 000 to 3 600
2 000 to 2 900
800 to 1 600
700
I
Tensile
~
Shearing
~~
10 to 30
30 to 80
30 to 60
..................
..................
.................
..................
30 to 80
260
60 to 160
100 to 200
150 to 300
....................
....................
....................
....................
100 to 300
160 to 260
General orientation of stresses
The diagrams in Figure 8-23 represent the various
stress states which may be encountered and the main
structures that will be formed by those stresses, as well as
the type of sedimentary basins associated with them.
DIRECTION OF STESSES
01
RESULTS
TENSION
Compression
COMPRESSION
Sup.rficiJm(btirtkroeb1
(minimalhorimul stress)
thrust fwb
Fold axis
Figure 8-22 - Deformation may produce change in volume without
change in shape or change in shape without change in volume, or a
combination of the two (adapted from Leet et al., 7982).
Rock strenath
Rocks are more or less resistant to stresses. The
strength of a rock corresponds to the stress at which the
rock starts a permanent deformation.
Rocks show different types of strength, because they
respond differently to various stresses. Hence, there is,
for each rock, a compressive, tensile and shear strength.
The compressive strength for a brittle rock is sometimes 10 to 30 times more than its tensile strength (Table
8-2).
Serralog 0 2003
Figure 8-23 - Relations between structures, the forces acting upon
them and the type of basin.
307
Previous Page
h = (b/4)2 = h2 = (1 + e)2
The results of stresses : strains
As previously seen, stress deforms the bodies, causing variations in dimensions and the angles between
their different lines or planes (Figs. 8-24 and 8-25).
(8-13)
- the logarithmic, natural or true strain, E, is the logarithm of h:
(8-14)
&=
h
-the angular shear, u,, and the shear strain, y, that express the angular variations, are related by:
y = tangy
A. Unstrained
state
8. Homog.mour
strain
C. lnhomogmour
strain
Figure 8-24 - Various types of two-dimensional deformation
(from Ramsay, 1967).
Undeformed
Deformed
(8-10)
With strain acting on three dimensions (Fig. 8-25), one
considers the coordinate variations of a point (x, y, z)
before and after strain (xi, y,, z,).
Normally, strains are classified in two types, according
to the following geometrical criteria
- Homogeneous strains (Fig. 8-24B), characterized by
the following properties:
. straight lines remain straight after strain;
. parallel lines remain parallel after strain (u,cons
tant);
. all lines in the same direction have the same
values of e.
- lnhomogeneous strains (Fig. 8-24C). In this case we
observe that:
. straight lines become curved after strain;
. parallel line lose their parallelism;
. the values of e and u, vary in any given direction
of the strained bodies.
Strain results can be classified in two important categories.
Figure 8-25 - Deformation due to a finite force in three dimensions
(from Ramsay, 1987).
The reaction of rocks to stress falls into two categories
- continuous strains : fold and flow;
- discontinuous strains which are fractures (studied in
Chapter 9),faults (studied in this chapter), and pressuresolution (stylolites) studied in Chapter 6.
The measurement of these changes enables determination of the type of strain.
When the strain acts on two dimensions (Fig. 8-24),
we have:
- the extension, e, that is defined as the change in
length;
If we express the ratio
e=h-I
b/&as equal to h we have:
(8-12)
- the quadratic elongation, 1,defined as the square of
the ratio of the lengths before and after the strain;
308
Continuous strain : fold
A fold is a flexible strain of materials submitted to compressive or tensile tectonic stresses.
Tensile stress
A normal fault in a basement may be reflected in the
cover by flexure of the overlying beds (Fig. 8-26a).
Differential compaction (Fig. 8-26b), or vertical movements: diapiric uplifts (diapiric shales, mud lump, or halokinesis), and magmatic intrusions, present the same
situation.
ComDressive stress
All types of folds are possible, according to the mechanism of their formation (flexure by compression or by
action of a torque or folding by shearing), the mechanical
behavior of rocks, and their competency. The latter is a
relative property that indicates the capacity of a rock to
withstand compression without variation in thickness. A
competent rock will transmit a compressive force much
farther than a weak, incompetent rock. Hence, more competent beds will bend under plastic strain and slip over
other beds, under the effect of compression. The incomSerralog 0 2003
Tectonics
petent layers that are more passive may be subject to
complex strains. In the most simple case, the fold will be
an isopach type (Fig. 8-27b see the definition, later). We
will have similar folds (Fig. 8-27c) with developing schistosity in the most complex case, or in connection with a
viscous flow.
A fold usually has a general cylindrical symmetry (Fig.
8-28). However, regardless of the extent of a fold - it cannot be infinite - each extremity must develop into a conical structure (Fig. 8-29). Domes (diapiric uplift) generally
have conical symmetry. The apical angle is the characteristic element of a conical fold.
1 Chapter 8 I 309
_I
Figure 8 - 2 7 ~ Example of similar (isogonal) fold [from Ramsay, 7967).
Hinge point
Trough
over a fault
Figure 8-28- The geometrical elements of a cylindrical fold.
over a sand lens
Photograph of a drapping
of a sand lens
Figure 8-26 - Example of tension folds.
Fold4 Cbss'2
Similar
Figure 8-29 - ,The geometrical elements of a conical fold.
-
DescriDtive element of a fold (Fig. 8-30)
- The crest is the highest point of a given stratum in
Fold 5 Clssdi3
Figure 8-27a - Basic fold types (from Ramsay, 1967).
East
Figure 8-27b - Example of
parallel [isopach) fold. The
thickness of the compact beds
is Constant not the anbe oftheir
surface
(from Ramsay, 1967).
L
Serralog Q 2003
any vertical section through a fold and from which the surface slopes downward in opposite directions.
- The trough is the lowest point of a given layer in any
section through a fold.
- The hinge corresponds to the locus of the fold's surface, where the curvature is at a maximum. A box fold has
two hinges, while a perfectly circular fold has no hinge.
- The hinge line is a line joining the points of flexure or
maximum curvature of the bedding planes in a fold.
- The crest line is "a line joining the crest points in a
given stratum" (Bates & Jackson, 1980).
- The crestal surface is "a surface that connects the
crest line of the beds of an anticline" (Bates & Jackson,
1980).
- The axial surface is a surface that joins the hinge
lines of the various beds in a fold. In a vertical fold, the
axial and the crestal surfaces are indistinguishable; in an
asymmetric fold those surfaces can be distinguished from
one another.
- The fold limbs or flanks are the sides of a fold. They
309
Plunge of hinge line
Figure 8-30 - Descriptive elements of a fold (from Ramsay, 1967).
correspond to that area of a fold between two successives
hinges. They generally have a greater radius of curvature
than the hinge region. They may be planar.
- The plunge is the angle between the fold axis and the
horizontal plane measured in the vertical plane. It may be
as great as 90". This is the case of folds with a vertical
axis, which are observed in vertical series deformed by
the passage of strike slip (Fig. 8-31).
Synfom
Figure 8-31 - A fold with a vertical Figure 8-32 - The different forms of
axis.
fold.
- The angle of folding or interlimb angle is the dihedral
angle formed by the two planes that are tangent to the two
limbs of a fold at their inflection points.
- The bisecting plane is the plane that divides the angle
of foldina into two equal parts.
Fold shaDe (Fig. 8-32)
We can distinguish two main shapes of folds
- antiform folds that present a convex curvature in an
upward direction : the limbs close upward; an anticline is
a fold, generally convex upward, whose core contains the
310
stratigraphically older rocks.
- synform folds that present a concave curvature
upward : the limbs close downward. A syncline is a fold,
generally concave upward, of which the core contains the
stratigraphically younger rocks.
If the folded layers retain their correct depositional
sequence in the structure, the same folds are termed antiformal anticline or synformal syncline. If they are in an
inverted sequence, they are termed synformal anticline or
antiformal syncline.
Fold t w e
We can observe, either a homogeneity in the thickness
of the beds (elastic, brittle rocks), or a variation in thickness owing to creep or flowage (plastic rocks), depending
on the mechanical behavior of rocks during folding.
- Parallel or isopach folding (Fig. 8-27a, fold 2 and Fig.
8-27b), in which the normal thickness of a bed, i.e. the
thickness measured perpendicularly from top to bottom,
remains constant. The radius of curvature decreases with
depth. This is the most frequent type of fold formed at
shallow or medium depths.
- Similar or isogonal folding (Fig. 8-27a, fold 4 and Fig.
8-27c) "in which the orthogonal thickness of the folded
strata is greater in the hinge than in the limbs, but the distance between any two folded surfaces is constant when
measured parallel to the axial surface" (Bates & Jackson,
1980). The form of the fold does not vary with depth (Fig.
8-33). These folds are formed at great depth and in the
zone of schistosity.
Nomenclature of folds
In addition to the above-mentioned classifications (by
curvature and by type), folds are also classified as a function of their geometrical shape through a transversal section. Figure 8-34 illustrates the main folds of this classifiSerralog 0 2003
Tectonics
cation with their nomenclature and with the theoretical
representationof dips in a plane perpendicular to the axis.
Figure 8-33 - In the similar fold
the thickness is greater in the
hinge than in the limbs, but the
distance between any two folded surfaces is constant when
measured parallel to the axial
surface which is vertical in this
figure.
AP
I
Chapter 8
I
311
Movements alona a fault
We have seen that faults are fractures with displacements. The displacement or throw of the opposite parts of
the fault ranges from a few centimetres to several hundred kilometres, while its length may be of the same
orders.
The relative movement along a fault may be translational or rotational (Fig. 8-35). When the movement is
translational all straight lines in each block that were
parallel before faulting remain so afterwards. The movement may be vertical (dip slip fault), horizontal (strike slip
fault), or mixed ( diagonal slip fault). The dip of beds on
either side of the fault remains unchanged.
With rotational movement, certain elements in each
block that were parallel before dislocation are no longer
so. The dip of the beds in each block is different.
Figure 8-34 - Nomenclature of folds
We can distinguish the following folds
- A symmetrical or upright fold is a fold with a vertical
axial surface.
- An asymmetrical fold has an inclined axial surface
with limbs that dip in opposite directions and with different
angles.
- An overturned fold has an axial surface, and its two
limbs, inclined in the same direction, usually at different
angles.
- A recumbent fold has a nearly horizontal axial surface.
- A fan fold has a broad hinge region and limbs that are
overturned and converge away from the hinge.
A fold is called isoclinal when its limbs are parallel to
one another.
Discontinuous strain : fractures and faults
These deformations represent the surfaces of a distinct discontinuity, along which cohesion has been lost.
The materials between these surfaces may not be deformed.
- Fracture is a general term that indicates all breaks or
ruptures in a rock, with or without displacement.
- Fault is "a fracture or a zone of fractures along which
there has been displacement of the sides relative to one
another parallel to the fracture" (Bates & Jackson, 1980).
Calling a fault a fracture depends on the scale of observation. Due to the importance of fractures on reservoir
properties (especially permeability), a complete chapter is
devoted (cf. Chapter 9) to the study of fractured reservoirs. Hereafter, our discussion will be limited to faults.
Serralog 0 2003
Figure 8-35 - Different types of faults as a function of the movement of
each block.
Fault description
Afault is described (Fig. 8-36) by defining the following
elements
- its orientation or strike which is the direction or trend
of the line of intersection of the fault "plane" with a horizontal plane, with respect to North;
- its net slip which defines the amount of displacement.
It is divided into:
. dip slip, " c b - "ad", the component of the movement or the net slip measured parallel to the dip of the
fault plane. Further subdivided into
. throw, "ae", corresponding to the displacement
measured along a vertical direction (vertical slip),
. heave, "ad", equal to the displacement in a horizontal direction (horizontal dip slip or horizontal throw);
. strike slip, "ac", which corresponds to horizontal
displacement measured in a direction parallel to the strike
of the fault;
311
- rotation, that is the angle formed by both parts of the
same bed after separation, and measured on a plane perpendicular to the axis of rotation;
- dip, that is the angle between a horizontal surface
and the plane of the fault.
Other elements may be used to describe a fault. Pitch,
for instance, is the angle, measured in some specified
plane, that a line on this plane makes with the horizontal
of the same plane. In Figure 8-36 the angle bac, measured on the fault plane, is the pitch of the net slip.
The two parts separated by the fault plane are the
blocks. The upper side is the hanging wall, the lower side
the footwall. The faulted surfaces of the two blocks parallel to the fault plane are the lips. They are sometimes
polished or striated in the direction of displacement. This
polished surface worn away by erosion is then known as
slickenside. The upthrow of a fault is the upthrown side of
a fault.
ab : net slip
ad = cb : dip slip
ae : vertical slip (a throw s)
ed : horizontal dip slip (a heave s)
ac : strike slip
Figure 8-37 - Classification of fault movements.
- longitudinal faults strike parallel to the strikes of the
regional structures;
- transverse fauts strike perpendicularly or diagonally
to the strikes of regional structures.
Another classification is based on the attitude of adjacent beds relative to the fault plane
- a conformable fault dips in the same direction as the
affected beds;
- an unconformable fault dips in the opposite direction
to the affected beds;
- a bedding fault has its plane parallel to the bedding.
Faults are also classified by shape, being plane or
warped. Warped faults include listric faults, which are
concave spoon shaped, usually pointing upwards. A
growth fault (Fig. 8-38) "forms contemporaneously and
continuously with deposition, so that the throw increases
with depth and the strata of the downthrown side are thicker than the correlative strata of the upthrown side" (Bates
& Jackson, 1980).
Figure 8-36 - The descriptive elements of a fault.
Fault classification
There are several methods of classifying faults. They
can, for example, be classified according to their relative
movement along the fault (Fig. 8-19).
Thus we have:
- gravity or normal faults, sometimes called direct
faults, in which the hanging wall appears to have moved
downwards relative to the footwall;
- thrust or reverse faults in which the hanging wall has
moved upward relative to the footwall. This situation leads
to repeating series;
- strike-slip faults. The movement is predominantly
horizontal and the terms dextral or sinistral are used if the
displacement of each block as seen from the other block
is to the right or to the left respectively;
- oblique or diagonal faults are those that strike obliquely or diagonally to the strike of the dominant structure;
312
1
Figure 8-38 - Example of growth fault and explanation of its formation.
On the side photograph of an experiment.
Serralog 0 2003
Finally, faults can be classified according to their layout
pattern caused by the stress system invoked. Parallel,
radial, en echelon and peripheral faults are shown in
Figure 8-39.
\
Figure 8-39 - Geometrical classification of faults
-
Determination of structure Traditional geological approach
On outcrops, in favorable cases the determination of
tectonic structures can be achieved by direct observation
(Figs. 8-40 to 8-42). But, due to the size of the structure
or the fact that the observation is on a plane or sometimes
partly masked by vegetation, this determination is only
possible through dip determination requiring measurement with compass in two perpendicular vertical planes
and the use of stereographic projection.
I
Figure 8-41 - Photograph of a normal fault filled by calcitic cement. The
drag is very small.
Figure 8-40 - Typical anticline at Durbuy, Belgium.
In a well, the structure determination requires dip measurement from core which is only possible if the core is
oriented and the borehole deviation well known. In the
other cases the measurement provides only an apparent
non-oriented dip which does not correspond to the real
structural dip.
It is the reason why, in subsurface, tectonic information
is generally extracted either from seismic data or from
measurements recorded in wells using dipmeters or
image tools or nuclear measurements in some cases.
Serralog 0 2003
Figure 8-42 - Other example of several small normal faults.
Determination of tectonic structures
from seismic data
This determination is well known and provides the first
information needed to locate the exploratory well.
Unfortunately, the subsurface images provided by seismic
data have not the necessary resolution allowing the
detection of all the faults or deformations which can exist
at the vertical of the well (Figs. 8-43 and 8-44).
313
Next Page
Determination of tectonic structures from
gamma ray
In some special cases the study of the gamma ray
curve can show either overturned features (Fig. 8-45 and
8-46) or reverse faults (8-47). These features can be
confirmed by dipmeter data but the correlations of gamma
ray depth intervals are sufficient to detect them. The
gamma ray response is not affected by the stresses
applied to the formations as the other curves may be.
-
Figure 8-43 - Faults detected on seismic cross-section.
W+E
s
Figure 8-45 - Example of overturned fold detected from the gamma
ray curve. The small section on
the top-right corresponds to the
lower section of the main curve on
the left after its reversal. As it can
be observed the pics correlate
perfectly
(courtesy of Schlumberger).
Figure 8-46 - Other example of
double overturned fold better
detected on gamma ray curve
than on sonic curve. The dipmeter
confirms the upper fold but does
not clearly reflect the lower fold
(courtesy of Schlumberger).
Figure 8-44 - The dipmeter allows the detection of more faults than the
ones recognized on the seismic cross-section shown above.
As illustrated by Figure 8-44, coming from the dipmeter recorded in the well indicated on the seismic section,
many other faults exist in this well. The seismic section
allows the detection of the major faults with a sufficient
throw.
As this book is fundamentally devoted to well logging
applications we refer the readers to the books on seismic
interpretation.
314
J
Serralog 0 2003
Previous Page
~~
Tectonics
Laterolog deep
Dips
Chapter8
I 315
muth of any surface crossing the well (Fig. 8-48).
Combining these two informations the real dip of any surface can be computed. The accuracy of the measurement
depends on the caliper, the apparent dip angle and deviation, the magnetic declination, intensity and inclination
and the processing. This accuracy is close to f 2" for azimuth and 0.2"for dip magnitude for dip higher than 5".
*
Azimuth of hole deviation
Angle of
hole deviation
Figure 8-48 - Principle of dip measurement. Non-alined three points of
the space determine a plane. These three points correspond to the
intersection of a plane with three generatrices of the cylinder corresponding to the well.
Figure 8-47 - A reverse or thrust fault. The slip (80 m) can be determined by defining the depth interval of the gamma ray cutve
that is repeated
(from Schlumberger, Well Evaluation Conference, Iran, 1976).
Determination of tectonic structures
from dipmeter or image data
Tectonic applications were the first geological application developed when dipmeters were introduced. The goal
was to analyze dips and dip evolution with depth to recognize possible rock deformation under stresses.
Ultimately, this could be used to identify tectonic or stratigraphic traps.
The tectonic application is fundamentally based on the
evolution with depth of the dips measured in low-energy
environments. In such environments, the sediments are
deposited by slow settling of fine particles in suspension
only by the force of gravity. This usually occurs at relatively great depths, beneath the action of waves, currents
or tides. The fine particles settle down on surfaces that
are approximately planar and subhorizontal. Consequently, any non-horizontal dip measured in such deposits is assumed to be the result of stresses (tectonic, differential compaction, subsidence, etc.) which have acted
on the deposits.
In wells, tectonic structures are easily detected and
analyzed thanks to dipmeter or image tools. These tools
allow the measurement firstly of the borehole deviation
and azimuth and secondly of the dip magnitude and aziSerralog 0 2003
As well known, dipmeter and image tools (Figs. 8-49 to
8-53) measure changes of the same parameter (conductivity, resistivity, ultra-sonic travel time and amplitude, density, radioactivity, magnetic susceptibility). They do not
directly measure the dip of bed boundaries. But, as the
petrophysical characteristics (lithology, texture, etc.)
generally vary from one bed or lamina to another, they are
reflected in the measured parameter, and they enable the
detection of parallel or oblique laminations, bed boundaries or faults as illustrated by Figures 8-54 to 8-56.
-I
Figure 8-49 - On the left: photograph of the SHDT tool (Stratigraphic
High resolution Dipmeter Tool) (couriesy of Schlumberger). On the
right photograph of the six -arms tool (courtesy of Halliburion).
315
Figure 8-52 - Sketch of the ARI tool (Azimuthal Resistivity Imager)
(courtesy of Schlumberger).
J
Figure 8-50 - On the top: photograph of the FMI (Fullbore Formation
Micmlmager tool (courtesy of Schlumberger). Below: photograph of the
STAR tool (courtesy of Baker-Atlas).
J
Rotating seal
Transducer
interchangeable
rotating sub
Figure 8-51 - On the left: the UBI
(Ultrasonic Borehole Imager) tool. On
the right: the RAB tool (Resistivity At
the Bit) (courtesy of Schlumberger).
Each electrode or button of the dipmeter and image
tools detects resistivity changes because of their sensitivity and high vertical resolution (approximately 1 cm for
dipmeters and 0.5 cm for image tools). If the boundary is
not perpendicular to the tool axis, this resistivity change
occurs at different depths on each recorded curve (cf. Fig.
8-48). The computation of the displacements in depth between buttons and pads, for a given resistivity contrast
(assumed to correspond to a lamina or a bed boundary),
is made by correlation techniques. It allows determination
of the dip of this boundary. Images tools have the advantage to allow the separation of sedimentary features from
tectonic or other features (stylolites, fractures) which is
not possible from the dip data provided by dipmeters tools
processed by interval correlation techniques.
316
b
a
Figure 8-53 - (a) On the left: photograph of the OBMI* (Oil-Base Mud
Imager). (b) On the right: sketch of the tool
(courtesy of Schlumberger).
ADN Tool
LlNC Tool
Figure 8-53c - The ADN
(Azimuthal-Density-Neutron) tool
(courtesy of Schlumberger).
But, because of the
shallow depth of investigation, each dip measurement has an individual
significance which can only
be local if it corresponds to
phenomena of small lateral
extent and with minimal
vertical effect (small scale
sedimentary features).
On the other hand, the vertical persistence of a nearly
parallel dip with slowly varying amplitude or azimuth, over
a certain vertical thickness will indicate the presence of
large structural entities.
Serralog Q 2003
Tectonics
I Chapter 8 I 317
Dipmeter and image tools thus determine the attitude
of the surfaces they traverse.
data. Thus, the absence of correlations between curves
and of dips on an arrow plot does not necessarely mean
that the recording is bad. Over intervals including such
deposits, the number of true structural dips calculated
from dipmeter data will be very low. In these cases a great
deal of skill and experience is required to identify these
few dips. Previous determination of the environment will
give useful clues to identify such cases. Furthermore, it
will also be useful in determining the environment for the
choice of the tectonic model (e.g. growth faults with rollover will be preferred choice in deltaic environments, etc.).
Figure 8-54 - Three types of image of the borehole wall (courtesy of
Schlumberger).
Figure 8-56 - Example of density and photoelectric (PEF) images
obtained by ADN tool in a nearly horizontal well
(courtesy of Schlumberger).
J
Figure 8-55 - Comparison of FMI images, on the left, with RAB images
on the right, and dip computation (courtesy of Schlumberger).
Other tools, such as the RAB tool (Fig. 8-55) or the
nuclear ADN (Fig 8-56) tool, can provide images during
the drilling, but with a much lower vertical resolution.
Should no such surfaces exist, due to the geological
environments (e.9. alluvial fan, conglomerates, breccia,
reefs, bioturbated beds, etc.), it will obviously be impossible to determine structural dips from dipmeter or image
Serralog 0 2003
In any case, in a structural (tectonic) interpretation of
dip data, it is necessary to choose, among all these surfaces, those that will show bed deformations caused by
stress, the other surfaces corresponding to sedimentary
features, stylolites, fractures, erosion, unconformities, etc.
This selection is only possible if the dip data is generated
by a processing based on break correlation (i.e. GEODIP
or LOCDIP for Schlumberger) and is related to a reconstructed lithological column and to the environment.
Hence, only dips related to the boundaries of the beds
that correspond to a low energy sediment, (e.9. to the
nearly horizontal surfaces at the time of sedimentation),
without any current related features, are retained. This
situation is best observed in the well-laminated shaly or
shaly-silty-sandy series, deposited by gravity action, or in
series formed by chemical precipitation (alternance of calcareous series, mudstone type, with shale deposits, etc.).
The type of processing of the dipmeter measurements
must consequently be taken into account. In any case, it
is highly recommended to always interpret dip data kno317
wing the lithology in which dips occur.
The interpretation of the dip or image data requires
also the knowledge of the nature of the well: vertical, inclined or horizontal (Fig. 8-57).
W i l Wll
Ima in
Hori?ontai Hole
Bed Dipping Away from
Kickoff Point
Shale layers
True dip
1.73190
Figure 8-57 - Influence of the well inclination on the images
(courtesy of Schlumberger).
Structural analysis of dip data
By reviewing the stresses and strains, and the structures that result from these forces, it is evident that analysis
of the dips, measured either on outcrop or in a well by dipmeter or image tools, makes a geometrical reconstruction
of the structure possible.
But, to be efficient and correct the interpretation of dipmeter or image data must be done by taking into account
other information and external data.
- Firstly, all the information related to the tool and the
processing used for the dip computation must be known.
1- The type of tool and principle of measurement:
three, four or more arms, number of electrodes per arm,
presence or not of a three-axes accelerometer, presence
or not of "speed buttons" allowing better corrections for
318
downhole movement of the tool;
2- The type of processing used in order to extract
dip data from dipmeter tools (i.e. MARK IV, CLUSTER
(Hepp et a/., 1975), or GEODIP (Vincent et a/., 1979) for
the HDT tool, MSD, CSB, or LOCDIP for the SHDT tool,
BorDip for any tool, OMNIDIP, etc.). Programs based on
correlation interval provide less data and compute a mean
dip which may not correspond to the structural dip but to
a meaningless dip, if the correlation interval contains features related to sedimentary processes, to erosion or fractures. The probability of such features increases with the
length of the correlation interval. Consequently, it is highly recommended to use a correlation interval between 2
and 4 feet. This also reduces the possibility of excessive
tool rotation over the correlation interval, which would, of
course, cause an erroneous dip computation. The quality
of the dips obtained from a correlation-interval processing
can be estimated by the sharpness of the correlogram
and the planarity : the sharper the correlogram and the
better planarity of the computed dip the higher is the probability of a representative structural dip, especially if, at
the same time, it corresponds to a low energy environment. In such environments, we can reasonably assume
that a succession of planar and parallel boundaries
occurs, sedimentary features due to current, tide or wave
action being generally absent. In such cases, the crosscorrelation technique used in programs favours structural
dips, all events on the resistivity curves that are due to
beds or laminations with planar and parallel boundaries
will be in phase for the same displacement across the
borehole. Consequently, this creates a sharp maximum in
the correlogram.
GEODIP or LOCDIP-type programs provide much
more dip data, the quality of which can be checked by looking at the correlation links, which must be selected to
extract dips computed at the boundaries of beds corresponding to low energy sediments. This selection can
be achieved by processing the results with the SYNDIP
program (Fig. 8-58 see next page) combined with a
LITHO-type results.
The dip determination from images can be obtained on
a work station using points of correlation or by fitting a
sine wave over the surface seen on image. Dips can be
computed either using a ScanDip-type program (Fig. 859) or BorDip. The advantage of images is that dips can
be classified as corresponding to sedimentary structures,
bed boundaries, fractures, faults, stylolites or unconformities. This classification can be achieved taking into
account resistivity contrast, depth interval between dips,
dip and azimuth consistency, planarity, etc.
3- Parameters selected for the computation: correlation interval, step distance, search angle, standard or
California option, for correlation interval processings;
number of samples for the computation of the derivative,
values selected for the definition of the size of the events
(small, medium, large), weights applied on each of the 9
Serralog 0 2003
Tectonics
parameters used for the computation of the pattern vector, likeness and discernability thresholds, search angle
and option, for GEODIP processing; number of samples
for the derivative computation, derivative threshold, decimation factor, search angle, for LOCDIP processing, etc..
These parameters should be printed on the heading of the
arrow-plot.
1 ChaDter 8 I 319
OFUENTATION
Figure 8-59 - Example of dip data extracted from FMS images using
the ScanDip program (courtesy of Schlumberger).
A quality control of the recording and dip computation
must also be made by looking carefully at the field monitor log (deviation, tool rotation, orientation, azimuth, relative bearing, resistivity curves, calipers, EMEX current),
and by controlling the dip measurements. A look at the
activity of the curves and the thickness of the events
allows a better determination of the parameters for dip
computation, especially for GEODIP and LOCDIP-type
processings.
- The arrow plot(s).
- The corresponding listing of the computation results
with the quality rating if any has been computed during
the processing.
Figure 8-58 - Example of interpretation of the dipmeter measurements
by the SYNDlP program. This processing ana/yzes curves in terms of
texture (homogeneous, heterogeneous and laminated), recognizes the
green patterns and displays the resistivity curves as a function of the
resistivity values seen by the electrodes. Blue spots correspond to calcitic nodules, yellow to sandstone and red to shale.
Serralog 0 2003
- Secondly, a composite-log of the open-hole logs with
the arrow plot at the same scale, is required. It will help to
give a rough idea of the lithology.
Any dipmeter or image interpretation must be done in
combination with the knowledge of the lithology.
Of course, it is highly recommended to have the geologist's mud-log, or the lithology column obtained by a
processing of the open-hole logs with a program such as
LITHO. The results Of a quantitative interpretation by
GLOBAL or ELAN-type programs can also be useful.
If the sonic waveform has been recorded with the
Long-Spacing Sonic tool or the Array Sonic Service, or
better with a dipole shear sonic measurement, the computation of the elastic properties is possible using the
319
320
Well Logging and Geology
MECHPRO-type program. These data define the mechanical properties of the rocks, information which is helpful
to determine the type of behavior under stress.
- Thirdly, a magnetic tape of the results is required, if
complementary processing of the data is requested .
Polar Plots
These plots are based on:
- either the WULFF stereonet (Fig. 8-61);
- or the SCHMIDT stereonet (Fig. 8-62).
Remark: All this information is not always available, especially for
exchanged wells.
Statistical study of dip data
When a large number of measurements are available
it is preferable to process them statistically. In this case
we will use graphical representations such as dip and azimuth histograms, density stereograms, etc., in order to
identify a preferential grouping or orientation, etc., and to
visualize the results.
DiD and azimuth histoarams
The dip values are plotted on the ordinate and the
number of samples on the abscissa (Fig. 8-60).
The dip values are generally subdivided into 5" or 10"
intervals while their azimuths are grouped into 10" intervals.
W
Figure 8-61 - The Wulff stereonet (see further for explanation of its
construction).
a
0
Figure 8-62 - The Schmidt stereonet (see further for explanation of its
construction).
b
Figure 8-60 - Example of dip (a) and azimuth (b) histograms allowing
the determination of the most probable structural dip.
320
On these projections the planes are represented by
their poles, and dip values are marked from 0" to 90" from
the center towards the circumference of the stereonet.
On a modified SCHMIDT polar plot (Fig. 8-63) the dip
graduations are equidistant, concentric circles increasing
from 0" to 90" at the center. The azimuth graduations are
at 10" intervals.
When the dip values are low this type of plot makes it
easier to plot and study the data.
This plot, which is not a stereographic projection, is
known as polar plot (Fig. 8-64).
Azimuthal freauencv plot
Here a polar plot is also used. Dip azimuths are plotSerralog @ 2003
Tectonics
ted on a circle in 5" or 10" categories, for example. The
lengths of the sector radii (Fig. 8-65) are proportional to
the frequency of each azimuth category. Polar plots and
azimuth frequency plots are commonly plotted together
and known as a polar frequency plot (Fig. 8-66).
1
Chapter 8
I 321
Densitv Stereoarams
As we can see in diagrams or stereograms (Fig. 8-67),
the zones of clustered points are easily perceived. In certain cases of very dispersed measurements it is impossible to find any group of points or trends. The density
stereogram is an aid that transforms the qualitative appreciation of the eye to a quantitative presentation which can
be interpreted more easily.
C
I
SOUTH
Figure 8-63 - A modified SCHMIDT net.
b
Figure 8-67 - Construction of a density stereogram using the
Dimitrijevic counting net. (a): Polar plot of the 150 dips. (b): Counting
the points of (a). (c): Corresponding percentages. (d): Lines of iso-density of the same stereogram (adapted from Henry, 1976).
Figure 8-64 - On the left: a typical polar plot
(courtesy of Schlumberger).
Figure 8-65 - On the right:: an azimuth frequency plot
(courtesy of Schlumberger).
Histograms, polar diagrams and azimuth frequency
plots are all computed automatically by zone in standard
computer processing of dip data.
The principle consists of counting the points (poles of
dips) inside each division, of a fixed area, distributed as a
uniform grid over the plot or stereogram. For this purpose
a counting overlay is used : the one introduced by Pronin
for a conformable projection (WULFF stereonet), or the
one created by Kalsbeek (1963, Fig. 8-68), or Dimitrijevic
(Fig. 8-69) for an equal-area projection (SCHMIDT stereonet).
a
140'
Figure 8-66 - Example of an polar frequency plot
(courtesy of Schlumberger).
Serralog 0 2003
b
Figure 8-68 - (b): Counting diagram obtained by drawing iso-density
points, using a counting net (a) defined by Kalsbeek, 1963
(from Ragan, 1973).
Construction of densitv stereoaram
As the SCHMIDT (azimuth equal-area) stereogram is
generally used, we will use the Dimitrijevic counting net.
The number of points in each small ellipse of the coun32I
ting overlay is counted. This number (or its percentage of
the total) is recorded on the overlay in the center of each
ellipse. Contours of iso-density (or iso-percentage) are
then traced. The azimuth and average dip values of one
or several areas of higher point density can then be read
from the base stereogram (Fig. 8-67). They can correspond to the structural dips if all the sedimentary dips
related to current activities have been eliminated from the
stereogram.
Stereographic plotting techniques
Plotting structures can be done using stereonets,
sometimes called stereographs.
However before discussing plotting techniques it is
useful to review the principal stereographic projections,
their construction and their use.
StereoaraDhic Drojections
"Stereographic projection is a presentation and
abstract geometrical construction which enables analysis
of the orientation of tectonic elements in space independently of their geographical position" ( J . Henry, 1976).
While this is not the place for a detailed account of the
theory and practice of stereographic projection (it is well
treated in texts on structural geology or on this technique
alone), it is very useful to summarize the principles before applying them.
Principle
Stereographic projection involves moving various observed structural elements (lines and planes), while keeping them parallel to themselves, until they touch or cut
the top half of a reference sphere resting on a flat horizontal plane (Fig. 8-71).
Figure 8-69 - The Dimitrijevic counting net.
Interpretation of these diaarams
When the polar diagram shows a distribution of dips
such as that in Figure 8-70, we can conclude that the
structural dip is low and that the azimuth dispersion is due
to these low values. Other dip values that are grouped in
a triangular form, with the apex oriented toward the center, may correspond to either tectonic deformations, or
sedimentary features.
When a diagram (such as the one shown in Fig.8-67)
illustrates a more concentrated distribution, the sectors
having high density and low dip values probably define
the structural dip, sedimentary features, or folds being
added to this.
North
Figure 8-70 -.
Example of an interpretation of a
SCHMIDT diagram
(courtesy of
Schlumberger).
322
Nadir
a
Nadir
b
Figure 8-71 - Stereographic image of a plane (a) :polar projection;
(b) :cyclographic projection.
This plane contains the North-South axis. The intersections of the lines and planes with the hemisphere can
be projected stereographically onto the horizontal plane
with reference to the lower pole or nadir of the whole
sphere. This inversion - which is a geometric transformation (Fig. 8-72) - conserves angles, but looses all information relating to the geographical situation of the sturctural elements. For example, two parallel fault planes
1000 m apart would, when translated to the reference
hemisphere, touch it tangentially at the same point, and
be indistinguishable from each other on the stereogram.
This procedure is useful in that the attitude (strike and
dip) of a plane or line can be represented as a single
point. It is thus possible to analyze the angular relation
between a large number of planes.
Serralog 0 2003
Tectonics
1 Chapter8 I 323
KMxKrn = KZxKO = Constant
C
Diredion ande
Figure 8-74 - Cyclographic projection of a plane.
K
(nadir)
KMxKm = KBxKb = KZxKO = Constant
I
7gure 8-72 - Stereographic projection is a particular case of inversion.
a): inversion of a curve; b): inversion of a circle;
c): inversion of an upper hemisphere.
Planes and lines on a stereogram
As Figure 8-71 indicates, there are two ways of representing a plane.
. Polar plot : this is the stereographic projection of the
point of tangency of the plane and the hemisphere (Fig. 873), or of the intersecting point of the diameter perpendicular to the plane, with the hemisphere. The projection of
this point onto the horizontal reference plane is the pole of
the plotted plane.
If the dip of the plane is higher, its polar representation
will be farther from the center of the sphere, but its cyclographic representation will be closer to the center.
Plotting a line
There is only one method of plotting a line. That is to
project the point of intersection of the straight line passing
through the center of the sphere with the upper hemisphere. Its projection onto the horizontal reference plane
is a point.
Take, for example, the straight line N 160" - 40" NW.
Turn the tracing paper in an anticlockwise direction so as
to align its North with the 160" division of the stereonet
(Fig. 8-75a). Then count 40" southward of the North on
the principal diameter of the stereonet and mark a point.
A tail is usually added, pointing towards the center, to distinguish the representation of a point from polar representation of a plane (Fig. 8-75b). This point now represents
the dip and strike of the straight line.
Dip azimuth
Figure 8-73 - Polar projection of a plane
. Cyclographic plot: this is the stereographic projection
of the intersection of the hemisphere with a plane passing
through its center (Fig. 8-74). This line represents half of
a great circle. Its projection onto the horizontal reference
plane is a curve, part of a great circle.
Serralog 0 2003
Figure 8-75 - (a) fmage of a straight fine N 160" - 40" NW
(b) Stereogram of this line; and (c) its representation in space.
323
The greater the plunge of a straight line, the closer to
the center the point its projection will be. The plunge is the
inclination of a line measured in the vertical plane. Its
pitch is the angle of the line with the horizontal measured
in the plane containing the linear feature (Fig. 8-76).
Vertical plane
Figure 8-76 - Representation of the pitch and the plunge.
Construction of the WULFF stereonet
The WULFF stereonet represents the stereographic
projection of parallels and meridians of the hemisphere of
reference onto the horizontal plane; its North-South axis
has the lower pole or nadir as the projection pole.
Diagrams of Figure 8-77 explain how the meridians
are represented by the arcs of great circles, named on
projection as the great circles, and the parallels by the
arcs of the small circles.
This projection retains the angles : the great and small
circles are orthogonal to one another. The surfaces are
delimited by two parallels and two successive meridians.
They vary according to their position.
This projection is in common usage. In fact, it serves
to determine the following elements : the true dip from
apparent dips, the angle between two planes and the
intersection of the beds with a fault (see later).
However, this projection has disadvantages. It does
not preserve surfaces, due to inversion, and it is thus difficult to draw a density contouring (Figs. 8-67 and 8-68)
from this stereogram. In these cases, it is preferable to
use a SCHMIDT equal-area stereonet.
Construction of the equal-area or SCHMIDT stereonet
The SCHMIDT stereonet corresponds to a Lambert
azimuthal equal-area projection. Relative to the center of
a sphere parallels and meridians are represented on a
SCHMIDT stereonet by the arcs of an ellipse. Any area
bound by two meridians has the same area, regardless of
its position on the projection (Fig. 8-79).
ParalleL
Nadir
Figure 8-79 - The equal-area, or SCHMIDT stereonet construction.
Figure 8-77 - Construction of a WULFF stereonet.
Figure 8-78 represents the isogonal WULFF stereonet
for meridians drawn at 2" intervals and for the parallels
crossing the North-South meridians also, at 2" intervals.
Its uses are much the same as the WULFF stereonet.
It has the advantage of high resolution when used to
represent planes of low dip, and is, hence, the preferred
stereonet for the study of gently folded structures.
Angles are not well preserved but this projection is
especially useful for statistical studies. As J. Henry stated:
with a large number of measurements we can study "the
dip and strike of structural elements from a statistical point
of view". In other words it is possible to consider variations
due to accessory and complex phenomena or measurement errors as negligible and use statistical averages as
the basis for analysis and interpretation.
Stereographic representation of structural elements
Straight lines, planes and folds are the structural elements that will be represented in this exercise.
Small circle
Figure 8-78 - The WULFF stereonet.
The external circle of the stereonet is called the fundamental or primitive circle.
324
A stereonet (WULFF or SCHMIDT) on a firm base is
necessary. Tracing paper should be placed over the
stereonet, and held in position b its center. The center,
major circle and the points of the compass should be marked on the tracing Paper (Fig. 8-80).
Serralog 0 2003
Tectonics
Stereogram
(transparentpaper)
Stereonet
Figure 8-80 - Board for the stereographic plotting of dip data.
- Stereographic representation of a plane
Take, as an example, a plane with a dip of 60" in a
direction N 25" (Fig. 8-81). As noted previously this plane
can be represented in two ways:
. Polar representation of a plane
Turn the tracing paper in an anti-clockwise direction,
so as to align its North with the 25" division on the stereonet, corresponding to the azimuth of the dip to be plotted
(Fig. 8-82). At this point the plane to be plotted should,
conventionally, pass through the E-W axis of the stereonet. The pole of the plane is plotted on the 0" line of the
N-S diameters at the 60" graduation.
Plane direction
or dip strike
60"
-Horizontalplane
Dip
magnitude
I
Chapter8
I
325
The first method of plotting a plane is used when manipulating dipmeter data as it is the azimuth of dips that is
calculated during the processing of the data.
. Cyclographic representation of a plane
The pole plotted previously (25" N 60")is transposed
to the E-W diameter and marked either: 90" from the pole
towards the center of the stereonet; or 60" from the edge
of the stereonet along the line opposite the one on which
the pole lies (Fig. 8-82 right).
The cyclographic trace is drawn as the corresponding
portion of a great circle on the stereonet.
- Stereographic representation of a straight line on a
plane
This is common in structural geology, when, for example, the apparent dips of a plane seen in cross-section
need to be plotted.
Firstly, it is necessary to know the direction and dip of
the plane containing the line, then the azimuth and pitch
of the straight line are required.
Consider a plane with a direction N 60" inclined 50"
NW, containing a line inclined 50" to the West. Construct
the cyclographic trace of the plane using the method illustrated in Figure 8-83. Count along the trace 50" starting
from the N-E edge of the stereonet and following the great
circle. Rotate the tracing paper to bring the great circle to
the principal N-S diameter and read the dip and azimuth
of the line in the plane defined by the point marked on the
great circle.
Vertical plane
-
Figure 8-81 Descriptive elements of a plane dipping at 60" with an
azimuth of N 25".
Figure 8-82 - On the left: polar image of the plane in Fig. 8-81; on the
right: its cyclographic representation.
Figure 8-83 - Stereographic image of a straight line with a "pitch"of
50" towards the west, in a plane N 60" - 50" NW [from Henv, 1978).
Planes with small dips will plot with their poles near the
center of the diagram, while those with greater dips plot
towards the edges of the stereonet. A vertical plane has
its pole on the fundamental circle.
- Stereographic representation of folds
As we have seen, folds are curved surfaces, with
either a generally cylindrical or conical symmetry. The
axis of the fold may be horizontal or inclined and its axial
surface vertical or inclined and planar or warped.
We have also seen that stereographic projection involves the parallel translation of geometrical elements onto a
Note: In certain cases a plane is defined by its direction (strike)
rather than its azimuth. Its direction is 90"from its azimuth. In this case
its pole is plotted on the 270" line of the E-W diameter (Fig. 8-83).
Serralog 0 2003
325
Next Page
reference hemisphere. As the translation of planes is
easier than that of curved surfaces, we will represent
these folds as an infinite number of planes tangent to their
surfaces. This is close to reality, because we are only describing curved surfaces in terms of the attitude of the
plane within the fold that has been subject to same stresses as the fold and intersects the borehole.
Firstly we will describe the theoretical stereographic
representation of a cylinder and a cone. We will then
assume that the features of the fold are known and see
how it can be reconstructed from the stereographic projection.
. Stereographic representation of a horizontal
cylinder (Fig. 8-84)
sentative point can be plotted; from this point rotate the
transparent overlay on the W-E diameter and draw the
great circle, 90" from this point.
Cylindrical fold
Axial plane of the fold
Inflexion points
fold limb
Fold angle
Pole of the axial
plane
Fold axis
Point of tangency
Its reDreSentatlonon
a'stereogram
Figure 8-85 - Image of a cylindrical fold, with inclined axial planes and
a plunging axis (adapted rom Henv, 1978).
Cylinder simulated by
planes tangent to its
upper surface
Cylinder
axis
Point of tangency
Translation of planes
to make them tangent
to the upper hemisphere
Cylinder axis
Cylinder axis
Stereonet representation
of the cylinder
Poles fall on a great diameter
Figure 8-84 - Stereographic representation of a horizontal cylinder (a) :
planes tangent to the cylinder; (b) :translation of these planes onto the
upper hemisphere; (c) :stereogram of the cylinder
(adapted rom Henv, 1978).
As explained, a cylindrical surface may be decomposed into an infinite number of planes tangent to that surface. The planes form a network in space around a common axis. Each plane is represented in its projection by its
pole. The poles representing this network lie in a zone on
a great diameter of the stereonet. The fold axis is situated
on the fundamental circle and is perpendicular to the
great diameter.
. Stereographic representation of a cylinder with
an inclined axis (Fig. 8-85)
In this case the poles of the tangent planes gather in a
zone on a great circle situated at 90" to the axis of tilt.
Knowing this great circle the axis of inclination is defined
by counting 90" on the W-E diameter perpendicular to it.
Having now defined the axis of the cylinder its repre326
Axial
plane
. Stereographic representation of a cylinder with a
vertical axis
In this case the poles of the tangent planes lie in a
zone on the fundamental circle and the axis of the fold is
in the center of the stereogram.
. Stereographic representation of a cylindrical fold
with a straight, inclined axis.
A cylinder, being symmetrical about its axis is not
represented by any one plane. However, in nature a fold
only represents part of a cylinder, this part represented by
the folding angle (the dihedral angle between planes at
tangents to the limbs of a fold). Thus the axis of a fold may
be horizontal (Fig. 8-86) or plunging (Fig. 8-85). If the
stereogram contains points representing the dip at the
points of inflection, A and 6 ,of the two limbs of the fold (a
and b on the stereograms), the folding angle is defined as
the angle between the two planes tangent at these points.
The axial plane is represented by the intersection, on a
great circle, of the axis of the fold, and the bisector of the
two outer beds, assuming that the axial and the bisecting
planes are close (case of isopach or similar folds). It is
easy to construct the pole of the axial plane, by shifting
the axial plane on a great circle and by counting an angular distance of 90" from this great circle on the E-W diameter.
Axial Dlane
Fold axis
Pole
'Axis
%$I?
plane
Figure 8-86 - Image of a cylindrical fold, with inclined axial plane but a
horizontal axis.
Serralog 0 2003
Previous Page
Tectonics
1
Chapter8
I 327
where the poles are aligned.
The inclination of the axial plane in relation to the vertical is determined by counting the angular distance between the center of the stereogram and the great circle
that represents this axial plane. Its azimuth is defined by
turning the transparent overlay to bring the pole onto the
northern half of the N-S diameter and then counting on
the fundamental circle, in an anticlockwise direction, the
number of divisions between the N of the stereonet and
that of the stereogram.
. Stereographic representation of a cone with an
inclined axis (Fig. 8-88)
In this case the planes tangent to the cone form distinct lines once translated onto the hemisphere of reference: they describe the arcs of two oblique circles that do
not lie directly on the small circles of a stereogram.
However, after rotation of the WULFF stereonet by a
value corresponding to the plunge of the axis of the cone
(40" in the example) they fall in a line on two small circles.
. Stereographic representation of a cone with a
horizontal axis (Fig. 8-87)
In this example we will plot a cone with a horizontal
axis oriented to N 150" with a 60" apical angle. The
stereographic projection of all planes tangent to this cone
gives a symmetrical figure the planes are, in fact, tangent
to two small circles of the hemisphere and their poles lie
along two small circles on the stereogram. The horizontal
axis is situated on the fundamental circle on a diameter
perpendicular to the small circles.
Point of tangency
Cons simulated by planes
tangent to ik sur(ace
Nadir
Figure 8-88 - Stereographic image of a cone with an inclined axis
(from Henry, 1976).
Upper hemisphere
Translationof planes
. Stereographic representation of a cone with a
vertical axis (Fig. 8-89)
to make them tangent
lo the *per hemsphere
Cone with vartical axis
end an apical angle of 90'
Stereogram of a ume
with verfical angle down
P d o d tho plane
Stereonet reprerentatbo
d the cone.
Poks fdl on a Smell drcle
a, b. c. d and e are polesd the plan63
Figure 8-87 - Stereographic image of a cone with a horizontal axis (a) :
planes tangent to the cone; (b) :translation of the planes onto the
upper hemisphere; (c) :stereogram of the cone
(adapted from Henv, 1978).
The position of the small circles is a function of the
apical angle. It may be determined by counting the angular distance between the center of the stereogram and
one of the small circles (here 30") and by multiplying it by
2, or directly, by measuring the angular distance between
the two small circles. One of the small circles corresponds
to a synform fold, the other to an antiform fold. The direction of the apex of the cone representing the fold is the
same as the direction of the concavity of the small circle,
Serralog 0 2003
TransJationbf the ume
on UIE upper hamisphere
Figure 8-89 - Stereographic image of a cone with a vertical axis
(adapted from Henv, 1978).
In this case the poles of the tangent planes are gathered in a zone on a circle, the center of which coincides
with that of the stereonet. The apical half-angle is defined
by the angular distance between this circle and the fundamental circle. The axis of the cone is in the center of the
stereonet.
Plotting of dip data
When plotting dip data (magnitude and azimuth) it is
actually the intersection of a plane (or surface) with the
327
borehole that is being represented.
As the borehole axis does not usually follow the axis of
the fold we will be making a cross-section through the
borehole showing the various strata at varying angles,
and naturally at different depths. Stereographic plots do
not take into account the geographical position of features
(in this case their depth) but do record the relative angles
of features. Thus we will have a plot representing planes
tangent to the fold surface and passing through the borehole, and from this information will be able to make a full
description of the fold. Plotting the relation in space between the dips of the representative planes enables us to
determine if the variations define, for example, a cylindrical or conical fold, and to determine the axis and plunge
of the fold as well as the inclination of the axial plane and
the angle of folding.
This technique can also be used to define the intersecting line of any two vertical planes, the apparent dip on
a cross-section of a plane, remove regional dip, make
rotations, etc.
Except for particular applications we will use the polar
representation of planes as it usually simplifies both the
reading and interpretation of stereograms.
Procedure
Normal plotting of dipmeter data is extremely easy.
Quite simply the values of dip, read from logs, or preferably listings, of processed dipmeter or image data are plotted on a WULFF or SCHMIDT stereonet using the previously described techniques.
As prospective oil-bearing structures usually only
involve low dips it may be easier to use a large scale version covering the central part of the stereonet only (Fig. 890). This leaves more space between plotted points
making it easier to analyze their relation. However manual
plotting of several hundred or even thousands of dips that
a dipmeter log will give is a laborious and tedious task.
The use of data, taken directly from the magnetic tape of
the computed dipmeter results, in a simple plotting program can produce stereograms on a video monitor or in
hard copy. The first stage in analyzing dip meter data is to
zone the arrow-plots into coherent intervals. This zoning
is partly done using the lithology, as indicated by other
well logs, but also using the arrow plots themselves, the
arrangement, evolution and patterns of dips influencing
the choice of zones.
H
Figure 8-90 - Enlargement of the center of a WULFF net.
328
H'
For tectonic applications dips computed by the interval
of correlations programs, such as CLUSTER for HDT
data, or MSD or BorDip program for SHDT data, are most
commonly used. If dips computed by the GEODIP or
LOCDIP-type programs are used, intervals corresponding
to low energy beds showing a series of dips with the same
dip and azimuth (green patterns) should be chosen. A
SYNDIP processing (cf. Fig. 8-46) of the GEODIP or
LOCDIP results enables "green" patterns to be selected
(Fig. 8-91). The addition of the lithology information provided by a processing of the open-hole logs by a program
similar to LITHO allows "green" patterns, corresponding
to low energy (shaly) environments, to be identified.
Solid angle = 5"
Figure 8-91 - Determination of "green patterns" on arrow-plot and
stereonet. A "greenpattern" corresponds to a minimum 5 successive
dips falling in a 5" solid angle.
Now, using the coherent intervals chosen by the above
methods, stereograms can be plotted, either manually or,
more easily, using computer programs.
Use of stereograms
At this stage of the analysis the objective is to reduce
the geometric information contained in the stereograms of
each interval to a more synthetic variable.
Several cases may occur.
- The simplest case is that of an interval of approximately constant dip and azimuth : in this case the stereographic projection enables two variables to be defined : the
mean dip over the interval and the angular dispersion of
the dips about this mean. To define the angular dispersion
a circle enclosing the majority of the plotted points and
centered close to their highest density must be drawn. A
Dimitrijevic counter (see Fig. 8-67) may also be used. The
dispersion is interesting for several reasons. Associated
with the percentage of values within an interval it gives an
estimate of the confidence of the mean dip. On the other
hand the value of the dispersion may be characteristic of
certain formations as it is a good indicator of the amount
of stratification.
- The studied interval corresponds to a regular change
of dip and/or azimuth on the arrow-plot. Such patterns
Serralog 0 2003
Tectonics
usually correspond to a cloud of points on a stereonet,
stretching along a line, such that defining an average dip
or azimuth is impossible. This is where stereographic projection can provide important information. It is, in fact, a
precise geometrical translation which can be viewed as a
succession of segments of cylinders or cones inside each
other, each dip corresponding to a plane tangent to these
successive cylinders or cones.
. If the dips are grouped in a line following a great
diameter of the stereonet we can conclude that the fold's
axis is horizontal.
. If the dips lie on a great circle they represent a
cylindrical fold with a plunging axis (orange pattern introduced by Etchecopar). Determination of the azimuth and
the plunge of the axis is carried out in the following manner:
- Draw the great circle passing through
the poles of the planes.
- Count 90" from this great circle on the
E-W diameter towards the center. This point represents
the axis of the cylindrical fold, while its plunge is defined
by counting angular distance from the edge to this point.
- To find its azimuth, turn the transparent
overlay to bring the axis onto the N half of the N-S diameter.
- Then count, in an anticlockwise direction, the angle between the N marked on the stereonet
and that on the stereogram.
. If the dips are grouped in a zone on a small circle we have a conical fold with a horizontal axis. The axis
is on the fundamental circle and on the N-S diameter. The
apical angle is defined by measuring the angular distance
between the center and the small circle and multiplying it
by two. The azimuth is determined by rotating the axis to
align it with the N of the stereonet, and then reading the
anticlockwise angle between the two North (stereonet and
stereogram).
. If the plotted points lie along a curve that does
not correspond to any great or small circles then the axis
of the fold represented may be inclined. By shifting the
stereogram around the stereonet it may be possible to
align the plotted points along a small circle (Fig. 8-92), the
shift corresponding to the pitch of the axis of the cone.
Chapter8
1
329
- In certain cases the arrow-plots may show no apparent organization (Fig. 8-93). However after plotting, the
stereogram shows a coherent pattern, representative of
stratification planes A and 6, indicating a fold, overturned
to the South. The other points (a, b, a and b) also fall perfectly on two small circles of the stereogram on the right,
obtained after rotation to bring the axis of the fold horizontal. These points correspond to conjugated orthogonal
joints. The plunge of the axis is 10". By convention these
patterns are called dispersed axes.
Armw olot
Figure 8-93 - Example of an incoherent arrow plot. When plotted on a
stereonet certain points (A and B), representative of bedding planes,
show a coherent grouping along the line of a great circle. The other
points (a and b, a and p) also represent a coherent feature on the
stereogram plotted on the right. This has been obtained by rotating the
projection to bring the fold axis horizontal. They correspond to the
conjugated orthogonal joints.
Another example of an apparently incoherent arrow
plot is shown in Figure 8-94.
I
Figure 8- 94 - Another example of a random arrow-plot corresponding
to a group of points along a great circle (contributed by J. Henry).
- The opposite case can also occur; an arrow plot
showing clear red and blue patterns giving a dispersed
stereogram with no apparent axes (Fig. 8-95).
Figure 8-92 - Stereogram o f a non-cylindrical fold with an inclined axis
and its plot on a shifted stereonet (adapted from ffenv, 1976).
Serralog 0 2003
- Certain folds have a strongly curved axis. The measurements carried out on this axis at different points are
grouped in a zone on a great circle of the stereogram.
This great circle represents the axial plane of fold hence,
its pole, pitch, and azimuth can be defined (Fig. 8-96a).
329
softwares developped by service companies for work stations.
"Quick-Look" Techniaue : Dip Patterns
Arrow plots or tadpole plots where the results of dip
calculations are presented versus depth, often show dip
arrangements (Fig. 8-97), called dip patterns that can be
organized into four groups and coded by color (Gilreath &
Maricelli (1964).
Dips
Figure 8-95 - The arrow-plot seems to show a "mega-blue pattern".
The stereogram clearly illustrates the dispersion of the plotted dips
suggesting a more complex structure (contributed by J. Henry).
Come otids fimdsmcutally to
di if observed on a
long h t e J with low energy
deposits
On a short interval, nlay
correspond to laminations
s b u c d
Green
Red
Yellow
Figure 8-96 - a) Defining the axial plane by the dispersion of the axes.
b) Summary of dipmeter data illustrating the average dips, the general
trend and the major and minor axes (contributed by J. Henry).
Methods ofh dip data interpretation
Generally, three methods of dip data interpretation are
used.
The first one, which can be qualified as a "Quick-Look"
technique, uses dip patterns observed on arrow plots.
The second, more sophisticated, uses techniques
introduced by Bengtson, an inventor formerly working for
Chevron (1980) and Sohio (1981).
The third, more scientifically based, uses work stations
and softwares in which the stereographic projection is the
base for dip analysis. As previously illustrated the stereographic projection allows a more detailed analysis of dip
data giving a complete and precise reconstruction of
structures crossed by a well.
Visual arrow-plot analysis
A dip, or tadpole, represents a surface cutting the
borehole. The dip and azimuth of this surface may have
two components, firstly a stratigraphic or sedimentary
component caused by current or wave activity, and
secondly a structural component caused by later, largescale tectonic movements.
The first step in the tectonic application of dip data is
to determine the structural dip. The arrow plot interpretation can be achieved either using a "quick-look technique
called dip patterns, or automatically using stereographic
projection. Dip data extracted from dipmeters or imagetool measurements can also be interpretated thanks to
330
Interpretation
Blue
FoUowiiq its length it comesponds
to :
a fault
a draping of a reef or a bar
a rilhg of a chsrmel
an unconformity
--
Emtic events which may
correspond to cross bcddin if
observed on a short intarf.
its length may
:
;:l;Yato:
a halt
foreset laminations
an uuconfonuity
--
Figure 8-97 - Typical colored dip patterns and their interpretation
(from Gilreath & Maricelli, 1964).
- Green patterns correspond to a set of dips with nearly constant magnitudes and azimuths.
- Blue patterns correspond to dip angles of similar azimuth that decrease with increasing depth.
- Red patterns are the inverse, the dip angle decreases with decreasing depth, the azimuth remaining essentially constant.
- Yellow patterns correspond to adjacent dips with random magnitudes and azimuths.
Green patterns, when observed in low energy deposits, usually correspond to a structural dip, a monocline or
a limb of a fold (Fig. 8-98, zone 7500-7580 ft).
Blue and red patterns, when also observed in a low
energy deposits, indicate the structural deformations of
the beds or layers, that may be related to folds, faults with
their associated phenomena (rollover, drag), or to an
unconformity.
The diagrams in Figure 8-99 show the theoretical
arrow plot responses for different cases.
The rapid analysis generally starts by determining the
structural dip from the arrow plots. The dip of the beds is
defined, assuming they were deposited horizontally and
parallel to each other. By extrapolating this dip the architecture of the beds and the shape of the structure they
Serralog 0 2003
Tectonics
constitute can be inferred. Structural dip is generally defined as a predominant dip, over a thick interval, of constant azimuth and magnitude. Therefore, a “base line”, or
succession of green patterns with the same angle and
ostensibly the same azimuth, can be determined on the
arrow plots if the dip is steep . Structural dip is thus determined by a visual, or graphic statistical analysis. In the
case of graphic statistical analysis, dip histograms and
azimuth frequency plots or rosettes are used. The determination of the azimuth of the structural dip may be very
precise as soon as the dip exceeds a few degrees.
Gamma ray
1 Chapter8 1 331
tern (Figs. 8-100 to 8-102) and the extent of the deformation can generally be related to the amount of slip.
Asymmetric foM -Anticline
Ownturned anwine
Dips
Asymmetric fold - syncline
Drag along the fault
plane upper block
Rollover in upper block
of a growth faun
Recumbent antidine
Drag along the fault plane
upper and lower blocks
Beds overturned by drag in
the upper blodc of a thrust fault
Figure 8-99 - Theoretical arrow plots, the four upper diagrams for
anticlines, the four lower diagrams for faults.
Figure 8-98 - Example of structural dip determination on a dipmeter
arrow plot. It corresponds to “green” patterns
(courtesy of Schlumberger).
Determinina structural diD
Structural dips must be determined in rocks that represent low energy environments. These rock types generally correspond to shales, silty shales or mark (Fig. 8-98)
and are identify by:- shaly zones from logs (high gamma
ray, high thorium, uranium and potassium [usually], neutron-density shale separation, high transit time values,
high attenuation of electromagnetic waves, etc.)
- consistent dips (green patterns) over relative long
intervals;
- the lowest dips (usually).
Detectina faults
Faults can be detected on dip arrow plots if drag or rollover was generated during the faulting. This type of
deformation corresponds generally to a red or blue patSerralog 0 2003
In the case of a reverse fault, a repeat section is present and the slip can be determined using the gamma ray
curve; i.e., part of the gamma ray curve will be duplicated
(cf. Figs. 8-46 and 8-47).
Dip
Figure 8-100 - A normal fault is suggested by the color paffems
associated with the drag.
331
Dips
Figure 8-101 - A reverse fault is suggested by the color patterns associated with the drag.
the crestal plane (CP), as well as the type of drag of a
fault (Fig. 8-105).
Figure 8-102 - Example of SCATplots. From the top left to the bottom
right are successively reproduced: the contour map, the dip versus azimuth plot, the croos-section, the DAPSA plot (aeimuth and dip separated but each versus depth, the transverse dip component and the longitudinal dip component (from Bengtson, 1981).
Faults that have no associated deformation can be
detected only with images (see later).
Detectina folds
Folds are characterized by a progressive change of
dip angle, and commonly dip azimuth, with depth over a
long interval (Fig. 8-47 from 2525 to 2675 m). The dips
generally plot on a great circle on a stereonet.
Bengtson Technique
This technique is a method of analyzing dip data, the
apparent advantages of being:
- more analytical than the colored pattern method,
- less time-consuming than the stereographic projections.
This technique uses several graphs or plots to geometrically describe most of the parameters that are used to
define a fold or a fault. An example of such graphs is
shown in Figure 8-102. They are known under the acronym SCAT (Statistical Curvature Analysis Techniques)
plots. The dip-azimuth histogram combined with the azimuth versus depth plot and dip versus depth plot - these
two lastplotsbeing known under the acronym of DAPSA
plots (Dip and Azimuth Plots for Structural Analysis of
dipmeter data) -, allow determination of the structural
transverse direction. This direction will be used to obtain
stick-plots which themselves will be used to draw a vertical cross-section through the well. A dip versus dip-azimuth scatter diagram (Fig. 8-103) enables determination
of the fold or fault type, the drag pattern and the "local"
transverse and longitudinal dip component. The DAPSA
plots combined with the "local" transverse dip component
determine the bearing, the crestal plane (CP), the axial
plane (AP) and the inflection plane (IP) as well as the
plunge for an axial plunging fold (Fig. 8-1O4), or the drag
zone, the precise depth of the fault, the trough plane (TP),
332
Figure 8-103 - Example of dip versus azimuth plot
(from Bengtson, 1981).
-1
Figure 8-104 - Example of analysis of SCATplots in the case of a fold
with plunging axial plane (from Bengtson, 1981).
J
Figure 8-105 - Example of analysis of SCATplots in the case of normal
faults (from Bengtson, 1981).
Serralog 0 2003
Tectonics
DIP AZIMUTH
DIP
S
0
40
D.vi.tion
W
N
I
I
E
S
I
Figure 8-106 - Example of SODA-Seria plot
(courtesy of Schlurnberger).
The DAPSA plots do approximately the same thing but
the presentation is slightly different (Fig. 8-102), the apparent dips being displayed versus depth in the transverse
and longitudinal directions which are determined from the
analysis of the azimuth versus depth plot or the azimuth
frequency plot. Countouring of the data concentration can
be done, the DAPSA apparent dip plots being continuous
equal-area presentations. As pointed out by Wise &
McCrory (1982) this technique can suppress noise and
reveal broader patterns and major trends.
SCAT plots correspond to the plots introduced by
Bengtson (1980, 1981). They include the dip versus azimuth plot (Fig. 8-103) and the azimuth versus depth and
dip versus depth plots associated to the DAPSA plots.
Automatic and interactive programs
The methodology for determining structural dip, faults
and folds has been implemented in two different programs
(Etchecopar & Dubas, 1992). The first program automatically extracts large-scale structural information from dip
data. The second program is interactive, allowing the user
to accurately compute the local value of the structural dip
using stereographic projection. Hereafter are briefly described the different programs
DiDTrend Droaram
The program that operates automatically is called
DipTrend. The processing consists of the following
steps.
Serralog @ 2003
I Chapter8 I 333
1 - Filter the dip data, taking into account the quality
and structural continuity of the data. The result of the filtering will depend on the sequence length selected by the
user (three or five tadpoles).
2 - Sequence the data into green, “orange”, and “yellow” patterns, and determine the axis of each orange pattern.
- A green pattern suggests structural dip, as discussed
previously. Tilted green sequences are differentiated from
the horizontal sequences because their tilt is necessarily
caused by a tectonic event.
-An “orange”pattern corresponds to dips fitting a great
circle on a stereoplot and is characterized by the pole of
this circle. Figure 8-107 (see next page) illustrates the
results of these two steps.
- A “yellow” pattern corresponds to dips fitting a small
circle and it has an axis and a conicity (do not confuse
with the yellow pattern of Gilreath & Maricelli)..
3 - Merge the sequences with consistent axes into
megasequences, assuming that each megasequencewill
characterize a single tectonic event. During this step, an
orange sequence can be merged with other orange or
green sequences into a megaorange, but never into a
megagreen.
4 - Smooth the data according to the previously determined structural axes. Smoothing avoids small discontinuities when building cross-sections or three-dimensional
models at seismic scale. This allows each structure to be
described in three dimensions, as if made of perfectly planar, cylindrical or conical sections.
- Megagreen sequences are smoothed by using either
a median or an averaging filter of the data. The length of
the window used for smoothing, is usually twice the minimum sequence length (expressed in number of dips) of
the initial sequencing, minus one dip.
- Megaorange sequences are smoothed in two steps.
The first step consists in an orthogonal projection of each
pole on the circle characterizing the structure (Fig. 8-108).
The second step consists of smoothing the dips along the
great circle, using a median filter. This smoothing is justified by the assumption that the small angle between each
dip and the circle is caused by microstructures, sedirnentary features or compaction.
5 - Remove the smoothed structural trend from the filtered dips. For each filtered dip, there is now an associated “structural” trend which can be removed. For a green
sequence, the removal is a single rotation around the strike (horizontal line) of the smoothed dip. The value of the
rotation is opposite the value of smoothed dip. For an
orange sequence, the trend removal is divided into two
rotations. The first one removes the dip of the axis of the
great or small circle and is applied to the filtered and
smoothed dips. The second one is, as for a green
sequence, a rotation that removes the smoothed dip to
the horizontal.
The trend removal described above is severe, because the removed structural dip is smoothed considerably to
allow comparison with seismic dip. For accurate sedimentary studies using images or using dipmeters inter333
Figure 8-107 - On the left part is presented the initial sequencing of the arrow plot. On the right part of the figure from left to right are displayed the
original data with the initial sequencing of the upper interval, the filtered dips, dips after structure trend removal, and local curvature axes
(from Etchecopar & Dubas, 1992).
preted by a CSB, GEODIP or LOCDIP - type processing,
a new interactive method for structural dip determination
was developed.
Automatic structural dip determination
The program for automatic structural dip determination
is called, in Schlumberger, SediView. Its application
allows the computation of the structural dip even in intervals showing a lot of sedimentary dips (cross bedding,
foreset bedding, etc.).
SediView derives accurate sedimentological information from dip and lithological results. The module helps
reservoir delineation by inferring geometrical information
from depositional features. The module also helps predict
locations and directions of permeability anisotropy.
SediView consists of four separate steps.
1. Data loading. Precomputed dip data are loaded
from the GeoFrame database. Lithological information
can also be loaded from the same database, but may be
manually input or computed from application of simple
thresholds to the openhole logs.
2. Data filtering. Avariety of filtering techniques may be
applied to focus the interpretation on meaningful dips. A
typical filter would be the dip quality index. In addition,
lithological zoning may be specified.
334
3. Structural dip evaluation. This phase evaluates and
removes the effects of post depositional deformation or tilting. An innovative and accurate approach, based on
Local Curvature Axis techniques, is used to derive the
structural dip which can then be removed in the standard
manner (Fig. 8-109).
The structural dip is traditionally determined from
zones where the dips are constant in magnitude and azimuth (green patterns) and that correspond to low energy
deposits (essentially shales or mark). The value of these
dips provide evidence of structural tilt.
However, in thick sandstone intervals with cross-bedding or foresets the determination of the structural dip is
more difficult. SediView makes use of the observation that
the cylindrical or conical bed surface axes within each
bedset are often co-planar. The dip and orientation of this
plane reflects the effect of the post-depositional deformation. The program for automatic structural dip determination in such cases is based on the two following observations, and one assumption (Beaudoin et a/., 1983). The
first observation is that usually a few poles of consecutive
dips that are close in depth appear aligned on a great circle on a SCHMIDT stereonet (Fig. 8-110). This may be
because of sedimentary structures, compaction or microstructures. Whatever their origin, these "local" great cirSerralog 0 2003
Chapter8
Tectonics
I 335
Local curvature axis
(Schmidt projeclion
-
Upper hemisphere)
Figure 8-108 - Stereographic Schmidt projection - upper hemisphere o f : - (a) raw dip data, - (b) dips after initial sequencing and filtering, - (c) dip
after local smoothing and structure delineation - and (g) determination of the local curvature axes, for the interval 142.0 m to 854.3 m,
radial extent = 90"
(adapted from Etchecopar & Dubas, 1992).
cles can be characterized through the orientation of their
poles (Local Curvature Axis).
The second observation concerns the poles of the
local great circles. Most of the consecutive poles of local
great circles, also called local curvature axes, are themselves aligned on a great circle (cf. Fig. 8-110).
The basic assumption is that the plane that corresponds to the local great circle poles, or local curvature
axes, was nearly horizontal during sedimentation and now
represents the structural dip. In other words, the axes of
microstructures or sedimentary structures, which are
recognized as locally cylindrical, are assumed to have
been nearly horizontal during sedimentation.
The first pass of this module computes the local curvature axis of each bedset or laminaset. These axes are
plotted on the stereonet and analyzed interactively to define structural dip common to a group of beds. This value
is assigned to the corresponding interval. The technique
detects the subtle changes in structural dip that can occur
Serralog 0 2003
between depositional episodes.
4. Sedimentary dip determination.After substraction of
the structural dip for each depositional unit, a dip dispersion analysis is performed to organize the bedsets into
groups. The thickness and the azimuth of each depositional unit are then computed. These values are integrated
by the interpreter with other data from images, logs and
cores as well as with general knowledge of the depositional environment in order to infer reservoir elongation and
transport current direction. Results are then displayed
(Fig. 8-109 right).
Figure 8-109 shows a Formation Microscanner dip file
(on the left) that has been filtered, but not smoothed, by
DipTrend processing. This dip file corresponds to a 6 m
zone where a fairly constant structural dip is assumed.
The Schmidt plot shows that the azimuth of the layers
varies significantly (Fig. 8-111 left). Figure 8-111 right
shows wide variation in the orientations of local great circle axes. As can be observed on the Schmidt projection of
these axes the latter fit a great circle, this great circle
335
Next Page
dip
Local great
circle axes
Dips after
Filtered
dip substraction of
structural dip
Cross-section
Dips measured in the well
Figure 8-109 - On the left: the original arrow plot and the determination of the local great circle axes.. In the middle filtered dips and removed-dip
file. On the right: cross-section of the sedimentary dips. The truncation surfaces can be readily determined. The direction of transport current and
its change with depth can be determined from this display. (from Etchecopar & Dubas, 1992).
representing the structural dip. This structural dip can be
removed, as shown in Figure 8-109 middle, after which a
sedimentary analysis is possible. Finally, a cross -section
of these sedimentary features is constructed (Fig. 8-109
right). The changes in the transport current direction were
almost impossible to detect without this dip removal.
(4)
J
-
Figure 8-110 - Sedimentary features. Their dips correspond to orange sequences of small extent fitting great circles that have clearly different axis orientations. The axes of these Sedimentary features fit a great
circle as they correspond to the same plane that characterizes the original depositional surface. This plane is the structural dip (SO)
(from Etchecopar & Dubas, 1992).
336
Figure 8-111 On the left: stereographic projection of dips of Figure 8109. From this plot the precise determination of the structural dip is not
easy. On the right: Schmidt plot of these axes. The fit of the local axes
with a great circle that represents the structural dip plane is precisely
determined in this case
(from Etchecopar & Dubas, 1992).
Serralog 0 2003
Previous Page
Tectonics
1
Chapter 8
I 337
Analysis of borehole images
-
Figure 8-112 - Possible classification
of surfaces seen on images. This
classification can be based on the
m No type
following attributes:
Cross bed
- resistivity contrast
rn Cemented f r a c t u r e - surface proximity
- planarity
Open f r a c t u r e
- dip magnitude
m Bed boundary
- dip azimuth
Unconformi t y
- lithology
m Others
Surfaces with low resistivity contrast,
high proximitx same dip in a sandstone or limestone corresDonds to
sedimentary features (foresets or cross-bedding). Low planarity with
high resistivity contrast may correspond either to erosional surface if a
lithology change is observed or to stylolites in a limestone Isolated
conductive or very resistive surfaces with a dip angle close to 30"or
60"may correspond either to a fracture or to a fault if a change in texture is observed from one side of the surface to the other.
1
Figure 8-114 -Another fault with small drag (close to 1 m). The fault is
confirmed by change in image texture on each side of the surface, and
by dip evolution with depth ("red pattern").
(from Schlumberger, Middle East Well Evaluation Review, no 9, 1990).
Combined with the lithology determination it is possible to determine the nature of the material filling the fault
(Fig. 8-116).
Fault with very small slip (Fig. 8-117), as well as reverse fault (Fig 8-118), are detectable on resistivity images.
X253
Figure 8-113 - On the left - Complex microfold with steeply dipping
limbs very well detected on the FMI images.
Such features are impossible to detect from dipmeter data
- On the right -Another example of normal fault with a small drag.
(courtesy of Schlumberged.
Serralog 0 2003
Figure 8-115 - A normal fault with no deformation is readily detected on
the images because of the change in texture on each side of the fault
surface (courtesy of Schlumberger).
337
Figure 8-116 - An anhydrite cemented fault (from Schlumberger, Well Evaluation Conference, Angola, 1991).
ORIENTATION
- X277
- X278
- x279
Figure 8-118 - Small reverse fault well detected on the images. On the
right a 3 D view of the borehole wall surface
(courtesy of Schlumberger).
Figure 8-117 - Normal fault with very small slip
(courtesy of Schlumberger).
Fault can be detected as well during the drilling with
the RAB tool as illustrated by image of Figure 8-119.
Images also allow the recognition of en echelon faults
(Fig. 8-120).
338
J
Figure 8-119 - Fault detected by the RAB tool during drilling. An apparent fault is seen in both the RAB (right) and FMI (left) images
(courtesy of Schlumberger).
Serralog 0 2003
Tectonics
I Chapter 8 I 339
ted (Fig. 8-122) the orientation of the minor axes clearly
coincides with those of the larger scale trends. The rotation of the major axes with depth (indicated by the arrow
on the frequency plot) is also significant.
Figure 8-120 - Example of en echelon faults detected on FMS images
(from Schlumberger, Well Evaluation Conference, Angola, 1991).
As illustrated by the previous examples it is obvious
that the image tools are much better than dipmeter tools
to achieve a precise description of the tectonic structure.
This is fundamental to precisely put back the reservoir in
its tectonic setting.
Structural dip versus regional or seismic dip
Usually, the dips determined in a well correspond to
the structural dip, and not regional dip, with the latter better defined by seismic surveys. However, where the structural dip remains generally constant over a large depth
interval, it can be assimilated into the regional dip.
Figure 8-121 - (a) Defining the axial plane by the dispersion of the
axes. (b) Summary of dipmeter data illustrating the average dips, the
general trend and the major and minor axes (contribution of J. Henry).
Naturally the plotting of stereogram summaries should
be adapted to the objective of the study and the geological characteristics of the area in which the well was
drilled. For example, it is evident that data plotted from
formations separated by an unconformity should be plotted on separate stereograms or with different marks on
the same stereogram. It is also useful to label or distinguish the plotted values corresponding to principal formations crossed by the borehole, or to label successive average values to illustrate the evolution of dips intersecting
the well.
Reconstruction o f the tectonic features o f a well
The use of stereograms by chosen intervals gives
information such as the average dip that can be used in
the construction of geological cross-sections for example.
But the geometrical analysis of such data can also provide information about the complete structure. Here, once
again, the stereogram will be a useful summary of the
information available. On the stereogram we can plot the
representative points from each interval (obtained from
the interval stereograms). If the well has been drilled
through a monoclinal structure the resulting cloud of
points plotted will give the average regional dip. In the
case of Figure 8-121b these points form a linear cloud,
enabling the axis of the general structure to be defined. In
other cases intervals will show an evolution of points
which define the major axes characterizing the structure
over an interval of several hundred meters. The orientation of such axes can vary with depth, and such information may also be relevant to the regional geometry. Finally
all the axes identified on the interval stereograms can be
plotted on a summary stereogram.
Another method provides a clear comparison between
the various orientations of the axes plotted. The major
axes and the general trend are added to an azimuth frequency plot of the minor axes, eventually taking into
account their direction of plunge. In the example presenSerralog Q 2003
46 minor axes
Figure 8-122 - Azimuth frequency plot of the minor axes
(contribution of J. Henry).
Dip analysis, containing the same information, but
conserving the relative depths of the structural elements
can be summarized in a table (Table 8-3). Some columns
require comment. In the aspect column the blank sections
represent zones of consistent dip, hatched sections indicate intervals showing an organized evolution of dips that
339
enable axes to be defined, and black corresponds to
zones where the dips were too dispersed to give an average dip or indicate an axis. The average dip column lists
the values read from each interval stereogram. In the
axial direction column all of the minor axes constructed
during analysis of the interval stereograms are listed.
Under the title % of good measurements is the proportion
of good dips to total dips on the arrow plots or listings. The
column dispersion factor in o the angular size of the circles containing the clouds of points are listed. Finally in
the last column the numbers attributed to each level that
have been used in the analytical stereograms are given.
Such tables are used as the basic reference document
when analyzing the geometry of different wells drilled on
the same structure.
Table 8-3
Example of a dip analysis table
(contribution of J.Henry).
Cross-section construction
Arrow plots and stereograms constitute a kind of presentation of the dip data. However, they are not enough
descriptive. Structural geologists are more familiar with 2
D (cross-section) or 3 D (block diagram) representations
of a structure. Consequently, to better visualize the tectonic features it is important to establish the cross-section
from the dips recorded in a well.
Cross-sections can be drawn manually either by adopting a parallel model, in which each of the layers keeps a
constant thickness, or the similar model, in which each
layer can be derived from the others by a simple translation. Figure 8-123 illustrates a cross-section drawn
manually assuming a similar fold.
NNE-
OBSERVATIONS
Figure 8-123 - Cross-section corresponding to the Figure 8-47.
The slip (80 m) can be determined by defining the depth interval of the
gamma ray curve that is repeated
(from Schlumberger, Well Evaluation Conference, Iran, 1976).
With work station and adapted softwares it is now possible to draw automatically the cross-sections.
Automatic cross-section construction
To describe structures in a basin, two main types of
information can be used: seismic sections and dipmeter
measurements or imaging techniques. Geologists are
familiar with seismic sections, which are, increasingly
accurate. However, there are several reasons why significant structures may be missed on seismic sections:
- the vertical resolution of seismics is at the best 20 rn,
- structures may be too small (less than 50 meters),
- average dip may be too high,
- velocity contrast may be too low.
In each of those cases, seismic techniques are insuffi340
Serralog 0 2003
Tectonics
cient to achieve an accurate structural description of a
basin. However, dip measurements are infrequently used
by geologists and geophysicists, even though great
improvements have been made in data acquisition with
the development of imaging techniques. The conventional
display of results (''arrow plots") may not be intuitive to
structural geologists, which may be one reason for their
infrequent use.
The final goal of a geologist is to transform the tadpole information into cross-sections and even block diagrams. Unfortunately, the manual construction of crosssections from large amounts of data is tedious and typically inaccurate. A great number of dipmeter measurements remains poorly interpreted because of the lack of
an appropriate software tool to facilitate this process.
To answer this need, special programs, such as
StrucView (Etchecopar & Bonnetain 1992), were developed for the construction of structural cross-sections from
dipmeter or image data. The StrucView program runs on
a SUN workstation using the dip files generated by
DipTrend processing.
An important constraint imposed on the cross-sections
built with the StrucView program is that they have to be
strictly consistent with the dips measured at the well.
Even when meeting this constraint, different cross-sections are possible for a single data set, depending on fundamental choices under the user's control. The main choices concern the zoning of the well and, for each zone, an
assumption about the type of structure. These decisions
are not always clear-cut and may have to be changed
during the course of the interpretation. Therefore, the software was designed to help the user by providing a completely automatic stereoplot analysis of the data, and it
warns when unusual situations are encountered.
Moreover, all important choices are easily modified, and
the consequences of those changes are quickly taken into
account.
I Chapter8 I
341
lysis.
The parallelism implies that in the plane perpendicular
to the fold axis, each layer is strictly parallel to its neighbours. The similarity implies that the position of a layer
boundary in space can be derived from the others by a
simple translation (Figs. 8-124 to 8-126).
Whatever the orientation of the cross-section plane
relative to the structure, it contains a direction called the
"translation direction" along which a layer boundary can
be derived from the others by a translation. For a given
structure the different directions of translation define a
"translation plane". This plane corresponds to the axial
plane for a fold, the fault plane for a fault and the detachment plane for a rollover.
cs2
Figure 8-124 - Structural axis and translation parameters for a fold.
This figure illustrates the determination of the translation direction for
any section across a similar fold. This fold is cut by two cross-section
planes normal (CS,)and oblique (CSd to the fold. In each of these
planes, the different layer boundaries are obtained from each other by
a translation in a particular direction (A or 6 ) which is the intersection
between the axial plane and the cross-sectionplane. SAX = structural
axis (Etchecopar and Bonnetain, 1992).
Hwotheses for cross-section construction
Cross-sections can be described in terms of folds and
monoclines that may be separated by faults. As already
seen, to describe a fold, geologists classically consider
two models (Ramsay, 1967):
- the parallel fold or isopach fold model - each layer
has a constant thickness within the fold,
- the similar fold or isogonal fold model, - each layer
can be derived from the others by a simple translation.
Similar folds versus parallel folds
Since 1929 (Busk, 1929), all the proposed methods for
cross-section construction, such as the kink method
(Suppe, 1985), have been based on the parallel fold
model. The StrucView program allows this method to be
used when the geologist determines that the dip evolution
corresponds to a fold. However, because this model is
appropriate only for fold analysis, and not for structures in
which thickness varies (such as rollover or drag), the similar fold model was also developed for the program. As a
first approximation, this model may be used for fold anaSerralog 0 2003
SAX
Figure 8-125 - Structural axis and translation parameters for normal
fault IF The movement along the fault plane has developed a drag fold
for which the axis is the structural axis (SAX).By contrast with folds,
the translation plane is not the axial plane of the drag, but the fault
plane itself The intersection of the cross section planes (CS1) and
(CSd with the fault plane defines the translation directions A and 6.
The fault plane, which has pole p, is assumed dipping at 65"
(Etchecopar & Bonnetain, 1992).
341
0 10 20
plane, which contains the structural axis (i.e. the fold axis)
and is the bisector of the two limbs (Fig. 3.2-17). Some
potential polarity changes can be detected between two
consecutive dip sequences, and the user can always validate or invalidate these automatic detections. Once the
decisions have been made about possible overturned
limbs, the axial plane is computed using classical stereoplot techniques.
Faults. Admittedly, the fault plane is the translation
plane (Figs. 8-125 & 8-126). Dipmeter data alone cannot
determine a fault plane, so other information (tectonic
style of the basin, tension or compression forces) is needed to specify the fault type (normal or reverse) and the
fault plane orientation. On images, however, the fault
plane can usually be easily detected and located (Fig. 8127) and the strike and dip easily determined.
Figure 8- 126 - Structural axis and translation parameters for reverse
fault F. The movement along the fault plane has developed a drag fold
for which the axis is the structural axis (SAX).By contrast with folds,
the translation plane is not the axial plane of the drag, but the fault
plane itself The intersection of the cross section planes (CS,)and
(CS2) with the fault plane defines the translation directions A and B.
The fault plane, which has pole p, is assumed dipping at 25" in an
azimuth opposite to that of the layers in the drag zone
(Etchecopar & Bonnetain, 1992).
Constructina cross-sections
To generate a StrucView cross-section, take the following four steps:
1. select a zone
2. determine the structural axis and possibly the translation plane
3. determine the cross-section plane and possibly the
translation direction
4. draw the cross-section.
Selecting a zone
Few wells exist where only a single structure is present. Therefore, the first task for the user is to split the well
into subzones, each one corresponding to a single structure. Correctly identifying each structure is not always
obvious, but the user can change the boundaries of the
zone at any time during the analysis. Furthermore, these
boundaries can vary depending on how much detail the
user needs in the cross section. The user can also choose the zoning determined by the DipTrend processing.
Determining the structural axis and translation plan
Both folds and drags are assumed to be cylindrical
structures, characterized on a stereoplot by the bed poles
aligning along a great circle. The pole of this great circle
is the structural axis. For a cross-section based on similarity, a translation plane must be defined that necessarily
contains the structural axis. Because a straight line is not
enough to determine a plane, some extra information is
needed, which is different for folds and faults.
folds. The translation plane for a fold is the axial
342
s 27 w
Figure 8-127 - Example of a fault that is well detected on FMS 2 pads
images. The change in texture on each side of the sine wave confirms
the slip between the two parts of the figure and the nature of the fault.
Its strike and dip are determined by the sine wave.
It is impossible to determine the fault type without
regional knowledge and the study of missing or repeat
sections (only repeat sections can be detected in a single
well, using a gamma ray for example). To approximate the
fault dip from dipmeter data, an assumption based on the
classical observation that the angle between the main
stress and the failure plane is usually about 25" is introduced. Therefore, where a fault appears, it is assumed to
dip about 65" for a normal fault (Fig. 8-128) and about 25"
for a reverse fault (Fig. 8-129). The fault plane can then
be computed from the stereonet, because it is the only
plane that satisfies the following two constraints:
- it contains the axis of the drag
- it dips 25" in the azimuth opposite to the drag for a
reverse fault, or 65" in the same azimuth as the drag for a
normal fault.
In practice, the angle between the main stress and the
failure plane varies with respect to different parameters,
Serralog 0 2003
Tectonics
such as the confining pressure. The user may modify this
angle if regional knowledge indicates that the angle is different from 25” (or 65”). If structural tilting of the bedding
exists, the user is asked to indicate whether it occurred
before or after the faulting. Fortunately, it is easy to compute and draw the cross-sections in both cases (Figs. 8130 and 8-131).
a
C
a
b
C
d
I ChaPter 8 I
343
b
d
Figure 8-130 - (a) Polar projection of the dips showing a “red pattern”
from the “green pattern” (G). (b) Determination of the structural axis
(SAX). (c) Construction of the polar projection of the fault plane by
counting 65”from the great diameter of the stereogram, point P which
must be on the great circle determined previously, or by counting 65”
from the primitive circle for the cyclographic projection which must go
through SAX.(d) Return to the original position
(adapted from Etchecopar,private communication).
Figure 8-128 - (a) Polar projection of the dips showing a “red pattern”
from the “green pattern” (G). (b) Determination of the structural axis
(SAX)and prolongation along the great circle in order to have 65”from
G (point P). (c) Construction of the cyclographic projection of the fault
plane with P a s its polar projection. The cyclographic projection of the
fault plane must go through the structural axis SAX. (d) Return to the
original position (adapted from Etchecopar, private communication).
a
b
C
d
Figure 8-131 - (a) Polar projection of the dips showing a “red pattern”
from the “green pattern” (G).(b) Determination of the structural axis
(SAX). (c) Construction of the polar projection of the fault plane by
counting 25” from thegreat diameter of the stereogram, point P which
must be on the great circle determined previously, or by counting 25”
from the primitive circle for the cyclographicprojection which must go
through SAX. (d) Return to the original position
(adapted from Etchecopar, private communication).
Figure 8-129- (a) Polar projection of the dips showing a “red pattern”
from the “green pattern” (G). (b) Determination of the structural axis
(SAX)and prolongation along the great circle by counting 25” from G in
the opposite direction (point P). (c) Construction of the cyclographic
projection of the fault plane with P a s its polar projection. The cyclographic projection of the fault plane must go through the structural axis
SAX (d) Return to the original position
(adapted from Etchecopar, private communication).
Serralog 0 2003
On images, the fault dip is easily determined either
manually or with workstations similar to FLIP or
GeoFrame for Schlumberger.
Of course, dipmeter data do not provide information
about the amount of slip. Exceptions include viewing the
343
Next Page
size of the drag, which is qualitatively related to the fault
movement, or identifying reverse fault with repetition of
formations (cf. Figs. 8-46 and 8-47). Extra information
from repeat or missing sections is required.
Rollovers. A direct modeling approach has been developed (Etchecopar and Bonnetain, 1992) on the basis of
recent work that shows rollovers as parts of similar structures (White eta/., 1986; Faure and Chermette, 1989). In
this model, the depths of the top and bottom of the subzone where dips are perturbated by the rollover must be
determined. The bottom is characterized by a clear break
in the dip value. The top, which is more subtle, corresponds to the zone where the fault has no further effect.
The distance between these two depths is taken into
account to determine the radius of the listric part of the
fault. The shear angle is about 60". After the depths have
been selected, the angle of the arc starting from the maximum dip variation in the rollover zone is computed. Figure
8-132 is a real example of rollover with dip data, crosssection and image of the listric fault itself. Figure 8-133 is
an example of rollover cross-section construction.
tion and the direction chosen for the cross section.
Because the dip information is acquired along the borehole, the cross-section plane should stay as close as possible to the borehole trajectory. In the StrucView program,
this plane is constrained to go through the first and last
points of the selected zone. The plane is vertical only if
computed along the azimuth of the borehole deviation. If
computed in the direction perpendicular to the azimuth of
the borehole deviation, the cross-section plane is deviated, as is the well itself, as shown in Figure 8-134.
a
R
Figure 8-133 - On the top, the
reconstruction method for rollover.
First, the intersection of each measured dip with the cross-section
plane is computed. Then each dip
trace is translated parallel to the
direction of translation until it is in angle of arc = 75". shear angle = 601
continuity with the others. On the b
right theoretical geometty of layers
infilling the half graben related to a
listric fault. (a) Well crossing the
detachement plane. (b) Well crossing the curved part of the listric
fault. The dipmeter response
depends only on 8'. (from maximum dip value of layer 37"
Etchecopar & Bonnetain, 1992).
W
Figure 8-132 - Arrow-plot from F M dips,
~
section across the rollover
zone and FMS image of the fault plane. The small deformation feature
on the top of the rollover is an assumed slumped zone
(from Etchecopar & Bonnetain, 1992).
Determining the cross-section plane.
The CrOSS-SeCtiOn plane iS a function Of the Well devia344
Figure 8-134 - CroSS-SeCtiOn Plane in Case Of a deviated Well. The dip
of the Ct-OSS-SectionPlanes D1 and D2 depends on their orientation
relative to the well deviation d. (from Etchecopar & Bonnetain, 1992).
In principle, the user can request a cross-section in
any direction, except for rollovers which cannot be consi&red cylindrical structures. Obviously, the most interesSerralog 0 2003
Previous Page
Tectonics
ting cross-section direction is the direction perpendicular
to the structural axis. The direction of translation is defined as the intersection of the cross-section plane with the
translation plane.
Constructing the cross-section
Once the cross-section direction has been selected,
the intersection of each dip plane with the cross-section
plane is computed (this is equivalent to building a "stick
plot" Fig. 8-135). The cross section is deduced from this
stick plot by shifting each dip trace parallel to the direction
of translation as described in the following.
I Chapter 8 I 345
ging the value of a parameter, it is possible to cancel such
an "anomaly". However, because these anomalies commonly indicate secondary features, exercise caution in
canceling them.
Structural
cross'sec tion
Figure 8-135 - Stick-plots in planes of various strikes of the dips shown
on the left
(from Schlumberger, Well Evaluation Conference, Iran, 1976).
Starting from a given dip trace, such as at the top of
the interval under study, the next trace is translated until it
appears aligned with the first. These two traces are then
linked together to form a piece of curve. A third dip trace
is then shifted and aligned with this new piece of curve
and so on for all dip traces down to the bottom of the interval. Finally, the curve obtained is replicated by translating
it down the well to fit all observed dip traces in the interval
(Fig. 8-136).
The curves under construction are interrupted if the
distance between consecutive dips becomes greater than
a user-defined distance, or if too sharp a break appears in
the average value of the dips (e.g., where the well crosses a major fault). In these cases, the structures are divided into different parts. If an anomalous dip is present, the
corresponding dip trace is shifted into place but not linked
to the others, causing a little break in the curves. By chanSerralog 0 2003
Figure 8-136 - Example of structural cross-section corresponding to
Fig. 8-107. The well was split into three main zones. From the bottom
to 460 m the main anticline. From 460 to 290 m a monocline over an
unconformity. The dip pattern at 325 m is interpreted as a reverse fault.
From 290 to the top a ramp anticline
(adapted from Etchecopar & Bonnetain, 1992).
Exploitation of dip data
Knowledge of the structural dip means that the relative
position of the well can be determined by reference to the
top of the structure. It also helps to predict, establish, and
confirm correlations between wells drilled near each other
(cf. Fig 8-2 and Fig. 8-137). To achieve these kinds of correlations, stick plots may be very helpful. Finally, the struc345
tural dips, determined in several wells of a field, are also
useful in tracing contours of given horizons around the
field (Fig. 8-138).
top of a feature that may be a sand lens or reef (Figs. 8139 and 8-140).
Well A
Figure 8-137 - Correlations between wells done with the help of the
dipmeter data.Red lines indicate extrapolation of structural dip from
wells (from Schlumberger, Well Evaluation Conference, Nigeria, 1974).
(Faun distorsion)
Figure 8-139 - Example of a “redpattern”above a sand bed.
Figure 8-138- Example of isobath map done with the use of the dipmeter data. In black from 5 wells. In blue the final structure (from
Schlumberger, Well Evaluation Conference, Nigeria, 1974).
The second stage of dip data analysis involves detecting the deformations of the beds that took place under
tectonic stress. Faults for example can be identified and
located in depth, but only if their formation was accompanied by associated phenomena such as tilting, stretching,
rotation, brecciation, etc.
Fortunately, this is often the case, and these phenomena manifest themselves as red, blue or yellow patterns. Faults are not, however, the sole cause of such patterns as they may in fact correspond to sedimentary features (foreset beds, large scale cross-bedding, etc.), a
fold, a draping or compaction fold, or an unconformity.
The lithology, the facies, the depositional environment,
the mechanical properties of the formations, fluid pressure measurements, the pattern itself and its position in the
stratigraphic series, all have to be taken into account.
when correlating patterns with such phenomena. Hence,
a red pattern at the top of a sandy or carbonate formation
generally corresponds to a draping or compaction fold on
346
In the case of a convex shape, we can, however, define the direction of elongation of that form that is perpendicular to the azimuth of the beds covering it. Its crest is
in the opposite direction to the dip.
A pattern without significant azimuth variation (& 5”)
may correspond to a cylindrical fold with a horizontal axis,
while one with a progressive change in azimuth could
represent a fold with a plunging axis. A succession of
green patterns can draw a “mega red” or “mega blue” pattern. Such “mega” patterns are often related to folds. For
a simple red or blue pattern, the reality of the hypothesis
establishing a fold is proportional to the length of the
depth interval along which this effect is observed : when
the depth interval is thicker, the hypothesis of a fold is
more valid and it can be established with greater certainty.
The drawing of the cross-section of a well may suggest
a wrong position of the well due to an erroneous interpretation of seismic section. In that case the drilling of a sidetrack well may be decided and may allow the crossing of
the reservoir in a better position, and the discovery of
hydrocarbon (Fig. 8-141).
Determination of the type of fold (cylindrical or conical), the plunge of its axis, etc., can only be done by
stereographic projection of the dips corresponding to the
Serralog Q 2003
Tectonics
patterns (Fig. 8-142). Knowledge of the fold type is important to better evaluate the volume of the reservoir and the
potential hydrocarbon reserve (Figs. 8-143 to 8-145).
The probability of the presence of a fault is higher
when the interval along which the patterns are observed
is short, but may still be ambiguous.Associating such patterns with a fold or a fault can be done:
- either by drawing cross-sections with the help of
stick-plots (Fig. 8-146). These plots give a three dimensional view.
- or using a program similar to StrucView (Fig. 8-147).
An extrapolation of the structural dip to the area adjacent
to the well by applying rules of isogone or isopach conservation of beds following a given direction, can help to
represent the tectonic feature of the well and suggest the
presence of faults (Fig. 8-148);
I Chapter 8 I 347
Side track
well
Main
well
Figure 8-141 - The cross-section drawn for the main well clearly show
that the well was positionned on the side of the main structure. A sidetrack well was decided and the cross-section drawn at the same scale.
It confirms the anticline and the reservoir was found oil-bearing
(from Etchecopar 8, Dubas, 1992).
Figure 8-140 - Example of “red patterns” above a reef
- by finding in the well, by correlation technique using
most of the time gamma ray (cf. Figs. 8-45 to 8-47), a
repeating series (reverse fault); or a gap in series (normal
fault). The latter can be detected when well to well correlations are available (Fig. 8-151 see next pages);
- by knowledge of the tectonic style of the basin;
- or by using stereogram plotting techniques, in which
dip data is analyzed in detail. This technique is described
in a following section.
Very often, in the case of tension that usually creates
normal faults, faulted zones, rather than single faults, are
observed. Thus major faults are generally accompanied,
within a short distance, by several, more or less parallel,
faults. The fault blocks between the minor faults played a
significant role during their respective movements. Thus,
a vertical section ’how’ a
Of closered and
blue patterns, usually accompanied by a change in azimuth from one block to another.
Serralog 0 2003
2
3
Figure 8-142 - The influence of fold axis and fold type on the volume
and the oil-water contact
(from Etchecopar).
347
identified as the series repeat, this phenomenon being
detected on gamma-ray logs as previously seen (cf. Figs.
8-45 to 8-47). In such cases it is also possible to determine the net slip. Image data allow the direct determination
of the fault angle and azimuth.
F.UT
a
Gaslwater
contact
Figure 8-143 - Block diagram corresponding to the lower part of Figure
8-136. It is obtained by interpolation of cross-sections from various
directions using a standard countouring package
(from Etchecopar & Bonnetain, 1992).
FAIAT 1
Figure 8-146 - Stick-plot can help to better visualize faults not so
obvious on the arrow-plot presented on the right
(from Schlumberger, Well Evaluation Conference, Egypt, 1984).
J
Figure 8-144 - Contour map of one layer
(from Etchecopar & Bonnetain, 1992).
Figure 8-145 - Stereogram of the
lower part (465 - 840 m) confirming the conical shape of the
fold.
(from Etchecopar & Bonnetain,
1992).
J
A fault’s direction can be determined from the dips it
has created, being perpendicular to the dip azimuth.
However, its inclination can rarely be determined unless
the fault plane itself causes correlatable resistivity features. Such features identify the fault as an isolated dip with
either the same azimuth as neighboring dips (Fig. 8-149),
or 180” out of phase. The first case corresponds to a
conformable normal fault, while the second case corresponds to an unconformable normal fault if there is no
repetition of series. If series are repeated the contrary
applies, the first case corresponds to a conformable
reverse fault and the second to an unconformable reverse fault (Fig. 8-150). Reverse faults are generally quickly
348
Figure 8-147 - On the left: typical Dip
Trend output. The dips are colored
according to their initial structural
classification (“green”, “orange”, ‘ye/low’y. Next: groups of tadpoles belonging to the same structural trend are
linked by colored bands.
On the upper right: StrucView based
on the Dip Trend results.
Stereogram indicates conical folds
(from Schlumberger,Middle East Well Evaluation Review, no 13, 1992).
Serralog 0 2003
Tectonics
Figure 8-148 - Extrapolation of the structural dip
to the area adjacent to the well by applying rules
of isogone conservation following a given direction. From this reconstructed geometty of the
beds it can be concluded that the crossing of a
fault with drag has a high probability (from
Etchecopar,private communication).
I
Chapter 8
I
349
of a well deviated at 10". The arrow-plot shows, from the
bottom up, a "green pattern" dipping to the north, a "blue
pattern" dipping to the south and a "blue pattern" dipping
to the north. The stereogram clearly shows that this corresponds to a cylinder with axis nearly horizontal and
lying ENE-WSW. This geometry could correspond to
several geological models. Four of these possibilities are
suggested in Figure 8-152:
- in a, two reverse faults,
- in b, a normal fault,
- in c, a disharmonic fold,
- in d, the fill of a paleomorphologicalfeature (a channel, an erosional surface on top of a regional disconformitY).
r
L!
Any fault that is not reversed is normal. But its nature
(tension fault, growth fault, etc.), as well as its net slip, can
only be deduced from complementary information such as
local geological knowledge (tectonic style of basin, depositional environment, etc.), and correlation with other
wells.
The methods discussed previously have given precise
and quantitative information on the geometry of formations that intersect the borehole. The goal in interpreting
this information is to describe geologically this geometry.
-I
Figure 8- 149 - Constructing cross-section using dip-data and stereographic projection for normal fault
(from Schlumberger, Middle East Well Evaluation Review, no 8, 1990).
Figure 8-150 - Constructing cross-section using dip-data and stereographic projection for reverse fault
(from Schlumberger, Middle East Well Evaluation Review, no 8, 1990).
Unfortunately a geometric configuration alone is not
enough, often indicating several possible geological
interpretations. To remove this ambiguity it is always
necessary to have additional information (such as stratigraphy, sedimentology, well log correlations, geophysics
and the regional geology). This is illustrated in the follOwing example.
Figure 8-152 shows an arrow-plot, its stereographic
representation and a stick-plot section normal to the axis
Serralog 0 2003
Figure 8-151 - well to well correlations a//owthe detection of normal
faults positively identified while some are confirmed by dipmeter data
(see Fig. 8-146 corresponding to well C)
(from Schlumberger, Well Evaluation Conference. Egypt, 1984).
349
compact brittle beds. One can however adopt an isogonal
fold-type and allow variations in thickness in plastic beds
(shale, sands, halite, etc.) at hinges or at inflection points.
In drawing the structures, based on measured values,
it is also necessary to consider the tectonic nature of the
region.
Finally for the representation of complex structures
one can use block diagrams. However drawing block diagrams is a long process and requires a certain skill at perspective drawing.
Work stations
C
d
Figure 8-152- Arrow-plot, stereogram and a stick-plot of part of a dipmeter log. Geological models compatible with the geometry
(contributed by J. Henry).
Well log correlation could confirm the solution, with the
exception of evidence indicating repeating series.
The structural style or geophysical (seismic) information, or correlation with nearby wells could indicate a missing series, tending to favor b as the probable solution.
Sedimentological knowledge, lithology, facies, seismic
sections or the regional context (embankment at the edge
of a platform for example) could imply that solution d was
the answer.
Although the solution c was selected, as the series
was a well stratified turbidite sequence that had been
compressed and folded in a N-S direction, the presence
of a normal fault seems to be correct as the extrapolation
of the structure based on the hypothesis of an isogonal
folding (translation of the dip along a direction parallel to
the general axis) seems to suggest (cf. Fig. 8-148).
As illustrated by the previous examples, cross-sections present the arrangement and the geometry of the
structures, deduced from dip data interpretation on the
whole well or a part of it.
The position of the cross-section must be chosen carefully, so as to change the picture of the structure as little
as possible. The represented dips will be the apparent
dips on the chosen plane (most often vertical).
One can start with the results of a stick plot, for which
the plane of cross-section must first be defined. It can be
perpendicular to the vertical plane that passes through the
borehole axis. But if this axis is inclined, this section will
alter the geometric picture (for example, over-estimate the
radius of curvature). In this case it will be better to have
an inclined cross-section with an orthogonal axis. The
position of the cross-section can vary along the same well
depending on the orientation of the structures traversed.
When drawing the beds an isopach fold can be preferred. These are the commonest type of fold, particularly in
350
The emergence of work stations in early 1980’s years
has considerably simplified the work of geoscientists. This
is due to several factors:
- easy importation of data from different origins in the
same environment
- set of softwares for processing and interpretation of
raw and computed data, and display and storage of
resuIts
- interactivity
- rapidity
Work stations have relieved geoscientists of all the
repetitive, tedious, sometimes difficult, and time-consuming part of the analytical aspect of their work. They simplify interpretation as they allow users to easily manipulate any kind of data and display results of processing. The
interpretations use softwares containing rules defined by
“experts”.
Integrating all well data (logs and images, cores, tests,
raw and computed data, well seismics) in the workstation
environment aids geologists, petrophysicists and reservoir engineers in their evaluation of formations.
Geologists can improve the structural, sedimentological,
diagenetic and stratigraphic analysis of logs calibrated on
core data. They can put back the reservoirs in their geological setting. Petrophysicists can refine log interpretation and minimize uncertainty taking into account the geological information provided by the logs. Reservoir engineers can better understand the dynamic potential of a
reservoir through recognition of petrophysical volume properties and surface transmissibility.
One of its most attractive features is the ease with
which all of the tedious dip-data processing techniques
can be performed (stereograms, arrow plots, azimuth frequency plots, cross-sections, etc.), freeing the user from
the onerous tasks of plotting the data and enabling him to
concentrate on the objective of all the previously described manipulations : an interpretation.
On the screen open-hole logs, dip results or resistivity
curves, images, and any kind of plots (azimuth, polar,
stereogram, etc.) can be displayed at any scale over a
chosen interval. Zooming in on a given feature or pattern,
and vertical scrolling is also possible. Elimination of
doubtful results, or addition of dips computed from correlations between dipmeter or image resistivity curves made
by hand on the screen, can be made.
Serralog 0 2003
Tectonics
Other applications of
stereographic projection
The stereographic projection is the main mean of
determining certain angular relations in space. Their
applications are explained in the following sections.
Find the line of intersection of two Dlanes
The line of intersection of two planes will be represented on the stereogram by the intersection point of the two
cyclographic representations of the planes. It is, therefore, sufficient to trace two great circles that are the cyclographic representations of the two planes (Fig. 8-153).
The intersection point P is the projection of this line. To
find its dip and azimuth we rotate the transparent overlay
to place P on the N branch of the N-S diameter. First read
from the edge to point P the angular distance that corresponds to the dip, then count, in an anticlockwise direction, the angle between the N of the stereonet and that of
the stereogram.
N2OVi
I Chapter 8 1
351
nes along two lines that define the apparent dip. It
contains, therefore, these two lines, represented on the
stereogram by two points.
The required plane is represented by the great circle
that passes through these two points (Fig. 8-155). The
method consists in tracing the cyclographic representation of the planes, the orientation of which is known (in this
case the vertical planes in Fig. 8-155, which are N 30" E
and N 40" W). In this case the planes are represented by
the diameters on which the apparent angles (pitches) are
plotted, measuring up from the south of the transparent
overlay because they are oriented towards the north. The
plane is represented by the great circle passing through
these two points. The construction of this great circle and
its pole enables its azimuth and dip to be plotted : N 3" E
28" in this case.
N 40' E
N40'E
N 20" W
Figure 8-153 - Definition of the line at the intersection of two planes
and the angles between them (courtesy of Schlumberger).
Determine the anale between two Dlanes
The intersection of two planes defines four dihedral
angles, opposite pairs of which are equal, the total making
360".
One of the angles will be measured on the plane which
is perpendicular to the intersection line and contains the
poles of the planes, hence it will be measured on the great
circle that passes through these poles .The angular distance between the poles P, and P, is measured along this
great circle.
The complementary dihedral angle (for control) is
determined by measuring the angular distance between
the cyclographic traces, also along the great circle (Fig. 8154).
~
?I
Figure 8-154 - Measurement of the
angle between two planes. Two dihedral angles exist. Their addition gives
180" (courtesy of Schlumberger).
Figure 8-155 - Determining the real dip of a bed from 'rte apparent dips
measured in two vertical planes.
Determination of the aDparent dip of a bed in anv azimuthal direction, when its true dip is known
This is the inverse problem of the previous case. It is
sufficient to plot the cyclographic trace of the bed and to
trace the diameter, the cyclographic representation of the
azimuth of the apparent dip, we wish to measure. The
point of intersection represents the stereographic projection of the straight line at the intersection of the bed and
the plane. Move this point to the N-S diameter and read
the value of apparent dip, the angular distance between
the circumference and the point (Fig. 8-156 : the apparent
dip of the previous bed in a direction N 70" W is 9"). This
type of operation is done automatically when constructing
stick-plots.
N70'W
Determination of the true diD of a bed from amarent
dip measured on two vertical planes
The plane that represents the bed cuts the vertical plaSerralog 0 2003
Figure 8-156 - Determining the apparent dip of a bed in any azimuthal
direction, when its true dip is known
(courtesy of Schlumberger)
351
DiD Removal
In determining the original dip of sedimentary formations, that have been subjected to deformation by folding,
it is important to remove the structural dip of the fold from
the measured dip. The problem consists in applying a
rotation of a fixed value and in a fixed direction to a plane,
the amount of this rotation being determined by another
plane. In this case the cyclographic representation of the
structural dip as well as the polar projection of the measured plane must both be plotted (Fig. 8-157).
N20"E
s40"w
Figure 8-157 -An example of structural dip removal. A dip of 30" N 20"
E is computed for a sedimentary feature found in a formation whose
general inclination is 15" S 40" W.Finding the sedimentary dip of the
feature at the time of deposition is equivalent to rotating the major system back to the horizontal. One finds 46" N 25" E
(courtesy of Schlumberger).
To bring the structural plane horizontal it is sufficient to
align its cyclographic trace with a great circle on the
stereonet. Then, by moving the stereograph by an angle
equal to the structural dip, bring this trace to the circumference. Displace all other poles by the same angle along
the small circles. It may occur that this displacement
along the small circles passes through a horizontal position and intersects the fundamental circle of the stereonet.
It is, in this case, necessary, to shift it diametrically on the
fundamental circle and complete the rotation on the opposite side of the stereogram. This type of operation, dip
removal, is also performed by computers.
References and Bibliography
ASQUITH, G.B. (1982). - Basic Well Log Analysis for
Geologists. Amer. Assoc. Petroleum Geol., Methods in
Exploration Series.
Association des Geologues du Sud-Ouest (1974). Divers aspects de la tectonique. Journees de Toulouse,
28-29 novembre.
BATES, R.L., & JACKSON, J.A. (1980). - Glossary of
Geology. 2nd ed. Amen Geol. InsL, Falls Church, Virginia.
BEAUDOIN, B, PINOTEAU, B., & DELHOMME, J.P.
(1983). - Geological aanalysis of logs and dipmeter data
for well-to-well correlation.
BENGTSON, C.A. (1980). - Statistical Curvature
Analysis methods for interpretation of dipmeter data. Oil &
Gas J., June 23, p. 172-190.
BENGTSON, C.A. (1980). - Structural use of tangent
diagrams. Geology, 8, p. 599-602.
352
BENGTSON, C.A. (1981). - Statistical Curvature
Analysis Techniques for structural interpretation of dipmeter data. Amer. Assoc. Petroleum Geol. Bull., 65, p. 312333.
BIGELOW, E.L. (1985). - Making more intelligent use
of log derived dip information. 5 Parts. The Log Analyst,
26, 1, 2, 3, 4, and 5.
BILLINGS, M.P. (1942, 1972). - Structural Geology. 1st
& 3rd ed. Prentice-Haft Inc., Englewood Cliffs, New
Jersey.
BUSK, H.G. (1929). - Earth flexures. Cambridge
University Press.
BUSCH, D.A. (1974). - Stratigraphic Traps in
Sandstones - Exploration Techniques. Amer. Assoc.
Petroleum Geol., Mem. 21.
Chambre Syndicale de la Recherche et de la
Production du Petrole et du Gaz Naturel. Comite des
Techniciens. (1976). - Methodes modernes de geologie
de terrain. Manuel d'analyse structurale. 2b. Traitement
des donnees. Ed. Technip, Paris.
Chambre Syndicale de la Recherche et de la
Production du Petrole et du Gaz Naturel. Comite des
Techniciens. (1983). - Methodes modernes de geologie
de terrain. Manuel d'analyse structurale. 2a. Methodes
d'observation, de mesure et de notation. Ed. Technip,
Paris.
DELFINER, P., PEYRET, O., & SERRA, 0. (1984).
Automatic determination of Lithology from Well Logs. 59th
Ann. fall Mtg. and Techn. Conf. of SPE of AIME, Houston,
Texas, paper SPE 132900.
DELHOMME, J.P., & SERRA, 0. (1984). - Dipmeterderived Logs for Sedimentological Analysis. SPWLA, 9th
Europ. Intern. Format. Eval. Trans., paper 50.
DICKEY, P.A. (1979). - Petroleum Development
Geology. Petroleum Publishing Co., Tulsa.
DICKINSON, W.R. (ed.) (1974). - Tectonics and
Sedimentation. SEPM, special publication 22.
DODGE, C.F., HOLLER, D.P., & MEYER, R.L. (1971).
- Reservoir heterogeneities in some cretaceous sandstones. Amer. Assoc. Petroleum Geol. Bull., 55, p. 18141828.
DONATH, F.A. (1970). - Some information squeezed
out of rock. American Scientist, 58, p. 54-72.
ETCHECOPAR,A. (1992). - Looking at dips from a different angle. Charismatique, 1, Schlumberger Data
Services, p. 16-29.
ETCHECOPAR, A., & BONNETAIN, J.L. (1989). Cross Sections Construction from Dipmeter Data.
SAID/SPWLA Logging Symposium, Paris.
ETCHECOPAR,A., & DUBAS, M.O. (1992). - Methods
for geological interpretation of dips. SPWLA, 33rd Ann.
Logg. Symp. Transactions, paper J.
ETCHECOPAR, A., & BONNETAIN, J.L. (1992). Cross Sections from Dipmeter Data. Amer. Assoc.
Petroleum Geol. Bull., 76, 5, p. 621-637.
ETCHECOPAR, A., & MOTET, D. (1992). - Structural
interpretation of dip data using the Dip Trend method.
Schlumberger, Reservoir Characterization in the Africa
and Mediterranean Region, no 3, p. 24-31.
Serralog 0 2003
Tectonics
I Chapter 8 I 353
~~
FAURE, J.L., & CHERMETTE, J.C. (1989). Deformation of tilted blocks, consequences on block geometry and extension measurements. Bull. SOC. Geol.
France, 5, 3, p. 451-476.
GARY, M., McAFEE, R.Jr., & WOLF, C.L. (1972).
Glossary of Geology. Amer. Geol. Institute, Washington,
D. C.
GILREATH, J.A., & MARICELLI, J.J. (1964). - Detailed
Stratigraphic Control through dip Computations. Amer.
Assoc. Petroleum Geol. Bull., 48, 12, p. 1902-1910.
GOGUEL, J. (1952, 1965). - Traite de Tectonique.
Masson, Paris.
HALBOUTY, M.T., et a/. (1970). - World's Giant Oil and
Gas Fields, Geological factors affecting their formation,
and basin classification. Amer. Assoc. Petroleum Geol.,
Mem. 14.
HATCHER, R.D.Jr. (1990). - Structural Geology.
Principles, Concepts, and Problems. Merril Publishing
Co., Columbus, Toronto, London, Melbourne.
HENRY, J. (1976). - see Chambre Syndical ... (1976).
HEPP, V., & DUMESTRE, A.C. (1975). - CLUST€R - A
method for selecting the most probable dip results from
dipmeter survey. 50th Ann. Fall Mtg. and Techn. Conf. of
SPE of AIME, Dallas, paper SPE 5543.
HOBBS, B.E., MEANS, W.D., & WILLIAMS, P.F.
(1976). - An outline of Structural Geology. John Wiley &
Sons, New York.
HOBSON, G.D., & TIRATSOO, E.N. (1975). lntroduction to Petroleum Geology. Scientific Press Ltd,
Beaconsfield, England.
JEAN, F. & MASSE, P. (1974). - Methodes d'analyse
structurale. Elf-R.E. - Departement Developpement
Formation.
KERZNER, M.G. (1983). - Formation dip determination - An arfificial intelligence approach. The Log Analyst,
24, 5, p. 10-30.
LANDES, K.K. (1951). - Petroleum Geology. John
Wiley & Sons, New York.
LEET, L.Don, JUDSON, S., & KAUFFMAN, M.E.
(1978). - Physical Geology. 5th ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
LINK, P.K. (1982). - Basic Petroleum Geology. OGCl
Publications, Tulsa.
PERRODON, A. (1980). - Geodynamique petroliere.
Genese et repartition des gisements d'hydrocarbures.
Masson, Paris.
PHILLIPS, C.F. (1955). - The use of the stereographic
projection in structural geology. Arnold edit. Ltd, London,
73 p.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Houston.
PRESS, F., & SIEVER, R. (1982). - Earth. 3rd ed. W.H
Freeman & Co., San Francisco.
RAGAN, D.M. (1973). - Structural Geology. John Wiley
& Sons, New York.
RAMSAY, J.G. (1967). - Folding and Fracturing of
Rocks. McGraw-Hill Book Co., Inc., New York.
RIDER, M.H. (1978). - Dipmeter Log Analysis - An
Assay. SPWLA, 19thAnn. Log. Symp. Trans., paper G.
Serralog 0 2003
RUSSELL, W.L. (1951). - Principles of Petroleum
Geology. McGraw-Hill Book Co., New York.
Schlumberger (1979). - Well Evaluation Conference.
Algeria.
Schlumberger (1983). - Well Evaluation Conference.
Afrique de I'Ouest.
Schlumberger (1985). - Well Evaluation conference.
Nigeria.
Schlumberger Ltd (1970). - Fundamentals of Dipmeter
Interpretation.
Schlumberger Ltd (1981, 1986). - Dipmeter
Interpretation. Volume 1 - Fundamentals.
Schlumberger Middle East S.A. (1981). - Well
Evaluation Conference. United Arab Emirates/ Qatar.
Schlumberger Middle East S.A. (1984). - Well
Evaluation Conference. Egypt.
SELLEY, R.C. (1985). - Elements of Petroleum
Geology. Freeman & Co., New York.
SERRA, 0. (1984). - Fundamentals of Well-Log
Interpretation. Volume I : The Acquisition of Logging
Data. Developments in Petroleum Science, 15A, 440 p.,
Elsevier, Amsterdam.
SERRA, 0. (1986). - Fundamentals of Well-Log
Interpretation. Volume 2 : The lnterpretation of Logging
Data. Developments in Petroleum Science, 15B, 684 p.
Elsevier, Amsterdam.
Services Techniques Schlumberger (1970). - Well
Evaluation Conference. Libya.
SMITH, R.G., & BAKER, J.D. (1983). - The Dipmeter
Advisor System: a case study in commercial expert system development. Proc. 38th Intern. Joint Conf. on
Artificial Intelligence, p. 122-129.
SMITH, R.G., & YOUNG, R.L. (1984). - The design of
the Dipmeter Advisor System. Proc. ACM ann Conf.,
ACM, New York, p. 15-23.
SUPPE, J. (1985). - Principles of Structural Geology.
Prentice-Hall, Inc., Englewood Cliffs, New Jersey.
VINCENT, P., GARTNER, J., & ATTALI, G. (1979).
GEODIP - An approach to detailed dip determination
using correlation by pattern recognition. J. Petroleum
Technol., Feb., p. 232-240.
WHITE, N.J., JACKSON, J.A., & McKENZIE, D.P.
(1986). - The relationships between the geometry of normal faults and that the sedimentary layers in their hanging
walls. Jour. Struct. Geology, 8, p. 897-909.
WILLIS, B., & WILLIS, R. (1934). - Geologic
Structures. McGraw-Hill, New York.
WISE, D.V., & Mc CRORY, T.A. (1982). - A new
method of fracture analysis :azimuth versus traverse distance plots. Geol. SOC.Amer. Bull., 93, p. 889-897.
353
WELL LOGGING AND FRACTURES
Introduction
Fracture is a general term that indicates all breaks or
ruptures in a rock, whether accompanied by a displacement or not. It corresponds to a surface along which there
is a loss of cohesion.
These ruptures are caused by tectonic forces (tension,
compression or torsion), by expansion of rocks upon
release of overburden load (i.e. erosion), unequal pressure, or excessive pressure and compression, or by changes of temperature (i.e. dessication), by leaching in the
plane of stratification or schistosity, and also by drilling.
Generally grouped in the category of fractures are:
- crack is a partial or incomplete fracture;
- fissure is “ a surface of fracture or a crack along which
there is a distinct separation. It is often filled with mineral
bearing material‘ (Bates & Jackson, 1980);
-joint is “ a surface of fracture without displacement;
the surface is usually plane and occurs with parallel joints
to form part of a joint set“ (Bates & Jackson, 1980);
- gash is a small-scale tension fissure of several centimeters to a few decimeters in length, and several millimeters to a few centimeters in width. It may be gaped or,
most often, filled with crystals. Several gashes are most
frequently arranged in en echelon (Fig. 9-1). They are
produced by simple shear;
Figure 9-1 - En echelon tension gashes produced by simple shear.
Top : Theory :deformation of a cube under couple.
Bottom : Photograph of an actual case (from Ramsay, 1967).
Serralog 0 2003
- fault is “ a fracture or a zone of fractures along which
there has been displacement of the sides relative to one
another parallel to the fracture” (Bates & Jackson, 1980).
Calling a joint or fault a fracture depends on the scale of
observation.
The fractures may be cemented (filled with crystalline
material: calcite, anhydrite, pyrite or by clay material) or
open. Clearly it is the open fractures which are of interest
for production, because they create substantial permeability, and a preferred flow path for the fluids. The latter are
largely caused by tension or torsion, while closed fractures are generally associated with compression.
Fractures are “generally” transverse to the plane of
stratification, and are usually more or less planar. The two
lips of the fracture plane are locally not separated.
Moreover, the occurrence of fractures is not random (Fig.
9-2). In a constrained formation, the fractures appear as
interconnected systems, each system consisting of a
group of more or less parallel fractures. They result in the
rock being broken up into small volumes or parallelepipeds which can be broken off by the drill-bit or the rotating
drill-pipe.
Figure 9-2 - Fracture systems related to folding. Outer layer in tension,
inner layer in compression, separated by a neutral layer.
L Longitudinal fractures, normal fractures and tension fissures in the
outer layer, inverse fractures in the inner layer.
T : Transverse tension fissures related to stylolitic peaks S’ in the inner
layer. Ds :Sinistral diagonal fractures. Dd :Dextral diagonal fractures.
S :Stylolites with vertical peaks (often more numerous in the outer
layer) and associated with small vertical fissures S’ :Stylolites with
peaks parallel to the bedding planes (subhorizontal).
355
The average gap of a fracture, or fracture aperture, is
often less than 0.1 mm, and so the porosity of fractures is
generally negligible (less than 2%). Boyeldieu et a/.
(1982) have estimated that, if the fracture system breaks
the rock into cubes with 10 cm edges, a gap of 1 mm
Would be necessary to create a Porosity of 1 (Fig. 9-3).
Borehole axis
Figure 9-3 - SuPPOsing a fracture
aperture equal to t, in a rectangular
parallelepiped, its volume is equal
to: t x (a / cosa ) x b , which must
be compared with the parallelepiped
volume V = a x b x h.
If one takes into account the intersection of the fracture plane with the
borehole, the fracture volume is
equal to :t xn x ( r / cosa) x r
and the borehole volume is equal
to: x ?h. For a horizontal fracture in
a cube (a=b=h=l m), the porosity of
fracture having an aperture of 1 mm
is equal to 0.1%. For a fracture with
an angle of 60" the porosity of fracture is equal to 0.2%. Comparing
with the borehole volume, the borehole having a radius of 10 cm, for a
fracture with a dip of 60" the porosity of fracture is equal to 0.2%. For a
vertical fracture crossing the borehole along a diameter the porosity
of fracture is equal to 0.636%
Fractures appear predominantly in brittle rocks, hence
in consolidated formations. Q~~~~~~~and dolomites are
the rocks in which fractures occur more frequently (Fig. 94). Very often fractures disappear on entering formations
which are more plastic (clays or halite), or friable (sands).
Review of general concepts
Relation of fracture to stress orientation
It has frequently been observed that the fracture system, or network, in a given region tends to have the same
orientation as the fault system and is related to the stress
orientation(Fig. 9-5a).
08
Figure 9-5a - The fracture orientation is linked to the principal stresse
orientation which control the fault type.
This is confirmed by the observations made on outcrops and in wells (Fig. 9-5b).
a
!
3
-
b
C
c
0
a
t
c
0
Figure 9-5b - Fractures and faults have almost identical strike patterns
as illustrated by the mapped faults (a) made on Buchan field, and the
strikes of measured fractures and veins in cores of wells 21/1-2 (b),
and 21/1-3 (c) (from Butler et al., 1976).
Figure 9-4 - Frequence of fractures as a function of rock composition.
Metamorphiques quartzites and dolostones fracture more easily than
quartz cemented sandstones and even limestones.
356
However, although the orientation may be statistically
significant, it must be remembered that there can be
considerable dispersion. This is a function of the fracture
types (Fig. 9-6).
Serralog 0 2003
Fractures
Chapter 9
357
strength varies continuously from a minimum in that direction to a maximum in planes perpendicular to it (Fig. 9-8).
The density of fractures (number of fractures per unit
area) is inversely proportional to the thickness of the rock
unit.
Maximum stress axis
F,
/
Angle of internal friction
I
Figure 9-6 - Schematic illustrations of the most common fractures
associated to fold. Type I fractures develop in gently folding anticline.
Type I1 fractures only appear on the anticline when the folding has progressed far enough to deform the rock elastically. Type 111 fractures
develop as a result of shear stresses. Type llla corresponds to tensional stress. Type lllb corresponds to compressional stress
(courtesy of Schlumberger).
Figure 9-8 - Relation of fractures to stress orientations. Planes of maximum shearing stress (S1 and Sd and planes of rupture (F, and FA
which form respectively an angle of 45" and an angle close to 25"-30"
to the maximum principal stress.
0.1
Ql
4
?I
As previously seen (cf. Chapter 8), in homogeneous
isotropic materials under compression, laboratory experiments have shown that compressive stress can be
expressed in terms of a set of three mutually perpendicular axes (Fig. 9-7).
Figure 9-9 - Occurrence of fractures under stress. Cases of single fractures :(a) and (d). Cases of symmetrically inclined sets (b) and (c)
(courtesy of Schlumberger).
B
C
D
Figure 9-7 - Principal stress axes, the stress ellipsoid (A) and conjugated shear fractures (B)or faults (C and 0).
When brittle rocks are stressed beyond their strength,
fracture results and takes the form of two sets of shear
plane. Their directions are bisected by the major stress
axis which forms an angle of approximately 25"-30" with
each of the shears (Fig. 9-8).
Strong anisotropy affects the rupture process. There is
probably a single plane of weakness and the shear
Serralog 0 2003
(a)
(b)
Figure 9-10 - Case of an anisotropic material. S :plane of weakness;
F :rupture plane oblique to S
(courtesy of Schlumberger).
357
Importance of fractures
In formations of low porosity and permeability, the production potential relies on an extensive system of open
fractures. The productivity will vary greatly according to
the number, extent and opening of the fractures and to the
porosity and permeability of the matrix.
As already mentioned, the porosity of fractures is insignificant in all but a few exceptional cases (highly compacted rocks), and makes no significant contribution to
the reserves. However, the presence of fractures may
significantly enhance the drainage surface, and thereby
the contribution of the matrix porosity to the production.
Open fractures considerably increase the permeability
but may cut the potential output of a reservoir if they are
not taken into account during the secondary recovery
phase.
A subvertical fracture system may be fed by an underlying reservoir.
Finally, in the case of fluid injection to maintain pressure, they act as preferred paths for the injected fluids
with the risk of isolating formation blocks which are still
hydrocarbon-saturated, and of having early production of
injected fluids.
Mechanical properties evaluation
Knowledge of the mechanical properties of a rock is
required in several domains.
where Pb, is the overburden pressure, assumed to be
equal to oz.In the simplified Terzaghi and "hard rock"
options a is assumed equal to unity.
Only elastic constrains determine o, = P&. The laws of
elasticity associate to this vertical stress a minimum horizontal stress o h , and the tectonic stresses are estimated
through the value of oHwhich can vary between o h (in a
non tectonic regime), and 0,.
J
The fracture initiation pressure Pb is a function of several parameters. It is expressed by the following relation:
where o h and oHare the minimum and maximum horizontal stresses respectively. oHis usually defined in terms
of the tectonic imbalance factor q / o h . Existence of tectonic imbalance can be inferred from borehole deformation
tests, or from break-out identification with the aid of multiple diameter caliper logs or, better, from borehole wall
images. Pore pressure is obtained from measurements
with the tester tools in new wells, or from pressure buildup tests in producing wells. T~ is the tensile strength. In
Terzaghi or "hard rock" options, a is assumed to be equal
to unity.
To compute the fracture re-opening pressure Pfr the
tensile strength is set equal to zero. So we obtain:
Mechanical behavior of the reservoirs - Stress comTIc)utations
To know if reservoirs require tubing or gravel packing,
or if they can be produced in open-hole conditions, or if
they will collapse, it is necessary to estimate the critical
wellbore pressure P., It can be demonstrated that P, is
expressed by the following relation using the MohrCoulomb failure criterion:
where a = 1 - c&b, C, and Cb being respectively the
rock compressibility (at zero porosity) and the bulk compressibility (with porosity), Ppis the pore pressure, zi is the
initial shear strength ( = zo), and v the Poisson's ratio. o h
is the minimum horizontal stress. It can be obtained assuming a horizontally constrained elastic model and is
expressed, following the Griffith and Mohr-Coulomb failure criteria, by:
7500
7600
Figure 9- 1I - Formation strength analysis (Schlurnberger's courtesy).
358
Serralog 0 2003
Fractures
The previous parameters are computed and displayed
in the MECHPRO program of Schlumberger (Fig. 9-12
two left tracks).
Chapter9
1
359
Table 9-1
Dynamic elastic parameters and how they can be
computed from wireline log data.
I
I
v
Poisson's Ratio (PR)
Longitudinal strain
G
shear Modulus
I
E Young's Modulus (YME)
Kb
cb
Bulk compressibility
C,
I
I
1/2(Ats/Atc)*
Volumetric deformation
Hydrostatic pressure
Rock compressibility
Change in matrix volume
(zero porosity)
Hydrostatic pressure
Biot Elastic Constant
Pore pressure
proportionality
-1
(Ats/Atc)2 - 1
I
Applied stress
Shear strain
( pb/ AtS2) X a
I
I
2G(1 + v)
Amlied uni-axial stress
Normal strain
Hydrostatic pressure
Volurnetric strain
Bulk Modulus
(with porosity)
YME
I
pb(l/AtC2 -4/3Ats2)a
I
PR
G
a
1 - (c&))
CB
Figure 9-12 - Display of mechanicalproperties (from Edwards, 1985).
Dvnamic elastic proDerties
Computation of some of the previous factors require
the knowledge of the dynamic elastic properties. If a sonic
waveform recording has been made using a Long
Spacing Sonic tool (LSS) or a dipole shear sonic tool, Atc
and Ats can be obtained. By combining these two data
with the corrected bulk density, it is possible to compute
the dynamic elastic parameters at each sampling level
(Table 9-1). This is achieved for instance by the MECHPRO program of Schlumberger.An example of the display
of the results is given in Figure 9-11 (right track).
Inherent strenath computations
The inherent rock strengths are computed, for instance, by the MECHPRO program of Schlumberger. They
are related to one another by simple functions expressed
below.
Initial shear strength zi
This parameter is derived by an empirical model based
on Deere & Miller's work (1969) and elaborated by Coates
& Denoo (1981).
'Ti
= 0.025E[
0.008Vclay
Tensile strength z,
The tensile strength is set at one-twelfth of Co as the
average value (Table 9-2): T' , = C0/12
In addition to these applications mechanical properties
evaluation can be used for:
- mud weight control to avoid hydraulic fracturing and
loss of circulation;
- drillability of the formation : adaptation of drilling
parameters, choice of rock bit, of the rotation speed,
weight on the rock bit ...;
- dip data interpretation by enabling a choice between
the faulting or folding of rocks.
Table 9-2
Uniaxial compressional and tensile strengths for rocks
Quartzite, Cheshire
Granite, Westerly
Diabase, Frederick
Sansdtone, Gosford
Marble, Carrara
CO
MPa
TO
461
229
486
50
90
28
21
40
3.6
6.9
(9-4)
- sin$)]
with $ = 30"
(9-5)
$ is the angle of friction in the Mohr-Coulomb failure
model. It is set at 30".
Serralog 0 2003
90
16.5
10.9
12.2
13.9
13.0
+ 0.0045(1 - Vclay)]/cb.106
I
Uniaxial compressive strength C,
Co = zi [2cos$/(l
CO
MPa
Granite
Sandstone
Marble
0.84
0.51
0.75
MPa
0.31
0.28
1.1
Detection of fractures on cores
This detection is generally quite difficult except if the
fractures are cemented. This difficulty is linked to the fact
that when open fractures exist into a formation, generally
the formation is more or less broken and, due to that, the
core barrel is quite often empty, the rock pieces falling
down in the hole.
Detection of fractures from well logs
With the exception of image tools, which can, in most
of the cases, see fractures directly, the responses of the
logging tools are affected only indirectly by the presence
of fractures, due to the “anomalies” on the normal sensor
response produced by fractures. It is only by these indirect effects that the fractures can be detected.
With this in mind, we will now examine, tool by tool, the
effects of fractures on their responses, and so get an idea
of the capacity of each tool for detecting them.
Natural Gamma Radioactivity
To the extent that the circulation of fluids may have
contributed to the precipitation of uranium in the fracture
system, the standard gamma ray tool, or the spectrometry of the natural gamma ray, will show increased activity
levels or increased uranium content in front of fractured
zones (Fig. 9-13).
NATURAL GAMMA RAY
SPECTROMFTRY
Similarly, a comparison between two successive
gamma ray measurements, the first with a non-radioactive mud, and the second over the same section after a
radioactive tracer with short lifetime has been circulated
briefly in the mud, may show up fractured zones. The tracers invade the permeable zones and cause the open
fractures to exhibit increased radioactivity. A further measurement made some time later, or after the start of production, should show decreased radioactivity over the
fractured zones.
NOTE: In cases of deep invasion, the start of production may cause
a temporary increase in activity by bringing the radioactive mud closer to
the borehole wall.
Caliper
Fractured zones may appear on the caliper curve as:
- either a reduction in hole diameter in compacted
zones which are in gauge, most probably due to a deposit of mud cake, especially if lost-circulation material has
been used (cf. Fig. 9-43);
- or an increase in hole diameter due to crumbling of
the fractured zone during drilling resulting in chunks of
various sizes falling away.
These phenomena can best be seen by a four arm
caliper or multi-fingers tool or dipmeters and image tools
rather than the standard two-arm calipers.
An increase on only one of the diameters is due to the
presence of fractures and follows their orientation (Fig. 914). The orientation can be obtained from the inclinometry measurement. The direction of elongation is often that
of a major system of faults and fissures, as has been
shown by various researchers (Babcock, 1978 (Fig. 9-15);
Bell & Gough, 1979; Cox, 1983).
Single steeply
Closely spaced steeply
. _
dipping fractures
Intersecting fractures
dipping fracture
Drill encounten vug
D
Figure 9-13 - Possible fractured zones in this Ordovician formation of
Algeria identified by uranium peaks not associated to thorium and
potassium peaks.
Those uranium peaks may also correspond to stylolites
(from Schlumberger, Well Evaluation Conference, Algeria, 1979).
360
Change in deviation
of hole
E
Figure 9-14 - Several reasons for hole ovalization in fractured zones,
most likely caused by the chipping of the formation by the drilling, and
the fall of the blocks of various sizes comprised between fractures
(courtesy of Schlumberger).
Another example (Fig. 9-16) coming from Timimoun
Serralog 0 2003
Fractures
I Chapter9
361
R 26
Figure 9-15 - On the left :remarkable consistency of hole ovalization over a large area.
On the right :relashionship between hole ovalization and direction of surface outcrop joints (Cretaceous to Devonian sandstones, Canada)
(from Babcock, 1978).
conductivity
Caliper
Figure 9-16 - On the left :dipmeter measurements allowing the analysis of the well geometv. Observe the borehole ovalization and also
the slowing down of the tool rotation. On the right :diagrams of tectonic ovalization in wells of the Timimoun Basin, Algeria
(from Schlumberger, Well Evaluation Conference, Algeria, 1995).
Basin in Algeria, illustrates this high consistency of hole
ovalization and their link with the direction of the current
major horizontal stress direction (N 135"-145").
Serralog 0 2003
Temperature
In a well the temperature gradient in the mud is affected by the presence of open fractures due to the invasion
361
of the fracture system by the drilling mud which has the
effect of cooling the formations. This phenomenon must
.not be confused with gas production which also causes a
drop in temperature.
The circulation of mud disrupts the normal distribution
of heat which depends partly on the difference in temperature between the mud and the formations, and partly on
the thermal conductivity of the rocks. The latter varies
considerably as each type of rock has its own thermal
conductivity (Table 9-3 and Fig. 9-17a). For this reason, a
thermometer log recorded immediately after drilling and
measured on the run-in can be a good indicator of the
types of rock encountered.
The mud at the bottom of the well is usually cooler
than the formations, while near the surface it is hotter (Fig.
9-17b). When circulation has been stopped for some time,
the mud temperature tends to homogenize by thermal
exchange, horizontally by conduction, and sometimes
vertically too, by convection. Thus, temperature changes
at all depths are slow, and some time is required before
the temperatures revert to their original values. Thus, the
mud becomes heated in the deeper part of the well. This
means that the temperature gradient of the mud intersects
the geothermal gradient at a certain depth (Fig. 9-17b).
Above this point of intersection the mud is hotter than the
formations, while below it is cooler. Consequently, mud
invasion in the upper zone increases the formation temperatures, while in the lower zones they are decreased.
Clearly, the interpretation of temperature logs must take
account of the position of this point of intersection.
Ternpemtdncreases
where there has been a partial or total loss of circulation.
Table 9-3
Thermal conductivity of the principal rocks.
Thermal Conductivity in 10-3 Calories/sec/cm/"C
2.8 - 5.6
Shale
3.5 - 7.7
Sand
Porous Lm. 4 - 7
Dense Lm.
6-8
0.9-13
Dolostone
13
Quartzite
1.2 - 1.4
Water
0.065
Gas
Gypsum
An hydrite
Halite
Sulphur
Steel
Cement
Air
Oil
3.1
13
12.75
0.6
110
0.7
0.06
0.35
b
Figure 9-18 - (a) : Temperature log showing
zones of cooling characterized by local
decrease in temperature compared to the
temperature gradient. (b) :Schematic diagram and logs after circulation of a cold fluid
(courtesy of Schlumberger).
Formation density
Figure 9-1 7 - (a) :normal earth temperature gradient. (b): temperature
profile after drilling of a well. Equilibrium is re-established after many
hours or even days (courtesy of Schlumberger).
When a cold fluid such as the drilling mud penetrates
the formation it displaces the formation fluid. The time
taken for the formation to revert to its normal temperature
will depend on the duration of circulation and on the degree of invasion (Fig. 9-18).
Zones which have been more deeply invaded will thus
appear as cooler zones on the temperature log. This will
be particularly noticeable in zones with open fractures
362
In the case of the compensated formation density tool,
two measurements may be considered : the density measurement itself, and the density correction.
- Being a skid-mounted device, the density tool may
face in different directions on two successive runs over a
fractured interval. One would then expect a drop in density if on one of these runs the skid was facing an open fracture. However, the dense, compact formations in which
fractures usually occur will produce low count rates on the
detectors, and hence a high level of statistical variations.
The resulting poor repeatability between successive runs,
which is a feature of high-density formations, whether
they are fractured or not, makes it impractical to look for a
variation in density as an indication of the presence of
fractures across one axis of the hole.
The fact that the tool is unidirectional and not free to
rotate does not simplify matters. However, it may be assumed that, if the hole is eccentric, the long axis will have
the same orientation as the vertical fractures, as long as
Serralog 0 2003
Fractures
these are more or less unidirectional.
- The readings of skid-mounted tools will be affected
by small depressions in the borehole wall which are the
result of small pieces of rock falling away. The short-spacing detector is more influenced by the mud filling these
small cavities than is the long-spacing detector.
- In zones where the caliper indicates a smooth borehole wall, the Ap curve will show a higher correction than
normal in the case of baryte muds (Fig. 9-19). This is
often accompanied by a very low density reading, but may
be localized, blurred or even hidden by the time constant
of the measurement circuit.
-The caliper may indicate sudden changes of hole diameter. When these changes are due to scaling of the formation wall, they can be "seen" by the short-spacing
detector.
I Chapter9 I 363
very high values (Figs. 9-19 and 9-20). This is due to the
high atomic number of barium compared to those of the
elements making up the majority of sedimentary rocks.
This property can be useful for estimating the porosity of
the fractures.
Caliper
Pb
Example 1
Figure 9-20 - Another example of fracture influence on density (Ap)
and photoelectric index measurement (Pe) (courtesy of Schlumberger).
Neutron Hydrogen Index
Figure 9-19 - Effect of baryte on the density correction Ap and on the
photoelectric capture cross-section index. Observe as well the caliper
(courtesy of Schlumberger).
Photoelectric capture cross-section
This measurement, which is made with the litho-density tool (LDT for Schlumberger), is more or less independent of porosity. Consequently it is of no use for detecting
fractures in normal muds.
However, the measurement is very sensitive to baryte,
and so can detect fractures which have been invaded by
baryte muds. When the pad of the tool passes a fractured
zone, the photoelectric capture cross-section will show
Serralog 0 2003
This measurement responds essentially to formation
content in hydrogen, and so it is a measurement of total
porosity. Since the porosity of fractures is usually small
compared to that of the matrix (e. g. in chalk or compacted clays), it is difficult to identify fractures because the
small variation in porosity is masked by statistical variations. In any case, because it is not a directional measurement, the neutron tools will give a more stable measurement. This is especially true in dense, compact formations because of higher count rates and lower statistical
variations.
Sonic measurements
Sonic measurements are known to be sensitive to
fractures in different ways, especially noticeable on full
waveform data. The effects of fractures on body wave
components (compressional and shear), and on Stoneley
waves are different.
Effects on bodv waves
The sensitivity of body waves to fractures depends on
the medium homogeneity on sound propagation. Any
heterogeneity whose dimension is not negligible with
363
respect to the wavelength will have an effect on its propagation, and therefore on the measurement. This is linked
to the large contrast between the elastic properties of the
fluid filling the fracture and the ones of the formation.
Fractures act as major discontinuities.
The effects on body waves include:
- time delay causing slowness differences between
receiver and transmitter modes realized with modern tools
with multiple receivers and transmitters including dipole
(Fig. 9-21);
- amplitude reduction (attenuation) of compressional
and shear waves;
- reflections causing chevron-patterns;
- mode conversions causing criss-cross patterns;
- borehole coupling enhancements causing amplitude
spikes.
Transmitter mode
time path. This is the case with subvertical fractures, or
more correctly fractures which are parallel to the tool axis,
and these are generally not detected by the sonic tool.
Common
receiver
position
Receiver mode
Interval
easured
ki
Fig 9-21 - Receiver and transmitter modes for computation of the
slowness of the sound waves and the Normalized Differential Energy
(courtesy of Schlumberger).
The Variable Density Log (VDL) plot on Figure 9-22
illustrates some of these effects indicated by arrows:
- a step in the arrival time of the compressional, representing a delay of about 20 ps on a depth interval of 3 m
which represents the transmitter-receiver spacing;
- a reduction of the shear amplitude;
- some distortion of the shear arrival caused by mode
conversion;
- a reduction of the Stoneley amplitude after 2500 ps;
- some interference patterns in the Stoneley wave caused by reflections.
Arrival time
In theory, the arrival time of the compressional wave is
unaffected by fractures which do not cross the shortest
364
Figure 9-22 - Effects of fractures on a Variable Density Log (VDL).
Observe the step in the arrival time of compressional wave, the shear
amplitude reduction and distortion, the Stoneley amplitude reduction
(courtesy of Schlumberger).
Whenever the fracture system is more complex, diffraction and reflection will attenuate the compressional
wave to such a degree that detection may not occur until
the second or third peak in the wave train, resulting in
erratic increases in the apparent travel time (so-called
"cycle -skips", Figs. 9-23 and 9-24). This phenomenon is
detected more easily with the older tools. New tools (i.e.
Schlumberger's Array Sonic Service, SDT, or the Dipole
Shear Sonic Imager, DSI) with multiple transmitters, including dipole technique, and multiple receivers are capable
of detecting cycle-skip conditions and may automatically
take steps necessary to avoid cycle skipping that may be
due to presence of fracture.
The shear wave velocity, on the other hand, is more
affected by fractures than that of the compressional wave.
It is seen to decrease while the compressional velocity
remains constant. Thus, by comparing Ats with Atc possible fractured zones can be identified when Ats increases
while Atc remains constant.
Attenuation of body waves
In general, the amplitude of an acoustic wave is
decreased when it crosses a fracture. This is the result of
a transfer of energy. The coefficient of transmission is a
function of the apparent dip of the fracture relative to the
direction of propagation (Fig. 9-25 left). Energy transmission across a fracture depends to a large extent on the
efficiency of mode conversions at the fracture interface.
Serralog 0 2003
Fractures
For acoustic energy to cross a fracture, a propagating
compressional or shear wave must be converted to a fluid
wave at the first fracture interface and then converted
back again at the second.
Transit time (pslft)
90
40
Figure 9-23 - Spikes and cycle skips on the sonic transit time measurement indicating a fractured zone.
Attenuated signal
TIME
Figure 9-24- Explanation of the cycle skipping development.
Fracture dip
Fracture dip
Figure 9-25 - On the left : variations of the coefficients of transmission
as a function of the apparent dip angle of a fracture plane with regard
to the propagation direction (blue and red lines from Knopoff et al.,
1957; green line from Mom's et al., 1964).
On the right :attenuation across a fracture as a function of the angle of
dip of the fracture for waves with a vertical trajectory
(from Moms et al., 1964).
Obviously, the inclination of the fracture is crucial here.
Figure 9-25 (right), from Morris et. al. (1963), is based on
Serralog 0 2003
1
Chapter9
I
365
experimental results and shows that compressional
waves suffer little attenuation on crossing fractures which
are parallel or perpendicular to the tool axis. The attenuation is high when the angle is between 35" and 80". Shear
waves on the other hand, are strongly attenuated by fractures at low angles. According to J. Gartner (in a personal
memorandum) this contrasting behavior could suggest a
conversion from one mode to the other (compressional to
shear) for certain values of inclination of the fractures. The
attenuation decreases with increasing dip. It becomes
very small when the dip of the fracture is above 65" (25"
to the axis of the tool or borehole).
Normalized Differential Energies (NDE)
Loss of acoustic energy, especially of shear and
Stoneley propagations, is also a fracture indicator as illustrated by Figure 9-26. This loss, expressed in dB/ft, is calculated as the difference, known as the Normalized
Differential Energy (NDE), between the 1 foot BHC NDE
and a baseline calculated as the average over 50 ft. The
computation of the difference between energies measured at two distinct receivers, divided by the distance between them is done, in practice, for 1 foot spacing between
the different receiver locations for a single position of the
transmitter (AERx).Similar computations can also be performed in transmitter mode (AERx) (cf. Fig. 9-21).
Receiver and transmitter mode NDE measurements are
affected in opposite direction by borehole enlargements
and formation coupling changes. Their average yields a
BHC Normalized Differencial Energies measurements,
AEBHc,which is practically free from effects of hole size
changes and lithology variations. These three curves are
displayed with a shading between receiver and transmitter curve to indicate the difference magnitude. Fractures
are detected as sharp negative peaks. The NDE computation with ompressional and shear are more noisy than
with Stoneley.
Amplitude spike analysis
Right at a fracture (Fig. 9-27) there is far better acoustic coupling between the borehole and and the formation
than elsewhere. In principle this causes two momentary
increases in amplitude as the tool passes a fracture. The
two peaks should be separated by the tool spacing and,
as previously discussed, the amplitude between them
should be reduced. The analysis starts by computing for
each amplitude a spike curve in order to pinpoint the
depth of local enhencements. For each amplitude curve,
a slow-moving average is computed and then substracted
from the amplitude curve itself, leaving positive and negative departures from the norm. Only positive departures
are kept. As seen on Figure 9-28, pair of spikes on each
curve are centered around the same depth (2655 m). The
spacing between spikes increases with transmitter-receiver spacing. The spikes at the top correspond to the receiver spikes, the spikes at the bottom to the transmitter spikes. The fracture depth is at the center of the two receiver-transmitter spikes.
This analysis has a sharp vertical resolution.
365
Next Page
Figure 9-26 - Energy losses of compressional, shear and Stoneley waves, by computation of the Normalized Differential Energies (NDE), reflection
coefficient and fracture aperture computation from the waveform reproduced on the right (courtesy of Schlumberger).
Figure 9-27- The acoustic coupling between the formation and the borehole is
improved when transmitter or receiver
passes a fracture. This generates an
amplitude spike
(courtesy od Schlumberger).
Crisscross patterns
They are the direct result of mode conversions at the
fracture interface, each set corresponding to a given pair
of propagation modes (Fig. 9-29).
The distinguishing characteristic of any crisscross pattern is the Slope of the two symmetrically juxtaposed heS.
This is a simple function of the velocities of the two propagating modes under consideration:
306
J
Figure 9-28 -Amplitude spike analysis. Compressional amplitude curves are displayed on the left track, spike curves on the middle track,
and spike analysis indicating fractured zones
(courtesy of Schlumberger).
Serralog 0 2003
Previous Page
Fractures
1
Chapter9
I 367
so it is more straightforward.
- The Stoneley wave, being mainly influenced by borehole fluid, does not react much to changes in lithology.
Thus, a strong Stoneley reflection most likely indicates an
open fracture, not a bed boundary.
Variable Density Display
Transmitterreceiver
spacing
Compressional
Shear
StoneIey
Figure 9-29 - Mode conversions at a fracture generate crisscross patterns in the VDL display, with length equal to transmitter-receiver spacing. Note interval transit time lengthening on compressional and
shear. Compressional/Stineley crisscross is not shown as it would
superimpose itself on top of what is already illustrated
(courtesy of Schlumberger).
Focussing on compressionallshear crisscross, which
is less cluttered by direct arrivals than later pairs, the first
step is to determine the compressional and shear speeds
using conventional waveform analysis. This enables the
computation of the crisscross slope. Then, for each given
depth, a window, equal to the transmitter-receiver spacing, is defined with that depth as its midpoint. The measurement of the crisscross energy contained in this windowed set of waveforms is achieved, first by normalization of the waveforms, second by applying a median filter
in order to remove the bulk of the direct arrivals, third, by
summing up the total energy following each crisscross
line knowing the crisscross slopes and a suitably chosen
time window. As the window moves up the well, a log of
crisscross energy is plotted. The vertical resolution of this
crisscross analysis is about two feet.
Effects on Stonelev waves
The Stoneley wave, and especially its low frequency
component known as the tube wave, is a borehole fluid
mode that propagates as a pressure wave along the borehole.
The way fractures affect the Stoneley wave is quite different compared to the way they affect compressional and
shear waves. Acoustic energy is not lost through inefficient mode conversions, but more as a result of moving
the fluid in the fracture system, resulting in a pressure
drop in the borehole. As a result, the direct Stoneley wave
is attenuated, and a reflected Stoneley is generated (Fig.
9-30).
Three advantages of the Stoneley wave analysis can
be considered.
- In fast formations, where we generally look for fractures, Stoneley wave amplitude is much higher than the
other two arrivals (compressional and shear, Fig. 9-31),
Serralog 0 2003
Borehde
Figure 9-30 - The Stoneley energy
attenuation is a function of the fluid
movement in the fracture, irrespective of dipping angle. The fluid is
pumped in and out of the fracture,
thus dissipating energy. The attenuation is a function of the fracture
permeability, therefore its width or
aperture, and the fluid viscosity
(generally mud).
T for transmitter, R for receiver
(courtesy of Schlumberger).
- The roughly constant Stoneley velocity eases the
signal processing task of measuring the reflected signal.
Figure 9-31 -Amplitude and frequency content of the sonic waves.
The amplitude of the Stoneley wave is much higher in the low frequen.
cy region allowing its detection
(courtesy of Schlumberger).
Stoneley reflections
A wave sent by the transmitter is capted by the receiver, but propagates further and can be reflected down by
a fracture and detected a second time later (Fig. 9-32).
Slope =
2
v
Time
Figure 9-32 - Effect of reflections on waveforms
(courtesy of Schlumberger).
367
368
Well Logging and Geology
The time difference between the direct and the reflected arrivals depends on the distance between the receiver
and the fracture, and on the slowness of the non fractured
formation. It changes linearly as the tool moves up closer
to the fracture, and becomes nil when the receiver faces
the fracture. In this case a fraction of the down going
signal can be reflected up and received later. Both effects
result in the chevron patterns visible on the VDL plots
(Fig. 9-33). They are not necessarily visible on compressional and shear waves because the reflection has a
smaller energy and may be masked by the direct arrivals.
The reflection coefficient, r, which is linked to the
transmission coefficient, t, is an important parameter related to the fracture aperture:
r+t=l
600
Frequency (Hz)
Figure 9-34 - Theoretical influence of fracture aperture and dip on the
reflection coefficient r (fcourtesy of Schlumberger).
Resistivities
The electrical system consisting of the formation, the
borehole and the fracture network is represented by the
diagram in Figure 9-35. The fractures are assumed to be
subparallel to the borehole axis and invaded by a conductive fluid.
Figure 9-35 - Fracture identification
from resistivity tools. The induction
“sees”the fractures in series, the
laterolog tools “see”the fractures in
parallel
(courtesy of Schlumberger).
700
800
Figure 9-33 - Effects of reflections on VDL plot. The chevron patterns
are well detected on Stoneley arrivals, exceptionally on shear. The
analysis of the waveform allows the computation of the reflection
coefficient r. The higher values of r mean more permeable fractures
(courtesy of Schlumberger).
The theoretical prediction of r for horizontal and dipping fractures is shown on Figure 9-34.
The reflected Stoneley arrival is separated from the
direct arrival thanks to,a velocity filter. Once the direct and
reflected Stoneley arrivals have been separated, they can
be stacked to obtain the direct and reflected wavelets at
this level. Finally, the reflection wavelet on a certain number of levels, below the fracture, are extrapolated to
obtain the reflection wavelet where the fracture intercepts
the well. The maximum value of the envelope at this location is the final reflection coefficient.
368
Taking into account the current distribution for each
type of device, it will be observed that, in the case of fractures which are subparallel to the borehole axis:
-the induction is unaffected by the fractures which only
constitute a negligible part of the whole circuit since they
are in series for the Foucault currents;
- the electrode tools will be strongly affected because
the fracture network presents paths of lowered resistance
which act as shunt resistances to the current.
In the case of fractures which are subperpendicular to
the borehole axis:
-the induction will be strongly influenced because now
the fractures are in parallel rather than in series, and their
conductivity is very high compared with that of the surrounding formations;
- for the other tools, these fractures continue to offer
paths of lowered resistance.
Thus, a comparison of resistivity values from induction
and electrode tools in zones containing subparallel open
fractures will show substantially lower resistivities on the
laterologs than on the induction (Fig. 9-36). However, we
must bear in mind that the induction measurement is not
recommended in resistive, compact formations because
of low signal level. The analysis will therefore rely on the
relative behavior of the two laterologs (deep and shallow)
and of the microdevices.
Serralog 0 2003
Fractures
~
Resistivity (ohm-m)
+
=igure 9-36 - Comparison between
!he responses of the induction and
laterolog in a fractured zone
(courtesy of Schlumberger).
)) Suau's model
3gure 9-37 - Current distribution in
he case of a fracture which is subparallel to the borehole axis. a) :
Clavier's model; b) Suau's model
(courtesy of Schlumberger).
Chapter 9
369
fluid (gas, oil or fresh water), the resistivity of the LLs will
be substantially less than that of the LLd (Fig. 9-37).
If the mud is less conductive than the original fluids in
the fractures, the separation of LLs and LLd is much less
and may even be inverted.
In compact zones of low porosity which are not fractured, and therefore with little invasion, the two measurements will read about the same resistivity (Fig. 9-38, top
interval).
Because they are pad-mounted, the microdevices only
respond to fractures in front of the pad. But because the
borehole wall tends to crumble near the fractures, it
becomes ovalised, and the pad tends to ride the low side
of the major axis. Hence, the probability of following the
fracture network is increased. Clearly the presence of
fractures will strongly influence these devices because of
their small volume of investigation. Moreover, this part of
the fracture system will be invaded by mud or mud filtrate, and so the resistivities will be much lower (Fig. 9-38,
bottom interval). In addition, crumbling of the borehole
wall will create zones of current leakage.All this enhances
the difference in the resistivity readings
Dipmeter
All dipmeter tools consist of arms, extending from the
tool in different azimuths, so that the pads attached to the
ends of the arms are in contact with the formation (Figs.
9-39 and 9-40). High-resolution sensors on the pads provide the logs necessary for heterogeneity or surface
detection.
HDT 1969
J
Figure 9-38 - Example showing the responses of the laterologs and the
MSFL in a fractured zone (courtesy of Schlumberger).
When the fratures are subparallel to the borehole axis,
the apparent drop in resistivity becomes more pronounced with decreasing depth of investigation although it
remains constant within a fracture. Consequently the deeper-reading device is less affected by the fracture than the
shallow-reading device. A ratio of 1.5 to 2 is commonly
observed between RLLd and RLLs. Moreover, if the
drilling mud is more conductive than the original formation
Serralog 0 2003
Figure 9-39 - The dipmeter tools
(courtesy of Schlumberger
The circuitry for the electrode output is arranged so
that the curve deflections are proportional to the button
current. This button current varies widely according to the
contrast between the resistivity of the formation in front of
the button and the formation surrounding the tool.
Because the voltage is ignored, the curves are recorded
with a "floating zero" on a nonlinear scale designed to
accommodate large variations in local resistivity.one pad
Several parameters must be analyzed with this type of
tool.
369
Figure 9-40 -Another type of dipmeter
tool :the Six Arm Dipmeter of
Halliburton
(courtesy of Halliburton)
Figure 9-42 - FIL presentation for two different orientations of the HDT
sonde. Observe also the caliper difference and the variation of the tool
rotation
(courtesy of Schlumberger).
Resistivitv Curves
As with all the pad-mounted microdevices, only the
pads which are in front of the fractures will be affected and
show a drop in resistivity (Fig. 9-41).
Figure 9-41 - Diagram explaining the detection of an open fracture by
one pad of the HDT (on the left) and the SHDT (on the right).
If the hole is ovalized because of fractures, the usual
orientation of the tool will be with two of the four arms
across the major axis, the other two being perpendicular.
Thus in compact, fractured formations, the two opposite
pads, which "see" the fractures, will show a drop in resistivity, while the other pair, which does not see them, show
a high resistivity value with little or no curve activity (Fig.
9-42), assuming that a low EMEX value has been used.
Superimposing the resistivity curves of two adjacent
(i.e. 90" apart) pads will reveal fractured zones whenever
there is a separation between the two curves. A visual
representation of the presence of fractures is obtained by
shading between the two pairs of adjacent curves (Fig. 943).
This technique was known as Fracture ldentification
Log (FIL), and this presentation could be obtained at the
well-site using the CSU system.
370
Figure 9-43 - Reduction in hole diameter and slowing down in the rate
of rotation of the tool in fractured zones (courtesy of Schlumberger).
Unfortunately, the FIL is often confused by sedimentary features such as laminations, flasers or pebbles, and
the majority of the shaded areas correspond to beds with
an apparent dip rather than to fractures (Fig. 9-44).
This problem has been eliminated with the introduction
by Schlumberger of a new program known as DCA
(Detection of Conductive Anomalies). Conductive events
which cannot be correlated are searched for, and only
these can be interpreted as possible fractures. The events
are defined during GEODIP processing. The conductive
anomaly is then reproduced only if the following conditions are satisfied
- the conductivity exceeds a certain value;
- there is a sufficient difference between the conductivity values;
- the anomaly is detected on a minimum number of
successive intervals.
Serralog 0 2003
Fractures
1
Chapter9
~
371
malies are then indicated along the corresponding azimuth curve. The available fracture indicators with this
presentation include:
- the conductive anomalies revealed by the DCA program;
- borehole rugosity and the axis of ovalisation;
- changes in the speed of rotation of the tool.
J
Figure 9-44 - Influence of apparent dip of beds on the FIL presentation
(courtesy of Schlumberger).
The two electrodes by pad existing in the SHDT tool
gives even better detection of fractures by comparing the
measurements of the two buttons on the same pad (Fig.
9-46). In certain favourable cases, the dip of the fracture
can even be determined (Fig. 9-47).
The three thresholds can be set by the log analyst and
so adapted to local conditions. The results are presented
in the form of a log. The azimuths of pads 1 and 2 are
displayed against depth in the left-hand track (Fig. 9-45).
The shaded areas indicate a difference between the
nominal hole diameter and the readings of the two calipers.
AZ=60"
Figure 9-46 - Display of the SHDT curves showing their interest to
detect fractures. Observe the very good repeatability of the curves
recorded by electrode 1 and the speed electrode. The fractures, indicated by f., correspond to intervals showing a local increase in conductivity of one electrode compared to the other of the same pad
(adapted from Schlumberger).
A polar frequency plot of the conductive anomalies is
also provided (Fig. 9-48). It is used to determine the direction of the fracture network or networks. This direction is
related to the axis of maximum constraint and to the general orientation of the faults in the region.
n
Figure 9-45 - Example of the DCA presentation
(courtesy of Schlumberger).
The azh-~uthsof Pads 1, 2, 3 and 4 are displayed
against depth in the right-hand track. The conductive anoSerralog 0 2003
Azimuth curve of Dad 1
When the hole is not very ovalised, the tool will rotate
because of the torque in the logging cable. The fractures
are then seen successively by the different pads (Fig. 9-49).
As we have seen, the tool normally rotates as it travels
uphole. Any slowing down, stopping or change of direction
in the rotation usually indicates the presence of fractures.
This phenomenon is the result of the pad following a sort
of subvertical or oblique pathway created by crumbling of
the fractured zone for a certain distance (Fig. 9-50). The
tool then resumes its normal rotation, usually after a brief
period of more rapid rotation to release the torsion which
has built up in the cable.
371
t
Stsaogrrun of fractuns
Of conductive
Figure 9-47 - (a):
malies detected by the SHDT tool.
J
Figure 9-49 - A fracture seen successively by two different pads. Note
that the azimuth is constant (courtesy of Schlumberger).
(b): They can be correlated to determine the
dip and the azimuth of the fractures
(courtesy of Schlumberger).
Neutron (pa.)
+
Scale 1% brbivision
Figure 9-48 - Example of a polar frequency plot which provides a
means of orienting the fracture network (courtesy of Schlumberger).
Caliper
Since the dipmeter tool has two measurements of diameter 90” apart, comparison between them will reveal
any hole ovalisation, sudden variations in diameter, or
restrictions due to deposits of mud cake or lost circulation
material in the fractured zones (cf. Figs. 9-16 and 9-50),
Figure 9-50 - Example of reversed tool rotation in fractured zones.
Observe as well the caliper and the Ap curves
(courtesy of Schlumberger).
When the GEODIP program is used for the HDT tool,
or the LOCDIP program for the SHDT tool, there is a noticeable absence of four-pad dips. There may, however, be
some dips which are erratic in dip angle and azimuth
which are due to three-pad correlations. In certain favourable cases (e. g. a single fracture), the conductive peaks
can be correlated to give the dip of the fracture (Fig. 9-47).
Spontaneous Potential
DiDs
In compact fractured formations, the fractured zones
can be identified from the CLUSTER program for the HDT
tool, or the MSD program for the SHDT tool by examining
the values of erratic dips or dips of poor quality.
Correlations which are due to conductivity peaks have no
reason to produce dips which are consistent in either dip
angle or azimuth.
372
Negative “anomalies”, especially if they are important,
are sometimes observed in fractured zones. This is often
explained by the development of an electrofiltration potential when the formation has been drilled with a fresh mud
(salinity of less than 5,000 ppm).
Pyrite filling a fracture could also generate SP deflections.
Serralog 0 2003
Fractures
I
Chapter 9
I
373
Ultrasonic image tools
BoreHole Televiewer (BHTV)
This tool (Zemanek et a/., 1969) provides an acoustic
image of the borehole wall (Fig. 9-51). It is obtained by
measuring part of the acoustic energy reflected from the
borehole wall. The same piezo-electric transducer acts as
both transmitter and receiver.
The formation is more reflective when the rock is
smooth and compact. When it is rugose, fractured or
vuggy, the acoustic energy is more dispersed and these
irregularities then appear as darkened areas on the film.
In the BHTV tool two parameters can be used for fracture detection, the amplitude of the received signal and its
transit time. The amplitude of the signal is reduced due to
the dispersion of energy at the edges of the fracture, while
the transit time will be increased (Fig. 9-52). This tool provides, then, not only a detection of all the open fractures,
but also their orientation and dip. The only requirement is
to minimize the amount of material in suspension in the
mud to avoid having a speckled image due to dispersion
of the energy. Other adverse conditions to be avoided are
excessive mud-cake, excessive hole ovalisation or gascut mud.
Amplitude
Borehole
"Tnsducer
Receiver
Figure 9-52 - Ovalization and breakouts generate reflections and diffractions which reduce the signal
amplitude received at the receiver
and increase the transit time as
shown by the multiple reflections
of a signal not perpendicular to the
borehole wall before its am'val at
the receiver.
An interval of 5 feet, recorded in 1 minute, contains
180 scan-levels, each level corresponding to 250 measured points. This corresponds to a measurement each 2.7
mrn horizontally and 8.38 mm vertically in a 8.5 in. holediameter.
The BHTV and its more recent version (Acoustic
TeleScanner de Schlumberger) has been replaced by
modern tools such as Ultra-sonic Borehole Imager (UBI)
of Schlumberger, the Circumferential Acoustic Scanning
Tool (CAST) of Halliburton, and the Circumferential
Borehole Imaging Log (CBIL) and Simultaneous Acoustic
and Resistivity Imager (STAR) of Baker Atlas (Fig. 9-53).
Filtered transit time
Compensating device
Motor assembly
Gear box assembly
Rotatin electrical
m n J m
Centralizer
,
&i=R?S#2
Rotating seal
Figure 9-53 - On the left
the UBI* tool
(courtesy of Schlumberger).
On the top right the STAR tool
(courtesy of Baker Atlas).
.Transducer
Interchangeable
mtating sub
-7.5 rps
-
Figure 9-51 Fractures can be detected by both the amplitude and the
filtered transit time recorded by the borehole televiewer
(courtesy of Schlumberger).
Serralog 0 2003
These tools are fundamentally sensitive to the borehole micro-rugosities. These micro-rugosities may either
reflect actual geological events (vugs, stratifications, fractures, etc.), or scratches generated by the rock-bit or the
turbine. In that case other information is necessary.
Figure 9-54 illustrates fractures seen by the UBI tool
compared to the resistivity images provided by the resistivity images tools.
373
FMI
ARI
UBI
The borehole coverage is a function of the borehole diameter (Fig. 9-56). In a 8.5 in. diameter the FMi covers
80% of the borehole surface and the EM1 60%.
Description of the
EM1 pad
Photograph of the
OBMl
Figure 9-54 - Three types of borehole Wall images. Observe the good
agreement between the FMI and UBI images. The ARI images
confirms the most important features (courtesy of Schlumberger).
With these tools and the resistivity image tools, for the
first time it was possible to “see” the formation and consequently better describe it. They provide information allowing a much better detection of fractures.
Figure 9-55 - The resistivity imaging tools. At the top, the Formation
Microlmager (FMI) tool of Schlumberger. Below, photograph of the
Halliburton’s half EM1 tool and of a pad. On the right, the photograph of
the Oil Base Mud Imager (OBMI) tool of Schlumberger.
Figure 9-56 - Evaluation of the
borehole coverage of the FMI
tool as a function of the borehole
diameter
(courtesy of Schlumberger).
2
Resistivity image tools
The first resistivity image tool, the Formation
Microscanner (FMS-2 pads), was introduced by
Schlumberger in 1986. It was followed by the FMS-4 pads
in 1988. In 1991 Schlumberger introduced the Formation
Micro-Imager tool (FMI), with 4 pads and 4 flaps. This tool
has 192 electrodes of 5 mm diameter, 24 in 2 rows on
each pad and flap (Fig. 9-55). More recently (2002),
Schlumberger has introduced on the market the Oil Base
Mud Imager (OBMI). Halliburton has developed the EM1
tool (Fig. 9-55), and BakerAtlas the STAR tool (Fig. 9-53).
The EM1 and STAR tools have150 electrodes divided in
two rows on 6 pads.
When one of the button electrodes, on these pads or
flaps, of these tool passes an open fracture in the formation, the current it emits will take the least resistive path.
This will be reflected on the corresponding conductivity
curve as a sharp increase, while the images will represent
open fractures as one or several dark irregular lines (Fig.
9-54).
One of the major advantages of these tools is the lateral coverage it provides due to the large number of electrodes and the number of pads with the shift of each row
compared to the other row by half the electrode diameter.
374
Figures 9-57 and 9-58 illustrate how an open fracture
can be identified by resistivity image tools.
Pad
Formation
Average surrounding
bed resistilty
R*
Areaofthepeak
Electrode’
Figure 9-57 - When the
electrode crosses an ooen
fracture the current will
preferably flow through the
fracture which is less resistive than the surrounding
non-fractured formation.
This creates a conductive
peak with a size similar to
the electrode diameter and
a deflection function of the
mud conductivity and the
fracture aperture.
Healed cemented fractures can also be detected, if the
resistivity contrast with the surrounding rock is sufficient.
These appear as white irregular lines on the images (Fig.
9-59).
Serralog 0 2003
Fractures
Cms-section
Chapter 9
375
Front view
Figure 9-60 - Two other resistivity
tools providing images of the
borehole wall (cf Fig. 9-54)
(courtesy of Schlumbergetj.
RAB
The images provided by these tools have not the vertical resolution of the previous tools but have a higher
depth of investigation giving a better idea of the fracture
prolongation inside the formation.
Borehole wall
Figure 9-58 - Diagram explaining how long an oblique open fracture
will be detected by the two rows of a pad.
Figure 9-59 - (a) Two cemented conjugated white features intersect on
the left strip with an angle of 130". They form an angle of 25" with the
vertical to the bedding. This set of observations indicates that they correspond to conjugated shear fractures with the greatest principal stress
axis perpendicular to the bedding (practically vertical). The strike of the
fracture planes is about N 160" which is also the direction of the mean
principal stress axis. The apparent dips of the fracture planes are 60"
N 170" and 70" N 230". (b) The core photograph confirms the
presence of cemented fractures (courtesy of Schlumberger).
Other resistivity measurements allow the detection of
fracture either during the drilling itself - Resistivity At the
Bit or RAB tool - or by wireline technique - Azimuthal
Resistivity Imager ARI and High-resolution Azimuthal
Laterolog Sonde (HALS) of Schlumberger (Fig. 9-60).
Serralog 0 2003
Other examples of fractures are given in Figures 9-61
to 9-63. They come from FMS, FMI, RAB or STAR tools.
Imager (Pai (L Flap)
Figure 9-61 - On the left :
conjugated shear open fractures in a granite. Few induced
fractures are also visible in
some places {courtesy of
Schlumberger). On the right :
cemented fractures in laminated formations(top :courtesy of
Baker Atlas. Bottom: courtesy
of Schlumbergetj.
As it can be easily understood, the image interpretation must take into account the well orientation (vertical,
deviated or horizontal). As illustrated by Figure 9-64 recorded in a horizontal well a fracture may appear practically
horizontal when in fact it is vertical, and shale layers with
high dip when practically horizontal. To be complete mention must be made of images provided by Azimuthal
Density Neutron (ADN) tool which measures density and
photoelectric index in four quadrans.
375
Next Page
' Figure 9-64 - The appa-
'
shale
layers
rent dip of the conductive
dark beds is in fact equal
to 1.7" N 90" and the
fracture is practically vertical in this horizontal
well. The actual orientation of the well is illustrated by the lower figure
(courtesy of
Schlumberger).
fracture
Figure 9-62 - On the left :open fractures in a dolostone with computation of their mean aperture. On the right :calcite-cemented fractures in
a limestone with determination of their dip. Histograms of the fracture
strikes are shown at the top. Fracture density is reproduced on the
right of each image (from Delhomme and Motet, 1993).
They rarely allow the detection of individual fractures,
only indicating the presence of fractured zones. But the
variations in tool response due to fractures could also be
caused by other phenomena. The following procedure is
recommended to be sure of the origin of these variations.
- It is necessary first of all to look for these variations
in intervals which are likely to be fractured. These may be
zones in which there has been a loss of circulation or an
inflow of fluids, or consolidated formations such as compact chalks, limestones or dolostones, quartzites, anhydrites, plutonic or metamorphic rocks. In general terms, it
is zones of high resistivity which are of interest, and not
porous, unconsolidated sands or plastic clays.
- The next step is to note all possible occurrences by
identifying on each available measurement all the phenomena which could be attributed to fractures (Fig. 9-65).
Tight bed
Open fracture
Figure 9-63 - Open fractures detected by the RAE tool, confirmed by
the FMI. Observe on the FMI images a local small shift between the
pad and the flap images (6 & 7 strips)
(courtesy of Schlumberger).
As illustrated by these numerous figures, image tools
allow a direct detection and analysis of the fractures.
Other logging tools are not capable of detecting fractures
themselves'
but by the effect that the fractures have On
the log measurements.
376
Figure 9-65 - Several effects on different measurements confirm the
presence of fractures (courtesy of Schlumberger).
Serralog 0 2003
Previous Page
Fractures
I Chapter 9 I 377
~~~
The probability of fractures is in fact much greater than
the phenomena observed on the logs may indicate. Thus,
if several of the phenomena already described are detected, it is reasonable to conclude that fractures are present.
Schlumberger had developped a program for the
detection and evaluation of fractures. It was commercially available under the name of DETFRA. This program
(Boyeldieu & Martin, 1984)grouped all the known fracture indicators into five categories : electrical, acoustic,
radioactive, electromagnetic and multi-pad.
Each log was analyzed, and a fracture probability was
estimated using certain criteria (threshold, median and
maximum probability (Fig. 9-66).The probabilities were
then combined using bayesian logic. Thus, two criteria
with individual probabilities PI and P, would have a combined probability which was given by:
P = 1 - (1 - P,)(1
- P2)
(9-7)
Fracture evaluation
The evaluation of fractured zones requires the following information
- depths of the fractured zones;
- types of fractures : natural (open or cemented) or
induced;
- orientation (dip and azimuth) of fractures;
- vertical and lateral extent of fractures;
- total fracture length per unit volume;
- fracture density: number of fractures per unit length;
- fracture aperture;
- fracture porosity;
- fracture permeability.
Well logs do not provide all of this information, only the
following being obtainable.
Depths of fractured zones
This is the simplest information to obtain from the logs,
especially from the image tools. So there is no need to
elaborate.
L
P ~ ~ Y D
Ovality 1-3 2-4 in inch
Type of fracture
The image tools can usually differentiate between
open fractures, fractures induced by the drilling process,
and healed cemented fractures (Figs. 9-54,9-59,9-61,962,9-63and 9-68and 9-69).
Figure 9-66 - Fracture probability in the case of a caliper measurement
(from Boyeldieu et al., 1984).
This rule was associative, and could be extended to an
unlimited number of probabilities. The results were presented in the form of a log (Fig. 9-67).
c
o
y
:&
v
*
"
Figure 9-68
Figure 9-69
Figure 9-68 - Open fracture and stylolites. Figure 9-69 - Few conjugated shear fractures and induced fracture in the middle and on the sides
of the images (courtesy of Schlumberger).
Figure 9-67 - Example of results obtained with the DETFRA program
(from Boyeldieu et al., 1984).
Serralog 0 2003
For the other tools, only open fractures will affect the
log responses and be detected. In any case, it is only
open fractures which are of interest for production. Hence
every fracture which is detected as a conductive anomaly
377
is by definition open. However, not every conductivity
peak is a fracture.
Construction
of the fracture
network
and
Fracture orientation
3D visualization
There are two parameters to be determined: dip and
azimuth. The image tools are the only tools which allow
the determination of both the orientation and the dip of
fractures (cf. Figs. 9-59, 9-61 & 9-62).
The dip cannot be determined with any certainty from
the other measurements, because even if a correlation is
made between conductivity peaks, there is no guarantee
that they all belong to the same fracture.
If one now considers the size of an event detected by
a pad (Fig. 9-70), one can attempt to define two possible
dips and select the ones which show the most constant
values. These data must also be plotted as a function of
pad azimuth.
First network
’
Second network’
Figure 9-71 - The fracture network can be constructed. This allows the
determination of the fracture and stress orientation
(from Delhomme & Motet, 1993).
The fracture trace length (F,,,,) is the cumulated fracture trace length seen per square foot of borehole wall. It
is computed by the following equation:
(9-8)
~~
with :
R = borehole radius
C = borehole wall coverage by pads as a fraction
H = height of the sliding window
z = depth
JzH(i)is equal to 1 if the fracture i lies within H centered at depth z, otherwise is equal to 0.
Figure 9-72 explains how fracture length can be detected and computed. Figure 9-73 illustrates a case of interupted fractures.
Figure 9-70 - Determination of fracture dip from the length of conductive anomalies on a single pad. Several explanations are possible
(courtesy of Schlumberger).
The azimuth can be determined if fracturing is accompanied by hole ovalisation, or from a polar frequency plot
of conductive anomalies detected by the DCA program.
Figures 9-15 and 9-16 show the consistency of results,
and their correlation with the predominant fracture or fault
directions.
The two buttons on each pad of the SHDT provide a
means of determining the apparent dip of the planes of
the fractures picked up by each pad. The dip and azimuth
of the fractures can then be defined if we assume that the
two anomalies correspond to the same fracture, or at
least to the same system of parallel fractures (cf. Figs. 947 & 9-48).
With image tools a fracture network can be realized
(Fig. 9-71) and the paleo-stress orientation determined.
Fracture length
As previously illustrated individual fractures can be
identified with the image tools. This allows the determination of the number of fractures in a given window, and of
their length (Fig. 9-72).
378
Figure 9-72 - Diagram
explaining how the fracture
length can be computed.
Fracture density
The corrected fracture density (FD,) is the number of
fractures observed per foot inside an interval of height H
corrected for the orientation bias. It is given by the following equation:
(9-9)
Serralog Q 2003
Chapter9
in which :
Bi = apparent dip of fracture i
IzH(i)is equal to 1 if the fracture i lies within H centered
at depth z, otherwise is equal to 0.
cubes summed over an interval of height H:
I-CJJ.h:
y
A F & ~
(9-11)
JZHOI,
The area corresponding to the conductive peak linked
to the fracture (cf. Fig. 9-57) allows the estimation of the
fracture aperture and through that the evaluation of the
fracture porosity (Fig. 9-75).
Figure 9-73 - Example of fractures interrupted when joining a plastic
(shaly) bed. They are detected and their aperture determined all along
their trace. Conductive layers are also isolated in the processing
(courtesy of Schlumberger).
Figure 9-75 - The additional current (A) created by an open fracture
filled by the conductive mud (R,) divided by the fracture aperture
(Win mm) is function of the ratio of R, (or RJ over R, .
Hole axis
Apparent
interval
between
real
interval
between
fractures
9-74 - Diagram explaining how the fracture density can be evaluated
from the images.
With the other tools this can be evaluated from the frequency at which the fracture indicators occur, notably on
the dipmeter and on the FIL (Fig. 9-42) and DCA (Fig. 945) presentations, and from the porosity of the fractures.
This can be evaluated by various means.
Fracture aperture
The mean fracture aperture (AF,) is the mean of fracture trace aperture averaged over an interval of height H:
Sibbit & Faivre (1985) related the opening (in pm) of
vertical and horizontal fractures to the conductivity measured by the Dual Laterolog (DLL) tool and the difference
between the deep (LLd) and shallow (LLs) resistivities.
They also showed a relation between their lateral extent
(depth into the formation) and the same Dual Laterolog
measurements. In the case of vertical fractures (parallel to
the tool axis) the two measurement curves separate (LLd
> LLs) and their difference is proportional to the product of
the fracture aperture, E , and the conductivity of the invading fluid, C
., For horizontal fractures (perpendicular to
the tool axis) the two curves show a resistivity decrease
over approximately 0.8 m (Fig. 9-76). Again the separation is proportional to the product of the fracture opening
and the invading fluid conductivity. The image tools enable to determine if the fractures are vertical or horizontal.
As previously seen an evaluation of the fracture aperture can be obtained from the Stoneley wave analysis (cf.
Fig. 9-26).
Fracture porosity from images
(9-10)
with :
lj = length of segment j
aj = local aperture of segment j
One determines the mean hydraulic fracture aperture
AF, which is the cubic root of the fracture trace aperture
Serralog 0 2003
The apparent fracture porosity (Qf) is estimated as the
ratio of two areas: the apparent area of fractures seen on
the borehole wall (Fig. 9-75), and the area of the borehole wall covered by the image:
(9-12)
379
380
Well Logging and Geology
The first term is always very small and can be ignored.
The matrix porosity of compact fractured rocks is also low
(usually less than 10 %) while Pef is also very small
(0.358 for water, 0.48 for oil and 0.807 b/e for salt water).
We can therefore write as a first approximation:
The porosity QXp is derived from the density-neutron
combination, and includes both matrix porosity and fracture porosity. This gives:
(9-16)
Figure 9-76 - Relationship between the fracture aperture e in mm (a) :
for vertical fracture and the conductivity; (b) :for horizontal fractures
and the resistivity (from Sibbit & Faivre, 1985).
Now, we can show that:
The results of the previous analyses are reproduced
as logs versus depth (Fig. 9-77).
PeBa
(Pe)Ba = 070
(9-17)
and further, as a first approximation, we can take:
which gives:
Note: The last equation only holds if the borehole wall is smooth,
so that the pad fits closely to the formation. Otherwise there may be a
cave due to crumbling of the borehole wall filled with barite mud. It is
necessary, therefore to examine the caliper and the density correction
before applying this formula. We must also bear in mind that, being a
unidirectional tool, it will only analyze the part of the formation in front of
the pad, and so it will not necessarily measure the total fracture porosity. In any case, if the hole is ovalized due to the presence of fractures,
the pad will usually ride the major axis of the hole, and so face the fractures. The measurement will thus be representative of the fracture porosity since it is unlikely that there is another fracture network at 90" to the
first when the hole is ovalised.
Figure 9-77- Example of display of resistivity image analysis. From the
left to the right are successively reproduced: fracture density, fracture
length, fracture aperture and fracture porosity
(courtesy of Schlumberger).
Fracture porosity from photoelectric index
We have already seen that the photoelectric capture
cross-section is strongly influenced by baryte muds, and
this feature can be used to evaluate fracture porosity.
The following equation introduces the electronic density
Pe pe = C Pei pei
(9-13)
Or, for the case of fractured rocks invaded by baryte
muds:
380
Fracture porosity from dual laterolog (DLL)
Boyeldieu et a/. (1982) proposed the following equation for fracture porosity after studying the effects of fractures on the deep and shallow laterologs, and making certain assumptions
where Qfr is the fracture porosity, (@fr)c
is the computed
fracture porosity and CLLsand CLLd are the conductivities
in mhos of the LLs and LLd; m is between 1.3 and 1.5.
The assumptions made by the authors are as follows
- The fracture system is seen by both laterologs as a
system of resistivities in parallel with the compact, nonfractured, formation (a perfectly reasonable assumption).
- There is no invasion of the non-fractured part of the
Serralog 0 2003
Next Page
Fractures
formation (the blocks contained within the fracture system), but only of the fracture system. This assumption is
justified by the very high permeability of the fracture system compared with that of the rock itself, so that the overpressure of the mud column will act preferentially on the
fracture network.
- The invasion of the fracture system is not too deep,
but sufficient to ensure that the LLd reads the virgin formation while the LLs reads the flushed zone. The validity
of this assumption will depend on the type of mud and on
the degree of opening of the fractures. If the losses observed during the drilling are low, it can be assumed that
the openings are small and that a mud-cake was able to
develop and limit the invasion. In this case the assumption is valid. If the losses were considerable, the invasion
will be deep, and we can no longer assume that the LLd
reads the virgin zone.
- The water saturation of the uninvaded fracture system is almost zero. This is a reasonable assumption given
the permeability of the fractures.
- The filtrate saturation of the invaded fracture system
is 100 %. Again, due to the high permeability of the fractures, we can assume that all the hydrocarbons have
been flushed.
The authors then derived the following inequalities:
I
Chapter9
I
381
Formation characteristics
(9-20)
and
(9-21)
where $ma is the matrix porosity, $fr is the fracture porosity, S,,, is the water saturation of the non-fractured,
uncontaminated formation. Subtracting equ. 9-13 from
equ. 9-14 gives:
(9-22)
The above hypotheses assume that SXofr= 1 and Sdr
= 0. This then gives equ. 9-16 by substituting conductivities for resistivities.
However, as the authors themselves pointed out, the
best results are obtained when the mud resistivity is about
equal to that of the formation water, and when the formation contains hydrocarbons.
In water-bearing sequences, on the other hand, the
two salinities (mud and formation water) should be very
different. In this case the authors proposed the following
equation:
(9-23)
Figure 9-78 shows an example of results from an
interpretation of very compact, fractured formations.
Serralog
2003
Figure 9-78 - Example of the computation of fracture porosity from the
dual laterolog in granite (courtesy of Schlumberger).
Determination of fracture permeability
Mathieu et a/. (1984) have estimated that fracture permeability can be determined from an analysis of Stoneley
wave detected by a tool which records the complete
acoustic wave train. The results they obtained in a solid
crystalline formation seem encouraging.
In any case, the fracture permeability is controlled by
the minimum aperture of the fracture.
Fracture effects on reservoir interpretation
Through their influence on log measurement it is
obvious that fractures affect the reservoir evaluation.
Tortuosity factor m
In a reservoir the influence of the pore geometry on
electrical measurements has been recognized for many
years (Fig. 9-79). The well known Archie’s equation links
381
Previous Page
the resistivity in the virgin reservoir, R,, to the porosity, @,
the saturation, S,, in formation water, R
,, and to the changes in pore geometry in the form of the tortuosity factor,
also known as the cementation or Archie's factor, rn :
R, I R, = a I @mS,n
(9-24)
Figure 9-79 - On the top left, the pore geometry is modelled as a
set of water-filled tubes within a cube of rock of length L, the
average length of each tube being La
Below, the flow path length (in red) and pore radius are dependent on
the variations in grain size.
The cross-sectional variations can be modelled as step changes.
On the right the diagram shows the relationships between path length,
porosity and Archie's factor m. The longer the path length, the greater
the value of LJL
(adapted from Waffa, in Schlumberger Middle East Well Evaluation
Review, no 2, p. 49-57, 1987).
The effects of pore geometry on current flow path can
be modelled as a set of tubes showing variations in their
cross section area due to variations in grain size. The
length of the current flow path, La,can be compared to the
length of the reservoir cube, L. For intergranular porosity
the ratio La/L is close to 3.5. An open fracture, filled by
mud, will create a short circuit, since the open fractures
are more or less rectilinear planes. Consequently, one
would expect for an horizontal fracture a ratio La/L= 1 and
a tortuosity factor to be close to 1, at least when the porosity is due to the fractures, and the current lines are parallel to the plane of the fractures. In fact, even if the fractures have not been healed, there will be crystals in the fractures which are not evenly distributed, and these will
increase the tortuosity. In addition, the fractures are not
always planar or indeed open, and they are frequently at
an angle to the borehole axis. Finally, there are often
several criss-crossing fracture systems. As a result, the
tortuosity factor, rn, is always greater than 1, but usually
well below 2 or 2.3, the values observed in compact formations, and more usually around 1.4. It depends on the
fracture porosity compared to the formation porosity (Fig.
9-80).
382
vugs
Figure 9-80 - The effects of fractures and non-connecting vugs on the
Archie's factor m (from Waffa, 1987).
Saturation evaluation
In fractured reservoirs, electrical current travels along
two paths, one through fractures and another through
intergranular porosity. The entire system can be modelled
as a parallel resistance network (Fig. 9-81).
Circuit in fracture
Circuit of equivalent resistances
Figure 9-81 - Modelling of the electrical current travel in a fractured
reservoir (adapted from Waffa, 1987).
In the fracture path one can write:
R,=%
(9-25)
4fI
and in the intergranular pore system:
Serralog 0 2003
Fractures
(9-26)
where index ma is for matrix with intergranular porosity, and fr for fracture.
The measured resistivity can be defined in two ways:
(9-27)
or
Rma+Rfr
Rt=RmaXRfr
and neutron hydrogen index combination. These different
values are introduced in the law of parallel resistances
which controls the current path into the formation. This
law is given by the following equation :
- Vt 4Sxo-Vf
Rin Rw
Rrnr
-+
46x0
(9-31)
From this relationship the value Rin is computed and
introduced in the Archie’s equation for the invaded zone
assuming n =2 :
(9-28)
This simplifies to:
I Chapter9 I 383
(9-32)
The only unknown is Sxo which can be computed
through the following equation:
(9-29)
(9-33)
It often happens that vugs are associated to fractures.
The rn factor being affected both by fractures and vugs
(Fig. 9-81), it is necessary to compute rn continuously.
This can be realized thanks to the electromagnetic measurement. As generally shale is absent or in very poor
amount in this type of reservoir, one can write the transit
time of the electromagnetic wave as follows :
MSFL
In this relationship the electromagnetic wave transit
times are generally known or obtained from tables or
charts:
- in filtrat,,,t(
is close to 25-32 ns/m as a function of
salinity),
- in hydrocarbon (tphcis between 4 , 7 4 2 ns/m for oil
and 3,3 ns/m for gas),
- in solid fraction (tpmavaries between 8,7-9,l nslm).
The porosity is evaluated by the density-neutron combination.
The only unknowns are finally the resistivity of the fluid
in the flushed zone (Rin) and S,, .
The electromagnetic wave travels in a zone very close
to the borehole wall (Fig. 9-82). Its travel time is sensitive
to the water content in this zone. In fact, in this zone fluids
are composed by original irreducible formation water,
retained by capillary forces, by mud filtrate and, possibly,
by hydrocarbon. The mixture formatio- water-filtrate has a
resistivity, Ri,, which depends on the salinity of the two
fluids and of their respective percentage: V, and $Sxo.The
estimation of V, creates a problem. The quantity of formation water retained by capillary forces is generally unknown, this quantity varying with the nature of the grain or
crystal surface (smooth or rugous). In the case of a drilling
with water-base mud one can estimate this volume as, for
instance, 10% of the filtrate volume in the invaded zone
(V, = O,l$Sxo). In the case of a drilling with oil-base mud
V, corresponds to @SWirr.
$Sxoor @Swirr
is computed from
Archie’s equation assuming rn = 2 at the beginning.
Moreover, the total porosity is deduced from the density
Serralog 0 2003
Figure 9-82 - Comparison of the formation volume investigated by
MSFL and EPT measurements. The
MSFL has generally a deeper depth
of investigation than the EPT.
Consequently, it can be slightly
affected by the virgin zone.
EPT
Saturation in the invaded zone is given by the previous
equation:
sx0=pd+ RXO
(9-34)
From this relationship, the rn value can be computed
and introduced in the Archie’s equation for the virgin zone
in order to evaluate the water saturation :
m=
loga+logRn-nlogsxa
~Og4,+10gRxa
(9-35)
and
(9-36)
$, V, and Rin are computed using the new value of rn
until a convergence on Sxo.Figure 9-83 illustrates results
383
of quantitative evaluation in a fractured formation.
ble. The lithology determination must use neutron-sonicthorium-potassium-photoelectric index combination.
Figure 9-84 - Example of cross-plot combining the formation factor and
the porosity derived from the neutron-density combination
(from Suau et al., 1978).
Figure 9-83 - Example of log interpretation in a low porosity formation.
Four zones were tested. Zone A had better production then the more
porous zone 0.Zone A appears to be fractured while zone D seems to
be made up of vuggy porosity. Fractures near zones B and C may
have contributed to the higher production though with lower vuggy
porosity. Observe the drop of m in front of fractured zones
(from Waffa, 1987).
100
1w
Porosity (%)
POmSity (%)
-
Formation factor Porosity relationship
As the fractures influence resistivity measurement and
have a very low effect on density-neutron measurements,
the relationship between the resistivity formation factor
and the porosity will be affected. If the porosity is plotted
on a logarithmic scale as a function of the formation facfractured zones will appear as zones
tor (FR = RJR,),
having the lowest values of FR for a given value of porosity in a low-porosity zone (Fig. 9-84). This is due to the
drop in resistivity associated with fractures.
Similar plots can also be made by replacing FR by R,
or RLLd (Fig. 9-85) and (I by At (Fig. 9-86). These plots
allow the computation of the rn factor
Figure 9-85 - Examples of crossplots of Rt vs porosity
(from Aguilera, 1980).
Lithology determination
In compact, non-fractured formations, the lithology and
the mineralogy of the reservoirs is easily determined from
the various logging measurements using Z-plot techniques (cf. Chapter 2). In the fractured zones, the readings of the density are frequently affected by caves or
borehole rugosities. Consequently, they are often unusa384
R,
Figure 9-86 - Plot of Rt vs At showing a low value for the cementation,
or tortuosity, factor m (from Aguilera, 1980).
Core orientation
Images of the borehole wall compared to photograph
Serralog 0 2003
Chapter9
I
385
of the external surface of a core allow the core orientation
as illustrated by Figure 9-87.
Closing
Figure 9-89 - Illustration of possible slippages
(courtesy of Schlumberger).
horizontal (q,)stresses (Fig. 9-90) and to pore-fluid pressure (Po).
The vertical stress corresponds to the weight of the
rock lying above. It is called the overburden. It is calculated by integrating the overlying rock's density with respect
to depth.
I
Figure 9-87 - Core orientation by comparison of the photograph of the
external surface of the core by the AUTOCAR apparatus with the FMS
image of the borehole wall (courtesy of ELF and Schlumberger).
Hole slippage
A careful examination of borehole shapes imaged on
UBI survey (Fig. 9-88) may reveal slippage or shear
displacement along fault. In few cases, the slippage may
be so large that the borehole can collapse and the tubing
or casing damaged (Fig. 9-89). The well is lost!
1
Figure 9-90 - Stresses and pressures undergone by a piece of rock at
great depth assuming a continuous fluid path from surface to
subsurface (adapted from Brie et al., 1985).
Figure 9-88 - A cross-section of the borehole shape may reveal slippage along the fault plane. The right figure corresponds to UBI data in
downward perspective (courtesy of Schlumberger).
Detection of borehole damaae in deviated wells
In subsurface, rocks are submitted to vertical (OV) and
Serralog 0 2003
The horizontal stresses, generally anisotropic, correspond to:
- elastic and plastic deformations caused by the overburden:
- tectonic stresses linked to the deformation of the
earth crust:
- stresses stored in the rock resulting from events linked to geological history.
The pore-fluid pressure may be equal to the hydrostatic pressure when exists a continuous fluid path from the
surface to the subsurface. When there is no continuous
path the pore-pressure, P, either exceeds the hydrostatic head in overpressured formation, or is lower in a deple385
ted or underpressured reservoir.
In vertical holes, the minimum horizontal stress direction (q,) is the only parameter that can be determined
directly from the breakout analysis, as the vertical stress
(6,)is parallel to the borehole axis and has no influence
on the breakout orientation.
In deviated holes, breakouts depend on the three main
stresses - vertical (o,,),
maximum horizontal (q,)and minimum horizontal (oh).
As previously mentioned, stress modifications caused
by drilling can lead to borehole stability problems. It is the
reason why it is important to clearly understand the tectonic forces at work before setting a well plan, assessing
fracture efficiency and formulating injection/production
optimization schemes. This can be achieved through the
stress orientation.
In deviated holes, the ultrasonic images of the borehole wall are providing the opportunity to identify the exact
nature and origin of the borehole damage (stress-induced
or drilling-induced breakouts).
Most of the time (80%) the ovalization is due to drilling
damage.
To determine the nature, geometry and orientation of
other types of damage than breakout, an analysis of the
borehole curvature, based on ultrasonic transit time data,
is performed at each level. This analysis is carried out
using either the HOSANA (Hole Shape ANAlysis) software or an interactive program called HSView (Hole Shape
View) developed by Schlumberger. Figure 9-91 is an
example of hole shape analysis results.
The curvature analysis is performed in three steps:
- at each level and at each azimuth, the radius and
center of curvature are computed in a sliding window
approximately 40" wide;
- if consecutive curvature computations match, the
corresponding angular zones are merged into a single arc
whose center is computed;
- according to the distribution of the arcs and their centers, the borehole damage is automatically identified as
breakouts (Fig. 9-91), shearing at the borehole wall (Fig.
9-92), keyseat (Fig. 9-93) or reaming (Fig. 9-94).
Bcnhcla radiw (In.)
Figure 9-91b- Borehole crosssection showing a breakout. Dots
are data ooints. solid lines reDresent the part of the borehole wall
that have a constant curvature.
The broken lines are the continuation of the largest solid
(courtesv of Schlumberaer).
"
Figure 9-91c - Perspective view.
(courtesy of Schlumberger).
I
~
X430
x460
X431
X432
Figure 9-91a - Example of Hole Shape Analysis results over 2.5
m. Track 3 from left presents the image of the borehole radius, the left
edge being the top of the hole. The value of the radius corresponding
to each color is reproduced in track 2 with the median, the lower 25
percentile and the upper 25 percentile radius values. Tracks 1, 4 and 5
show respectively the well deviation, the orientation of damage relative
to top of borehole, the extent of the feature relative to the hole size
(courtesy of Schlumberger).
386
Borehole cross-section
Figure 9-92 - Example of shear at a
pre-existing fracture and asymmetric
reaming. Observe the sudden enlargement of the hole at the shear
plane (top image). Bottom image
illustrates cross-section of a shear at
a pre-existing fracture
(see caption Figures 9-91 for notation)
(courtesy of Schlumberger).
Serralog 0 2003
Chapter9
f
387
to faulting on a pre-existing plane. The projection onto the
shear plane of the two vectors joining the centers of the
two arcs corresponds to the slip vector (Fig. 9-96).
-
x406
Figure 9-93 - Example of zone of
keyseat as illustrated by the HOSANA program, and cross-section of
the borehole
(courtesy of Schlumberger).
Brehole radius (in.)
Bonhols radius (in.)
Figure 9-96 - On the left :borehole cross-section showing a shear at a
pre-existing fracture. On the right :a perspective view of this shear.
Observe the displacement of the center of the two arcs
(courtesy of Schlumberger).
To avoid confusing this to reaming, a plane crossing
the level must be observed on the image or in a perspective view of this part of the well (cf. Fig. 9-96 right). The
obvious slip shown in Figure 9-96 is partly masked by
subsequent reaming creating a sudden borehole enlargement at the slip plane itself, followed by a smooth decrease in the average hole radius (Figs. 9-92 and 9-97).
Slip occurring
during drilling
Figure 9-94 - Example of zone of
reaming as illustrated by the
HOSANA program and cross-section of the borehole. Observe the
enlargement of the borehole and
the displacement of the centers of
the two arcs (green and red)
(courtesy of Schlumberger).
Preserved
plane
,
-
mdlm(In.)
Breakouts correspond to two undeformed arcs (i.e.
whose radii are close to the radius of the bit) separated by
two damaged zones 180" apart. If the centers of the two
arcs do not coincide this is due to a closing movement in
the direction of the main stress (Fig.9-95).
Figure 9-95 - Other example of
breakout well illustrated by the
borehole cross-section. Observe
the two damaged zones 180"
apart. Observe the displacement of
the center of thr two arcs and the
1.5 cm closure (courtesy of
Schlumberger).
-
.E
1
Reamed
volume
Figure 9-97 - Sketch of a shear at a pre-existing fracture followed by a
reaming. Observe the asymmetric enlargement at the shear plane and
the smooth upwards decrease in the average borehole radius
(courtesy of Schlumberger).
Keyseat can be assumed when a radius of curvature
smaller than the bit size is detected at the bottom of the
borehole (cf. Fig. 9-93).
Reaming can be recognized when two arcs, whose
radii correspond to the bit size, exhibit centers clearly
separated. An enlargement of the well must be observed
at the same depth (cf. Fig. 9-94).
RelationshiD between stress-induced features and
stress axis
Bwehoi8 radius (In.)
Shearing corresponds to two undeformed arcs displaced relative to each other. Such movement corresponds
Serralog 0 2003
The previous described stress-induced features can
be related to the full stress tensor: o,,oHand oh.
387
lnduced tensile fractures
In a vertical well they are usually parallel to the borehole axis and are generated along planes perpendicular
to the minimum stress.
In deviated wells they are en echelon as they are inclined relative to each of the principal stress axes;
Breakouts
A breakout is generated when there is a maximum tangential stress at the borehole wall.
In vertical wells they indicate the b h azimuth.
In deviated wells, the well axis is oblique to the three
main stresses which influence the maximum tangential
stress.. The breakout orientation depends essentially on
the following parameters (Fig. 9-98):
-the direction of the minimum horizontal stress (sh)
- the stress regime -which principal stress (maximum,
intermediate or minimum) is vertical
- the ratio R of the stress magnitude, given by:
Stress determination
The complete state of stress parameters cannot be
determined from a single breakout or fault movement.
Fortunately, the number of features is generally much
greater. If breakouts are observed at several borehole
orientation, the observations can be inverted to give the
main stress parameters. A method for breakout inversion
is proposed by Etchecopar eta/. (1981).
The unknown are o h and Q.
The solution lies on:
- the analysis of breakout orientation data relative to
the well,
- the well orientation.
The method consists to determine the parameters
which give the lowest value for the misfit M (Mmin).This
misfit is defined as the sum of the angular misfit between
actual and "computed" breakouts. Determination of the
minimum misfit value (Mmin)for all the data involves investigating a range of solutions using a plot of q, versus Q
(Fig. 9-99).
The two last parameters can be combined into a single
parameter Q in the range O< = Q < 3 such that:
O < = Q < 1 impliesov>oH>oh,R = Q
1 < = Q < 2 implies oH> (s, > b h , R = 2 - Q
2 < = Q < 3 implies oH> b h > a, R = Q - 2
Figure 9-98 - Theoretical sketch
showing the change in breakout
orientation due to change in intermediate stress magnitude. At the
top the intermediate stress is equal
to the minimum; at the bottom it is
equal to the maximum
(here vertical)
(courtesy of Schlumberger).
The parameters <Th and Q are the same as those constraining the slip direction of microfaults, or the geometry of
nodal planes of focal mechanisms.
Shear movements
During or after drilling some fracture or fault planes
can act as slipping surfaces. Those planes normally lie at
an oblique angle to the current stress axes. If the fracture
or fault plane is reactivated a reltive displacement of Zwo
borehole sections may occur (Fig. 9-96).
The shear-vector orientation depends on the natural
state of stress but not on the well trajectory, as is the case
for the breakout orientation. So, the amount of breakouts
is a function of the well trajectory, but the occurrence of
slippage at pre-existing planes is independent of the well
trajectory.
388
Figure 9-99 - The cross-plot, on the left, shows the mean absolute
deviation to all data points as a function of the stress parameters oh
and Q. Each value of the cross-plot represents a stress state. The
stress state giving the best fit to the data point is marked by a red
square. The digits 1, 2...9 mark the stress state where the mean absolute deviation is 10, 20...90% above the best-fit value. The stress states with no marking have an aboslute deviation less than 10% in the
area around the square and more than 100% in the other area.
The histogram on the right shows the absolute deviation delta of the
individual data points for the best-fit stress state. The mean absolute
deviation is 35" (courtesy of Schlumberger).
The numbers N around the best solution indicate, for
the corresponding tensors, the misfit Mf as a function of
Mmin
1 means the fit Mf is within 10% of the minimum fit
Mmin,etc. If M, < 2 * Mmin,N is not displayed.
Many incorrect data, due to misinterpretation of the
borehole damage, may disturb the interpretation of the
cross-plot. To improve the interpretation, the method miniSerralog 0 2003
Fractures
mizes for the best N%, where N is generally fixed to 80%.
In order to determine the minimum misfit, each possible
tensor is examined as follows:
- for each stress tensor solution, the misfit for each
individual data point is computed
- the individual misfits are sorted in increasing order
- the total misfit Mf is taken to be the sum of the N%
smallest individual misfits
- the best solution is the tensor for which Mf is minimum.
Figure 9-100 reproduces the results of this filtering of
the data using these criteria. This new processing is constrained in orientation (N 150), but less in Q (0.6<Q<0.9).
This Q range corresponds to 0.6<R<0.9 with 6, being
Chapter 9
389
Q
omax.
Figure 9-101 - Stress states compatible with shearing of the borehole
at pre-existing fracture. Th star corresponds to the best-fit stress state
obtained from breakout analysis (courtesy of Schlumberger).
Q
N
Figure 9-100 - The cross-plot, on the left, shows the trimmed (best
80%) mean absolute deviation as a function of the stress parameters
oh and Q. Each value of the cross-plot represents a stress state. The
stress state giving the best fit to the data point is marked by a red
square. The digits I, 2...9 mark the stress state where the mean absolute deviation is 10, 20...90% above the best-fit value. The fit is computed for the best 80% of points at each stress state. Different points
may be rejected at different stress states. A much precise best-fit
stress state is obtained. The histogram on the right is also narrower if
we compare to Figure 9-99. The trimmed mean absolute deviation is
17” for the best-fit stress state (courtesy of Schlumberger).
oHis much closer to ovthan to oh. Using the misfit histogram associated to this new processing, it is possible to
locate zones where the orientation of the stress varies.
These can often be confirmed by changes in the orientation of induced fractures.
Interpreting the last solution, two type of microstructures support it:
- tensile fractures oriented N 150 as the maximum
horizontal stress determined,
- single slip movement in which it has been possible to
measure a pre-existing fracture.
Figure 9-101 shows that the stress is compatible with
both types.
Using stress information, stress components tangential to the borehole wall can be computed for all well orientation (Fig. 9-102).
Serralog 0 2003
Figure 9-102- Schmidt diagram showing the tangential stress state at
the borehole wall for different well deviations and azimuths. The center
of this diagram corresponds to a vertical well. A point on the outer circle corresponds to a horizontal well. The longuest length of the crosses indicates the maximum and the shortest the minimum tangential
stresses and their orientation. The numbers correspond to the ratio
between the maximum tangential stress and the vertical one (the ratio
a/av is assumed to be 0.6 and the fluid pressure is not taken into
account). The squares and circles indicate respectively the well orientation minimizing or maximizing the tangential stress at fhe borehole
wall. When the maximum tangential stress reaches its minimum value
it is also identical all around the borehole wall. This also indicates the
direction where drilling is safer (courtesy of Schlumberger).
This information of this diagram allows the determination of the well direction in order to limit the breakout
extension, and reduce the risks of hydraulic fracturing and
slip along pre-existing fractures. In addition, the mud
weight can be reduced as well, and also the drilling cost.
389
Prediction of the fracture efficiency
RECAP
As previously seen, the knowledge of the compressional and shear slownesses, and of the bulk and grain densities allows the determination of the mechanical rock properties. Through these parameters it is possible to calculate the vertical fracture migration. At each level up and
down the perforated interval, the additional pressure
required above extension pressure is calculated. With
delta pressure plotted versus depth, the relative breakdown pressures of the different zones may be determined
(Fig. 9-103).
In a certain extent, the pore-fluid pressure conteracts
the stresses imposed on the rock. Consequently, the rock
sustains an effective stress, 0,which is given by:
OV,H
= sV,H -
(9-39)
pp
Knowing this in-situ stress in and around the hydrocarbon bearing zones, the fluid density and volume can
be determined to predict the fracture migration behavior.
Then, The total fracture height for a given injection pressure can be calculated (Fig. 9-103).
Depth
Qualitative frac geometry
a 2 D representation
Before and after
frat gamma ray
One can conclude that the image tools are the best
technique for detection and a complete analysis of the
fractures.
In the absence of these measurements, the existence
of fractures can be concluded if several of the following
phenomena are observed simultaneously at about the
same depth:
- a change in temperature gradient;
- a change in hole diameter;
- a localised decrease in density, accompanied by a
variation in Ap while Pe, At and @N, remain steady, but not
if there is a cave, or the mud contains baryte;
- a very slight increase in porosity;
- secondary porosity;
- a reduction in the value of the rn factor;
- a change in the ratio LLd/LLs;
- sudden drops in resistivity on the microdevices;
- high Pe values when the mud contains baryte;
- conductivity peaks on the FIL;
- DCA showing conductive anomalies;
- a pause in tool rotation;
- strong attenuation of acoustic waves;
- a blurred zone on the VDL, or a lack of vertical coherence on the wave train;
- radioactivity peaks or uranium peaks;
- strong negative SP deflections.
8100
Figure 9-103 - Example of Fracture prediction log. In this example
each color represents the fracture height in 200 psi increments of frac
pressure. The maximum frac pressure must not exceed 800 psi. The
upper zone is weaker than the lower zone and will have a tendency to
take more of the treatment. Therefore, fluid entry must be limited to
this zone by increasing perforation density in the lower zone. The frac
job is crontrolled by the recording of the gamma ray before and after
the frac with radioactive tracers. One can observe that the predicted
frac height is within two feet of the actual height
(courtesy of Schlumberger).
390
Serralog 0 2003
References and Bibliography
ADAMS, J., BOURKE, L., & BUCK, S. (1990). Integrating Formation Microscanner images and cores.
Oilfield Review, 2, 1, p. 52-65.
AGUILERA, R. (1976). - Analysis of naturally fractured
reservoirs from conventional well logs. Journ. Petroleum
Technol., 28, July, p.764-772.
AGUILERA, R., & KIVAU POOLLEN, M., (1977). Current status on the study of naturally fractured reservoirs. The Log Analyst, 18, 3, May-June.
AGUILERA, R. (1980). - Naturally Fractured
Reservoirs. Petroleum Publishing Co., Tulsa, Oklahoma.
ANDERSON, T., & WALKER, T. (1972). - Log derived
rock properties for use in well stimulation design. SPE of
AIME, SPE paper 4095.
ATKINSON, A. (1977). - Fracture pressure gradients
from acoustic and density logs : an updated approach.
SPWLA, 18th Ann. Log. Symp. Trans.
BABCOCK, E.A. (1978). - Measurements of subsurface fractures from dipmeter logs. Bull. Amer. Assoc.
Petroleum Geol, 62, p. 1119-1126.
BARNES, R.P., DEE, S.J., & SANDERSON, D.J.
(2002). - Characterization of Fracture Networks Using
Data from Boreholes: An Example from the Concealed
Borrowdale Volcanic Group in West Cumbria, U.K. In:
LOVELL, M., & PARKINSON, N. (ed.) (2002). - Geological
Applications of Well Logs. AAPG Methods in Exploration,
604 p.
BARTON, C.A., & ZOBACK, M.D. (2002). - Wellbore
Imaging
Technologies
Applied
to
Reservoir
Geomechanics and Environmental Engineering. In:
LOVELL, M., & PARKINSON, N. (ed.) (2002). - Geological
Applications of Well Logs. AAPG Methods in Exploration,
604 p.
BATES, R.L., & JACKSON, J.A. (1980). - Glossary of
Geology. 2nd ed. Amer. Geol. Inst., Falls Church, Virginia.
BECK, J., SCHULTZ, A., & FITZGERALD, D. (1977). Reservoir evaluation of fractured cretaceous carbonates
in South Texas. SPWLA, 18th Ann. Log. Symp. Trans.
BELL, J.S., & GOUGH, D.I. (1979). - Northeast-southwest compressive stress in Alberta : evidence from oil
wells. Earth Planet. Sci. Let., 45, p. 475-482.
BEYDOUN, W.B., CHENG, C.H., & TOKSOZ, M.N.
(1983). - Detection of subsurface fractures and permeable
zones by the analysis of tube waves. 8th Europ. Form.
Eval. Trans., London, paper I.
BIGELOW, E.L. (1985). - Making more intelligent use
of log derived dip information. 5 Parts. The Log Analyst,
26, 1, 2, 3, 4, and 5.
BILLINGS, M.P. (1942). - Structural Geology. 1st ed.
Prentice-Hall Inc. Englewood Cliffs, New Jersey.
BILLINGS, M.P. (1972). - Structural Geology. 3rd ed.
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
BOND, L.O., ALGER, R.P., & SCHMIDT, A.W. (1971).
- Well log Applications in Coal Mining and Rock
Mechanics. Trans. SME 250.
BOURKE, L., DELFINER, P., FETT, T.H., GRACE,
L.M., LUTHI, S., SERRA, O., STANDEN, E., &
Serralog 0 2003
TROUILLER, J.-C. (1989). - Using Formation
Microscanner images. The Technical Review, 37, 1, p.
16-40.
BOYELDIEU, C., & MARTIN, C. (1984). - Fracture
detection and evaluation. SAID-SPWLA, 9th Europ. Form.
Eval. Trans., Paris, paper 21.
BOYELDIEU, C., &WINCHESTER, A. (1982). - Use of
the Dual Laterolog for the Evaluation of the Fracture
Porosity in Hard Carbonate Formations. Offshore South
East Asia 1982 Conference, 9-12 Feb., Singapore.
BRIE, A., HSU, K., & ECKERSLEY, C. (1988). - Using
the Stoneley normalized differential energies for fractured
reservoir evaluation. SPWLA, 29th Ann. Log. Symp.
Trans., paper XX.
BUTLER, M., PHELAN, M.J., & WIGHT, A.W.R.
(1976). - Buchan field :evaluation of a fractured sandstone reservoir. SPWLA, 4th Europ. Symp. Trans., London,
paper A.
BURKE, J.A., CAMPBELL, Jr. R.L, & SCHMIDT, A.W.
(1969). - The Litho-Porosity Cross Plot. SPWLA, 10thAnn.
Log. Symp. Trans., paper Y.
CHEUNG, Ph. (1984). - Fracture detection using the
sonic tool .SAID-SPWLA, 9th Europ. Intern. Form. Eval.
Trans., Paris, paper 42.
CLAVIER, C., & RUST, D.H. (1976). - MID-PLOT a
new Lithology Technique. The Log Analyst, 17, 6.
COATES, G.R., & DUMANOIR, J.L. (1974). - A new
approach to log-derived permeability. SPWLA, 14th Ann.
Log. Symp. Trans.
COATES, G.R., & DENOO, S.A. (1980). - Log Derived
Mechanical Properties and Rock Stress. SPWLA, 21s'
Ann. Log. Symp. Trans.
COATES, G.R., & DENOO, S.A. (1981). -Mechanical
properties programs using borehole analysis and Mohr's
circle. SPWLA, 22nd Ann. Log. Symp. Trans., paper DD.
COX, J.W. (1970). - The high resolution dipmeter
reveals dip-related borehole and formation characteristics. SPWLA, 11th Ann. Log. Symp. Trans.
COX, J.W. (1983). - Long axis orientation in elongated
boreholes and its correlation with rock stress data.
SPWLA, 24th Ann. Log. Symp. Trans., paper J.
CRARY, S. et a/. (1987) - Fracture Detection with Logs.
The Technical Review, 35, 1.
CURRIE, J.B., & NWACHUKWU, S.O. (1974). Evidence on incipient fracture porosity in reservoir rocks
at depth. Bull. Canad. Petrol. Geol., 22, 1.
DEERE, D.U., & MILLER, R.P. (1969). - Engineering
Classification and lndex Properties for lntact Rock. US.
Air Force Systems Command Air Force Weapons Lab.,
Kirtland Air Force Base, New Mexico, Tech. Rep.
DELFINER, P., PEYRET, O., & SERRA, 0. (1984). Automatic determination of Lithology from Well Logs. 59th
Ann. Techn. Conf. SPE of AIME, Houston, Texas; paper
SPE 13290.
DELHOMME, J.P., PILENKO, T., CHERUVIER, E., &
CULL, R. (1988). - Reservoir applications of dipmeter
logs. Journ. Petroleum Technol., 40, 2, p. 180-186.
DELHOMME, J.P., and MOTET, D. (1993). -Reservoir
description and characterization from quantitative boreho391
le
image
analysis.
Schlumberger
Reservoir
Characterization. Africa and Mediterranean Region, no 5.
DENNIS, B., STANDEN, E., GEORGI, D.T., &
CALOW, G.O. (1987). - ldentification and productivity predictions in a carbonate reef complex. 9th Europ. Form.
Eval. Trans., Paris, paper 21.
DICKEY, P.A. (1979). - Petroleum Development
Geology. Petroleum Publishing Co., Tulsa.
DONATH, F.A. (1970). - Some information squeezed
out of rocks. Amer. Sci., 58, p. 54-72.
DRUMMOND, J.M. (1964). - An appraisal of fracture
porosity. Bull. Canad. Petrol. Geol., 12 , 2.
EDWARDS, D. (1985). - Mechanical Properties
Evaluation and its applications. Schlumberger Technical
Services Inc., Europe Unit.
EKSTROM, M.P., CHEN, M.Y., ROSSI, D.J., LOCKE,
S., & ARON, J. (1986). -High Resolution Microelectrical
Borehole Wall Imaging. SPWLA, 10th Europ. Symp.
Trans., Aberdeen.
EKSTROM, M.P., DAHAN, C.A., CHEN; M.Y., LLOYD,
P.M., & ROSSI, D.J. (1986). - Formation Imaging with
Micro-Electrical Scanning Arrays. SPWLA, 27th Ann. Log.
Symp. Trans., Houston.
FAIVRE, 0. (1993). - Fracture evaluation from quantitative azimuthal resistivities. SPE of AIME, 68th Ann.
Techn. Conf. & Exhib., SPE paper 26434.
FAIVRE, O., & SIBBIT, A. (1984). - Dual Laterolog
Response in Fractures.
FOCKE, J.W., & MUNN, D. (1987). - Cementation
exponents in Middle Eastern carbonate reservoirs. SPE
Formation Evaluation, p. 155-167.
FONS, L. Sr. (1969). - Geological application of well
logs. SPWLA, 10th Ann. Log. Symp. Trans.
FRISINGER, M.R., & GYLLENSTEN, A. (1986). Fracture detection in North Sea reservoirs. SPWLA, 10th
Europ. Form. Evaluation Symp.
GARY, M., McAFEE, R.Jr., & WOLF, C.L. (1972). Glossary of Geology. Amer. Geol. Institute, Washington,
D.C.
GAUER, P., RUHLAND, M., BEAUCOURT, F. de, &
JANOT, P. (1984). - Ovalisation des trous de forage evaluee a partir des pendagemetres 4-patins; caracteres
geologiques de ces variations; leur utilisation pour la
determination de /'orientation des contraintes horizontales
actuelles dans le Fosse Rhenan. SAID-SPWLA, 9th
Europ. Intern. Form. Eval. Trans., Paris, paper 31.
GEORGI, D.T. (1985). - Geometrical aspects of
BoreHole Televiewer images. SPWLA, 26th Ann. Log.
Symp. Trans., paper 0.
GEORGI, D.T. (1986). - Guides for the interpretation of
dipmeter fracture logs. SPWLA, 27th Ann. Log. Symp.
Trans., paper TT.
GOGUEL, J. (1952). - Traite de Tectonique. Masson,
Paris.
GRIFFITH, A.A. (1920). - The phenomena of rupture
and flow in solids. Phil. Trans. Roy. SOC.London, ser. A
221, p. 163-197.
GRIFFITH, A.A. (1924). - The theory of rupture. Proc.
(1st) Intern. Congr. Appl. Mech., p. 55-63.
392
GRIGGS, D.T. (1936). - Deformation of rocks under
high confining pressure. J. Geol., 44, p. 541-577.
GRIGGS, D.T. (1939). - Creep in rocks. J. Geol., 47, p.
225-251.
GRIGGS, D.T. et al. (1951). - Deformation of Yule
Marble: part lV, effects at 150°C. Geol. SOC.Amer. Bull.,
62, p. 1385-1406.
GRIGGS, D.T., & HANDIN, J. (eds) (1960). - Rock
deformation. Geol. SOC.Amer. Mem., 79 , 382 p.
GRIGGS, D.T., TURNER, F.J., & HEARD, N.C. (1960).
- Deformation of rocks at 500"to 800".Geol. SOC.Amer.
Mem., 79, p. 39-104.
HATCHER, R.D. Jr. (1990). - Structural Geology.
Principles, Concepts, and Problems. Merril Publishing
Co., Columbus.
HEPP, V., & DUMESTRE, A.C. (1975). - CLUSTER -A
method for selecting the most probable dip results from
dipmeter survey. SPE of AIME 50th Ann. Fall Mtg., Dallas,
SPE
paper
5543.
HOBBS, B.E., MEANS, W.D., & WILLIAMS, P.F.
(1976). - An outline of Structural Geology. John Wiley &
Sons, New York.
HORNBY, B.E., JOHNSON, D.L., WINKLER, K.W., &
PLUMB, R.A. (1989). - Fracture evaluation using reflected
Stoneley wave arrivals. Geophysics, 54, 10.
HORNBY, B.E., LUTHI, S.M., & PLUMB, R.A. (1990).Comparison of fracture apertures computed from electrical borehole scans and reflected Stoneley waves: an integrated interpretation. SPWLA, 31s' Ann. Log. Symp.
Trans., paper L.
HUANG, C.F., HUNTER, J.A. (1981). - The correlation
of "tube wave" events with open fractures in fluid-filled
boreholes: current research, Part A. Geol. Survey of
Canada, paper 81-1A, p. 361-376.
INGRAM, D.S., ANDERSON, R.A., & ZANIER, A.M.
(1972). - Fracture pressure Gradient Determination from
well Logs (SPE paper 4135).
KLEINBERG, R.L., CHOW, E.Y., PLONA, T.J.,
ORTON, M., & CANADY, W.J. (1982). - Sensitivity and
reliability of Fracture Detection Techniques for Borehole
Application. SPE of AIME, 57th Ann. Fall Mtg., New
Orleans, SPE paper 11036.
KOERPERICH, E.A. (1977). - lnvestigation of acoustic
boundary waves and interference patterns as techniques
for detecting fractures. SPE of AIME, SPE paper 6820.
LEET, L.Don, JUDSON, S., & KAUFFMAN, M.E.
(1978). - Physical Geology. 5th ed. Prentice-HOY Inc.,
Englewood Cliffs, New Jersey.
LESLIE, H.D., & MONS, F. (1982). - Sonic Waveform
Analysis: Applications. SPWLA, 23rd Ann. Log. Symp.
Trans., paper GG.
LINK, P.K. (1982). - Basic Petroleum Geology. OGCl
Publications, Tulsa.
LIU, O.Y. (1985). - Fracture evaluation using borehole
sonic velocity measurements. SPE of AIME, 60th Ann.
Techn. Conf. & Exhib., SPE paper 14399.
LLOYD, P.M. (1986). - The Formation Microscanner :
A New Generation High Resolution Logging Technique.
LLOYD, P.M., DAHAN, C.A., & HUTIN, R. (1986). Serralog 0 2003
Fractures
Formation Imaging from Micro-Electrical Scanning Arrays
: A New Generation of Stratigraphic High Resolution
Dipmeter Logging Tool. SPWLA, 10th Europ. Syrnp.
Trans., Aberdeen.
LOVELL, M., & PARKINSON, N. (ed.) (2002). Geological Applications of Well Logs. AAPG Methods in
Exploration, 604 p.
LUTHI, S., & SOUHAITE, P. (1990). - Fracture apertures from electrical borehole scans. Geophysics, 55, 7, p.
821-833.
MATHIEU, F., & TOKSOZ, M.N. (1984). Determination of fracture permeability using acoustic
logs. SAID-SPWLA, 9th Europ. Intern. Form. Eval. Trans.,
Paris, paper 47.
MAURY, V., & ETCHECOPAR,A. (1992). - Shear induced by oil and gas wells drilling and production along
faults and discontinuities. 29th Intern. Geol. Congress,
Kyoto, Japan.
MAXWELL- GARNETT, J.C. (1904). - Transactions of
the Royal Society, 203 , p. 385, London.
MAYER, C., & SIBBIT, A. (1980). - GLOBAL, a new
Approach to Computer-processed Log Interpretation.
SPE of AIME, 55th ann. Fall Mtg., Dallas, SPE 9341.
MINNE, J.C., & GARTNER, J. (1979). - Fracture
Detection in the Middle East. Middle East Oil Techn. Conf
of SPE, Bahrain, 25-29 Mar.
MORRIS, R.L., GRINE, D.R., & A R K F E L ~ ,T.E.
(1963). - The use of compressional and shear acoustic
amplitudes for the location of fractures. SPE of AIME, 38th
ann. fall Mtg., New Orleans, Louisiana, SPE paper 723.
Also in J. Petrol. Techn., June 1964, pp. 623-635.
NOREL, G. (1984). - Detection des fractures et des
pendages par ultra-sons. SAID-SPWLA, 9th Europ. Form.
Eval. Trans., Paris, paper 16.
NURMI, R., CHARARA, M., WATERHOUSE, M., &
PARK, R. (1990). - Heterogeneities in carbonate reservoirs: detection and analysis using borehole electrical
images. In: Hurst, A, Lovell, M.A., & Morton, A.C. (eds) :
Geological Applications of Wireline Logs, Geological
Society, special publication, 48, p. 95-111, London.
PAILLET, F.L. (1985). - Problems in fractured-reservoir
evaluation and possible routes to their solutions. The Log
Analyst, 24, 1, p.26-41.
PERRIER, P., & QUIBLIER, L. (1973). - Orientations
des fractures observees sur carottes a I’aide de la pendagemetrie. Revue IFP, XXVIII, 6.
PEZARD, P.A., & LUTHI, S. (1988). - Borehole electrical images in the basement of the Cajon Pass scientific
drillhole, California: fracture identification and tectonic
implications. Geophysical Research Letters 15 (August
supplement), p. 1017-1020.
PICKETT, G.R. (1963). - Acoustic Character Logs and
their Applications in Formation Evaluation. J . Per.
Technol., 15, 6.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Houston.
PRESS, F., & SIEVER, R. (1982). - Earth. 3rd ed. W H.
Freeman & Co., San Francisco.
RAGAN, D.M. (1973). - Structural Geology. John Wiley
Serralog 0 2003
I
Chapter9
I 393
& Sons, New York.
RAMBOW, F.H.K. (1984). - The BoreHole Televiewer:
some field examples. SPWLA, 25th Ann. Log. Symp.
Trans., paper C.
RAMSAY, J.G. (1967). - Folding and Fracturing of
Rocks. McGraw-Hill Book Co., Inc., New York.
RAMSAY, J.G., & HUBER, M.I. (1987). - Modern structural Geology vol 2 : Folds and Fractures. Academic
Press, New York.
RASMUS, J.C. (1982). - Determining the type of fluid
contained in the fractures of the Twin Creek Limestone by
using the Dual Laterolog tool. J. Petrol. Techn., Nov.
RASMUS, J.C. (1983). - A variable cementation exponent, m, for fractured carbonates. The Log Analyst, 24, 6,
p. 13-23.
RIDER, M.H. (1978). - Dipmeter Log Analysis - An
Assay. SPWLA, 19thAnn. Log. Syrnp. Trans., paper G.
RODGERS, S.F., & EVANS, C.J. (2002). - Stressdependent Flow in Fractured Rocks at Sellafield, United
Kingdom. In: LOVELL, M., & PARKINSON, N. (ed.)
(2002). - Geological Applications of Well Logs. AAPG
Methods in Exploration, 604 p.
RUSSELL, W.L. (1951). - Principles of Petroleum
Geology. McGraw-Hill Book Co., New York.
Schlurnberger Ltd (1970). - Fundamentals of Dipmeter
Interpretation.
Schlurnberger (1979). - Well Evaluation Conference.
Algeria.
Schlurnberger Ltd (1981). - Dipmeter Interpretation.
Volume 1 -Fundamentals.
Schlurnberger Middle East S.A. (1981). - Well
Evaluation Conference. United Arab Emirates/ Qatar.
Schlurnberger Middle East S.A. (1984). - Well
Evaluation Conference. Egypt.
Schlurnberger & Sonatrach (1995). - Well Evaluation
Conference, Algeria .
Schlurnberger and Agip (2000). - Italy 2000 - Valueadded reservoir characterization.
SERRA, 0. (1989). - Formation Microscanner Image
Interpretation. Schlurnberger, SMP-7028.
Services Techniques Schlurnberger (1970). - Well
Evaluation Conference. Libya.
SHANKS, R.T., KWON, B.S., VRIES, M.R. de, &
WICHMAN, P.A. (1976). - A review of fracture detection
with well logs. SPE of AIME, paper SPE 6159.
SIBBIT, A.M., & FAIVRE, 0. (1985). - The Dual
Laterolog Response in Fractured Rocks. SPWLA, 26th
Ann. Log. Syrnp. Trans., paper T.
SMEKHOV, E.M. (1969). - Etude de la fracturation des
roches et des reservoirs fractures d’hydrocarbures.
Travaux du VNIGRI. Traduction en franpais d’ELF R.E.
SNYDER, R.H., & CRAFT, M. (1977). - Evaluation of
Austin and Buda formations from core and fracture analysis. Gulf Coast Assoc. Geol. SOC.,27.
SUAU, J., BOYELDIEU, C., ROCCABIANCA, R.,
CIGNI, M., & SPILA, M. (1978). - Evaluation of very low
porosity Carbonates (Malossa, Italy). SPWLA., 19th Ann.
Log. Syrnp. Trans.
SUAU, J., & GARTNER, J. (1979). -Fracture detection
393
394
Well Logging and Geology
from the logs. SPWLA, 6th Europ. Symp. Trans., paper L.
SUPPE, J. (1985). - Principles of Structural Geology.
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
THEYS, P., LUTHI, S., & SERRA, 0. (1983). - Use of
dipmeter in Carbonates for detailed sedimentology and
reservoir engineering studies.
TIMER, M.P., LOVELESS, G.W., 81ANDERSON, R.A.
(1973). - Estimation of Formation Strength from the
Mechanical Properties Log. 48th Ann. Meeting of SPE of
AIME, paper SPE 4532
VINCENT, P., GARTNER, J., & ATTALI, G. (1979). GEODIP - An approach to detailed dip determination
using correlation by pattern recognition. J. Petroleum
Technol., Feb., p. 232-240.
WALDSCHMIDT, W.A., FITZGERALD, P.E., & LUNSFORD, C.L. (1956). - Classification of porosity and fractures in reservoirrocks. Bull. Amer. assoc. Petroleum Geol.,
40, 5.
WALKER, T., & KESSLER, C. - Defection of natural
fractures with well logs. Welex, L 38.
WATFA, M. (1987). - Current affairs in saturation.
Schlurnberger Middle East Well Evaluation Review, no 2.
WATFA, M. (1987). Seeking the saturation solution.
Schlumberger Middle East Well Evaluation Review, no 3.
ZEMANEK, J. et al. (1969). - The Borehole Televiewer
: A New Logging Concept for Fracture Location and other
types of Borehole Inspection. SPWLA, 9th Ann. Log.
Syrnp. Trans
-
394
Serralog 0 2003
WELL LOGGING AND SOURCE ROCK
Introduction
The interest of a sedimentary basin for petroleum
exploration requires the conjunction of four characteristic
properties:
- oil or gas potential of the sedimentary basin because
if the basin has no suitable source rock there can be no
hydrocarbon accumulation and consequently no interest
to pursue the exploration of this basin (Fig. 10-1);
- presence of reservoir rocks capable of accumulating
hydrocarbon during their migration from source rocks;
- existence of traps in which hydrocarbon can accumulate;
- existence of seal rocks avoiding any later dismigration of hydrocarbon.
The geological description of reservoir rocks and of
their transformation by diagenesis, and their deformation
under tectonic stresses has been done in the previous
chapters
The estimation of the hydrocarbon potential of a sedimentary basin is now analyzed.
This analysis starts by the location of source rocks and
the evaluation of their geochemical characteristics - organic content and potential to generate oil or gas.
Figure 10-1 - Potential source rocks depend on the organic carbon
content and their potential to produce hydrocarbon
(from Perrodon, 1980) .
Traditional geological approach
The source rock evaluation of a sedimentary basin is
carried out under laboratory conditions by direct geochemical analysis of rock samples collected on outcrops or in
wells in shaly intervals (Fig. 10-2). Unfortunately, in wells
the samples are rare as generally shales are not cored.
Sidewall cores give representative results but are discrete points of data. Cuttings collected in zones appearing
shaly may be contaminated by material falling from layers
further up the well or by the mud if the latter is oil-based.
The original organic matter - humic, sapropelic or
planktonic - present within a sediment evolves as a function of the conditions of oxydo-reduction into the depositional environment. Oxygen transforms the organic material into C02. The reducing conditions that exist fundamentally in these environments - swamps, lagoons, lakes,
fjords, deep seas - preserve it from this transformation in
Serralog 0 2003
J
Figure 10-2 - Comparison of the chromatograms of source rocks with
the chromatograms of oil samples. The two upper source rocks of
well 2 have not generated the oils produced in well 2. The lower source rock corresponds to the lower oil.
395
C02. The bacteria living in anaerobic conditions act on
the organic matter. Under their action it is transformed
and can become a source rock which can generate biogenic methane, oil, bitumen or thermal gas as a function
of its maturation under temperature and pressure influence.
Analysis of the organic matter and comparison with oil
collected in a well allows the determination of theorigin of
the source rock
Several authors, Beers & Goodman (1944) (Fig. 10-4),
Russel (1945), Swanson (1960) (Fig 1 0 4 , Hassan
(1973), Supernaw et a/. (1978) (Fig. 10-6), have observed
a strong correlation between uranium and organic matter.
Well logging approach
The evaluation of the source rock potential can be
achieved through well log analysis. Different approaches
will be studied.
The Carbolog method
This technique, proposed by Carpentier et a/. ( I 989),
is a quick, quantitative and qualitative method that uses
measurements which have been recorded in old wells. It
requires a calibration for each sedimentary basin. There
are also some difficulties in defining the 100% point for
organic material.
The uranium concentration
Figure 10-4 - Relation between
uranium and organic carbon in
sedimentary rocks
(from Beers 8, Goodman, 1944).
So, after calibration with core data, it is possible to
evaluate the organic carbon content of a source rock from
its uranium content and from that its hydrocarbon potential if the nature of the organic material is known (Fig 10-
7).
The reducing conditions which are favorable for the
transformation of organic matter are also favorable to the
preservation of uranium which will precipitate or accumulate into sediments under certain conditions. In presence
of organic matter uranium is irreversibly adsorbed from
uranyl solutions in the presence of bacteria and humic
fractions. Organic matter and acidic pH are favorable to
convert U022+ into the insoluble quadrivalent ion U02.
The same reducing conditions generate the transformation of sulfates in sulfides (pyrite for example). In acidic
pH environments, humic and fulvic acids, ethers, alcohols, aldehydes, favour the precipitation of uranium by the
reduction of US+to U4+,forming urano-organic complexes
or chelates. Clays also encourage the formation of schoepite by hydrolysis of the uranyl ions (Fig. 10-3).
Figure 10-5 - On the left: diagram showing possible relation of uranium
content to total organic matter as controlled by the proportion of humic
and sapropelic material making up the organic matter. On the right: oil
yield of a marine black shale as a function of the total organic matter
(from Swanson, 1960).
Consequently, the uranium measured through the
spectrometry of the natural radioactivity is a good indicator of the source rock potential (Fig.10-8).
The A log R method
J
Figure 10-3 - Diagram showing possible association and time of
emplacement of uranium with common constituents of marine black
shales. Uranium is represented by black squares
(from Swanson, 1960).
396
This method was introduced by Exxon geologists. It is
based on a A log R which is obtained by superimposing
sonic and resistivity logs on the correct scales and using
the value of the level of organic metamorphism units
(LOM).
A log R represents the area of separation between the
sonic transit time and the resistivity curves. The sonic
curve is superimposed on the resistivity using a logarithmic scale for resistivity and a linear scale for sonic (Fig.
Serralog 0 2003
Source rock
10-8). A base line is determined for resistivity and sonicby
overlaying the two curves in a non-source rock interval.
A log R is given by the following relation:
1 Chapter 10
The LOM values vary as follows:
- 0 to 7.5 : immature
- 7.5 to 11.5 : oil phase
- > 11.5 : gas phase.
NATURAL GAMMA RAY SPECTROSCOPYDATA
Figure 10-8 Example of formation very rich in uranium as shown by
the natural gamma
ray spectrometry
data. It may correspond to a potential source rock
(courtesy of
Schlumberger).
0
0
1
2
3
4
5
6
Organic carbon (Oh)
Figure 10-6 - Relation between uranium (expressed as a
uranium/potassium ratio) and organic carbon
(from Supernaw et al., 1978).
To determine the Total Organic Carbon (TOC) from the
A log R, the level of maturation (LOM) of the organic
material, must be determined. The LOM values vary from
0 to 12. It is evaluated from the pyrolysis maturation parameters (TmaX)by optical methods: determination of the
Vitrinite Reflectance (VR or PRV, Table 10-I), and
Thermal Alteration Index (TAI), and by the nuclear fission
traces or the estimation of the paleotemperatures or the
subsurface temperatures (Fig. 10-10).
a
Organic matter, percent
30
Humic, percent
15
Sapropelic. percent
85
Uranium, percent 0.0030
Oil, gallons per ton
26.6
6
25
25
75
0.0041
20.3
Knowing A log R and LO the TOC value can i e
determined from the chart (Fig. 10-11). If TOC measurements have been carried out from core samples in laboratory, the values can be plotted to confirm the exact
value of LOM.
A second parameter is required. It correspond to the
hydrocarbon generating potential S2. This parameter is
deduced from the TOC and LOM values as illustrated by
the chart of Figure 10-12.
C1
C2
c3
c4
20
50
17
50
50
15
10
50
50
0.0031
6.2
50
0.0063
12.5
10
50
50
0.0048
0.0054
10.6
9.4
0
I l I l I
d
15
75
25
0.0070
6.6
I
f
5
95
5
0.0051 0.0031
3.6
1.4
20 Miles
10
I
e
10
85
15
I
J
Figure 10-7 - Sketch showing theoretical distribution of humic and sapropelic materials in a shallow sea in which black muds are accumulating, and
the estimated uranium content and oil yield of the resulting black shale. Increase in total organic matter seaward due chiefly to seaward decrease
in amount of detrital sediment; proportional increase in sapropelic matter seaward due to decrease of humic land-plant debris and predominance of
planktonic matter (from Swanson, 1960).
Serralog 0 2003
397
DJOUI
W.1
Figure 10-11- Chart to allow the
determination of TOC from A log
R separation and LOM
(from Malla & Baci, 1995).
Figure 10-12 - Determination of S2
from TOC and LOM for type I1
(oil-prone) kerogen
(from Malla & Baci, 1995).
Methods based on cross-plots
Figure 10-9 - Two example Of A log R, TOC and S2curves determined
on 2 wells of Algeria (from Malla & Baci, 1995).
The validity
Of
this approach is controlled by plotting
the
deduced from logs
sured on cores (Fig. 10-13).
the
values
mea-
Table 10-1
Vitrinite reflectance (PRV) values for hydrocarbon
and coal
These methods are based on the theoretical influence
of organic matter on the log responses.
Organic matter is considered to be fundamentally nonconductive. Therefore, increases in resistivity may indicate the presence of organic matter in the shaley beds.
Organic matter has a specific density in the range of
0.95 to 1.O5g/cm3. This is comparable to the density of oil
or water. If one considers that organic matter replaces
part of the matrix, reduction in bulk density can be expected. Unfortunately, at the same moment influence of pyrite, often present in reducing conditions favorable to the
preservation of organic matter, will increase the apparent
bulk density. One has also to remember that shales generally break and cave and, consequently, the density measurement will be affected.
Wl1
Figure 10-13 - Comparison of the TOC measured on cores and TOC
log, at the same depth for the two wells (from Malla & Baci, 1995).
Oil phase
.Gas phase
Figure 10-10- The value of LOM, determined from measured PRV
equivalents for well DJD-1, was calculated at 11.5
(from Malla & Baci, 1995).
398
The average sonic transit time of organic matter is
assumed (Mendelson & Toksoz, 1985) to be close to 180
@ft, so comparable to that of water.
The neutron response of organic matter is estimated to
average 67 p.u. (Mendelson & Toksoz) whereas that of
clay minerals composing essentially shales is generally
lower than 50 p.u. Consequently, one can expect an
increase of neutron index in organic-rich rocks.
As previously seen one can expect an increase of
gamma ray response and uranium content in organic rich
-
Serralog 0 2003
Source rock
rocks. The proportion of uranium showed the best relationship with TOC (Mann et a/., 1986) confirming the
works mentionned previously.
Autric & Dumesnil (1985) suggested the use of sonic
or gamma ray versus resistivity, these measurements
being less affected by borehole rugosity (Fig. 10-14).
I Chapter 10 I 399
Mendelson & Toksoz (1985) modelled expected log
responses in source rocks assuming a simple formation
model of matrix, organic matter and porosity and comparing the calculated total organic carbon (TOC) value with
the measurements made on cores. They used multiple
regression analysis to define the relationship between
TOC and the various responses.
The methods based on cross-plots allow a qualitative
detection of possible source rocks but they do not allow
the precise evaluation of the organic carbon concentration.
Method based on nuclear measurements
Rl
Figure 10-14 - Characterization of source rocks by cross-plots combining sonic transit time and resistivity (left diagram), or gamma ray and
resistivity (right diagram) (from Autric & Dumesnil, 1985).
Meyer & Nederlof (1984) suggested same type of
cross-plots but they corrected resistivities to standard
temperature (Fig. 10-15).
Source Rocks
Other attemps to evaluate directly the source-rock
potential are based on the measurement of the carbonoxygen ratio (C/O) by spectrometry of the gamma rays
induced by inelastic collisions of fast neutrons (Figs. 1016 and 10-17). This measurement requires either a stationary run or several runs with very low recording speed.
This method was proposed by S. Herron (1986).
Knowing the C/O ratio, the determination of the organic
carbon content needs two determinations:
- a measurement of the oxygen concentration in rock.
This evaluation is relatively easy if the lithology is well
known. In any case the average percentage of oxygen in
the earth’s crust is 46% in weigth. Knowing the ratio and
the 0 content it is easy to compute the total C content;
- a correction for inorganic carbon influence on the
measurement. The later is possible if the lithology is
known and especially the percentage of calcite and dolomite.
Inelastic
collision of
neutrons
Source Rocks
Figure 70-16 - Example of interactions between high energy neutrons
and C and 0 atoms (courtesy of Schlumberger).
Resistivity at 75°F (omh-m)
Figure 10-15 - Bulk density-resistivity cross-plot and sonic-resistivity
cross-plots allowing the identification of source rocks
(from Meyer & Nederlof, 1984).
Serralog 8 2003
After corrections for these two parameters, the remaining carbon corresponds to the organic carbon present
within the rock.
As illustrated by Figure 10-17 a very good match exists
between the core measurements and the log measurements.
399
Reference and Bibliography
Energy (MeV)
Figure 10-17 - Typical Inelastic collision spectrum. Observe the peaks
of Oxygen and Carbon. They are more characteristic and so can be
easily detected by window energy detection collimated around these
peaks (courtesy of Schlumberger).
This method does not require a calibration on core
measurement. Unfortunately, this technique is expensive
and due to that not often used.
Figure 10-18 Evaluation of the organic carbon content
through the measurevent of the carbon-oxygen ratio
(from Herron et al.,
1988).
Total Organic Carbon (%)
400
AUTRIC, A., & DUMESNIL, P. (1985). - Resistivity,
radioactivity and sonic transit time logs to evaluate the
organic content of low permeability rocks. The Log
Analyst, 26, 3 , p. 36
BATES, R.L., & JACKSON, J.A., eds (1980). Glossary of Geology. 2nd ed., Amer. Geol. Institute, Falls
Church, Virginia, U.S.A.
BEERS, R.F., & GOODMAN, C. (1944). - Distribution
of radioactivity in ancient sediments. Bull. Geol. SOC.
America, 55.
CARPENTIER, B. et a/. (1989). - Diagraphies et
roches-mere : Estimation des teneurs en carbone organique par la methode du Carbolog. Revue I.F.P., 44, p.
600-719.
DELLENBACH, J., ESPITALIE, J., & LEBRETON, F.
(1983) - Source rock logging. SPWLA 8th Eur. Form.
Evaluation Symp. Trans., paper D.
HASSAN, M. (1973). - The use of radioelements in
diagenetic studies of shale and carbonates. Int. Symp.
Petrography of Organic Material in Sediments. Centre
Nat. Rech. Scientifique, Paris, September.
HERRON, S.L. (1986). - Derivation of the total organic
carbon log for source rock evaluation. SPWLA, 27th Ann.
Log. Symp. Trans., paper HH.
HERRON, S., Le TENDRE, L., & DUFOUR, M. (1988).
- Rock evaluation using geochemical information from
wireline logs and cores. Bull. amer. Assoc. Petroleum
Geol., 72, 8.
MALLA, M.S., & BACI, S. (1995). - Organic
Geochemistry and Well Logging. In: Schlumberger &
Sonatrach, Well Evaluation Conference, Algeria.
MANN, U., LEYTHAEUSER, D., & MULLER, P.J.
(1986). - Relation between source rock properties and
wireline log parameters: An example from Lower Jurassic
Posidonia shale, N. W. Germany. Advances in Organic
Geochemistry, 10, p. 1105-1112.
MEISSNER, F.F. (1978). - Petroleum geology of the
Bakken Formation, Williston, North Dakota and Montana.
In: The Economic of the Williston Basin. Montana Geol.
SOC.,1978 Williston Basin Symp. p. 207-227.
MENDELSON, J.D., & TOKSOZ, M.H. (1985). Source rock characterisation using multivariate analysis
of log data. SPWLA 26th ann. Logging Symp. Trans.,
paper UU.
MEYER, B.I., & NEDERLOF, M.H. (1984). ldentification of source rocks on wireli,?e logs by density/
resistivity and sonic transit time/resistivity crossplot. Bull.
amer. Assoc. Petroleum Geol., 68, 1, p. 121-129.
PASSEY, Q.R., et a/. (1990). - A practical model for
organic richness from porosity and resistivity logs. Bull.
amer. Assoc. Petroleum Geol., 74, 12, p. 1777-1794.
PERRODON, A. (1980). - Geodynamique petroliere.
Genese et repartition des gisements d’hydrocarbures.
Masson, Paris.
RUSSELL, W.L., (1945). - Relation of radioactivity,
organic content, and sedimentation. Bull. amer. Assoc.
Petroleum Geol., 29, 10.
Serralog 0 2003
Chapter 10 I 401
SCHMOCKER, J.W. (1979). - Determination of organic
content of Appalachian Devonian Shales from FormationDensity iogs. Bull. arner. Assoc. Petroleum Geol., 63.
SCHMOCKER, J.W. (1981). - Determination of organic
matter content of Appalachian Devonian Shales from
gamma ray logs. Bull. arner. Assoc. Petroleum Geol., 65,
p. 1285-1298.
SMAGALA, T.M.C., et a/. (1984). - Log-derived indicator of thermal maturity Niobrara Formation, Denver basin,
Colorado, Nebraska, Wyoming. In WOODWARD, J. et a/.
(eds), Hydrocarbon Souce Rocks of the greater Rocky
Mountain Region. Rocky Mountain Assoc. geologists, p.
355-363.
STOCKS, A.E., & LAWRENCE, S.R. (1990). ldentification of source rocks from wireline logs. In: Hurst,
A., Lovell, M.A., & Morton, A.C. (eds), Geological
Applications of Wireline Logs, Geological Society, Special
Publication, no 48, London.
SUPERNAW, I.R., McCOY, A.D., & LINK, A.J. (1978).
- Method for in-situ evaluation of the source rock potential
of each formation. U.S. patent 4,071,744, Jan. 31.
SWANSON, V.E. (1960). - Oil yield and Uranium
content of black shales. Geol. Survey, Prof. paper 356-A.
WILLIAMS, P.F.V. (1986). - Petroleum geochemistry of
the Kimmeridge Clay of Onshore Southern and Eastern
England. Marine and Petroleum Geology, 3, p. 258-281.
Serralog 0 2003
401
WELL LOGGING AND
SEQUENCE STRATIGRAPHY
Introduction
As defined by Gignoux (1936), stratigraphy is that
branch of geology which “studies the layers of the earth‘s
crust and rocks from the point of view of their chronological succession and their geographical distribution”.
For american geologists, “stratigraphy is the science
of rock strata. It is concerned not only with the original
succession and age relations of rock strata but also with
their form, distribution, lithologic composition, fossil
content, geophysical and geochemical properties indeed, with all characters and attributes of rocks as strata; and their interpretation in terms of environment or
mode of origin, and geologic histor)/‘ (Bates 8,Jackson,
1980).
Stratigraphic sequence is “a chronologic succession of
sedimentary rocks from older below to younger above,
essentially without interruption; e.g. a sequence of bedded rocks of interregional scope, bounded by unconformities” (Bates & Jackson, 1980).
Sequence stratigraphy “is the study of genetically related facies within a framework of chronostratigraphically
significant surfaces’’ (van Wagoner et a/., 1990).
than from any other method (Figs. 11-1 and 11-2).
Furthermore, logs provide much greater precision in
determining the stratigraphic distribution. Because of the
degree of detail they provide, largely due to the very good
vertical resolution of microdevices and images of the
borehole wall, the logs can provide relative dating on a
scale of the order of decimeters or even centimeters. This
cannot be achieved by any other stratigraphic technique,
Darticularlv in uniform deposits.
Figure 11-1 - Example of overturned fold
detected from the gamma ray curve. The
small section on the top-right corresponds
to the lower section of the main curve on
the left after its reversal. As it can be observed the pics correlate perfectly
(courtesy of Schlumberger).
Figure 11-2 - Other example of double
overturned fold better detected on
gamma ray curve than on sonic curve.
The dipmeter confirms the upper fold but
does not clearly reflect the lower fold
(courtesy of Schlumberger).
The modern suite of logs allows us to determine the
succession and relative age of rock strata as well as their
characters and attributes. As well admitted, well logs provide a continuous image of the sequence of formations
encountered during a logging run and, from correlations
between wells, the logs can also give an idea of geographical distribution. Given this, it is clear that well logs can
be a rich source of stratigraphic information, a fact which
has long been recognized by geologists.
The stratigraphic information which can be derived
from well logging data can be classified into several categories.
Relative dating
Because the logs show, from top to bottom, the
sequence of formations in reverse chronologic order, they
are generally useful for providing a relative dating. This is
true even in the case of folded or overturned sequences,
which can in fact be identified much more easily from logs
Serralog 0 2003
J
403
Definition of parastratigraphic units
The recorded formations can be subdivided into units
which may be termed parastratigraphic (Busson, 1972).
These units can be directly dated if characteristic flora or
fauna can be detected in core samples or cuttings, or indirectly, from surrounding sequences which have themselves been dated. These units are delimited by marker
levels above and below. Of substantial extent, these markers are approximately parallel, suggesting a continuous
sequence of deposition.
Depending on the refinement of the division into stratigraphic units, it may be possible to identify units which
correspond to very specific periods in the geological history of a basin (e.9. a volcanic episode, a coal bed, a
transgressive level), thus providing stratigraphic markers
of considerable significance.
Stratigraphic phenomena
Stratigraphic phenomena such as cessation of sedimentation, erosion, unconformities or disconformities are
usually indicated by abrupt changes in the values and
appearance of the curves on at least one log, usually
several. The use of the logs to identify these phenomena
will thus involve analyzing each discontinuity on a curve in
order to establish the most likely cause for its occurrence
and its hierarchy.
boundary, an erosional surface or a fault.
- Surfaces related to a textural change reflect either an
abrupt grain-size change in siliciclastic deposits or an
abrupt textural change in carbonates.
- Surfaces generated by currents can correspond to
planar parallel laminations, cross bedding, foreset bedding, etc.
- Surfaces can relate to erosion, hardgrounds, paraconformities (horizons of roots, burrows, mud cracks, or
flooding surface), disconformities, angular unconformities
or transgression.
- Surfaces can be caused by tectonic stresses, such
as faults or fractures.
- Surfaces can correspond to compaction or diagenetic influences, such as change in the degree of compaction, pressure solution (stylolite), cementation (calcrete or
dolocrete).
- Surfaces can be linked to a fluid change, such as a
gas-water contact.
For bedding surfaces, the criteria proposed by van
Wagoner et a/., (1990) can be extended to the borehole
images, because most of the features are easily recognized on images (Fig. 11-3).
Break detection
Break definition
A curve discontinuity or break is any significant
response change occurring over a depth interval not
exceeding the vertical resolution of the tool. This break
will appear sharper when the depth scale is more compressed and the resolution of the tool is good. Thus, it will
be easier to identify it on a 1/1000 log scale than on a
1/200 or 1/40 log scale, and if the measurement is by a
microdevice than a macrodevice.
Break sianificance
Any break is an indication of a major change in at least
one of the factors affecting the response of the tool. This
is why a break is so significant and why one must try to
determine the reason for it. However, any major change in
one of the geological parameters will provoke a response
change, and thus a discontinuity, but only on those logs
which measure parameters which are susceptible to such
a change.
Breaks on curves correspond generally to surfaces.
The best tools for detection of these surfaces are image
tools. Effectively, images of the borehole wall can detect
nearly all the surfaces crossing a well and they allow their
classification.
These surfaces can have different origins, as listed
below.
- Surfaces related to an abrupt lithologic change,
which can be marked by a plane, can correspond to a bed
404
GRAVEL
MUDSTONE
SANDSTONE
ORGANIC DEBRIS
BEDDING SURFACE
Figure 11-3 - Typical bedding surfaces and how to recognize them
(adapted from van Wagoner et al., 1990).
Break classification
These surfaces are classified by their attributes and
lithology, and those that correspond to breaks in sedimentation are extracted and placed into a hierarchy.
Breaks without maior litholoaical chanae
When the lithology, reflected by the log responses essentially GR, thorium, potassium and Pe measurements - does not change but certain curves show existence of breaks, the origin of these breaks must be estaSerralog Q 2003
Sequence stratigraphy
I Chapter 11
405
blished from among the following possible causes.
S p o n l a m s potential
- Change in the type of fluid - In this case, the break
appears mainly on the resistivity curves and possibly on
the density, neutron and sonic curves if there is gas-oil or
gas-water contact.
- Textural change - A change in sorting or in the percentage of cement will affect the porosity and consequently all the measurements which depend on it, such as
density, neutron, sonic slowness and resistivity. An example is given in Figure 11-4. These textural changes which
occur suddenly may indicate either an erosion or a transgression.
Figure 11-5 - Examples of unconformities identified by the displacement of the shale base line of the SP curve.
(a) and (b) from Doll, 1948; (c) from Serra, 1972).
Figure 11-4 - Example of textural change in a sand, detected on the
density, neutron and resistivity curves, and, to a lesser degree, on the
gamma ray. Sand (e) is more coarse and poorly sorted than the lower
sand (d). This is interpreted as an erosional surface linked to the distributary channel cutting a mouth bar. Observe also the transgressive
levels at the top of the sand bodies. They correspond to a cemented
sandstone rich in shell debris partly dissolved, the calcite or dolomite
being reprecipitated due to diagenetic phenomenum
(from Serra & Sulpice, 1975).
- Diagenesis - Diagenetic phenomena usually result in
a variation in porosity. They can correspond to local
cementation due to drying periods in a channel (calcrete
or dolocrete) or to transgression marqued by an accumulation of shell fragments associated to the sand (Fig. 114). Later, due to their partial dissolution and, ultimately,
the precipitation of the dissolved calcite linked to change
in temperature and pression, the sand is cemented.
- Tectonic accident - It may bring into contact two identical lithologies with different petrophysical properties.
Diprneter or image analysis, possibly supplemented by
correlations between wells, should identify such an occurrence.
- Unconformitv - For instance a change in the shale
base line (Fig. l l - i ) , possibly associated with a change in
radioactivity or thorium/potassium ratio (Fig. 11-6) ITIay
indicate an unconformity. A sudden change in the Sonic
slowness trend with depth is also a good indicator of
unconformity (Fig. 11-7).
Serralog 0 2003
Figure 11-6 - Detection of an unconformity using the natural gamma
ray spectrometry measurement. Note the very marked change in the
thorium/potassiumratio on the two sides of the unconformity (from
Schlumberger, Well Evaluation Conference, Venezuela, 1980).
405
C
Figure 11-7 - Three examples of the detection of unconformities from compaction profiles using slowness measurements
(F from Fertl, 1976; S from Serra, 1972; C from Chiarelli et al., 1973).
Breaks correspondina to major litholoaical chanae
A change in lithology represents a major change in
sedimentological conditions and may or may not be part
of a sequential pattern.
1) In the first case, the lithological change only represents the passage from one element of the sequence to
the next one; e.g. dolomite-,
anhydrite + halite +
potassium salt.
The nature of the lithological change will shed light on
the polarity of the sedimentary sequence. Figure 11-8
gives an example of a sequence generated in an increasingly restricted basin, while Figure 11-9 provides an
example of the converse.
2) In the second case, several reasons may be advanced to explain such a change. The choice between them
will depend on:
- a detailed analysis of the various logs, covering lithology, facies and sequential analysis, as well as interpretation of dipmeter or image data and correlations with seismic measurements in the vicinity;
- complementary information, such as analysis of samples or cuttings and knowledge of the geology of the
region.
The possible reasons for a lithological break are now
studied. For each reason the controlling elements which
support the hypothesis will also be analyzed.
Diagenesis together with a mineralogical change
This phenomenon occurs frequently in
sequences. There is a response change in the measurements which are sensitive to mineralogy, photoelectric
index, density, neutron index, sonic slowness (Fig. 11-10).
406
caliper
Figure 11-8 - Breaks observed on the curves indicate major lithological
changes in sequences characterizing an increasingly restricted basin:
limestone - do/ostone - anhydrite - halite - polyhalite (radioactive zone)
(from Serra, 1980).
Serralog 0 2003
Sequence stratigraphy
Density
w-7
2.5
I Chapter 11 I 407
Erosion
The analysis of dipmeter resistivity curve (Fig. 11-11)
or better image of the borehole wall may reveal either a
reduction in the thickness of a bed from one curve to another, or a variation in dip angle or azimuth between the
upper and lower boundaries, or sometimes both. The analysis may also reveal an absence of planarity. Correlations
with nearby wells can reveal also a gap in the sequence
(Figs. 11-12 and 11-17)
Figure 11-9 - Breaks on curves indicating lithological changes which
characterize an opening of the basin and increasingly marked marine
influences: from the bottom to the top halite anhydrite - dolostone sandstone shale (from Sera, 1980).
-
-
Figure 11-11 - Evidence of erosional surface as shown by this old HDT
dipmeter processed by GEODIP. Observe the abrupt contact below the
upper sand, the erosion of the lower shale (pad 2 resistivity curve thinner than the other curves) and the filling of the trough by the sand
(curves 1 and 4 thinner than 2 and 3) (courtesy of Schlumberger).
Figure 11-70 - Example of lithological changes due to diagenesis.
Observe from the bottom to the top the succession of limestone - shale
- dolostone - limestone - shale - dolostone. The dolostone at the origin
was a limstone. Circulation of less salty water has transformed partly
the limestone in dolostone. The part of the limestone not transformed
at the top has been protected from diagenetic change by the formation
of an anticline before the migration of oil.
Serralog 0 2003
Unconformity
The unconformity may be due to a prolonged interruption in the process of sedimentation, that is, the upper
sequence being deposited on the eroded upper surface of
the existing beds which have been exposed to various
types of erosion.
When two successive beds have been deposited horizontally, the change in the surface of erosion is the only
indication of an unconformity. In this case, the surface of
erosion usually shows up as either (a) an interval without
dip or with no dip values of coherent azimuth which indicates a heterogeneous formation whose original stratification has been disrupted by alteration and bioturbation,
or (b) by a grouping of "blue" dip pattern (Fig. 11-13).
When the unconformity is angular it is generally much
to detect Since the beds do not have the Same dip
on either side of the UnConformitY: the upper sequence
rests on the upset beds of the lower sequences affected
by tectonic movement. This condition is shown on an
arrow plot by a fairly abrupt and significant change of dip
and azimuth (Fig. 11-14).
407
Figure 11-12 - Example of unconformity identified by correlations between wells (from Lardenois & Serra, 1967).
E m
DIP ANGLE L
MAECWN
UNCONfORYITY
WEATHERE0 ZONE
Figure 11-13 - Example of an unconformity characterized by a “blue”
pattern. Observe also the change in dip magnitude and azimuth
(courtesy of Schlumberger).
Figure 11-14 - Angular unconformity identified by a change of dip
magnitude and azimuth
(from Schlumberger, Well Evaluation Conference, Iran, 1976).
408
The erosion connected with the unconformity may
have resulted in a degree of a topographical relief whose
hollows may have been the first to be filled when sedimentation restarted. This is usually indicated by dips
incressing with depth, showing up on the arrow-plot as a
“red” pattern right up to the boundary of the unconformity.
As previously seen, an unconformity may also be
revealed by a change in the shale base line on a SP curve
(cf. Fig. 11-5). Another indication of an unconformity may
be a change in the average radioactivity level of the clays,
or a change in the thorium/potassium ratio (cf Fig. 11-6).
These phenomena may be explained by a combination of
several factors:
- a change in the mineralogical composition of the
clays (e.g. transition to an illite or montmorillonite) and of
the associated elements (e.g. organic matter and heavy
minerals). This may be due either to a change in the
depositional environment, or to different conditions on
either side of the unconformity (e.9. deep burial of the
lower sequence);
- variation in the salinity of the water bound to the
clays.
As also previously seen, a sudden change in the compaction gradient may be detected by analyzing the variations of the acoustic slowness in the shales with dept.
This will represent a change in the conditions of burial or
sedimentation at the unconformity, and so indicate its presence (cf. Fig. 11-7)
Ashes, lava flows or volcanic intrusions
A bed of volcanic ashes may be characterized by a
radioactive peak due to an increase in the thorium
concentration. Lava flows or volcanic intrusions (andesite
or basalt) can be more closely identified by a detailed
lithological analysis. Basalt is dense, sometimes vesicular, poor in radioactive elements.
Emersion
A period of emersion can be detected on images which
show traces of rootlets (Fig 11-15).
Transgression
This may appear in various forms:
- as a thin calcareous bed, rich in shell debris, at the
head of a prograding sequence, topped by fine prograding
sediments (cf. Fig. 11-4);
- as a condensation level rich in phosphate debris,
glauconite and in organic matter rich in uranium characterized by a radioactive peak (Fig. 11-16);
- as deposits over an erosional surface well recognized by correlations between wells (Fig. 11-17). This figure shows the transgression of the Cretaceous over the
Upper Jurassic in the Bays de Cau, Paris Basin, France.
The quality of the correlations reveals the gradual east-towest disapprearance of the Portlandian, as well as almost
the whole of the Kimmeridgian, with only the extreme
base of the limestones with Astartes remaining at
Preuseville 16.
Serralog 0 2003
Sequence stratigraphy
1 Chapter 11 I 409
Sequence stratigraphy from well logs
Nomenclature issues
Bioturbated
ZOM
Root trams
Figure 11-15 - Example of a period of emersion detected by the observation of root traces on borehole wall image
(courtesy of Schlumberger).
Sequence stratigraphy is a method that subdivides the
geological record in terms of transgressive-regressive
facies cycles bounded by unconformities or their correlative conformities. This approach is fundamental to the
subdivision, correlation and mapping of Sedimentary
rocks.
Seismic sequence stratigraphy is a powerful application of the sequence stratigraphic method that is most
applicable to basin exploration. One can determine:
- lines or surfaces (seismic reflectors) that are considered by geophysicists as chronostratigraphic markers,
which is debated by many geologists,
- basic stratigraphic units that represent depositional
sequences, systems tracts and periodic parasequences
(Vail eta/., 1977).
However, "seismic stratigraphy [does] not offer the
necessary precision to analyze sedimentary strata at the
reservoir scale" (van Wagoner et a/., 1990). This is becau-
Figure 11-16 - Example of unconformities marked by a peak of radioactivity on the resumption of sedimentation (Serra, 1972).
Preuseville 14
Pr 15
Pr 16
Pr 10 bis
Pr 9
Pr 8 bis
Eawy 101
Ea 2
CRETACEOUS
Portalandian
Turonian
Cenomanian
CRETACEOUS
-
Sequanian
Sequanian
Figure 11-1 7 - Example of erosion and an unconformity in the Paris Basin, which can be followed from the resistivity logs. The lower Cretaceous
transgression; which is difficult to differentiate, and the Gault shales can also be followed easily (from Lardenois 8, Serra, 1967).
Serralog 0 2003
409
se of the present-day resolution of seismic surveys. Even
with improved resolution, it would not be sufficient to
detect all surfaces. Consequently, the use of seismic data
must be restricted to determining the overall geometry
and geological setting of the unit that includes the reservoir. Van Wagoner et a/., (1990) also mentioned that
"using well logs and cores, a very high resolution chronostratigraphic framework of sequence and parasequence boundaries... could be constructed to analyze stratigraphy and facies at the reservoir scale".
It is essential to calibrate the seismic sequence stratigraphy with a higher-resolution technique that is better
able to detect and interpret each sedimentation break.
This way, the surfaces that bound rock masses at different
scales can be arranged in a hierarchy.
If we refer to the example of facies, depositional environment and sequence stratigraphy analysis (Fig. 11-18)
given by Van Wagoner et a/. (1990), a modern logging set
and their interpretation can provide practically all the information needed to achieve such an analysis of formations.
As indicated in Figure 11-18, lithology, bedding type, biologic activities, partly grain size, facies, and depositional
environment can be extracted from a complete set of
modern logs including images of borehole wall recorded
in a well. This is the reason why sequence stratigraphy
and correlations between wells are possible and must be
achieved from well logs.
The rock masses can be classified from the thinnest to
the thickest (Table 11-1) as:
10) Lamina: the thinnest strata unit of uniform composition and texture, which is never internally layered
(Campbell, 1967)
9) Laminaset: a conformable succession of laminae
(Campbell, 1967)
8) Bed: a succession of laminae or lamina set bounded by surfaces (called bedding surfaces) of erosion or
nondeposition (not all beds contain laminasets)
(Campbell, 1967)
7) Genetic increment of strata (Fig. 11-19): a succession of strata (or beds) with no interruption of sedimentary processes (e.g., Bouma's sequence, varves) (Busch,
1958, 1971)
6) Bed set: a relatively conformable succession of
beds, each one separated from the others by small
breaks in sedimentary processes, and bounded by surfaces of truncation, erosion or nondeposition (Campbell,
1967). A bed set can be classified as a simple bed set if
composed of several superimposed beds of
same lithology separated by truncations or erosional surfaces, or as a composite bed set if composed of several
beds of differing lithologies (Reineck & Singh, 1975)
5) Genetic sequence of strata: a succession of genetic
increments of strata (e.g., several turbidite deposits)
(Busch, 1958, 1971; Selley, 1976)
4) Parasequence: a "succession of genetic sequences
or "relatively conformable succession of genetically related bedsets, bounded by marine flooding surfaces or their
correlative surfaces" (van Wagoner et a/., 1990). The floo410
ding surfaces indicate an abrupt increase in water depth
generally caused by external events (tectonic, climatic,
etc.)
3) Parasequenceset: a "succession of genetically related parasequences forming a distinctive stacking pattern
bounded by major marine flooding surfaces and their correlative surfaces" (van Wagoner et al., 1990);
2) System tract: a linkage of contemporaneous depositional systems (Brown & Fisher, 1977) or three-dimensional assemblages of lithofacies (Fisher & McGowen,
1967). The four recognized system tracts are: lowstand
system tract, shelf margin system tract, transgressive system tract and highstand system tract.
1) Sequence: a "relatively conformable succession of
genetically related strata bounded by unconformities or
their correlative conformities" (Mitchum, 1977). Two types
of sequences are recognized (van Wagoner eta/., 1988).
A Type-I sequence comprises lowstand, transgressive
and highstand systems tracts bounded below by Type-I
unconformities and their correlative conformities. Type-2
sequence comprises shelf-margin, transgressive and
highstand systems tracts bounded below by Type-2
unconformities and their correlative conformities.
Nonmafirmcoal subfades
(swamp subenvironment)
Carbonaceous
sand facies
(deltaic environment)
Carbonaceoussand subfacies
(delta front subenvironment)
Mnrine shale subfadeg
(otfshon, subenvironment)
Figure 11-19 - Illustration of genetic increment and genetic sequence
(adapted from Selley, 1985).
The numbers of this classification, adapted from van
Wagoner et a/., (1990), indicate the relative hierarchy
from the thickest volumes, corresponding to the longest
time interval (more than 106 yr), down to the thinnest volumes, which may be deposited in a very short period (few
days, hours or even minutes). Two intermediate hierarchies are introduced to incorporate the geological phenomena that generate intermediate stratal unit.
Most hydrocarbon bearing reservoirs fall in the category of beds to parasequences, exceptionally to parasequencesets. Consequently, they usually represent relatively short time periods, and their limited thickness cannot
be precisely studied by seismic methods. In addition,
paleontological data, where available, are generally not
precise enough to provide reliable time markers.
Because time markers are essential for determining
lateral facies evolution, they must be determined from well
logs.
Serralog Ci 2003
Sequence stratigraphy
I Chapter 11 I 41 1
TROUGH CROSS REDS
TROUGH CROSS BEDS - SMALL SCALE
SIGMOIDAL CROSS BEDS
CURRENT RIPRE OEOOING
CONTORTED BEDS
PLANAR BEDS
HUMMOCKY BEDS
WAVP RIPPLE BEDDING
CLAY CLASTS
ABUNDANT
COMMON
RARE
ABSENT
CHURNED BY eunmws
Figure 11-18 - Example of facies and sequence stratigraphy of formations (from van Wagoner et al., 1990).
Serralog 0 2003
411
Next Page
Table 11-1
Scales of geological events from laminations to macroscale cyclic sequences. Most reservoirs correspond to parasequences or parasequence sets, as their thickness ranges from few meters to hundred of meters
(modified from Van Wagoner et a/., 1990).
Stratal Units
Thickness range
Area range
Time range
(feet)
(Sq. miles)
(Years)
Using log patterns
Well logs have long been us d for facie recognition
and sequence stratigraphic analysis. Most of the time,
users base their interpretation on patterns in standard
logs such as SP, gamma ray, resistivity and sonic. In a
typical basin, this approach is valid when the patterns are
sufficiently calibrated on core data. It is not valid when
used without control, because the same patterns may be
found within different environments or reflect different
phenomena.
A reduced logging set is not considered accurate
enough for precisely determine lithology, facies, or the
nature of surfaces or correlations, as will be demonstrated
later. The first step for well log-seismic sequence stratigraphic analysis is "to interpret lithology from log character. Any error in this step can lead to misinterpretations in
the later stages of the procedure, thus confirmation with
cores and cuttings, where possible, is very important"
(Vail & Wornardt 1990). Consequently, and especially
where no core data are available, it is desirable to use the
complete logging set, including dipmeter and borehole
images. Then, one can determine lithology, facies and the
nature of their correlative surfaces, the vertical and lateral
organization in genetic sequences, and parasequences at
the reservoir scale without compromising the interpretation. One can also better detect and interpret the breaks
observed on logs and images (e.g., rootlets or burrows)
for a more accurate sequence stratigraphic analysis. The
analysis of dipmeter and borehole images allows the user
to define hierarchical relationship between surfaces,
412
Tool Resolution
Use
using the same concepts that are applied to reflector terminations in the seismic sequence analysis (Fig. 11-20).
Of course, the interpretation of these surfaces must be
guided by the facies of the two surrounding beds.
The integration of:
- seismic data to resolve sequences to parasequence
sets,
- standard logs and dipmeter to resolve parasequence
sets to beds,
- high-resolution logs with borehole imaging to resolve
bed sets to laminae and to precisely determine the nature and origin of surfaces and their lateral extent defines
the whole range of sequence units (van Wagoner et a/.,
1990) and even some units in the range of beds to parasequences (Flint & O'Byrne,l992; Busch,l971).
The proposed sequence units were inherited from
seismic interpretation and later modified by information
from well logs, cores and outcrops (van Wagoner et a/.,
(1990). Seismic attributes, lithologies, log properties,
facies, and surfaces can be traced between rock masses.
Each stratum boundary, with a minimum of two contacts
on each surface, may record evidence of the processes
active during the diastem between beds (sedimentary
impact structures, traces of fossils, burrows, rootlets,
etc.). This concept has been clearly addressed by several
workers trying to interpret different contacts (Doglioni et
a/., 1990). In summary, a hierarchy of surfaces intersects
a hierarchy of rock masses, as defined using the reflector
terminations in seismic, bedding arrangement in outcrops,
Serralog Q 2003
Previous Page
Sequence stratigraphy
lo's km
L
32 My reflector (dated)
1's km
0.1's km
1 Chapter 111 413
0.001's km
Bomhole
reflector
I O.l'sm
37My
40 to 44 My
reflector
t
Paleornagnetism
Calibrated
Paleomagnetic Seismic
Detailed
Seismic Tracing
Dipmeter Tracing
Images
(Tomography )
Basin Geometry
External Reservoir
geometry
internal Reservoir
geometry
Figure 11-20 - The nesting concept of geometry from basin to internal reservoir geometry and from seismic scale to electric imaging.
dipmeter angular relationship or borehole images.
Table 11-2
Drawbacks of simplistic use of log patterns
Simplistic use of log patterns
Facies analysis is often attempted by interpreting log
patterns from a reduced set of logs (e.g., SP, gamma ray
and resistivity). This kind of approach involves "the matching of a series of unique log patterns between wells. It
works well in marine areas where there are small
contrasts in bathymetric relief... However, in nonmarine
and marginally marine areas and where there are large
differences in bathymetric relief, as occurs along a basin
margin, correlations are difficult at besf" (Vail et al., 1990
or Vail & Wornardt, 1990).
In many cases, even in the favorable cases mentioned
by Vail, these reduced logging sets are not reliable
enough to correctly determine facies (Serra & Sulpice,
1975). There are several drawbacks associated with the
simplistic use of these tools (Table 11-2). These include
the lack of vertical resolution for SP and resistivity, the
influence of salinity contrasts between mud and formation
water on the SP deflection, radioactivity that does not
necessarily reflect the shaliness of the formation, the
influence of diagenetic effects on sonic logs, their lack of
significant response in carbonates and the inability of
these tools to detect faults, emersion surfaces, etc.. For
these reasons, using a complete set of logs is recommended (Serra & Sulpice,l975). Many examples of misinterpretationof a reduced set of logs may be shown. This
will be illustrated by two examples.
Serralog 0 2003
SP :
Lack of resolution
Flat if R,f = R,
Poor in carbonates
GR :
Influence of radioactive minerals other than clays
such as: feldspars, micas, zircon, monazite,
phosphates, organic matter...
R:
Lack of resolution
Influence of fluids
Influence of pore geometry and distribution
Influence of conductive minerals other than clays
Sonic :
Influence of compaction
Influence of pore geometry and distribution
Influence of fluids
In the first example (Fig. 11-21), the evolution of the
gamma ray, resistivity and sonic curves might be interpreted as a fining-upward sequence. However, by incorporating the photoelectric index and potassium and neutrondensity curves, the correct lithologic interpretation is a
dolostone to feldspathic sandstone with practically no
shale. This sequence corresponds to an upward increase
in energy, the opposite of the interpretation based on the
reduced set of logs. This lithologic sequence results from
episodic discharges of detrital immature material into a
brackish lake.
The second example (Fig. 11-22) illustrates a similar
case, but this could be interpreted as a coarseningupward sequence. Instead, it corresponds to an arkosic to
subarkosic sandstone grading to a dolostone, as illustra413
414
DEPTH
Well Logging and Geology
Track 1
Track 2
Track 3
Track 4
Figure 11-21 - On the basis of the evolution of the GR, resistivity and sonic, this interval would have been interpreted as a succession of fining- and
coarsening-upward sequences. The complete logging suite, especially the natural gamma ray spectrometv, photoelectric index and neutron-density curves, allows a more precise determination of the lithology and indicates that the patterns correspond to the evolution of dolostones to arkosetype sandstones with practically no shale except at the lower part at 1026 m. The sandstones are interpreted as episodic discharges of detrital
material in a brackish lake.
interpretation and even identify anhydrite nodules and
vugs.
Figure 11-22a - The upward evolution of the gamma ray suggests a
coarsening-upwardsequence. In fact, the complete logging suite clearIy indicates a dolostone abruptly superposed by an arkosic to subarkosic sand (flash flood) and a progressive return to a chemical deposit.
The Formation Microscanner images on the right (Fig. 11-22b) illustrate the bed boundary types with heterogeneous internal bed organization. Conductive (dark) patches are vuggy porosity (Vu). Resistive
(white) patches are anhydrite cement growth in the sand
(from Schlumberger Well Evaluation Conference,Angola, 1991).
ted by the lithologic reconstruction from the complete logging set. Formation Microscanner images complete the
414
J
Figure 11-22b - Images show boundaries with heterogeneous internal
bed organization. Conductive patches are vuggy pores (Vu), resistive
patches are anhydrite cement growth
(from Schlumberger, Well Evaluation Conference,Angola, 1991).
Serralog 0 2003
Sequence stratigraphy
Other misinterpretations exist. For instance, in carbonates the gamma ray curve cannot be used as a shale
indicator, because the radioactivity is essentially linked to
uranium content, or to phosphates or even to glauconite.
Natural gamma ray spectrometry is much more accurate
for carbonates, enabling the recognition of more shaly
layers which will be associated with at hence high thorium
and potassium values.
Thinly layered formations cannot be correctly interpreted without high-resolution logs such as electromagnetic
measurement type EPT or ADEPT, dipmeters or image
tools.
Surface hierarchy from well logs
The identification criteria of surfaces too small to be
resolved by seismic methods (covering the range from
parasequence sets to laminae) are discussed in this section. Because the sequence analysis of rock masses is
well defined, an analysis based upon a hierarchy of surfaces should also be established. This analysis should
take into account facies distributions and the geometric
relationships between the rock units (baselap, toplap,
truncation, concordance). Table 11-3 proposes a hierarchy of some of the surfaces and facies that are described
in the literature.
Table 11-3
Proposed hierarchy of surfaces and facies.
Surfaces
Unconformities
Paraconformities
Aggradation surfaces
Progradation surfaces
Flooding surfaces
Erosional surfaces
Hardgrounds
Burrowed surfaces
Root surfaces
Reactivation surfaces
Bedding surfaces
Amalgamation surfaces
Avalanche surfaces
Truncation surfaces
Bedding surfaces
Facies
Basin Succession
Facies Systems
Facies Cycles
recognition is obvious in outcrops, where the bedding is
visible and needs no interpretation. Because of the significant loss of geometry in the subsurface, especially below
the scale of seismic resolution, this type of identification is
critical for obtaining a detailed reservoir description.
Fortunately, the vertical arrangement of surfaces, combined with facies determination and thickness data, allows
the geometry of the rock units to be inferred.
For higher-resolution sequence stratigraphy, well logs
must be used to determine surfaces and geometries.
However, surfaces defined by log pattern correlations are
not adequate time markers, in that they correspond to
facies correlations that are generally diachronous.
Accurate time markers must be defined by analyzing the
facies succession to determine genetic increments, bed
sets, genetic sequences and parasequences. Through
this analysis, the major breaks in sedimentary processes
are determined and understood and their correlative
conformities inferred at the reservoir scale.
Building a conceptual geological model
Conceptual geological models for seismic data processing can be generated using the logging approach.
This is fundamental to obtain accurate seismic sections
that are correctly interpreted in terms of sequence stratigraphy, even in continental or carbonate deposits. It is
well known that in continental deposits, the breaks in sedimentation are not necessary linked to changes in sea
level and not easily detected on seismic sections. It is also
understood that above a certain value of dip, beds are no
longer detected vertically below the source.
Some type of scaling-up can be applied using surface
hierarchy and angular relationships between the different
hierarchies and assuming log properties like density and
sonic in consistent geometric and facies intervals to create an appropriate seismic model. Automatically detecting
geometric breaks helps to define the lateral extent of sedimentary bodies by using the angular relationships between different sequence units in a given area.
Methodology
Facies Sequences
Facies
The hierarchy of a surface may change, but the related rock mass retains its definition. The play of surfaces
and rock masses defines the geometry of subsurface
bodies, allowing us to better predict the development of
sedimentary units at the reservoir scale. Of course, this
Semlog 0 2003
I Chapter 11I 415
The methodology presented in Figure 11-23 is proposed to maximize the use of well logs in sequence stratigraphy. By applying these techniques, all the recommended (van Wagoner et a/., 1990) data are used, because
the complete set of logs provides most of the information
regarding lithology, grain size, bedding types and organic
activities (cf. Fig. 11-18). No other data source, including
cores in some instances, provides this information. These
data are essential for sequence stratigraphic studies at
the reservoir scale and also for an accurate and detailed
basin history description.
This methodology must be applied to a network of key
wells in which a complete logging set has been recorded.
This "calibration" enables a correct interpretation of log
responses in other wells with a reduced logging set.
415
METHODOLOGY
Images
Dipmeter data
Openholelogs
MINERALOGY
POROSITY
DIAGENETIC EFFECTS
SURFACE TYPES 8 ORIGIN
REALTYKNESS
,
.Facies succession
.
-
FACIES OFEACH BED
Genetic sequence
Break hierarchy
I)
w
Depositionalenvironment
Surfaces of higher rank
Imagedata
Tectonic events
(folds, faults, fractures)
Succession of surfaces in terms of reflection coefficients
and convolution with a wavelet
w
SYNTHETIC SEISMOGRAM
.Comparison with seismic section
Figure 11-23 - Diagram explaining the methodology proposed for a
complete and accurate sequence stratigraphy analysis.
Correlations from well logs
At the origin correlations were realized from core analysis using paleontological data. Since their introduction in
1927, well logs have always been used for correlation
purpose. It is the reason why well log was first called electrical coring because the resistivity logs replaced cores.
Electrical coring was replaced by the american term log,
but should be returned to common use with the emergence of images which complete the description of the formation providing a view of the borehole wall surface similar to a picture of the external core surface.
The first stratigraphic applications of well logs started
with geologists, who observed similar resistivity features
in different wells at approximately the same depth. This
suggested that these features corresponded to the same
formation. Subsequently, experience has shown that correlations established with well log features or log patterns
correspond to facies correlations (more precisely, electrofacies).
If the resistivity was historically the first log type used
for correlations, and is still in use, gamma ray should probably be preferred as this measurement is less sensitive
to fluid. Figure 11-24 is a good example of correlations
between wells using this measurement converted in color
as a function of the radioactivity level. The color image
has a better impact and allows a higher perception of the
vertical variations and lateral continuity of the electrofacies.
Modern log data and processing allow geologists to
make good facies correlations between wells. However,
only under certain conditions do the correlations have a
chronostratigraphic value.
In order to establish chronostratigraphic correlations, it
is more appropriate to correlate geologic events of higher
rank than facies. Chronostratigraphic events are reflected
by important breaks in sedimentation that are marked on
the logs by abrupt changes in lithology. These changes
Figure 11-24 - Correlations of the Permian-Cretaceous section of western Kansas based on gamma-ray logs converted to a false-color spectrum
scaled to gamma-ray values as indicated by the colorred bar (from Collins et al., 1992).
416
Serralog 0 2003
Sequence stratigraphy
indicate sudden sea-level rise (flooding, marked by overlying deeper-water facies) or fall (receding, marked by
overlying shallow-water facies or emergence features
[e.g., rootlets]). Consequently, making chronostratigraphic
correlations require the considerations described in the
following sections.
In order to correlate a discontinuous sequence of surface outcrops, we generally look for certain characteristic
patterns such as type of lithology, color, texture, sedimentary features, and the sequence of flora and fauna.
Chronostratigraphic correlations between outcrops are
generally based on characteristic fossil species or faunal
assemblage.
In subsurface datations are esentially based on microfauna or microflora collected from cores. In their absence
correlations between wells are based on logs and electrofacies. The equivalent characteristics shown by events or
groups of events on the logs are used. This technique has
been known for a long time and the early log analysts
recognized that similar characteristics were often observed between wells, sometimes over wide areas. Well logs
have the added advantage of giving a continuous, objective and quantitative evaluation of the formations. In addition, they have very good vertical resolution which makes
it possible to detect fine detail, as well as very small changes which might be missed on cores. Furthermore, some
parameters are virtually unobtainable by any other
means. Some of these parameters have a considerable
bearing on the geological history, as well as spanning
great distances and variations in facies (for example,
radioactive levels associated with cinerite or volcanic
ashes).
However, most electrofacies are diachronous because
they correspond to deposits that migrated laterally during
appreciable periods of geologic time. Consequently, electrofacies correlations differ often from the chronostratigraphic framework.
Recent introduction of the measurement of the magnetic susceptibility and the natural remanent magnetization
in a well (Lalanne eta/.,1991; Pozzi eta/., 1993; Bouisset
& Augustin, 1993; Etchecopar et a/., 1993) allows the
datation based on inversion of geomagnetic polarity (Fig.
11-25).
To detect the reversals (polarity inversion), the problem is to find the zones where the TOTal-Field signal
(TOTF) variations are higher than the SUS ones (normal
zones), or smaller (reverse zones). SUS is the
Susceptibility Signal (converted in equivalent magnetic
field).
To that end, a cross-plot (Fig. 11-26) is made for each
zone, and the slope of the best straight line is computed.
The zone can slide over the whole log, and a C curve can
be plotted (Fig. 11-27). The length of the sliding zone is
set to a compromise between a good resolution and a
high frequency noise filtering effect. The C computation
algorithm is the classical least mean square one
C=
ZTOTFi x XSUSi - M x C(T0TFi x SUSi)
( CSUSi)’ - M x CSUSi’
Serralog 0 2003
(1
-,,
)
IChapter 11 I 417
Cae data
p
3
Figure 11-25 - Comparison between polarity inversion measured on
cores and polarity inversion deduced from Natural Remanent
Magnetization (NRMT) and Magnetic Susceptibility (SUMT) measurements realized in a well (from Etchecopar et al., 1993).
where:
TOTFi is the total field signal at level i,
SUS, is the magnetic susceptibility signal (converted in
equivalent magnetic field) at level i,
M is the number of levels in the sliding-zone length.
Imem
polarlty
Figure 11-26 - The cross-plot of the corrected total field (TOTF) in
nanoteslas versus the induced field (SUS) in nanoteslas allows the
determination of the polarity of the inversions of the earth’s magnetic
field (courtesy of Schlumberger).
Figure 11-27 - Logs of the TOTF and SUS and of the C coefficient for
the research of the reversals of the magnetic Earth’s field
(courtesy of Schlumberger).
417
Total oil company used these measurements to make
correlations between wells in the Paris Basin (Fig. 11-28).
The magnetostratigraphic time scale is calibrated on the
scale published by Haq et a/. (1987). As it can be observed the correlations based on electrofacies (in fact
gamma ray) compared to the ones realized with the
magnetism logs show the diachronism of the electrofacies
correlations.
duce the same log responses just as they produce the
same lithology and facies.
Sedimentatiar
rate
Sedimenlalb
rate
Figure 11-29 - Sedimentation rates deduced from the magnetostratigraphic scales allowing the verification of the calibrations. A wrong
interpretation of the inversion sequences (right part of the figure) generates erratic sedimentation rates, especially at the reversals
(from Etchecoparet al., 7993).
J
Figure 11-28 - Correlations based on gamma ray (top) compared to
correlations based on magnetism (bottom). Induced magnetic intensity
(in nanoteslas) and the Well Magnetostratigraphic Sequence (WMS) of
polarities are shown together with interpreted absolute age time lines
based on the Haq et al. time scale. The base of limestone marked by
depths A, B and C on gamma-ray curves are indicated by same letters
on the magnetostratigraphic curves .
Observe that the induced magnetism log (in nanoteslas) is an approximate inverse image of the gamma-ray log. Both tools are sensitive to
shale content. Radioactive isotopes and magnetic minerals are both
more abundant in shale than in limestone.
(from Bouisset & Augustin, 1993; Etchecopar et al., 7993).
The determination of the inversions of polarity can be
used for the estimation of the sedimentation rate (Fig. 1129). However, wrong determination of the inversion
sequences generates over estimation of the sedimentation rate.
Correlations based on magnetostratigraphic measurements must be made taking into account other log data,
especially dipmeter or image data, as any fault may eliminate inversion sequences making the correlations more
complex.
Principle of causality
The principle of causality states that the same causes
produce the same effects. Thus, the same set of depositional conditions in a given geological period should pro418
The application of this principle allows us to assert that
the persistence of certain criteria from one point of observation (a well) to another is proof that the original causes
were the same at both points. Thus if we observe similar
log features in different wells we may conclude that:
- the depositional conditions were the same at both
locations,
- it is probably the same formation, unless the phenomenon is not isolated in time (vertically on the log), or is
repeated in the same stratigraphic interval.
Geological phenomena of considerable importance
such as periods of burial, erosion, transgression or tectonic movement will all leave their mark on the log measurements, just as they do on rocks and formations, regardless of facies and environment. These features will therefore indicate the presence of these geological phenomena. It is so true that in many places well logs provide the
basis for lithostratigraphic formations, and each lithological unit in a basin can be designated by a type well which
is used as a reference for both lithology and log characteristics.
Concepts of log correlation
There are three fundamental concepts used in the
process of log correlation.
ConceDt of similarity
This is the first concept to apply simply because it is
the most obvious and the most intuitive. It is essentially
based on the shape of the curves, that is, the frequency,
amplitude and position of log events in vertical sequences. Clearly, for each event the value of all the log parameters must be considered, otherwise there can be incorrect correlations (Fig. 11-30). This concept will be used for
correlations of fine detail, for very close and precise studies, e.g., of a field. It may be useful to consult logs with
very good vertical resolution, such as the microlaterolog
and the dipmeter or images.
Care must be taken when applying this concept for
Serralog 0 2003
Sequence stratigraphy
Sonic
Sonic
Chapter 11
I 419
Sonic
Sonic
Figure 11-30 - Example of erroneous correlations (in red) established without taking into account the measurement values. Following the initial correlations, a condensation and unconformity were proposed for the well on the right. Without rejecting this hypothesis out of hand, the dashed correlations seem to point clearly to the presence of faults (from Serra, 1972).
correlations over a considerable distance, especially
when the similarity of shape is not perfect. If, on the other
hand, the similarity is perfect, then it can be safely
concluded that the correlation is valid and chronostratigraphic.
ConceDt of rhythmicity
Sedimentation takes place in sequences, rhythms or
cycles related to geological phenomena of some importance, and will thus be characterized regionally regardless
of the type of deposition (rhythmostratigraphy as defined
in Pomerol et a/., 1980) : "Sequences are separated from
each other by discontinuity surfaces or limit-surfaces
which reveal a break in sedimentation before a return to
the conditions of deposition similar to those which formed
the basis of the preceding sequence").
This often produces similar general evolution or, put in
another way, "electrosequences" which are comparable.
This is a very important concept, since it enables us to
identify geological phenomena showing close synchronization, even allowing for a certain "delay" between one
part of the basin and another. Such phenomena include
breaks in sedimentation due to tectonic movement, transgressive periods or eustatic cycles, erosion, gaps in
deposition and "hard-ground".
Concept of lateral variability
This concept is based on two sets of evidence:
-the lateral linking of facies which is not random since,
according to the Walther's law, there is a relationship between the juxtaposed and superposed elementary
sequences over the scale of the sedimentary sequence.
In other words, at the same instant in a given basin, for
example, there will be deposition of sands, silts, clays and
coals;
- the thickness of the deposits during the same time
period depends on:
. the type of lithology and depositional environment as well as on its compaction capacity,
. the phenomenon of subsidence,
Serralog 0 2003
. a combination of the two previous phenomena.
Lateral variability will be preferred in certain types of
deposit and especially in basins (deltaic, evaporitic), all
the more so if chronostratigraphic correlations are to be
established.
Intrinsic value of correlations.
Concept of reliability
From the quantitative information given by the well
logs we can evaluate the quality of the correlations by
introducing the concepts of correlation coefficients and
reliability.
A correlation coefficient may be computed for any correlation between two curves of the same type, just as dip
is calculated using the correlogram established from the
cross-correlation technique of the resistivity curves recorded by the dipmeter tools. In a given window, that is, over
a certain interval, the greater the similarity between the
curves in terms of shape, parameter values and frequency of events, the higher will be the coefficient of correlation and the greater the reliability of the correlation.
The measure of similarity between similar curves of
two wells is computed by the standard (Pearson) correlation coefficient, r:
c o w 1 L2)
r=
SLI -SL2
(11-2)
where Cov and s are the covariance and the standard
deviation of logs L, and L, . A segment of the curve of the
first well is then moved by small increments past the similar curve of the second well. At each step a correlation
coefficient is computed and plotted as a function of lag in
a correlogram (Fig. 11-31). The lag is the depth shift between the two curves at each position of comparison.
Programs for automatic correlations using this
approach have been written. The parametersthat must be
set are similar to those used in dipmeter processing:
depth interval for correlations, step length, and search
419
length.
Well A
I
Well B
Correlogramm
Well A vs Well B
negatlve R '- positive R
Figure 11-31 - Correlogram produced by cross-correlation of a curvesegment from well I , moved by increment over similar curve from well
2. The best correlation between the two curves is selected as the peak
of the correlogram with the maximum value and the smallest lag
(from Poelchau, 1987).
Though this technique has been successfully tested by
certain companies, it is difficult to apply since changes in
the type of tool, as well as possible variations in thickness
or in the amplitude of different events from one well to the
next all have to be taken into account. Hence, a subjective judgment as to the quality of the correlation is generally based on:
- the degree of similarity between the shape of each
log and the corresponding or equivalent one in the other
well,
-the number of logs showing this characteristic similarity,
- the interval over which this similarity is observed.
Thus the correlation coefficient will be high if the degree of similarity is high on each log over a sufficiently long
interval, say several tens of meters. By applying these
concepts to log correlations between wells a specific chronostratigraphic value for the correlations can be established.
Stratigraphic value of log correlations
With a few exceptions, such as cinerite or tuff levels, or
radioactive markers, log correlations are lithological or
facies correlations. Thus, the chronostratigraphic value of
these correlations lies in the answer to the following question : do the lithological correlations or facies correlations
follow the time-lines, that is, are they synchronous, or do
they cut across these lines ?
The answer lies on both the intrinsic value of the correlations, that is, their reliability, and, in the area covered,
on the type of basin, the type of lithology, and the sedimentation model. If, in a given interval there is nearly
420
exact repetition of the shapes of the curves and of the
measured parameter values between wells, thereby providing a maximum correlation coefficient, then there is
such a low probability that there will be another such period and within a similar time span another such sedimentary cycle showing the same exact repetition of characteristics that it can be assumed that the correlation reliability is maximal and the synchronism is confirmed.
When considering the type of basin, it is clear that in a
deltaic basin, a certain large-scale similarity of sequences
can be observed. There are successions of sequences of
similar type due to the fact that the depositional process
remained more or less constant over several epochs.
There is not, however, any exact similarity, except possibly in the homogeneous pro delta shales. In the direction
of progradation of the delta there is every chance of finding diachronous facies correlations. Moving along isopic
aureoles perpendicular to the direction of progradation
should give a more or less synchronous path. Some peat
or limestone beds can constitute good time markers especially in a limited extended area (a field for instance).
In the case of a calm intracratonic basin, such as the
Paris Basin at the Kimmeridgian stage, with marly limestone type deposits, an overall similarity of sequences and
often a peak-to-peak similarity of curves can be observed
over thicknesses of more than 100 m between wells separated by several tens of kilometres. The correlations are
easily established regardless of the direction taken.
Furthermore, since the correlations are surrounded by
excellent and well-dated markers, there can be no doubt
about their chronostratigraphic value.
Whereas in some detrital sequences it is almost
impossible to establish valid correlations in massive
sands, this is not the case in shaly groups where very
good correlations are frequently found (Fig. 11-32 see
next page). In the example given (Serra, 1972). the correlations show this kind of similarity and continuity, and it
is quite reasonable to give them a chronostratigraphic
value. It follows that:
- the lateral variation of the facies and the variations in
the thickness of shale and especially sand deposits can
be traced;
- in the absence of any fauna, the space-time development of the Cretaceous transgression can be traced
very precisely, and relative dating can be established for
its lower boundary at each point;
- the chronostratigraphic non-validity of each break
can be established on the basis of lithology changes.
Correlations based on the concepts of rhytmicity and
lateral variability are by definition more chronostratigraphic. This is because they correlate events of greater
amplitude, such as major discontinuities in sedimentation
cycles, and therefore represent a higher probability of
having a certain degree of synchronism.
Search for time-markers
Chronostratigraphic correlations must be based on
time constant horizons which are rare, because they
Serralog C 2003
Sequence stratigraphy
I Chapter 11 1 421
Figure 11-32 - Correlations in shale-sand sequences. The existence of shaley episodes in the center of this group of “green sands” in the AlboAptian of the Paris Basin provides excellent correlations which more than likely have very good chronostratigraphic value. In addition, they enable
space-time variations in the Creataceous transgression to be dated accurately, in a relative way, for each correlation point. Finally, as this figure
shows, any chronostratigraphic limit based on the lithology reference is worthless (from Serra, 1972).
generally correspond to more or less instantaneous phenomena, in terms of geological time, and they must be
independent of environment. To be usefull, they also must
be of wide geographical extent. The most common timemarkers which fulfill these criteria are those corresponding to cinerite or volcanic ashes. They are often difficult
to detect from cuttings, except with a microscope, but they
are often easily seen on well logs. Other time-markers
correspond to deposits in deep-water environments related to chemical changes in water masses due to changes
in ocean water-circulation patterns. Finally, good timemarkers are those corresponding to important breaks in
sedimentary and stratigraphy records. Such breaks are
usually easily detected on logs. They correspond to either
dip changes, or change of compaction, and/or “anomalous” peaks such as radioactive peaks or very dense and
compact levels (hardground), or even breaks of sequential evolution.
Extension to seismic sections
P
Figure 11-34 - Synthetic seismogram realized taking into account density and sonic travel time from which are computed the reflection coefficient. The convolution with a wavelet of this coefficient generates the
synthetic seismogram on which the seismic section can be correlated
and calibrated
(courtesy of Schlumberger).
The high resolution of sequence stratigraphy realized
from well logs must be transfered to seismic sections in
order to extend the markers detected on well logs to the
field or the basin. This extension is achieved through vertical seismic profile (VSP) recorded in wells and calibrated
on well log data (Fig. 11-33 see next page). In absence of
VSP a synthetic seismogram can be generated and compared to the seismic sections (Fig. 11-34).
Serralog 0 2003
421
SEISMIC
SECTION
VSP
SEISMIC
SECTION
Figure 11-33 - The left image is a calibration of VSP data with formation analysis from well logs. In the center the Vertical Seismic Section recorded
in the well. On the right the comparison of the VSP with the seismic section. This succession of processing allows a better interpretation of the
seismic section
(courtesy of Schlumberger).
References and Bibliography
BATES, R.L. & JACKSON, J.A. (1980).- Glossary of
Geology. 2nd ed. Amer. Geol. Inst., Falls Church, Virginia.
BOUISSET, P.M., & AUGUSTIN, A.M. (1993). Borehole magnetostratigraphx absolute age dating, and
correlation of sedimentary rocks, with examples from the
Paris Basin. Bull. amer. Assoc. Petrol. Geologists, 77,4,
pp. 569-587.
BROWN, L.F., & FISHER, W.L. (1977).- Seismi-stratigraphic interpretation of depositional systems: examples
from Brazil rift and pull-apart basins. In C.E. Payton, (ed.),
Seismic Stratigraphy - applications to hydrocarbon exploration. AAPG Memoir, 26,p. 213-248.
BUSCN D.A. (1971).- Genetic units in delta prospecting. Bull. amer. Assoc. Petroleum GeoL, 55, p. 1137-
1154.
BUSCH, D.A. (1974). - Stratigraphic Traps in
Sandstones-Exploration Techniques. Amer. Assoc.
Petroleum Geol., Mem. 21,164 p.
BUSSON,. G. (1972).- Principes, methodes et resultats d'une etude stratigraphique du Mesozoi'que saharien.
Mem. Museum Nat. Hist Natur., nouvelle serie, serie C,
Sci. de la Terre, 26.
CAMPBELL, C.V. (1967).- Lamina, laminaset, bed
and bedset. Sedimentology, 8,p. 7-26.
CAMPBELL, R.L. (1968). - Stratigraphic applications
of dipmeter data in Mid-Continent. Bull. amer. Assoc.
Petroleum GeoL, 52,p. 1700-1719.
422
CHIARELLI, A., SERRA, O., GRAS, C., MASS, P. &
TISON, J. (1973).- Etude automatique de la sous-compaction des argiles par diagraphies differees. Rev. Inst.
Franq. Petrole, 28,1, p. 19-36.
COLLINS, D.R., DOVETON, J.H., & MacFARLANE,
P.A. (1992). - Regional gamma-ray gray-tone intensity
images of the Permian-Cretaceous sequence of western
Kansas [Abstract]. SEPM 1992, Theme Meeting, Fort
Collins, CO, p. 19.
DESVIGNES, G., BARTHES, V., & TABBAGH, A.
(1992).- Direct determination of the natural remanent
magnetization effect in a hole drilled in layered ground
from magnetic field and susceptibility logs. Geophysics,
57, 7,pp. 872-884.
DICKEY, P.A. (1979). - Petroleum Development
Geology. Petroleum Publishing Co., Tulsa.
DICKINSON, W.R. (ed) (1974). - Tectonics and
Sedimentation. SEPM, special publication 22.
DOLL, H.G. (1948). - The S.P. Log: Theoretical
Analysis and Principles of Interpretation. AIME, Technical
Publication No. 2463.
DOVETON, J.H. (1994). - Geologic Log Analysis
Using Computer Methods. AAPG Computer Applications
in Geology, no 2.
DUFF, P.Mc L.D., HALLAM, A. & WALTON, E.X.
(1967). - Cyclic Sedimentation. Developments in
Sedimentology, 10,Elsevier, Amsterdam.
EINSELE, G., & SEILACHER, A, eds. (1982).-Cyclic
and event stratification. Springer-Verlag, New York, 536 p.
ETCHECOPAR, A., MOULIN, P., PAGES, G., &
Serralog 0 2003
Sequence stratigraphy
POCACHARD, J. (1993). - Harnessing paleomagnetics
for logging. Oilfield Review, 5, 4, pp. 4-13.
FERTL, W.H. (1976). - Abnormal formation pressures.
Development in Petroleum Sciences, 2, Elsevier
Amsterdam
FISHER, W.L., & McGOWEN, J.H. (1967). Depositional systems in the Wilcox Group of Texas and
their relationship to occurrence of oil and gas. Trans. Gulf
Coast Assoc. Geol. SOC.,17, p. 105-125.
FRIEDMAN, G.M. & SANDERS, J.E. (1978). - Principles of Sedimentology. John Wiley & Sons, New York.
GARY, M., Mc AFEE, R.Jr. & WOLF, C.L. (1972). Glossary of Geology. Amer. Geol. Inst., Washington, D.C.
GIGNOUX, M. (1950). - Geologie stratigraphique.
Masson, Paris.
GILREATH, J.A. & MARICELLI, J.J. (1964). - Detailed
Stratigraphic Control through dip Computations. Bull.
amer. Assoc. Petroleum Geol., 48, 12, p. 1902-1910.
HOBSON, G.D. & TIRATSOO, E.N. (1975). lntroduction to Petroleum Geology. Scientific Press Ltd,
Beaconsfield, England.
KRUMBEIN, W.C. & SLOSS, L.L. (1963). Stratigraphy and Sedimentation. 2nd ed. W H Freeman &
Co., San Francisco.
LALANNES, B., BOUISSET, P., PAGES, G., & POCACHARD, J. (1991). - Magnetic Logging: borehole magnetostratigraphy and absolute datation in sedimentary rocks.
paper SPE 21437, presented at the SPE Middle East Oil
Show, Bahrain, Nov. 16-19.
LANDES, K.K. (1951). - Petroleum Geology. John
Wiley & Sons, New York.
LARDENOIS, J. & SERRA, 0. (1967). - Series aux
multiples visages. Rev. Inst. Franq. Petrole, 22, 12, p.
1793-1817.
LEET, L. Don, JUDSON, S. & KAUFFMAN, M.E.
(1978). - Physical Geology. 5th ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
LINK, P.K. (1982). - Basic Petroleum Geology. OGCl
Publications, Tulsa.
LOCK, G.A. & HOYER, W.A. (1971). - Natural Gamma
Ray Spectral Logging. The Log Analyst, 12, 5, p. 3-9.
LOVELL, M., & PARKINSON, N. (ed.) (2002). Geological Applications of Well Logs. AAPG Methods in
Exploration, 604 p.
MACQUAKER, J.H.S., & JONES, C.R. (2002). - A
Sequence-stratigraphic
Study
of
Mudstone
Heterogeneity: A Combined PetrographidWireline Log
lnvestigation of Upper Jurassic Mudstones from The
North Sea (U.K.). In: LOVELL, M., & PARKINSON, N.
(ed.) (2002). - Geological Applications of Well Logs.
AAPG Methods in Exploration, 604 p.
MATTHEWS, R.K. (1974, 1984). - Dynamic
Stratigraphy. 1st and 2nd ed. Prentice-Hall Inc., Englewood
Cliffs, New Jersey.
MITCHUM, R.M. (1977). - Seismic stratigraphy and
global changes of sea level, Part 1: Glossary of terms
used in seismic stratigraphy. In C.E. Payton, ed., Seismic
Stratigraphy - applications to hydrocarbon exploration.
AAPG Memoir, 26, p. 205-212.
Serralog 0 2003
1 Chapter 11 1 423-
PERRODON, A. (1980). - Geodynamique petroliere.
Genese et repartition des gisements d'hydrocarbures. ElfAquitaine et Masson, Pau et Paris.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Houston.
POELCHAU, H.S. (1987).- Coherence mapping - an
automated approach to display goodness-of-correlation
between wells in a field. Mathematical Geology, v. 19, no
8, pp. 833-850.
POZZI, J.P., MARTIN, J.P., POCACHARD, J., FEINBERG, J., & GALDEANO, H. (1988). - In situ magnetostratigraphy: interpretation of magnetic logging in sediments. Earth & Planetary Science Letters, 88, pp. 357373.
POZZI, J.P., BARTHES, V., THIBAL, J., POCACHARD, J., LIM, M, THOMAS, Th., & PAGES, G. (1993).
- Downhole magnetostratigraphyin sediments: comparison with paleomagnetism of a core. J. Geophysical
Research, 98, pp. 1939-1957.
PRESS, F. & SIEVER, R. (1978). - Earth. 2nd ed. W H.
Freeman & Co., San Francisco.
REINECK, H.E., & SINGH, I.B. (1975). - Depositional
Sedimentary Environments. 2nd ed., Springer-Verlag,
Berlin., 439 p.
RUSSELL, W.L. (1951). - Principles of Petroleum
Geology. McGraw-Hill Book Co., New York.
Services Techniques Schlumberger (1976). - Well
Evaluation Conference, Iran.
Schlumberger Surenco S.A. (1980). - Evaluacionde
Formaciones en Venezuela.
Schlumberger (1991). - Avalia@o de forma@oes de
Angola, Well Evaluation Conference.
Schlumberger Ltd (1981). - Dipmeter Interpretation.
Volume l- Fundamentals.
Schlumberger Technical Services, Inc. (1982). Essentials of NGS Interpretation.
SELLEY, R.C. (1976). - An introduction to
Sedimentology. Academic Press, London, 408 p.
SERRA, 0. (1972). - Diagraphies et Stratigraphie. In :
Mem. B.R.G.M, 77, p. 775-832.
SERRA, 0. & SULPICE, L. (1975). - Sedimentological
analysis of shale-sand series from well logs. SPWLA, 16th
Ann. Log. Symp. Trans., paper W
SERRA, 0. (1980). - Aspects diagraphiques des evaporites. Bull. Centres Rech. SNEA (P), 4, 1, p. 411-431.
VAIL, P.R., MITCHUM, R.M., & THOMPSON, S.
(1977). - Seismic stratigraphy and global changes of sea
level, Part 3 :Relative changes of sea level from coastal
onlap. In C.E. Payton, (ed.), Seismic Stratigraphy Applications to hydrocarbon exploration. AAPG Memoir,
26, p. 63-97.
WAGONER, J.C. van, MITCHUM, R.M., CAMPION,
K.M., & RAHMANIAN, V.D. (1990). - Siliciclastic
Sequence Stratigraphy in Well Logs, Cores, and
Outcrops. AAPG Methods in Exploration Series, no 7.
423
Acknowledgements
We wish to express our gratitude to the companies which have encouraged us to publish this book. Thus, we are
deeply indebted to TOTAL, Schlumberger, ENSPM (IFP), and Georex which have made a pre-order of this book allowing thereby its printing.
Thanks also to Schlumberger, Halliburton and Baker Atlas to have authorized the reproduction of many of their
figures, most of the time after redrawing, in order to support and illustrate the text.
Finally, for use of previously published illustrations, after redrawing and coloring, we are grateful to all the cited
authors and to the following : Academic Press, The American Association of Petroleum Geologists, American
Geological Institute, Applied Science Publishers, Blackwell Scientific Publications, The Canadian Association of
Petroleum Geologists, Chapman and Hall, Elsevier Publishing Company, W.H. Freeman and Company, The
Geologists Association of London, The Geological Society of London, Geosciences Canada, Gulf Coast Association
of Geological Societies, The Institute of Petroleum, the Journal of Geochemical Exploration, the Journal of Geology,
the Journal of Petroleum Geology, the Journal of Petroleum Technology, McGraw-Hill, Masson, The Offshore
Technology Conference, Oil & Gas Journal, Petroleum Publishing Company, Prentice-Hall Inc., The Society of
Economic Paleontologists and Mineralogists, The Society of Petroleum Engineers, The Society of Professional Well
Log Analysts, Springer, John Wiley & Sons, Wilson and World Oil.
Oberto & Lorenzo SERRA
ii
Serralog 0 2003
INDEX OF REFERENCED AUTHORS
Index Terms
Links
A
Abbott
8
35
210
211
214
219
224
238
56
114
115
Aguilera
384
391
Alger
133
155
Allen
122
125
Adams
155
213
164
168
193
Anderson
82
115
Athy
271
274
Atwater
268
294
Augustin
417
418
Autric
399
400
Babcock
360
361
Baci
398
400
Baker
145
156
Baldwin
112
113
115
272
294
1
34
119
121
155
159
171
180
193
197
236
239
255
259
262
265
294
299
301
306
309
310
312
352
355
391
400
403
422
123
124
155
Beaudoin
33
34
334
352
Beers
57
115
260
262
352
294
422
B
Bates
Beard
391
396
400
Bell
360
391
Bengtson
330
332
333
81
82
117
304
305
306
Biggs
Billings
391
Bissell
244
262
This page has been reformatted by Knovel to provide easier navigation.
307
352
Index Terms
Blatt
Links
48
115
160
193
200
242
254
262
268
294
82
83
116
Bonnetain
190
193
Borg
267
294
Bourke
13
34
Bousset
417
418
422
Boyeldieu
356
377
380
391
Bram
85
115
Brie
145
146
155
385
391
Brown
198
410
422
Burst
272
294
Busch
197
199
236
410
412
191
193
199
Blunden
422
Busk
341
352
Busson
404
422
Butler
356
391
161
162
410
422
Carpentier
396
400
Chermette
344
353
Cheung
27
34
Chiarelli
278
280
285
287
288
289
291
293
294
406
262
263
271
155
244
245
45
115
C
Campbell
422
Chilingar
275
294
Chilingarian
240
250
274
294
127
128
246
262
Christensen
67
85
115
Clarke
38
39
42
Coleman
203
232
237
Collins
416
422
Coogan
270
294
Cox
360
391
Crain
82
115
Curry
206
207
Choquette
231
This page has been reformatted by Knovel to provide easier navigation.
236
Index Terms
Links
D
Daly
39
44
63
66
115
Dallmus
276
294
Dapples
243
262
Daukoru
235
238
Davis
201
Deer
49
115
Delhomme
19
34
147
148
149
151
153
155
186
192
193
237
255
258
376
378
391
56
70
74
115
Dodge
123
155
Doll
405
422
Donath
306
352
Dragoset
81
115
Draxler
85
115
Dubas
333
335
336
347
352
Dumesnil
399
400
Dunoyer de Segonsac
244
262
Dzulinski
170
193
359
392
85
116
124
155
166
167
275
276
295
190
194
333
335
336
341
342
343
344
345
347
348
352
417
418
284
289
Desbrandes
E
Edwards
Emmermann
van Engelhard
Etchecopar
422
Evamy
243
262
Fairbridge
240
244
295
Faivre
380
392
393
Faure
344
353
Fertl
265
271
278
295
406
423
F
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Fisher
198
232
422
423
Folk
157
Ford
233
237
248
249
263
79
80
116
Foster
283
295
Fraser
122
123
125
Frayssinet
112
114
116
Friedman
201
237
Fuchtbauer
250
263
Gaida
154
275
276
295
Galloway
162
194
201
203
204
232
237
Garrels
39
116
Gasparini
56
114
Ghibaudo
200
Gignoux
403
423
Gilreath
191
194
237
330
353
Goetz
191
194
237
4
34
57
115
260
262
396
267
295
267
410
295
G
Golyer
Goodman
400
Gough
360
391
Graton
122
123
85
116
Gressly
197
237
Griggs
306
Gutschick
171
Guy
164
194
82
83
Halbouty
299
353
Haq
418
Grau
H
Haile
Harker
15
34
Hassan
285
295
116
258
400
Hatcher
304
353
Haug
197
237
This page has been reformatted by Knovel to provide easier navigation.
260
396
Index Terms
Links
Head
124
125
Heckel
171
194
Hedberg
271
295
Heliot
27
34
Henkel
76
77
116
Henry
322
324
325
326
327
329
330
339
340
350
400
203
204
353
Herron
262
263
Hitchon
275
295
Hjulstroms
166
167
194
Hobday
162
194
201
237
Holmes
39
116
Hossin
285
295
Hottman
283
288
1
201
277
296
Illing
4
35
Islam
4
35
1
34
119
121
155
159
171
180
193
197
236
239
255
259
262
265
294
299
301
306
309
310
312
352
355
391
400
403
422
Johnson
283
288
295
Jopling
168
169
194
Hutton
295
I
Ibrahim
J
Jackson
K
Karker
139
Katz
277
296
Kenyon
145
156
Knight
275
294
Kohonen
112
113
116
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Krumbein
Krynine
Links
47
48
116
119
120
121
122
136
154
160
194
197
201
230
237
239
241
242
246
263
45
46
48
116
162
296
194
L
Lahee
4
35
Lalanne
417
423
Land
249
263
Lang
289
290
291
Lardenois
408
409
423
Lauger
277
297
Laurier
225
227
Lee
124
156
Leet
3
52
Leith
45
117
Lennon
206
237
Lewis
276
296
Lombard
198
199
Longman
240
263
Luthi
27
35
Lyell
201
237
Malla
398
400
Mann
399
400
Manus
270
294
Maricelli
330
353
Martin
377
384
Mathieu
381
393
Maxwell
267
268
99
117
117
200
237
294
296
M
Mayer
McClelland
113
McCulloh
268
269
McGowen
410
423
McKee
160
163
McKenzie
39
116
Mead
39
45
296
168
117
This page has been reformatted by Knovel to provide easier navigation.
181
194
Index Terms
Links
Mendelson
398
400
Meyer
399
400
Miall
228
237
Middleton
197
201
Miller
268
294
Mitchum
198
Monk
136
154
Moore
197
206
Morris
365
393
Motet
19
34
192
228
236
237
151
155
186
193
255
376
378
156
253
254
237
391
Mutti
200
Neasham
126
127
Nederlof
399
400
129
155
160
195
N
Nurmi
156
O
Otto
P
Payne
199
Peck
273
276
285
297
Perrodon
299
353
395
400
Pettijohn
39
40
41
46
117
121
159
160
161
164
170
195
250
251
263
Philipp
290
296
Pickett
66
67
117
Pilenko
152
Pirson
201
232
237
Pittman
253
Poelchau
420
423
Poster
5
35
Potter
123
156
160
161
169
195
272
296
Powers
159
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Pozzi
417
423
Pray
127
128
246
262
40
155
244
245
41
42
117
241
263
303
353
124
125
156
de Raaf
201
227
237
Ragan
321
353
393
Ramsay
302
304
309
353
355
393
Rao
133
156
Rasmus
145
156
393
Raymer
81
82
117
Reading
197
200
201
237
Reineck
163
164
167
168
169
195
199
201
410
423
Press
Pryor
203
R
Renault
277
Ricci-Lucchi
200
Richardson
165
167
Rider
225
227
237
Rieke
271
274
294
Rivest
153
156
Roberts
268
269
Robertson
269
297
Rodriguez
171
Rogers
114
117
Rose
276
296
Rukhin
277
Rumelhart
113
Russel
260
263
Ryer
277
297
296
124
125
396
400
S
Sabins
235
Sanders
201
Sanyal
85
86
Sarma
133
156
Sawabini
265
297
310
117
This page has been reformatted by Knovel to provide easier navigation.
341
Index Terms
Links
Scheidegger
161
Scholle
201
Schwarzacher
161
195
Scott
161
195
Seilacher
171
195
Selley
125
157
159
160
165
166
170
171
191
194
197
198
199
200
227
229
231
237
238
260
263
410
423
140
157
7
8
13
35
58
69
79
117
131
132
139
140
147
148
155
182
183
184
185
186
187
188
191
195
203
207
210
211
213
214
219
223
224
226
228
233
238
259
263
286
289
292
297
405
406
407
409
412
413
419
420
421
423
Sibbit
99
117
380
392
393
Siever
40
41
42
117
241
263
303
353
Simons
164
165
196
Singh
163
164
167
168
169
195
199
201
410
423
Sippel
250
263
Sloss
48
116
119
120
121
160
194
197
198
201
230
237
239
263
122
157
Souhaitié
27
35
Spearing
201
Spurlin
100
Stadler
254
Suau
100
118
236
384
393
Sulpice
132
139
149
157
203
233
238
405
413
423
Sen
Serra
Sneed
195
201
118
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Sundborg
166
Supernaw
57
118
260
396
397
401
396
397
401
Taylor
267
268
297
Teichert
197
238
Terzaghi
273
276
Theys
141
157
Timko
278
284
Toksoz
398
400
Trask
121
157
Vail
198
409
Verral
267
296
Visher
231
232
122
157
16
35
199
Swanson
261
263
T
285
297
295
V
412
413
423
163
196
198
200
201
202
205
229
238
403
404
409
410
411
412
415
423
197
200
201
227
228
231
232
238
394
238
W
Wadell
van Wagoner
Walker
Walther
199
Walton
170
Washington
42
Watfa
382
384
Weber
235
238
39
46
118
Weir
160
163
168
194
Weller
197
250
263
272
155
Wedepohl
297
Weyl
123
124
Whalen
283
295
39
118
Wickman
This page has been reformatted by Knovel to provide easier navigation.
277
Index Terms
Links
Wilson
253
263
Wolf
240
250
Wyllie
283
297
Zemanek
373
394
Zingg
122
157
Zobell
239
241
262
Z
264
This page has been reformatted by Knovel to provide easier navigation.
263
INDEX – GLOSSARY
Here below are listed terms or expressions used in geology and well logging, followed by a short definition or explanation. When they
are explained more fully in this book, only the relevant chapter and paragraph is referred to. See also the Index and Glossary of
Volume 1 for all the usual well logging terms. The figure refers to the chapter and the pages.
Index Terms
Links
A
Absorption
- the process of taking up by capillary, osmotic, chemical or solvent
action.
- the process by which energy, such as that of electromagnetic or
acoustic waves, is converted into other forms of energy.
Abyssal plain : flat regions at the bottom of major ocean basins with
a water depth greater than 4 000 m.
Accessory minerals : minerals present in such small amount (i.e.
less than 1 %) that their presence or absence is not significant when
considering the mineral composition for classification purposes, but
which can affect some logging measurements if their inherent properties
are considerably different from those of the principal minerals (i.e. zircon
and monazite which have a so high content in thorium and uranium that
they affect the total radioactivity even with a percentage less than 1 %).
Accretion : a gradual increase in size of an inorganic body by the
external addition of new particles deposited by a stream.
Acidic : a descriptive term applied to those igneous rocks that
contain more than 60 % SiO2
Acoustic : of or pertaining to sound.
Acoustic impedance : the product of acoustic velocity and density.
Activation : technique in which the rocks are irradiated with neutrons
that transmute some nuclei into radioisotopes which are characterized
by the energy of the induced gamma rays and by their decay time
schemes.
Actualism :see Uniformitarianism.
Adsorption : adherence of gas molecules, or of ions or molecules
in solution, to the surface of solids with which they are in contact.
Aeolian : pertaining to the wind.
Agglomerate : a pyroclastic rock composed mostly of bombs.
Aggradation :the building-up of the Earth's surface by deposition of
detrital material by a stream.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Aggregate : a mass or body of rock particles or mineral grains or
both.
Albite : pure sodium-feldspar end member in the plagioclase series.
Allochthonous : formed or produced elsewhere than in its present
place.
Allogenic : formed or generated elsewhere, usually at a distant
place.
Alluvial : pertaining to or composed of alluvium, or deposited by a
stream or running water.
Alluvial fan : mass of loose rock material, shaped like an open fan,
deposited by a stream.
Alluvium : a general term for detrital material deposited by a stream
or running water in the bed of the stream or on its flood plain or
delta ,or as a cone at the base of a mountain slope.
Alteration : any change in the chemical or mineralogical composition
of a rock produced by weathering or by the action of hydrothermal
solutions.
Amphibole : a group of dark rock-forming ferromagnesian inosilicates
having the general formula :A2-3 B5 (Si,AI)8 O22 (OH)2, where A =
Mg, Fe2+, Ca, or Na, and B = Mg, Fe2+, Fe3+, Al orTi.
Amphibolite : a metamorphic rock consisting mainly of amphibole
and plagioclase with little or no quartz.
Amplitude: half the height of the crest of a wave above the adjacent
troughs. The maximum value of the displacement in an oscillatory
motion.
Anadiagenesis
Chapter 6, p. 244
Anaerobic : said of organisms that can live in the absence of free
oxygen, or of conditions that exist only in the absence of free oxygen.
Anastomosis [streams] : a product of braiding; esp. an interlacing
network of branching and reuniting channels.
Andesite : a dark-colored, fine-grained extrusive (volcanic) rock.
Anhydrite : anhydrous calcium sulfate of the evaporite group.
Anion : a negatively charged ion.
Anisotropy : the condition of having different properties in different
directions.
Anorthite : pure calcium-feldspar end member of the plagioclase
series.
Anorthoclase : sodi-potassic alkali feldspar.
Anticline : a convex upward fold.
Chapter 8, p. 310
Antiform: a fold whose limbs close upward
Chapter 8, p. 310
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Antithetic : pertaining to minor faults that are oriented opposite to
the major fault with which they are associated.
Aragonite : orthorhombic calcium carbonate with greater density
and hardness and less stability than calcite.
Arcuate delta : a curved or bowed delta with its convex outer
margin facing the sea or lake.
Arenaceous : said of a sediment or sedimentary rock consisting
wholly or partly of sand-size fragments.
Arenite : a general name used for consolidated sedimentary rocks
of sand-size fragments irrespective of composition.
Argillaceous : pertaining to, largely composed of, or containing
clay-size particles or clay minerals.
ARI*: Schlumberger's tool. Acronym for Azimuthal Resistivity
Imager.
Chapter 3, p. 132
Arkose : a feldspar rich, coarse-grained sandstone, pink or reddish.
Arrow plot : a display of dipmeter data.
Chapter 8, p. 330
Ash :fine (< 2 mm in diameter) pyroclastic material.
Asthenosphere : the layer or shell of the Earth below the
lithosphere.
Authigenesis : the process by which new minerals form in place
within a sediment or sedimentary rock.
Chapter 6, p. 252
Authigenic : formed or generated in place.
Autochthonous : formed or produced in the place where now
found.
Avulsion : an abrupt abandonment of a segment of a river channel.
Axial surface
Chapter 8, p. 309
Axis : the line which, moved parallel to itself, generates the form of
a fold.
Azimuth : direction of a horizontal line as measured clockwise from
North on an imaginary horizontal circle.
Azimuth frequency plot
Chapter 8, p. 320
Azoic : said of an environment that is devoid of life.
B
Back roof : the landward side of a reef.
Bar : a generic term for any of various elongate offshore ridges,
banks or mounds of sand, gravel, or other unconsolidated material.
submerged at least at high tide.
Barchan : an isolated crescent-shaped sand dune lying transverse
to the direction of the prevailing wind.
This page has been reformatted by Knovel to provide easier navigation.
Chapter 9, p. 375
Index Terms
Links
Barite : sulfate of barium.
Barrier : an elongate offshore ridge or mass rising above the hightide
level, generally extending parallel to, and at a some distance from,
the shore ,and built up by the action of waves or currents, or by organisms.
Basal-conglomerate : a conglomerate that forms the bottom stratigraphic
unit of a sedimentary series and that rests on a surface of
erosion, thereby marking an unconformity.
Basalt : a general term for dark-colored basic and mafic igneous
rocks, commonly extrusive but locally intrusive.
Basement: the undifferentiated complex of rocks that underlies the
rocks of interest in an area.
Basic : said of an igneous rock having a relatively low silica content,
relatively rich in iron, magnesium and/or calcium, and thus includes most
mafic minerals.
Basin : a low area in the Earth's crust, of tectonic origin, in which
sediments have accumulated.
Bathyal : pertaining to the ocean environment or depth zone
between 200 and 2 000 metres.
Bauxite : a rock composed of a mixture of various amorphous or
crystalline hydrous aluminium oxides and hydroxides. A common
residual of clay deposits in tropical and subtropical regions.
Bay : a wide, curving open indentation, recess, or inlet of a sea into
the land.
Beach : a shore of a body of water, formed and washed by waves
or tides, usually covered by sandy or pebbly material.
Bed : the smallest formal unit in the hierarchy of lithostratigraphic
units, distinguishable from layers above and below.
Chapter 4, p. 160
Bedding : the arrangement of a sedimentary rock in beds or layers
of varying thickness and character.
Chapter 4, p. 161
161
168
169
Bed load : the part of the total stream load that is moved on or
immediately above the stream bed, such as the larger or heavier
particles transported by traction or saltation along the bottom.
Bedset : a group of strata bounded by stratification surfaces.
Chapter 4, p. 163
Bell shape : an evolution of a curve (i.e. SP or resistivity) with depth
drawing the shape of a bell.
Chapter 5, p. 203
Benthic : pertaining to the benthos.
Benthos : those aquatic organisms that live on or within the
sediment at the bottom of a body of water.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Bentonite : a soft, plastic, porous, light-colored rock composed
essentially of clay minerals of the montmorillonite group plus colloid silica,
produced by devitrification and accompagnying chemical alteration
of a glassy igneous material, usually a tuff or volcanic ash.
Biochemical : characterized by, or resulting directly or indirectly
from, the chemical processes and activities of living organisms.
Bioclastic : consisting primarily of fragments of organisms.
Biogenic : produced directly by the physiological activities of
organisms.
Bioherm : a moundlike, domelike, lenslike, or reeflike mass of rock
built up by sedentary organisms, composed almost exclusively of their
calcareous remains.
Biostrome : a distinctly bedded and widely extensive blanketlike
mass of rock built by and composed mainly of the remains of sedentary
organisms.
Biotite : a dark and dense mineral of the mica group.
Bird's-eye fabric : a common pattern in supratidal carbonates in
which former gas bubbles become preserved as open or calcite-filled
cavities. These cavities are typically 2 to 5 mm in diameter and may
constitute 50 % of the rock.
Block
- [part. size] a large, angular rock fragment having a diameter
greater than 256 mm; it may be nearly in place or transported
by gravity or ice.
- [volc] a pyroclastic particle larger than 64 mm ejected from a
volcano in a solid state.
"Blue pattern ": a convention used in dipmeter interpretation. It
corresponds to an increasing dip magnitude with decreasing
depth with nearly uniform azimuth.
Body Force
Chapter 8, p. 330
Chapter 8, p. 301
Bog : waterlogged, spongy ground, consisting primarily of mosses,
containing acidic, decaying vegetation that may develop into peat.
Bomb : a pyroclastic particle larger than 64 mm ejected from a volcano
while viscous but solidified and received its more or less rounded
shape while in flight.
Bone bed : a sedimentary layer characterized by a high proportion
of fossil bones, scales, teeth, coprolites (phosphatic deposits).
Bottomset : a nearly horizontal layer of sediment deposited in front
of the advancing foreset beds.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Boulder : a rock fragment or particle having a diameter greater than
256 mm.
Boundstone : (Dunham's classification] a term used for a sedimentary
carbonate rock whose original components were bound together
during deposition and remained substantially in the position of growth.
Bound water
- water which has become adsorbed to the surfaces of solid particles
or grains. Under natural conditions this water tends to be viscous and
immobile but might not have lost its electrolytic properties.
- water which is chemically bound by becoming part of a crystal lattice.
This water cannot be removed without changing the structure or
composition of the material. It has lost its electrolytic properties.
Break : syn.: discontinuity.
Chapter 11, p. 404
Breccia : a general term for a coarse-grained clastic rock consisting
of angular, broken rock fragments held together by a mineral cement or
in a fine-grained matrix. This implies a minimum transport of fragments.
Brine : a term used for highly saline waters present in restricted
basins.
Brittle : "said of a rock that fractures at less than 3-5% deformation
or strain" (Bates & Jackson, 1980).
Bulk density : the weight of a material divided by its volume
including the volume of its pore spaces.
Bulk modulus
Chapter 8, p. 305
Burrow : a tubular or cylindrical hole made by a mud-eating animal.
Chapter 4, p. 170
Button : a small disc-shaped, button-like electrode used in microresistivity pads (ML, MLL, SHDT, FMS/FMI, EMI, STAR).
C
Cable : a wireline.
Calcarenite : [Grabau's classification] a limestone consisting predominantly
(more than 50 % ) of detrital calcite particles of sand size.
Calcareous : said of a substance that contains more than 10 % and
less than 50 % calcium carbonate.
Calcilutite : [Grabau's classification] a limestone consisting predominantly
(more than 50 %) of detrital calcite particles of silt and/or clay
size.
Calcirudite : [Grabau's classification] a limestone consisting predominantly
(more than 50 % ) of detrital calcite particles larger than sand
size.
Calcite : a calcium carbonate.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Calibration : the process wherein the scale and sensitivity of the
measuring circuit is adjusted to meaningful units.
Chapter 1
Caliper : a wireline logging tool which measures hole diameter.
Capillary forces : .
Cap rock : an impervious rock overlying a reservoir.
Carbonate : a sediment formed by the organic or inorganic precipitation
from aqueous solution of carbonates of calcium, magnesium, or
iron.
Carnallite : an evaporite mineral.
Cast : a sedimentary structure representing the infilling of an original
mark or depression made on top of a soft bed, and preserved as a
solid form on the underside of the overlying stratum.
CAST-V™ : Halliburton's tool. Acronym for Circumferential Acoustic
Scanning Tool.
Chapter 3, p. 132
Cation : a positively charged ion.
Cation exchange : the displacement of a cation bound to a site on
the surface of a solid, as in clay-minerals, by a cation in solution.
Cation Exchange Capacity : symbol CEC: a measure of the extent
to which a substance will supply exchange cations.
Cave : part of a borehole where the hole diameter becomes larger
than the drill bit diameter.
CBIL : Baker Atlas's tool. Acronym for Circumferential Borehole
Imaging Log.
Chapter 3, p. 132
Celestite : sulfate of strontium occurring in deposits of salt, gypsum,
and associated dolomite and shale, and in residual clays.
Cement : mineral material usually chemically precipitated in the spaces
between the individual grains or crystals (pores), thereby binding
them together as a rigid, coherent mass.
Cementation
Chapter 6, p. 242
Cementation factor : the porosity exponent m in Archie's formula.
Syn.: tortuosity factor.
Chalk : a soft, friable, pure, earthy, fine-textured limestone of marine
origin consisting almost wholly (90-99 %) of calcite, formed mainly by
shallow-water accumulation of calcareous tests of floating
microorganisms.
Chamosite : an hydro-alumino-silicate of the chlorite group, rich in
iron. An important constituant of many oolitic and other bedded iron ores.
Channel : an elongate depression where a natural body of water
flows; an abandoned or buried watercourse represented by stream
deposits of gravel and sand.
This page has been reformatted by Knovel to provide easier navigation.
251
Index Terms
Links
Channel lag : a deposit consisting of the coarsest material that settles
out and accumulates along the deepest part of a river channel.
Chemical rock : a sedimentary rock composed primarily of material
formed directly by precipitation from solution or colloidal suspension.
Chert : a hard, extremely dense or compact, dull to semivitreous,
microcrystalline or cryptocrystalline sedimentary rock, consisting
dominantly of interlocking crystals of quartz less than about 30 gm
indiameter. It may contain amorphous silica (opal). It occurs principally
as nodules, less commonly as areally extensive layers.
Chlorite : an hydrous-alumino-silicate of iron and magnesium.
Clastic : pertaining to a rock composed principally of broken fragments
that are derived from preexisting rocks or minerals and that have
been transported some distance from their place of origin.
Clay
- a rock or mineral fragment or a detrital particle of any composition
having a diameter less than 1/256 mm.
- a loose, earthy, extremely fine-grained, natural sediment or soft
rock composed primarily of clay-size or colloidal particles and characterized
by high plasticity and by a considerable content of clay minerals.
- a clay mineral.
Clean : containing no appreciable amount of clay or shale.
Cluster analysis : a procedure for arranging a number of objects in
homogeneous subgroups based on their mutual similarities and
hierarchical relationships.
CLUSTER* : a Schlumberger trade mark for a dip computation
technique.
Coal: a readily combustible rock containing more than 50 % by
weight and more than 70 % by volume of carbonaceous material, formed
from compaction of altered plant remains similar to those in peat.
Cobble : a rock fragment or sediment particle having a diameter in
the range of 64-256 mm.
Chapter 3
Cohesiveness : a mass property of unconsolidated, fine-grained
sediments by which like or unlike particles (having diameters less than
0.01 mm )cohere or stick together by surface forces.
Cohesive strength
Chapter 8, p 303
Compaction
Chapter 6, p. 242
Competent : said of a layer which, in contrast to adjacent layers,
has formed more nearly parallel folds (the adjacent layers being more
nearly in similar folds).
This page has been reformatted by Knovel to provide easier navigation.
Chapter 7
Index Terms
Links
Component : one of a set of chemical compositions the relative
masses of which may be varied to describe all compositions within it.
Composite log
Chapter 2
Chapter 1, p. 21
Composition
Chapter 2
Compressibility :the reciprocal of bulk modulus.
Chapter 8, p. 305
Compression : "a system of forces or stresses that tends to decrease
the volume of a subsatnee" (Blates & jackson", 1980).
Compressive strength
Chapter 9, p. 359
Compressive stress : a normal stress.
Chapter 8, p. 301
305
308
Cone : a structure shaped like a cone.
Chapter 8, p. 327
Confining pressure
Chapter 8, p. 301
Conformable : said of strata or stratification characterized by an
unbroken sequence in which the layers are formed one above the other
in parallel order by regular, uninterrupted deposition under the same
general conditions.
Conglomerate : a coarse-grained clastic sedimentary rock, composed
of rounded to subangular fragments larger than 2 mm in diameter.
Conical fold : a fold model that can be described geometrically by
the rotation of a line about one of its ends, which is fixed.
Chapter 8, p
Consolidated : pertains to a rock framework provided with a degree
of cohesiveness or rigidity by cementation or other binding means.
Core
- a cylindrical section of rock.
Chapter 1, p. 5
- the central zone or nucleus of the Earth's interior.
Correlations
Chapter 11, p. 418
CORIBAND*: a Schlumberger mark for a program of interpretation
for complex lithologies.
Coset: a sedimentary unit composed of two or more sets.
Chapter 4, p. 168
Crack
Chapter 9, p. 355
Craton : a part of the Earth's crust that has attained stability, and has
been little deformed for a prolonged period.
Creep : continuously increasing strain resulting from a small
constant stress acting over a long period of time.
Crest: the highest point or line of a landform.
Chapter 8, p. 309
Crestal line
Chapter 8, p. 309
Crestal surface
Chapter 8, p. 309
Crevasse :
- a wide breach or crack in the bank of a river.
- cracks in the top of a glacier.
This page has been reformatted by Knovel to provide easier navigation.
327
Index Terms
Links
Cross-bedding : cross-stratification in which the cross-beds are
more than 1 cm in thickness.
Crossplot: a graphic plot of one parameter versus another.
Chapter 4, p. 168
Chapter 2, p. 67
Crust : the outermost layer or shell of the Earth.
Crystal : a homogeneous, solid body of a chemical element,
compound, or mixture, having a regularly repeating atomic
arrangement that may be outwardly expressed by plane faces.
Cuttings : rock chips cut by a bit in the process of well drilling.
Cycle
Chapter 1, p. 6
Chapter 5, p. 198
Cyclic sedimentation
Chapter 5
Cyclographic projection
Chapter 8, p. 323
Cylinder
Chapter 8, p. 326
Cylindrical fold : a fold model that can be described geometrically
by the rotation of a line through space parallel to itself.
Chapter 8, p. 326
D
DAPSA plot
Chapter 8, p. 332
DCA* : a Schlumberger mark for a program of Detection of
Conductive Anomalies.
Chapter 9, p. 370
Debris : any surficial accumulation of rock fragments, soil material,
and sometimes organic matter detached from rock masses by chemical
and mechanical means.
Debris flow : a moving mass of rock fragments, soil and mud, more
than half of the particles being larger than sand size.
Decompaction
Chapter 7, p. 293
Deep-sea fan : a submarine equivalent of an alluvial fan. Syn.
turbidite.
Deformation : syn. strain.
Chapter 8, p. 303
Deltaic : pertaining to or characterized by a delta.
Dendrogram : a treelike diagram depicting the mutual relationships
of a group of items sharing a common set of variables.
Chapter 5, p.216
Density current : any current that flows downslope because of it is
denser than the fluid around.
Density stereogram
Chapter 8, p. 321
Depositional environment : a natural geographic entity in which
sediments accumulate.
Depth of investigation : the radial distance from the measure point
of a sonde within which material contributes significantly to the readings
from the sonde.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
DETFRA* : a Schlumberger mark for a program of Detection of
Fractures.
Chapter 9, p. 377
Deviation : departure of a borehole from vertical. Syn. : drift.
Detrital : pertaining to or formed from detritus.
Dextral
Chapter 8, p. 312
Diagenesis
Chapter 6
Diapir : a dome or anticlinal fold in which the overlying rocks have
been ruptured by the squeezing-out of plastic core material (salt, shale,
or igneous intrusions). Syn. : salt dome.
Diapirism :the process by which a diapir is formed.
Diatomite : a siliceous sedimentary rock consisting chiefly of opaline
frustules of the diatom, a unicellular aquatic plant related to the algae.
Dike : a tabular igneous intrusion that cuts across the bedding or
foliation of the country rock.
Chapter 4, p. 171
Dilation : deformation by a change in volume but not shape.
Diorite : a group of plutonic rocks intermediate in composition
between acidic and basic.
Dip : the angle that a structural surface makes with the horizontal,
measured perpendicular to the strike of the structure and in the vertical
plane.
Chapter 8, p. 315
Dipmeter
Chapter 8, p. 315
Dip pattern
Chapter 8, p. 330
Dip slip
Chapter 8, p. 312
Discontinuity : any interruption in sedimentation, whatever its
cause or length, usually a manifestation of nondeposition and
accompanying erosion.
Discordance : lack of parallelism between adjacent strata. Angular
unconformity.
Dispersed : a term used to refer to particles (clays) distributed
within the interstices of the rock framework.
Distal: said of a sedimentary deposit consisting of fine clastics and
formed farthest from the source area.
Distributary : a divergent stream flowing away from the main
stream and not returning to it, as in a delta or on an alluvial plain.
Dolomite : a carbonate of calcium and magnesium.
Dolomitic : said of a rock that contains 10-50 % the mineral
dolomite in the form of a cement and/or grains or crystals.
Dolomitization : the process by which limestone is wholly or partly
converted to dolomite rock or dolomitic limestone.
Chapter 6, p. 255
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Dome : an uplift or anticline structure, either circular or elliptical in
outline, in which the beds dip gently away in all directions.
Drag : the bending of strata on either side of a fault, caused by the
friction of the moving blocks along the fault surface.
Chapter 8
Drift
- [drilling ] the attitude of the borehole; the drift angle or deviation is
the angle between the borehole axis and the vertical.
-(glacial geol.].;
- [geophysics] .
Drumlin : a low, smoothly rounded, elongate oval hill, mound or
ridge of compact glacial till.
DUALDIP*: a Schlumberger mark for a program of dip computation
for SHDT dipmeter tool.
Ductile : said of a rock that is able to sustain 5-10% of strain
before fracturing.
Chapter 8, p. 304
Ductility : a measure of the degree to which a rock is ductile.
Chapter 8, p. 306
E
Effusive : see extrusive.
Eh : oxidation-reduction potential.
Elastic : said of a body in which strains are instantly and totally
recoverable and in which deformation is independent of time.
Elastic behavior
Chapter 8
Chapter 8, p 303
Elastic limit:" the greatest stress that can be developed in a material
without permanent deformation remaining when the stress is
released" (Bates & Jackson, 1980).
Chapter 8, p. 304
Electrobed : corresponds to an interval of depth in which log
responses are nearly constant.
Electrofacies
Chapter 5, p. 206
Electrosequence
Chapter 5, p. 223
Eluvium :fine soil or sand moved and deposited by wind, as in a
sand dune.
EMI™ : Halliburton's trade mark for Electrical Micro Imaging.
End member : one of the two or more pure components of a mixtue.
Chapter 3, p. 132-133
Chapter 2, p. 48
Endogenetic : "derived from within; said of a geological process, or
of its resultant feature or rock, that originates within the Earth. The term
is also applied to chemical precipitates (evaporites) that originate within
the rocks that contain them" (Blates & Jackson, 1980).
Chapter 6
Endogenous : endogenetic.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Entropy : a measure of the degree of mixing of the different kinds of
rock components in a stratigraphic unit.
Environment : "a geographically restricted complex where a sediment
accumulates, described in geomorphic terms and characterized by
physical, chemical and biological conditions, influences or forces "
(Blates & Jackson, 1980).
Chpater 5, p. 228
Environmental effects : effects related to the influence of the
borehole on the measurements made by wireline tools.
Chapter 1
Eogenetic : a term proposed by Choquette & Pray (1970) for the
period of time between final deposition of a sediment and its burial below
the depth to which surface or near surface processes are effective.
Chapter 6, p. 245
Eolian : see aeolian.
Epidiagenesis
Chapter 6, p. 244
Epigenetic : said of a sedimentary mineral, texture, or structure
formed after the deposition of the sediment.
Chapter 6
Epsomite : a hydrous sulfate of magnesium.
Eruptive : said of a rock formed by the solidification of magma.
Esker : a long, narrow, sinuous, steep-sided ridge composed of irregularly
stratified sand and gravel that was deposited by a subglacial or
englacial stream flowing between ice walls or in an ice tunnel of a
stagnant or retreating glacier.
Estuary : the seaward end or the widened funnel-shaped tidal
mouth of a river valley where fresh water comes into contact with
seawater and where tidal effects are evident.
Eustasy : the worldwide sea-level regime and its fluctuations,
caused by absolute changes in the quantity of seawater.
Eustatism : syn. : eustasy.
Euxinic : pertaining to an environment of restricted circulation and
stagnant or anaerobic conditions.
Evaporite : a non clastic sedimentary rock composed primarily of
minerals produced from a saline solution.
Excavation effect : a decrease in the neutron log apparent porosity
reading below that expected on the basis of the hydrogen indices of
the formation components.
Exogenetic : said of processes originating at or near the surface of
the Earth, such as weathering and denudation, and to rocks and
landforms that owe their origin to such processes.
Chapter 6
Exogenous : exogenetic.
Extrusive : said of igneous rocks that has been erupted onto the
surface of the Earth.
This page has been reformatted by Knovel to provide easier navigation.
254
Index Terms
Links
F
Fabric : the orientation in space of the elements composing a
sedimentary rock.
Chapter 3
Facies
Chapter 5
Facies model :
FACIOLOG* : a Schlumberger mark for a program of facies
analysis.
Chapter 5, p. 211
Factor :
- cementation- : the porosity exponent m in Archie's formula.
- formation- : symbol F;
Failure : fracture or rupture of a rock that has been stressed beyond
its ultimate strength.
Chapter 9
FAST plot : a contraction for Formation Anomaly Simulation Trace,
a plot versus depth obtained by the intersection of dip planes with the
borehole considered as a cylinder in space. Dip presentation introduced
by Schlumberger.
Fault
Chapter 8, p. 301
311
Feldspar: a group of abundant rock-forming minerals of general formula
MAI(Al,Si)3O8 where M = K, Na, Ca, Be, Rb, Sr, and Fe. Feldspars
are the most widespread of any mineral group and constitute 60 % of the
Earth's crust. On decomposition, they yield a large part of the clays.
Feldspathic : said of a rock containing feldspar.
Felsic : a mnemonic adjective derived from feldspar + /inad (feldspathoid) + silica + c, and applied to an igneous rock having abundant
fight-colored minerals in its mode; also, applied to those minerals
(quartz, feldspars, feldspathoids, muscovite) as a group.
Ferromagnesian : containing iron and magnesium.
Ferruginous : pertaining to or containing iron.
FIL* : a Schlumberger mark for the Fracture Identification Log.
Chapter 9, p. 370
Fissure
Chapter 9, p. 355
Flank : limb.
Chapter 8, p. 309
Flaser : ripple cross-lamination in which mud streaks are preserved
in the troughs but incompletely or not at all on the crests.
Flexure : syn.: hinge.
Chapter 4, p.169
Chapter 8, p. 309
Fluvial : of or pertaining to a river.
FMI* : acronym for Formation Micro Imager
FMS* : acronyme for Formation MicroScanner tool.
Fold
Chapter 8, p. 301
Footwall : the underlying side of a fault or the wall rock beneath an
inclined fault.
Chapter 8, p. 312
This page has been reformatted by Knovel to provide easier navigation.
308
Index Terms
Links
Fore roof : the seaward side of a reef.
Foreset : pertaining to or forming a steep and advancing frontal
slope, or the sediments deposited on such a slope.
Chapter 4, p. 164
Formation : a general term applied in well logging to the external
environment of the drilled well bore without stratigraphic connotation.
Formation factor: expresses the relationship between the porosity
and the resistivity of the water formation: FR = R0/Rw = a/Φm.
Formation Micro Imager* (FMI): Schlumberger's tool.
Formation MicroScanner* tool (FMS) :Schlumberger's tool.
Chapter 3, p. 132-133
Chapter 3, p. 132
Formation water : water present in the virgin formation under natural
conditions, as opposed to introduced fluids such as mud filtrate.
Fossil : any remains, trace, or imprint of a plant or animal that has
been preserved in the Earth's crust since some past geologic time.
Fracture
Chapter 9
Funnel shape : an evolution of a curve (i.e. SP or resistivity )with
depth drawing the shape of a funnel.
Chapter 5, p. 203
G
Gabbro : a group of dark-colored basic intrusive igneous rocks.
Gash : a small scale tension fracture.
Chapter 9, p. 355
Geochemistry : the study of the distribution and amounts of the
chemical elements in minerals, rocks, ores, soils, waters, and the
atmosphere.
GeoColumn* : a Schlumberger mark for a program of automatic
lithofacies determination. Syn. : LITHO*.
Chapter 2, p. 94
GEODIP*: a Schlumberger mark for a dip computation program by
pattern recognition technique written for the HDT dipmeter tool.
GEOGRAM* : a Schlumberger mark for a program of synthetic
seismogram.
Geology : the study of the planet Earth.
Geometry : the three dimensional (length, width and thickness)
shape of a sedimentary body.
Geophysics : study of the Earth by quantitative physical methods.
Geostatic pressure : the vertical pressure at a point equal to the
pressure caused by the weight of the column of overlying rock.
Chapter 8, p. 301
Glacial : of or relating to the presence and activities of ice or
glaciers.
Glacier: a large mass of ice.
Glass : an amorphous product of the rapid cooling of a magma.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Glauconite : a dull-green earthy or granular mineral of the mica
group. It occurs abundantly in greensand, and seems to be forming in
the marine environment. It is an indicator of very slow sedimentation.
Gneiss : a foliated metamorphic rock.
Chapter 2, p. 82
Goethte : an hydrous oxide of iron, the commonest constituent of limonite.
Graben : an elongate, relatively depressed crustal unit or block that
is bounded by faults on its long sides.
Graded bedding : a gradual and progressive change in grain size.
Chapter 4, p. 169
Grain : a mineral or rock particle of all sizes more or less rounded.
Chapter 3, p. 120
Grain size : the general dimensions of grains or particles in a
sediment or rock.
Chapter 3, p. 120
Grainstone : [Dunham's classification] a term used for a mud-free,
grain-supported, carbonate sedimentary rock.
Chapter 3, p. 128
Granite : a plutonic acidic rock in which quartz constitutes 10 to 50
% of the felsic components.
Granulite : a metamorphic rock.
Chapter 2, p. 172
Chapter 2, p. 82
Granulometry : the measurement of grain size.
Gravel : a particle having a diameter in the range of 2-20 mm. Syn.
pebble.
Graywacke
Chapter 2, p. 38
Green pattern : a convention in dipmeter interpretation. It represents
a succession of dips of relative constant azimuth and magnitude.
Chapter 8, p. 332
Greensand : a sand having a greenish colour, consisting largely of
dark greenish grains of glauconite.
Growth fault : a fault in sedimentary rock that forms
contemporaneously and continuously with deposition.
Chapter 8, p. 314
Gypsum : hydrous calcium sulfate of the evaporite group.
H
Haematite: iron oxide: Fe2O3.
Halite : sodium chloride of the evaporite group. Syn. : salt.
Halmyrolysis : the geochemical reaction of sea water and
sediments in an area of little or no sedimentation.
Chapter 6, p. 244
Halokinesis : a general term for the study of the mechanism of salt
movement and related structures.
Hanging wall : the overlying side of a fault or the wall rock above a
fault.
Chapter 8, p. 312
Hard-ground : a zone at the sea bottom, usually a few cm thick, the
sediment of which is lithified to form a hardened surface, often
encrusted, bored, and solution-ridden. It implies a gap in sedimentation.
This page has been reformatted by Knovel to provide easier navigation.
141
Index Terms
Links
Hardness : the resistance of a mineral to scratching.
Heave : "in a fault, the horizontal component of deparation or
displacement" (Bates & Jackson, 1980).
Chapter 8, p. 312
Heavy mineral : a detrital mineral from a sedimentary rock, having
a specific gravity higher than a standard (usually 2.85), and commonly
forming as an accessory mineral (less than 1 % ).
Hemipelagic : deep-sea sediment in which more than 25 % the fraction
coarser than 5 μm is of terrigenous, volcanogenic, and/or neritic origin.
Heterogeneous : said of a bed showing numerous uncorrected
events on dipmeter or FMS resistivity curves.
Hinge : syn. : flexure.
Chapter 8, p. 310
Homogeneous : said of a bed without any resistivity variations on
the dipmeter or FMS resistivity curves.
Horizontal slip : in a fault, the horizontal component of the net slip.
Syn. : heave.
Chapter 8, p. 312
Hornblende : the commonest mineral of the amphibole group.
Horst : an elongate, relatively uplifted crustal unit or block that is
bounded by faults on its long sides.
Humic : pertaining to or derived from humus.
Humus : the generally dark, more or less stable part of the organic
matter of the soil.
Hydrostatic pressure : stress that is uniform in all directions. The
pressure exerted by the water at any given point in a body of water at
rest.
Chapter 7, p. 266
I
Iceberg : a large, massive piece of floating or stranded glacier ice of
any shape, detached from the front of a glacier into a body of water.
Igneous : said of a rock or mineral that solidified from molten or partly
molten material, i.e. from a magma.
Chapter 2, p. 37
Illte : a clay mineral containing less potassium and more water than
true mica.
Imbibition : the absorption of a fluid by a granular rock or any other
porous material under the force of capillary attraction in the absence of
any pressure.
Immature : said of a clastic sediment characterized by unstable
minerals (i.e. feldspars and plagioclases), abundance of mobile oxides
(i.e. alumina), presence of weatherable material (such as clay), and
poorly sorted and angular grains, indicating processes (i.e. transport and
weathering) acting over a short time and/or with a low intensity.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Injection : the forcing, under abnormal pressure, of sedimentary
material (downward, upward, or laterally) into a pre-existing deposit or
rock.
Intertidal : pertaining to the benthic ocean environment or depth
zone between high and low tide. Syn. : littoral.
Intrusion : the process of emplacement of magma in pre-existing
rock.
Intrusive : of or pertaining to intrusion, both the processes and the
rock so formed.
Invasion : the process by which the mud filtrate penetrates in a
porous rock.
Irreducible saturation : it corresponds to the minimum saturation of
a fluid when the fluid is displaced from a porous medium by another fluid
immiscible with the first.
Isobar : a line on a map connecting points of equal pressure.
Isobath : a line on a map connecting points of equal depth.
Isochore : a line on a map connecting points of equal drilled thickness.
Isochrone : a line on a map connecting points of equal travel time.
Isohypse : a line on a map connecting points of equal elevation.
Isolith : a line on a map connecting points of equal aggregate thickness of a given lithologic facies within a formation.
Isomorphic : having identical or similar form.
Isopach : a line on a map connecting points of equal true thickness.
Isopic : said of sedimentary rocks of the same facies.
Isopycnic : a line on a map connecting points of equal density.
Isorad : a line on a map connecting points of equal radioactivity.
Isostasy : the condition of equilibrium, comparable to floating, of the
units of the lithosphere above the asthenosphere.
Isotope : one of the two or more species of the same element
having the same number of protons in the nucleus but differing from one
another by the number of neutrons.
Isotropy : the condition of having properties that are uniform in all
directions.
J
Joint : a surface of fracture in a rock without displacement.
Chapter 9, p. 355
K
Kainite : a mineral of the evaporite group.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Kame : a low mound, knob, hummock, or short irregular ridge,
composed of stratified sand and gravel deposited by a subglacial
stream as a fan or delta at the margin of a melting glacier.
Kaolinite : a common clay mineral of the kaolin group, generally
derived from alteration of alkali feldspars and micas.
Karst : a type of topography formed on carbonate or gypsum rocks
by dissolution.
Kettle : a steep-sided, usually basin- or bowl-shaped hole or
depression in glacial-drift deposits.
Kieserite : a hydrous sulfate of magnesium of the evaporite group.
L
Labile : said of rocks and minerals that are mechanically and
chemically unstable.
Lacustrine : pertaining to, produced by, or formed in a lake.
Lamina : the thinnest recognizable unit layer.
Chapter 4, p. 160
Laminated : said of a rock that consists of laminae.
Lamination : the finest stratification or bedding.
Langbeinite : a sulfate of potassium and magnesium of the
evaporate group.
Layer : a general term for any tabular body of rock.
Leaching : selective removal of soluble minerals by throughgoing
water.
Chapter 6, p. 243
Lens : a geologic deposit bounded by converging surfaces.
Lenticular bedding : a form of interbedded mud and ripple
crosslaminated sand, in which the ripples or lenses are
discontinuous vertically and horizontally.
Chapter 4, p. 163
Lignite : a brownish-black organic rock that is intermediate between
peat and coal.
Limb : that area of a fold between adjacent fold hinges.
Limestone : a sedimentary rock consisting of more than 50 %
calcium carbonate.
Limnic : said of coal deposits formed inland in freshwater basins,
peat bogs, or swamps.
Limonite : a general term for a group of brown, amorphous naturally
occurring hydrous ferric oxides.
Listric : a curvilinear, concave spoon-shaped, usually pointing
upward, surface of fracture or fault, which becomes less steep as one
goes deeper, becoming nearly horizontal at some depth.
Chapter 8
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Lithification : the conversion of a newly deposited unconsolidated
sediment into a coherent solid rock.
LITHO* : see GeoColumn.
Lithofacies : a facies characterized by particular lithologic features.
Lithology : the description of rocks on the basis of colour,
mineralogic composition and texture (grain size).
Lithosphere : a layer of strength relative to the underlying asthenosphere.
It corresponds to the relatively rigid outer shell of the Earth
comprising the crust and upper mantle.
Lithostatic pressure
Chapter 7, p. 266
Littoral : syn. : intertidal.
Load : the material that is moved or carried by a stream, a glacier,
the wind, or waves, tides and currents.
Lobate delta : syn. : arcuate.
LOCDIP*: a Schlumberger mark for a dip computation program by
derivative technique written for the SHDT dipmeter tool.
Log : a continuous record of a parameter as a function of depth.
Longshore bar : a low, elongate sand ridge, built chiefly by wave
action, occurring at some distance from, and extending generally
parallel with, the shoreline.
Lutite : a general term used for consolidated rocks composed of silt
and/or clay.
M
Mafic : a mnemonic term derived from magnesium + ferric + is to
denote ferromagnesian minerals.
Magma : a naturally molten mass, formed within the crust or upper
mantle, which may solidify to form an igneous rock.
Mantle : this portion of the Earth's interior lying between the crust
and the core.
Marble : a metamorphosed carbonate (chiefly limestone).
Marker
Chapter 2, p. 82
Chapter 11, p. 420
Marl : an argillaceous limestone.
Marsh : a water-saturated, poorly drained area, intermittently or
permanently water-covered, having aquatic and grasslike
vegetation, without the formation of peat.
Massive : said of a rock that occurs in very thick homogeneous
beds.
Chapter 4, p. 164
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Matrix
- for a log analyst the solid framework of rock, except shale, which
surrounds pore volume.
- for a geologist the smaller or finer-grained material filling the
interstices between the coarser grains or particles of a sediment
or sedimentary rock.
Chapter 3
Mature : said of a clastic sediment characterized by stable minerals
(i.e. quartz), deficiency of the more mobile oxides (such as soda),
absence of weatherable minerals (such as clay), and well sorted but
subangular to angular grains, indicating processes acting over a long
time and with a high intensity.
Meander : one of a series of regular freely developing sinuous
curves, bends, loops, turns, or windings in the course of a stream.
Meandering stream : a stream having a pattern of successive
meanders.
Mechanical behavior
Chapter 8, p. 301
Mesogenetic
Chapter 6, p. 245
Metamorphism :the processes by which changes in solid rocks
under influence of heat, pressure and chemically active fluids.
Mica : a group of minerals pertaining to phyllosilicates of general
formula (K,Na,Ca) (Mg,Fe,Li,AI)2-3(Al,Si)4O10(OH,F)2
Micrite : a descriptive term for carbonate mud with crystals less than
4 microns in diameter.
Microcline : the triclinic form of potassium feldspar.
Migration : the movement of liquid and gaseous hydrocarbons fromtheir source rocks through permeable formations into reservoir rocks.
Mineral : a naturally occurring inorganic element or compound
having an orderly internal structure and characteristic chemical
composition, crystal form, and physical properties.
Chapter 2, p. 45
Mineralogy : the study of minerals.
Mixed-layer mineral : a mineral whose structure consists of
alternating layers of clays minerals and/or mica minerals.
Mode : the mode is the percentage (by weight) of the individual
minerals which make up a rock.
Modulus of elasticity : the ratio of stress to its corresponding strain
under given conditions of load for materials that deform elastically.
Mohr stress circle
Chapter 8, p. 303
Mohr stress envelope
Chapter 8, p. 303
Monogenetic : said of a conglomerate composed of a single type of
rock.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Montmorillonite : a group of expanding-lattice clay minerals. Syn.
smectite.
Moraine : an accumulation of material which has been transported
or deposited by ice.
Mouth : the place of discharge of a stream.
Mud cake : the residue deposited on the borehole wall as the mud
looses filtrate into porous, permeable formations.
Mud filtrate : the effluent of the continuous phase liquid of drilling
mud which penetrates (invades) porous and permeable formations.
Mudflow : a general term for a mass-movement landform and a process
characterized by a flowing mass of predominantly fine-grained
earth material possessing a high degree of fluidity during movement.
Mudstone
- an indurated mud having the texture and composition of shale.
- [Dunham' s classification] a term used for a mud-supported
carbonate sedimentary rock containing less than 10 % grains.
Chapter 3, p. 128
Muscovite : a white mineral of the mica group.
N
Nadir : the point on the celestial sphere that is directly beneath the
observer and directly opposite the zenith.
Chapter 8, p. 322
Natron : hydrous sodium carbonate occurring mainly in solution in
soda lakes or in saline residues.
Natural levee : a long broad low ridge or embankment of sand and
coarse silt, built by a stream on its flood plain and along both banks of
its channel.
Neogenesis : the formation of new minerals by diagenesis or
metamorphism.
Neomorphism : term suggested by Folk (1965) for all transformations
between one mineral and itself or a polymorph.
Chapter 6, p. 255
Neritic : Pertaining to the ocean environment or depth zone between
low-tide level and approximately the edge of the continental shelf.
Net : a stereographic or an equal-area projection of a sphere in
which the network of meridians and parallels forms a coordinate system.
Net slip : the distance between two formerly adjacent points on
either side of a fault, measured on the fault surface or parallel to it.
Chapter 8, p. 311
Nodule : a small, irregularly rounded knot, mass, or lump of a mineral
or mineral aggregate (i.e. pyritic nodules in a coal bed, chert nodules
in limestone, anhydritic nodules in limestone or dolomite, phosphatic
nodules in marine strata).
Chapter 6, p. 252
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Norm
- the theoretical mineral composition of a rock expressed in terms of
normative mineral molecules that have been determined by specific
chemical analyses for the purpose of classification and comparison.
- recognized type as reference.
Normal stress : that component of stress which is perpendicular to
a given plane.
Chapter 8, p. 301
Normative mineral : a mineral whose presence in a rock is
theoretically possible on the basis of certain chemical analysis.
O
Offlap :the progressive offshore regression of the updip terminations
of the sedimentary units within a conformable sequence of rocks.
Oligomictic : said of a clastic sedimentary rock composed of a
single rock type.
Olistostrome : a sedimentary deposit accumulated as a semifluid
body by submarine gravity sliding or slumping.
Olivine : a group of common rock forming minerals of basic, ultrabasic, low silica igneous rocks (gabbro, basalt) of formula
(Mg,Fe,Mn,Ca)2SiO4
Onlap : an overlap characterized by the regular and progressive pinching
out, toward the margins or shores of a depositional basin, of the
sedimentary units within a conformable sequence of rocks.
Oolite : a sedimentary rock, usually a limestone, made up chiefly of
ooliths.
Oolith : one of the small round or ovate accretionary bodies in a
sedimentary rock, having the size of a sand.
Opal : a mineral or mineral gel of the silica group, having a varying
proportion of water.
Orogeny : the process by which structures within fold-belt
mountainous areas were formed.
Orthoquartzite
Chapter 2
Orthose : the monoclinic form of the potassium feldspar. Syn.
orthoclase.
Outcrop : that part of a geologic formation or structure that appears
at the surface of the Earth.
Outwash : stratified detritus removed or a washed outs from a glacier
by meltwater streams and deposited in front of or beyond the end
moraine or the margin of an active glacier.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Overburden : the upper part of a sedimentary deposit compressing
and consolidating the material below.
Overburden pressure : syn. geostatic pressure.
Chapter 7, p. 267
Overlap : a general term referring to the extension of marine,
lacustrine, or terrestrial strata beyond underlying rocks.
Overlay : graphic data on a transparent sheet to be superimposed
on another sheet.
Overpressure : pressure in excess of lithostatic pressure, from
tectonic stress.
Oxbow : the abandoned crescent- or bow-shaped channel of a
former meander.
P
Packing :the manner of arrangement of the solid clastic particles in
a sediment or sedimentary rock.
Chapter 3, p. 122
Packstone : [Dunham's classification] a term used for a sedimentary
carbonate rock whose granular material is arranged in a
selfsupporting framework.
Chapter 3, p. 128
Paralic : said of coal deposits formed along the margin of the sea.
Particle : a general term for a separate or distinct unit in a rock
without restriction as size, shape, composition, or internal structure.
Peat: an unconsolidated deposit of semicarbonized plant remains in
a watersaturated environment such as bog or fen, and of persistently
high moisture content.
Pebble : a general term for a small roundish rock fragment having a
diameter in the range of 4-64 mm.
Pelagic : said of marine organisms whose environment is the open
ocean, rather than the bottom or shore areas.
Pelite : syn.: lutite.
Pellet: a silt or sand-sized aggregation of carbonate mud, generally
fecal in origin.
Permeability : the property of a porous rock for transmitting fluid.
pH : the negative logo of the hydrogen-ion activity in a solution; a
measure of the acidity or basicity of a solution.
Pitch : the angle between the horizontal and any linear feature.
Chapter 8, p. 312
Plagioclase : a group of triclinic feldspars of general formula
(Na,Ca)Al(Si,Al)Si2O8, which form a complete solid-solution series from
albite (pure Na) to anorthite (pure Ca) .They are among the commonest
rock-forming minerals of igneous rocks.
Planktonic : said of that type of pelagic organism which floats.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Plastic behavior
Chapter 8, p.305
Platform :that part of a continent that is covered by flat-lying or gently
tilted strata, mainly sedimentary.
Playa lake : a shallow, intermittent lake in an arid or semiarid region.
Plunge : the inclination of a fold axis or other linear structure,
measured in the vertical plane.
Plutonic : pertaining to igneous rocks formed at great depth.
Chapter 8, p. 310
Chapter 2, p. 37
Point bar : one of a series of low, arcuate ridges of sand and gravel
developed on the inside of a growing meander by the slow addition of
individual accretions accompagnying migration of the channel toward
the outer bank.
Poisson's ratio
Chapter 8, p. 305
Polar plot
Chapter 8, p. 320
Polyhalite : a mineral of the evaporite group.
Polymictic : said of a clastic sedimentary rock composed of
manyrock types such as arkose, graywacke, conglomerate.
Pore : a small to minute opening or passageway in a rock. Syn.
interstice.
Pore-bridging
Chapter 3, p. 126
Pore-fillingChapter 3, p. 126
Pore-lining
Chapter 3, p. 126
Porogenesis
Chapter 6, p. 245
Poronecrosis
Chapter 6, p. 245
Porosity : the percentage of the bulk volume of a rock that is
occupied by interstices, whether isolated or connected.
Chapters 1
3
Pressure : the force exerted across a real or imaginary surface
divided by the area of that surface.
Chapter 8
Principal axis of stress
Chapter 8, p. 302
Principal plane of stress
Chapter 8, p. 302
Proximal : said of a sedimentary deposit formed nearest the source
area.
Pyroclastic : pertaining to clastic rock material formed by volcanic
explosion.
Pyroxene : a group of dark rock-forming minerals having the general
formula ABSixOe, where A = Ca, Na, Mg, or Fe2+, and B = Mg, Fe2+,
Fe3+, Fe, Cr, Mn, or Al. They constitute a common constituent of
igneous rocks.
This page has been reformatted by Knovel to provide easier navigation.
2
Index Terms
Links
Q
Quartz : crystalline silica, an important rock forming mineral. It is,
next to feldspar, the commonest mineral and has a widespread
distribution in igneous, metamorphic and sedimentary rocks.
Quartzarenite : a sandstone that is composed primarily of quartz
(more than 95 %).
Chapter 2
Quartzite : a very hard but not metamorphosed sandstone, consisting
chiefly of quartz grains that have been completely cemented.
Quartzose : containing quartz as a principal constituent.
Quartz wacke : a moderately well-sorted, commonly fine-grained
sandstone containing up to 90 % quartz and chert, and with more than
10 % argillaceous matrix, less than 10 % feldspar, and less than 10 %
rock fragments.
Quick Look : a general term for a rapid survey of logs.
R
Radiolarite : a comparatively hard fine-grained chertlike homogeneous
consolidated rock composed predominantly of the remains of
Radiolaria.
Recrystallization : the formation, essentially in the solid state, of
new crystalline mineral grains in a rock.
Recumbent fold : an overturn fold.
Chapter 6, p. 242
Chapter 8, p. 311
Redoxomorphism : a diagenetic phenomenum characterized by
mineral changes primarily due to oxidation and reduction. It is typical of
unlithified sediments.
"Red pattern" : a convention used in dipmeter interpretation to
denote decreasing formation dip with decreasing depth with a near
constant azimuth.
Chapter 8, p. 330
Reef : a ridgelike or moundlike structure, layered or massive, built
by sedentary calcareous organisms.
Regression : the retreat or contraction of the sea from land areas.
Repeat section : a short section of a log that is recorded in addition
to the main survey section in order to provide an inter-run comparison of
log similarity, and, therefore, instrument stability and repeatability.
Chapter 1, p. 18
Reservoir rock : a porous and permeable rock.
Rhyolite : a group of extrusive igneous rocks.
Rhythm
Chapter 5, p. 198
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Rhythmic sedimentation : the consistent repetition of a regular
sequence of two or more sedimentation units organized in a particular
order and indicating a frequent and predictable recurrence or pattern of
the same sequence of conditions.
Chapter 5, p. 198
Rift : a long, narrow continental trough that is bounded by normal
faults.
Rigidity : the property of a material to resist applied stress that
would tend to distort it.
Chapter 8, p. 305
Ripple mark : an undulatory surface consisting of alternating subparallel small-scale ridges and hollows formed at the interface between
a fluid and incoherent sedimentary material.
Chapter 4, p. 164-168
Rotation : angle formed by both parts of the same bed after separation
and measured on a plane perpendicular to the axis of rotation.
Chapter 8, p. 312
Roundness : the degree of abrasion of a clastic particle as shown
by the sharpness of its edges and corners.
Chapter 3, p. 122
Rudite : a general term used for consolidated sedimentary rocks
composed of rounded and angular fragments coarser than sand.
Rupture point
Chapter 9
Rupture strength : the differential stress that a material sustains at
the instant of rupture.
S
Sabkha : a supratidal environment of sedimentation, formed under
arid or semiarid conditions.
Salt : syn. : halite.
Saltation : a mode of sediment transport in which the particles are
moved forward in a series of short intermittent jumps.
Salt dome : syn.: diapir.
Sand
- a rock fragment or detrital particle having a diameter in the range
of 1/16 to 2 mm.
Chapter 3
- a loose aggregate of unlithified mineral or rock particle of sand
size.
Sandstone : a lithified, consolidated sand.
Sapropel : an unconsolidated, jellylike ooze or sludge composed of
plant remains, mostly algae, macerating and putrefying in an anaerobic
environment on the shallow bottom of lakes and seas. It may be a
source material for hydrocarbons.
Sapropelic : pertaining to or derived from sapropel, indicating a high
sulfate or reducing environment.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
SARABAND* : a Schlumberger mark for a program of interpretation
of shaly sand.
Saturation : the percentage of the pore volume occupied by a
specific fluid.
SCAT plot : acronym for Statistical Curvature Analysis Techniques
intorduce by Bengtson, 1981.
Chapter 8, p. 332
Schist : a strongly foliated crystalline rock, formed by dynamic
metamorphism.
Schmidt stereonet
Chapter 8, p. 324
Scour and fill : a process of aternate excavation and refilling of a
channel, as by a stream or the tides.
Sedimentary rock
Chapter 4
Chapter 2, p. 38
Sedimentary structure : a structure in a sedimentary rock.
Sedimentation unit :
Chapter 4
Chapters 1
4
Chapter 5, p. 198
Chapter 11, p. 410
Seif : a very large, sharp-crested tapering longitudinal dune.
Seismogram : a seismic record.
Sequence
Shale : a fine-grained detrital sedimentary rock, formed by the
consolidation of clay, silt, or mud. It is characterized by finely laminated
structure, and by an appreciable content of clay minerals and detrital
quartz.
Shear modulus :syn. modulus of rigidity.
Chapter 8, p. 305
Shear strength : the internal resistance of a body to shear stress.
Chapter 9, p. 359
Shear stress : that component of stress which acts tangential to a
plane through any given point in a body.
Chapter 8, p. 301
Shelf : a stable cratonic area that was periodically flooded by
shallow marine waters and received a relatively thin,
well-winnowed cover of sediment.
Siderite : carbonate of iron.
Sill: a tabular igneous intrusion that parallels the planar structure of
the surrounding rock.
Silt
- a rock fragment or detrital particle having a diameter in the range
of 1/256 to 1/16 mm.
- a loose aggregate of unlithified mineral or rock particles of silt size.
Siltstone : an indurated silt.
Sinistral: pertaining to the left.
Slate : a metamorphic rock.
Chapter 9, p. 355
Chapter 2, p. 38
Slickenside : a polished and smoothly striated surface that results
from friction along a fault plane.
Chapter 8, p. 312
This page has been reformatted by Knovel to provide easier navigation.
82
Index Terms
Links
Slip : the relative displacement of formerly adjacent points on
opposite sides of a fault.
Chapter 8, p. 312
Slump : a landslide characterized by a shearing and rotary
movement of a generally independent mass of rock.
Chapter 4, p. 171
SODA plot: acronym for Separation Of Dip and Azimuth.
Chapter 8, p. 333
Solution : a process of chemical weathering by which mineral and
rock material passes into solution.
Sorting : the spread or range of particle-size distribution.
Source rock
Chapter 3 p. 121
Chapter 10
- the parent rock from which other sediments or rocks are derived.
- sedimentary rock in which organic material was transformed in
hydrocarbons under pressure, heat and time influences.
Sparite : a descriptive term for clean, relatively coarse-grained
calcite accumulated during deposition or introduced later as a cement.
Sphericity : the relation to each other of the various diameters of a
particle.
Chapter 3, p. 122
Stack : the sum of several seismic traces or log runs.
State of stress
Chapter 8, p. 302
Stereogram :a graphic diagram on a plane surface, giving a threedimensional representation.
Chapter 8, p. 321-329
Stereographic projection
Chapter 8, p. 322
Stereonet
Chapter 8, p . 320
Stick plot
Chapter 8, p. 345
Stoneley wave : a type of guided wave.
Chapter 9, p. 367
Strain : change in the shape or volume of a body as a result of
stress.
Chapter 8, p. 303
308
Stratification : the formation, accumulation, or deposition of
material in layers.
Chapter 4
Stratigraphy: the science of rock strata. It is concerned with all
characters and attributes of rocks (succession, age, form, distribution,
composition, fossil content, geophysical and geochemical properties.
STRATIM* : a presentation of the borehole image obtained from the
SHDT dipmeter data.
Chapter 8
Stratum : a tabular or sheetlike body or layer of sedimentary rock.
Strength : the ability to withstand differential stress.
Chapter 8, p. 303
Chapter 9, p. 359
Stress : the force per unit area acting on any surface within it.
Chapter 8, p. 301
Stress components
Chapter 8, p. 302
Stress difference : the difference between the greatest and the
least of the three principal stresses.
Chapter 9, p. 357
This page has been reformatted by Knovel to provide easier navigation.
307
Index Terms
Links
Stress ellipsoid : a geometric representation of the state of stress
at a point.
Chapter 8, p. 302
Stress field : the state of stress.
Chapter 8, p. 302
Stress-strain diagram
Chapter 8, p. 304
Stress-strain ratio
Chapter 8, p. 305
Chapter 9, p. 357
Stretch : the measure of the change in length of a line.
Strike : the direction or trend taken by a structural surface (bedding
or fault plane) as it intersects the horizontal.
Strike slip
Chapter 8
Chapter 8, p. 311
Structural geology : the branch of geology that deals with the form,
arrangement, and internal structure of the rocks.
Chapter 8
Structure
- a megascopic feature of a rock mass or rock unit.
Chapter 4
- the general disposition, attitude, arrangement, or relative positions
of the rock masses of a region or area.
Chapter 8
Stylolite : a surface or contact marked by an irregular and
interlocking penetration of the two sides.
Stylolitization : a pressure-controlled solution phenomenon.
Chapter 6, p. 243
Chapter 6, p. 243
257
Chapter 8, p. 301
Chapter 9
Subsidence : the sudden sinking or gradual downward settling of
the Earth's surface with little or no horizontal motion.
Subtidal : below low tide level.
Supermature : said of a mature clastic sediment whose well-sorted
grains have become subrounded to well-rounded.
Supratidal: above high tide level.
Surface force
Suspension : a mode of sediment transport in which the upward
currents in eddies of turbulent flow are capable of supporting the weight
of sediment particles and keeping them indefinitely held in the
surrounding fluid.
Swamp : an area intermittently or permanently covered with water,
having shrubs and trees but essentially without the accumulation of
peat.
Sylvite : potassium chloride of the evaporite group.
Syncline : a concave upward fold.
Chapter 8, p. 310
Syndiagenesis
Chapter 6, p. 244
SYNDIP*
Chapters 3 and 4
Synform : a fold whose limbs close downward.
Chapter 8, p. 310
Syngenetic : said of a primary sedimentary structure formed
contemporaneously with the deposition of the sediment.
Chapter 4, p.159
Synsedimentary : accompanying sedimentation.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
T
Tectonics : a branch of geology dealing with the broad architecture
of the outer part of the Earth.
Chapter 8
Telogenetic
Chapter 6, p. 245
Tensil strength
Chapter 9, p. 359
Tension : a state of stress in which tensile stresses predominate.
Chapter 8, p. 301
Terrigenous : derived from the land or continent.
Texture : the general physical appearance or character of a rock.
Chapter 2
Throw : the amount of vertical displacement on a fault.
Chapter 8, p. 311
Thrust : an overriding movement of one crustal unit over another.
Chapter 8, p. 312
Tidal flat : an extensive, nearly horizontal, marshy or barren tract of
land that is alternately covered and uncovered by the tide.
Till : dominantly unsorted and unstratified drift.
Toe : the lowest part of a slope.
Toeset : the forward part of a tangential foreset bed.
Topset : one of the nearly horizontal layers of sediments deposited
on the top surface of an advancing delta and continuous with the
landward alluvial plain.
Torsion
Chapter 8, p. 301
Tortuosity factor : syn. : cementation factor.
Traction
- the stress vector acting across a particular plane in a body.
- a mode of sediment transport in which the particles are swept
along and parallel to a bottom surface by rolling, sliding, dragging,
pushing or saltation.
Transgression : the spread or extension of the sea over land areas.
Transportation : a phase of sedimentation that includes the movement
by natural agents of sediment, either as solid particles or in solution,
from one place to another.
Trap : any barrier to the upward movement of hydrocarbons
allowing them to accumulate.
Chapter 8
Trona : a bicarbonate of sodium ourring in saline residues.
Trough
Chapter 8
Trough cross-bedding : cross-bedding in which the lower
bounding surfaces are curved surfaces of erosion; it results
from local scour and subsequent deposition.
Chapter 4, p. 168
Tuff : a general term for all consolidated pyroclastic rocks.
Turbidite : a sediment or rock deposited from a turbidity current.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
U
Unconformable : said of strata or stratification exhibiting the
relation of unconformity to the older underlying rocks.
Unconformity : a substantial break or gap in the geologic record.
Chapter 11, p. 405
Uniformitarianism : the fundamental principle or doctrine that geologic
processes and natural laws now operating to modify the Earth's
crust have acted in the same regular manner and with essentially the
same intensity throughout geologic time, and that past geologic events
can be explained by phenomena and forces observable today.
Introduction.
Unstable : said of a constituent of a sedimentary rock or mineral that
does not resist further mineralogic change under weathering.
Upthrow :
- the amount of upward vertical displacement of a fault.
- the upthrown side of a fault.
Chapter 8, p. 312
V
Vadose zone : zone of aeration.
Chapter 6
Varve : a glaciolacustrine layer seasonally deposited in a glacial
lake.
Vertical resolution : the minimum thickness of formation that can
be distinguished and fully characterized by a tool under operating
conditions.
Viscosity : the property of a substance to offer internal resistance to
flow.
Viscous behaviour
Chapter 8, p. 305
VOLAN* : a Schlumberger mark for a program of interpretation of
shaly sand.
Volcanic : pertaining to the activities, structures, or rock types of a
volcano.
Vug : a small cavity in a rock.
Chapter 6, p. 243
W
Wacke : a durty sandstone containing more than 10 % argillaceous
matrix.
Wackestone : [Dunham's classification] a term used for a mudsupported
carbonate sedimentary rock containing more than 10 % grains.
Wadi: the channel in an arid region that is usually dry except during
the rainy season.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Washing : the selective sorting, and removal, of fine-grained
sediments by water currents.
Wavy bedding : bedding characterized by undulatory bounding
surfaces.
Chapter 4, p. 169
Weathering : the destructive processes (physical disintegration and
chemical decomposition) which transform earthy and rock materials on
exposure to atmospheric agents, and prepare sediments for transportation.
Wettability : the ability of a liquid to form a coherent film on a surface
owing to the dominance of molecular attraction between the liquid
and the surface over the cohesive force of the liquid itself.
Winnowing : the selective sorting, or removal, of fine particles by
wind action, leaving the coarser grains behind.
Wireline : a general term for any flexible steel line or cable connecting
a surface winch to a tool assembly lowered in a well bore.
Wulff stereonet
Chapter 8, p. 320
Y
Young's modulus : a modulus of elasticity in tension or compression,
involving a change of length.
Chapter 8, p. 305
Z
Zenith : the point on the celestial sphere that is directly above the
observer and directly opposite to the nadir.
Zeolite : a generic term for a large group of hydrous aluminosilicates.
Zircon : the silicate of zirconium. A common accessory mineral.
References
BATES, R.L., & JACKSON, J.A. (1980). - Glossary of Geology.
Amer. Geol. Inst., Falls Church, Virginia.
SHERIFF, R.E. (1973). - Encyclopedic Dictionary of Exploration
Geophysics. Soc. Explor. Geophysicists, Tulsa, Oklahoma, U.S.A.
Society of Professional Well Log Analysts (1975). - Glossary of
terms and expressions used in well logging.
WHITTEN, D.G.A., & BROOKS, J.R.V. (1972). – Dictionary of
Geology. Penguin Books Ltd., Harmondworth, Middlesex, England.
This page has been reformatted by Knovel to provide easier navigation.
324
As my previous books got a very good reception by reviewers (G.B. Asquith from AAPG, H.M. Johnson from
Tulane University, A.V. Messineo from SPWLA, M. Verdier from IFP), I have been encouraged to work on a complete
revision and updating.
As a geologist by education, I started my career with a natural turn for geological applications of well logs.
I am more and more convinced of the fundamental link between logging data and geological parameters. I observed
ever since that many geologists do not realize sufficiently the true geological value of well logging data. Well logs
constitute a fantastic source of information on subsurface geology with a vertical resolution that can even reach 1
cm. Needless to say that logs are the eyes and tools of the subsurface geologist. It is the reason why "Well Logging
and Geology" is the first book of a series of three.
Moreover, too often petrophysicists and log analysts tend to restrict the applications of well logs to the determination of effective porosity, water saturation and permeability. The reservoir volume, and subsequently the hydrocarbon
volume in place, and the reservoir complexity are not correctly evaluated as the reservoir is not put back in its precise geological setting. The latter requires the precise determination of its original depositional environment, its diagenetic environment, the nature of its surrounding facies and finally a detailed and accurate description of the tectonic
structure that seismics cannot provide with sufficient resolution and accuracy. This can explain that the success rate
in field development reaches only 80% in favorable cases.
Unfortunately most of the time, production geologists do not have the logging data set which would help them
improving their reservoir description. Too often, oil company managers want to reduce costs on logging acquisition
as much as possible. The objective in itself is reasonable because reducing costs is always a good thing in itself and
in the bottom line companies want to make profit. The problem is that before selecting an asset reduction of 25% of
the logging cost will save only 2.5% of the well cost assuming the logging cost would represent statistically only
10% of the well cost! As a consequence, reduction of logging data, (most of the time natural gamma ray spectrometry and dipmeter or image data) will not allow the precise and accurate 3D description and evaluation of the reservoir. In addition, the error on volume estimation will be certainly much higher than the error on porosity measurement! In the end, the short term economy realized on logging can prove to be disastrous on a long term. An incomplete and imprecise knowledge of the local reservoir geology may result in much higher expenses due to wrong
location of production and injection wells! Do not consider the logs as a commodity but as a very valuable source of
information. And even if the logging measurements are affected by errors, remember, as pointed out by Philippe
Theys (Petrophysics, vol. 44, no 1, p.14, 2003),that "when you evaluate the magnitude of the errors that affect logging measurements, you find that the ones induced by the limitations of physics or technology are smaller than the
human errors". You would not rely on a diagnostic made by a doctor who would only examine your throat, take your
temperature and measure your blood pressure! Why would you trust a reservoir evaluation based on a reduced logging set?
Consequently, do not reduce too much the logging data set! But, at the same time exploit completely the logging
information. Do not leave the logs "sleeping" in their drawers! Analyze and interpret them in depth! Squeeze the
lemon!
The objective of this book is to demonstrate to geologists, petroleum engineers and log analysts what kind of
geological information can be extracted from logging data (a great deal actually!). For that purpose, I compare the
geological and well logging approaches, showing that one can extract from log data the information that generally
one asks to cores.
This "Well Logging and Geology" book will be completed by the book on "Well Logging - Data Acquisition and
Applications" and the book on "Well Logging - Reservoir Evaluation". These three books should demonstrate that
logs are a fundamental source of information for geologists and petrophysicists.
Oberto SERRA
Serralog 8 2003
i
TABLE OF CONTENTS
Foreword
Acknowledgements
Chapter 1
i
ii
GeneraIities
Introduction
Definitions
Rock properties
Rock physical properties
Rock chemical properties
Rock geological attributes
Sources of information
Outcropts
Information provided by subsurface data
Information provided by surface seismics
Information provided by drilling
Goal of well log interpretation
Interpretationmethodology
Log quality control
Tool calibrations
Repeat section
Depth match
Composite-Log
Segmentation in electrobeds
Well and Logging Types
Open holes
Vertical wells
Deviated wells
Horizontal to subhorizontal wells
Cased holes
Conclusion
Workstations
Formation Image Examiner
FracView fracture synergy log
DIAMAGE
GeoFrame Geology
Data preparation
Geology
Petrophysics
References and Bibliography
1
1
1
1
1
2
2
2
2
4
4
5
11
11
18
18
18
20
21
21
23
23
23
23
23
26
26
26
27
27
28
29
29
29
34
34
Chapter 2
Well Logging and Rock Composition
Introduction - Review of general geological concepts
The three major rock types
Igneous rocks
Sedimentary rocks
Detrital or clastic rocks
Chemical and biochemical rocks
Metamorphicrocks
Relative abundance of rocks
Rock classification
Rock composition
Elemental composition
Mineralogicalcomposition
Determination of the rock composition - Traditional method
Elemental and chemical composition
Mineralogicalcomposition
Determination of the rock composition - Well logging approach
Elemental composition
Natural gamma ray spectrometry
Induced gamma ray spectrometry
Determinationof the mineral composition
Exploitation of the elemental composition
Measurements involving interactions of photons
Measurements of the sound wave velocities
Cross-plot interpretation
Cross-plot definition
Cross-plot analysis
37
37
37
37
38
38
38
38
38
40
40
42
45
50
50
52
54
55
55
57
60
61
64
66
67
67
71
2003
Serralog i.:
Formations of very high resistivities
Identification of igneous rocks
Identification of evaporites
Identification of metamorphic rocks
Formations with variable resistivities
Identification of detrital rock type
Identification of biochemical rock type
Automatic lithology determination - LITHO-ROCKLASS
Introduction
Construction of the database
Logs used in LITHO-ROCKLASS
Quantitative interpretation
SERRALOG approach
Platform Express lithology display
Neural network approach
References and Bibliography
73
74
76
85
86
86
92
97
97
98
99
99
101
106
112
114
Chapter 3
Well Logging and Rock Texture
119
Introduction - Review of petrographic concepts
119
Definition
119
Textural components
120
Texture of detrital rocks
120
Grain size
120
Sorting
121
Grain shape
122
Packing
122
Orientation
123
Influence of grain properties
123
Influence of grain size
123
Influence of sorting
124
Influence of shape and roundness of grains
125
Influence of the orientation of grains
125
Influence of packing
125
Influence of the mineral composition of grains
126
Influence of other textural components
126
Texture of carbonate rocks
127
Determinationof textural parameters - Traditional approach
129
Determinationof textural parameters - Well logging approach
131
Detrital rock texture
133
Grain size
133
Sorting
137
Grain orientation
138
Arrangement or packing
139
Grain shape
140
Cement
140
Carbonate rock texture
141
Porosity
141
Porosity and pore size determination from NMR
145
Automatic extraction of textural informationfrom dipmeter 147
148
Quantitative analysis of image data - BorTex program
Influence of textural parameters on reservoir properties
154
References and Bibliography
155
Chapter 4
Well Logging and Rock Structure
Introduction - Review of general geological concepts
Definition
Importance of Sedimentary structures
Classification of sedimentary structures
Primary sedimentary structures
Stratum, bed, lamina description
Shape of sedimentation unit
Nature of bed boundaries
Groups of beds (“bedsets”)
Internal organization of beds
Secondary sedimentary structures
Disconformities
159
159
159
159
159
160
160
160
162
163
163
171
171
1
Intrusions- Dikes
Diapiric structures
Determination of rock structure - Traditional approach
Determination of rock structure - Well logging approach
Bed description
Bed shape and bedding planes
Bed thickness
Nature of bed boundaries
Internal organization of beds
Illustrationof structures by images
Bed boundaries
Internal organization
Graded bedding
Automatic analysis of dipmeter data - The SYNDIP program
The SYNDIP program
Descriptionof a SYNDIP display
Rules for interpretationof dipmeter and image data
Quantitative analysis of image data
Sedimentological applications of sedimentary structure detection
References and Bibliography
171
171
171
172
174
174
175
175
175
180
180
181
183
185
187
188
189
191
191
193
Chapter 5
Well Logging and Facies, Sequence and Environment
Introduction - Review of general geological concepts
Definitions
Importance of facies and sequence analysis
Traditional approach of facies analysis
Well logging approach
Historical
The Electrofaciesconcept
Electrofaciesanalysis from well logs
Manual identification of electrofacies
Automatic electrofacies identification - FACIOLOG
Automatic segmentation of logs
Clustering techniques
Final electrofacies determination
Principal component analysis
Translation of electrofacies to facies
Result presentation
The Electrosequence concept
Sequence analysis from well logs
Gradational transition - Elementary sequence
Abrupt contact
Sequences of electrofacies
Applications of facies and sequence analysis
Reconstitutionof the depositional environment
Interest of the reconstitution of the environment
Recognitionof the environments from logging data
Illustration of environments from logging data
Reconstruction of the geometry of the facies
Mapping of electrofacies
Constitution of a data base of electrofacies
Quantitative interpretation
Choice of a more judicious core sampling for analysis
References and Bibliography
197
197
197
200
201
201
203
206
207
208
211
212
214
216
217
221
223
223
224
225
226
227
227
228
231
231
232
234
235
236
236
236
236
Chapter 6
Well Logging and Diagenesis
Introduction - Review of geological concepts
Definition
The diagenetic environment
Factors affecting diagenesis
The fundamental processes
Physical processes
Chemical processes
Biochemicaland organic processes
Diagenetic changes
Rock components
Compaction
Cementation
Recrystallization
Transformation
Mineralogical replacement
Selective leaching
239
239
239
239
240
241
241
241
241
241
242
242
242
242
242
242
243
11
Authigenesis
Stylolitization
The diagenetic phases
Effects of diagenesis on the sediment properties
The need for diagenetic studies
Diagenetic Studies - Traditional methods
Well logging approach
Diagenesis in detrital sequences
Compaction
Pressure-controlledsolution
Cementation
Authigenesis
Nodules
Pyrite crystals
Epidiagenesis
Diagenesis in carbonate sequences
Crystallization
Neomorphism
Selective leaching
Stylolitization
Cherts and anhydritic nodules
Hard-ground
Diagenesis in volcaniclastic sequences
Diagenesis in shales
Diagenesis of organic material
References and Bibliography
~~
Chapter 7
Well Logging and Compaction
Introduction - Review of geological concepts
Definition
Compaction study
Sand compaction
Carbonate compaction
Shale compaction
Principal stages in shale compaction
Compaction mechanisms
The Hubbert 8 Rubey law
Compaction effects
Effects on shale porosity
Effects on shale density
Effects on chemistry of interstitial fluids
Effects on geothermal gradient
Effects on mineralogicaltransformation
The influence of time
Compaction of organic sediments (peats)
Compaction anomalies
Definition
Origin of compaction anomalies
Zones exhibiting undercompaction
Traditional geological approach
Well logging approach
Associated phenomena
Construction of the compaction profile
Manual plot
Construction of the profile at the well site
DENSON automatic program
Conduct of a study ..
Advantages of the DENSON program
Construction of the normal compaction trend
Detection of undercompacted zones by VSP
Study of sand compaction using well logs
Study of carbonate compaction using well logs
Study of shale compaction using well logs
Sonic velocity in shales
Resistivity of shales
Formation factor
Hydrogen index of shales
Capture cross-section of shales
Natural gamma ray spectroscopy log
Applications of compaction studies
Hydrodynamic applications
Detection of entry into an undercompactedformation
Evaluation of formation pressure
243
243
244
245
245
246
250
250
250
250
251
252
252
253
254
254
255
255
256
257
258
259
259
260
260
262
265
265
265
265
267
270
271
271
273
273
274
274
275
275
276
276
276
277
277
277
277
278
278
278
278
278
278
278
278
279
280
281
281
281
283
283
283
283
283
284
284
284
285
285
285
285
Serralog @ 2003
Evaluation of pressure gradients
Geological applications
Detection of unconformities
Reconstructionof maximum burial depths
Differential compaction
Correlations between wells
Sedimentologicalapplications
Applications to reservoir studies
Geophysical applications
Geochemical applications
Decompaction
Objectives
Procedures
Applications
References and Bibliography
287
289
289
289
290
290
290
292
292
293
293
293
294
294
294
Chapter 8
Well Logging and Tectonics
Introduction
Review of geological concepts
Definition
Concepts of stress
Mechanical behavior of rocks
Elastic behavior
Plastic behavior
Viscous behavior
Factors controlling rock behavior
Confining pressure
Temperature
Time
The actual behavior of rocks
Types of state of stress
Rock strength
General orientation of stresses
The results of stresses : strains
Continuous strain : fold
Tensile stress
Compressive stress
Descriptive elements of a fold
Fold shape
Fold type
Nomenclatureof folds
Discontinuousstrain : fractures and faults
Movements along a fault
Fault description
Fault classification
Determination of tectonic structures - Traditional approach
Determination of tectonic structures from seismic data
Determination of tectonic structures from gamma ray
Determination of tectonic structures from dipmeter or image data
Structural analysis of dip data
Statistical study of dip data
Dip and azimuth histograms
Polar plots
Azimuthal frequency plot
Density stereograms
Construction of density stereograms
Interpretationof these diagrams
Stereographic plotting techniques
Stereographic projection
Principle
Planes and lines on a stereogram
Polar plot
Cyclographic plot
Plotting a line
Construction of the WULFF stereonet
Construction of the SCHMIDT stereonet
Representationof structural elements
Plotting of dip data
Use of stereograms
Methods of Dip data interpretation
Visual arrow-plot analysis
“Quick-look technique : dip patterns
Determining structural dip
299
299
299
299
301
303
303
305
305
306
306
306
306
306
307
307
307
308
308
308
308
309
310
310
310
311
311
311
312
313
31 3
314
315
31 8
320
320
320
320
321
321
322
322
322
322
323
323
323
323
324
324
324
327
328
330
330
330
331
Serralog C 2003
Detecting faults
331
Detecting folds
332
Bengtson technique
332
Automatic and interactive programs
333
DipTrend program
333
Automatic structural dip determination
334
Analysis of borehole images
337
Structural dip versus regional or seismic dip
339
Reconstructionof the tectonic features of a well
339
Cross-section construction
340
Automatic cross-section construction
340
Hypotheses for cross-section construction
341
Similar folds versus parallel folds
341
Constructing cross-sections
342
Selecting a zone
342
Determiningthe tructural axis and translation Dlane 342
Determining the cross-section plane
344
Constructing the cross-section
345
Exploitation of dip data
345
Workstations
350
Other applications of stereographic projection
351
Find the line of intersection of two planes
351
Determine the angle between two planes
351
Determinationof true dip of a bed __.
351
Determination of the apparent dip ...
351
Dip removal
352
References and Bibliography
352
Chapter 9
Well Logging and Fractures
Introduction
Review of general concepts
Relation of fracture to stress orientation
Importance of fractures
Mechanical properties evaluation
Mechanical behavior of the reservoirs Stress computations
Fracture-Pressurecomputations
Dynamic elastic properties
Inherent strength computations
Detection of fractures on cores
Detection of fractures from well logs
Natural gamma radioactivity
Caliper
Temperature
Formation density
Photoelectric capture cross-section
Neutron hydrogen index
Sonic measurements
Effects on body waves
Arrival time
Attenuation of body waves
Normalized Differential Energies (NDE)
Amplitude spike analysis
Crisscross patterns
Effects on Stoneley wave
Resistivities
Dipmeter
Resistivity curves
Azimuth curve of pad 1
Caliper
Dips
Spontaneous potential
Ultrasonic image tools
Resistivity image tools
Fracture evaluation
Depths of fractured zones
Type of fracture
Fracture orientation
Fracture length
Fracture density
Fracture aperture
Fracture porosity from images
Fracture porosity from photoelectric index
355
355
356
356
358
358
358
358
359
359
360
360
360
360
361
362
363
363
363
363
364
364
365
365
366
367
368
369
370
371
372
372
372
373
374
377
377
377
378
378
378
379
379
380
III
Well Logging and Geology
IV
Fracture porosity from dual laterolog (DLL)
Determinationof fracture permeability
Fracture effects on reservoir interpretation
Tortuosity factor m
Saturation evaluation
Formation factor - Porosity relationship
Lithology determination
Core orientation
Hole slippage
Detection of borehole damage in deviated wells
Relationship between stress-induced features and
stress axis
Stress determination
Prediction of the fracture efficiency
RECAP
References and Bibliography
380
381
381
38 I
382
384
384
384
385
385
387
388
390
390
39 1
Chapter 10
Well Logging and Source Rock
Introduction
Traditional geological approach
Well logging approach
The Carbolog method
The uranium concentration
The A log R method
Methods based on cross-plots
Methods based on nuclear measurements
References and Bibliography
395
395
395
396
396
396
396
398
399
400
Chapter I 1
Well Logging and Sequence Stratigraphy
Introduction
Relative dating
Definition of parastratigraphic units
Stratigraphic phenomena
Break detection
Break definition
Break significance
Break classification
Breaks without major lithological change
Change in type of fluid
Textural change
Diagenesis
Tectonic accident
Unconformity
Breaks corresponding to major litholonical chanae
Diagenesis together witha mineraiogical change
Erosion
Unconformity
Ashes, lava flows or volcanic intrusions
Emersion
Transgression
Sequence stratigraphy from well logs
Nomenclature issues
Using log patterns
Simplistic use of log patterns
Surface hierarchy from well logs
Building a conceptual geological model
Methodology
Correlations from well logs
Principe of causality
Concepts of log correlation
Concept of similarity
Concept of rhythmicity
Concept of lateral variability
Intrinsic value of correlations. Concept of reliability
Stratigraphic value of log correlations
Search for time markers
Extension to seismic sections
References and Bibliography
403
403
403
404
404
404
404
404
404
404
405
405
405
405
405
406
406
407
407
408
408
408
409
409
412
413
415
415
415
416
418
418
418
419
419
419
420
420
421
422
-
Index Glossary
Index of referenced authors
Iv
423
425
Serralog 0 2003
ct
W
I-
n
a
I
0
1
GENERALITIES
Introduction
Humanity has constant need for resources provided by
the planet Earth: foods, water, raw materials, ores and
energy, either natural (sun, wind, hydraulic, tidal, geothermal, nuclear) or fossil (coal, petroleum, gas). An on-going
effort in research and exploration is therefore necessary
to discover, exploit and develop them. And, in this effort, it
is better to rely on research, technology and geology than
on hazard and luck!
Definitions
Geology is by definition "the study of the planet Earth.
It is concerned with the origin of the planet, the material
and morphology of the Earth, and its history and the processes that acted (and act) upon it to affect its historic and
present forms" (Bates & Jackson, 1980).
This study, even when it includes fundamental
research, is aimed at a better understanding of the
influences that have formed, transformed and modeled
our planet, of the laws ruling the formation of rocks, their
distribution, their transformation, their deformation, and of
the laws governing the accumulation of those raw materials and ores which are of economic interest. This scientific guidance is available to help us discover the low cost
minerals and energy resources that humanity needs so
much.
The geologist's work consists (Table 1-1):
- in observing and describing, as much as possible
completely and objectively, the rocks attributes and the
geological phenomena;
- in interpreting these observations by comparing them
with those made on recent series or phenomena (existing
models), those coming from the study of reconstructed
models or from laboratory experiments, observations
made on ancient formations which have been studied in
detail and are well understood (ancient models and application of theories of actualism or uniformitarianism as
developed by Hutton, 1788)
In the case of the study of sedimentary rocks, this
interpretation must lead to a reconstitution of the geographic and climatic frame and, consequently, to the understanding of conditions under which the sediments and
rocks are formed. From this reconstitution the geologist
will try:
- to predict those zones most favorable for the accu-
Serralog 0 2003
mulation of mineral resources,
- to specify the spacial extent of these resources,
- to evaluate their volume from the rock mineral resource content estimation.
Petrophysics is the "study of the physical properties of
reservoir rocks" (Bates & Jackson, 1980).
The petrophysicist's work, or log analyst's work
consists - in petroleum industry - in interpreting the physical properties of the reservoir rock crossed by a well, in
order to evaluate the volume of hydrocarbon both in place
and extractible inside a field. This evaluation is necessarily based on a reduced number of points of observation
and measurement (the drilled wells). In each well is measured the local reservoir thickness, and are evaluated its
local effective porosity, saturation, and permeability.
These local data must be extended from each well in
order to evaluate the final hydrocarbon volume.
Therefore, to be more accurate this volume evaluation
requires to put back the reservoir in its precise geological
setting. The latter requires the determination of its original
depositional environment, its diagenetic environment, the
nature of its surrounding facies and a detailed and accurate description of the tectonic structure. This study must
be achieved by a team composed of a geologist, a geophysicist, a petrophysicist and a reservoir engineer. This
team will base its study on the rock properties and the different sources of information.
Rock properties
The rock properties can be classified in three principal
groups.
Rock physical properties
The rock physical properties are numerous. They are
listed hereafter :
hardness, density, ductility, elasticity, compressibility,
magnetism, magnetic susceptibility, magnetic permeability, electrical conductivity, dielectric permittivity, natural
radioactivity, sound travel time, thermal conductivity,
power of gamma ray absorption, neutron interactions,
proton precession, porosity, capillarity, interfacial tension,
permeability, water saturation.
Most of these properties are measured in borehole by
well logging tools. The physical principles of these tools,
1
Table 1-1
Geologist’s work
-
1 Observation and description
of sedimentary bodies
Composltlon
elemental
mineralogical
Texture
-
-
-
2 Comparison with present
or well known old models
4 Prediction
Depositional
Environment
Palaeogeography
Location and
extent of
economical
objects
Determination
of mineral
resources
Volume
estimation
Lithology
Facies
paleocurrent
Thickness
Width
Length
Geometry
Relationship with
surrounding
sedimentary bodies
Evolution
-vertical
lateral
-
Diagenetic phenomena
Structural dip
Folds
Faults
Fractures
Stylolites
-
Sequences
Rhythms
Cycles
Geologic history
Diagenetic sequences
I
External structure
Stresses
their applications and, as well as, the interpretation of the
data they record, are explained by numerous papers and
books (cf. Volume 1).
Rock chemical properties
These properties are fundamentally linked to the minerals and materials which compose the rock. Rocks are
composed of minerals that are more or less stable.
The rock decomposition is due to weathering which is
a twofold process:
- mechanical weathering or rock fragmentation,
- chemical weathering by decay of unstable minerals,
each helping and reinforcing the other, the smaller the
pieces, the greater the surface area available for chemical attack, and the faster the pieces decay.
Some minerals are also easily dissolved under certain
conditions but can be regenerated as such by reverse
conditions: for instance salt, calcite.. .: the less soluble,
the more resistant. Quartz wins hands down.
Stability, as it affects resistance to weathering, is also
a relative property.
The mineral stability follows the reverse of the
Bowen’s reaction series (Fig. 1-1 next page). Fe-oxides
are the most stable. They are successively followed by Aloxides, Quartz, Clay minerals, Muscovite, K-feldspar
(orthoclase), Biotite, Na-feldspar (albite), Amphibole,
Pyroxene, Ca-feldspar (anorthite), and Olivine which is
2
-
3 Interpretation
~
Tectonic style
Structural traps
Location,
nature and
distribution
of permeability
barriers or of
permeability
pathways
the least stable.
Rock geological attributes
The geological attributes of sedimentary rocks include:
- the facies of each depositional unit. This facies is
defined by :
- the composition,
- the texture,
- the sedimentary features or structures,
- the colour,
- the fossils
of the depositional unit;
- the facies succession or sequence,
- the transformations undergone since the deposition
of the sediment, in other terms the observed diagenetic
phenomena,
- the deformations undergone under stress action.
Sources of information
As previously indicated the team in charge of the formation analysis will use several sources of information to
achieve the goals of their studies. These sources are
reviewed hereafter.
Outcrops (Cliffs, quarries, trenches, ditches, tunnels,
mines)
Serralog 0 2003
Generalities
I Chapter 1 I
3
ROCK
TYPE
MB)I
1II)c.
MOST
UNSTABLE
HIGH
FIRST TO
CRYSTALLIZE
[XMC
0
ANORTHrn
0
fca kldrp.f)
\
/
I
HIGH
LOW
R1
DUNITE
PERlDOTlTE
WEAT
PVROXENITE
GABBRO
n.s.0
DlORlTE
>
t
.o
Andrnsn.
B
OUARTZ DlORlTE
0.ClI.
ORANODIORITE
I I
LOU
TEMP.
LASTTO
CIYSTALLIZE
4
1
GRAWITE
UlsCbVrrE
Rhylrolll.
0
I
t
WRTZ
@
LOW
MOST
STlBLE
WEIS:
LOW
NtW
1: single tetrahedron. 2: single chain. 3: double chain.4: sheet. 5: 3-D lattice
Figure 1-1 - The Bowen's reaction series and its interpretation both for rock type formation and for weathering intensity and results
(adapted from Leet et al., 1982).
In certain types of research (metallic ores), natural or
man-made outcrops are still the essential source of geological information (Fig. 1-2). In others (petroleum, coal,
geothermal resources) outcrop information has been progressively replaced by drilling data or completed by surface geophysics (gravimetry, seismic surveying, magnetism, electric and telluric measurements), or borehole geophysics, the latter including Wireline Logging (WL) or
Logging While Drilling (LWD).
On land, the basin exploration starts generally by a
sedimentologic, stratigraphic and tectonic study of outcrops surrounding the basin in order to determine its
hydrocarbon potential, it means the co-existence of reservoir-rocks, source-rocks, seal-rocks and traps. The analysis of rock samples allows the verification of this co-existence. The samples are collected with the required size.
They are easily located in their environment. It is generally easy to laterally follow the beds from which they are
extracted, and determine their geometry and the one of
their surrounding beds. Aerial and satellite photographs
can help in this work. The samples can be replaced in the
vertical sequence of deposition. At the end, it is possible
to practice all the analyses necessary to describe, evaluate and date them. It is also quite easy to come back, resample and re-evaluate them.
In sea, the basin evaluation requires fundamentally
geophysical techniques completed by a few drillings.
In any case, more generally, the geological knowledge
of a basin, on land or offshore, would not be as complete
and precise if it did not included information provided by
specialties such as surface geophysics, geochemistry,
petrology, mineralogy, paleontology, sedimentology, or
Serralog 0 2003
I
Figure 1-2 - Observation and description of outcrop
(Mt Saint-Odile, Vosges, France, photo 0. Serra).
3
petrophysical measurements realized in a well, such as
well logs, borehole seismics and tests. This is because in
modern exploration, drilling and geophysical techniques
are more often used, not only in the petroleum industry,
but also in subsurface storage, or to discover coal, uranium or even some metallic ores, and in geothermal
resources. With deeper targets, outcrops, as a source of
information, are less frequently used because extrapolations established from them are less reliable. Moreover,
the geological complexity of targets increases (stratigraphic traps, fluid permeability barriers, size and depth of
structures, ...).
As a result, at the present time, most knowledge concerning geological sedimentary basins (especially deep
basins) comes from drilling and geophysics. As mentioned by de Golyer (1921), Lahee (1932) and after them
by llling (1946) : "if geology has contributed greatly to the
growth of the oil industry the debt is not one-sided one.
Geology owes a great deal to the oil industry in the expansion of its knowledge and the increased efficiency of its
methods".
Information provided by subsurface data
Information provided bv surface seismics
This branch of Earth's science is fundamental for many
reasons. It is the only one that allows continuous study of
formations in subsurface. It completes our perception of
these formations on the outcrops. Two and, more often,
three dimensional pictures of subsurface can be obtained
by today's surface geophysical techniques (Fig. 1-3).
They are extremely important tools for the exploration of
subsurface, since it gives direct information, not only on
the shape and arrangement of beds (Fig. I-4), but also on
their nature, their petrophysical properties and even
sometimes their fluid content (seismofacies, "bright spots"
Fig. 1-5).
Unfortunately, seismic vertical resolution is close to 20
meters (Islam, 2000), 10 meters at best at low depth. In
many cases, at the reservoir scale, it is not sufficient to
describe it correctly. In addition, the hypotheses drawn by
the interpretation of this information must be verified by
Figure 1-3 - Example of geological interpretation of surface seismic data in terms of depositional environment and tectonic setting
(courtesy of GECO).
4
Serralog 0 2003
Generalities
drilling. The translation of surface geophysical data into a
geological interpretation will be considerably easier and
more reliable if it is supported by well log measurements.
In other words, well log data provide the link between geophysics and geology (Fig. 1-6). The former has thus to be
correlated with the latter. Well logs are the only means for
providing an accurate transfer of time data to depth data.
I Chapter 1 I
5
a
b
Figure 1-4 - From this time slice representation of the seismic data, it is
easy to recognize a meandering channel system.
C
Figure 1-5 - On this seismic cross-section the gadoil and oil/water
contact are detected thanks to the changes in wave attributes
(amplitude, phase).
They allow for the transfer of amplitude and signal frequency data to sedimentological or economic data
(facies, porosity, fluid content, Fig. 1-7).
d
information provided bv drillinq
Two kinds of data are obtained during drilling
those linked to rock samples (full diameter cores,
sidewall cores, cuttings), and to fluid samples;
- those provided by physical measurements made in
drilled holes essentially by logging tools.
-
Core information
Standard cores, due to their size and if they are continuous, Constitute a sampling Of good quality similar to
the one collected on outcrops. They will allow all types of
analyses, providing information at microscopic scale
electron microscope
is
(grain and pore size) if scanning
used (Fig. 1-8).
Serralog 0 2003
Figure 1-6- a) Surface seismic section (minimum phase). 6) Acoustic
impedance vs porosity. c) Correlation between logs, formation impedances, stratigraphy and VSP (minimum phase). d) Acoustic impedanve vs porosity, the colour coding indicating the depth of each point,
shallow depth in blue/green, deeper depth as red/yellow
(adapted from Poster, 1988).
5
Figure 1-7a - The well log impedance model, extracted wavelet, and
cross-plot information have been combined to indicate the probable
effect on the surface seismic section of a porosity increase.
1500
Figure 1-7b - The seismic events now correlate with the main formation
boundaries with a higher resolution than the original data of Fig. 1-5a
(adapted from Poster, 1988).
However, the bed geometry will not be determined due
to the lack of dip and azimuth measurements of bed and
borehole, their size and the lack of lateral continuity.
Moreover, cores may also have drawbacks such as
partial recovery or none at all, sample damages linked to
release of the stress after their extraction, and the types
of processing they support during all the steps of analysis
which may modify significantly their petrophysical properties (cleaning, drying, ...).
Figure 1-8 - SEM photograph showing quartz grains and “books”of
kaolinite. It gives also an idea of the pore size, pore geometry and clay
distribution.
In some cases, there is considerable uncertainty over
the depth of a given cuttings sampling and furthermore it
can be difficult to restore the constituents and thickness of
the lithological column from cuttings alone. This is due to
mud swirling, caving, the loss of some constituents (siltysize grains, salts) during washing or by total lost-circulation events. The reduced dimension of this type of rock
samples does not generally allow a full analysis, and
makes an observation incomplete.
Therefore, geologists and, in particular, subsurface
geologists are often short of, or deprived of representative or good quality samples. Sidewall cores can partly
compensate for this shortcoming but, because of their
small size, observations or measurements may still be
inaccurate compared to those carried out on larger samples.
Consequently:
- Geologists and petrophysicists are unable to respond
to the fundamental question that one asks in hydrocarbon
exploration: does it exist economic hydrocarbon accumulations in a well?
- They cannot make certain studies yet necessary for
the basin exploration or field development pursue.
Well loaaina information
Cuttina information
Unfortunately, for economic and technical reasons,
coring can be a rare operation, particularly under certain
drilling conditions or in certain types of formations.
Therefore, often the only rock sample available consists
of the cuttings obtained during drilling, or sidewall cores
cut with the help of a rotary or of a bullet core barrel which
runs with a wireline.
6
Fortunately, borehole geophysical techniques allow
geologists to fill the gap represented by the lack of representative rock samples. Between those techniques well
logging occupies a special and very important place. This
is due to the quality and diversity of tools of acquisition
and methods of interpretation. Even, thanks to technical
progresses it is now possible to obtain physical measurements during drilling. The vertical resolution of these
measurements ranges from few meters (Vertical Seismic
Serralog 0 2003
Generalities
Profile) to about one centimeter (resistivity imaging
tool).
Well logs are of special interest in that:
-they provide the only source of data to give accurate
information on the depth and the apparent, and even real,
thickness of beds if a dipmeter or image data have been
recorded, thanks to the borehole deviation and dip of bed
boundaries they provide.
- They give a nearly continuous analysis of the formations (one sampling every 15 cm for traditional open hole
logs; this frequency can be increased to one sample
every 2.5 cm if required, and one sampling every 5 mm or
even 2.5 mm for dipmeter and image tools). In contrast,
information extracted from cores is more discontinuous,
and often scattered in depth. Even in the case of continuous coring, all analyses are not systematically made on
each plug selected on the cores (a plug each foot [30
cml).
-They generally analyze a volume of rock that is oftengreater than the one represented by a core or plug, and
consequently than a cutting. Consequently, they are more
representative of the mean properties of the rock, especially in heterogeneous rocks. But, at the same moment
they provide a very detail description of the formations if
images are recorded (resistive objects of 1 cm size and
conductive features as thin as 1 micron [lpm] are detected).
- They give a vision of the formations, at the scale of
meter for standard logs and centimeter for images, indeed
particular but nearly continuous and objective.
The information logs provide is
. quantitative and, consequently, it allows us to think
about geological objects represented by well logging
measurements by using the full the computer's capacity to
process the information;
. precise, even if, sometimes, errors are present as in
any measurement;
- objective and repetitive;
- permanent; whereas the cores are destroyed for analysis, preventing any further study, log data can be reinterpreted with new ideas, new techniques, or new parameters;
- obtained rapidly, even on the well site;
- economic. Coring and core analysis and description
are expensive and time consuming, and the desired information is obtained only several weeks later. Do not forget
that the wireline logging cost represents generally between 5 and 10% of the drilling cost and that logs provide
approximately 90% of the information extracted from a
well (Fig. 1-9). A reduce logging set represents certainly a
short-term economy, but can generate much more development cost on a long-term basis!
- The measurements made with logging tools are
strongly dependent on geological parameters (Table 1-2).
Consequently, the information they provide is of the
utmost interest for geologists.
In conclusion, well logs contain information that can
interest geologists and consequently can be used. Well
serralog Q 2003
1 Chapter 1 I
7
logs "photograph" the drilled formations and record real
facts that cannot be modified (Fig. 1-10). "They provide a
spectral picture, albeit particular and incomplete, which is
practically continuous and always permanent, objective
and quantified. It is easily understood that the "picture" will
be clearer when the number and the diversity of the log
measurements are greatest. One can say that logging
tools are to the subsurface rock description what the
eyes, geological instruments and products (hammer,
magnifying glass, compass, chloridric acid, alizarin...) are
to the surface outcrop. Thus, the logs can be considered
as the "signature" of the rocks since they depend on their
physical properties" (Serra, 1986). The physical characteristics of drilled formations that logging tools measure
result in fact on the one hand from the physical, chemical
and biological conditions - hence also geographical and
climatic - that existed in the environment of deposition of
the sediment and which determine the original facies and
characterize the environment, and, on the other hand,
from the evolution that these sediments have undergone
during their geological history to become sedimentary
rocks.
I
Figure 1-9 - Logging represents in average 10% of the well cost but if
provides approximately 90% of the information linked to the wellbore.
This is the reason why log data are so important. It is
no more possible to conceive any geological synthesis
and reservoir evaluation without the exploitation of well
log data. Doing so would correspond practically to base
our knowledge of the basins and reservoirs on the one
hand on outcrops, consequently on a small surface and
volume fraction of the formations composing them, on
surface geophysics but without calibration and scaling on
well logs, and, on the other hand, on cuttings, exceptionally on cores.
In addition, log interpretation should not be limited to
reservoir evaluation as too often practiced. The reservoir
must be put back in its geological setting (sedimentological, diagenetic and tectonic) to better evaluate its extent
and its volume. This is also the reason why geologists
must achieve well log analysis. You wouldn't think of a
radiologist analyzing medical images without knowing the
anatomy! Why would you entrust your log-interpretationto
a log-analyst who wouldn't know about geology? Only
geologists have sufficient geological knowledge to be able
to extract the required geological information. But, of cour7
se, they must also have a good understanding of the PhYsical parameters recorded by logging tools and the link
existing between the logging parameters and the geological attributes (Table 1-2).
(WL). In addition, logging techniques allow the sampling
of cores and fluids.
Table 1-2
Relative influence of geological attributes on well logging measurements
(adapted from Serra et Abbott, 1980).
Type of measurement
“Log interpretation consists, thus, of a data “translation” from log parameters to geological data. To do this we
need a good “dictionary” or an “interpreter” who knows
the two “languages“ welr (Serra, 1986).
Well logging data can be acquired either during the
drilling itself or after a period of drilling by tools gone down
at the end of a cable. The first one is called LWD (logging
while drilling), the second is known as Wireline Logging
8
Well seismic information
The Vertical Seismic Profile (VSP) obtained in a well
(Fig. 1-11) provides the best vision and exact depth of
each reflector, its signature and the effect over the underlying reflectors. It enables the transformation to be made
from reflectors to beds, providing a precise measurement
to go from depth to time and vice versa.
Senalog 0 2003
Generalities
I
Chaoter 1
I
9
Foreset laminations
Transport current
direction : N 200'
-
Drilling induced fracture
Foreset laminations
Transport current
direction : N 40'
-
Drilling induced
fractures
.
Foreset laminations
Transportcurrent
direction : N 120
-
Foreset laminations
Transport current
direction: N MI'
Ebsbnal contest
-
Figure 1-1Oa - Subdivision of a 2.2 meters bed in its depositional units from the borehole image analysis.
GR (API)
CAL (in)
SP (mV)
0
4
-80
150
14
20
2
10 ohm-m
MSFL
LLD
LLS
100
0
Pe (b/e) 10
45
NPHl (%)
1,95
Pb (g/cm3)
140
At (ps/ft)
-15
2,95
40
~~
Figure 1-lob - The 2.2 meters, corresponding to the image of Fig. 1-1 Oa, are indicated by a red strip on the standard logs.
Semlog 0 2003
9
Next Page
10
Well logging and Geology
seismic section and seismogram (Fig. 1-12).
0.9
I
.+
Figure 1-12 - Display of seismic section, seismogram and logs with
reflection coefficient and conversion of two way time in depth
(courtesy of Schlumberger).
Figure 1 - l l a - Example of VSP (courtesy of Schlumberger).
For example, it is possible to "see" the beds which
underlie formations with strong reflectance (anhydrite,
halite, compact dolomite or limestone), or with strong attenuation (undercompacted shales), or formations below
the bottom well depth. It also makes it possible to remove
the multiples that often complicate the interpretation of
seismic sections. VSP can also be used to analyze the
rock properties through the study of seismic waves (Direct
Signal Analysis method).
Wd1
Figure 1 - l l b - The VSP recorded in the well is inserted into the seismic
section allowing a better interpretation and the precise location of the
faults (courtesy of Schlumberger).
A graphic display of logs versus time instead of depth
can also be easily obtained and reproduced alongside
10
In fact, measurements of the density and the traveltime of acoustic sound derived from boreholes with well
logging tools, make it possible, after corrections, to determine the accurate acoustic impedance and reflection
coefficient for each boundary in formations. A reflectivity
log made from acoustic and density data provides a basic
document for the establishment of a theoretical seismic
section through a program, which, when correlated with
real seismic sections, permits its conversion into depth
and makes a composite log possible (Fig. 1-12).
Even if well logs constitute a fundamental source of
geological and petrophysical data, it is however important
to never under-estimate the problems that can occur
during their interpretation. These problems are linked to
the following conditions.
- Logging tools have different vertical resolution and
depth of investigation. Consequently, combination of their
data is not always valid as the different log responses do
not correspond necessarily to identical volume of rock.
- A perfect depth-matching of log data is sometimes
difficult to achieve if logging speed variations are not
recorded by an accelerometer tool added to the logging
string.
- Due to their lack of resolution, some of the recorded
data, especially those corresponding to the transition from
one bed to another, have no geological meaning. This is
especially crucial in thin beds. To avoid any misinterpretation the data corresponding to bed transition should be eliminated.
- The geological "bed" notion must be adapted to the
type of logging data and replace by the "electrobed"
concept. An electrobed is "the smallest unit that can be
recognized on logs from the surrounding units due to
significant changes in logging parameters". The minimal
thickness of an electrobed varies as a function of the logging type of measurement (Fig. 1-13). It depends on the
Serralog 0 2003
Previous Page
Chapter 1
intrinsic and qualitative vertical resolution. In order to
detect the thinnest units of deposition, imaging tools are
fundamental as they have the highest vertical resolution:
approximately 1 cm.
Sonic BHC
(ilyni
Figure 1-13 - Example of segmentation of log responses in electrobeds.
- What is generally called a bed may often be subdivided into several depositional units as illustrated by Figure
1-14.
Lithology
DePositional
units
Geological objects
11
- Most of the equations used for the quantitative evaluation of mineralogical and petrophysical properties are
em piricaI.
In fact, the properties of a bed essentially depend on
the own properties of each unit that composes it. Each
unit shows variable thickness and extent as a function of
its depositional environment. In other terms, a formation is
composed of a vertical and lateral succession of two
types of geological objects: volumes (beds, strata, layers)
delimited by surfaces. Too often, standard logs, due to
their lack of resolution, reflect imperfectly the average
petrophysical properties of a bed, not the real ones of
each depositional unit that composes it. Due to their very
high vertical resolution, images allow the detection of
each unit. It is also fundamental to quantify the information they provide (see further). An electrobed recognized
on a log with lower vertical resolution may in fact be composed of several smaller units, each of them having their
own petrophysical properties. The interpretation should
be based on this new concept.
As a conclusion we must keep in mind these problems
and proceed first to a log quality control. We must detect
the deficiencies when they occur, be able to make the
appropriate corrections, when they are possible, eventually reject the unreliable data, in order to correctly interpret the well logging information. But, in any case, extract
all the information included in logging data. Squeeze the
lemon to its ultimate drop! Accumulate real facts of observation in order to improve the accuracy and reliability of
your interpretation. You would not thrust medecine doctors who will base their diagnosis only on measurement of
your fever, your blood-pressure and examination of your
throat and ears. Why would you believe in an interpretation based on a reduced set of data?
Goals of well log interpretation
Figure 1-14 - The geological reality : generally several units compose
a bed and two types of geological objects : volume and surface must
be taken into account and their properties evaluated
(adapted from Blatt et al., 1980).
- Certain logging measurements are more sensitive to
borehole environmental effects (mud type, caving, temperature...) than others.
- A deep invasion by the mud filtrate may have a strong
effect on certain measurements and consequently obscure the original fluid in place.
- Very often the measurements are not corrected for
the influence linked to the apparent angle between the
measuring tool and the formation. This is very important in
deviated or horizontal wells.
Serralog 0 2003
Extracting geological information, and petrophysical as
well, related to bed thickness, lithology, composition, texture, structure (internal and external), diagenesis, stratigraphy and reservoir characteristics are the goals of subsurface data analysis. Characterizing these formation properties will assist geologists in determining the facies, the
genetic sequence, the depositional environment and,
consequently, the volume of the reservoir beds.
Geologists can also detect diagenetic effects and provide
a very detailed and accurate sequence stratigraphy analysis from those data.
To achieve these goals the analyst must use all the
available information provided by surface and subsurface
geological and geophysical data.
Interpretation methodology
An interpretation methodology is proposed in Table 13 (see next page). It integrates all available data in two
11
Well logging and Geology
Table 1-3
Interpretation methodology
(modified from Serra et a/., 1993).
Third step
steps for a single well. The third step corresponds to the
study of several wells of a field or a basin, each well
having been previously interpreted as indicated in the
two-step procedure.
Fundamentally, in a first step the subsurface geologist
extracts from well logs recorded in each well of a basin or
field the information he generally obtains, on surface, from
outcrop observation and rock analysis. For that purpose
he must apply to the well log analysis the same approach
12
he uses to observe and describe a core or an outcrop. He
must consider any well log object as a geological object.
His approach, similar to the one of the surface geologist, summarized in Table 1-1, consists of a first analytical
phase during which he accumulates the maximum data
and information coming from an objective, detailed, meticulous and complete observation and description of all
available well log data relative to each electrobed.
Serralog 0 2003
Generalities
He interprets this set of data in order to extract information relative to:
- lithology: elemental and mineralogical composition;
- texture: grain size, sorting, packing, cement type and
percentage, porosity, pore size, pore type and pore distribution;
- internal structure: sedimentary features, direction of
transport current and progradation, nature of bed boundaries, biological activity, energy in the depositional environment, figures of slump, slide, or avalanche;
- external structure or geometry: volume of each unit of
deposition: thickness, length, width, shape of limiting surfaces, dip and azimuth; determination of deformations
undergone since the sediment deposition: structural dip,
folds, faults, fractures; direction of stresses;
- facies and sequence: spatio-temporal relationship
between volumes (electrofacies) defining sequences of
deposition;
- depositional environment: the succession of facies
and their geometry allows the determination of the depositional environment and through that of the extent of
each unit of deposition;
- stratigraphy: relative datation, strata succession;
detection of reverse or recumbent series, erosion, emersion, transgression, unconformities;
- diagenesis: transformations undergone by sediments, dolomitization, compaction, cementation, dissolution; organic matter maturation;
- petrophysics: porosity (percentage, type, pore size,
pore distribution and geometry); permeability; anisotropy;
cation exchange capacity; irreducible water saturation,
wettability;
- economics: nature of mineral resources; extension;
volume; depth; proximity of consumption zones.
In a second phase, the geologist collects all the data of
each well he analyzed and with the help of the other members of the team he synthesizes the extracted observa-
Fundamenta I pre1iminary remark
Being conscious or not - it should be better to be
conscious! - logging tool measurements depend strongly
on the geological parameters which have controlled the
rock properties since its deposition.
Consequenfly:
- any log interpretation is by nature a geological interpretation;
- logs constitute a fundamental source of geological
and petrophysical information in subsurface.
The goals of this book are:
- to convince the reader that many types of geological
information can be extracted from logging data;
- to show how this information can be obtained from
the logging measurements.
In order to extract this information from logs one must
apply to well logs the same approach which a geologist
applies to outcrops and core examination.
Serralog 8 2003
I Chapter 1
13
tions and descriptions. Applying a continuous change of
scale of observation by taking into account the seismics,
logs, core and test data (Fig. 1-15), together they interpret them by reference to well known modern or ancient
models (experience) and to laws that control deposition,
transformation and deformation of sediments. In sedimentary rocks, this consists in the reconstruction of the
paleogeography, the diagenetic history and the tectonic
and stratigraphic setting of the formation and reservoir.
It is obvious that the accuracy of interpretation will be
improved by integrating the information provided by all
available data. Indeed, accumulation of observations and
real facts of different natures allows the interpreter to
avoid erroneous conclusions. Any interpretation that
would be based on a reduced set of information would
suffer of a lack of reliability. In the early stage of a field or
basin study, integration of log and image data calibrated
on core analysis, results of processing of well and surface seismic surveys, guided by information provided by
logs and dips, and test measurements are a must to ensure an accurate and reliable evaluation of the economic
potential of formations.
The accuracy will be, as well, improved by eliminating
all the artifacts generated by the lack of resolution of standard tools. This is the reason why the electrobed concept
is important and why the use of images is fundamental.
Workstations simplify interpretation as they allow the
user to easily process raw data, manipulate a variety of
data types, and display the results of processing and
interpretation at any scale (see further Fig. 1-16).
Interpretation of images is a multi-step process of observation, description, interpretation, and implication or
exploitation (Serra, 1989; Bourke et a/., 1989). The following set of guidelines for image interpretation is recommended:
For observation and description of sedimentary features, use images that have the same horizontal and vertical scales.
Interpretation of enhanced images should be done in
conjunction with static normalized images to check the
resistivity range of the observation. A calibrated microresistivity curve, displayed on the side, may replace the static images.
All images should be interpreted in context of other
open-hole logs, preferably recorded at a high sampling
rate, to relate events seen on the images to petrophysical
properties and vice versa.
The geological interpretation must use all the available data and especially, as much as possible, images or
dipmeter data. Due to their very high vertical resolution
(Fig. 1-17 and Table 1-4) images are fundamental to
determine the internal organization of any depositional
unit, especially reservoir, to better understand its complexity and/or its dynamic behavior. It is obvious that a
reduce set of logging data will not allow a reliable description of the formations.
13
14
Well logging and Geology
Figure 1-15 - From seismic data to thin section analysis, through standard logs and images. This change of scale allows a much better link between the different sources of information and, consequently,a better understanding of the attributes and properties of each depositional unit recognized with techniques of higher resolution
(courtesy of Baker Hughes).
Figurel-16 - Example of data reproduced on the screen of a workstation. One can see quantitative interpretation results, borehole wall images
indicating a highly laminated reservoic dip data and cross-plot of density vs neutron measurements. This plot indicates as well the laminated sandshale sequence of this reservoir. The shale percentage computed from the standard logs is more important than the one shown by the images.
Gas effect can also be observed (courtesy of Schlumberger).
14
Serralog 0 2003
Generalities
D
m
c
Y
0
E
I-
Figure 1-1 7 - Vertical resolution of logging tools.
Only bed boundaries, laminations and fractures that
are not oblique to the borehole axis will be common to
both images and core photographs (Fig. 1-18). In heterogeneous formations, features present in the core will not
occur on the borehole wall at the same depth and perfect
correspondence between core and image is unlikely.
Any geological interpretation of well logs must extract
the fundamental geological attributes of each depositional
unit as it is done from outcrops or cores (see next page,
Fig. 1-19).
These geological attributes are:
- Thickness.
- Elemental and mineralogical composition.
- Texture.
- Internal structure.
- Facies.
- Sequence or facies succession.
- Depositional environment determination.
- Diagenetic effects undergone by the sediment since
its deposition.
- Deformations undergone since the sediment deposition.
- Stratigraphic history.
Although the details of images are crucial for formation
description, analysis of the data can be very time consuming. A multi-step, systematic approach is recommended
to achieve optimum results in the least amount of time
(Harker eta/., 1988).
1. Display the images at a compressed scale (e.g.,
1:lo0 or 150) together with relevant companion logs (Fig.
1-16). This display is used first to perform depth matching
and correlate images with open-hole logs and arrow plots.
2. Zone the major depositional units using this composite log. A lithologic column obtained by processing openhole data may be preferred to raw data, because it directSerralog 8 2003
I Chapter 1 I
15
ly indicates the main lithofacies.
3. Identify major sedimentary discontinuities and tectonic features using an arrow plot. Use seismic information
and local knowledge to narrow the possibilities.
4. Define the most probable depositional environments
from all available information - lithofacies, thickness,
genetic sequences, nature of bed boundaries, etc.
5. Display the images at 1 5 or 1:lO both horizontal
and vertical scales. At these scales, image details can be
interpreted in terms of texture, sedimentary features, etc.,
within the interpreted environmental context. Local
enhancement may be necessary to highlight faint features.
6. If cores are available, describe them in detail.
Comparing sedimentary features between cores and images permits assessment of lateral variation of thin beds
and other features at the scale of the borehole. Images
also provide better visualization of minor faults and
slumps that are ambiguous or absent on dipmeter logs or
may be too large to be interpreted from core data alone.
7. Summarize all the information provided by the images and focus on the nature of bed boundaries, bedding
types, dips, fractures, stylolites and pore types and distributions.
U
Figure 1-18 - Depth-match of core and FMS images with the
DlAMAGE program. Observe the traces of fracture and
laminations. Their vertical extent is not the same as soon as their
apparent angle is not nil. This is due to the higher diameter of the
borehole than the core diameter
(courtesy of Schlumberger and ELF).
15
16
Well logging and Geology
GRAIN
SIZE
Figure 1-19 - Typical geological analysis that must be achieved from well logging data as the latter provide practically all the information necessaty
to make this geological description as done by Van Wagoner et al., 1990.
16
Serralog 0 2003
Generalities
I Chapter 1 I
17
-
Table 1-4
Vertical resolution of the principal logging tools compared to the geological objects.
,E!ec!W!.im?P. * . -. I
.Dipmeters
. .. . .. . .
Ultrasonic images
e
icrolog
Electromagnetic (GHz) W
Micro SFL
Litho-density - - - t---+,
Radioactivity- - - - Acoustic *-\
Azimuthal resistivitye-1
Laterolog I-*
Spherical log +e
Neutron - - - +e
Spectrometry
Array induction
Dual induction- - sp
t
I
1
1
1
1
I
0.05
0.0 1
o~ooo6.001
I
I
1
I
I
I
-
I
0.05
I
0.5
0. I
VSP I*
I
I
1
I
1
5
I
.+,
I
10
Scale (m)
c'
Y Crystals
8
3
cn 0
- - -Rock fragments - - - - - - - - - -
rd)
8
's"
Silts
w
I Sands I 1
1
Pebbles Cobbles1 Boulders
$
m
Shell debis _-------_-Nodules - - - - - - - - - - Sand lenses - - - - - - - - Crystals - - - - .r(
CI
09
....1
2
m
8
.rl
w
0
m
Fractures
od)*
g
d)
- -
-Small sedimentary features - - - - - - -Large sedimentary features
sedimentary bodies
Faults
. . . . Higher sensitivity than resolution
the purpose of interpretation.
Grade 1 - A unique interpretation of the image is
possible independent of other data. Trained analysts can
interpret the images from their prior experience with confiSerralog (i3 2003
--
- - - -Higher resolution techniques
contoured features with well-defined shapes. Other logs
may offer information to aid the interpretation.
Grade 2 - Ambiguous. A conceptual model or other
log data are necessary to interpret the images correctly
and completely.
17
18
Well logging and Geology
Grade 3 - Core calibration needed. Images are used
to complement core interpretation. After calibration to core
data, the images may be used to evaluate zones where
cores have not been recovered.
the analysis of borehole images and nuclear magnetic
resonance data for texture. As a reliable knowledge of the
lithology is fundamental, a particular care must be applied
to log interpretation to obtain this information.
To be used in any interpretation the information provided by image data needs sometimes to be quantified and
classified based on their origin (see Fig. 1-20 next page).
This is achieved, for instance, through the TexScan or
BorScan programs of Schlumberger. Its logic will be
explained later (Chapter 3).
The log study continues by determination of the sedimentary features, from borehole images or dipmeter data,
which will inform us about the facies and the depositional
environment through the analysis of the facies succession
or sequence.
Geological interpretation of well logs requires the
same type of work than generally done when studying
outcrops or cores (cf. Table 1-1). As logs strongly depend
on geological parameters, detailed and precise observation and description of events seen on logs and images,
comparison with previous studies, interpretation based on
experience, and prediction must be also the successive
steps of geologists working on subsurface data.
During their studies, workstations will help considerably the geologists in 'their geological interpretation as all
the data can be shown and studied at the same time as
they can be reproduced on the screen at the required
scale (Fig. 1-16).
As previously mentioned, and this must be well
understood, any log interpretation is by nature a geological interpretation as log responses depend strongly on
geological parameters. We insist again, it is important to
understand this fact. The goal of the geological interpretation of well logs is to convert the logging data into geological attributes. This requires a good knowledge of the
link between the log measurements and the geological
parameters. This link is shown in Table 1-2 (see above)
and will be developed and illustrated in the next chapters.
Thus, once more, extracting information related to
lithology, composition, texture, structure (sedimentary and
tectonic), thickness, geometry, diagenetic effects and
reservoir characteristics of each depositional unit crossed
by a well is the main goal of geological interpretation of
well logs, images and dip data.
Characterizing these formation properties assists geologists in the description of the facies, genetic sequence,
depositional environment, types of traps and stratigraphy.
In addition, reservoir engineers can determine the presence of permeability pathways or barriers and reservoir
flow properties, allowing them to better determine the
scheme of the field development and the calculation of
the net pay.
As we can understand, the study of any formation
starts with the determination of the lithology of each bed
which composes it, it means its mineral composition and
its texture. This first step requires the interpretation of the
log data with the help of cross-plots for composition, and
18
This study will be completed by the determination of
the transformations (diagenesis) and deformations (tectonics) undergone by the formations since their deposition.
Finally, formation evaluation and reservoir interpretation will take into account the geological data provided by
the logs to more accurately determine the petrophysical
properties of the reservoirs and their extent. In other
terms, the understanding of the reservoir properties and
behavior will be improved if the reservoir can be put back
in its geological setting (environmental, diagenetic and
tectonic).
Before covering these different steps, preliminary
works must be achieved.
Log Quality Control
Previously to any log interpretation it is very important
to check the quality of the logging data.
This starts by the control of the calibrations.
Tool calibrations
Each tool when it goes out from the factory must.be
calibrated. This is achieved by laboratory measurements
either in special holes like the pits in Houston, or in calibrating blocks existing in each service company. But, as
the tool-component properties can vary with time (age of
the sources or of the detectors, electronics influenced by
temperature...), it is of the utmost importance to check at
the shop and on the well itself the good functioning of
each tool and to recalibrate it.
Repeat section
After each run, a part of the well (generally the most
interesting section) is again recorded. This repeated
interval is called the repeat section. Comparison of the
two runs allows the detection of possible problems (Fig. 121 next pages). The repeat cannot be perfect, especially
for nuclear data. This is linked to the statistics but also to
the fact that the tool (litho-density for instance) does not
follow the same trajectory. In addition, it can occur that the
gamma ray of the repeat run shows higher values. This is
generally due to delayed gamma rays induced by neutron
Serralog 0 2003
Generalities
I Chapter 1 1
19
QUANTITATIVE BOREHOLE IMAGE ANALYSIS
OEOLOG~ICAL
OBJECTS DOWN TO A FEW MILLIMETERS SIZE
MATHEMAT~CAL
TOOLS ALLOW TO AUTOMATICALLY
A FAST ZONATION CAN BE EXTRACTEDFROM
BINARIZED IMAGES DERIVED WITH AN
ARE IDENTIFIED WITH BOREHOLE IMAGES
EXTRACT AND Q U A ~ F Y
THEM
Landscape view of the
conductivity image.
Landscape view of the
gradient image.
ADAPTAT WE BACKGROUND
FOREACH ZONE, GEOLOGICALOBJECTS ARE AUTOMATICALLY
RESISTIVITY FILTER
EXTRACTED AND THEIR ATTRIBUTES ARE COMPUTED
THIS METHOD INVOLVES A PRIMARY STEP OF INTERPRETATION AND THE IMAQE ANALYSIS
IS GUIDED B Y THE QEOLOQICAL KNOWLEDGE OF THE FORMAWONS
DEPOSITIONAL
SURFACE ATTRIBUTES:
- Resistivity contrast
- Dip end azimuth
- Planarity
- Proximity
LAYER
ATTRIBUTES:
- Medianconductivity
- Thickness
DIAOENETIC
HETEROGENEITY
TYPE AND DISTRIBUTION:
- Size, shape, area
- Frequency
- Connectedness
- Median conductivity
STRESS INDUCED
FRACTURE
ATTRIBUTES:
- Polarity (resistive,conductive) - Dip, azimuth
- Length
- Electrical aperture
- Density
- Planarity
THEINTEGRATIONW m i STANDARD LOGGING MEASUREMENTS, CORE AND FORMATIONTESTS PROVIDE
VOLUMES AND
A QUANTITATWE ClIARAClERUATlON OF THE INT€RNAL ORQANlzATloNOF RESEWTHEIR BOUNM#Q SURFACES TO BE?TER DEFENE THE FLOW PROPERTIESAND RESERVOIR MOMLS
Figure 1-20 - Image analysis and quantification
(from Delhomme & Motet, 1993).
Serrabg 8 2003
19
20
Well logging and Geology
E
P
Figure 1-21 - Comparison of main log (left) and repeat section (right) allowing the control of the log quality. Obsetve the small depth shift on
resistivity and neutron-density between the main log and its repeat section.
interactions with elements.
Another test of the data quality is to cross-plot the
repeat-section data versus the main-run data on the same
depth interval (Fig. 1-22).
Comparison of histograms of log data recorded during
the main run and the repeat run is also another way to
check the logging qualities (Fig. 1-23).
Depth match
As the tools are either run one after the other or combined (triple-combo or Platform Express) but superposed,
depth match of the log data are needed to put back at the
same referenced depth the different measurements corresponding to the same bed. As already mentionned,
tools can stick and suddently jump due to the cable elasticity. In the absence of accelerometer data, this step of
the preliminary works is sometimes difficult to achieve
correctly. Analysis of the tension curve, the caliper and the
Dr curve, allows sometimes the determination of the
depths where tool-sticks can occur. With accelerometer
data, this depth match is generally automatic as the variations of logging speed are recorded and the length between measurement points are well known (Fig. 1-24).
Figure 1-22 - Cross-plots of repeat section data versus main log data. GR of repeat run (GR-R) is very often higher than GR main run. Density
repeat (RHOB-R) is closer to the density of the main run but still with variations.
20
Serralog 0 2003
Generalities
---
I
Chapter 1
I
21
Figure 1-23 - Comparison of histograms of log data, on the same depth interval, corresponding to the main log (on left) and to the repeat section
(on right). Some differences can be observed even if the general aspect is preserved.
Mr--
Composite-Log
As soon as the depth match of logs is correctly achieved, the composite-log must be realized (Fig. 1-25 which
provides an example of composite-log including image
data). This document combines all the information recorded in a well and will help the analyser in his work as he
will have all the available data under his glance.
Segmentation in electrobeds
Figure 1-24 - Comparison of logs before (second track) and after (third
track on the right) corrections for running speed variations recorded by
the accelerometer included in the tool string
(courtesy of Schlurnberger).
Serralog 0 2003
The next step of the preliminary work consists to segment the interval to analyse in electrobeds. As already
mentionned, due to their lack of resolution, some data are
not representative of geological realities. The segmentation of logs in electrobeds allows the elimination of these
non-representative data and the correction of the log data
for error measurements and statistics. The logic of segmentation is explained (Fig. 1-26, next pages). Example
of log segmentation is shown in Figure 1-27 next pages.
Several log analysts think that clustering technique or
neural network approach eliminate automatically the nonrepresentative data. In certain conditions - thick beds and
high resolution tools -those data are partly included in the
cluster. In thin bed conditions, as the data corresponding
to the transition zone from one bed to the following can be
numerous and generate clusters which have no meaning.
In any case, the non-representative data affect the log
responses corresponding to each bed modifying the
representative values of each bed.
21
Next Page
22
Well logging and Geology
Gamma Ray
0
( W
LLd
LLS
MSFL
200
0 Pe(b/e) 10 -----20 Th (ppm) -20
-10 K(%)
10
20 U (ppm) 0 -----1.95 Density (g/cm3) - 2.95
45
Neutron (P.u.)
-15
200 Transit time (ps/ft)
0
---
--------
Hole
drift
0" 10"
Caliper
1.25 C 1-3
11.25 C2-4
(in.)
11.25
1.25
0"
- 0.5
.......&
.... 0.5.
1
FMS Images
ohm-m 1000
360"
II
-
Figure 1-25 Example of composite log combining standard logs with image data repmduced at the same verfical scale.
22
Semlog Q 2003
Previous Page
Generalities
I Chapter 1 I 23
which is illustrated by Figure 1-31.
-
Figure 1-28 Zone investigated by a
tool in a vertical well. In such a case,
the vdume investigated may be parallel to the bedding and may cornspond
to a unique bed if this bed is sufficiently thick compared to the vertical
resoltion of the tools
sy of Schlumbetyer).
L
-
Figure 1-29 Case of beds not perpendicular to the borehole axis. In that case, if
the beds are relatively thin compared to
the tool vertical resolution, the measurement is not representative of any bed
physical ptvpe@.
+
-
Figure 1-26 Explanation of the log-segmentationlogic taking into
account vertical resolution, statistics and measurement emrs.
Well and Logging Types
In petroleum industry several well types exist. The logging data must be associated to the well type in which
they have been recorded before any interpretation.
Open holes
In those types of wells, generally any type of logging
data can be recorded but one must know several parameters related to the well: the borehole size, the orientation and the reached depth. In addition, the apparent dip
between the borehole axis and the beds must be known
as well.
F
N
s
Fardetector
Neardetector
Swm
-
Figure 1-30 Case of a deviated well in the direction of the dipping. As
it can be imagined the volume analyzed by the near and the far detector of this litho-density tool does not cornspond to the same volume
of rock at the bed boundary.
Vertical wells
The spatial geometries between tool and beds are
illustrated by Figures 1-28 and 1-29. They correspond to
many old situations as, initially, the wells were fundamentally vertical and the logging acquisitionwas essentially by
wireline techniques.
Deviated wells
In that case the geometry is most of the time the one
illustrated by Figure 1-30.
Horizontal to subhorizontal wells
The recorded data are influenced by a different volume
serralog 82003
c.--
-
Figure 1-31 Case of a horizontal
well. As it can be observed, the zone
investigated has a symmetry perpendicular compared to the bedding
(courtesy of Schlumbetger).
J
Interpretationof the logging data, and more especially
of the resistivity data and the images of the borehole wall,
must take into account the relative angle between the
bedding and the tool axis (Fig. 1-32).
23
24
Well logging and Geology
Figure 1-27 - Comparison of standad log presentation (top) with segmented log presentation (bottom). This segmentation eliminates the non
representative data and generates a geological display of the interval in electrobeds closer to the bed notion.
24
Serralog 0 2003
Generalities
I Chapter 1 I
25
Table 1-5
Storage of the standard logging data corresponding to each electrobed.
4
n
When recorded in a deviated or horizontal well the
investigation of the borehole by any tool is different from
the one realized in a vertical well (Figs. 1-28 to 32). This
must be taken into account when analyzing the data.
Figure 1-32 - The volume investigated by a tool is not affected by the
same type of formation in a vertical, deviated and horizontal well. The
tools are analyzing in adifferent way the beds composing the formations. This generates an anisotropy and a different type of images.
This must be taken into account when analyzing logging data
(courtesy of Schlumberger).
Since more than 20 years, thanks to the technological progresses, the acquisition of logging data is now
possible during the drilling itself (Logging While Drilling or
LWD).
The development of deviated and horizontal wells has
also been possible taking into account the information
provided by Measurements While Drilling (MWD) and
sensors close to the rock bit (well trajectory, drilling effi
ciency and formation properties). Even when deviated or
horizontal, the wells start vertically, with a conventional
rotary bottomhole assembly.
Figure 1-33 - RAB image in a horizontal well
(courtesy of Schlumberger)
CDNtool PowapUku,
The drillstring and bit are rotated from the surface
either by a rotary table on the derrick floor or a motor in
the traveling block, called a topdrive or “locked” assembly.
The rotary assembly can be replaced with a steerable
motor, driven by mud flow. When mud is flowing, the
motor rotates the bit, but not the drillstring which slides
along. This type of drilling is called sliding mode. The
direction of the bit is measured and sent to surface by
MWD equipment. These measurements include azimuth
and inclination with respect to the vertical.
To the MWD measurements are added different LWD
measurements such as resistivity, density, neutron, photoelectric index Pe, sonic slowness, radioactivity and if
required Resistivity-At-the-Bit tool (Fig. 1-33) which can
provide images of the borehole. These measurements will
allow a better formation evaluation during the drilling itself
Serralog 0 2003
Figure 1-34 - On the top and the left, example of surface and
downhole components of LWD tools (IDEAL Integrated Drilling
Evaluation And Logging system of Schlumberger).
On the bottom right, sketch of the ADN tool of Schlumberger.
(courtesy of Schlumberger).
25
The recording of any wireline logging data in horizontal wells requires a special technique called tubing
conveyed system or Tough Logging Conditions (TLC* ) for
Schlumberger. This technique uses a drill pipe to push
logging tools and a side-entry sub for the entry of the
cable in the drill pipe. The side-entry sub precludes the
circulation of drilling mud to loosen or advance the drill
pipe.
The LWD measurements do not require any special
equipment other than the sensors inserted in the drill collars or the stabilizers. Another advantage is the acquisition of data during the drilling itself allowing a rapid reaction in order to maintain the drill bit inside the reservoir by
comparing top, right, bottom and left measurements of
density and photoelectric sensors (Fig. 1-35).
Based on this set of log data one can start an interpretation in order to completely describe these electrobeds in terms of composition, texture, structure, facies,
sequence, environment, diagenetic effects, tectonic deformations, petrophysical properties, etc.
This interpretation can be achieved “manually” or with
the help of specific softwares written for computers or
workstations (Table 1-6). In any case, the results of this
processing by any software must be controlled.
Table 1-6
Interpretation programs of Schlumberger.
4pplications
~~
Quality control
Lithology
I I
Wellsite
I
LOGOS
I
I
I
Reservoir
modeling
LQMS
FLIP*
Facies
FACI0LOG
DimQC
II
BorTex:
ScanTex
FasTex
SYNDIP
(SYNDIP)
(STRATIM
DUADIM)
FLIP*
I
GeoFrame*
Rockclass
RockCell
Litho QL
LITHO
LITHO
Platform FACIOLOG
Express ESTIME
Quicklool
Sedimentary
patterns
I
SediView
DipFan
analyzer
Stratigraphy
Figure 1-35 - On the left comparison of the sensors put into the drill
collar (bottom image) and on the stabilizer (top image). On the right:
ADN images acquired in a horizontal well (courtesy of Schlumberger).
FLIP*
DipTrend
FIL
DCA HDT
HDT AN
SHDT AN
FracView*
Cased holes
In such wells the presence of casing and cement between the tools and the formations reduce the possibility of
measurements. Nuclear measurements, such as natural
radioactivity, neutron interactions (thermal neutron decay
time, induced gamma ray spectrometry), and density can
be recorded. Resistivity measurements through casing is
also possible using a new tool, CHFR* (Cased Hole
Formation Resistivity), introduced by Schlumberger. In
any case, the interpretation of the recorded data is more
difficult as the nature and amount of cement is not precisely known. Images and dip data are not available.
StratLog II
StrucView
Cross-section
BorView
FasTex
Workstation programs
As workstations are now available in many offices the
next paragraph will describe the one developed by
Schlumberger.
The other service companies and the major petroleum
research companies have their own workstations based
fundamentally on the same principles.
Conclusion
Workstations
At the end of this preliminary work, it can be useful to
collect all the logging data corresponding to each electrobed into a table (Table 1-5 previous page). This part of the
work corresponds to the analytical step which, generally,
will not be argued.
The emergence of workstations in early 1980’s years
has considerably simplified the work of geoscientists. This
is due to several factors:
- import data from different origins in the same environment;
- set of softwares for processing and interpretation of
26
Serralog 0 2003
Generalities
raw and computed data, and display and storage of
results;
- interactivity;
- rapidity;
Workstations have relieved geoscientists of all the
repetitive, tedious, sometimes difficult, and time-consuming part of the analytical aspect of their work. They simplify interpretation as they allow users to easily manipulate any kind of data and display results of processing.
Integrating all well data (logs and images, cores, tests,
raw and computed data, well seismics) in the workstation
environment aids geologists, geophysicists, petrophysicists and reservoir engineers in their evaluation of formations. Geologists can improve the structural, sedimentological, diagenetic and stratigraphic analysis of logs calibrated on core data. Geophysicists can improve the processing and the interpretation of their seismic sections using
data provided by well logs. Petrophysicists can refine log
interpretation and minimize uncertainty taking into
account the geological information provided by the logs.
Reservoir engineers can better understand the dynamic
potential of a reservoir through recognition of petrophysical volume properties and surface transmissibility.
Chapter 1
I
27
could locally enhance part of an image to emphasize
details, make stereographic projections of selected dips
or have a three-dimensional view in any direction for better interpretation of the images. They could also select
thresholds for the images to make a vuggy porosity estimation or a sand bed count with cumulative bed thickness.
During the work session, users could make color
screen dumps of selected intervals. These dumps can be
updated and customized with TOUCH software. When the
interpretation is complete, a continuous color hard copy of
the images, plus the computed dips and the annotations,
can be made over the entire interval (Fig. 1-36).
A rapid description of the different processings available in the Schlumberger-GeoQuest workstations will allow
geoscientists to better know the type of information they
can extract from logs and images and what they can
achieve in terms of interpretation.
Formation Image Examiner
J
Figure 1-36 - Examples of functionalities existing in FLIP Oriented
images, curves, dips, three-dimensional views, azimuth frequency plots
and annotations can be generated and displayed on the screen or reproduced as a color screen dump
(courtesy of Schlumberger).
The Formation Log Imaging Product (FLIP) was the
first workstation introduced to interpret images. FLIP was
a fast, user-friendly, flexible tool used on a SUN workstation. This Formation Image Examiner displayed all the
information necessary for quality control and interpretation of Formation Microscanner*, FMI*, ARI* or acoustic
images. The display was easy to use and fast enough to
allow extensive browsing through the available information, allowing users to concentrate on the data rather than
on the program. Users could display the images at different horizontal and vertical scales, scroll up and down the
log, and view the borehole and its azimuth in three dimensions. They could use their expertise to detect, analyze
and interpret features on the images, and could isolate
individual features and analyze the interesting parts of the
images.
The interactive, user-friendly FracView program is
designed to derive quantitative information relating to
structural and stratigraphic information (Cheung & Heliot,
1990) as well as fracture dip, fracture orientation, fracture
density, fracture aperture and porosity (Luthi & Souhaite,
1990). Healed (Fig. 1-37) or open natural fractures (Fig.
1-38) can be recognized and separated from induced
fractures (Fig. 1-39) and bed boundaries (cf. Chapter 9).
Users had a set of built-in interpretation facilities at
their disposal. They could document their observations
directly by annotating the images or logs, but interactive
interpretationtools are available. Interactive dip computation allowed users to indicate directly on the image what
they recognized as bedding, laminations, erosional surfaces, stylolites, fractures or fault planes. In addition, they
Handpicking fractures may be time consuming and
tedious. This motivated the development of a batch-oriented product (SPOT) that automatically extracts conductive
fractures from a set of electrical images. FMI, Formation
Microscanner and ARI images can be processed.
Serralog 0 2003
FracView* fracture synergy log
Stereographic projections can be generated to improve the interpretation or to analyze the fracture network in
conjugated sets.
The image-derived quantitative information is displayed by a set of synthetic logs (Fig. 1-40).
27
28
Well logging and Geology
on a workstation. DIAMAGE is available on Sun workstations. It uses the X Windows SystemTM (MOTIF) interface.
Core photograph
Figure 1-37 - Conjugated cemented fractures in an eolian sandstone
(courtesy of Schlumberger).
0.50
Figure 1-40 - Example of quantitative logs generated by the Fracview
program. The density of fractures, fracture trace length, fracture aperture and fracture porosity are displayed as a function of depth. The
fracture dips are also reproduced and separated from the dips related
to bed boundaries (courtesy of Schlumberger).
Core photographs are taken by the AUTOCARTM
photography system developed by the Societe
Europeenne de Systemes Optiques (seso), which captures the image in colors of a cylindrical core's outer surface - without distortion - before any destructive intervention
occurs. The core is sprayed with water for a better color
contrast, and then is rotated on its axis (Fig. 1-41). A device synchronized with the rotation moves a film perpendicular to the core axis. The system controls photographic
focus, sprinkler startup, rotation, synchronization,
sequence and shutter action.
Figure 1-39 - Drilling induced
fractures associated with natural open fractures. The imgular features of the drilling induced fractures, which are not
continuous, generally parallel
to the borehole axis and
concentrated in a direction perpendicular to the natural fracture dip planes. The open conjugated shear fractures have
Figure 7-38 - open natural conjuga- dips equal to N270" and N73"r
with a strike approximately N-S
ted fractures in a granite
(courtesy of Schlumberger).
(courtesy of Schlumberger).
DIAMAGETM
D~AMAGEis a user-friendly software, written by ELFAquitaine and commercialized by Schlumberger, that
allows import, display and interpretation of core photographs and borehole wall images with other borehole data
28
Figure 1-41 - Photograph and sketch explaining the AUTOCAR system
(courtesy of Elf-Aquitaine).
The core photograph is scanned, and the image is loaSerralog 0 2003
Generalities
ded and displayed on the workstation using DIAMAGE
software (Fig. 1-42). Integration with all other well data
allows the following:
- precise and accurate depth match of the core with
borehole data
- automatic core orientation (depth and azimuth) by
selecting representative dips
- merging, splitting, flipping, rotating and shifting core
images for comparison with oriented borehole wall images
I Chapter 1 I
29
- statistical studies of fractures and borehole ovalization
- identification of drilling-induced fractures
- interactive dip computation and classification by fitting a sine wave or selecting points
- modifying image-color restitution using a variable
palette for enhancement of details
- interactive drafting of lithologic column
- detection on borehole images of fine details not visible on core photographs
- adding textual and graphics comments referenced in
depth
- simultaneous display of openhole logs and test
results (RFT* Repeat Formation Tester, MDT* Modular
Formation Dynamics Tester and DST [drill stem test])
GeoFrame* Geology
The GeoQuest GeoFrame integrated reservoir characterization system delivers the softwares users need to
develop fast and reliable, risk-reducing solutions that will
help geoscientists and engineers with prospect interpretation, well planning, reservoir behavior understanding and
field development management. GeoFrame comprises
applications specifically designed for project data management, seismics, geology, petrophysics, borehole geology, visualization and interpretation, and reservoir analysis,
launched from the GeoFrame Application Manager.
GeoFrame software applications have standardized features for software and data loading, as well as transfer for
moving back and forth between projects. This tight integration workflow process and ensures that changes to
one display are reflected in another. In addition to the processing softwares for dipmeters and imaging tools, described previously (cf. Volume 1 - Chapter 24), GeoFrame
offers a set of softwares for preparation of the standard
log data and for their interpretation that are described
now.
R
Figure 1-42 - Example of DIAMAGE display. On the left the FMS
image, in the center the oriented, depth matched core photograph.
On the right the dips computed from the images
(courtesy of Elf-Aquitaine).
- mirror imaging of core photographs to obtain a view
from inside for comparison with borehole wall images
- typology and identification of features and events
seen on borehole images
- precise location on images of plugs collected from
core if core photograph is taken after collecting the plugs
- detailed description of formations
- calibration of images for quantitative interpretation
- analysis of texture, sedimentary structure, facies and
sequence for sedimentology and stratigraphy applications
- three-dimensional analysis of sedimentary features
- classification of the hierarchy of surfaces (stratifications, bed boundaries, truncation, erosion, fractures, stylolites, etc.) and validation of permeability barriers
Serralog 0 2003
Data PreDaration
Data Preparation products - WellEdir and PrePlus* bring a powerful set of user-friendly tools for data editing
and environmental correction.
The WellEdit module comprises:
- general log curve editing,
- log depth match by stretch and squeeze depth shifting,
- core data and core image editing.
The PrePlus module applies appropriate environmental corrections to log data.
Geoloay
GeoFrame geological applications softwares include
StratLog /I*, Wellpix*, WellSkefch*, BorView* , BorTex
and DipFan.
29
30
Well logging and Geology
Stratlog I/ delivers all the tools needed to interpret
complex geology. It offers numerous easy-to-use features:
- single-well interpretation and display that incorporates any type of exploration and production data - geological, geophysical, petrophysical, engineering and production data into a single-well display to support your own
interpretation process;
- creation of structural and stratigraphic cross-sections;
- creation and manipulation of multi-horizon contour
maps.
-
The Wellpix module allows the user to rapidly pick
geologic markers and faults. Any marker picked in WellPix
is available in any GeoFrame application. To interpret
complex areas, WellPix offers a variety of features including fault gapping, flattening, independent scrolling and
log drag, stretch and squeeze.
Wellsketch brings powerful spreadsheet capabilities
and graphic editors to foster quick access, effortless editing and robust functionality.
The BorView program is the equivalent of the FLIP and
FracView programs written under LOGOS previously described. The BorView program offers a group of interactive
modules for interpreting Formation Microscanner and
FMI images. BorView incorporatesthe following modules:
- The lmageView interactive module supports the
display of high-resolution borehole images. Functions
include:
- interactive dip picking by selecting points or fitting a sine wave template directly on the images. The user
can immediately classify the computed dip as a bed
boundary, lamination, cross bedding, erosional surface,
stylolite, fracture or fault plane.
- histogram interaction and normalization
- fracture aperture analysis
- The Attributeview module computes a certain number of object attributes, such as dip, azimuth, planarity,
fracture density, fracture length, fracture aperture and
fracture porosity.
- Stereonetviewallows the user to make stereographic
projection (Wulff and Schmidt plots), histograms and
rosette diagrams, and all the manipulations on stereonet.
- StrucView allows the user to construct cross section
either from a dip file or through the automatic picking of
dips on the images. This module uses DipTrend logic.
The purpose of the BorTex program is twofold:
- extract and summarize the information about the
internal organization of the formation encountered by a
well, based on the analysis of dipmeter fast channel or
image microresistivity curves
- provide a zonation of the well into significantly different "morphofacies" using hierarchical clustering techniques, based on the information previously extracted and
summarized.
30
The information derived from the dipmeter curve or
image morphology is summarized by BorTex as logs, called BorTex "summary logs", so that it can be easily combined with traditional logs.
Intervals characterized by the same summary log
signatures define dipmeter curve- or image-derived
facies. Such facies are called "morphofacies" because
they are based on morphology, to distinguish them from,
e.g., "lithofacies" that are based on lithology.
BorTex thus combines in a single module new functionalities for electrical SCANner image TEXture analysis,
which is referred to as ScanTex,with the functionalities for
dipmeter FASt channel TEXture analysis offered by
FasTex of the DipFan program for dipmeter curve analysis (see below).
ScanTex inherits from some algorithms and concepts
originally developed for the SPOT software developed
under LOGOS. It is a geology-driven image analysis. Its
goal is a precise and complete extraction of the information provided by the images.
A common user interface gives access to both FasTex
and ScanTex algorithms. In wells where electrical imaging
tools such as FMI or Formation Microscanner tools have
been run, ScanTex algorithms is preferably used.
However, it is also possible to use the FasTex algorithms,
either to speed up the processing or for the sake of
consistency in a field study context. In such a case,
pseudo-dipmeter curves are first computed from
FMVFormation Microscanner data to replicate SHDT or
HDT fast channels.
For each pad or flap image, an image analysis technique reconstructsthe part of the image that extends from
right to left (i.e., the image-crossingcomponent). Then, a
reconstruction across the borehole, guided by the bedding dip information input to BorTex, is applied. This gives
the borehole-crossing component of the images.
Conductive and resistive heterogeneities are sorted into
spots and patches. Both spot and patches have boundaries that are located at maximum-contrast lines in the images but spots are local regions flagged by a local maximum and with normally a high contrast with respect to the
background. Patches are larger and more diffuse, showing only a medium contrast with respect to the background. In some instances, spots are small regions that
only show a medium contrast.
The results are displayed over the images, providing
boundaries of beds or layers, and contours of isolated
events. In addition, layer conductivity and high-resolution
textural curves are generated (Fig. 143).
A statistical study of the results is also achieved. This
allows a summary of the information and a data compression for comparison with other logs. Sliding statistics of
attributes (percentage, thickness, contrast, connectedness, size, aperture, etc.) are done. The results are
displayed as summary logs .
Based on these results, an automatic interval segmentation is achieved applying clustering techniques.
Serralog 8 2003
Generalities
814B.8
8741.2
8741.6
Figure 1-43 - Example of object attribute computation (a) original
image. (b) Vug area percentage and median size log (computed with a
0.5 ft [15.24 cm] sliding window). (c) Vug contours. (d) Vug connectedness log and boundaries of the vugs and conductivity paths (in white)
(from Delhomme, 1992).
DipFan (Dipmeter Facies Analysis) is a complete software package for geological interpretation. It is integrated
to GeoFrame. DipFan has been developed jointly by
Schlumberger and Agip SPA. Integration of dipmeter data
with other openhole logs allows exploitation of the sedimentological and stratigraphic information contain in the
complete logging suite. The interactivity of the chain permits a user-oriented interpretation of the well. The processing chain is built from six modules that can be run to
suit each particular case. Three primary modules, FasTex,
SediView and Sequence, are geologically oriented. They
can be used separately but overall results are enhanced
considerably by the three DimQC, StatPack and RockCell
that are optional utilities modules.
- DimQC performs basic quality control on the dipmeter input data. It represents the first step in the reliable
interpretation of dipmeter data. This module consists of
three parts.
1 - lndividual curve analysis. Individual curves, such as
calipers, tension or navigation (inclinometry), are tested
following user-defined thresholds. These tests allow the
detection of anomalies during the recording.
The dip data are tested in the same manner. All tests
are applied on either a level-by-level or on a window
basis.
The module can be used interactively or in batch mode
with default threshold values. The results of this quality
control can be displayed with doubtful data points indicated by changes in color.
2 - Fast channel qualification. In this part fast channel
data are controlled and any degraded data are crosschecked against those attached to the tension, caliper
and navigation channels. This process allows the degraded data to be classified as tool- or hole-related.
Serralog 0 2003
I
Chapter 1
31
3 - Severity curve computation. Two severity curves,
RawQC and TadQC, are computed in the final phase of
the DimQC module. The RawQC curve is computed from
the fast channel qualification results and the user can
either set the weight of each defect for the severity computation or use the default values. The TadQC severity
curve is then computed combining RawQC values with
the results of dip data and navigation curve checks.
The results are displayed as flagged zones and as
severity curves.
- StatPack is designed to provide the user with a number of basic data handling utilities including analysis techniques (stereoplots, frequency plots), statistical and
mathematical functions together with common graphic
presentations such as azimuth vector plots (Fig. 1-44).
This last plot displays the progression of dip azimuth with
depth and is useful in the identification of major breaks in
bedding.
well test
'
interval. 2900.21 - 3300.00m
Cut-off : Dip quality factor
Cut off (mirdmax) 12 00 / 20 00
Number of points 747
Depth wlor scale
R 2900-2940
2WO-3105
-_.-
__
3105-3173
3 3173-3275
3 3275-3300
Figure 1-44 - Example of azimuth vector plot
(courtesy of Schlurnberger).
For instance, FasTex uses the StatPack library in order
to run Cluster or principal component analysis, and results
are displayed under the form of dendrogram, cross-plot or
spider plot.
When running SediView the library provides statistics
on dips such as azimuth frequency plot, azimuth or dip
histograms, stereonet projection, Bengtson plot, dip vector plot and cross-plot of dip magnitude versus dip azimuth.
Other statistical analyses can be done such as minimum variance analysis or Markov analysis. Other utilities
(e.g. user-defined formulae) are also included.
- RockCell performs multi-well facies analyses based
on the assumption that the facies information is known
and contained in the well logs. One or more wells of a field
are selected as the "key wells" and are used to define the
set of facies estimators. The input data may be the open31
32
Well logging and Geology
hole logs, logs derived from dipmeter data, or images, or
indeed any other depth-indexed facies indicator. A model
is established which relates the estimators to each other
and to facies based on multi-dimensional histogram
techniques, neural network, or multidimensional ellipsoids.
Three phases form the RockCell module.
1 - Key well data analysis. This phase provides the
user with powerful working environment in which he can
display the results from an unlimited number of wells.
Data can be analyzed versus depth or in cross-plots.
Any user interaction in one view is visualized in the other.
Intervals representingvarious facies can be selected and
labeled on both the depth and the cross-plot views. Log
curves which represent a facies description from the key
wells can also be used.
2 - Model building. The model consists of the list of
estimators available in the different key wells and the
multi-dimensional histogram which relates the facies
information to the estimator value distribution. The log
curve axes of the multi-dimensional histogram are userdefined with each axis divided into equal bins. However,
the axes may also be functioned to suit specific needs.
3 Facies estimation. The goal of this final phase is to
assign the most likely facies to each level of the target
well. In practice, a number of facies will be found, and
these are retained in decreasing order of likelihood.
One or more facies models can be used for the interpretation. In this case, each model and the relevant set of
parameters are associated with a particular depth interval.
The results are displayed in a user-defined format
together with the estimators for quality control purposes.
In practice, only the first three facies estimated by the
model are displayed.
-
- FasTex analyses the character of the dipmeter microresistivity curves, or fast channels - electrical imaging tool
data sets may be input to the FasTex module since
pseudo-dipmeter curves that replicate the dipmeter fast
channels can be computed (Fig. 1-45).
As previously explained, these curves reflect the internal organization of the formations encountered by the
well. The module identifies intervals with similar curve
response and proposes a zonation of the formations into
"morphofacies": homogeneous, heterogeneous or layered.
In order to ensure the reliability of the answers, FasTex
is normally run after DimQC, this module having identified
bad hole or floating pad intervals.
The program performs a geologydriven analysis of
the dipmeter fast channels. FasTex algorithms basically
compare dipmeter fast channels with each other along the
apparent bedding planes. The borehole-crossing part is
called the background conductivity log. Conductive and
resistive anomalies (i.e., uncorrelated events) are extracted by detecting peaks and troughs that are present in the
fast channels but not in the background.
Summary logs are used to express the information
contained within the dipmeter fast channels. These curves reflect the layering and the degree of heterogeneityof
the formations. Attributes such as thickness, resistivity
contrast and polarity are used to describe raw curve
events.
Using the StatPack library, a clustering technique is
applied to the summary logs. The result is a segmentation
of the formations in different zones. The user can monitor
the aggregation process through various statistical plots
such as the dendrogram. The number of clusters, or
"morphofacies", is set via the input parameters and the
process is fine-tuned by using a vertical logic which takes
into account a minimum zone thickness.
-
Sediview derives accurate sedimentologicalinformation from dip and lithological results. The module helps
reservoir delineation by inferring geometrical information
from depositional features. The module also helps predict
locations and directions of permeability anisotropy.
SediView consists of four separate steps.
1 Data loading. Pre-computed dip data are loaded
from the GeoFrame database. Lithological information
can also be loaded from the same database, but may be
manually input or computed from application of simple
thresholds to the openhole logs.
2 - Data filtering. A variety of filtering techniques may
be applied to focus the interpretation on meaningful dips.
A typical filter would be the dip quality index, output by
DimQC. In addition, lithological zoning may be specified.
3 Structural dip evaluation. This phase evaluates and
removes the effects of post depositional deformationor tilting. An innovative and accurate approach, based on
Local Curvature Axis techniques, is used to derive the
structural dip which can then be removed in the standard
manner.
-
-
-
Figure 1-45 Image data (center) may replicate fast channels of HDT
(lefi) or SHDT (right) data in order to compare with dipmeter data
recorded in other wells (courtesy of Schlumberger).
32
The structural dip is traditionally determined from
zones where the dips are constant in magnitude and azimuth (green patterns) and that correspond to low energy
deposits (essentially shales or mark). The value of these
dips provide evidence of structural tilt. However, in thick
Serralog Q 2003
Generalities
sandstone intervals with cross-bedding or foresets the
determination of the structural dip is more difficult.
SediView makes use of the observation that the cylindrical or conical bed surface axes within each bedset are
often co-planar. The dip and orientation of this plane
reflects the effect of the post-depositional deformation.
The program for automatic structural dip determination in
such cases is based on the two following observations,
and one assumption (Beaudoin eta/., 1983). The first observation is that usually a few poles of consecutive dips
that are close in depth appear aligned on a great circle on
a Schmidt net (Fig. 1-46). This may be because of sedimentary structures, compaction or micro-structures.
-
Figure 1-46 Sedimentary features. (1 to 3) Their dips correspond to
orange sequences of small extent ffiing great circles that have cleady
different axis orientations.
(4) The axes of these sedimentary features fit a great circle as they
correspond to the same plane that charecterizesthe original depositional surface. This plane is the structural dip (SO)
(from Etchecopar & Bonnetain, 1992).
Whatever their origin, these "local" great circles can be
characterized through the orientation of their poles (Local
CurvatureAxis).
The second observation concerns the poles of the
local great circles. Most of the consecutive poles of local
great circles, also called local curvature axes, are themselves aligned on a great circle (cf. Fig. 1-46 (4)).
The basic assumption is that the plane that corresponds to the local great circle poles, or local curvature
axes, was nearly horizontalduring sedimentation and now
represents the structural dip. In other words, the axes of
microstructures or sedimentary structures, which are
Serrakg Q 2003
I Chapter 1 1
33
recognized as locally cylindrical, are assumed to have
been nearly horizontal during sedimentation.
The first pass of this module computes the local curvature axis of each bedset or laminaset. These axes are
plotted on the stereonet and analyzed interactivelyto define structural dip common to a group of beds. This value
is assigned to the corresponding interval. The technique
detects the subtle changes in structural dip that can occur
between depositional episodes.
4 - Sedimentary dip determination.After subtraction of
the structural dip for each depositional unit, a dip dispersion analysis is performed to organize the bedsets into
groups. The thickness and the azimuth of each depositional unit are then computed. These values are integrated by the interpreter with other data from images, logs
and cores as well as with general knowledge of the depositional environment in order to infer reservoir elongation
and transport current direction. Results are then
displayed.
- Sequence. The interactive Sequence module
consists of a number of independent interpretation submodules. The outputs can be combined with other DipFan
results to provide stratigraphic interpretation capability
with the maximum in user-friendliness.
Bed boundaries, defined either by the DipFan modules, or by stratigraphy and core analysis, are consolidated
and ranked into an organized geological sequence
display.
Three sub-modules form the Sequence processing
chain.
1 - Curve shape analysis. Sequence performs a semiautomatic analysis of the different logs for identificationof
bell, funnel and cylinder shape. The curve-shape analysis
is a two-stage process. Breakpoints are detected by
searching for levels at which the first derivative exceeds a
given threshold. Then, zones defined by two successive
breakpoints are classified in accordance with the slope of
the line that approximates the input log.
2 - Thickness analysis. The thickness analysis recognizes and describes thickening or thinning upward
trends in the analyzed sequence. Three algorithms have
been developed that enable the user to perform this analysis at different scales. A large scale, quick-look
approach over the whole sequence can be run using gravity analysis. For each of the facies, the average location
along the sequence is computed. This position, with
respect to the mid-point of the interval under study, indicates whether the facies is randomly distributed or shows
a preferred trend.
On a finer scale, regressionanalysis can be performed
within zones defined either automatically, or by the user.
Alternatively, parasequence stacking pattern analysis can
be inferred through interpretation of Fisher plots. The
Fisher plot is a simple graphic presentation of the vertical
trend of parasequence thickness. The resulting patterns
illustrate the deviation from the average throughout the
33
34
Well logging and Geology
stratigraphic interval.
-
3 Lithology analysis. This analysis statistically describes the lithology distribution within a sequence.
Petrophvsics
Included in the petrophysical product line are
PetroView Plus*, Rockclass* or Rockcell, Quickview,
ELANPlus* and MultiWell ResSum*.
Petroview Plus delivers accurate detailed petrophysical analyses quickly. This application guides you through
a deterministic petrophysical evaluation step-by-step.
PetroView Plus offers interactive parameters selection
from crossplots, histograms and log displays, and provides graphic presentations of input and output data.
Rockclass produces a zoned lithological display and
describes the gross mineralogy of a lithofacies, using
wireline log measurementsor computed mineral volumes
with a lithofacies database (cf. Chapter 2).
The lithologic display can be particularly helpful in correlating non-depth-indexed data such as seismic, or poorly-indexed well data such as cores, mud logs or cuttings.
The log derived lithological column becomes an especially powerful geologic tool when combined with dipmeter
correlations and high-resolution images.
The lithofacies database may be created interactively
by the user, or recalled from a collection of previously
saved files.
The Quickview petrophysical program provides quick
and easy analysis and all the necessary graphics and
tabular output to make accurate well decisions. This
module offers a standard methodology based on a well
log analysis model used since 15 years - for analyzing
openhole logs, providing both Archie and dual-water saturation solutions, effective porosity, shale volume, and
apparent grain density.
-
ELANPlus is an advanced petrophysical analysis
module that is derived from the well known ELAN software.
MultiWell ResSum is a module that computes reservoir thickness and associated propertiesfor multiple wells
and layers. To increase the precision of the interpretation,
true vertical thickness and true stratigraphic thickness calculations are available.
References and bibliography
ACKERT, D., BOETEL, M., MARSZALEK, T., CLAVIER, Ch., GOODE, P., THAMBYNAYAGAM, M., &
STAGG, T. (1988). Looking Sideways for Oil. The
Technical Review, 36, 1.
ALLEN, D. et a/. (1995). Modeling Logs for Horizontal
Well - Planning and Evaluation. Oilfield Review, 7, 4.
-
-
34
AULIA, K. et al. (2001). - Resistivity behind casing.
Oilfield Review, 13, 1, p. 2-25.
BABOUR, K., JOLI, F., LANDGREN, K., & PIAZZA,
J.L. (1987). Elements of Borehole Seismics illustrated
with three case studies. The Technical Review, 35, 2.
BATES, R.L. & JACKSON, J.A. (1980). - Glossary of
Geology. Amer. Geol. Institute, Falls Chuch, Virginia.
BEAUDOIN, B., PINOTEAU, B., DELHOMME, J.P.
(1983). - Geological analysis of logs and dipmeter data for
well-to-well correlation.
BEGUIN, P., BENIMELI, D, BOYD, A, DUBOURG, I.,
FERREIRA, A., McDOUGALL, A., ROUAULT, G. & van
der WAL, P. (2000). Recent progress on formation resistivity measurement through casing. SPWLA, 41th Ann.
Log. Symp. Trans., paper CC.
BENIMELI, D., LEVESQUE, C., ROUAULT, G.,
DUBOURG, I.,PEHLIVAN, H., McKEON, D, FAIVRE, 0, &
REBOLLEDO, K. (2002).- A new technique for faster
resistivity measurements in cased holes. Schlumberger.
BLATT, H. MIDDLETON, G., & MURRAY, R. (1972,
1980). Origin of sedimentary rocks. 1st and 2nd ed.
Prentice-Hall Inc. Englewood Cliffs, New Jersey.
BONNER, S. et a/. (1993). Measurements at the Bit:
A New Generation of MWD Tools. Oilfield Review, 5, 2/3.
BOURKE, L., DELFINER, P., TROUILLER, J.C., FETT,
T, GRACE, M., LUTHI, S. SERRA, O., & STANDEN, E.
(1989). - Using Formation Microscanner* Images. The
Technical Review, 37, 1.
CHEUNG, Ph., & HELIOT, D. (1990). - Workstationbased fracture evaluation using borehole images and
wireline logs. SPE Bull., no20573.
De GOLYER, E. (1921). - Debt of Geology to
Petroleum Industry. Bull. amer. Assoc. Petroleum Geol.,
5, p. 394-398.
DELHOMME, J.-P. (1992). A quantitative characterization of formation heterogeneities based on borehole
image analysis. SPWLA, 33d Ann. Log. Symp. Trans.,
paper T.
DELHOMME, J.-P., & MOTET, D. (1993). - Reservoir
description and characterization from quantitative bomhole image analysis. Schlumberger, Reservoir Characteriza.
tion, Africa and Mediterranean Region, no 5.
DESBRANDES, R. (1985). - Encyclopaedia of Well
Logging. Editions Technip, Paris.
DOVETON, J.H. (1994). - Geologic Log Analysis
Using Computer Methods. AAPG Computer Applications
in Geology, no.2.
DOVETON, J.H. (2002). The Geological application
of Wreline Logs: A Keynote Perspective. In: LOVELL, M.,
& PARKINSON, N. (ed.) (2002). - Geological Applications
of Well Logs. AAPG Methods in Exploration, 604 p.
ELLIS, D.V. (1987). - Well Logging for Earth Scientists.
Elsevier, Amsterdam.
HARKER; S.D., McGANN, G.J., BOURKE, L.T,. &
ADAMS, J.T. (1988). - Formation Microscanner Applications in Integrated Reservoir Description in
Claymore and Scapa Fields (North Sea). Geological
Applications of Wireline Logs, London, June 1-2, 1988.
Geological Society of London.
-
-
-
-
-
-
Serralog Q 2003
Generalities
HARKER, S.D., McGANN, G.J., BOURKE, L.T., &
ADAMS, J.T. (1988). - Methodology of Formation
MicroScanner lmage lnterpretation in Claymore and
Scapa Field (North Sea). Journ. Geol. Society of London.
HURST, A., LOVELL, M.A., & MORTON, A.C. (eds)
(1990). - Geological Applications of Wireline Logs.
Geological Society, Special Publication no 48, London.
HURST, A., GRIFFITHS, C.M., & WORTHINGTON,
P.F. (eds) (1992). - Geological Applications of Wireline
Logs /I. Geological Society, Special Publication no 65,
London.
HUTTON, J. (1788). - Theory of the Earth; or an investigation of the laws observable in the composition, dissolution, and restoration of land upon the globe. Royal
Society of Edinburgh. Transactions, v. 1, p. 209-304.
ILLING, V.C. (1946). - lnfluence of Oil on Geology. Oil
Weekly, September, p. 176.
ISLAM, M.R. (2000). - Emerging Technologies in
Subsurface Monitoring of Petroleum Reservoirs. 1st
Symposium on Well Log Analysis & Formation Evaluation
(WLA&FE), 29-31 Oct. 2000, Tripoli, Libya.
KHALIL, Mi, FERRARIS, P. & DAWOUD, A. (2001). Saturation monitoring with cased-hole formation resistivify. SPE Annual Technical Conf. and Exhib., New Orleans,
Louisiana, paper SPE 71716.
LAHEE, F.H. (1932). - Contribution of Petroleum
Geology to pure Geology in the Southern Mid-Continent
Area. Bull. amer. Assoc. Petroleum Geol., 43, p. 953-964.
LOVELL, M., & PARKINSON, N. (ed.) (2002). Geological Applications of Well Logs. AAPG Methods in
Exploration, 604 p.
LUTHI, S. (2001). Geological Well Logs. Their use in
reservoir modeling. Springer.
LUTHI, S., & SOUHAITE, Ph. (1990). - Fracture aperture from electrical borehole scans. Geophysics, 55, 7, p.
821-833.
LYNCH, E.J. (1962). - Formation Evaluation. Harper &
Row, New York.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed., Gulf Publishing Co., Houston.
POSTER, C. (1988). - Lateral Thinking in Seismics.
Schlumberger, Middle East Well Evaluation Review,
Special supplement: Borehole Seismics.
RIDER, M.H. (1986). - The Geological lnterpretation of
Well Logs. Blackie, Glasgow & London, Halsted Press,
New York.
Schlumberger (1991). - Log Quality Control
Reference Manual. M-090209.
SERRA, 0. (1984). - Fundamental of well-log interpretation. 7. The acquisition of logging data. Elsevier,
Amsterdam, Developments in Petroleum Science 15A.
SERRA, 0. (1986). - Fundamental of well-log interpretation. 2. The interpretation of logging data. Elsevier,
Amsterdam, Developments in Petroleum Science 158.
SERRA, 0. (1989). - Formation MicroScanner lmage
lnterpretation. Schlumberger, SMP-7028.
SERRA, 0. (1993). - Relations entre les donnees diagraphiques et les parametres geologiques. Les differents
domaines d'applications geologiques des diagraphies.
I ChaRter 1 I
35
Gkologues, no 99.
SERRA, 0. (1998). Role of well logging in the study
of sedimentary basins. In: Purser, B.H. (ed), Dynamics
and methods of study of sedimentary basins, Oxford and
IBH Publishing Co., New Delhi, India. Editions
Technip,Paris, p. 11-57.
SERRA, O., & ABBOTT, H. (1980). - The contribution
of logging data to sedimentology and stratigraphy. 55th
Ann. Fall Techn. Conf. SPE of AIME, paper SPE 9270.
SERRA, O., STOWE, I., & MOTET, D. (1993). - True
integrated interpretation. SPWLA - CWLS 34th Ann. Log.
Symp. Trans., paper 2.
THEYS, Ph. (1991). - Log data acquisition and qualify
control. Editions Technip, Paris.
TITTMAN, J. (1986).
Geophysical Well Logging.
Academic Press, Inc., Orlando, Florida.
Van WAGONER, J.C., MITCHUM, R.M., CAMPION,
K.M., & RAHMANIAN, V.D. (1990). - Siliciclastic
Sequence Stratigraphy in Well Logs, Cores, and
Outcrops. AAPG Methods in Exploration Series, no 7.
-
-
Remark : In all the chapters a * refers to Schlumberger's
mark.
-
-
Serralog Q 2003
35
WELL LOGGING AND ROCK COMPOSITION
Introduction
Review of general geological concepts
edly be useful to recall some basic concepts of geology,
petrography and mineralogy to non-geologists.
The three major rock types
The nature of a rock and its composition are the first
characteristics which the geologist attempts to determine,
and a knowledge of these enables him to name the rock.
This is also his first concern in the study of well logs.
He will therefore attempt to reconstruct the vertical
lithological profile from an analysis of the well logs. This
reconstruction is aimed at defining:
-the apparent thickness and the real thickness of each
electrobed, having an electrolithofacies and electrofacies,
or of each electrosequence,
- the type of rock and the elemental and mineralogical
composition of each electrobed.
In order to obtain the most reliable reconstructionpossible, the interpreter, or log analyst, must first of all procure a suite of logs which is as complete as possible, and
in addition, a good description of cores and cuttings from
the drilling process. Secondly, he must possess the basic
concepts which are necessary for a good understanding
of the problems involved in analyzing logs, and the implications that these will have in the choice of an interpretation model, be it manual or automatic.
Clearly, the interpretationof a volcanic or granitic rock
will proceed differently, and therefore will require a different model, from the interpretation of a sedimentary rock
such as an argillaceous sand made up of a mixture of
quartz, feldspar, kaolinite, illite and montmorillonite.
Moreover, taking account of the frequently limited availability of data and recorded logs, it is not always possible
to simultaneously determine the nature of the minerals
present and their proportions, especially in the case of
complex composition.
It is often necessary, therefore, to obtain additional
information from examinations of cuttings, core analysis
and local geological knowledge to establish the mineralogy model, and thereafter to restrict the investigation to
determining the percentages of each of the principal minerals assumed to be present in the rock.
Before proceeding with the problem of determining
rock type and composition from log data, it will undoubt-
Serralog 0 2003
Geologists divide rocks into three main categories
according to their mode of formation.
Igneous Rocks
These rocks arise from the solidification of a molten
mass known as magma. Depending on whether the solidification takes place at depth or on the surface, the rock
can have a coarse texture (slow crystallization), or a fine
or even glassy texture (rapid crystallization). The first type
is known as plutonic or intrusive igneous rock, and the
second as volcanic or extrusive igneous rock.
Plutonic rocks have the distinction of exhibiting practically no porosity, the crystals being tightly packed.
Alteration and fracturing of plutonic bodies usually favour
the development of some porosity and permeability,
thereby conferring some of the characteristic of a reservoir. Examples of hydrocarbon production from granitic
rocks exist in several countries: India, Egypt, Vietnam,
Libya.. .
Volcanic rocks, on the other hand, can be porous or
even very porous, but the pores are not always connected (pumice, vesicular basalt). They correspond to bubbles
separated by a thin lining of volcanic glass, which are
formed by violent depressurizationwhen fragments of viscous magma are projected into the atmosphere.
Plutonic rocks generally form the basement rock of
sedimentary sequences. Their upper zones are often
weathered, or broken down and may constitute a reservoir. Sometimes they can be found as intrusions, such as
dikes or sills, in the sedimentary formations (Fig. 2-1).
Volcanic rocks may be intercalated in the sedimentary sequence, and can therefore be encountered at any
level of the stratigraphic column. Dependingon their characteristics, they may constitute a reservoir rock.
37
38
Well logging and Geology
Figure 2-1 - Different types of intrusions of igneous rocks in sedimentary rocks (adapted from Schmidt & Shaw in Press & Sieve6 1982).
Sedimentary Rocks
These rocks arise from the consolidation of sediment
formed on the surface of the Earth or on the seabed by
the deposition of various materials, usually under the
action of gravity, acting on rock fragments or minerals of
any size transported by water, wind, or ice from their
source or parent-rock, but also by chemical precipitation
from solution or by secretion from living organisms. More
often, they are deposited in laminae, beds or strata. This
family of rocks is divided into several groups based on the
origin of the sediment.
Detrital or Clastic Rocks
These are formed from debris arising from the alteration and decomposition of pre-existing rocks, and may
be transported, often a considerable distance, by wind,
water or ice from the site of erosion and weathering to the
site of deposition.
Sediments which settle under the action of gravity at a
distance from their source are termed allochthonous or
exogenous. The particles are usually bound together by a
cement of chemical or biochemical origin formed subsequent to the deposition. It occupies part of the pore space.
This group is divided into several subgroups:
- Terrigenous rocks are formed from accumulations of rock debris from the alteration and erosion of
land-based outcrops.
- Pyroclasticrocks result from the accumulationof
fragments of solidified magma expelled into the atmosphere from volcanoes, deposited under the action of gravity, and then re-worked or altered to varying degrees by
the action of water.
- Bioclastic rocks result from the accumulation of
skeletons and other animal remains, typically carbonate
shells, and sometimes remains of vegetation.
which fall out of solution following changes of pressure,
temperature or concentration (evaporates), or in response
to chemical changes within the water, due to the activity
of organisms such as plankton, algae or bacteria (carbonates). Such rocks are termed autochthonous or endogenous because the site of deposition is usually the same as
the site of formation.
Frequently included in this group are the biochemical
rocks which result from the action of organisms such as
reef-building coral, or from an accumulation of organisms
having calcareous shells, or of siliceous organisms (giving
chert, radiolarites, diatomites), or from the transformation
of vegetable debris (humic or sapropelic) under the action
of anaerobic bacteria (giving peat, lignite, coal, hydrocarbons).
With the exception of purely chemical rocks of the
evaporate type, sedimentary rocks more often than not
exhibit connected intergranular, intraparticle and vuggy
porosity, which renders them potential reservoir rocks in
which fluids can accumulate (water, oil, gas). As a result,
they represent a major factor in the search for these substances.
Metamorphic Rocks
These result from the chemical, mineralogical, textural
and structural transformation of rocks under the action of
high temperatures, and frequently of high pressures. They
are divided into two groups depending on the type of
metamorphism by which they are created.
- Rocks associated with a general or regional metamorphism result from deep burial, and hence the simultaneous action of heat and pressure on pre-existing rocks,
and facilitate the modification of both texture and structure, and the formation of new minerals. This type of
metamorphism affects bodies of rock over large areas
and depths.
- Rocks associated with contact metamorphism are
produced by a mineralogical transformation of formations
in the vicinity of igneous intrusions, usually under the
influence of temperature alone.
The type of rock formed will depend on that of the original rock. Metamorphic rocks present any porosity or permeability other than that associated with the existence of
fractures. Regional metamorphic rocks sometimes form
the basement rock of sedimentary sequences.
Relative abundance of rocks
According to Clarke (1924), igneous rocks represent
95 % of the volume of the Earth's crust (lithosphere up to
a depth of 16 km), sedimentary rocks accounting for only
5 % (Table 2-1). If, however, one only considers the
exposed surfaces, sedimentary rocks account for 75 %,
while igneous rocks account for 25 % (Fig. 2-2).
Chemical and Biochemical Rocks
These are formed by the accumulation of precipitates
38
Serralog 0 2003
Composition
Table 2-1
Estimated volume of sedimentary rocks
(from Pettijohn, 1975).
kteference
Cubic kilometers
larke (1924)
oldschmidt (1933)
uenen (1941)
ickman (1954)
oklervaart (1955)
om 8 Adams (1966)
onov (1968)
3.7 x lo8
3.0 x lo8
13.0~10~
4.1 f 0.6 x 10'
6.3 x 10'
1 0 . 6 ~10'
I Chapter2 I 39
alone 90 % of the total.
As far as the sedimentary rocks are concerned,
Garrels 81MacKenzie (1971) have estimated that the proportions of clays, sands or sandstones, and carbonates
are 81 %, 11 % and 8 % respectively.
These values can be compared to those given by
Mead (1907), Clarke (1924), Holmes (1913), and
Wickman (1954) (cf.Table 2-4).
Table 2-3
Representationof the abundance of various types of volcanic rock in terms of the area they occupy
(from Daly, 1933).
9.0 x 10'
Exbusive Rocks
4.8 x 10'
According to the data of Wedepohl (1969, Table 2-2),
the granites, granodiorites, and quartz diorites represent
86 % of all plutonic rocks, the remainder being made up
of gabbros (13 %).
Pacific
'otal squan
Cordillera, Belt, square
miles
iquare miles
miles
Rhyolite
Dacite
2,145.7
82.1
Mica andesite
Hornblende andesite
Pyroxene andesite
(Ch@W)
3.0
21.6
3,966.0
Augite porphyrite
Basalt
Trachyte
Laiite
Phonoliie
Trachydolerite
eschenite
Nepherie basalt
(Texas)
Nephelite-meleliie
basalt (Texas)
Limburgite
Quah basalt
Totals
255.0
3,079.0
6.5
4.6
1.o
.....
.......
.......
.......
.......
130.0
5.5
0.3
02
2,146.7
82.1
3.0
21.6
3,966.0
255.0
3,209.0
0.3
1.2
.......
1.2
2.8
.......
2.8
2.5
8.0
9,584.0
.......
8.0
9,715.0
131.0
-
Figure 2-2 Histograms showing the relative abundance of igneous
and sedimentary rocks :(a) as a volumetric percentage of the Earth's
crust to a depth of 16 km, and (b) as a percentage of the surface exposure (fivm Pettijohn, 1949, based on data by Clarke, 1924, p. 34).
Table 2-4
Computed proportions of sedimentary rocks
(completed from Pettijohn, 1975).
I
Plutonic rocks
Granite and quartz monzonite
Granodiorite
Quartz diorite
Diorite
Gabbro
Others
TOM
Serrabg @ 2003
Percentage
I MtADm
A
-L
CI
(1907)a (1924)b (1913)~ (1954)d MACKENZIE
(1971)
Shale
82
80
70
83
81
Sandstone
12
15
16
8
11
Limestone
6
5
14
9
8
44
34
8
1
13
1
100
39
40
Well logging and Geology
Rock classification
In general terms, the classification of rocks is based on
the one hand on composition in terms of essential mineral percentage (e.9. quartz, feldspars, micas, amphiboles,
pyroxenes; calcite, dolomite; clays, etc.), and on the other
hand on textural properties such as the size of the crystals or particles and their arrangement.
Figs. 2-3 to 2-8, taken from Press & Siever (1982) and
Pettijohn (1975), give a summary of the classification of
rocks pertaining to the three fundamental families. Table
2-12 provides the composition, texture, parent rock, and
type of metamorphism associated with each of the metamorphic rocks.
Figure 2-4 - Classification of sedimentary rocks
(from Pettijohn, 1975, fig.2-3).
Metamorphic rocks are subdivided as a function of
temperature, pressure and foliation or granulation (Fig. 27). The composition of the rock which has been metamorphosed can also be taken into acount (marble from carbonates, quartzite from sandstones, gneiss from granite,
slate from shale, etc.).
Figure 2-3 - Classification of igneous rocks
(adapted from Press & Siever, 1982).
Sedimentary rocks are subdivided into exogenetic or
clastic rocks and endogenetic or chemical and biochemical rocks (Fig. 2-4)Pettijohn (1975). Clastic or detrital
(exogenetic) rocks are subdivided into cataclastic, pyroclastic, residues and epiclastic sedimentary rocks the latter including conglomerates, sandstones and mudstones
or shales as a function of the grain size (Fig. 2-5).
Figure 2-5 - Classification of detrital sedimentary rocks
(from Press & Siever, 1982, fig. 3-18a).
Rock composition
Sandstones themselves are subdivided into arenites
and wackes (Fig. 2-6 next page) as a function of the composition of the grains, their percentage and their size.
Chemical and biochemical rocks, called by Pettijohn
endogenetic rocks, are subdivided into non-evaporitic
rocks, which include limestones, dolostones, ironstones
and cherts, and evaporates which include gypsum, anhydrite, halite and many other salts such as polyhalite, carnallite, langbeinite, kainite, kieserite, sylvite, tachydrite,
bischofite, epsomite, trona and natron (Fig. 2-4 and Table
2-5).
40
The solid fraction of a rock is composed of minerals,
under the shape of grains or crystals, themselves com.posed d...moleGu198..m&-.I~ Qf.elementsQr-afoms(Fig.
2-8). The elements are listed in Mendeleiev’s periodic
table (Fig. 2-9).
Elements are fundamentally composed of a nucleus,
made of protons and neutrons, and electrons revolving
around it. Elements can combine to form molecules of
more or less complex composition.
Serralog 0 2003
Composition
I Chapter2 I
41
LlTNlC ARENITES
-
Figure 2-6 Classification of sandstones (adapted from Pettijohn, 1975, fig. 7-6).
Greenschist
facies
Zeolite
facies
Amphibolite
facies
Pyroxene
hornfels
facies
Temperature increase
Sanidinite
f8cies
t
Poorly foliated
Marble
Amphifflite
Quartzite
Foliated
Gneiss (coarse grained)
Schist (medium grained)
Sate (fine grained)
Tectonic granulation
Granulated
Mylonites
Cataclasites
-
Figure 2-7 Classification of metamorphic rocks. The textural gradations (bottom of diagram) do not necessarily correlate with the mineralogical
facies controlled by temperature and pressure (from Press & Sieve6 1982, fig. 3-19).
Serralog 82003
41
42
.
Well logging and Geology
Table 2-5
Classification of chemical or biochemical rocks (from Press & Siever, 1982, fig. 3-18b).
IrOnaOnOS
Evaporates
/Chefl
/Organic
Phosphates
I
Mineral
Calci
(aeonW
Dolomite
Ankerite
Magnesite
Coal
Oil
Gllr
Carbon +
hydroeen
Chemical
mporition
<1%
2.1%
2.3%
2.4%
4%
696
8%
28%
46%
Figure 2-10 - Relative abundance of elements in the earth's crust
(from Press & Siever, 1982). Compare with Table 2-6.
-
Figure 2-8 From element to outcrop through mineral and rock
(redrawn from Press & Siever, 1982).
Elemental covmposition
Table 2-6
Elemental composition of the Earth's crust
(igneous and sedimentary rocks)
recomputed from Clarke & Washington, 1924.
If 93 natural elements have been recognized, one has
to consider that only 8 are abundant as constituents of the
Earth's crust (Table 2-6 and Fig. 2-10). Elements are classified on the basis of electronic structure (Fig. 2-9).
The 8 most abundant elements represent more than
99% of the total mass of the Earth's crust. Also, as it can
be observed, oxygen is the most abundant component
both in weight percentage, atom percentage and in volume percentage. Oxygen is associated to a lot of other
elements to compose molecules and minerals. Table 2-7
lists the oxygen content of the most abundant minerals. In
average its weight percentage is close to 50%.
42
2.08
3.65
2.75
2.58
0.62
0.14
Atom, %
60.5
20.5
6.2
1.9
1.8
1.9
2.5
1.4
0.3
3.0
0.51
0.03
0.44
0.37
028
1.04
1.21
1.88
....
0.36
0.56
0.70
0.70
0.65
0.99
0.95
1.33
....
Serralog 0 2003
Composition
Table 2-7
Oxygen percentage of the most abundant minerals compbsing the Earth's crust.
Mineral
Quartz
Calcite
Domolite
Anhydrite
Orthose
Albite
Anorthite
Muscovite
Biotite
Glauconite
lllite
Kaolinite
Chlorite
Montmorillonite
43
them.
The chemical compositions, expressed as oxides, of
the more abundant rocks making up the Earth's crust are
presented in Tables 2-8 to 2-12 and Figure 2-11.
Oxygen (weight percentage)
53
48
52
47
46
48
48
50
55.7
52
53
By gain or loss of electrons, atoms become cations
(lost electrons) and anions (gained electrons). Cations
and anions bond to form molecules. Most of the crystal
space is occupied by anions that are generally larger than
cations, cations (smaller in size) fit into spaces between
Atome
1 Chapter 2 I
I
Figure 2- 11 - Oxide composition of some sandstones.
Less abundant but major geological importance
Metal
U:Solid at 25°C under 1 bar
Semi-conductor ~e : Gas at 25°C under 1 bar
Non metallic
Br: Liquid at 25°C under 1 bar
TC: Obtained by synthesis
Noble gases
Lanthanides & Actinides
Radioactive
The most abundant
Figure 2-9 - Periodic table of elements. The most abundant elements are indicated with a blue circle. The radioactive elements are indicated by a
red circle. The elements of lesser abundance in the earth's crust but of major geologic importance are indicated by an orange circle. They are constituents of common rock-foning minerals or of economic significance.
serralog 8 2003
43
Table 2-8
Chemical composition of the Earth's crust
(percentages by weight)
(from US Geol. SUN. paper 127, 1924).
Table 2-10
Chemical composition of the principal volcanic rocks
(from Daly, 1933).
Bfusive
Wde
Clarke and
Washington
59.07
15.22
3.10
3.71
0.11
3.45
5.10
3.71
3.11
1.30
1.03
0.30
72.m
0.33
13.49
1.45
0.88
0.08
0.38
1.P
3.38
4.46
1.47
0.08
Bfusive
Wte
6fuswe
jfusive Basal
Andeste
66.68
0.57
58.58
16.25
2.38
1.90
0.06
1.41
3.46
3.97
2.67
1.50
0.15
17.31
3.33
3.13
0.18
275
5.80
3.58
2.04
1.26
0.26
49.06
1.36
15.70
5.38
6.37
0.31
6.17
8.55
3.1 1
1.52
1.62
0.77
0.45
0.35
0.05
0.06
0.04
0.03
Table 2-9
Chemical composition of the principal plutonic rocks (from Daly, 1933).
FlutonicGranite
=2
Ti02
A403
Fe203
FeO
MlO
Me0
CaO
k20
60
YO
125
'
44
70.18
0.39
14.47
1.57
1.78
0.12
0.88
1.99
3.48
4.11
0.84
0.19
PkROnic
Gramdorite
65.01
0.57
15.94
1.74
2.65
0.07
1.91
4.42
3.70
2.75
1.04
0.20
Rutonic Quartz I Flutonic M
oioriie
61.59
I 56.77
0.66
0.84'
16.21
16.67
2.54
3.16
3.77
4.40
0.10
0.13
2.80
4.17
5.38
6.74
3.37
3.39
2.10
2.12
1.22
1.36
0.26
0.25
I
eI
I
I
Autonic
GaMro
48.24
0.97
17.88
3.16
5.95
0.13
7.51
10.99
2.55
0.89
1.45
0.28
Flutonic
Anorthosite
50.40
0.15
28.30
1.06
1.12
0.05
1.25
12.48
3.67
0.74
0.75
0.05
Flutonic lhnite
40.49
0.02
0.86
2.84
5.54
0.16
46.32
0.70
0.10
0.04
2.88
0.05
Serralog 0 2003
Composition
Table 2-11
Chemical composition of the principal sedimentary rocks
(from Clarke; Leith & Mead; eta/.).
Average
Shale
58.90
%O
A$%
15.63
4.07
2.48
2.47
3.15
1.32
3.28
3.72
F824
FeO
w
-0
4 0
60
YO
+
0.86
Ti2
0.17
2.87
1.48
p205
(32
MSce#aneoUS
lw.w
Average
sandstone
Average
lmtm
78.64
4.77
1.08
0.30
1.17
5.51
0.45
1.32
1.33
0.25
0.08
5.03
0.07
100.00
5.20
Table 2-13
Elemental composition of ocean waters.
Chbrine
sodim
Magnesium
0.81
0.54
7.92
42.74
0.05
0.33
I Chapter 2 1 45
I
Ppm
18980
10501
1272
Sulfur
884
Calcium
Rdassium
Others
400
380
118
Mineralogical composition
0.56
0.08
As previously mentionned rocks are essentially composed of minerals. Consequently, it is fundamental to
determine them to recognize the rock type.
0.04
41.70
0.05
Approxymately 3,300 minerals have been identified,
but the vast majority of them are rare and only found
either as trace minerals or occasionally, especially in
igneous and metamorphic rocks.
lw.w
Table 2-12
Chemical composition of the principal metamorphic rocks.
1
7293
12.67
blm
2.08
0.62
1.W
3.19
4.55
050
0.11
0.54
0.11
2
6133
1524
3.31
3.43
2.6l
4.79
329
2.05
2.38
3
5195
12.56
0.80
6.m
7.00
2.79
131)
2.67
0.14
1.a3
9940
1.02
0.43
9958
2.665
.....
.....
021
4
5153
1081
290
9.93
8.74
9.97
2.62
0.W
0.67
0.00
8.n
0.50
0.70
0.02
02%
99.96
0.38
9.
c
I
OM
024
0.57
0.71
lW.16
2.95%
5
16.45
132
4.89
5.04
3.17
2.62
2.14
0.68
0.03
0.54
Tl8Ce
6
6348
17.70
3.02
1.a
3.72
2.72
1.78
3.34
.0.87
0.23
0.52
0.58
0.48
100.38
2.m
0.45
100.81
2.680
63.04
7
8
on
8329
0.42)
1.or
5487
1. Granite gneiss, Argentina. M. Dittrich, analyst.
3. Chlorite schist, Lake Superior greenstone. S. Darling, analyst.
4. Hornblende schist, Lake Superior greenstone, near granite. F. Grout, analyst.
5. Biotitc schist, Lake Superior, from slate. F. Grout and S. Darling, analysis.
2.75
,
......
....
...........
......
) 4.40
......
...
......
100.35
......
......
10090
10
92M
421
1.w
427
3595
....
0.88
8.04
0.16
1.I6
0.96
......
43.49
2. Diorite gneiss, British Guiana. J. B. Harrison analyst.
124
0.16
4.68
27.13
TNC~
9
40.42
1Ad
0.18
1051
021
MW
Trace
1M
124
90.47
"
0.14
021
.....
100.68
11
4283
1.53
31.41
0.30
Mile
2337
b W
Trace
0.43
9987
In fact, and fortunately, the various rock types are
made up of a reduced number of minerals, and if one
considers the igneous rocks, not more than ten essential
minerals (or major components) are required for each
rock (Table 2-14).
6. Slate, Lake Superior, Knife Lake slate. D. Manuel, analyst.
7. Marble. Georgia. W. H. Emerson, analyst. Other marbles may be dolomitic.
8. Soapstone, steatite, South Africa. Van Riesen, analyst.
9. Serpentine, Massachusetts. George Steiger, analyst.
10. Quartzite, South Mountain, Pennsylvania. F. A. Genth, analyst.
Table 2-15 (next pages) lists the ninety most common
minerals which may be principal components, secondary
constituents or occasional inclusions in the composition of
a rock.
11. Garnet rock, Clifton, Arizona. George Steiger, analyst.
To be complete it is also important to mention the
average elemental composition of the ocean water (Table
2-13) as several rock types can be formed from those elements
According to Krynine (1948), "althoughover one hundred and sixty different minerals have so far been identified in sediments, less than twenty mineral species form
well over 99 per cent of the bulk of sedimentary rocks...
Table 2-14
Mineralogical composition of some plutonic rocks
(from Wedepohl, 1969).
Granite
Granodiorite
Quartz
diOrite
Diorite
Gabbro
Av. Igneous
rock(')
30
37
46
53
63
47
35
21
15
22
8
2
3
56
11
16
1
5
13
3
12
5
12
5
1
6-4
3
8
16
2
16
2
% In volume of:
Plagiodase
Quartz
Potassium feldspar
(including micro-perthite)
Amphbole
Biotite
Orthopyroxene
Clinopyroxene
Olie
Magnetite, ilmenite
Apatite
2
2
2
3
5
4
0.6
2-5
0.5
0.5
0.5
0.8
0.6
0.5- 1
n Clarke eta/., 1924.
They are known as "major" or "main" constituents".
Krynine adds: "it is rare indeed that more than five or six
minerals occur in sizeable amounts in any one rock ... In
addition, approximately twenty other minerals, known as
"accessory minerals", occur in very small amounts in sediments, although locally they may be of great importance".
These minerals are listed in Table 2-16, and the mineralogical compositions of the principal sedimentary rocks
are given in Table 2-17.
Reducing the number of fundamental minerals to 10,
still 95% of all the sedimentary rocks can be composed.
The important point is that sedimentary rocks are usually composed of a mixture of at most four minerals or
major contituents, that is, having a content of more than 5
%. The concept of "end-members", or major constituents,
introduced by Krynine (1948) and extended by Pettijohn
(1949) to describe the composition of a rock, is based on
this observation.
The extended form of the concept proposes that the
composition of any sedimentary rock in terms of minerals
can be represented by a point within a triangle or tetrahedron whose apexes correspond to the end-members in a
pure state (Fig. 2-12).
46
Table 2-16
The principal minerals found in sedimentary rocks
(from Krynine, 1948).
Major Constituents
>10%ofrock
I
< 10%ofrock
Accessory Minerals
I<l%ofrock
QUARTZ
Microdine
CLAY
MINERALS
(kaolin-bauxite)
FINEGRAINED
MICAS
(illite, sericite
muscovite)
DETRITAL CHERT
Sodic plagiodase
(albite-oligodase)
Coarse-grained
micas: muscovite
biotite chlorite
Hematite
Limonite
"IRON ORES":
MAGNETITE,
ilmenite,
DETRITAL
LEUCOXENE
STABLE GROUP:
ZIRCON TOURMALINE,
rutile
UNSTABLE GROUP:
APATITE, EPIDOTE
GARNET, HORNBLENDE
kyanite, sillimanite,
staurolite, titanite, zoisite
MICAS: frequently occu
as accessories rather thai
as major constituents
CALCITE
DOLOMITE
ANKERITE
ANATASE;
CHERT and opal
authigenic rutiie
"SECONDARY" QUARTZ
GYPSUM and anhydrite, and leucoxene
halie
Some hydromicas of the illite
sericite-chloriteseries
Phosphates and glauconite
Siderite and some iron ores
Serralog 0 2003
Composition
I Chapter2 I 47
Table 2-15
The 90 most abundant minerals.
Goethie
Hematite
Carbonate
group
Sulfide group
Sulfate
group
Chloride
group
Clay
group
Feldspar
group
Pyroxene
group
Amphbole
PUP
Mica
group
Olivine
group
Garnet
group
EDidote
group
Nepheline
Qroup
2eoIine
group
Zircon
,group
Phosphate
group
I
Alumina (bauxite)
Magnetite
lRAe
ICalcite, aragonite
Dolomite
Ankerite, magnesite
Siderite
Stroniianite
Trona, natron
Pynte
Marcasite
Galena
Chalcopyrite
Gypsum
Anhydrite
Epsomite
Barite, Celestine, Kieserite,
Langbeinite, Polyhalie
Halie
Syhrie, Bischofite, Carnallite, Tachydrite
Kaoiinite, dickite, halloysite
Smectite (montmorillonite), nontronite
beidellie. saponite
llliie
Chlorite (chlinochlore, thuringite)
Chamosite
Glauconite
Potassium Feldspars:orthoclase,
microline. anorthoclase
Alkali plagioclases: albie
otigocIase
andesite
Calcic plagioclases: labradorite
bytormite
anorthite
Enstatite, hypersthene,
diopside, augite
Hornblende,
antophyllie,
tremolie, actinolite,
glaucophane
Muscovite, biotite, lepidoliie, phlogopite
chlorite, pyrophillite
Oliiine. favalite. forsterite,
monticetiiie
Calcic garnets = grossularite, andradiie
IN0 calcic = almandine, pyrope, spessartite
End member
I
Figure 2-12 - Representation of rock composition by means of a
tetrahedron (from Krumbein, 1954).
This method of classification is based on the subdivision of a tetrahedron. Sectioning a triangle (or tetrahedron) into lines of equal percentage or equal ratio will
define areas (or volumes) corresponding to rocks with
well defined names (Fig. 2-13).
A
A
D
B
C
RATIO- CUT
A
A
PERCENTAGE CUT
I
I
IEDidote. zoisite
(aiinite '
(Nepheline
\Leucite
I Mordenite, heulandite, analcime, natrolite
Laumontite
Zircon
Thorite
Apatite
Monazite
D
A
A
A
B
C
A
D
A
RATIO-PERCENTAGE CUT
I
-
Figure 2-13 Division of a tetrahedron (or triangle) into lines of equal
ratio, equal percentage, or a combination of both
(from Krumbein, 1954).
Serralog Q 2003
47
Figures 2-14 and 2-15 illustrate the division of tetrahedra into different rock types.
Table 2-17
Sediment composition (from Krynine, 1948).
Mineral
Percentage
Quartz
31.5
Chalcedony
(chert)
9.0
7.5
Feldspars
Micas
chlorite
and
19.0
Clay minerals
7.5
Carbonates
20.0
3.0
3.0
Iron oxides
All others
Figure 2-14 - The fundamental tetrahedron for classifying sedimentav
rocks. (from Krumbein & Sloss, 1963).
Calcite
Pure limestone
95%
Shaly limestone
Figure 2-16 - Composition of detrital rocks as a function of grain size
(from Blatf et al., 1980).
65%
Marl
35%
calcardolomitic
claystone
the current. Coarse fragments settle rapidly under the
gravity action depending on the flow energy. Finest particles are maintained in suspension and carried far away,
settling down only when the energy decreases.
Claystone
Clay
Anhydrite
I
I
CONGLOMERATE
Calcareoanhydritic
dolostone
I
I
I
I
SHALE
Siltstone
Claystone
Dolomite
Rock F,.agments
Figure 2-15 - Example of subdivision of a tetrahedron for carbonate
classification.
Krynine (1948) evaluated the sediments composition
as consisting of the minerals and mineral families listed in
Table 2-17.
The composition of detrital rocks also depends on the
grain size as illustrated by Figures 2-16 and 2-17.
Pebbles and coarser particles will have generally the
composition of the parent rock, consequently will be composed of several minerals, while fine sands, silts and
clays will be composed of monocrystalline minerals.
This is due to the fact that, during weathering and
transport, rock fragments are broken down and abraded
reducing progressively the size of the particles carried by
40
I
SAND
...-.
Detrilsl carbonates
.,.._
---..-..
Feldspar
Figure 2- 17 - Other way to link the composition of particles to their
size.
Serralog 0 2003
Composition
It is also important to remember that every mineral
may have either a definite or a slightly variable chemical
composition (Table 2-18). This is due either to impurities
or substitution (cations having similar coordination numbers and similar ionic radii tend to substitute for each
other) or due to atoms or ions missing. By reference to
Table 2-19, which lists the ion radii of the most common
cations and anions, it is easy to understand what kind of
ionic substitution can occur in some minerals.
Table 2-18
Elemental composition of 3 orthoclases
(from Deer et a/., 1978).
=2
4 0 3
F%03
FeO
64.76
19.960.08
''
64.68
19.72
0.08
63.68
19.54
0.10
tr.
MO
tr.
tr.
CaO
,40
0.84
0.34
0.50
5.54
8.12
0.54
3.42
11.72
0.18
0.80
15.60
ro
KO
-
Minerals are classified as a function of their defining
anions (Table 2-20). Silicates are the most abundant as
they represent about 90% of the rock forming minerals.
They are classified as a function of their structure that is
based on the arrangement of silica tetrahedron: isolated,
ring, single and double chains, sheet and framework (Fig.
2-18 and Table 2-21 next page).
-Element Name 3ymbc
I
Sodium
I
Figure 2-18 - On the left, crystal structure (cubic) of halite: CI in blue,
Na in red. On the right, structure of the silica S O p quartz: silicon in
orange and oxygen in blue are set inside a qua& crystal form.
If very few minerals can be amorphous most of them
crystallize in systems with planes and folds of symmetry.
6 fundamental crystal systems exist (see further Fig. 2-20
next pages).
z
A
-
0
ion
Ion
barns
:harg
-
Radius E
+ I
0.01
C=2UA
Hydrogen
H
1
1.0080
0.33200
Helium
He
2
4.0026
>0.00700
Lithium
Li
3
6.9410
71.OOOOO
Beryllium
Be
4
9.0122
9.20000
+2
0.35
0.88769
5 10.8110
759.00000
+3
0.23
0.92498
0.99907
B
1.98413
0.99935
+ I
0.68
0.86443
Carbon
C
6 12.0112
0.00340
+4
0.16
Nitrogen
N
7 14.0067
0.00000
.5,-3
0.13.1.7
0.99952
Oxygen
0
8 16.0000
0.00027
-2
1.32
1.ooooo
Fluorine
F
9
18.9984
0.00980
-1
1.33
0.94745
Sodium
Na
I1 22.9898
0.53000
+I
0.97
0.95695
Magnesium
w
12 24.3120
0.06300
+2
0.66
0.98717
Aluminium
41
13 26.981 5
0.23000
+3
0.51
0.96362
Silicon
Si
14 28.0860
0.16000
+4
0.42
0.99694
Phosphorous
P
15 30.9738
0.01900
+5
0.35
0.96856
Sulphur
S
16 32.0640
0.52000
.2,+6
84,0.3
0.99800
Chlorine
CI
1;
33.2000C
-1
1.81
0.95902
35.453(
Potassium
K
I!
39.098:
2.10000
+ I
1.33
0.97191
Calcium
Ca
2(
40.080C
0.43000
+2
0.99
0.99800
Titanium
Ti
2;
47.900C
6.41000
+4
0.68
Vanadium
V
2
50.941;
5.06000 +4,+2
Chromium
Cr
2d
51.996(
3.10000
Manganese
Mn
2
54.938(
13.3000( *4,+i
Iron
Fe
21
55.847(
2.56000
Cobalt
co
2
58.933:
Ni
+3
1.63,0.81
0.91856
0.90299
0.63
0.92315
0.6,O.B
0.91012
+2,+? 1.74,O.P
0.93112
37.50000 +2,+3 1.72.0.6:
0.91629
2f
58.710C
4.54000
+2
Copper
U
:
25
63.5400
3.81000
+2,+'
Zinc
Zn
3(
65.370C
1.10000
Bromine
Br
35 79.9090
6.8000
-1
Nickel
Chloring
49
Table 2-19
Ion charge and ion radius of elements.
Boron
Orthoclase 2 Orthoclase 3
Orthoclase 1
I Chapter 2 1
+2
0.69
0.95384
3.72.098
0.91281
0.74
0.91785
1.96
0.87599
Strontium
Sr
38 87.620C
1.21000
+2
1.12
Zirconium
Zr
$0
91.220(
0.18200
+1
0.79
0.87700
Niobium
Nb
11
92.906C
0.15000
+5,+'
l.69,0.74
0.88261
Molybdenum
Mo
12
95.940C
2.6500
+6.+,
).62.0.7(
0.87555
Sylver
17 107.87Ol
63.8000
+2,+
1.89, 1.2
0.87142
Cadmium
Ag
Cd
18 112.4001 2450.000
+2
0.97
0.85409
Tin
Sn
50 118.6901
0.6300C
+4
0.71
0.84253
Barium
Ba
56 137.340
1.20000
+2
1.34
0.81549
Lanthanum
La
i7 138.9101
8.900
+3
1.02
0.82068
Cerium.
:e
58 140.120(
0.730
+3
1.03
Samarium
im
62 150.350( 5820.000
+3
0.96
0.86738
0.82786
0.82474
Europium
ZU
6:
+3,+2
.95, 1.0!
0.82917
Gadolinium
>d
61 157,2501 19000.000
+3
0.94
0.81399
Tantalum
Ta
7:
180.948(
22.000
+5
0.68
0.80686
Tungsten
W
74 183.850(
18.500
+6
0.62
0.80500
Lead
Pb
32 207.200
0.18800
+2
1.2
0.79151
Thorium
Th
9
232.038
7.40000
Uranium
U
92 238.050f
2.72000
151.9601 4100.000
+4
1.02
+4,+€ .94,0.80
0.77573
0.77294
Serralog 0 2003
49
50
Well logging and Geology
Determination of the rock composition
Traditional method
Table 2-20
Chemical classes of minerals.
aass
Defining anions
Exanples
Native elements
None : no charge ion
Cu,Au, Ag, diamond
Sulfides
Sz8 sirilar anions
write: FeS,
Oxides
0 2
Hemtite: F%O,
Hydroxides
(ow-
Brucite: Mg(OH),
Halies
Carbonates
sulfates
Ct, F, Br, t
Halite: Nacl
coy2
SO;2 8 sinilar anions
Calcite: C
a
m
,
Anhydrite: CaSO,
Phosphates
FO,J 8 sinilar anions
Apatite: Ca5F(F0,),
Silicates
SQ-2
Pyroxene: MgSii
Table 2-21
Structure of silicates.
As previously indicated, the rock composition can be
expressed in two different ways: either by elemental
analysis, or by mineralogical determination.
Elemental and Chemical Composition
This is provided by a chemical analysis of rock samples in the laboratory or by using X-ray diffraction (Fig. 219), alpha spectrometry for thorium (Th) and uranium (U),
or thermal neutron activation, the latter being the most
precise and accurate. The results can be expressed either
in terms of percentages of elements present or, more
often, in terms of oxides of these elements, (Tables 2-8 to
2-12), the oxygen being often strongly bonded to each of
the abundant elements.
N.B. The analysis must be repeated on several samples in order to
have a more precise evaluation of the element percentage of the rock
samples through statistical study.
.
-
Figure 2-19 SEM photograph of an anorthite coated by smectite and
below the spectrographic analysis by EDAX of this mineral.
50
Serralog 0 2003
Next Page
Composition
Isometk :3 equi-kngth
axes, all at right angles
c
u
r
Galena
Garnet
Diamond
Tetragonal:2 equal axes
and a thinl either longer
or shodec all at right
angles
Tetmgonal
Cassiterite
Wulferite
Barite
Olivine
Epidote
1
51
Magnetite
Pyrite
Halite
Fluorite
Zircon
Sulfur
Mamite
Epsomite
Celestite
Anhydrite
Monoclinc :3 unequal
axes, 2 at right angles,
a thinl perpendicular
to one but oblique to
others
Monoclinic
Chapter 2
Leucite
Rutile
Chalcopyrite
Ofthodmmbic: 3 unequal
axes, all at right angles
Olthorhombic
I
*psum
TMinic :3 unequal
axes, all at oblique
angles
Sphene
Hornblende
orthoclase
Anorthoclase
Oligoclase
Polyhalite
Triclinic-
Axinite
Rhodonite
Hexagonal: 3 equal
axes in same plane
intersecting at So',
a kwfth perpendicular
to other three
Calcite
Dolomite
Hematite
Corundum
Hegagonal & rhombohedral
Apate
QdlFiZ
-
Figure 2-20 The different basic crystal systems.
Serrabg Q 2003
51
Previous Page
Well logging and Geology
52
YAGYA
TVPE
MIOH
TEMP
FIRST TO
CRYSTALLIZE
ROCK
TVPE
DUNITE
PERIDOTITE
PVROXENITE
GABBRO
Basan
DlORlTE
cIn
Andwsitw
x
W A R T 2 DlORlTE
D.ClIW
ORANODlORlTE
GRANITE
WsEoYilE
*
LOW
TEMP
LAST TO
CRYSTALLIZE
*
Rhyolllw
0
I
t
WAATZ
@
LOW
YOST
SLOW
WEATt4ERING
STABLE
1: single tetrahedron. 2: single chain. 3: double chain. 4: sheet. 5: 3-D lattice.
Figure 2-21 - The Bowen's reaction series and its interpretation both for rock type formation and for weathering intensity and results
(adapted from Leet et al., 7982).
Mineralogical Composition
In spite of its usefulness - we shall see later the application of tools such as the natural or induced gamma ray
spectroscopy tools for this - the elemental composition of
a rock is not the best expression of rock composition,
either from a sedimentological or from a log analysis
standpoint. In fact, a rock is a mixture of minerals. These
are what gives its petrophysical characteristics, which are
generally those measured by the logging tools - density,
resistivity, sonic travel time, compressibility, etc. These
properties therefore depend on:
- the individual characteristics of each of the constituent minerals forming the rock,
-the relative percentages of each mineral,
- their distribution and bonding.
For this reason, it is preferable to express the composition of a rock in mineralogical terms.
One way is to convert oxide composition into mineral
composition. This is based, for igneous rocks, on the
ClPW (for Cross, Iddings, Pirsson and Washington) norm
or the Niggli's norm. The explanation of the rules for computation of the normative minerals is not the goal of this
booklet. One can say that the norms are based on the
Bowen's reaction series (Fig. 2-21) which admits that the
first to crystallize is olivine, or anorthite, then pyroxenes
and amphiboles or plagioclases poorer in Ca and so on.
The other way to determine the rock type and its min-
52
eral composition is by using different mineral properties color, density (Table 2-22), hardness (Fig. 2-22a), magnetism, radioactivity.. . -, techniques - staining (Fig. 222b), examination of thin sections under microscope
using cross nicols and optical properties of minerals (Figs.
2-23a to 2-23d), scanning electron microscope or SEM
(Figs. 2-23e to 2-23h) and X-ray or electron diffraction...and products - chloridric acid, alizarine, copper nitrate,
benzidine, safranine"y"... -.
Injection of colored fluid (epoxy), which polymerizes in
the pore space with time or temperature, allows the visualization of the pore system (Fig. 2-23c).
I
Figure 2-221, - Calcite crystals are stained in red by alizarine.
Sermlog 0 2003
Composition
I Chapter2 I
Table 2-22
Specific gravity (density) of the principal minerals.
Specific
I
Mineral
gravity
(9/cm3)
Coal
1.O-1.8
Lignite
1.15-1.3
Bituminous coal 1.14-1.40
1.32-1.7
Anthracite
1.47
Natron
1.60
Bischofite
1.592-1.602
Carnallie
1.66
Tachhydrite
1.68
Epsomite
2.07
Sulfur
2.1-2.2
Syhrite
1.9-2.3
Opal
Montmorillonite 2.0-2.1
2.16
Halite
2.13
Trona
2.15-2.24
Kainite
2.2-2.4
Glauconite
2.31-2.33
Gypsum
2.45-2.50
Leucite
2.4-2.63
Kaolinite
25 2 . 6
Kieserite
2.57-2.60
Anorthodase
2.55-2.58
Microcline
2.56-2.58
Orthoclase
2.55-2.65
Nepheline
2.59-2.64
Chalcedony
2.6-2.9
lllie
2.62-2.64
Albite
2.65
Quartz
2.65-2.90
Chlorite
2.71
Calcite
2.74-2.76
Anomie
2.76-3.1
Muscovite
I
Mineral
I Polyhalie
Dolbmite
Langbeinite
Pnkerite
Biotite
Anhydrite
Tourmaline
Apatite
Andalusite
Fluorite
Hornblende
Augite
Enstatite
Silimanite
Olivine
Charnosite
Goethiie
Garnet
Limonite
Pynhotite
Siderite
Sphalerite
Celestite
Chalcopyrite
Rutile
Barite
Zircon
Marcasite
pyrite
Magnetite
Hematite
Galena
Uraninite
Specific
gravy
Wcm 1
2.77-2.78
2.8-2.9
2.83
2.95-3.1
2.8-3.1
2.899-2.985
2.98-3.25
3.15-3.23
3.16-3.20
3.01-3.25
3.05-3.47
3.2-3.4
3.1-3.3
3.23-3.24
3.27-3.37
3.3-3.4
3.3-4.37
3.154.3
3.64.0
3.54.5
3.83-3.88
3.9-4.1
3.95-3.97
4.14.3
4.18-4.25
4.34.6
4.684.70
4.854.90
4.95-5.10
5.168-5.1 8
4.9-5.3
7.4-7.6
9 .O-9.7
-
Figure 2-23a Polish section of a granite.
Feldspar
-
Figure 2-22a Mohs scale of hardness.
Quartz
Figure 2-231, - Thin section of a granite.
Serralog 8 2003
53
Well logging and Geology
54
-
Figure 2-23f Another SEM photograph of "books" of kaolinite.
-
Figure 2-23c Thin section of a dolostone. Porosity is in blue.
m
-
Figure 2-23d Thin section of a compact sandstone. Quartz grains are
interlocked due to pressure solution.
-
Figure 2-239 SEM photograph of fibrous like crystals of illite.
I
Figure 2-23h - SEM photograph of chlorite crystals. On the right bottom comer one can see a quartz crystal overgrowth.
Determination of the rock composition
Well logging approach
-
Figure 2-23e SEM photograph of a "book" of kaolinite.
54
There are many different well logging measurements
which respond to the rock composition. Table 2-23 indicates the relationship that exists between the geological
attributes of rock composition and the different logging
parameters. As a result, an adequate logging suite will
Semlog 0 2003
Composition
I Chapter2 I 55
Table 2-23
Link between geological attributes of rock composition and well logging parameters.
Geological Parameters
Well Logging
Parameters
affected
i
Relevant
Logging Tools
Nalum
make the determination of rock composition much easier.
As rocks are made of beds composed of minerals
themselves composed of molecules and atoms (cf. Fig. 28), log responses are affected by those geological attributes. Two approaches are also possible in using well logs
for determining the rock composition.
Determination of elemental composition
Several logging techniques provide data related to the
elemental composition of the formations crossed by a
well.
Natural aamma rav sDectrometry
Natural gamma ray spectrometry (Fig. 2-24) allows the
measurement of the potassium (K), thorium (Th) and uranium (U) percentage.
Potassium is the seventh most abundant element of
the Earth's crust.
Potassium, thorium and uranium are relatively abundant in acidic igneous rocks, like in syenite, granite, granodiorite, diorite, but are rare in basic and ultra basic rocks
(Table 2-24).
Serralog 8 2003
J
-
Figure 2-24 Natural gamma ray spectrometry allowing the determination of K, Th and U percentage.
Using cross-plot Th vs U (Fig. 2-25) allows the determination of the igneous rock type. This determination
must be confirmed by using other log data (see further).
Potassium is associated to alumino-silicates such as
feldspars, micas, feldspathoids (leucite, nepheline), clay
minerals. Consequently, it is abundant in arkoses, subarkoses, graywackes, claystones. It exists as well in some
55
evaporites such as polyhalite, kainite, langbeinite, carnallite, sylvite (Table 2-25).
Table 2-24
Potassium, thorium and uranium percentage in igneous
rocks
(from U.S.G.S. in Adams & Gasparini, 1970).
I
I
Igneous rocks
Potassium
(%)
Granite G1
average
Granite 0 2
average
Precambrian
granite
Granite'
Svenite
4.28.14
4.55
3.25-3.90
3.74
24
2.74.7
1.95-2.77
2.21
24
-
. Rhyolite'
I
Trachyte'
Granodiorite
average
Diabase
(dolerite)
-..
57
I
I
I
I
I
1I
6.5-80
52
215-40
25.2
14-27
I
I
I
I
Uranium
@pm)
I
11-26
610-1800
1305
6 15
- --
a75
I
2.174.71
2.39
0.49-0.59
0.53
zE
3
I
Thorium
(ppm)
I
Kus U (ignww rocks)
99-125
110.6
1.3-4.6
2.4
I
I
I
1.34.42
4
1.32-3.4
1.99
3.24.6
3.6-6.9
1900-3000
2521
2.5-5
2-7
0.19-2.68
1.98
0.28-1
0.5
0
-
Figure 2-24a Th vs U (bottom lee) and K vs U cross-plot for the
determination of the igneous rock type
(based on data from U.S.G.S. in Adams & Gasparini, 1970).
T h w U (Spnitm)
P
3
Basic rocks
Andesite
average
I
0.46-0.58
0.5-0.6
1.16-2.68
1.39
0.61.7
2.16-3.55
2.39
Peridotite
Dunite
I
i
0-0.07(0.019)
0.02
Gabbro
Gabbro'
Basalt
average
Basalt.
I
3.85-27
2.7-3.9
5.8-10
6.81
2-7
5-10
6.96
I
I
0.018
0.0092
0.84-0.9
0.8-0.9
1.2-2.17
1.73
1-1.7
1.4-2.6
1.94
I
I
Uranium (ppm)
0.0048
0.0037
zE
a from Desbrandes(1982).
I
0
Uranium (ppm)
n
E
LL
Q.
u
-
F@um 2-24b Th vs U and K vs U cross-plot for the determination of
syenite (based on data from U.S.G.S. in A d a m & Gasparini, 1970).
Uranlum (ppm)
56
Thorium is fundamentally an indicator of the detrital
origin of the sediments as its solubility product is very low.
It is associated to some frequent stable minerals (zircon,
monazite) and to some clay minerals that adsorb thorium
ions on their platelets.
Serralog Q 2003
Composition
I Chapter2 I
57
In sedimentary rocks, uranium is essentially associated to uraninite which exists only in reducing environments. Consequently, it is often associated to source
rocks rich in organic matter (Fig. 2-25). However, it exists
also in heavy minerals and in phosphates where U
replaces Ca.
generate gamma rays of which the energy spectrum is
characteristic of the target nucleus (Fig. 2-27).
Figure 2-25 - Uranium and uranium potassium ratio vs percentage of
organic cabon (fmm Beers & Goodman, 1944 on left; from Supernaw
et al., 1978 on right).
I
Adams & Weaver (1958) proposed to use the thoriumuranium ratio as an indicator of the type of sedimentary
rocks (Fig. 2-26).
Table 2-25
Potassium bearing minerals.
Mineral name
FeMsPars
Microcline
orthoclase
Anorthoclase
Plagiodase
McaS
Muscovite
Bitite
Phlogopite
lllite
Glauconite
Fcldrpalhdds
Leucite
Nepheline
Kalsilite
Kaliophilie
olhcr clay minerals
Montrnorilonite
Chlorite
Kaoliite
s-
Polyhalite
Camalie
Kainite
Langbeinite
Glaserite
SyMte
I K content (weight %)
I
16 (ideal) to 10.9
14 (ideal) to 11.8
2.94 to 4.39
-
Figure 2-26 Distinction of the sedimentary rock type through the Th
content and the Th/U ratio (from Adams & Weaver, 1958).
Several logging tools allow the detection and evaluation of several fundamental elements.
Tools based on inelastic collisions of fast and high
energy neutrons (Fig. 2-28) provide information about the
carbon (C) and oxygen (0),and also the yield of some
other elements silicon (Si), iron (Fe), calcium (Ca), sulfur
(S), and magnesium (Mg) (Fig. 2-29).
O U
Inelastic collisions
offast neutrons
9.8 (ideal) 6.2 to 9.37
5.42 to 7.61 (average: 6.6)
5.38 to 9.9 (average: 8.5)
2.7 to 6.65 (average: 5)
T h e m neuron capture
17.9 (ideal)
3 to 10.14
21.29 to 23.51
24.7 (ideal)
0 to 1.32
0 to 0.35 (average: 0.1)
0 to 0.73 (average 0.35)
a,
.c.
E
p
=I
8
13.37
14.07
15.7
18.84
24.7
52.44
Induced aamma rav spectrometry
Figure 2-27 - Elements that can be determined from neutron interactions with nuclei and from natural gamma ray spectrometv.
Several types of interactions of neutrons with nuclei
can exist. They depend on the neutron energy. All of them
Serralog 0 2003
57
58
Well logging and Geology
Table 2-26
Different types of interactions of neutrons with elements with indication of the resulting reaction-product and energy of
the produced gamma rays (from Serra, 1993).
we
Inelastic collision
(,, nl)
e
'S0
Fast reaction
(n, a)
'to
Inelasticcollision
>
ip
w (r: 4,441
* ':c*
(n, n')
w (r: 3.68)
W
1:0*
(r: 6,131
I$
15h
*
15h
:pg
Thermal capture
(n,a)*
p-
:iNe
11%
Delayed thermal activation
7%
>
(n, 1 =$ yiMg'*
*
(n,)
Delayed thermal adivation
=3
38'
p-
3
11%
@,5
* @J*
*
23
100%
3
9.5'
=b
1%
3
23
Thermal capture
>
(n.
j
w (v:
:si
4,93,3,54,2,03
,...)
I ....)
Delayed thermal activation
3
a
*
(n.
3
gh%
b
p-
* $ 3 ~ * p- * ETi
8s
58
\
57,s
(r:
5.10,3.10.2.00)
(y 1,78)
(y 4,11,3.10,
...)
serrabg 82003
Composition
1 ChirpterZ I
59
ments composing the Earth's crust are detected and evaluated. Comparisons with core measurements show the
accuracy of this logging approach (Figs. 2-34to 2-37)and
its validity.
Therefore, it proves their interest for the direct determination of the content of certain elements, and, above
all, for a better analysis of composition in conjunction with
the other measurements, in particular hydrogen index and
photoelectric cross-section.
Figure 2-28 - Inelastic collision of fast and high energy neutron with
nucleus (courtesy of Schlumberger).
Figure 2-30 - Schematic of fast-neutron particle reaction
(courtesy of Schlumberger).
-
Figure 2-29 Inelastic spectra of elemental standards for the RST-A
tool (courtesy of Schlumberger).
Several other tools, based on fast-neutron particle
reaction (Fig. 2-30)or on thermal neutron capture and
activation (Fig. 2-31)allow the evaluation of the percentage of fundamental elements such as Si, Fe, Ca, s, Ti,
Gd, CI and H, and in some case Al and Mg (Fig. 2-32).
Elemental yields are processed from standard spectra for
the different tools, obtained using laboratory measurements. Table 26 lists certain types of reactions between
neutrons and elements and the nature of the result and
the energy of gamma rays emitted during these interactions.
Tools based on elastic collisions of neutrons (Fig. 233) and spectrometry of the induced gamma ray due to
thermal neutron absorption allow mesurement of Si, Fe,
Ca, S, Ti, Gd, CI and H (Fig. 2-32).
One can realize that, thanks to the spectrometry
measurement techniques, all the most abundant ele-
Serrabg 63 2003
-
Figure 2-37 Schematic of thermal neutron capture (first two steps)
and activation (second step on) (courtesy of Schlumberger).
59
60
Well logging and Geology
Neutron-based tools provide a breakdown either in
terms of light elements (H, C, 0 and Si) which have a
slowing effect (seen by the epithermal tools), or in terms
of thermal neutron absorbers (Gd, B, Li, CI, Fe) seen by
the thermal tools.
The litho-density tool provides a measure of the mean
atomic number of the rock components.
Determination of the mineral composition
This determination can be obtained from the elemental composition, assuming certain associations and rules,
or from other logging measurements.
-
Figure 2-35 Example of element percentage evaluation from the RST
tool and comparison with core measurements
(courtesy of Schlumbeger).
Figure 2-32 - Capture spectra of elemental standards for the RST-A
tool (courtesy of Schlumbeger).
1
Figure 2-33 - Schematic of the neutron scattering with a taget nucleus
in elastic collision.
-
Figure 2-36 Example of element percentage evaluation from the GLT
tool and comparison with core measurements. One can observe a
very good agreement between the measurements and the core data
(courtesy of Schlumbeger).
1
Figure 2-34 - Example of element percentage evaluation from the ECS
tool and comparison with core measurements
(courtesy of Schlumbeger).
80
Serralog 0 2003
Composition
I Chapter 2 I 61
Remark
It is highly recommended to never base the mineral
identification on a single cross-plot. Use all the reliable
information to avoid misinterpretation.
Silicon (wt %)
Figure 2-38 - Mineral identification combining Si and K.
-
Figure 2-37 Another example of elemental evaluation obtained with
the GLT tool. Observe the very good agreement between logging and
core measurements in this scientific well.
No point in this zone
EXDlOitatiOn of the elemental ComDoSition
For instance Si, Ca, Fe and S are generally only present in dry rocks. K may be in mud filtrate, if KCI has been
put in the mud, otherwise it is associated to feldspars,
micas and clays. CI is assumed to be present in water and
mud filtrate if evaporites minerals are assumed to be
absent. The sum of the weight percentage of all elements
in the dry rocks is equal to 100%. The model of conversion of element into minerals uses oxides and hydroxyl
content of clay minerals to predict the weight of unmeasured elements (0 and H).
Silicon (wt %)
-
Figure 2-39 Mineral identification combining Si and Ca.
100 = SiOg + CaC03 + CaSO4 + a OH(A1203 + FeqOg) + K20
Cross-plotsof elements can help selecting the mineral
model (Fig. 2-38 to 2-43). 2-plot technique (a third element on the 2 axis) can improve this selection.
Serralag @ 2003
Remark
On these plots several minerals are indicated by an
ellipse. These ellipses reflect the variations in elemental
composition of those minerals.
61
62
Well logging and Geology
No point in
this zone
Silicon (wt %)
Aluminum (wt %)
Figure 2-40a - Mineral identification combining Si and Fe.
-
Figure 2-42 Mineral identification combining A1 and K.
E
2
Y
8
K (Wgt Yo)
Figure 2-40b - Mineral identification combining Th and K.
A1 (wgt %)
-
Figure 2-43 Mineral identification combining Ca and Al.
s
No point in
this zone
r
Y
5
‘5
C
a
Silicon (wt %)
Remark
On these cross-plots only the principal minerals are
represented through their elemental composition. The
mineral composition of a rock can be evaluated by taking
into account the minerals, surrounding the point representing the elemental composition of the rock, common
on all cross-plots.
Another example of exploitation of such plots for rock
type determination is illustrated by Figures 2-44a and 244b.
Figure 2-41 - Mineral identification combining Si and A/.
62
Serralog (B 2003
Next Page
Composition
I
II ChaDter 2 II
63
/
/ -
Figure 2-44a - Igneous rock type identification through association of
Si and K. Each rock is represented by a point. In fact these points correspond to the mean values. Around them must be drawn an ellipse
with elongation following the amws.
I
x
16 18
14
12
10
-1
-
Figure 2-45 - In the leff track is reproduced the determinationof the
mineral composition of the formations crossed by this scientific well.
This determination is based on the elements reproduced on the other
tracks.
8 6 4 -
2 0-
sro,
.--.
I
Figum 2-44b - Determination of volcanic rocks using silica associated
to other oxides (from Dalx 1933).
Exploitation of the elemental composition in sedimentary rocks is illustrated by Figure 2-45. A model with 12
minerals has been used for quantitative computation by
the ELAN program of Schlumberger.
Another example 'of mineral percentage evaluation
from elemental analysis is provided by Figure 2-46.
Serralog Q 2003
In Figure 2-47, a quantitative interpretationof standard
logs and elemental analysis provided by natural and
induced gamma ray (spectrometry measurement) allows
the determinationof the mineral nature and percentage of
grains and crystals, the evaluation of the matrix material
(clay minerals) and the determination of the cement
nature in this detrital siliciclastic formation.
In Figure 2-49 are reproduced firstly the results of elemental analysis, secondly the results of quantitative interpretation mixing standard log data and elemental analysis. The mineral percentage evaluation, based on a 11mineral model, is displayed with the porosity computation.
All these examples show the interest of this elemental
analysis to better determine the lithology and the mineralogy of the formations crossed by a well.
63
Previous Page
Well logging and Geology
64
Other logging measurements can be used for mineral
or rock type identification.
Measurements involvina interactions of photons
gamma with electrons
The first type of such interaction corresponds to measurements linked to photons gamma colliding electrons. As
one knows, depending on the photon energy the photons
gamma either are deflected by their encounters and scatter with a reduced energy (Fig. 2-48), or transfer all their
energy to the electrons in the form of kinetic energy, the
electrons being ejected from their atoms (Fig. 2-50).
These interactions correspond to two types of measurements which are respectively: the electron density measurement, by Compton scattering, and the photoelectric
index measurement by absorption of photons gamma.
_I
Figure 2-46 - Mineral percentage determination from elemental composition (courtesy of Schlumberger).
The electron density is linked to the bulk density by the
following equation:
with :
pe = electron density
pb = bulk density (g/cm3)
Z = atomic number
A = atomic mass
N = Avogadro's number (6.02~1023).
-
J
-
Figure 2-48 Compton scattering of a photon gamma (or gamma ray)
with an electon. The gamma ray is deflected with a reduced energy
and the electron recoils
(courtesy of Schlumberger).
Figure 2-47 Example of a quantitative interpretation including elemental analysis coming from spectrometty measurement of natural and
induced gamma ray (courtesy of Schlumberger).
64
Serralog Q 2003
Composition I Chapter 2
65
.-
Figure 2-49 - Another example of formation evaluation integrating data provided by spectrometry of natural and induced gamma ray. The elemental
analysis results are on the left with comparison with core data. Mineral percentage and porosity are reproduced on the right
(courtesy of Schlumberger).
The attenuation of an incident gamma ray flux, Qo, by
photoelectric absorption alone can be written as:
where E is the gamma ray energy in keV and Z is the
atomic number.
where na is the number of atomes per cm3, x the attenuation depth and z the cross section for photoelectric
absorption, expressed in barndatom, z is given by the following relationship:
Serralog 0 2003
By analogy with Compton scattering, the counts N
seen by a detector at a distance L from a source radiating
No gamma ray are:
N = N0 e - n a T L
65
66
Well logging and Geology
their relative percentages in the zone of investigation on
the one hand, but also on the nature and percentages of
the fluids occupying the pore spaces on the other hand.
So, for a mixture of i atoms one can write:
where:
ne is the electronic density number
nei the electronic density number of the atom i
Vi the volumetric fraction of atom i
Pei the photoelectric index of atom i.
In general the volumetric photoelectric absorption
index is used. It is given by the following equation:
U = Pe pe
Figure 2-50 - Absorption of a photon gamma (or gamma fay) by an
electron that is ejected from its atom (courtesy of Schlumberger).
The photoelectric absorption index Pe is proportional to
the "average cross-section by electron", z /Z
The density measurement can be used for rock identification. Table 2-27 lists the mean densities for igneous
rocks. Refer as well to Tables 2-28and 2-29for the bulk
density of the most abundant minerals composing rocks.
Table 2-27
Mean densities of igneous rocks
(from Daly, 1933).
Pe- l Z
-K Z
K is a constant coefficient characteristic of the energy E:
Rock
Plutonic rocks
155
syenite
24
21
13
Diorite
11
Gabbro
27
40
3
15
8
12
Peridotite (fresh)
Dunite
pyroxenite
z and Ki are respectively the cross-section and the coefficient at energy Ei.
Finally, as the energy dependences upon z and K cancel out, Pe does not depend on the energy. It is well
approximated by :
Pe is expressed in barns/electron.
As we have seen, rocks are made up of a mixture of
minerals. Consequently, it is the physical properties of
these minerals, and ultimately the atomic properties of
their constituent elements which influence the log
responses. In fact, the responses are a function of the
characteristics of each mineral present in the rock and of
06
11
Norite
Diabase ( k h )
z=Zzi andK=ZKi
I
Numberof
sample3
Granite
Granodiorite
Quertz diorite
Practically the absorption is made over a range of
energies and :
I
Anorthosite
Meall
density
Range of
density
2.667
2.716
2.757
2.806
2.839
2.984
2.976
2.%5
3.234
3.277
3.211
2.734
2.5 16-2.809
2.668-2.785
2.630-2.899
2.680-2.960
2.721-2.W
2.720-3.020
2.850-3.120
2.804-3.110
3.152-3.276
3.204-3.314
3.100-3.318
2.640-2.920
2.370
2.450
2.474
2.550
2.772
2.330-2.413
2.435-2.467
2.400-2.573
2.520-2.580
2.704-2.851
Natural glasses
Rhyolite obsidian
Trachyte obsidian
Andesite glass
Leuoite tephrite glass
Basalt glass
15
3
3
2
-
The introduction of tools able to measure the slowness
of compressional and shear waves allows the determination of the rock elastic properties which can be exploited
to determine the rock type. For instance, a cross-plot, proposed by Pickett (1963),associating compressional and
shear slownesses (Fig. 2-51)allows the recognition of the
main sedimentary rock types.
serralog Q 2003
Composition
I Chapter 2 I
67
160
-
Figure 2-51 Cross-plot of compressional and shear slownesses for
identification of the lithology (from Pickett, 1963).
The velocity (or transit time) alone can be used for rock
identification as illustrated by Figure 2-52.Combined with
other log data, it allows the recognition of evaporite minerals (cf. Figs. 2-61 to 2-63).For instance, combination of
density and sonic travel time allows the determination of
many non-porous rock types such as igneous rocks (Fig.
2-53)and evaporites (cf. Fig. 2-61).Don't forget that non
porous rocks, igneous, metamorphic and evaporites, are
also characterized by a very high resistivity generally without change of value between micro- and macro-devices.
The logging parameters of the most abundant minerals are listed in Table 2-28.
At
Figure 2-53 - Cross-plot of density vs sonic travel time for igneous
rocks.
that compose them is better achieved from cross-plot
analysis.
Cross-plot interpretation
The lithology can be determined by a "manual" or
"visual" interpretation, which is fundamentally based on
cross-plot analysis hereafter explained.
Cross-plot definition
A cross-plot is a graphical representation of one
parameter, plotted on the X axis, versus another, plotted
on the Y axis, from a sample at a given depth. This represents a 2 dimensional space. Cross-plots are convenient
to evaluate any existing relationship between two parameters. Those parameters must correspond to the same
volume at the same depth.
Figure 2-52 - Compressional sound velocity of the major igneous rocks
(from Christensen, 7965).
The logging parameters of some accessory minerals
are listed in Table 2-29and the logging parameters of the
micas and clay minerals in Table 2-30.
Logging parameters of major igneous and metamorphic rocks are listed in Tables 2-31 and 2-32.
A third dimension can be added by plotting on an axis,
perpendicular to the plane made by the two first, a third
parameter. In that case the plot is named a Z-plot. One
can plot on the 2 axis either the frequency or the average
value of a third parameter. A fourth dimension can possibly be added, if a color graphic plotter or screen is available, by superposing a fourth parameter using a color
scale which is a function of this parameter value.
The determination of the rock type and of the minerals
Serralog 0 2003
67
Table 2-28
Logging parameters of the most abundant minerals.
08
Serralog Q 2003
Chapter 2
Composition
69
Table 2-29
Logging parameters of accessory minerals, coals and fluids.
a May contain up to 2 500 ppm of thorium
Table 2-30
Logging parameters of the principal phyllrte minerals.
Composition
Name
Group
Kaolinitc
Dickite
Hdloysite
Kao lioitc
Muacovdc
AI,[S,O,J(OH),
~
,
(A)
F
PA)-
7.2
i
0-06
AI,[Si,O,d(OH),4H2O
10 1
I
iSib5~~~0,,,1~0~~9'F,c2
KCfi,[,S&@JO,,(OH),
,
70
with 1.0 < x.< 1 5
-(K,Na,Ca),JJ(Fe
FC.M&~
-*
i
A1
Montmorillonite
I
Chlorltc
'
(MsFc,Al),
ThkinBtc
Clmochlorc
-Chamositc
I(~,A!)&.&OH)La
Rich U-I iron
Rich m magnesium
FeleA<d(MeFc,&4,
__
~
-4.30
.
(WC)
U
(trlrm3)
16-1.8
15-1.9
-1.5
4-5
4-5
3-3.5
13-16
13 5-16
12-12.7
GR
(API)
80-130
80-130
-100
' 15-25
20-37
2 6-35
7-12
15-23
130-235
-
-2 85
20-35
24-40
5 8-7.3
IOz18
18-26
155-210
2.0-2.2
50+
60+
1.6-2.3
3 5-5
11-15
- 140-200
_
-37
-5 2
-6.3
-17
-25
8-11
-2G35
--27-40
12-25
16-40
3 5 3 3 ' 2.6-2 8-6.0
L
(p.~)
40-46
42-50+
33-38
(g/crn3)
-2.60
-2.60
-2.08
Pe
-_
*s&L
(p.~)
-34
-35
(CU)
I
( 1 0 CaNa)o,(AI,M_&Fc>+
I(%W800,J(OH), nH,O
hton1tc
Bcidellilc
Nontronite
Chlorite
*me
- Pb
~
1- -
[B,.,,AI,,,4O,oI(OH)4
Smcctite
K
~ ~ 4 1 ~ . 0 i o l ~ 7 ~2 ~ ;~ -0.20
~
IllllC
Glaucon-itc
Sleet
1' 0.5-1.5
9.717 2
I
(2&l&LdOH)s
0-03)
'
-276
'
-3.1
-2 8
-3.19
*
45-50+
47-56+
'
so+
50-250
L
,
. *
.
14.
:
_-
- 4 -
-0 13
-
-
-
5Oi
44-50+
so+
18-c8-_ 5-13.
A -
25-32
8-10
-
2
For more detail refer to document ELEMENT, MINERAL, ROCK CATALOG (SERRA. 1990)
type, for instance carbonate formations must be separated from siliciclastic detrital formations, formations of difSerrabg 0 2003
ferent ages as well. This helps the analyst to avoid any
confusion,
09
70
Well logging and Geology
Table 2-31
Log responses of the principal igneous rocks ( completed from Desbrandes, 1982).
ROCb
PMonlc
Syenite
Granite
Diorite
Gabbro
Peridotite
Dunite
Vokanic
Rhyolite
Trachyte
Andesitc
Basalt
Th
PI
Pe
4%
c
@pm)
(g/Cm?
(we)
@.uJ
(C.U.)
3.59
2.86
3.93
4.47
3.66
3.32
6
3
8
9
9
17
I300
14-80
8.5
2.7-3.9
0.05
0.01
2500
1.3-6.9
2
0.8-0.9
0.01
0.01
2.7-4.5
2.7-6.14
1.1
0.46-0.6
6-15
9-25
1.9-2
2-7
2.5-5
2-7
0.8-2
1-1.7
2-4
5.7
1.7-2.9
0.6-1.7
2.6-2.8
25000
107-234 2.56-2.68
2.8-2.9
68
2.9-3.1
25-32
2.3-3.3
3.5
3.0-3.7
0.5
0.2
0.02
2.3-2.5
2.4-2.5
2.4-2.6
2.7-2.9
76-164
143-247
4 1-70
26-69
Vpl
vp4
(1o3dS)
(~dm/~)
6.7
6.1-6.5
6.6
6.7-7.1
3.9-8.1
7.8-8.3
6.6
5.8-6.4
6.4
6.6-7.0
3.6-7.2
7.5-8.2
12
21
3.2-7.0
5.36
12
5.4-6.2
5.2-6.0
Table 2-32
Log responses of the principal metamorphic rocks (from Desbrandes, 1982).
Pr
(g/cm3)
____-_--
Igneous
-
_-7-21
-
..
3-4
5-8
1
5-10
I_____.-___._._-.
I-I.
-
_
I
_
_
_
-
-____..I._
13-14
5.9-6.0
-..--I_-_--
Calcitic marble
Dolomitic marble
__-.
-
0.9
0.5
~
2.62-2.70
2.62-2.70
40-60
196-252
-
---
I
_
I
_
_
1.8
2.3
-
____-_
2.7-2.8
5-5.2
ll_l_-_-_-__-___
_.__
6.2-6.4
_
I
_
-2
-2
4.2
5.5
.I___
_ l
6.0-6.1
6.0-6.1
i ^ I ___
_
_
_
I
_
--_I
_
_I
_
_
I
I
-
Carbonates
.__l__.._.l_____...______I-
'
-
~
_
.
I
_._..I_-
I
_
_
3
c Vp 1 kbar Vp4 kbars
(c.u.) (103m/s)
(103m/s)
4-7
5.0-5.2
2.5-3 128-260 2.75-2.85
_
.._I_.-.
_
I
_
4.1
3
18-19
6.2-6.5
6.4-7.0
2.9-3.1
3-4 -172-216
___-___I---.---I_.
3.7
6-10 ~_.___.____.I__
17
5.2
3-3.5 108-156 2.75-2.8 - __-_
__
.I__
___^I
Micaschist
8-11
I-"--------2-5=Schist
.I-__.
..
..__Phyllites
5-9
_________._.I____.I
Sand8
_._I-_
..
Whitequartzite
2-4
l__l_._._.___._l
._
Red quartzite
37-41
- _-
-
2.65-2.90- 4-4.5
_I_-._---"
-____I.-._
I
_
_
Shales
4rr.n.
(p.u)
__ __-
3-5
Gneiss
Pe
(We)
___ 10-30
--2.8-2.9
.____I
10-30
A
4.5-4.1
,--..I-.I_
0_7
2.5 '-6-6.5
6.0-6.2
6.9-7.1
6.2-6.3
6.2-6.3
I
_
.-.____I___
-.
has meant that
I
- A minimum thickness for zones is needed to base the
interpretation on sufficient statistical data. This also
decreases to a reasonable level the number of cross-plots
to analyze,
- Never base your interpretation on one type of crossplot (i.e. density-neutron), as too often carried out,
- Use all the possibilities of the Z-plot technique. By
adding a third parameter one can validate the data or
determine the influence of typical minerals.
. For instance CALI (or Cal) on the Z-axis enables to
validate the data or, in other terms, not take into account
samples which are affected by cave or rough hole; allows
the recognition of porous and permeable zones (mudcake deposits).
. For instance GR, POTA (K) and THOR (Th) on the Zaxis enable to recognize the influence of radioactive minerals (i.e. clays, orthoclase, mica or zircon).
70
. For instance SP or SSP, if sufficient contrasts exist
between the mud and the formation water salinity, helps
to determine where are the reservoirs, even if they are
radioactive.
If you plot data corresponding to a systematic sampling (for instance samples each half a foot) do not forget
that some of the data do not represent any geological
reality due to the lack of vertical resolution of standard
logs. Eliminate data corresponding to an isolated value
(frequency 1 or 2). Those data correspond either to transition from one bed to another or to thin beds.
Consequently, it is highly recommended to adopt a segmentation of log data in electro-beds before any crossplot (Fig. 2-54).
Serralog Q 2003
Composition [Chapter 2
- __
I
71
J
Figure 2-54 - (a) On the left standard presentation of logs. (b) On the right presentation of logs after segmentation in electro-beds. This segmentation takes into account the vertical resolution of each log measurement, the statistics and the uncertainties on the measurements.
Cross-plot analvsis
Several cross-plots are essential for the determination
of the principal minerals composing a rock. The use of the
Z-plot technique is even preferred as it introduces a third
parameter allowing us to avoid misinterpretation. Never
try to determine the mineralogical composition from a single cross-plot or a reduce set of log data.
For lithological and mineralogical determination, the
most useful cross-plots are listed hereafter:
- Density versus (vs) neutron with GR, Th, K, Pe, SP,
eatt, Cal and frequency on the Z axis. This plot is interesting and popular because it gives a good evaluation of
the porosity if the rock is clean (without clay or shale),
even if one does not know perfectly the lithology.
- Pe vs K with Th, rma, SP, eatt, Cal on the Z axis.
- Pma vs Uma with Th, K, GR, eatt, SP, Cal on the Z
axis.
- Th vs K with SP, Pma, Pe, eatt, Cal on the Z axis.
- At vs PHlN with SP, Pe, Th, K, eatt and Cal on the Z
axis.
- AtC VS AtS.
These plots are reproduced hereafter with the position
of the main minerals. Some of them are represented by
an ellipse. This ellipse reflects the variations in elemental
composition of those minerals (Fig. 2-55 to 2-59).
Many other cross-plots can be used. Remember that
Z-plots are very useful for a more precise and accurate
determination of the rocks and their mineral constituents.
Semlog 8 2003
-
Figure 2-55 Density vs neutron porosity croos-plot. This plot is very
popular as it gives an idea of the porosity even if the lithology is not
known correctly (courtesy of Schlumberger).
This will be illustrated by the determination of rocks
without porosity for instance evaporites (Fig. 2-60 to 2-64)
or igneous rocks, which are characterized by the absence
of effective porosity (Figs. 2-65 to 2-68). The interpretation of cross-plots corresponding to porous rocks requires
sometimes more data as the fluid filling the pore space
influences the log responses.
71
1.1
110
0.8
-
Figure 2-58a M vs N cross-plot for mineral identification
(courtesy of Schlumberger).
40
Figure 2-56 - Sonic slowness vs neutron porosity cross-plot.
(courtesy of Schlumberger).
If geochemical logging data exist, cross-plots of Al vs Si,
Ca vs Si, Fe vs Si, K vs Si, Fe vs Ca, Al vs K, Al vs Fe will
be very useful for a more precise determination of the
mineral composition.
T3
B
Q
Q,
a
K (wgt Yo)
Figure 2-57 - Pe vs Potassium (K) cross-plot. Several minerals are represented by ellipses to reflect their variations in elemental composition.
72
Figure 2-58b - Matrix Identification (MID) plot for identification of the
minerals composing a rock
(courtesy of Schlumberger).
Serralog 0 2003
Next Page
Composition
I Chapter2 I
73
Table 2-33
Tests for identification of rocks from resistivity, gamma
ray spectrometry, photoelectric index and matrix density
measurements.
A
n
6
.
INPUT LOG DATA
0
N
-
Figure 2-59 pma vs Uma cross-plot for mineral identification.
From a logging point of view, rock types can be divided into two main categories:
- rocks without interconnected porosity, and therefore incapable of constituting a reservoir. These always
exhibit very high resistivity. Igneous and metamorphic
rocks and evaporites belong to this category.
- Rocks with interconnected porosity which can
constitute potential reservoirs. They can sometimes
exhibit high resistivity due either to low porosity, or to a
high content of non-conductive fluids (hydrocarbons, bitumen or fresh water), or they can exhibit variable conductivity because of a proportion of the pore space is occupied by a conductive fluid (water with variable concentrations of salts). Sandstones, limestones, dolostones
belong to this category. Shales belong also to this category even if the porosity is not effective. Table 2-33 summarizes the tests that can be done to identify the rocks.
The preliminary remarks suggest the use of resistivity
values to break up the formations in these two categories.
1 - Formations of verv hiah resistivity (> 600 ohm-m)
As previously indicated, these formations can correspond to one of the following rock types:
- igneous rocks, either plutonic or volcanic, which have
not been altered or fractured;
- metamorphic rocks which have not been fractured or
altered;
- sedimentary rock of zero porosity (chemical rocks of
the evaporite class: gypsum, anhydrite, halite, polyhalite,
sylvite.. .) or of very low porosity (quartzites, compact and
well cemented carbonates which are not fractured);
- rocks which are porous (reservoirs), but filled with
non-conductive fluids (hydrocarbons, fresh water, bitumen, asphalt).
The selection of one or the other among the above
possibilities requires complementary measurements.
Serrabg Q 2003
1-1 - Separation of Resistivitv Curves
Quite often in sedimentary rocks the resistivity curves
measured by tools with different depths of investigation
(macro-devices: deep and shallow laterologs, deep and
medium inductions, and micro-devices: microlaterolog
and microspherically-focussed logs) show a continuous
separation. This can be attributed to invasion, and hence
will indicate a porous and permeable rock.
1-2 - No seoaration on Resistivitv Curves
When very high resistivity intervals show practically no
separation between the various resistivity curves, one can
reasonably conclude that the rock has no porosity, or at
least no connected porosity. The corresponding interval
can correspond to a non-porous rock.
If the radioactivity is low, and the hydrogen index is
between -2 and 2 P.u., the probability that this rock is sedimentary, be it of chemical origin or strongly cemented or
re-crystallized by diagenesis, is very high, especially if the
surrounding formations are of variable resistivity. If not, it
is more likely to be a metamorphic rock such as marble or
quartzite.
73
Previous Page
74
Well logging and Geology
Weak radioactivity and a high neutron hydrogen index
(> 35 P.u.) indicate a sedimentary rock of chemical origin
which is rich in water molecules within the crystals (gypsum, bischofite, tachydrite, epsomite, kieserite, trona).
The techniques for determining mineralogical composition
will be described later.
Weak radioactivity and a hydrogen index in the range
of 10 to 25 p.u. may indicate a basic volcanic rock such
as andesite or basalt.
Cases of variable but significant radioactivity, and
hydrogen index varying with radioactivity where the beds
are intercalated with non-radioactive beds of zero or high
values of hydrogen index most probably correspond to
sedimentary evaporites with varying amounts of potassium salts. If, on the other hand, the beds are fairly homogeneous and thick, the indications are for an igneous rock
(plutonic or volcanic), or for a foliated metamorphic rock
such as slate or micaschist.
It is important to recognize them. As indicated, generally
this determination starts by the fact that plutonic rocks
never present porosity, and therefore exhibit very high
resistivity, both on micro- and macro-devices, if they are
not fractured or weathered (Fig. 2-60).
Identification of igneous rocks
Plutonic rocks can be encountered either as sills or
dikes or even lacoliths, intercalated in sedimentary rocks,
and sometimes as batholiths constituting basement rocks.
-
Figure 2-62 Cross-plot GR vs density for identification of igneous
rocks (from Desbrandes, 1982).
1
-
Figure 2-60a Typical log responses in granite. The resistivity drops on MSFL are due to fracture influence generating boreholebreakout. The
peaks on Pe and U are due to pyrite influence.
74
Serralog 8 2003
Composition
-
POTA MOR
I Chapter 2 1 75
/WTA-PW
-
Figure 2-606 Several cross-plots corresponding to the previous log intewal showing the log responses in granite.
Some data are affected by weathering and by fractures filled with pyrite (higher Pe values).
Volcanic rocks exist as ashes, lava flows or tuffs.
Sometimes, they can show high porosity, most of the time
not connected (pumice, vesicular basalt). In that case
they are very resistive (Fig.2- 61). If the porosity is connected, due to fracturing or weathering, the resistivity can
be low and the rock can constitute a reservoir.
The radioactivity of the igneous rocks depends on the
Serralog 6 2003
content of radioactive minerals such as potassium
feldspars, feldspathoids, micas or trace minerals such as
zircon, allanite, xenotime or monazite. The potassium,
thorium and uranium contents decrease from acidic to
ultra-basic rocks (Fig. 2-24a). As a result, the radioactivity also decreases from acidic rocks to basic rocks (Table
2-31 and Fig. 2-62).
75
The speed of sonic waves is usually high, above 6000
m/s for plutonic rocks and 5250 m/s for volcanic rocks
without connected porosity.
The density of igneous rocks increases from acidic to
ultra-basic rocks (Fig. 2-63). Works made by Henkel
(1976), based on 30,000 samples from Northern Sweden,
links the density with the silica content (Fig. 2-64), and
with the ratio (Ca + Mg)/(K + Na) (Fig. 2-65). The silica
content decreases from acidic to basic rocks. On the contrary, Ca and Mg increase from acidic to basic rocks and
simultaneously K and Na decrease.
The sonic velocity increases from acidic to ultra-basic
rocks (cf. Fig. 2-52).
The determination of the igneous rocks must be based
on the analysis of the cross-plots illustrated by Figures 252 and 2-53, and Figures 2-62 and 2-63, completed by
the cross-plots of Figures 2-44a, 2-64 and 2-65 if elemental analysis by spectrometry has been obtained.
h
Identification of evaporites
CNL Neutron porosity index (P.u.)(apparent limestone porosity)
With the exception of certain carbonates of chemical
Figure 2-63 - Density vs neutron porosity cross-plot for igneous rocks.
I
Figure 2-61 - Typical log responses in basalt.
origin, the evaporites are characterized above all by high
resistivities of about equal magnitude on all the resistivity
measurements. This is a natural consequence of the fact
that after compaction these rocks have no porosity, and
do not, therefore facilitate invasion, and also that none of
76
the constituent minerals are conductive.
Their radioactivity will depend on the salts present.
Only potassium salts are radioactive, the activity being a
function of the potassium content. When potassium salts
Serralog Q 2003
Composition
I Chapter2 1
77
and will depend partly on the types of potassium salts
present and partly on their volumetric percentages.
Si02
Determination of the tvpe of evaDoritic salt
If the thickness of the salt bed is greater than the vertical resolution of the measuring device, which is the usual
case, it can be easily identified by its log response.
One can assume that, because of the way in which
these formations originate - chemical precipitation from a
brine -the constituent minerals are usually pure, and have
mineralogical, chemical and physical characteristics
which are well-defined and constant. It is only to be
expected then, that their log characteristics are also welldefined and constant. Table 2-28 sets out the characteristics of the more abundant salts and minerals.
It is clear from this table that ambiguities in identifying
an evaporite are practically non-existent given an adequate logging suite and a bed thickness greater than the
vertical resolutions of the tools. In such a case, the nature
of an evaporite salt is defined by the log value attained by
each measurement.
Density (g/cm3)
0
WhlU
1.64 Tuhhydde
Figure 2-64 - Density vs Si02 cross-plot for igneous rocks
2.0
(from Henkel, 1976).
Ca + l2Mp
K + l.UNa
DEVIATION
CAUSEDBY
SERPENTINISATION
2.2
p
O^
.
Fl
2.4
s
XI
2.6
Q
2.8
3.0
0
10
20
30
40
50
K (wgt %)
-
Figure 2-66 Density vs potassium cross-plot for determination of
evaporite minerals.
Density (g/cm3)
Figure 2-65 - Density vs ratio (Ca + 1.2 Mg)/(K + 1.43 Na) cross-plot
for igneous rocks (from Henkel, 1976).
form laminations of varying thickness and frequency
among deposits of non-potassium salts such as gypsum,
halite and anhydrite, the overall activity will be variable
Semlog Q 2003
Thus, if At = 67 f 3 p / f t (from the Borehole
Compensated Sonic, BHC), or Pb = 2.04 f 0.05 g/cm3 and
Pe = 4.5 f 0.3 (from the litho-density measurement), the
immediate conclusion is for halite. However, cross-plots
combining two or three log measurements are recommended for a more certain diagnosis. Figs. 2-66 to 2-70
show such plots, with the positions of the pure salts
marked.
Finally, one can employ the electro-lithofacies method.
The method consists of comparing the electrolithofacies
for each interval of stable response (facies) with the theoretical electro-lithofacies for each evaporitic salt.
77
78
Well logging and Geology
Several real examples of log responses in evaporites
will now be presented. They will demonstrate the validity
of the logging approach for the lithology and mineralogy
determination.
9
Q
tn
3.
W
Figure 2-67 - Density vs sonic travel time cross-plot for identification of
evaporite minerals.
Neutron porosity
Figure 2-69 - Cross-plot of sonic transit time vs neutron porosity for
identification of evaporites
K (wgt %)
Figure 2-68 - Cross-plot of sonic travel time vs potassium for identification of evaporite minerals.
Figure 2-71 reproduces a reduced set of log data
recorded in an evaporitic formation of North Sea essentially composed of anhydrite, halite and polyhalite, the latter being characterized by radioactivity pics. The crossplot (Fig. 2-72) reproduces the real position of sampled
levels of Figure 2-71. This plot confirms the mineral mixture but allows the precise determination of the logging
Darameters of those minerals. Polyhalite has a sonic travel time close to 58 p / f t and a gamma ray close to 220
API. The sonic travel time for halite shoud be 68 ps/ft.
On the previous cross-plots the mineral positions correspond to their theoretical values.
78
9"
Neutron porosity
Figure 2-70 - Cross-plot of density vs neutron porosity for identification
of evaporites.
In Figure 2-73 are reproduced the log responses,
recorded by the Platform Express equipment, and the
lithology determination realized at the well site.
Serralog Q 2003
Composition
The next example corresponds to a similar mineral
association. Based on cross-plot interpretation (Fig. 2-75)
the authors (Ford et a/., 1974)suggest a small shift of the
logging parameters of the three minerals. For polyhalite
the density is equal to the theoretical value, but the SNP
neutron porosity should be increased by 2 P.u.. Its sonic
value is similar to the previous one (58ps/ft).
I
Chapter 2 I 79
1
The sonic value for halite is also close to 68 ps/ft.
A
-
Figure 2-71 Gamma ray, neutron, sonic responses in
evaporite series in North Sea (Serra, 1980).
Figure 2-73 - Example of log responses in evaporites. In the lithology
track is reproduced the rock type determination made by interpretation
of the data provided by the Platform Express equipment
(courtesy of Schlumberger).
Based on these new values, the log interpretation
computes mineral percentages very close to the one evaluated by the geologist (Fig. 2-74).
The next five cross-plots (Figs. 2-76 to 2-80)allow a
more precise determination of the real logging parameters
of other evaporitic minerals.
Figure 2-72 - Cross-plots of some log values (points I to 20) extracted
from log data reproduced in Figure 2-70 (from Serra, 7980).
Serralog Q 2003
79
1TWIEl.N.1
Geologist's evaluation
LTW1CI n
.
1
Log interpretation
SNP Porosity (P.u.)
Figure 2-74
- Comparison of log interpretation with geologist's interpretation (from Ford et al., 1974).
Figure 2-75 - Density vs SNP porosity (plot on the bottom left) and
sonic travel time vs SNP porosity cross-plot (above), corresponding to
the previous log data. They suggest a small shift of the theoretical
parameters of evaporite minerals (from Ford et al., 1974).
.80
Figure 2-76 - Cross-plot density vs neutron porosity corresponding to
real log data. Accumulation of points close to typical minerals and
trends between minerals suggest the variations in composition of this
evaporitic formation composed of anhydrite, halite, tachydrite, camallite
with a marked trend toward sylvite.
SNP porosity (P.u.)
80
They illustrate also the mineral mixtures (thin intercalated layers of different mineral composition in this evaporitic environment) and the trends toward the principal
minerals. These plots, based on a standard sampling (one
Serralog 0 2003
Composition
I Chapter2 I 81
value each half a foot), reflect as well the lack of vertical
resolution of the logging tools, many values corresponding
to artefacts linked to transition values from one bed to the
following one.
Pe
( bfej
Figure 2-79 - Pe vs density cross-plot corresponding to real log data
allowing the statistical determination of the mineral logging parameters.
Figure 2-77 - Potassium vs neutron porosity cross-plot corresponding
to the same real log data. Accumulation of points close to typical minerals and trends between minerals suggest the variations in composition
of this evaporitic formation. Trends from halite to sylvite and
camallite are obvious.
-
1
3
Figure 2-80 - Another example of real log data combining density and
potassium (from Dragoset, 1977).
fi
E
Y
cb
v
The next example (Fig. 2-81) illustrates the log
responses in Prairie Evaporite Formation rich in sylvite,
marked by a very high gamma ray response.
GR
Figure 2-78 - Density vs potassium cross-plot corresponding to the
same real log data. Accumulation of points close to typical minerals
and trends between minerals suggest the variations in composition of
this evaporitic formation. Trends toward sylvite and camallite are
obvious.
Serralog Q 2003
NEUTRON
DENSITY
SONIC
Figure 2-81 - Composite-log of a section of the Prairie Evaporite
Formation of Saskatchewan, Canada, illustrating the log responses in
evaporitic formation rich in sylvite confirmed by the cross-plot of Figure
2-82 (from Raymer & Biggs, 1963).
81
The corresponding density vs neutron-hydrogen index
cross-plot (Fig. 2-82) reflects the sylvite influence.
'1
1
\
Neutron H index (%)
--"
/
0
I
Figure 2-82 - Density vs neutron-hydrogen index cross-plot corresponding to the sampling (points 1 to 20) of the previous data
(from Raymer & Biggs, 1963).
Figure 2-86 (next page). The log data (gamma ray, neutron hydrogen index, density and sonic) were interpreted
by Schlumberger's Global program. Z-plots, shown in
Figure 2-85, were used to statistically define the real logging parameters of the five minerals present in this formation: halite, bischofite, carnallite, kieserite and sylvite.
Good agreement is observed between the results
obtained by solving a system of linear equations and
those obtained by core analysis (thin section and X-ray
fluorescence analysis). This example confirms once more
the validity of the logging approach.
Finally, the log characteristics of each mineral are not
always known exactly, or can vary as a result of small elemental composition variations, or even calibrations errors
or borehole or mud effects. For this reason, several precautions are necessary: proceed to a prior check of log
quality and calibration accuracy, exact definition of the
mineralogical model and verification of the log characteristics of each mineral. For all these reasons, the actual
values do not correspond exactly to the theoretical values.
Of course, this must be taken into account, and crossplots should be used to define them more precisely.
The following example, from Crain and Anderson
(1966) on the same Prairie Evaporite Formation, compares log interpretation results with core analysis results.
One can observe a relatively good agreement if one takes
into account the lack of resolution of logging data (Fig. 283).
Figure 2-83 - Log interpretation results and core data comparison on
the Prairie Evaporite Formation
(from Crain & Anderson, 1966).
The next example (see Fig. 2-84 next page) illustrates
the logging responses in an evaporitic formation rich in
tachydrite, carnallite and halite with some clay intercalations. The logging parameters of tachydrite are well determined: the sonic travel time # 91 ps/ft, the density # 1.66
g/cm3 and the CNL neutron index between 68 and 58 p.u.
and the gamma ray close to 5 API.
f
-?2
f
Neutron porosity (P.u.)
Figure 2-85 - Z-plots corresponding to the previous log example (16301775 m). Positions of carnallite, halite and bischofite are indicated on
each plot and confirmed by the Z values
(from Haile & Blunden, 1984).
Another example from the Zechstein formation of
Netherlands (Haile and Blundein, 1984) is illustrated by
82
Serralog 0 2003
Composition
I Chapter 2 I
83
GR
Figure 2-84 - Composite log of an evaporitic formation essentially composed of halite, carnallite and tachydrite with some intercalation of clays.
SGR (API)
Figure 2-86 - Another example of log responses in an evaporitic formation from Zechstein of Netherlands. Comparison of core description and
Global processing results. A good agreement is observed between the two methods (from Haile & Blunden, 1984).
Serralog 0 2003
a3
Next Page
J
Figure 2-86c - Another example of Global interpretation results with the illustration of the log measurements and their manual interpretation;
Compare the log responses with the theoretical values of the evaporitic rock types. All of the evaporitic rocks show a very high resistivity
(courtesy of Schlumberger).
a4
Serralog 0 2003
Previous Page
Composition
Identification of metamorphic rocks
As we have already observed, these rocks have no
interconnected porosity other than that arising later from
fracturing or alteration. As a result, the resistivity is usually high, except in the case of phyllites which are rich in
phyllitic minerals such as mica and chlorite, and because
of twinning, which provides a large specific area and a
high cation exchange capacity (CEC). The presence of
metallic sulfides may also lead to high conductivities.
If an adequate logging suite is available, it is usually
possible to determine the type of metamorphic rock present.
The quartzites usually exhibit moderate radioactivity
(40-60 API). Their neutron hydrogen index is about zero,
their density and photoelectric capture cross-sections
close to those of quartz (Pb = 2.66 g/cm3 and Pe = 1.81
b/e respectively), and the sonic travel time is between 50
and 55 ps/ft. The presence of feldspars, micas or heavy
minerals will "pull" the cross-plot points towards these
minerals proportionally to their content, and this results in
a dispersion of the points.
Marbles are generally characterized by weak radioactivity (10-30 API), and a neutron hydrogen index of almost
zero. The density and sonic velocity increase, and the
capture cross-section decreases with increasing levels of
dolomite in the marble (Pb 2.72 to 2.85 g/cm3; At : 50 to
42 ps/ft; Pe 5.08 to 3.14 b/e). The presence of impurities
such as silicates, iron or sulfur will, of course, slightly
modify these parameters.
Like the quartzites, the granulites show moderate
radioactivity (50-55 API), but their density is higher and
varies between 2.73 and 2.93 g/cm3 according to the
amount of pyroxene (hypersthene). The potassium, thorium and uranium contents distinguish these rocks according to the degree of metamorphism to which they have
been subjected. The amphibolites, being rich in thorium,
are noticeably more radioactive (70-80 API). They are
also denser (2.8 to 3.2 g/cm3), and the sonic travel time is
always below 47 ps/ft.
Gneiss has a composition varying between that of
granite and gabbro, depending on the origin. The log
responses are therefore similar. Their foliated texture
affects both the resistivity and the sonic travel time and
acoustic wave attenuation. The latter is the result of
anisotropy, which is itself a function of the mica content
(as shown by Christensen, 1965), and also of the angle
between the plane of foliation and the axis of the measuring device. Thus, the travel time and attenuation are less
when this angle is go", but increases as the angle
approaches zero. This has not been confirmed, however,
by a comparison between measurements and core observations.
Serralog 0 2003
1
Chapter2
1
85
Slates and schists (e.g. micaschists, chloritoschists),
also have a foliated texture, and should therefore exhibit
the same phenomenon. This may help to explain the difference observed by Sanyal eta/. (1980) on the sonic logs
recorded in two neighboring wells over two identical and
perfectly-correlated sequences of schists, micaschists
and quartzite, when there were no notable differences on
the other logs.
The identification of the various types of schists should
be possible by combining the log data, in particular the
density, photoelectric capture cross-section, and radioactivity, or better, the potassium content. Thus, the chloritoschists should show a low potassium content (less than
1 %) and quite a high value of Pe (between 3 and 5 b/e),
while a micaschist will have a potassium content close to
4 or 5 %, with a value of Pe varying between 2.5 and 5 b/e
depending on the type of mica present (muscovite or
biotite).
The theoretical positions of the principal metamorphic
rocks are shown in the cross-plots of Fig. 2-87 (see next
page).
Once the type of metamorphic rock has been identified, the choice of mineralogical model, and hence of the
response equations and parameters to be used, becomes
a simple matter.
Several scientific wells were drilled in Germany (KTB
project - which stands for Kontinentales Tiefbohrprogramm der Bundesrepublik Deutschaland or German
Continental Deep Drilling Program) through metamorphic
rocks. Practically all the logging measurements available
at this period (from 1987 to 1994) were recorded in these
wells. Nearly 400 logging runs were made in KTB-VB
(4000 m deep) and 266 runs in KTB-HB (9101 m deep).
One of the most interesting tool was the GLT
(Geochemical Logging Tool). This tool provided concentrations of silicon, calcium, iron, titanium, gadolinium, sulfur, aluminium, potassium, thorium and uranium.
Anothertool with a semiconductor detector - germanium was also used, which gave a higher sensitivity. Sodium,
magnesium, manganese, chromium and vanadium concentrations were the other data available. The uranium
was used as a graphite indicator as it tends to concentrate
at graphite accumulations. Several rock types were recognized and identified: different gneissic rocks, paragneiss from a sedimentary parent and orthogneiss from an
igneous parent, amphibolites composed mainly of
feldspars and amphiboles, marbles, metagabbro and
metabasites.
The set of data recorded during this project allows a
much better determination of the logging parameters of
these rocks. The reader can refer to the numerous scientific publications covering this study (Emmermann et a/.,
1993; Bram & Draxler, 1993; Grau et a/., 1993; Bram et
a/., 1995).
85
F
3
E
P
M
Neutron porosity (limestone)
Figure 2-87a - Example of cross-plots showing the positioning of certain metamorphic rocks (from Sanyal et al., 1980).
2 - Formations with variable resistivities
These formations generally correspond to reservoir
rocks, or to claystones or shales, or to coal and lignite.
As previously indicated, multi cross-plots and specially Z-plot technique will allow the recognition of the rock
type and the nature of the fundamental minerals composing this rock. This type of analysis will be illustrated by the
following examples.
Identification of detrital rock type
The following example corresponds to a siliciclastic
formation, a succession of sandstones and claystones
with few intercalations of calcareo-dolomitic sandstones.
The sandstones are well characterized by a typical separation of neutron and density curves (in yellow on Figure
2-88 next page), a relatively low thorium content and an
average potassium content close to 2% corresponding to
approximately 20% of potassium feldspar (microcline or
orthoclase). Claystones are marked by high thorium values.
86
N
Figure 2-87b - Example of cross-plots showing the positioning of certain metamorphic rocks (from Sanyal et al., 1980).
Multi cross-plot (fundamentally Z-plot) analysis (cf. Fig.
2-89 following pages) allows the recognition of the principal minerals composing the rocks: quartz and potassium
feldspar, the latter being characterized by the points along
the sandstone line on the neutron-density cross-plot
showing a relatively high potassium and a low thorium
content, and the marked trend towards the orthoclase
position on the Pe vs K and Th vs K cross-plots. The
presence of calcite and dolomite as cement is marked by
thin layers (resistive peaks on resistivity measurements,
density, neutron and sonic curves as well) and low thorium content and low gamma ray. The clay type corresponds to a mixed layer illite-montmorillonite (high thorium content, close to 25 ppm, and potassium content
between 2.5 and 3%).
Another example of dolostone and arkose beds is illustrated by Figures 2-90 and 2-91 (next pages). The orthoclase (or microcline) presence is marked by a high potassium and a low thorium content.
Serralog 0 2003
Composition
DEPTH
Track 1
Track 2
Track 3
I Chapter2 I
87
Track 1
Figure 2-88 - Typical log responses in a siliciclastic formation.
Serralog (0 2003
87
88
Well logging and Geology
_I
Figure 2-89a - Density vs neutron porosity Z-plots corresponding to the example of Figure 2-88. Observe the low thorium content close to the
sandstone line that increases progressively toward the right indicating the position of claystone which present also the highest potassium percentage indicating the main clay type :possibly a mixed layer illite-montmorillonite. The sandstones content between 7.9 and 2.5% of potassium
indicating a subarkose to arkose nature for these sandstones.
F
E
Figure 2-896 - Cross-plots associating potassium and Pe values for the mineral identification.
88
Serralog 0 2003
Composition
I Chapter 2 I
89
Figure 2-89c - Example of Z-plots allowing the determination of the principal minerals and rock parameters. Observe the trend towards the orfhoclase position and the general trend towards the top indicating the increase in clay
Figure 2-89d - Example of Z-plots associating neutron and sonic allowing the determination of rock parameters. Observe the position of shale indicated by high values (red dots) of thorium and potassium.
Serralog 0 2003
89
90
Well logging and Geology
DEPTH
-
Figure 2-90 Example of log response in arkose intercalated in dolostone. The arkose is characterized by high gamma ray value linked to a high
potassium content (reaching approximately 5%), a low thorium (between 2 and 6 ppm), a Pe value close to 2-2. I We, and, sometimes, a curve
separation between neutron and density (neutron on right) in a limestone compatible scale presentation. Dolostone is characterized by low gamma
r a thorium
~
and potassium, a Pe close to 3 and a neutron curve on the left of the density curve. In that example, do not interpret the gamma ray
curve evolution as a clay indicator as too offen done.
-
Figure 2-91a Density vs neutron Z-plots with hquencv, thorium, potassium and Pe values on the Z axis.
Serralog Q 2003
Next Page
Composition
I Chapter2 I
91
WTW
-
Figure 2-91b Pe vs potassium Z-plots with frequency, density, neutron and sonic travel time on the Z axis.
mtm
-
Figure 2-91c Thorium vs potassium Z-plots with frequency, density, neutron and Pe on the Z axis.
Serrabg (b 2003
91
Previous Page
Identification of biochemical rock type
This identification is based on neutron and density
curves recorded in limestone compatible scale.
Limestones are characterized by neutron and density
curves practically superposed in a limestone compatible
scale presentation. Dolostones are characterized by a
small left shift of the neutron curve compared to the density curve. At the same moment thorium and potassium
curves must show very low values. The same type of
cross-plot analysis will confirm the mineralogy. Figure 292 (see next pages) illustrates a typical log response in
carbonate rocks.
DEPTH
The corresponding cross-plots are reproduced in
Figures 2-93 (see next pages). One can observe two
main populations of points close to respectively the limestone and dolostone lines.
Track 1
Figure 2-92 - Typical log response in a carbonate formation with successively from the bottom to the top: limestone, shale layers, dolostone, limestone, shale layers and dolostone at the top. Observe that the limestones towards the top are oil bearing while the dolostones below are water
bearing. This can be interpreted in the following way: at the origin, dolostone beds were limestone beds. An anticline was created in which oil was
accumulated by migration. The same formation outcropped and received rain falls which, by gravity, circulated into the lower part. As less salty it
generated a dolomitization of the lower limestone beds by diagenesis, the upper limestone oil bearing beds being protected by the closure of the
anticline.
92
Serralog Q 2003
NPHI(WI
Figure 2-93a - Density vs neutron Z-plots. Observe the two populations of points, one close to the limestone line, the other close to the dolostone
line, confirmed by the respective Pe values. Both populations show low values in potassium and thorium.
Figure 2-936 - Z-plots density vs Pe with frequency, thorium, neutron and sonic travel time on the Z axis. The marked trends towards lower density
and Pe indicate the porosity increase for limestone and dolostone.
Sernlog Q 2003
93
94
Well logging and Geology
-
Figure 2-93c Pe vs potassium 2-plots with frequency, neutron, thorium and sonic travel time on the 2 axis. Observe the trends towards the day
characterized by high potassium and thorium. The clay mineral is essentially illite.
Figure 2-93d - Sonic vs neutron 2-plots with frequency, thorium, potassium and Pe on the Z axis. The two populations of limestone and dolostone
are well marked specially on the 2-plot with Pe.
94
Serralog Cf) 2003
Composition
Another example in carbonate is illustrated by Figure
2-94 and cross-plots of Figures 2-95. Their interpretation
indicates clearly the limestone, dolostone and anhydrite
layers with sometimes some quartz or feldspar influence.
One can also observe some peaks of Pe indicating probably the presence of glauconitic pellets. This interval may
I Ch
be interpreted as starting with intertidal limestone
deposits composed of grainstones (more porous) and
mudstones (less porous zones), superposed by supratidal
dolostone deposits with anhydrite layers or nodules. The
dolomitic intervals above 8155 ft are oil bearing.
DEPTH
-
Figure 2-94 Example of log responses in carbonate formation with intercalations of anhydrite. Anhydrite beds are well Characterized by high resistivity and density values and low neutron values (close to 0).
Serralog 8 2003
95
Next Page
96
Well logging and Geology
-
Figure 2-95a Density vs neutron Z-plots with fmquencg sonic travel time, gamma ray and Pe on the Z axis. On the last Z-plot limestone and
anhydrite are well characterized by high Pe values.
Figure 2 - 9 s - Sonic travel time vs neutron Z-plots with fmquencg densiiy, gamma ray and Pe on the Z axis. On the last Z-plot limestone and
anhydrite are well characterized by high Pe values.
96
Serralog Q 2003
Previous Page
Composition
Automatic lithology determination:
The LITHO-ROCKLASS Programs
Introduction
As it seems obvious, the recognition of the rock type
requires a multi-cross-plots and histograms analysis. This
type of analysis is time consuming and, in the end,
tedious.
For many geologists and log analysts, interpretation of
log data, in order to give an accurate and reliable lithologic description, is not an easy job. To help them, this work
was done by geologists, experts in log and cross-plot
interpretation. The rock classification based on log
responses, or in other words the building of the electrolithofacies database, was achieved by interpretation of
more than 3,500 cross-plots with their histograms. These
log data came from more than 25 wells, distributed all
around the world and having crossed all types of sedimentary rocks, and on which existed also core analyses.
The goals of these programs are numerous:
- give an automatic, complete and continuous lithologic description of the formations crossed by a well, by comparing, at each sample level, the set of standard log
responses (gamma ray, thorium, uranium, potassium,
density, neutron, photoelectric index, sonic travel time.. .)
recorded in a well to an electro-lithofaciesdata base, previously built, for its attribution;
- provide the first step of a complete facies analysis;
- generate data for:
- sequence analysis
determination of the major geological events
which will be used as time markers for
correlations
mapping purpose
1. lithofacies maps
2. iso-liths maps
3. iso-percentage maps
4. iso-ratio maps ...
generate a more appropriate mineralogical model for
quantitative formation evaluation.
I Chapter2 I 97
The LITHO program uses a pre-built electro-lithofacies database (Fig. 2-96).
The construction of this data base relies on the fact
that a gaussian distribution generally describes quite well
natural sciences (Fig. 2-97). Each rock type is characterized by a gaussian distribution of its elemental and mineral compositionand consequently of its log responses (Fig.
2-98). The combination of several distributions generates
an hyper-ellipsoid, or hyper-volume, in the n dimensional
space, n representing the number of different log measurements used for this lithologic determination.
tv
Mean value
I
Standard Mation
Figure 2-97
- Gaussian distribution describing natural sciences.
-
-
-
.E
cn
C
B
Pe (b/e)
-
Figure 2-98 Combination of log distributions (here density and Pe)
characterizing a typical rock. In two dimensions this generates an
ellipse.
-
Figure 2-96 Schematic representation of the location of the main rock
types on cross-plots.
Serralog (b 2003
ROCKLASS is the name of the new software written
for workstations. It allows the interactive building of an
electro-lithofacies database by direct interpretation of
CrOSS-plots.
97
98
Well logging and Geology
Fluid point
Construction of the database
Several approaches exist.
1. Interpretive approach
This one corresponds to a multi cross-plot and histogram analysis of actual logs. It is based on a good geological knowledge of their interpretation and requires, as
much as possible, core data for validation. The original
electro-lithofacies database of LITHO was realized this
way.
2. Empirical approach
This approach uses clustering techniques similar to
those used in FACIOLOG. It requires a zoning of log data,
after their depth matching, and the elimination of too thin
beds or of ramps that can cre.ate artificial electro-lithofacies without any geological validity. A previous segmentation of logs is recommended.
3. Interactive approach
This one can be used with the help of workstation
(GeoFrame). In that case the ellipses are built directly on
cross-plots reproduced on the screen. To avoid any problem, beds must be selected on the composite log reproduced on the right side of the screen. Then, the ellipses
must be aligned along the equi-lithology lines (Fig. 2-99).
As much as possible, for lithological purpose alone and
for decreasing the number of electro-lithofacies, try to
build ellipses including the total porosity range (Fig. 2100).
-
Figure 2-99 The long axis of the ellipse must be aligned along the
equi-lithology line.
98
-
Figure 2-100 When building the ellipses include the total porosity
range and for that purpose go from -5% and 55% with a coefficient of
correlation equal to -0.9 for RHOB vs PHIN. This will reduce the number of electro-lithofacies.
When building ellipses, try to generate ellipses of
approximately same size to avoid conflicts between big
and small ellipses (Fig.2-99). As all the space must be
covered by hyper-ellipsoids to avoid any interval without
lithologic attribution, some hyper-ellipsoids cover the
same hyper-volume. A point, representing a bed, may be
closer to the center of the small ellipsoid but will be attributed to the bigger one due to the fact that the logic of attribution is based on the ratio of the distance to the center
to the distance from the ellipse external limit to the center
(Fig. 2-99).
-
Figure 2-101 Construct, as much as possible, ellipses of same size to
avoid affectation of the bed to a wrong lithology.
Serralog Q 2003
Composition
Equilithology line
I Chapter2 I 99
breaks due to fractures,
- Ap: a high value may indicate problems on the
density measurement due to rough hole conditions
T2 from CMR to discriminate between non-porous
and porous and permeable rocks, etc.
Threshold values from any log can also be used to orient the search.
-
Coefficient of correlation P 0
Example of LITHO results is reproduced in Figure 2103. One can observe a general good agreement
between this lithological description from logs and the one
made from core analysis. Too thin layers are not well
detected due to the lack of resolution of the logging tools.
Coefficientof correlation = 0.6
LITHOFACIES
CORE DESCRIPTION
Coefficient of correlation = 0.8
SUB
definition
-
Figure 2-102 Select the appropriate coefficient of correlation
between logs, when it exists, to be sure that the ellipse will be comctly
aligned along the equi-lithology line.
Select the appropriate coefficient of correlation
between logs, when it exists. This is fundamental for
measurements sensitive to porosity like density, neutron
and sound slowness.
Lous used in LlTHO-ROCKLASS
All the logs are useful for a reliable and accurate lithologic attribution. But, for the construction of the database,
measurements sensitive fundamentally to lithology and
composition (elemental and mineral) are used. They are:
density
- hydrogen index or neutron-porosity
- photoelectric index
- compressional and shear travel time of sound
- total gamma ray
potassium amount
- thorium and uranium.
-
-
Other measurements can be introduced if they exist:
weight percentage of Si, Ca, Fe, S, Al from GLT,
GST or RST logs
- yields ratios from GST
eatt and tpl from EPT tool.
-
-
The other logs will be used in a preprocessing step to
better orient the search of the appropriate ellipsoid:
- resistivity: very high resistivity may suggest either
an hydrocarbon bearing reservoir or a
- non reservoir rock such as halite, anhydrite, gypsum...
- spontaneous potential: if the SP curve is valid, a
SP deflection may indicate a porous permeable reservoir;
- caliper: mud-cake indicates a reservoir, cave may
indicate a shale or an unconsolidated sand, or borehole
serrakg ca 2003
Figure 2- 103 - Example of LlTHO processing results. Observe the
good agreement between the lithology from logs and the lithological
description from core. Too thin layers are not seen by the logs.
Quantitative interpretation
The determination of the mineralogical composition of
formations is based on the solution of a system of equations.
A detailed and correct description of a reservoir often
requires more parameters than the number of available
measurements. Thus the model usually has to be simplified in order to obtain the fundamental parameters of
porosity, saturation, hydrocarbon type, clay content and
lithology. In so doing it is of interest to estimate the quality of the interpretation and thus the validity of the chosen
model. This can be done by using the GLOBAL or ELAN
programs developed by Schlumberger (Mayer & Sibbit,
1980). GLOBAL may be described as series of processes
using the response equations given by the tools and introducing a degree of uncertainty to the measurements and
to the zoned parameters used, together with pre-defined,
geological and local constraints, as well as calculating an
incoherence function.
The tool response model is expressed as a system of
equations as follows
(ai = fi(X)>
where the ai corresponds to the tools or inputs, 11 is the
99
Well logging and Geology
100
vector of the unknowns or outputs. Thus in the RTGLOB
program we get:
a = (RLLD~RLLS, RILD? RMSFL)
x = (Rtl Rxo, di)
the chances of obtaining correct results are increased by
reducing the number of unknowns per lithology type, as
well as by adapting the equations to the actual models,
eg. type of distribution of clay or porosity. We are then in
a favourable situation, having a model which is overdetermined and adapted to reality and not dependent on
chance.
UIIY
Uncertainty or measurement dispersion is caused by
several factors : the equipment itself (statistics, electronic
noise, etc.), corrections made by the surface recording
system and environmental corrections. This uncertainty is
defined for each tool by a spread which is designated Oi.
Uncertainty in the response function of the tool
fi is also determined and has a spread which is designated 7.
The constraints are expressed as inequalities on the
results and are independent of the measurements
gj (XI 0
’
The constraints may themselves allow a certain tolerance and their spread is designated Zj.
The incoherence funcion is expressed as:
A(a,g) =
- fi(X))2
(gj(X))2
z (ai
___________
+ z----i
Oi2
+ Ti2
j
zt
where gj(X) is the negative element of the function.
Once established, the incoherence function is minimized
by an iterative algorithm using its partial derivatives.
Iteration proceeds until convergence is achieved. Quality
control of the result is achieved by comparing the original
logs with the ones recomputed from GLOBAL or ELAN
results. The incoherence curve is also reproduced and
when it is higher than 1, then the model chosen is unable
to correctly represent the tool responses. A low incoherence is only representativewhen the system is overdetermined. This does not mean that the result is correct but
only that the model satisfies the measurements made.
Discussion
These programs are powerful mathematical tools
adaptable to all types of interpretation models, since data
from any new tool or exterior parameter may be introduced. One just needs to enter the response equations.
Thus, for a recording by a GLT tool one can enter the
measurements made by this tool and thereby get a much
sharper and more exact interpretation (Fig. 2-104).
Comparison with the measurements made on core samples verifies the quality of this interpretation. If the model
introducedcan be guided by evidence derived either from
the geology or the results of processing by a LITHO and
SYNDIP or by a FACIOLOG (Suau & Spurlin, 1982) then
100
-
Figure 2-104 Example of ELAN results (courtesy of Schlumberger).
Examination of the incoherence curve can help to
check the quality of interpretation even though it only represents the coherence between measurements and
model. If incoherence is higher locally this may indicate
that at this point the model is not wholly adequate. One
must then analyze a log or the logs in order to try to establish the reason for discrepancy. Thus, if the recomputed
Serralog 8 2003
Composition
1-
101
density and the Pe value are too low compared with the
measurements made by the tools, this may indicate the
presence of a heavy mineral not included in the model
(eg. pyrite or siderite). The interpretation of the facies and
the environmentwill be helpful in choosing between these
hypotheses. We can introduce an increasing percentage
of the new mineral and first check if this improves the
incoherenceor not. If it does then we stop when the incoherence begins to increase once more.
can appropriate action be taken in the event of a tool malfunction, either by repairing the tool itself, or by re-running
the log with another tool of the same type or by running
one which provides comparable information. This is why
the various quality control and calibration control procedures were reviewed in detail in the first volume on log
acquisition and it may be useful at this stage to refer back
to that section.
Program selection depends on the complexity of the
geological model. It is however, useful to interpret an
interval using several models: a mineralogical model to
determine the percentage of minerals and to give a more
precise idea of the rock composition (Fig. 2-105 next
page); a textural model (Fig. 2-106 next pages) giving an
idea of grain size (sand, silt and clay) to obtain an idea of
sorting and permeability.
The goal of this filtering is to eliminate from the interpretation all the non representativedata linked to bad hole
conditions, lack of resolution of standard logs, measurement and statistical errors. This is achieved through a
segmentation of log data in electrobeds and ramps (Fig.
2-108 next pages) taking into account the tool vertical resolution, the measurement errors provided by service companies, nuclear statistics, depth match problems. This filtering has also the advantage to produce a document
closer to the geological reality as it generates electrobeds
and natural electrosequences or ramps.
Another example of well log interpretation using the
ELAN program of Schlumberger is reproduced in Figure
2-107 (next pages). In this processing and presentation, a
10 mineral model was used and the computation results
displayed as framework minerals (quartz, orthoclase,
albite and volcaniclastic), matrix (clay minerals), cement
(calcite) and accessory mineral (rutile).
SERRA LOG Approach
Previous mathematical methods have limitations as
the number of log measurements control the number of
unknowns that can be computed. In addition, quite often
the mineralogical model introduced in the computation
process is imposed for an interval in which lithology
changes may exist.
The basic logic of the SERRA LOG approach is that
the rock types existing in the Earth's crust are not too
numerous, and are well classified and described in books
of petrography and petrology. This classification is fundamentally based on the principal mineral association that
compose them. These minerals have generally typical log
responses. Consequently, it seems logical to classify the
rocks as a function of the complete set of log responses.
In fact the SERRA LOG approach is a processing using
inversion technique.
- Firstly, the rocks are created by an association of
minerals with defined ranges of percentage.
- Secondly, each rock is converted in log responses as soon as one knows the logging attributes of each
mineral entering into the rock composition and the
response equations for each tool.
The program SQWIZLOG, developed by the SERRA
LOG Company, consists in several steps.
quality control.
The various controls of log quality must be performed
at the wellsite as soon as the logs are recorded. Only then
- Filterina of the loa data
- Depth Matchina and Composite Loq
This stage in the preparation of the data is very important. During recording the tools can be momentarilystuck.
The cable elasticity generates a jump of the tools after
their sticking. This event is not at the same depth for each
measurement provided by the run (cf. Chapter 1, paragraph Depth match).
- Litholoaical determination
For lithological determination of each electro-bed the
program consults a data base that includes:
approximately the 150 most abundant minerals
with all their geological and logging attributes taking into
account their variations in elemental composition,
approximately 650 rock types with all their logging
attributes taking into account their mineral and elemental
composition with their range of percentage variation,
- filtering based on typical log responses of rocks
before consultation of the data base.
A better knowledge of the lithology and mineral
composition allows a more accurate evaluation of the
petrophysical properties of each sedimentation unit composing a reservoir. For instance, a wrong evaluation of the
clay content, based generally on the Gamma Ray
response, can induce an error on porosity and on saturation computation.
-
-
- Constructionof the electro-lithofaciesdata base
It consists to convert in log responses the rocks
described in books of petrography. This requires first to
know either the elemental composition of the rock or its
mineral composition, secondly the tool's response equations for each element or mineral.
For instance, assuming a tetrahedral classification of
biochemical rocks (Fig. 2-109) and that we want to characterize a pure limestone (with more than 95% of calcite),
101
Well logging and Geology
102
otmc
-
Figure 2-705 Example of results obtained with the GLOBAL program using a mineralogical model including quartz, orthoclase, biotite and illite as
clay. Compare with results reprvduced in Figure 2-706
(courtesy of Schlumberger, in Well Evaluation Conference, India, 7983).
102
senrrlog Q 2003
Composition
I Chapter 2 I 103
-
Figure 2-106 Example of results obtained with the GLOBAL program using a tectural model including sand, silt and clay.
Compare with the previous results shown in Figure 2-105
(courtesy of Schlumberger, in Well Evaluation Conference, India, 1983).
serralog 0 2003
103
Well logging and Geology
104
Permeability
Saturation
Total fluid volume
Matrix
Framework
Cements
-
Figure 2-107 Example of results of computation using the Schlumberger’s ELAN program. A 10 mineral model, induding GLT data, was used.
The presentation displays permeability data in track 1 (lefi), then saturation, then fluid content and mineralogical composition in terms of matrix,
framework and cement
(courtesy of Schlumberger).
104
Serralog 8 2003
Next Page
Composition
DEPTH
Track 1
Track2
Track 3
I Chapter 2 I
105
Track 4
Figure 2-108 - Segmentation of log data in electro-beds and natural ramps. Compare with raw logs reproduced in Figure 2-92.
the impurities being chert, dolomite, anhydrite or different
clay types, we will obtain for each log measurement the
following range of variations (cf. Table 2-34).
These different rock types can be positionned on a
cross-plot associating density and photoelectric index
(Fig. 2-110).
Shaly llmatone
Mul
- Consultation of the data base for attribution of the
rock tvDe to each electro-bed
Anhydrlte
The set of log responses corresponding to each electro-bed is compared to the electro-lithofacies data base
for attribution. To speed up and avoid possible conflict,
several thresholdings are applied to logging parameters.
-
Figure 2-109 Classification of biochemical rocks using a tetrahedron.
Serralog 8 2003
105
Previous Page
106
Well logging and Geology
This can be related to porosity and mineral changes.
Most of the time, these rock types correspond to the
same rock class (sandstones, limestones or dolostones)
with the same mineral association but with different mineral percentages. In that case the probability for the rock
class is equal to 1.
-I
Figure 2-110 - Classification of biochemical rocks using a tetrahedron.
Location of each class on a Pe vs m a cross-plot.
These thresholding are applied to resistivity, thorium,
potassium, density, photoelectric index, neutron porosity,
sonic values.
This type of processing is illustrated by the following
figures and cross-plots (2-111 to 2-119 next pages) completed by borehole wall images on the same interval
(Figs. 2-120 and 2-121 next pages).
Table 2-34
Range of variations of well logging responses for a pure
limestone (more than 95% calcite) without porosity.
If the rock classes are different, the probability of the
selected rock class will be less than 1. This probability is
computed as a function of the distance of the point - representing the data set (correspondingto the electrobed) in
the n-dimensional space (n being the number of logs
involved in the processing) - to the line joining the fluid
point to the point corresponding to the mean value of the
rock type, this value being computed assuming no porosity.
A similar approach is used to compute the probability
of the rock types corresponding to the same rock class.
The two probabilities (from 0 to 1) are reproduced as a
function of the depth on the side of the results.
PLATFORM EXPRESS* Lithology display
Parameters
Density
Slowing down length
Diffusion length
Thermal porosity
Epithermal porosity
Comp. transit time
Shear transit time
Photoelectric index
Volume photo. index
Gamma ray
Hydrogen wgt %
Carbon wgt %
Oxygen wgt %
Sodium wgt %
Magnesium wgt %
Aluminum wgt %
Silicon wgt YO
Potassium wgt %
Calcium wgt %
Iron wgt %
Sulphur wgt %
Thorium ppm
Uranium ppm
Symbol
Pb
Ls
Ld
f th
f epi
Dtc
Dts
Pe
Ue
GR
H
C
0
Na
Mg
Al
Si
K
Ca
Fe
S
Th
U
Minimum value
Maximum valul
2.6328
16.6
10.6
0
- 0.1
46.9
86
4.88
12.84
0
0
11.17
47.9
0
0
0
0
0
37.3
0
0
0
0
2.71 75
24.3
13.5
0.7
0.6
54.2
103
5.08
13.83
15.4
0.02
12
48.7
0.1
0.7
0.77
1.49
0.4
40
0.14
1.28
1.25
0.5
- Computation of the rock-type Drobability
As each rock type is represented by an hypervolume,
it may happen that a given set of log data may correspond
to several rock types of the electro-lithofacies data base.
106
In the PLATFORM EXPRESS lithology display, an
approach based on log segmentation is applied. Five log
data are used : density, neutron and photoelectric indices,
gamma ray and spontaneous potential. The three first log
data are converted in two new data: pmaand Uma.
The two
other data (GR and SP) are used to compute the clay
volume V,,. In fact, only three data are introduced. The
color change depends on Vcl and the position of each
point on a pma-U,,,,cross-plot. The colors assigned to the
plot are derived from a color cube with corners of yellow,
green, cyan and white on one face, which represents zero
volume of clay, and corners of red, black, blue and
magenta on the opposite face, which represents 100%
volume of clay. The colors on each V,, slice of the cube
show a gradation across the slice, and all slices show a
gradation of colors from V,, = 0% slice to V,, = 100% slice
(Fig. 2-124a & 2-124b next pages). Anhydrite is reproduced by a special blue (violet) added to the face of the cube
with less than 5% of clay. Halite plots in the yellow-white
region.
This lithology determination is based on an interpretation of cross-plots and clay indicators. As obvious if log
responses are affected by environmental factors (caves,
barite in the mud, etc.), by the bed thickness (surrounding
bed influence), or by special minerals (feldspars, micas,
pyrite, etc.), the lithology display will not represent the
geological reality. Other clay indicators must be used (i.e.
thorium and potassium) and environmental corrections
applied.
Serralog Q 2003
Composition
Figure 2-111
I Chapter 2 I 107
- Typical log responses in a detrital deposit.
-
Figure 2-112 The previous log data have been segmented in electmbeds and ramps.
Serralog Q 2003
107
108
Well logging and Geology
Figure 2-113 - Cross-plot associating neutron and density with on the Z-axis thorium. Observe the trend towards dolomite indicating the porosity
decrease due to dolomitic cement (indicated by the arrow).
I
Figure 2-114 - Cross-plot associating neutron and sonic with thorium on the Z-axis.
Serralog 0 2003
Composition
Figure 2-1 15 - Cross-plot associating potassium and photoelectric
index Pe with thorium on the Z-axis. Observe the trend towards position of orthoclase.
I Chapter 2 1 109
I
3
-
Figure 2-116 Cross-plot associating potassium and thorium with Pe
on the Z-axis.
3
Figure 2-117 - Interpretation in tern of lithology of the previous set of log data reproduced on the left.
Serralog d 2003
110
Well logging and Geology
0
Figure 2-118 - Interpretation of the previous data in terms of mineral percentages with on the right side the porosities from sonic (Wyllie equation),
and density
1
-
Figure 2-119 Borehole images on the same interval than the previous example and the results of their analysis in terms of more resistive and conductive isolated events compared to the resistivity background. These events comspond to microconglomerates. These pebbles are essentially
composed of chert (opal) as confirmed by the Uma vs Rhoma cross-plot (Fig. 2-120). The microconglomerate nature is confirmed by the borehole
images of Figure 2-121 which comspond to the interval marked by a red strip.
110
Serralog Q 2003
Composition
I Chapter 2
11 1
m
Figure 2-120 - Cross-plot Uma (X axis) vs Rhoma (Y axis) on the same
interval than the other cross-plots, with thorium on the Z-axis.
Observe the position of the points essentially in the triangle quartzmicrocline-opal, and some points in the quadrangle quartz-microclinecalcite-dolomiteindicating levels cemented by calcite or dolomite. The
red points with high thorium values correspond to shale but their position is influenced by cave effects on density
Figure 2- 123 - Color coding for pma-Uma cross-plot as a function of
Vcl. These colors are used in the lithology display realized on
PLATFORM EXPRESS answer product (courtesy of Schlumberger).
Pad orientation
-
Figure 2-121 Borehole image of a 1.5 m interval showing the microconglomertae nature of the sandstone.
3
Figure 2-124 - Example of lithology display made by Platform Express
(courtesy of Schlumberger).
-
Figure 2-122 Color cube of primary colors represent lithology classes
(courtesy of Schlumberger).
Serralog Q 2003
Ill
112
Well logging and Geology
Neural network approach
This approach uses a Simulated Neural Network
(SNN) to identify electrolithofacies.
A neural network is a lattice of neurons that are interconnected by synapses. It is inspired by the organization
of the human brain which is a specific architecture of
neurons connected via synapses with axons and dendrites (Fig. 2-125).
In the hardware structure of a neural network, neurons
are replaced by transistors, axons and dendrites by wires,
and synapses by capacitors, resistors and inductors.
As stated by Baldwin et al. (1990) “the simultaneous
operation of highly interconnected neurons produces
mindlike synergetic characteristics and qualities that can
be exploited to perform sophisticated log analysis and
interpretation”.
Several SNN types exist.
- The Self-organizing Maps (SOM) neural networks
(Fig. 2-126), originally devised by Kohonen, consists of
two layers of processing units: an input layer (log data)
fully connected to a competitive output layer (artificial
lithofacies).
-
Figure 2-125 Typical neural network of a human brain.
Each neuron may be connected to 10,000 other
neurons. Each neuron has one extension, called axon,
which may have numerous ramifications,and one or more
axon endings. The axon conducts the information only in
one way. The axon ending is composed of a certain number of very small bags which contains molecules. These
molecules transmit the information through the synapse
slit. The axon ending is the location where an electric
charge, coming from a neuron through the axon, is transformed into a chemical signal thanks to the neurotransmitters. The synapse slit is the space where the chemical
signals are transmitted from a neuron to the other. The
neurotransmitters are substances which match the molecules to the receivers in the same way than only one key
corresponds to a bolt. The receivers are located on the
sensitive surface of the transfer zone. The dendrites
receive information from the other neurons.
Each neuron operates very simply and processes
information in simple fashion. A neuron receives the information from the other neurons and “decides” to transmit it
or not as a function of the electric charge coming from the
cellular surface. However, each neuron of the lattice operates simultaneously and in parallel.
A Simulated Neural Network is a software equivalent
of a hardware structure designed to perform some task.
112
_I
Figure 2-126 - The selected lithofacies (Winner) collects information from all input data (preceding neurons). The output of each preceding neuron (input data) is modulated by a corresponding weight (wi)
before affecting the activity of the neuron.
Frayssinet et al. (2002) uses this approach to represent lithofacies. The map is a discrete set of formal
neurons. Each neuron of the map is associated to a referent vector in the data space. A rectangular map of size
(13x7) was generated (Fig. 2-127).
_I
Figure 2-127 Topologic map established with 4 log data (pb, Pe, GR,
4d.L=limestone; M=marls, gS=glauconitic sandstone; Sha=shale;
Sha 1=other shales; Si=silt; Sil =other silts; cS=coarse grain sandstone;
S=strict sandstone; SL=sandy limestone; B=sandy breccia; Lig=lignite
(from Frayssinet et al., 2002).
-
Serralog 0 2003
Composition
The labelling of the cells (neurons) was done by reference to the core data. Grey level between two neighboring neurons represents the distance between them. Rock
classes are represented by the gathering of neurons .
- The SNN can use Self-Organizing-Activation (SOA)
principles devised by Kohonen (1984) to describe observations of how neurons in an optic nerve operate and
interact. This is based on the fact that log analysts make
visual observation of logs, cross-plots and images in the
initial stage of well log interpretation. Baldwin et al. (1989)
had successfully applied 2D SOA principles to interpretation requiring visual correlation of dipmeter and full waveform sonic curve traces. They extended from two to eight
dimensions the SOA rules. They generated 8D inhibitory
and excitatory hyperspheres (Fig. 2-128). Each log measurement can be subdivided in a reduced number of segments (8). One neuron is assigned to each segment.
When combining log measurements one generates n8
hypercubes, n being the number of different log data
taken into account. Each hypercube corresponds to an
artificial lithofacies. This type of process generates a lot of
hypercubes or artificial lithofacies and requires large
memory and long processing time to obtain a solution. In
addition, many hypercubes do not correspond to any real
lithofacies.
When the 8D hypercube completes its operation a set
of hyperclusters of active neurons results. Each hypercluster in n dimensions represents a lithofacies type that
the log analyst can determine by visual observation of the
raw n data.
Surface neurons
Internal neurons
Synapses
Figure 2- 128 - Self-organizing hypercube (only three dimensions are
shown) containing excitation and inhibition hyperspheres
(from Baldwin et al., 1990).
To be sure that each lithofacies cluster represents only
a single neuron in n dimensions, a final stage was introduced by Baldwin et al. This stage is known as OnCenter-Off-Surrounding (OCOS) competitive activation.
“OCOS principles operate such that individual neurons
within a group compete among themselves until the strongest neuron eventually suppresses all others within that
group. The operation of the hypercube and OCOS processing completes the identification portion of the interpretation task.” (Baldwin et al., 1989).
The neural network structure is completed by a
Serralog $3 2003
Chapter 2
.
113
Competitive-Activation Pattern Classification (CAPC)
paradigm (McClelland & Rumelhart, 1988).
To explain the logics of the SNN and CAPC network
the following lines are extracted from the paper written by
Baldwin et al. (1989).
Figure 2-128 “shows the S N N during hypercube training of the CAPC network... It illustrates how the S N N
connects the hypercube and CAPC networks and describes the internal architecture of the CAPC network”
(Baldwin et al., 1989)... The CAPC network consists of
pools of input log neurons and one pool of lithofacies output neurons. The number of input pools equals the number of input logs. The number of neurons per input log
pool equals the number of segments into which each log
measurement has been broken... The number of neurons
in the output pool equals the number of lithofacies that the
SOA hypercube discovered. Every neuron in each input
pool is connected to every neuron in the output pool.
Additionally, these connections are bidirectional. That is,
every connection from an input pool neuron to an output
pool neuron is reciprocated as is shown for the third
neuron of the Cxo pool (Fig. 2-129)...
0 InhibitedNeuron
C Excited Neuron
* Inhibitory Synapse
Excitatory Synapse
-
Figure 2-129 - Lithofacies pattern identification and recognition by a
SNN trained from inputs of an auto-associated SOA hypercube. Due to
the complexity of the structure only few connections are reproduced
(i.e. the eight neurons of the K pool). This pattern illustrates how the
SNN connects the hypercube and CAPC networks and describes the
’
internal architecture of the CAPC network
(from Baldwin el al., 1990).
Input pools, however, are not allowed to connect to
other input pools. All within-pool neurons are interconnected with Off-Surrounding (0s)architecture (illustrated for
the fourth neuron of the K curve pool in Figure 2-128)...
The CAPC paradigm functions in the following manner. When an input signal consisting of one value per log
113
curve per depth is received by the CAPC network, only
one or at most two neurons within each log pool become
active... Neurons activated by the input signal begin activating lithofacies neurons via synaptic connections to the
output pool. Because these connections are bidirectionnal, activated lithofacies neurons begin sending activating
signals back to the log pools. Within-pool neurons also
interact via 0s competition. Thus, within-pool neurons in
each pool compete with each other for domination.
Simultaneouslx neurons in log pools activate neurons in
other log pools indirectly through activation of the lithofacies neuron pool”.
- Neural network systems can be trained to perform
particular tasks based on well log data. The Back
Propagation Neural Network (BPNN) is done by presenting the system with a representativeset of examples describing the problems to solve. Input samples (set of log
data) are connected to output samples (lithofacies description provided by core analysis). BPNN is in fact a
supervised training method. It requires the knowledge of
the output to a given input. After training, the output of the
Neural Network is compared to the desired output. The
observed difference is used to adapt the weights. The
adaptation of the weights starts at the output neurons and
continues towards the input data. After training, the neural
network approach can be used to recognize data that are
similar to those of the examples shown during the training
phase.
- The SNN approach is also used to generate a missing logging curve from available logging data. For instance the neutron curve must be predicted from a combination of density, gamma ray and photoelectric index.
Frayssinet et a/. (2000) based their method on a MultiLayer Perceptron (MLP). This method “acts as a non
linear regression method for the prediction task and as a
probability density distribution approximation for the outlier rejection task.... The MLP is made of three kinds of
neuron’s layers (similar to the back-propagation neural
network): an input layer, one or several hidden layers and
an output layer”.
Results of this method is shown in Figure 2-131.
J
Figure 2-131 - Reconstruction of the neutron curve from three other log
data and comparison with the recorded neutron curve
(from Frayssinet et al., 2000).
Discussions
The SNN approach has distinct advantages:
- computer codes are very short,
- they do not use a special programming language,
- they do not require the services of special knowledge
engineers or experts.
The SNN approach is able to learn interpretation techniques performed by experts.
However, most of the time core data are needed to
determine the type of lithofacies that the analyst wants to
predict from the logging data.
Quite often the logging measurements are not corrected for their lack of resolution. This can generate misinterpretation as the non-representative data, corresponding
to the transition depth between successive beds, represent clusters that can be confused with lithofacies.
Some logs used for the lithofacies determination (for
instance resistivity, SP) represent more often the porosity
and the fluid content than the lithology or the mineralogy.
R eferences and bibliography
Gamma ray
Neutron
Density
Figure 2-130 - Schematic of the back-propagation neural mtwork
(from Rogers et al., 1992).
114
ADAMS, J.A.S. (1954). - Uranium and Thorium contents of volcanic rocks. In: Nuclear Geology. J. Wiley &
Sons, New York.
ADAMS, J.A.S. & GASPARINI, P. (1970). - Gammaray spectrometry of rocks. Elsevier Publ. Co.,Amsterdam.
Serralog 0 2003
Composition
ADAMS, J.A.S. & WEAVER, R. (1958). - Thorium to
Uranium ratio as indicator of Sedimentary Processes:
examples of concept of Geochemical facies. Bull. amer.
Assoc. Petroleum Geol., 42 , 2.
AUBERT, G., GUILLEMIN, C. & PIERROT, R. (1978).Precis de mineralogie. Masson, Paris.
BALDWIN, J.L., OTTE, D.N. & BATEMAN, R.M.
(1989). - Computer emulation of human mental processes: application of neural network simulators to problems in
well log pattern recognition. Conference on Artificial
Intelligence in Petroleum Exploration and Production,
Texas A&M University, May 17-19, pp. 145-175.
BALDWIN, J.L., OTTE, D.N. & WHEATLEY, C.L.
(1989). - Computer Emulation of Human Mental
Processes: Application of Neural Network Simulators to
Problems in Well Log Interpretation. SPE 64th Annual
Technical Conference and Exhibition, San Antonio, Texas,
paper SPE 19619, pp. 481-493.
BALDWIN, J.L., BATEMAN, R.M. & WHEATLEY, C.L.
(1990). - Application of a Neural Network to the Problem
of Mineral Identification from Well Logs. The Log Analyst,
31, 5, pp. 279-293. Originally presented at M.G. .S., 3rd
International Symposium on borehole geophysic for minerals, geotechnical, and groundwater applications, Las
Vegas, Nevada, Oct. 2-5, 1989, paper U.
BARIAND, P., CESBRON, F. & GEFFROY, J. (1978). Les Mineraux: leurs gisements, leurs associations.
Editions du B.R.G.M..
BARTH, T.F.W. (1952). - Theoretical Petrology. John
Wiley & Sons, Inc., New York.
BAYLY, B.(1976). - lntroduction a la petrologie.
Masson, Paris.
BEERS, R.F. & GOODMAN, C. (1944). -Distribution of
radioactivity in ancient sediments. Bull. Geol. SOC.
America, 55.
BERRY, L.G. & MASON, B. (1959). - Mineralogy
Freeman & Co., San Francisco.
BISHOP, C. (1995). - Neural Networks for Pattern
Recognition. Oxford University Press.
BLATT, H., MIDDLETON, G. & MURRAY, R. (1980). Origin of Sedimentary Rocks. 2nd ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
BOLVIKEN, E., STORVIK, G., NILSEN, D.E., SIRING,
E. & van der WEL, D. (1992). - Automated prediction of
sedimentary facies from wireline logs. In: Geological
Applications of Wireline Logs II, HURST, A., GRIFFITHS,
C.M. & WORTHINGTON, P.F. (eds), Geol. SOC.London,
Special Publication no 65, pp. 123-140.
BOYER, S., ANDRIEUX, P. & MICHAUT, B. (1993). Principes physiques de fonctionnement des outils de diagraphie. Geologues, 99.
BRAM, K. & DRAXLER, J.K. (1993). - KTB Report 931: Basic Research and Borehole Geophysics (Report 14).
Stuttgart, Germany.
BRAM, K. et a/. (1995). - The KTB BoreholeGermany’s Superdeep Telescope into the Earth’s Crust.
Oilfield Review, 7, 1, p. 4-22.
BURKE, J.A., CURTIS, M.R. & COX, J.P. (1967). -
1
Serralog 0 2003
I Chapter 2 I
115
Computer Processing of Log Data enables better
Production in Chaveroo Field. J. Petroleum Technol.
BURKE, J.A., SCHMIDT, A.W., & CAMPBELL, R.L. Jr.
(1969). - The litho-porosity crossplot :a method of determining rock characteristics from the computation of log
data. SPWLA, 10th Ann. Log. Symp. Trans.
BUTTGENBACH, H. (1953). - Les mineraux et les
roches. Dunod, Paris.
CAILLERE, S. & HENIN, S. (1963). - Mineralogie des
argiles. Masson, Paris.
CAROZZI, A.V. (Ed.) (1975). - Sedimentary Rocks.
Benchmark Papers in Geology, 1 5 , Dowden, Hutchinson
& Ross, Inc., Stroudsburg, Pennsylvania.
CHRISTENSEN, N.I. (1965). - Compressional wave
velocity in metamorphic rocks at pressure to 10 kbars. J.
Geophys. Research. Vol. 70, no. 24.
CLARK, S.P.Jr. ed. (1966). - Handbook of Physical
constants. The Geol. Society of America.
CLARKE, F.W. (1924). - The data of geochemistry.
Bull. U S . Geol. Surv., 770, 841 p.
CLARKE, F.W. & WASHINGTON, H.S. (1924). - The
composition of the Earth‘s Crust. U. S. Geol. Survey,
Profess. Paper 127.
CLAVIER, C. & RUST, D.H. (1976). - MID-Plot:A new
Lithology Technique. The Log Analyst, 17, 6, p. 16.
CRAIN, E.R. & ANDERSON, W.B. (1966). Quantitative log evaluation of the Prairie Evaporite
Formation in Saskatchewan. J. Can. Petrol. Techn., 5, 3,
p. 145-152.
DALY, R.A. (1933). - Igneous Rocks and the Depth of
the Earth. McGraw-Hill Book Co., New York.
DALY, R.A., MANGER, G.E., & CLARK, S.P. Jr.
(1966). - Density of rocks. In: Handbook of Physical
Constants.
DEER, W.A., HOWIE, R.A. & ZUSSMAN, J. (1962,
1976). - Rock-forming Minerals. Longman, London.
DEER, W.A., HOWIE, R.A. & ZUSSMAN, J. (1978). An introduction to the Rock-forming Minerals. Longman,
London.
DELFINER, P., PEYRET, O., & SERRA, 0. (1984). Automatic determination of lithology from well logs. 59th
ann. Techn. Conf. of SPE of AIME, Houston, paper SPE
13290.
DESBRANDES, R. (1985). - Encyclopedia of Well
Logging. Editions Technip, Paris.
DOVETON, J.H. (1994). - Geologic Log Analysis
Using Computer Methods. AAPG Computer Applications
in Geology, no 2.
DRAGOSET, M.A. (1977). - An Evaporite Analysis in
Southeastern New Mexico. Dresser Atlas internal Report.
DREYFUS, G. (1997). - Les reseaux de neuronas, une
technique operationnelle pour le traitement des donnees
industrielles, economiques et financieres. Mesures, no
699, Novembre.
DREYFUS, G. (1998). - Les reseaux de neurones.
Mecanique lndustrielle et Materiaux, no 51, Septembre.
DUNHAM, R.J. (1962). - Classification of Carbonate
Rocks according to Depositional Texture. Amer. assoc.
115
Petroleum Geol., Mem. 1, pp. 108-121.
EDMUNDSON, H. & RAYMER, L.L. (1979). Radioactive parameters for Common Minerals. SPWLA,
20th Ann. Log. Symp. Trans., paper 0 .
ELLIS, D., HOWARD, J., FLAUM, C., Mc KEON, D.,
SCOTT, H., SERRA, O., & SIMMONS, G. (1988). Mineral Logging Parameters: Nuclear and Acoustic. The
Thechnical Review, 36, 1, pp. 38-53.
EMMERMANN, R., LAUTERJUNG, J. & UMSONST,
T. (1993). - KTB Report 93-2: Contribution to the 6th Ann.
KTB Colloquium. Geoscientific Results, Stuttgart,
Germany.
FAUSETT, L. (1994). - Fundamentals of Neural
Networks. Prentice-Hall International Inc., New Jersey.
FELDER, B. & BOYELDIEU, C. (1979). - The
Lithodensity Log. SPWLA, 6th Europ. Symp. Trans., paper
0.
FISCHESSER, R. (1977). - Donnees des principales
especes minerales. Societe de I’lndustrie Minerale, SaintEtienne, France.
FOLK, R.L. (1959). - Practical Petrographic
Classification of Limestones. Bull. Amer. assoc.
Petroleum Geol., 43, 1, pp. 1-38.
FORD, W.E. (1955). - Dana’s Textbook of Mineralogy.
John Wiley & Sons, Inc., New York.
FORD, M.E., BAINS, A.J. & TARRON, R.D. (1974). Log analysis by linear programming. An application to the
exploration for salt cavity storage locations. SPWLA 3rd
Europ. Log. Symp. Trans., London, paper B.
FRAYSSINET, D., THIRIA, S., BADRAN, F. &
BRIQUEU, L. (2000). - Use of neural networks in log’s
data processing: prediction and rebuilding of lithologic
facies. Petrophysics meets Geophysics, Paris, 6-8 Nov.
GARRELS, R.M. & McKENSIE, F.T. (1971). - Evolution
of Sedimentary rocks. Norton, W. W. & Co, New York, 397
P.
GOTTLIB-ZEH, S. (2000). - Synthese des donnees
geologiques et petrophysiques acquises en forages
petroliers B /’aide de traitements statistiques et neuronaux. Applications aux depbts turbiditiques et aux
pla tes-formes mixtes. These, MontpelIier University, 265
P.
GOTTLIB-ZEH, S., BRIQUEU, L., & VEILLERETTE, A.
(1999). - lndexed Self-organizing Map: a new calibration
system for a geological interpretation of logs. In: Proc.
IAMG, pp. 183-188.
GOTTLIB-ZEH, S., BRIQUEU, L., WLODARCZYK, S.,
& MOSS, B. (2002). - Neural applications to Geological
and petrophysical data for improved flow-unit identification. SPWLA Europ. Symp. Transaction, paper T2C.
GRAU, J.A., SCHWEITZER, J.S., DRAXLER, J.K.,
GATTO, H. & LAUTERJUNG, J. (1993). - Elemental
Logging in the KTB Pilot Hole - 1. Nal based spectrometry. Nuclear Geophysics, 2, p. 19-73-187.
GRIM, R.E. (1968). - Clay Mineralogy. McGrawHill
Co., New York.
GROUT, F.F. (1932). - Petrography and Petrology.
McGraw-Hill Book Co., New York.
116
HAILE, P.M. & BLUNDEN, H.A. (1984). - Zechstein
Magnesium rich Evaporite Deposits of northern
Netherlands and their volumetric analysis by GLOBAL.
SAID-SPWLA, 9th International Log. Symp. Trans., Paris,
paper 37.
HASSAN, M. & HOSSIN, A. (1975). - Contribution a
/’etude des comportements du thorium et du potassium
dans les roches sedimentaires. Comptes rendus
Academie des Sciences, Paris, 280.
HEIER, K.S. & ADAMS, J.A.S. (1965). - Concentration
of radioactive elements in deep crustal materials.
Geochim. Cosmochim. Acta, 29, p. 53?62.
HENKEL, H. (1976). - Studies of density and magnetic properties of Rocks from Northern Sweden. Padeagh
214.
HERRON, M.M. (1986). - Mineralogy from
Geochemical Well Logging. Clays and Clays Minerals, 34,
2, pp. 204-213.
HOLMES, A. (1913). - The age of the earth. Harper &
Row, New York.
HOSSIN, A. (1969). - Interpretation des diagraphies en
zones carbonatees et evaporitiques. Bull. A. F. T. P., 193,
p. 47-69.
JUNG, J. (1958). - Precis de petrographie. Masson,
Paris.
KELLER, G.V., MURRAY, J.C. & TOWLE, G.H.
(1974).- Geophysical Logs from the Kilauea Geothermal
Research Drill Hole. SPWLA, 15th Ann. Log. Symp.
Trans., paper L.
KENNEDY, M.C. (2002). - Solutions to Some
Problems in the Analysis of Well Logs in Carbonate
Rocks. In: LOVELL, M., & PARKINSON, N. (ed.) (2002). Geological Applications of Well Logs. AAPG Methods in
Exploration, 604 p.
KERHERVE, J. (1977). - Notes prepared for an
advanced seminar on interpretation of well logs in volcanic rocks.
KHATCHIKIAN, A. (1982). - Log Evaluation of OilBearing lgneous Rocks. SPWLA, 23rd Ann. Log. Symp.
Trans., paper AA.
KLEIN, G. deVries (1980). - Sandstone Depositional
Models for Exploration for Fossils Fuels. 2nd ed. CEPCO
Div., Burgess Publishing Co., Champaign, Illinois.
KOHONEN, T. (1984). - Self-organization and associative memory. Springer-Verlag, New York, 255 p.
KOHONEN, T. (1992). - The Self-organizing Map.
Information Sciences, Springer-Verlag, New York., vol.
30, 312 p.
KRAUS, E.H., HUNT, W.F. & RAMSDELL, L.S. (1936).
- Mineralogy. An introduction to the study of Minerals and
Crystals. McGraw-Hill Book Co., New York.
KRUMBEIN, W.C. (1959). - The tetrahedron as a
facies mapping device. J. Sed. Petrol., 24 , 1, p. 3-19.
KRUMBEIN, W.C. & SLOSS, L.L. (1963). Stratigraphy and Sedimentation. 2nd ed. WH. Freeman &
Co, San Francisco.
KRYNINE, P.D. (1948). - The megascopic study and
field classification of sedimentary rocks. J. Geology, 56 ,
Serralog 0 2003
p. 130-165.
LARSEN, E.S. Jr. & PHAIR, G. (1954). - The distribution of uranium and thorium in igneous rocks. In: Nuclear
Geology. J. Wiley & Sons, New York.
LEET, L.Don, JUDSON, S. & KAUFFMAN, M.E. (1978,
1982). - Physical Geology. 5th ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
LEITH, C.K. & MEAD, W.J. (1951). - Metamorphic
rocks. Henry Holt & Co., New York.
LINK, P.K. (1982). - Basic Petroleum Geology. OGCl
Publications, TuIsa.
LOMBARD, A. (1956). - Geologie Sedimentaire. Les
series marines. Masson, Paris.
LOMBARD, A. (1972). - Series sedimentaires.
Genese-Evolution. Masson, Paris.
LOVELL, M.A., & PEZARD, P.A. (1990). - Electrical
properties of basalts from DSDP Hole 504B: a key to the
evaluation ofpore space morphology. In: Hurst, A., Lovell,
M.A., & Morton, A.C. (eds), Geological Applications of
Wireline Logs, Geological Society, Special Publication no
28, London.
LOVELL, M.A., HARVEY, P.K., JACKSON, P.D.,
BREWER, T.S., WILLIAMSON, G., & WILLIAMS, C.G.
(1998). - lnterpretation of core and log data - integration
orcalibration? In: Harvey, P.K., & Lovell, M.A. (eds), CoreLog Integration, Geol. SOC. London, Special publication
136, pp 39-52.
LOVELL, M., & PARKINSON, N. (ed.) (2002). Geological Applications of Well Logs. AAPG Methods in
Exploration, 604 p.
MASON, B. (1952). - Principles of Geochemistry.John
Wiley & Sons, Inc., New York
MATTHEWS, R.K. (1974). - Dynamic Stratigraphy.
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
MAYER, C. & SIBBIT, A. (1980). - GLOBAL, a new
Approach to Computer-processed Log Interpretation.
SPE of AIME, Fall Mtg., Dallas, SPE 9341.
MEAD, W.J. (1907). - Redistribution of elements in the
formation of sedimentary rocks. Jour. Geol., 15, pp. 238256.
MILLWARD, D., YOUNG, S.R.,
BEDDOESTEPHENS, B., PHILLIPS, E.R., & EVANS, C.J. (2002).
- Gamma-ray, Spectral Gamma-ray and Neutron-density
Logs for lnterpretation of Ordovician Volcanic Rocks,
West Cumbria, England. In: Lovell, M., & Parkinson, N.
(ed.) (2002). - Geological Applications of Well Logs.
AAPG Methods in Exploration, 604 p.
MOSS, B. (1997). - The partitioning of Petrophysical
Data: a Review. In: Lovell, M.A., & Harvey, P.K. (eds),
Developments in Petrophysics, Geol. SOC. London,
Special publication 122, pp 181-252.
PETTIJOHN, F.J. (1975). - Sedimentary Rocks. 3d ed.
Harper & Row, Publishers, New York.
PETTIJOHN, F.J., POTTER, P.E. & SIEVER, R.
(1972). - Sand and Sandstone. Springer, New York.
PICKETT, G.R. (1963). - Acoustic character logs and
their applications in formation evaluation. J. Petrol.
Technol., 15, 6.
Serralog 0 2003
PIRSSON, L.V. (1949). - Principles of Geochemistry.
John Wiley & Sons, Inc., New York.
PIRSSON, L.V & KNOPF, A. (1947). - Rocks and Rock
Minerals. 3rd ed. John Wiley & Sons, Inc., New York.
PRESS, F. (1966). - Seismic Velocities. In: Handbook
of Physical Constants.
PRESS, F. & SIEVER, R. (1978, 1982). - Earth. 2nd &
3rd ed. WH. Freeman & Co, San Francisco.
RAYMER, L.L. & BIGGS, W.P. (1963) . - Matrix characteristics defined by porosity computations. SPWLA, 4th
Ann. Log. Symp., Trans., paper 10.
RITCH, H.J. (1975). - A n open hole logging evaluation
in metamorphic rocks. SPWLA, Ann. Log. Symp. Trans.,
Paper V.
ROGERS, S.J., FANG, J.H., KARR, C.L. & STANLEY,
D.A. (1992). - Determination of Lithology from Well Logs
Using a Neural Network. Bull. Amer. Assoc. Petroleum
Geol., 76, 5, pp. 731-739.
Schlumberger Ltd (1974). - Log Interpretation. Volume
/I - Applications.
Schlumberger Technical Services, Inc. (1982). Essentials of NGS Interpretation.
SANYAL, S.K., WELLS, L.E. & BICKHAM, R.E.
(1980). - Geothermal Well Log Interpretation. State of the
Art. final Report. Los Alamos Scientific Laboratory, LA8211-MS.
SCHOLE, P.A. & SPEARING, D. (Ed.) (1982). Sandstone Depositional Environments. Amer. Assoc.
Petroleum Geol., Mem. 31.
SELLEY, R.C. (1978). - Ancient Sedimentary
,Environments.2d ed. Chapman & Hall, London.
SELLEY, R.C. (1976). - An lntroduction to
Sedimentology.Academic Press, London.
SERRA, 0. (1980). - Aspects diagraphiques des evaporites. Bull. Cent. Rech. Expl. Prod. Elf-Aquitaine, Pau,
4, 1, p. 411-431.
SERRA, 0. (1986). - fundamental of well log interpretation. 2 - lnterpretation of logging data. Elsevier,
Developments in Petroleum Sciences 15B.
SERRA, 0. (1990). - Element, Mineral, Rock Catalog.
Schlumberger.
SERRA, 0. (1993). - Relations entre les donnees diagraphiques et les parametres geologiques. Les differents
domaines d’applications geologiques des diagraphies.
Geologues, 99.
SERRA, O., DELFINER, P. & LEVERT, J.C. (1985).Lithology determination from well logs: case studies.
SPWLA, 26th Ann. Log. Symp. Trans., Paper WW.
SHAND, S.J. (1951). - Eruptive rocks. 4th ed. John
Wiley & Sons, Inc., New York.
SMITHSON, S.B. (1971). - Density of metamorphic
rocks. Geophysics, 36, 4.
STUNDNER, M. & LECHNER, J. (2000). - Delineation
of Lithofacies based on Well Logs - Application of
Stochastic and Neural Network Models for Analyzing and
Interpretation. 1st Symposium on Well Log Analysis &
Formation Evaluation, 29-31 Oct., Tripoli, Libya, paper
HI1.
117
SULEIMAN, O.E. & KELEMEN, Z. (2000). - The Role
of Principal Component Analysis (PCA) in the Delineation
of Lithofacies based on Well Logs. 1st Symposium on Well
Log Analysis & Formation Evaluation, 29-31 Oct., Tripoli,
Libya, paper H/2.
SUPERNAW, I.R., McCOY, A.D. & LINK,A.J. (1978). Method for in-situ evaluation of the source rock potential
of each formation. U.S. Patent 4,071,744, Jan. 31.
TUREKIAN, K.K. & WEDEPOHL, K.H. (1959). Distribution of the elements in some major units of the
earth’s crust. Geol. SOC.Am. Bull., 72, pp. 175-192.
TURNER, F.J. & VERHOOGEN, J. (1951). - Igneous
and Metamorphic Petrology. McGraw-Hill Book & Co.
Inc., New York.
WAHLSTROM, E.E. (1947). - Igneous Minerals and
Rocks. John Wiley & Sons, Inc., New York.
WEAST, R.C. ed. (1981-82). - Handbook of Chemistry
and Physics. 62nd ed. CRC Press Inc.
WEDEPOHL, K.H (1969). - Handbook of
Geochemistry.
WESTPHAL, H & BORNHOLDT, S. (1996). Lithofacies prediction from wireline logs with genetic algorithms and neural networks. E. Schweizerbart’sche
Verlagsbuchhandlung, pp. 465-474.
WEST, F.G., & LAUGHLIN, A.W. (1976). - Spectral
Gamma Logging in crystalline basement rocks. Geology,
VOI. 4, pp. 617-618.
WHITFIELD, J.M., ROGERS, J.J.W. &ADAMS, J.A.S.
(1959). - The relationship between the Petrology and the
Thorium and Uranium contents of some granitic rocks.
Geochim. & Cosmochim. Acta, 17.
WICKMAN, F.E. (1954). - The “total” amount of sediment and the composition of the “average igneous rocks”.
Geochim. Cosmochim. Acta, 5, pp. 97-110.
WILSON, J.L. (1975). - Carbonate Facies in Geologic
History. Springer, New York.
ZHOU, C.D., WU, X.L. & CHENG, J.A. (1993). Determining reservoir properties in reservoir studies using
a fuzzy neural network. SPE Annual Technical
Conference and Exhibition, Houston, Texas, paper SPE
26430, pp. 141-150.
118
Serralog 0 2003
WELL LOGGING AND ROCK TEXTURE
Int roduction
Review of petrographic concepts
Before explaining how to extract textural information
from well logs it is important to review some concepts
which will help in the understanding of the links between
well logging data and textural parameters.
Definition
Texture is " the general physical appearance or character of a rock, including the geometric aspects of, and
the mutual relations among, its component particles or
crystals; e.9. the size, shape and arrangement of the
constituent elements of a sedimentary rock, or the crystallinity, granularity and fabric of the constituent elements
of an igneous rock. The term is applied to the smaller
ROCK
COMPONENTS
COMPOSITION
Figure 3-1
Serralog 0 2003
(megascopic or microscopic) features as seen on a
smooth surface of a homogeneous rock or mineral aggregate" (Bates & Jackson, 1980).
So, texture covers the geometrical aspects of the
constituent components of rocks : grains or particles and
crystals: their size, shape, appearance, their arrangements and sorting; also the grain-grain, grain-matrix or
grain-cement boundings.
Texture plays a very important part in sedimentary
rocks, because the petrophysical properties of a rock,
hence its porosity and permeability, depend essentially on
texture.
According to the origin of the rocks, Krumbein & Sloss
(1963) distinguish:
- fragmental texture, more specific to detrital rocks;
- crystalline texture, more specific to chemical or eruptive rocks.
- The components of texture.
119
Textural components
Regardless of the type of texture, sedimentary rocks
are marked by the textural characteristics of four principal
components (Fig. 3-1):
grains, or particles, and crystals;
matrix, that corresponds to the finest-grained
material enclosing, or filling the interstices or space between the larger grains, particles or crystals of a sediment.
It is deposited at the same time than the other elements;
- cement, binding the grains and the matrix;
porosity, or void, generally filled by fluid.
These components may have different mineralogical
compositions. Figure 3-1 shows the constituent minerals
of the different textural components, depending on the
type of detrital rocks.
These components are not necessarily present all
together in the same rock. For example in medium-grained and very well-sorted sand there will be neither matrix
nor cement. A consolidated, medium-grained, well-sorted
sandstone has no matrix, but contains a cement that can
be siliceous, calcitic or dolomitic, sideritic, or halitic.
The grains and matrix are generally associated,
because they are most often deposited together (sand
grains in an argillaceous matrix, or pebbles in a sandy or
shaly matrix). The cement, being the result of chemical
precipitation in the pore space, is always post-depositional.
As shown by Krumbein & Sloss (1963), the study of
texture must be subdivided into three categories depending on the nature of sedimentary rocks:
- purely chemical rocks, such as halite, gypsum, anhydrite... The texture of these rocks is characterized only by
the crystalline system, the size and imbrication of crystals;
- partly chemical or biochemical and partly detrital
rocks, such as carbonates, for which the three textural
components play, in turn or simultaneously, a very important part;
- detrital rocks, for which the three components play a
very important part.
ties
- grain size and its variation that will control the sorting;
- shape (or sphericity) of grains (Fig. 3-2);
- roundness;
- surface texture;
- grain orientation or fabric;
- mineralogical composition. This is not, properly spea-
king, a textural parameter, but depends on:
- density, hence the rate of sedimentation of each
cornponent ,
- possibility of dissolution (solubility) or of alteration, consequently the ulterior formation of vugs
or cement,
- wettability of rock.
Grain size
This parameter is very important as it is used to subdivide detrital sedimentary rocks into conglomerates,
sandstones, siltstones and claystones (Fig. 3-2). It is also
a guide to the proximity of the source area: coarser sediments are generally closer to the source, the grain size
decreasing with the transport distance due to the breaking
and abrasion of the particles during the transport (Fig. 33). Also, grain size is related to the dynamic conditions of
transportation and deposition.
Combined with the sorting of grain size, it indicates the
competence and efficiency of the transporting agent. The
agents (water, wind, ice) and modes of transport (traction,
saltation, suspension) differ materially in their sorting and
transporting ability.
Detrital rocks will be discussed first.
Texture of detrital rocks
According to Krumbein & Sloss (1963), the final texture of a sediment is influenced by six fundamental proper-
MILES FROM SOURCE
Figure 3-3 - Variation of beach sand size and sorting with distance
from source (from Krumbein & Sloss, 1963).
J
Figure 3-2 - Common terms used to describe grains as a function of their size either in mm or the Wentwodh-Lane phi scale.
120
Serralog 0 2003
Sorting
It characterizes “the dynamic process by which sedimentary particles having some particular characteristics
(such as similarity of size, shape, or specific gravity) are
naturally selected and separated from associated but dissimilar particles by the agents of transportation” (Bates &
Jackson, 1980).
A
Figure 3-4 - Visual aspect of sorting
(adapted from Trask, 1932 and Folk, 1965; in Pettijohn et al., 1972)).
A sorting coefficient is sometimes computed as an indicator of grain size variation. It corresponds to the ratio of
the maximum grain diameter to the ‘minimum grain diameter. This ratio is equal to l if all the spheres or grains
have the same diameter; it is higher if the grain size
varies. The sorting is also expressed through the standards deviation of grain size distribution (Fig. 3-4).
J
Figure 3-5 - Histograms of grain size distribution corresponding to
different types of environments
(adapted from Lung, 1927; Pettijohn, 1975; Wentworth, 1931; Krumbein
& Griffiths, 1938; Krumbein & Tisdel, 1940; Rosenfeld).
Grain size distribution is well described by histograms
(Fig. 3 4 , or cumulative percentage curve (Fig. 3-6), that
reflects partly the depositional environment as illustrated
by Figures 3-5 to 3-7. Histograms present a factual picture of the grain size abundance in each grade size, however they cannot allow the determination of the average
particle size and other properties that cumulative curve
allows (Fig. 3-7).
A
_I
Figure 3-6 - Cumulative curve of grain size with explanation of percentiles 10 and 90, median and lrst and 3d quartiles
( h n Kmmbein 8 Slow 1963).
Serralog 0 2003
Figure 3-7 - Grain size and sorting as a function of the depositional
environment.
121
Well Logging and Geology
122
Grain shape
The geometric form of a particle involves several separate but related geometric concepts.
- Measurement of three shape factors describing its
geometry (Fig. 3-8);
- Measurement of the sphericity and roudness of the
particle (Fig. 3-9).
Figure 3- 10 - Geometrical description of roundness of two different particles (from Krumbein, 1940).
The three previous grain parameters evolve with grain
transport as illustrated by Figure 3-11.
J
Figure 3-8 - Measurement of the three axes of a particle and its classification according to its shape (from Krumbein, 1941 and Zingg, 1935).
Shape
Size
Sorting
ROUNDNESS
Length of transport increases in that direction
-
Figure 3-9 Classification of grain shape as a function of their
sphericity and roundness
(from Krumbein & Sloss, 1963, fig. 4-10).
Sphericity expresses a relation between the surface
area of a particle and its corresponding sphere. It can be
expressed as the cube root of cVab (Sneed &Folk, 1958).
Roundness is expressed as the ratio of : (average
radius of corners and edges)/(radius of maximum inscribed circle) (Wadell, 1932). Because this is difficult to
apply, it is more convenient to work with a two-dimensional figure (Fig. 3-10) (projected image of the particle). In
this case the roundness is defined as the average radius
of curvature of the corners of the particle image divided by
the radius of the maximum inscribed circle. It is expressed
as :
p = Z(q/R)/N
where ri are the individual radii of the corners, N the
number of corners, and R the radius of the maximum
inscribed circle (Fig. 3-10).
122
Figure 3-11 - Modification of the shape, size and sorting with the distance of transport.
Packing
According to Graton & Fraser (1933, simple geometrical packing of equal-sized spheres are made in six different manners (Fig. 3-12). They proved that porosity varies
according to packing from 47.64 % for the most "open"
arrangement to 25.95 % for the most compact or "closed".
Allen (1984) made a complete review of several types
of packing (ordered, random or haphazard) of particles of
different shapes (spheres, prolate and oblate spheroids).
He concluded that "regular particles ... may form packing
of all three kinds, whereas natural particles, which are
irregular, can only form packings of random or haphazard
kinds." In fact, regular particles can exist in nature. They
correspond to oolites.
Semlog 0 2003
Gravity
a)
Gravity + current
Figure 3-12 - The six possible geometric arrangements of equal size
spheres (from Graton & Fraser, 1935).
-
Figure 3-14 Orientation of flakes. a) Flakes deposited under gravity
action. b) Flakes deposited under action of gravity and cunent
(from Potter & Pettijohn, 1977, fig. 3-2).
Orientation
Influence of arain size
The orientation of particles is defined by reference to a
horizontal plane and to the direction of current.
The orientation of pebbles is generally well defined,
because their size makes observation relatively easy (Fig.
3-13). The measurement and quantification of the orientation of small sized grains (sand, silt...) is much more difficult. However, for non-spherical grains it is generally
observed that the orientation of grains is the same as the
orientation of their axis of maximum elongation and is
parallel to the direction of current (Fig. 3-14)..
-
Figure 3-13 Schematic representation of the pebble orientation in different environments (from Rukhin, 1958).
Porosity is theoretically independent of grain size. An
arrangement of spheres with uniform size, which present
the same organization, will have the same porosity,
regardless of size. This ideal situation, which corresponds
to a maximum sorting, rarely occurs in nature. It can
sometimes be observed in washed or winnowed sands,
and more frequently in oolitic sands. Figure 3-15, from
Dodge et a/. (1971), seems to confirm that, above a certain level of porosity reflecting the best sorting and the
absence of cement, the porosity is more or less independent of grain size.
P
8
is
n
These different textural parameters affect reservoir
characteristics.
Influence of grain properties
Porosity and permeability, which are the main petrophysical characteristics of reservoir, depend on several
factors.
Beard and Weyl (1973) showed that the primary porosity and permeability of a detrital sediment, which has just
been
depend On five
size?
shape, roundness, orientation and arrangement of grains.
serralog Q 2003
Mean grain size (@ unit)
Figure 3-15 Relationship beheen porosity and mean grain size for a//
samples coming from sandstones of Paluxy Formation, Texas
(from Dodge et al., 1971).
123
124
Well Logging and Geology
In fact, Lee (1919), Von Engelhardt (1960) (Fig. 3-16)
discussing ancient sedimentary rocks, and Rogers &
Head (1961) (Fig. 3-17), and Pryor (1973), on the subject
of contemporary sands, show that porosity decreases
slightly, when grain size increases. This evolution is probably due to a number of factors which have only an indirect connection to grain size. Finer sands have a tendency to be more angular and are likely to organize according
to a less dense arrangement. Thus, they present a higher
porosity than sands with coarser grains.
matrix invades the large pores and fills up the big canals.
The combined influence of grain size and sorting on
porosity and permeability has been studied by Beard &
Weyl (1973). It is illustrated by Fig. 3-20.
Mean grain size ($ unit)
Mean diameter (Fm)
Figure 3-16 - Relationship between porosity and mean diameter of
grains in Bentheimer sandstones (from von Engelhardt, 1960).
Figure 3-18a - Relationship between permeability and mean grain size
for all the samples coming from sandstones of Paluxy Formation,
Texas (from Dodge et at, 1971).
r
5b
z
Q
a
Porosity (%)
Figure 3-17 - Relationship between porosity and mean diameter of
sand grains for several different sorting coefficients
A:So=2.086;B:So= 1.625;C:So=1.279;D:So=l.128;E:So = 1.061
(from Rogers & Head, 1961).
Contrarily, as demonstrated by Dodge et a/., (1971)
(Fig. 3-18), in a well-calibrated sand, permeability increases when the size of the grain increases. This is easily
understood because the pore size and the canals
(throats) which connect the pores to one another are
governed by grain size: the smaller the grains, the smaller the pores and the section of the canals will be. Thus,
capillary attraction will be stronger and permeabilitywill be
less.
Influence of sortinq
As investigated by Rogers & Head (1961) porosity and
permeability increase when sorting increases (Fig. 3-19).
In fact, in a poorly-sorted sand, the small grains (matrix)
are set in interstices left by the coarser grains. Thus, the
124
Figure 3-18b - Relationship Figure 3-18c - Influence of grain size on
between Wrmeabil#Yand
rock properties; surface area, throat
grain size Even if the rela- radius, tortuos@, Archie’s m factoc pertionship is not perfect, as
meability and irreducible water
expected, when the grain
(courtesy of Schlumberger).
size increases the permeability increases
(from von Engelhad,
1960).
Serralog 0 2003
Texture
E
E
v
E'
.-
POAOSITY, %
Figure 3-19 - Relationship between porosity and sorting coefficient of
sands for different grain sizes. A :median diameter = 0.106 mm;
B := 0.151 mm; C := 0.213 mm; D = 0.335 mm
(from Rogers & Head, 1961).
E
cu
g1
kj
n1
Influence of shape and roundness of arains
It seems that shape and roundness affect intergranular porosity. Fraser (1935) came to the conclusion that
sediments composed of spherical grains have lower porosity than those formed by grains with less sphericity He
attributed this to the fact that in the first type the grains
tend to settle according to a denser arrangement than in
the second type. Less spherical grains can pack together
in a way that creates wider volumes between them.
I.
The influence of shape and roundness on permeability is not still well known, but we observe that permeability
follows the fluctuations of porosity in connection with
variations in shape and roundness.
Porosity (%)
Figure 3-20 - Relationship between porosity and permeability for different grain size and sorting (from Beard & Weyl, 1973).
Influence of the orientation of arains
On the whole, the orientation of non-phyllitic (nonshaly) grains has no influence on porosity. On the contrary, it has a very strong influence on permeability or more
precisely on anisotropy or the direction of highest permeability.
Thus, in channel sands, the direction of maximum permeability is parallel to the axis of the elongation of sand
bodies.
In a littoral sand bar, maximum permeability is perpendicular to the axis of elongation of sand bodies, but is
parallel to the dominant direction of currents (Fig. 3-21).
The orientation of phyllitic particles (shales) will be the
same as the orientation of their large sides that are parallel to the plane of stratification (Fig. 3-14).
Serralog 0 2003
_I
Figure 3-21 - Grain orientation and directions of maximum permeability
in channel and bar sand bodies, upper and lower respectively
(from Pwor, 1973, in Selley, 1976, fig. 15).
Influence of packing
Allen (1984) showed that porosity is decreased by a
"widening of the range of particle sizes present in mixture", but is increased "by an increase of particle angularity
and surface mundness, and by the inclf.Jsionof strongly
anisometricparticles".
It seems obvious that permeability must follow a combecause the section Of pores and
parable
capillaries in compact arrangement is smaller than in
others. Figure 3-22 shows the theoretical variations in
permeability for ideal geometric packings of 500 pm-diameter spheres.
125
Next Page
126
I
Well Logging and Geology
Porosity (%)
I
Figure 3-22 - Porosity vs permeability for some ideal geometric packings of 500 pm-diameter spheres
(courtesy of R. Nurmi).
Naturally, the most "open" arrangements do not, in
fact, ever occur in ancient formations; very soon under the
influence of compaction the grains are organized according to the most "close" arrangement. The influence of
compaction will be studied later as it also affects the form
of grains.
Influence of the mineraloaical comDosition of arains
Grains composed of heavy and denser minerals will be
deposited with the minerals of the same weight, i.e. with
less density, but with bigger size.
This situation will lead to a decreasing of sorting, therefore of porosity and permeability.
Grains composed of unstable or chemically immature
minerals (pyroxene, amphibole, mica, feldspar,...) will
influence porosity of the sediments in which they occur.
Their ulterior alteration will cause the formation of authigenic clay minerals (kaolinite, montmorillonite, illite, chlorite,...) which will surround the grains or invade the pore
space, thus, causing major reductions in porosity and permeability (Fig. 3-23). But the type and distribution of authigenic clay minerals must also be considered because
they affect the permeability in a different way (Fig. 3-24).
Their influence has been studied by Neasham (1977).
Discrefe-particule or pore-filling clay minerals (for
example kaolinite "books") are characterized by uneven
distribution in the pore space. Crystals can reach large
size (more than 10 pm). As crystals grow, pore space and
permeability decrease, although there is always a small
amount of microporosity between individual plates.
Pore-lining clay minerals, essentially illite, chlorite and
montmorillonite, coat the pore walls with a thin layer of flakes that are parallel or perpendicular to the pore wall (Fig.
3-23b), but the growth does not reach far into the pore
space. A large amount of microporosity can be present
between flakes, although these pores are less than
micron in diameter. This type of authigenic clay greatly
reduces permeability and also influences many of the rock
electrical properties because it can considerably increases the surface area.
Pore-bridging clays, fundamentally illite fibers (Fig. 323c), are connected across the pore space. This type causes major reductions in permeability, since bridging is
126
-
Figure 3-23 The three types and distributions of dispersed authigenic
clay minerals in pore space. (a): pore filling; (b): pore lining; (c): pore
bridging (fmm Neasham, 1977).
most easily attained across the throats, and also it reduces the size of the pores. Porosity is less affected because microporosity is preserved between the very fine
fibers.
From the previous remarks, it seems obvious that the
knowledge of the type of distribution (laminated, structural
or dispersed), and of the nature of the clay minerals is of
the utmost importance to predict the permeability range
and the existence and distribution of permeability barriers.
Particles composed of soluble minerals (calcite, dolomite,...) lead to the creation of a secondary porosity by
their ulterior dissolution and the elimination of the solution
by hydrodynamism. They can also obliterate porosity and
consequently permeability by solution and formation of
cement by precipitation of new crystals in the pore space.
The diagenetic influences will be analyzed later
Influence of other textural comDonents on reservoir
characteristics
The other textural components, matrix and cement
also have an important influence on the petrophysical
characteristics of detrital reservoirs.
When the percentage of matrix and/or cement increases, the porosity and permeability diminish, since the fine
particles, that make up the matrix, and the cement tend to
occupy the pore space between the coarser elements.
Serralog 0 2003
Previous Page
Texture
I Chapter 3 I 127
Texture of carbonate rocks
!
Discrete Particle
0
0
The texture of carbonate rocks depends on the relative percentage of three components (particles, matrix and
cement), and on the type of pore distribution (Fig. 3-25).
Carbonate rocks have different components described
in Table 3-2.
Table 3-2
The textural components of carbonate rocks.
Pore Lining
bhodula LL (- rxtndw)
grain8 of non carbonate mmcirb: ex : quam
grains, c h ~ anhydritr,
.
phphaes,...
Imnclac;g
frrpnwmt of morM carbonrte rocka which
originated within thr drpo*riwl basin.
S l u l ~ a frsgmmt
l
(- biodrotr)
0
0
Palticlea
or
6
grains
Lumpr
or
(wou.
Allocherm
1
1 w
Porosity, percent
-
Poloid8 00.t.d by ab8l laminae.
Qdm
Phorth8,
grains
P s r d i s form in caverns (cavepearls) or in
CdiiChO CNStS bOl9OMh W&lOlOd ZOnOS
The matrix is really an important component only in
conglomerates, whereas in sands it is present only in very
poorly sorted sand bodies.
(Vadoss pirditht)
Algal encrusrod grains (- oncolii)
oncoliihs am famed by primitin Muo-gmn
algae (schizophycre or Qanophycae)
growing on a grain a d mncting urbonru
mud to their stickv sudaw.
Cement is developed after deposition either by chemical interaction between unstable grains and formation
water, or by circulation in the pore space of solutions
under hydrodynamic forces.
1.
algal lumps
Nucleus (of quam or skeletal grain) with
secretion of calcite (or silica or hematite)
with conantrio or radial struchm. am
grown in susp.mion in an agitating medium
~
(high O I I O ~ Oenvironment)
Figure 3-24 Influences on porosity and permeability of the distribution
type of authigenic clay minerals. Data from Neasham's study of 14
sandstones (from Neasham, 1977).
These different influences are synthesized in Table 3-
I
Composite arains (- gnpestone)
P.(oidsbondedbymicrite
&&Q
:carbonate mud (maxi. c w l size :
10 IM)
-
my k enmmtomd in mallquantities within
a grain-rupportrdcarbonate
mykrorkmbntmnitformraca*rt.
mudrock tanned :
micrit.oruloiknitr
Otthochorni
(FOLK, 1968
Table 3-1
Influence of textural parameters on porosity and permeability of detrital rocks.
ddolniaolpwitr
Dokmite
ddorprritr
Po-
These different parameters are used to classify the
carbonate rocks (Folk's classification). But in this chapter
the Dunham's classification will be adopted (Fig. 3-25).
Table 3-3 from Choquette and Pray (1970)illustrates
the different types of porosity in carbonate rocks, and links
them with original phenomenon and with time of pore formation.
Senlog Q 2003
127
DEPOSITIONALTEXTURE RECOGNIZABLE
Table 3-4
Comparison between porosity in sandstones and limestones (from Choquette & Pray, 1970).
Sandstone
Amount ofprinury porosity
Commonly 2540%
Cartmuart
Commonly 40-700/,
in wdimrnlr
Amount of
Commonly
Commonly none or only small
ultimate pros- half or more of fraction of initial porosity; 5ity in rocks
initialporosity; ISTO common in raavoir
IS-30%
mon
kiyure 3-25 - The Dunham's classification of carbonate rocks based on
com- facia
depositional texture.
It is obvious that porosity and permeability depend on
texture as illustrated by Fig. 3-26 which shows their
dependence on the texture of the carbonate rocks.
But, in these rocks we have to distinguish the original
characteristics, i.e. those existing at the time of deposition, from present characteristics. In fact, the original texture may have been deeply modified as a result of diagenetic phenomena that are often precocious and more
important in these formations than in detrital series.
Consequently, these modifications in texture bring about a
change in the reservoir characteristics themselves.
This is why these characteristics arise more from diagenetic phenomena than from texture, and why the study
of porosity is so essential.
Table 3-3
Different types of porosity in carbonate rocks
(adapted from Choquette & Pray, 1970).
rype(r) of pri- Almost exdumry porosity d w l y interpanide
interpanick commonly prcdominates, but intrapankle
and other t y p a arc imporrant
Typa(r) of ulti- A l W udu- Widdy varied bouur af portmate porosity d v d y primary depositio~lmodifiutionr
interpanick
Sim of'pores
Diameter and Dumer and throat rimcanthroat rim
d
y show littk relation to
closely d 8 t 4 aalimcnury panMc Mac or
t o d m t n c . r y rortin#
panick rim
and m
i
n
g
Shape ofpomg S t r o y &pen- G m t l y varicd. nrya from
darc on par- a r w y dcpmd+nt "paitiw"
tick s h a m or "nqativc" af pnicks to
"nytivc"
of form canpkcdy i
w
panida
of *pa
of dcporlunul or
dLpncciccmponau
Variable. rangiq from fairly
Uniformity of C a n m o d y
dzc.rhpe,and fairly uniform uniform to c x t r m d y hetwo-
distribution
I n f l u e m of
diapcncrir
within homo- gemus,
even within body
rock type
of a*
p n a body
~
nude up
Minor: u w l l y
minor reductionofprimary
porosity. by
Major: a n auk oblilcncc
or c a n p k t d y modify porority;
cmwnution and rdtuion imporunt
compaalon
and cementation
TIME
TYPE
lnflucna of
fracturins
Generally not Of major imponancc in maexof nu* im- voir propertia if proms
poruncc in
rcauvoir propCrtiCS
128
Variable. mniquantiutive vtrual a c i m a t a ransc from e u y
to vinurlly impossible; inslrumcnt mearunmcntr of poroc
ity. permeability and capillary
prruurccommonly nscdcd
Visual evaluation of porosity
and permcability
Semiquantiutivc visual minutes mmonly rd.tivdy easy
Adequacy of
core analysis
for rcrcrvoir
evaluation
Core plugs of Core plugs commonly itude-
Permeabilityporosity inter.
reliltions
Relatively con- Grcacly varicd; commonly insistent; com- dependent of panide size and
monly drpcn- w i n s
dent pn panicle
size and wnins
I-in. &ma quPre;mnw hok cora(-3-in.
commonlyadr- &meter) n u y be inadequate
quate for "nu- for large pom
trix" porosity
Serralog 0 2003
Texture
I Chanter 3 I
129
determination of size of grains or pores as small as a
micron.
-
Fig. 3-28 Thin section of a compact sandstone.
-
-
Figure 3-26 Porosity permeability relationship as a function of
the carbonate texture and its modification by cementation, dissolution
and fracturation (fmm Nurmi, 1986).
Determination of textural parameters
Traditional approach
Geologists determine the textural parameters through
different methods summarized in Figure 3-27.
Grain diameter in millimeters(kgarithmic scale)
-
Figure 3-27 Grain-size analysis methods
(adapted fmm KNmbein 8, Sloss, 1963).
The most common method for the determinationof the
grain size is sieving. This method is essentially used to
determine the grain size distribution or the sorting.
If the grains are large enough, direct measurements
are possible. In that case the measurement is realized following three perpendicular directions, called intercepts
(Fig. 3-8):longest dimension (a), intermediate (b) and
shortest dimension (c).
For smaller particles, their size can be also estimated
from their settling velocity which varies as the square of
the particle diameter (Stokes’s law), but also as a function
of their density.
For very small grains, examination of thin sections
through microscope (Fig. 3-28) or scanning electron
microscope (Fig. 3-29)are used. The latter allows the
Serralog Q 2003
-
Fig. 3-29 SEM photograph showing the grain size, the sorting, the
presence of authigenic clays coating grains or filling partly the pore
space.
Staining of thin section by Alizarin Red S (Fig. 3-30)
allows the distinction between calcite and dolomite grains
or cement. This technique is fast, easy and cheap.
Through thin sections or SEM photographs many textural as well as mineralogical parameters can be determined (shape of grains, grain overgrowth, sorting and packing, orientation... Fig. 3-31).This study can be completed
by energy dispersive analysis (EDAX) or electron microprobe analysis.
By injection of colored fluid (epoxy) the pore system
can be visualized (Fig. 3-32)and analyzed.
Capillary pressure measurementscan also be used for
evaluation of pore size and consequently, grain size.
Cathodoluminescencetechnique is also used to obtain
129
130
Well Logging and Geology
information on the spatial distribution of trace elements in
clastic grains and cements (Fig. 3-33). The differences in
luminescence is a function of subtle differences in the
trace elements composition.
In carbonate rocks, particle size and nature, cement
mineralogy (calcite, dolomite...) and type (micritic or sparitic) can be well described through thin sections and SEM
photographs (Figs. 3-34 to 3-36).
-
Figure 3-32 The porosity of this mature sandstone is well shown by
the epoxy (in green).
J
Figure 3-30 - Staining of a thin section allowing distinction between
calcite (stained in red) and dolomite remaining
unaffected (photo 0.06 mm large).
Figure 3-31 - (a) On this thin section quartz overgrowth is well detected. The original quartz grain, well rounded, is emphasized by dark
dots or “dust rims”. The overgrowth is marked by the euhedral crystal
shape around this grain. Calcite cement is also present (orange crystals). Pore space is in blue.
(b) This SEM photograph shows very clearly the quartz overgrowth
characterized by the euhedral crystal shape.
-
PI
Figure 3-33 On the left, a transmitted light through a thin section
shows a sutured angular quartz arenite which can be interpreted as
due to compaction. On the right, the photo, taken with cathodoluminescence, shows quartz overgrowth of well rounded quartz grains (in red
or light blue).
9
Figure 3-34 a): Thin section showing intra- and inter-granular porosity
(in blue epoxy) in an oolitic limestone. Micrite cement is around oolites.
Some sparitic cement (in white) is also present. b): SEM photo of oolitic sand. c): SEM photo of chalk. d): Thin section showing intergranular
porosity (in black) in an oolitic limestone.
-
130
Senalog Q 2003
Texture
I Chapter 3 I
d
Figure 3-35 - Thin section showing a very high porosity with large pore
throats indicating a high permeability in this carbonate.
Determination of textural parameters
Well logging approach
The influence of textural components on log parameters is summarized in Table 3-5.
Thus, it clearly appears that some log measurements
contain within themselves textural information. But, it is
131
3
-
Figure 3-36 SEM photograph of a dolostone. Shape and size of
crystals and pore size are easily determined.
not always simple to determine the origin of influence
because several textural parameters may have similar
effects. However, if we reexamine the problem, according
to the type of rock, it is sometimes possible to perceive
the privileged influence of certain parameters.
In order to extract information on the textural parameters themselves, tools of high vertical resolution are nee-
Table 3-5
Relationship between textural parameters and well logging attributes (from Serra, 1984).
Textural Parameters
fsiie
Petrophysical reservoir
characteristics depending
on textural parameters
Well logging
parameters
affected
7
Relevant
Logging tools
IL. AIT, LL. ARI, SFL, ML, MLL, MSFL. FMSFMI
FDC, LOT
CNL
Sonic, BHTV
TDT, GST, GLT
EPT
LOT
FMMl
FMSIFMI, EPT, Sonic
EFT, MSFL
IL AlT, LL. ARI, SFL, ML MLL, MSFL, W F M I
Tortuosity
Sonic
EPT
Sizeof pores and throats
which ccnbds:
IL, AIT, LL, ARI, SFL. ML. MLL MSFL. FMS/FMl
LDT, CNL, Sonic,EPT, TDT, GST
Total penne&lity
Relative permeability
lneduable saturabbn
CMR
IL, AIT, LL. ARI, SFL, ML, MLL, MSFL
-Anisolrow
Horizontal pemeability
Veldical permeability
Weaabilily
A
F W M I , UBI
n. (%)in
b
1 :geologist's definition
Serralog Q 2003
131
132
Well Logging and Geology
ded as the geological textural objects (grains, particles,
crystals) have a small size.
With standard tools, quantitative well log information is
acquired along one generatrix of the borehole (microlog,
microlaterolog, microsphericallyfocused log, litho-density,
electromagnetic propagation tool) or volumetrically
around the borehole (spontaneous potential, resistivity,
conductivity, radioactivity, neutron, sonic). The recorded
measurements by the standard tools correspond to an
averaged measurementthat views the formation as apparently homogeneous. They are essentially affected by
grain and matrix composition but also by porosity and permeability which depend on the textural parameters. It is
the reason why it was possible to extract information on
grain size evolution with depth by looking at ramps on the
SP, gamma ray and other logs (Fig. 3-37), or information
on sorting by taking into account the porosity level : high
porosity in a sand indicates a well sorted sand without
cement (Fig. 3-38). But, due to their lack of vertical resolution, standard logs will not inform directly on textural
parameters. In fact, the effects of those parameters on
porosity and permeability fundamentally affect the tool
responses.
1
-
Figure 3-38 Change of sorting detected from porosity change.
With dipmeter tools, which analyze along 4, 6, or 8
generatrices (Figs. 3-39 & 3-40), or with density which
analyze in 4 sectors (Fig. 3-43d), and more especially
with imaging technology, which analyze formations following 12 sectors'(AR1or HALS, Fig. 3-43b), or along a lot
of generatrices (FMI Formation Micro Imager, Formation
Microscanner FMS, STAR and EM1 tools Figs. 3-41),
even during drilling (RAB tool, Fig. 3-43a), or all around
the borehole wall (UBI [Fig. 3-43~1,CBIL, CAST), it is possible to detect heterogeneities either radially out into the
formation or circumferentially around the wellbore. The
size of the electrodes (1 cm for SHDT or 0.5 cm for FMI)
or of the ultrasonic beam allows the detection of very
small objects.
J
Figure 3-39 - Schematic of the SHOT-B sonde
(courtesy of Schlumberger).
MQI
Figure 3-37 -Grain size evolution detected from gamma ray log
(from Sera 8 Sulpice, 1975).
I32
d
Figure 3-40 - Photographs of dipmeter tools.
Serralog @ 2003
Texture
C
I
Figure 3-41 - Photographs of imagery tools.
I Chapter 3 I 133
d
-
Figure 3-43 c): Schematic of the UBI (Ultrasonic Borehole Imager)
tool; d): Schematic of the ADN fool (courtesy of Schlumberger).
The determination of the textural parameters from
standard and high resolution logging tools will be illustrated now.
PAD
Detrital rock texture
Grain Size
There is no general universal relationship between the
grain size and standard log measurement. Nevertheless,
in a number of cases we often observe, regionally, a very
clear correlation between log and grain size.
Different authors (Sarma et a/., 1963; Alger, 1966) showed the existence of a correlation between resistivity, or
formation factor, and grain size (Fig. 3-44).
FLAP
-
Figure 3-42 Pad and Flap photo and description of the FMI tool
(courtesy of Schlumberger).
a
b
Figure 3-44 - Relationship between formation factor, resistivity and
mean grain diameter (from Sanna & Rao, 1963).
Remark
-
Figure 3-43 a): Schematic of the
tool. b): schematic of the ARI
tool (courtesy of Schlumberger).
Serrelog Q 2003
In the example of this figure, for the same value of R, the formation factor increases when the arain size increases. This seems to
contradict the commonly accepted relation. This situation is undoubtedly due to the fact that infresh water formations the importanceof surface conductance increases when the grain size decreases.
133
Well Logging and Geology
134
Other authors have observed relationship between a
well logging parameter and grain size.
In the example of Fig. 3-37it is obvious that a correlation exists between gamma-ray and grain size measured
on core samples. Gamma-ray increases when grain size
decreases, because radioactivity is linked with the finest
grains. These fine grains consist of clay minerals and silt
size heavy minerals (zircon for instance). Further analysis
shows that these minerals are fundamentally of detrital (or
allogenic) origin and are of structural or laminated type.
Indeed, it is hardly conceivable that authigenic clays occupying the pore space would show such evolution, because in this case the percentage in the rock does not significantly evolve with granulometry.
trometry of natural gamma radiation), Pb, or SP (Figs. 351 and Fig. 3-52,see next pages), at the intersection
point of the two other values on the cross-plot.
*E
Sb
8
D(PS
CORRELATIONS
RESISTIVITY
CURVES
The "bell or "funnel shapes of spontaneous potential, gamma-ray, or resistivity curves (Fig. 3-45) introduced by SHELL geologists around 1956,are another application of this relation between a curve and a textural parameter.
In these cases if we cannot precisely define the absolute grain size without preliminary calibration, we are
nevertheless always able to determine a relative size.
These shapes (Fig. 3-45),in fact, explain the normal graded-bedding ("fining upward") or reverse graded-bedding
("coarsening upward"). In other words we can conclude
that shale distribution is essentially of structural andlor
laminated type.
I'
wcMQIII.cToIu)(o
Figure 3-46 - Example of structural and partly laminated shales as they
can be recognized on dipmeter resistivity curves. The curve evolution
indicates fining-up sequences confirmed by core description.
Mpo
-
Figure 3-45 Relationship between SP curve shape and grain size or
shaliness (adapted from SHELL'S documents).
We can also convert these curves into shale percentage and into permeability if a preliminary calibration was
done with the help of core measurements (Fig. 3-48).
The relation between radioactivity and granulometry is
not always proved. It may happen that the silty levels are
more radioactive than shales (Fig. 3-49). In this case it is
necessary to have several log recordings, enabling the
grain size determination. The "cross-plot" combining two
logs, such as gamma ray and spontaneous potential
helps in this analysis, particularly if the technique of "Zplot" is used. This technique presents the plotting of the
value of a third parameter, such as GR (gamma ray) Th or
K (amount of thorium or potassium obtained from a spec134
-
Figure 3-47 This figure is an enlargement Of One SeqUenCe Of Figure
3-46 (between 834.5 and 837 ff).
Serralog @ 2003
Texture
I Chapter 3 I
135
Figure 3 4 8 - Schematic relationships between dipmeter resistivity on
one hand, and grain size, shaliness and permeability on the other
hand. These relationships are applicable to the case illustrated
by Figure 3-46.
If the thickness of each granulometric sequence is
very small the logging macrodevices cannot detect these
tools having a resolution respectively of 1 cm and 0.5 cm.
tant value of neutron and sonic’ at the same time.
-
Figure 3 4 9 Example of silts (green shading) being more radioactive
than shales. The real grain size evolution can still be detected on resistivity, SP and neutron curves ( h m Sera & Sulpice, 1975).
It is preferable, then, to use the resistivity curves of
dipmeter tools (HDT or SHDT) or image tools. As previously seen, their high vertical resolution allows the
detection of sequences of a few centimeters thickness
(Fig.
, - 3-46).
If a poorly-consolidated detrital reservoir contains
hydrocarbons* we can Observe the existence Of a good
correlation between grain size and irreducible water-saturation (Fig. 3-53). The latter depends, in fact, on permeability that depends, in turn, on grain size (cf. Fig. 3-18b).
Serralog 0 2003
K
Figure 3-52 - Examples of cross-plots of SSP vs thorium (a), and
potassium (b) contents. They define the Ths. and Ksh for shales. The
and potassium
of ucleanmsands suggest the presence of radioactive minerals such as feldspars, micas and zircon.
They show the existence of silts more radioactive than shales.
135
Next Page
136
Well Logging and Geology
Fig. 3-51 - Crossplots of pb vs @ with frequency (a), thorium content (b), potassium content (c), and thorium-potassium ratio (d), on the same
interval than Fig. 3-50a, and their interpretation.
Figure 3-55b illustrates the relationship that exists between grain size, its geometric mean and the permeability.
Lastly, the invasion diameter, as it can be deduced
from a microlog, may constitute a good indicator of grain
size evolution. The example of Fig. 3-54 shows an evolution of microlog resistivity curves which can be correlated
with changes in grain size (normal graded-bedding in this
case), - the porosity being the same in this interval -,
these changes affecting the diameter of invasion.
But, the best tools for determination of grain size and
grain shape are the dipmeter tools and more fundamentaly the image tools. Their very high vertical resolution
allows the detection of objects as small as 1 cm for dipmeters and 0.5 cm for image tools. This is illustrated by
Figures 3-55 to 3-57.
When pebbles are present and larger than the electrode size (1 cm for dipmeters, 0.5 for imagery tools) they
can be detected by dipmeter tools or by the imagery tools.
The pebbles are generally more resistive than the surrounding matrix in which they are embedded.
Consequently, each pebble is distinguished by a peak of
resistivitywhose shape varies with the size, the proportion
and the arrangement of the pebbles. This gives a heterogeneous aspect to the curves with a practically total
absence of correlations between them (Fig. 3-55a at the
bottom). When the pebbles touch each other (grain supported conglomerate), the peaks are very close. When the
136
pebbles are isolated in a sandy or shaly-sandy matrix
(mud supported conglomerate), the peaks are isolated
(Fig. 3-55b). The other open-hole logs may indicate a
detritic formation dominated by quartz and often feldspars
and micas, or by pebbles originating from igneous rocks.
The NGS tool may be very useful to determine the type of
radioactive minerals.
a
b
Fig. 3-53 - a): Relationship between irreducible water saturation, porosity and grain size. b): relationship between grain size (in 4 scale), its
geometric mean and permeability (from Krumbein & Monk, 1942).
Serralog 0 2003
Previous Page
Figure 3-54 - Observe the microlog resistivity curves. Their evolution reflects changes in invasion diameter, possibly in relation with grain size
decrease. Observe also the number of sequences which can be detected in this massive sandstone reservoir thanks to the microlog. The average
thickness of sequence (between I and 2 m) suggests a turbidite deposit.
Isolated resistive peaks showing variable thicknesses
from button to button (Fig. 3-5513) indicate irregular particle size and angular shape with rapid thickness change.
or to a reef or packstone if it is observed in a carbonate
deposit.
When grains are coarser than 5 mm, electrical images
allow their detection and the determination of their shape,
size and frequency.
For example, very resistive well rounded white spots,
often in contact with others (Fig. 3-56), correspond generally to grain-supported conglomerates if the other standard log data indicate detrital siliciclastic deposits.
When the white spots have an irregular shape and size
and are isolated in a more conductive background they
indicate a mud-supported conglomerate or a debris flow
(Fig. 3-57).
Very conductive peaks (black spots or cubes) correspond to very conductive crystals, generally of pyrite or
galena (Fig. 3-58).
Black rounded spots in carbonates correspond most
probably to molds of oolites which have been dissolved by
selected leaching of particles (oolites) composed of more
unstable minerals (aragonite or impure calcite) (Fig. 3Serralog 02003
59).
Sorting
One can approach this parameter with a study of porosity evolution. In fact in sandy formation, at a given depth,
we can reasonably assume that arrangement and packing
are the same for all grain sizes, the porosity generally
decreasing when sorting decreases.
Fig. 3-60 gives an example of the change in sorting.
Levels 9 and 10 present an average porosity close to 35
%. Taking into account their depth (7000 feet) this high
porosity is certainly due to a good sorting. The low radioactivity of these levels combined with a strong deflection
of spontaneous potential seems to indicate the presence
of radioactive minerals other than shales. Granulometry of
these levels must be fine to very fine. Level 11 shows a
lower porosity (25 %) with less radioactivity and an identical spontaneous potential. This drop in porosity, seems,
therefore to indicate a poorly sorted sand (coarse to fine
grained).
Again, the best way to analyze this phenomenon is to
make a cross-plot (type Z-plot), combining the hydrogen137
neutron index and the density with radioactivity, spontaneous potential values, or the amount of thorium or potassium. On such cross-plot (Fig. 3-61), from the point defining the maximum porosity for a given interval (corresponding to the best sorting), a drop in porosity along the
sandstone line (not less than 15%), represents a decrease in the level of sorting.
measurements and from "red" or "blue" patterns on arrow
plots.
Varve?
CORE DESCRIPTION
Glacial
outwash
deposit
Boulder
Figure 3-55b - Observe towards 144 m on pads I , 2 and 3 the thickness variations of the peaks. Their thickness evolves from 1 m on pad
2 to 0 m on pad 4. This indicates a block or boulder which does not
cross the well. Such blocks are more probably carried by ice and deposited when the ice melts
Pad orientation
Figure 3-55a - Grain-supported conglomerate at the bottom, evolving to
sand with pebbles, sands laminated and homogeneous, and shale at the
top. The core description confirms this interpretation of the HDT curves.
Grain orienfafion
A preferential orientation of grains must theoretically
be indicated by an anisotropy in resistivity reflecting an
anisotropy in the permeability from which it is derived.
But, the contrast between vertical and horizontal resistivities being small (around 1 3 , it seems tenuous to attribute any variation in the two horizontal axes to anisotropy
because other explanations can be responsible of this
variation (pad contact, mud-cake thickness...).
It does not Seem possible at the present time to
approach this phenomenon quantitatively, H
~ we
can approach it by means of relative deflections or, more
specifically, by directions of currents determined from dip
138
Figure 3-56 - FMS example of grain-supported conglomerate. Observe
the well rounded white spots which are locally in contact with similar
others.
They
to 4 or 5 cm.
~
~ correspond
~ to pebbles
~ with a ~diameter close
,
The dark spots correspond to pebbles pulled out from the formation by
rock bit. The background corresponds to a sandy formation.
Serralog
2003
Texture
I Chapter 3 I 139
U
Figure 3-59 - Thin section showing oomoldic porosity (left). On the right
observe the shape and the size of the black spots on the FMI images
(courtesy of Schlumberger).
A
Figure 3-57 - Example of mud-supported conglomerate confined by
core photographs reproduced on the side. A coarsening-up sequence
can be deduced from the increase in size of the resistive features from
XI27 to X118 m. This vertical grading could not be seen in core since
large cobbles look like beds. Observe on images the rapid changes in
thickness of the white spots and their shape. This intervalk can be
interpreted as corresponding to debris flow
(from Karker et al., 1988).
J
Fig. 3-60 -Observe the change in porosity from levels 9 and 10 and 11.
It corresponds to a change in sorting. The two lower sands are fine,
clean, well sorted, slightly radioactive; the upper sand is coarser and
badly sorted (from Serra 8 Sulpice, 1975).
4
Figure 3-58 - Example of pyrite crystals detected on FMS resistivity
curves (conductive troughs) and on image (very dark cubic spots).
Orientation of pebbles can be sometimes observed on
images (Fig. 3-62). This can be interpreted for the recognition of the transport current direction (cf. Fig. 3-1 3).
Arrangement or Packing
This parameter is not accessible, because we can reasonably admit that after a burying of several hundred metres, packing of sediment is such that arrangement of
grains becomes more compact (or closed).
Serralog (8 2003
J
Figure 3-61 - Sorting decrease, grain size evolution and type of shale
distribution can be predicted from cross-plot analysis.
Remark
This porosity drop could correspond to a cementation by precipitation of silica. Even if this hypothesis cannot be totally rejected, it is however unlikely if we consider the existing high porosity.
139
Well Logging and Geology
140
tool should enable determinationof the orientation of mica
flakes and consequently give an indication of the existence of horizontal or vertical permeability barriers in a micaceous sand.
As we have seen, electrical images allow the recognition of the object shape as soon as their size is higher
than 5 mm. For instance, well rounded pebbles, rock fragments or cubic crystals of pyrite can be recognized (cf.
Figs. 3-56 to 3-58). In chalk, chert nodules are well recognized on image thanks to their irregular shape (Fig. 363).
J
-
Figure 3-62 On the left, inclined pebbles (white spots) can be observed in this mud-supported conglomerate confirmed by the core photograph. On the right another FMS example of mud supported conglomerat (fmm Serra, 1989).
The study of porosity within a short interval, or even
more understandably at any given point, cannot reveal
information on packing. All porosity variations can be
explained, in a plausible manner, by a change in sorting
or by diagenetic effects (cementation, dissolution).
On the other hand the evolution of porosity with depth
on a long interval will explain the modification of packing
under the effect of compaction, and/or of diagenesis. This
aspect will be analysed with the study of compaction.
Grain shape
If the sandy formation is chemically immature and, therefore, rich in feldspars, mica etc .... shown by recordings
of litho-density and by natural gamma ray spectrometry
(relatively high amounts of potassium), we can deduce
the existence of a textural immaturity and consequently of
angular grains. On the other hand, if the sand appears to
be very clean, with a very low radioactivity and high porosity, we can suppose the existence of a chemically and
texturally mature sand, hence probably well-washed or
winnowed and well-sorted sand with round spherical
grains.
From Sen (1980, 1981), the grain shape has a strong
influence when the electromagnetic field is applied perpendicularly to the flakes (micas). Thus, sands rich in
mica flakes should show a higher dielectric constant than
the one expected from the measured values for quartz
and mica. The analysis of the data recorded by the EPT
140
d
-
Figure 3-63 Chert nodules (white spots) in a chalk. Observe the
inegular shape of the white spots (courtesy of Schlumberger).
Cement
The cement proportion is more or less important. As
well known, porosity will be lower due to cementation. In
case of cementation, it is, therefore, not always easy to
draw certain obvious conclusions concerning the grains.
However, for the same content of cement, a decrease in
porosity must be attributed to a diminution of sorting.
The type of cement can be determined from the mineralogical composition of the rock, obtained from log analysis. In the case of ambiguity (calcitic or dolomitic
cement), we have to remember that because cementation
is always a postdepositional phenomenon, the volume of
cement added to the porosity value cannot exceed the
maximum porosity value that existed at the beginning of
the cementation process.
Thus, if we compute from cross-plots a percentage of
calcite cement which is higher than usual, either a hypothesis of dolomitic cement must be accepted, or, if the
existence of calcitic cement is confirmed from other data
(Pe from lithodensity tool LDT, cutting analysis), the presence of detrital limestone particles (bioclasts, oolites...),
associated with grains of quartz or feldspars must be
considered.
Serralog Q 2003
Texture
I Chapter3 I 141
Carbonate rock texture
We know that in this type of formation the precocious
diagenetic effects have a tendency to completely modify
the initial texture. Therefore, in this particular case, it will,
exceptionally, be possible to obtain information on texture.
The dipmeter resistivity curves after correlation with
cores, may indicate some types of texture. Fig. 3-64
shows that calcirudites or boundstones, calcarenites or
grainstones and calcilutites or mudstones can be recognized from the shape of resistivity curves.
As shown by Fig. 3-65 an excellent agreement between the core description given in terms of constituent
percentages and one resistivity curve is obvious. It allows
a very detailed description of the formation in terms of
mudstone, wackestone, packstone, grainstone and even
boundstone. It is of considerable interest for the palaeogeographic reconstructionof a reefal environment, especially if natural gamma ray spectrometry (NGS tool) data,
the photoelectric index (Pe), and the density measurement (LDT tool) are available. The uranium content measured by the NGS tool will reflect reducing (uranium present), or oxidizing (uranium absent) conditions, which in
turn may help to locate the back- or the fore-reef. The Pe
will indicate the importance of dolomitization.
By comparison with core analysis an empirical relationship has been established between (a) curve and dip
parameters, and (b) facies in carbonates. It is reproduced
in Table 3-6.
A high porosity accompanied by shallow invasion, if it
is continuous along a certain interval, can correspond to
chalk. Of course, the other open-hole logs must indicate a
limestone composition.
Foreset laminae detected on images in carbonates
indicate grainstone texture as foreset beds correspond to
grains, consequently to oolites or shell fragments carried
by current of high energy (Fig. 3-66 see next pages).
Other carbonate textures can be detected from image
analysis as illustrated by Figures 3-67and 3-68.
Porosity
The porosity of detrital siliciclastic deposits results
from textural parameter influence. As it is obvious, a very
well sorted sand will present a higher porosity than a
badly sorted sand; a cemented sandstone will be less
porous than a sand. Consequently, as soon as the porosity is evaluated it is important to extract as much as possible information linked to textural parameters: sorting,
and cementation.
The existence of vugs is sometimes easier to determine. With the help of the electrical images, the vugs can be
even seen and their percentage evaluated (Fig. 3-69).
They correspond to conductive peaks on the resistivity
curves, or to dark irregular spots on the images. The other
open-hole logs will confirm the carbonate nature of the
rock.
Semlog (9 2003
I
Ern
I*)
TEXTURAL
CURVES
RESISTWITY
INTERPRETATION
a
-
82
84
1
-
figure 3-64 Exam. I of HOT response in a m f a l environment. Note
that the aspect of the dipmeter resistiwify curves enables separation of
boundstone at the base (characterizedby peaks of variable thicknesses and absence of correlations), grainstone at the center (marked by
very small curve activities) and mudstone (characterized by high resistivity without variations).
Table 3-6
Relationship between dipmeter signatures and
carbonate texture (from Theys et a/., 1983).
141
porous
coral branch
not porous
Figure 3-65 - Comparison of textural interpretation of HOT micro-resistivity curves with core description (from Schlumberger, Well Evaluation
Conference, India, 1983).
XX15
2R
XX17
Figure 3-68 - Image which can be interpreted as representing boundstone texture. Observe the vertical distribution of white spots which can
correspond to reefal branchs.
Analysis of the images allows in many cases to determine the type of porosity distribution as illustrated by
Figures 3-69 to 3-72.
The vugs can be connected to secondary porosity due
to the dissolution of calcite or dolomite crystals by water
circulation. We know that, in this case, the sonic tool does
not (( see >) this secondary porosity. Consequently, a comparison of porosity determined from the combination of
density and neutron tools (tools that (( see >> the total porosity connected or not), with porosity deduced from sonic
tool will indicate this vuggy secondary porosity (Fig. 373a). This secondary porosity index (SPI) is computed
and reproduced alongside the left track in computational
result display (Fig. 3-75). For this purpose we can also
use the M-N-plot, or the MID-plot techniques (Fig. 3-73b).
Figure 3-67 - Example of mudstone texture in which several stylolites
can be observed.
142
Serralog 0 2003
Texture
I Chapter3 I 143
North
Figure 3-66 - Example of images in grainstone. The foresets observed in this limestone cannot be explained if they are not associated to particles
carried by a current. These particles are oolites or shell fragments. The direction of the transport currents is determined from the azimuth of the
foreset dips. This can be exploited to deduce the direction of the elongation of the oolitic bars and the permeability anisotropy
Serralog 0 2003
143
Next Page
A
Figure 3-69 - On the left, FMI example of pore distribution which can
be interpreted as illustrating fenestral porosity and alga mound deposit.
On the right, photo of algal mat with bird’s-eye pores.
Figure 3-70 - On the left, FMI example of vuggy porosity (black spots)
with anhydrite nodules (white spots) in a dolostone. On the right photo
of similar rock type.
Figure 3-72 - Vugs (black spots on the left images) are isolated by
thresholding of the resistivity (right images). The percentage of the
surface occupied by the vugs can be computed ( 72% in this case)
(courtesy of Schlumberger).
I
I
Figure 3-71 - On the left-up, FMI image of vuggy porosity. The shape
of the vugs suggests an oomoldic porosity most of the time connected
to others and to background porosity. On the right, photo of a thin
section showing a similar oomoldic porosity (in pink).
The lower figures show a SEM photo of the same oomoldic porosity
and a vuggy porosity on the right.
In fact, the first arrival corresponds to the one, the trajectory of
which, has missed most of the vugs. At best, the secondary porosity
index (sp~)
represents a minimal
value of the secondary porosity,
It
would be more correct to refer to an index of heterogeneity of pore distribution, which, of course is dependent on the presence of secondary
porosity. A regular distribution of small vugs would generate a nil secondary porosity index which would mean that the sonic tool would see the
total porosity.
144
Fig. 375% - Detection of secondary porosity by comparison of porosities computed from sonic tool, with those deduced from the neutrondensity combination.
Serralog 0 2003
Previous Page
cate an interpretation method of the measurement and
extract an evolution of the micro-geometry and of the
cementation factor with depth in carbonate series.
Rasmus & Kenyon (1985) developed a technique for
estimating separately the amounts of intergranular and
oomoldic water. They used the results to predict the presence of oil and its flow rate in carbonates.
Primary Medium
Distributed
porosity
Jncluded Spherical Pores
0.8
Figure 3-73b - Cross-plot M vs N for mineral and secondaty porosity
determination (courtesy of Schlumberger).
Electrically
isolated pores Electrically
connected pores -
Similarly to the sonic travel time, the resistivity is also
affected by the distribution of the pores in the rocks. The
resistivity does not "see" the isolated or not connected
pores (Fig. 3-74). Neither is it influenced by the size of the
pores, but only by the network of the capillaries which
connect them.
Brie et a/., (1985) developed a method of analysis of
acoustic and electric measurements, based on KusterToksoz model for acoustic velocities, and the MaxwellGarnett model for resistivity, which determines the
amount of spherical (oomoldic) porosity in carbonate
rocks. In addition, they proposed a way of evaluating the
cementation factor level by level from the sonic indication
of spherical porosity (Fig. 3-75). This value can be used to
obtain a more accurate value of the water saturation in
such reservoirs.They also mentioned that the accuracy of
the model could be increased by combining it with other
texture sensitive measurements such as dielectric permittivity.
From recent works it seems that dielectric measurements made by the electromagnetic propagation tool
(EPT tool), combined with data extracted from microresistivity devices, can help to determine the microgeometry of
a carbonate rock.
Kenyon (1984), Kenyon & Baker (1984) demonstrated
the need to introduce a "bimodal model" taking into
account the micro-geometry of the rock (shape of pores
and grains) to explain the EPT tool response. They advoSerralog 0 2003
Fig. 3-74 - Schematic representation of the effect of vugs (supposed to
be more or less spherical or oomoldic) on the electric properties.
Isolated pores are in blue (from Brie et al., 1985).
In conclusion, one can emphasize the importance of
the knowledge of the texture for permeability estimation
and for a more accurate saturation computation, especially in carbonate reservoirs. This knowledge will also
help in the quantitative interpretation by favouring the
choice of the value of the rn factor of the Archie's equation.
Porosity and pore size determination from nuclear
magnetic resonance measurements
It is important to remember that the transverse relaxation time T, measured by a nuclear magnetic resonance
tool is lithology independant and is a function of the hydrogen nuclei content in the pore fluid. T, is proportional to
V/S (V = volume, S = surface) which in turn is proportional to pore size (Fig. 3-76). For instance, for a spherical
pore SN = 3/r, where r is the radius of the pore.
145
Figure 3-76 - The decaying signal amplitude (top) is the sum of all the
decaying T2 signals generated by hydrogen protons in the measurement volume. Separating out ranges of T2 values by a mathematical
inversion process produces the T2 distribution (bottom). The curve
represents the distribution of pore sizes, and the area under the curve
is the porosity (courtesy of Schlumberger).
Fig. 3-75 - Example of results of spherical porosity computation
(from Brie et al., 1985).
Interpretation of the pore size distribution and the logarithmic mean T2 are used to calculate parameters such as
permeability and free-fluid porosity.
In siliciclastic deposits, the threshold is generally put at
33 msec (Fig. 3-77). for irreducible water saturation. In
carbonate rocks the threshold for vuggy porosity is put at
100 msec (Fig. 3-78). Figure 3-79 illustrates the influence
of pore size on the relaxation time.
146
3
Figure 3-77 - The area under the T2 curves to the left of 33-msec cutoff corresponds to very small pores with irreducible water saturation.
From 33 to 210 msec, the raea under the curve represents producible
fluid and pore size increase. Above 210 msec, the area represents oil
(courtesy of Schlumberger).
Serralog 0 2003
shows the validity of this technique (Fig. 3-80).
Pore diameter, microns
Time, msec
Figure 3-80 - The T2 time is calibrated in pore diameter
(courtesy of Schlumberger).
Automatic extraction of textural information from
dipmeter curves
J
Figure 3-78 - In this carbonate, the area below 100 msec represents
very small pores with irreducible water saturation. The area between
100 and 750 msec represents largerpores and vugs. Above 750 msec
the area may correspond to oil (courtesy of Schlumberger).
As previously seen, dipmeters reflect the internal organization of formations. This is illustrated by Figures 3-82.
In order to extract and quantify this information a program
was written (Delhomme & Serra, 1984). Its name is SYNDIP. An example of its results is given in Figure 3-83.
J
Figure 3-79 - Grain surface relaxation is the most important process affecting TI and T2 relaxation times. When the pores are small the relaxation
is rapid as the probability of collisions with grains is higher. When the pores are large this probability is lower.
The T, curve can be converted into pore diameter (Fig.
3-80). The width can be interpreted as reflecting the sorting. A sharp peak on T, above 33 msec, may represent
a deposit with well sorted grains and pores of similar size.
The threshold value at 33 msec has been established
from experiments.A large number of water-saturated core
samples from two wells was analyzed for porosity determination. The cores were centrifuged under 100 psi pressure to simulate draining down to typical reservoir capillary pressures. The amount of fluid drained (Fig. 3-81)
equals the free-fluid porosity, which is converted into an
equivalent area on the T, distribution curve. The area shaded from the right - determines
values.
A
parison of T, distribution taken before (curve in green)
and after (curve spun sample in purple) centrifuging
Serralog 0 2003
J
Figure 3-81 - Comparison of porosity extracted by centrifuging core
samples of two wells and free fluidporosity. Observe the good correspondance specially above 6 p.u. (courtesy of Schlumberger).
147
and dark when it reads the mud). Such effects are easily
recognized, and the intervals can be zoned out on the
affected pads during the image analysis.
Depth
(m)
Calibrated
dipmeter
resistivity
1
Figure 3-82 - The resistivity curves of dipmeter tools reflect the internal
organization of formations. Cutves flat, without character, indicate an
apparent homogeneous formation. Curves showing high activity
(numerous deflections or breaks, FBR) without evident correlations
indicate a heterogeneous formation (pebbly, vuggy or reefal, bioturbations). Curves showing high activity with high density of correlations
(DCL) indicate a thinly laminated formation.
(from Delhomme 8 Serra, 1984).
But, even if the resistivity dipmeter curves can reflect
formations heterogeneities like pebbles or vugs, images
are the corner stone of the textural determination. They
inform us about the internal organization of each depositional unit reflecting partly the energy in the medium of
deposition. Their analysis may be visual but can be as
well realized by their processing using a software. The latter simulates what the eyes can achieve. It is why this processing is now explained in detail and is commercialized
under the name of BorTex (previously ScanTex or
TexScan, or SPOT) of which the goals and approach are
explained here below and in a paper written by J.P.
Delhomme.
BorTex Program
The gray- or color-level changes relate only to conductivity changes. Interpreting ambiguous images may require knowledge of the geological context or other log data
(Serra, 1989). Dark rounded areas may be vugs filled with
conductive drilling mud, clay clasts or pyrite crystals, or
even pebbles pulled out by the drilling bit. Dark linear features may indicate fractures filled with mud or another
conductive material such as clay or pyrite. Large-scale
image level variations may relate to changes in shaliness
or porosity. The same holds true for automated image
analysis - the image must be interpreted in context.
Although the current is strongly focused into the formation, electrical images can be adversely affected by
bad borehole conditions. In very rough holes, a blurred
image can result (white when the pad touches the wall
148
Figure 3-83 - Example of SYNDIP result display. In the left track are
reproduced in thin lines the minimum and maximum resistivities recorded by SHDT buttons, in heavy line the average value. The second
track displays the internal organization of the beds. The limiting curve
is the frequency of inflexion points. The three columns on the side, are
flags indicating: detection of non planar surfaces, non parallelism between consecutive bed boundaries, and correlated parallel planes. In
green strip are indicated the average dip determined from a minimum
of 5 consecutive dips with angular spherical dispersion lower than 5".
The next track shows the conductivity curve shaded with a color scale
to reinforce the sand/shale opposition in sandkhale series. The blue
strip width indicates the number of buttons having showed high resistivity. The last track displays thickness of conductive or resistive beds in
a two mirror logarithmic scales. The cumulated resistive thickness is
indicated on the left (courtesy of Schlumberger).
Serralog Q 2003
Texture
I
Chaoter3
Table 3-7
Determination of carbonate texture from well logs and images.
Standard t d response
lmage characters
texture
bhck spots (faU of chert nodules, mrcasite)
W close to 4 - 4.2 We
-ll___l
Variable porosity
Nbst of thetime visible laminations, foresets or crossbedding
Stylolites may be present if corqacted
-.
Faelatiiely hom~geneouspale yellow or w hte knages.
Few white spots on dynarric normalized knages (shell
fragments) .
Few dark spots or patches corresponding to isolated
-
-
"WS
___-
___--__
Styldies m y be present
(cf. Fg.3-67)
__
._
-
Irregular orange or while patches on enhanced images
Imagery tools provide a cylindrical view of the formation.
This view can be considered two-dimensional either by
unrolling the cylinder or separately visualizing the strips
covered by each pad. The measurement resolution and
spacing are small enough with respect to the size and
spacing of most formation heterogeneities that one can
expect to image these heterogeneities. The shapes and
spatial arrangements that emerge with the images complement traditional log analysis. The effect of heterogeneities on pad tool readings (e.g., density) and the lateral
extent of one-dimensional high-resolution logs can be
evaluated.
In siliciclastic rocks, images allow us to recognize
debris-flow deposits or grain-supported conglomerates
(cf. Fig. 3-56 and 3-57), the siliciclastic nature of the rock
being confirmed by standard open hole logs. The size of
the rock fragments or pebbles and the volume percentage
they represent (Fig. 3-84) are determined by an automatic analysis of the image through the computation of the
area they cover on the image and an assumption about
their shape (cubes, parallelepipeds, spheres or ellip
Serralog Q 2003
soids).
In carbonate rocks, the images allow us to classify
them in terms of grainstone (cf. Fig. 3-66), mudstone (cf.
Fig. 3-67) or boundstone (cf. Fig. 3-68), or in terms of
debris flow or breccia (Fig. 3-85). The reasoning to apply
for carbonate texture determination are summarized in
Table 3-7. It includes both standard logs and images.
The following text is a summary of a paper written by
J.P. Delhomme.
Zonation based on imaae texture
The computerized analysis of borehole images leads
to zonation of the formations by their internal structure
(Fig. 3-86). Five zone types can be identified from electrical images: massive, laminated, interwoven heterogeneous, heterogeneous with conductive inclusions, and
heterogeneous with resistive inclusions. This classification is frequently used in the literature (Nurmi et a/. 1990).
This zonation can be obtained from electrical images
and automated using image topological properties. More
149
specifically, one can refer to the feature that extends
across the image. In a full-coverage image, this imagecrossing part obviously describes a feature of the formation that crosses the well. When a pad tool records several image stripes, further sophistication is required to bridge the inter image gaps.
Assume that the borehole electrical image can first be
made binary with reference to some adaptive threshold.
This threshold should track the evolution of the median
conductivity versus depth (cf. Fig. 3-67).
To get a zonation, the next step is to compute sliding
parameters that characterize the resistive and conductive
components and capture their mutual spatial arrangement. Specifically, the median conductivity values of both
components, their area percentages before and after eliminating non-crossing regions, and the lateral variation of
the apparent thickness of the image-crossing regions
must be captured.
The final step is to apply a set of rules.
- A massive zone is an interval where the median
conductivity values of the two components - if both coexist
- are not significantly different.
- A laminated zone is an interval where the area
percentages of, for example, the conductive component
are not significantly different before and after elimination
of the non-crossing regions. This also includes intervals
where the apparent thickness of the image-crossing
regions does not show significant lateral variations.
- An interwoven heterogeneous zone is an interval where the area percentages of, for example, the
conductive component are not significantly different befo-
re and after elimination of the no-crossing regions, but
where the apparent thickness of the image-crossing
regions shows significant lateral variation.
- A heterogeneous zone with conductive inclusions within a resistive background is an interval where
the area percentage of the conductive component is significantly smaller after eliminating the non-crossing regions
than before.
- A heterogeneous zone with resistive inclusions
within a conductive background is an interval where the
area percentage of the conductive component is significantly greater after eliminating the non-crossing regions
than before.
More elaborate rules exist that avoid the binarization
stage and are directly applicable to gray-scale images,
but the basic principles stay the same as those explained
here.
Quantitative analvsis.
When the zonation is performed, further quantitative
information can be derived from images. For instance, in
an electrical image of a typical carbonate rock, one identifies local objects (e.g., vugs or molds, hairline cracks or
fractures, chert or anhydritic nodules). The host rock
containing the vugs or molds is sometimes called matrix
rock in logging terms. In this publication, it will be referred
to as background rock to avoid any confusion with the
geological term "matrix" (defined as the finest-grained
material [not cement] filling the spaces between coarser
material [grains]). This background may reveal layering,
large patches and interwoven patterns.
1
Figure 3-84a - Open-hole logs corresponding to the images reproduced in Figure 3-846.
150
Serralog 0 2003
Texture
1 Chapter 3 1 151
il
Figure 3-84b - Example of image analysis in order to evaluate the area covered by conductive (C) and resistive (R) objects. On the left, images
with identification of the pebbles (more resistive spots than the background). On the right, analysis of the entire interval. The red strip locates the
image reproduced on the left [from Delhomme 8 Motet, 1993).
One of the first goals of borehole electrical image analysis was to extract and to quantify voids such as vugs or
molds in carbonate rocks. Another goal was to study their
interconnections, which generally constitute only a small
fraction of the total pore volume, but exert a decisive
control on the permeability. This approach is similar to that
one proposed for core analysis (Lucia 1983). The portion
of the total pore volume that comprises vugs, cracks and
fractures is called secondary porosity. Hence, the name
given to the image processing prototype is SPOT, which
stands for Secondary Porosity Typing (Delhomme, 1992).
However, the technique is equally applicable to heterogeneous siliciclastic reservoirs such as sandstones with
nodules, clay chips (flasers), clay galls or clay lenses, and
to conglomerates in which the rock heterogeneities usually determine a high- or low-quality reservoir.
The image analysis is accomplished through a series
of image transforms (also called filters). The transformation of images into other images (called image processing) and the transformation of images into data (called
image analysis) are both used.the program uses mathematical morphology transforms that simplify image data
while preserving essential shape characteristics.
Figure 3-84c - Details of FMS images showing the microconglomeratic
nature of the sandstone. The microconglomerate is composed of small
irregular granules.
Serralog 0 2003
151
2689
background. This change can be detected by applying
derivation operators. Higher contrast lines in the original
image correspond to crest lines in the gradient image,
which is also considered a relief (Fig. 3-87).
Among all the crest lines, the watershed, (divide) lines
are the water-parting lines that separate two different altitude minima. In the gradient landscape, such divide lines
can clearly be tracked along the crests of truncated coneshaped relieves (volcanoes) (Fig. 3-88). The morphological dual of the watershed line network is the thalweg network, which corresponds to water-collecting lines of the
landscape metaphor.
The watershed extraction is a gray-scale morphology
transformation that, when combined with other gray-scale
morphology tools, can be used in many gray-scale segmentation problems. It is a nonparametric and global
algorithm.
Figure 3-85 - Example of breccia. The angular shape of the white
spots are fragments of rock. Th core photograph on the side comes
from a nearby well in the same formation (from Pilenko, 1988).
The individual geological features (conductives or
resistives) are first extracted from the images. Then, the
features are characterized with attributes related to size
(area, perimeter), shape and connectedness. Finally, the
information is analyzed and summarized through summary logs to compare and integrate the data with standard
(e.g., porosity) logs (cf. Fig. 3-84), well test data and production logging surveys. Obviously, the first stage is critical to this analysis.
The simplest way to segment the image is by establishing thresholds (cf. Fig. 3-72). The landscape metaphor,
often used for gray-scale images, consists of topographic
representationof the image. The numerical value (i.e., the
gray-level) at each pixel stands for the elevation or terrain
altitude at that point. If the objects to be extracted have a
distinctive gray-level range, the points with an altitude
lower (more resistive) or higher (more conductive) than a
given threshold can be cut off to separate objects from the
background. The threshold is critical because it dictates
the contour and size of all objects. When the objects have
different gray-levels, the thresholding method fails. The
threshold value that would be required to extract every
object will outline some objects much larger than expected.
A second approach identifies the change in gray-level
that occurs at the boundary between objects and the
152
J
Figure 3-86 - Example of zonation of a formation, 2.5 meters thick,
based on image.
To avoid over-segmentation related to low-contrast
boundaries, a strategy called marker-controlled segmentation has been used. Rather than removing irrelevant
contour elements, the gradient image is modified so that
the resulting divides correspond only to the desired
objects. This strategy makes it easier to use external
knowledge, such as geological knowledge about the features extracted from the images.
The automated image analysis is performed in three
stages:
marking and outlining objects, computing object attributes, and calculating attribute statistics and summarizing
the results as logs. These stages are outlined in the following sections.
Serralog 0 2003
I
Figure 3-87 - 3 0 landscape representation of conductive peaks.)( axis
= pad width, Y axis = depth, Z axis = conductivity or gradient
(from Delhomme, 1992).
The adaptive threshold is used to reject low-contrast
local extrema. It is set at the value of image plateaus
obtained from morphologically eliminating highs and lows
(Rivest eta/., 1991). Then, the objects are contoured by
tracking the major contrast lines (i.e., the major gradient
modulus divides) between markers.
The gradient image is modified using a morphological
reconstruction operation (Rivest et a/., 1991). This operation fills the lows in the gradient image between object
markers and background marker, leaving only the highest
gradient crest.
In the final step of contour detection, the standard
watershed extraction is performed on the morphologically
reconstructed gradient image to produce one close
contour per marker The quality of the segmentation
depends on the reliability of the markers, but the marking
procedure may vary with the specific segmentation problem. The parameter tuning, if needed, is restricted to the
marker generation step.
In summary, to extract any geological object (vugs,
nodules, lenses or fractures), six processing steps are
successively performed:
1. if needed, apply a pre-filter to remove noise
from the image
2. determine the object marker image
3. determine the background marker image
4. compute the gradient modulus image
5. reconstruct the gradient modulus image from
object and background markers
6. extract the gradient watershed to find the
objectlbackground boundaries.
I
Figure 3-88 - 3 0 landscape representation of maximum gradient. M can
be compared to a volcano (from Delhomme, 1992).
Markina and outlinina objects.
A marker, which is a connected set of pixels included
in the region, first is automatically extracted for each of
the different regions to be segmented. The object markers
must roughly indicate the location of the objects, and the
background marker must surround the objects. Object
and background markers cannot overlap. Different prefilters and object and background marking procedures are
specified for each class of geological objects. The choice
is based on a priori geological knowledge of the expected
attributes of the objects.
For conductive spots (e.g., vugs filled with conductive
mud), the object markers are the image local maxima
(summits) that are higher than an adaptive threshold. The
background marker is the image thalweg network. For
resistive spots (e.g., nodules or pebbles in conglomerates), the object markers are local minima (sinks) that are
lower than an adaptive threshold. The background marker
is the divide network.
Serralog 0 2003
Computina object attributes.
To characterize each object in the "cylindrical twodimensional" view at the borehole wall, object attributes,
(e.g., area percentage in the image, apparent size, shape,
contrast and orientation in the image) are computed for
each object.
Connection between objects is also a very important
attribute, and it raises a theoretical problem for gray-scale
images. In the landscape metaphor, the whole landscape
is a totally connected surface. Only by establishing thresholds will the intuitive notion of connectedness appear.
For instance, if the sea level rises, the emerged lands are
no longer contiguous, and islands and continents are isolated. In fact, watershed and thalweg lines summarize the
gray-scale connectedness. The altitude drops can
approach this, from summits to saddle points along the
watersheds or by the jumps from sinks to saddle points
along the thalwegs. This principle is used to define the
vug connectedness from electrical images. Conductive
and resistive paths characterize connectedness. The
conductive paths used to track inter-vug channels or fractures are defined as arcs of the image divide network for
which the minimum (saddle-point value) is higher than an
adaptive threshold. Conversely, the resistive paths are
arcs of the image thalweg network for which the maximum
(saddle-point value) is lower than an adaptive threshold.
153
154
Well Logging and Geology
Some attributes can be used directly to infer threedimensional characteristics, such as area percentages
that can be translated into volume. Other attributes, such
as orientation characteristics, can be converted to three
dimensions
using
geometrical
transforms.
Connectedness attributes are not directly translatable
from two to three dimensions, apart from convex or spherical models that do not apply to many geological objects.
For such attributes, the cylindrical view is used to predict
three-dimensional connectedness and anisotropy.
Calculatina attribute statistics and summarizina results
as Ioas.
analysis of the images, in conjunction with the other openhole logs, allows this determination (Fig. 3-92). Images
showing no change in color will suggest a homogeneous
formation. Dark or white spots in a homogeneous background will suggest either clay galls or granules, nodules,
pebbles as a function of their size and resistivity. Dark
laminae in a sandstone may correspond to laminated shales. Vertical evolution of a homogeneous background will
suggest either a fining-up or a coarsening-up sequence in
grain size which in turn will suggest depending on the
thickness of this variation either fluvial channel fill, turbidite deposits or progradation of deltaic deposits.
To display results at a compressed scale for zoning
and comparing with standard logs, sliding window filters
are applied to output summary logs such as percentage
area of vugs, median vug size and vug connectedness
index, which can be expressed as the cumulative apparent fracture length in borehole wall per unit length of the
well.
Influence of textural parameters on
petrophysical properties
It is obvious and well known that textural parameters
control the petrophysical rock properties. This is the reason why it is so important to determine as much as possible this parameters.
In siliciclastic deposits determination of the grain size
and the sorting will help to link the permeability to the
porosity (cf Fig. 3-18, 3-20, Fig. 3-89 and Fig. 3-90). The
permeability also depends on the pore radius (Fig. 3-91).
J
Figure 3-90 - Influence of grain-size (expressed in 4 scale) and its geometric mean on the permeability (from Krumbein & Monk, 1942).
Porosity (%)
1
Figure 3-89 - Relationship between permeability and porosity as a
function of the grain size and sorting (from Chilingar, 1964).
In carbonates, the determination of the textura will
allow the determination of the rn factor of the Archie’s
equation (Fig. 3-92), and the link between the permeability and the porosity as illustrated by Fig. 3-26.
It is also important to determine the type of clay and
their distribution into the formations, as they will affect the
porosity, the permeability and the resistivity as well. The
154
Figure 3-91 - Influence of pore radius and porosity on permeability
(from Gaida et a/., 1973).
Serralog 0 2003
Texture
Porosity (%)
00
Figure 3-92 - Relationship between the formation factor and the prosity
allowing the determination of the m factor of the Archie's equation
(from Nurmi et a/., in Middle East Well Evaluation Review, 1986).
Figure 3-93 - Effects of the clay type and of their distribution on the
petrophysical properties of detrital deposits.
References and Bibliography
ALGER, R.P. (1966). - lnterpretation of Electric Logs in
fresh Water Wells in Unconsolidated Formations.
SPWLA, 7fiAnn. Log. Symp. Trans., Tulsa.
ALLEN, J.R.L. (1984). - Sedimentary structures. Their
character and physical basis. Developments in
Sedimentology, 30, Elsevier, Amsterdam.
BALOSSINO, P., & ANXIONNAZ, H. (1994).- Deriving
textural and geometrical information from dipmeter data
Serralog 0 2003
I Chapter 3 I
155
as a help to define subsurface geological models. Agip
s.p.a. & Schlumberger paper.
BATES, R.L. & JACKSON, J.A. (1980). - Glossary of
Geology. Amer. Geol. Institute, Falls Chuch, Virginia.
BEARD, D.C., & WEYL, P.K. (1973). - lnfluence oftexture on porosity and permeability of unconsolidated sand.
Bull. Amer. Assoc. Petroleum Geol., 57, p. 349-369.
BLATT, H., MIDDLETON, G., & MURRAY, R. (1972,
1980). - Origin of Sedimentary Rocks. 1st and 2nd ed.
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
BOURKE, L., DELFINER, P., TROUILLER, J.C., FETT,
T., GRACE, M., LUTHI, S., SERRA, 0. & STANDEN, E.
(1989). - Using Formation Microscanner Images. The
Technical Review, vol. 37, n o l .
BRIE, A., JOHNSON, D.L., & NURMI, R.D. (1985).
Effect of spherical pores on sonic and resistivity measurements. SPWLA, 26th Ann. Log. Symp. Trans., paper W.
CAROZZI, A.V. (Ed.) (1975). - Sedimentary Rocks.
Benchmark Papers in Geology, 15, Dowden, Hutchinson,
& Ross, Inc., Stroudsburg, Pennsylvania.
CHILINGAR, G.V. (1969). - Deltaic and shallow marine deposits. L.M.J.U. van Straaten, ed., Elsevier,
Amsterdam.
CHOQUETTE, P.W., & PRAY, L.C. (1970). - Geological
nomenclature and classification of porosity in sedimentary carbonates. Bull. Amer. Assoc. Petroleum GeoL, 54, p.
207-250.
CLAVIER, C., & RUST, D.H. (1976). - MID-PLOT: a
new Lithology Technique. The Log Analyst, 17, 6.
DELHOMME, J.P. (1992). - A quantitative characterization of formation heterogeneities based on borehole
image analysis. SPWLA, 33rd Ann. Log. Symp. Trans.,
paper T
DELHOMME, J.P. & MOTET, D. (1993). - Reservoir
description and characterization from quantitative borehole image analysis. Reservoir Characterization Africa and
Mediterranean Region, no 5, Schlumberger Wireline &
Testing.
DELHOMME, J.P. & SERRA, 0. (1984). - Dipmeterderived logs for sedimentological analysis. 9th Colloque
International de Diagraphies, SAID-SPWLA, Paris, paper
50.
DELHOMME, J.P., PILENKO, T., CHERUVIER, E. &
CULL, R. (1986). - Reservoir Applications of Dipmeter
Logs. 6lstAnn Fall Techn. Conf. SPE of AIME, paper SPE
15485, and in J. Petrol. Techn., 29, 2, pp. 180-186.
DICKEY, P.A. (1979). - Petroleum Development
Geology. Petroleum Publishing Co., Tulsa.
DODGE, C.F., HOLLER, D.P., & MEYER, R.L. (1971).
- Reservoir heterogeneities of some Cretaceous sandstones. Bull. amer. Assoc. Petroleum Geol., 55, p. 18141828.
DUNHAM, R.J. (1962). - Classification of Carbonate
Rocks according to Depositional Texture. Amer. Assoc.
Petroleum Geol., Mem. 1, p. 108-121.
ENGELHARDT, W.V. von (1960). - Der porenraum der
sedimente. Springer, Berlin, 207 p.
FOLK, R.L. (1959). - Practical Petrographic
Classification of Limestones. Bull. Amer. Assoc.
155
Petroleum Geol., 43, 1, p. 1-38.
FOLK, R.L. (1962). - Spectral subdivision of Limestone
Types.Amer. Assoc. Petroleum Geol., Mem. 1, p. 62-84.
FRASER, H.J. (1935). - Experimental study of porosity and permeability of clastic sediments. J. Geol., 43, p.
9 10-10 10.
FRIEDMAN, G.M., & SANDERS, J.E. (1978). Principles of Sedimentology. John Wiley & Sons, New
York.
GAIDA, K.H., RUHL, W. & ZIMMERLE, W. (1973). Examination of the voids in sandstones with the aid of
scanning electron microscope. Erdoel & Gas, 89, 9, p.
336-343 (in german).
GRATON, L.C., & FRASER, H.J. (1935). - Systematic
packing of spheres with particular relation to porosity and
permeability. J. Geol., 43, p. 785-909.
GRABAU, A.W. (1903). - Paleozoic coral reefs. Bull.
Geol. SOC.Amer., 14, p.337-352.
HOBSON, G.D., & TIRATSOO, E.N. (1975). lntroduction to Petroleum Geology. Scientific Press Ltd,
Beaconsfield, England.
KENYON, W.E. (1984). - Texture effects on megahertz
dielectric properties of calcite rock samples. J. Appl.
Phys., 55, 8, p. 31533159.
KENYON, W.E., & BAKER, P.L. (1984). - EPT interpretation in Carbonates drilled with salty muds. 59th Ann.
Tech. Conf. SPE of RIME, paper SPE 13192.
KENYON, W.E., KLEINBERG, R., STRALEY, Ch.,
GUBELIN, G. & MORRISS, Ch. (1995). - Nuclear
Magnetic Resonance lmaging - Technology for the 21s'
Century. Oilfield Review, Autumn, vol. 7, no 3.
KRUMBEIN, W.C. (1941). - Measurement and geological significance of shape and roundness of sedimentary
particles. J. Sed. Petrology, vol. 11, pp. 64-72.
KRUMBEIN, W.C., & MONK, G.D. (1942). Permeability as a function of the size parameters of
unconsolidated sand. Petroleum Technol., AIME Techn.
Publ. 1492, 5, p.1-11.
KRUMBEIN, W.C., & SLOSS, L.L. (1963). Stratigraphy and Sedimentation. 2nd ed. W.H. Freeman, &
Co., San Francisco.
KUSTER, G.T., & TOKSOZ, M.N. (1974). - Velocity
and Attenuation of Seismic Waves in TwoPhase Media :
Part I, Theoretical Formulations; Part ll, Experimental
results. Geophysics, 39, 5.
LANDES, K.K. (1951). - Petroleum Geology. John
Wiley, & Sons, New York.
LEE, C.H. (1919). - Geology and groundwaters of the
western part of San Diego County, California. Wat.-Supply
Irrig. Pap. Wash., 446, 121 p.
LEET, L.Don, JUDSON, S., & KAUFFMAN, M.E.
(1978). - Physical Geology. 5th ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
LINK, P.K. (1982). - Basic Petroleum Geology. DGCl
Publications, Tulsa.
LOMBARD, A. (1956). - Geologie Sedimentaire. Les
series marines. Masson, Paris.
LOMBARD,A. (1972). - Series sedimentaires. Genese
- Evolution. Masson, Paris.
156
LONGMAN, M.W. (1980). - Carbonate diagenetic textures from nearsurface diagenetic environments. Bull.
Amer. Assoc. Petroleum Geol., 64, p. 461-487.
MAXWELL-GARNETT, J.C. (1904). - Transactions of
the Royal Society, 203, p. 385, London.
MIDDLETON, G.V. (1976). - Hydraulic interpretation of
sand size distributions. J. Geology, 84, p. 405-426.
NEASHAM, J.W. (1977). - The morphology of dispersed clays in sandstone reservoirs and its effect on sandstone shaliness, pore space and fluid flow properties. 52nd
Ann Fall Meet SPE of AIME, paper SPE 6858.
NURMI, R.D., & FRISINGER, M.R. (1983) .- Synergy
of Core petrophysical measurements, log data, and rock
examination in carbonate reservoir studies. 58th Ann. Fall
Mtg. and Techn. Conf, of SPE ofAIME, paper SPE 11969.
NURMI, R. (1986). Carbonate Close up. Middle East
Well Evaluation Review, no 1.
PETTIJOHN, F.J. (1930). - Imbricate arrangement of
pebbles in a pre-Cambrian conglomerate. Jour. GeoL, 38,
p. 568-573.
PETTIJOHN, F.J. (1975). - Sedimentary Rocks. 3rd ed.
Harped & Row, Publishers, New York.
PETTIJOHN, F.J., POTTER, P.E., & SIEVER, R.
(1972). - Sand and Sandstone. Springer, New York.
PICKETT, G.R. (1960). - The use of acoustic logs in
the evaluation of sandstone reservoirs. Geophysics, 25,
1, p. 250-274.
PICKETT, G.R. (1963). - Acoustic Character Logs and
their Applications in Formation Evaluation. J. Pet.
TechnoL, 15, 6.
POTTER, P.E., & PETTIJOHN, F.J. (1977). Paleocurrents and Basin Analysis. 2nd ed. Springer, New
York.
PRYOR, W.A. (1973). - Permeability-porosity patterns
and variations in some Holocene sand bodies. Bull. amer.
Assoc. Petroleum Geol., 57, p. 162-189.
RASMUS, J.C., & KENYON, W.E. (1985). - An improved petrophysical evaluation of oomoldic Lansing-Kansas
City Formation utilizing conductivity and dielectric log
measurements. SPWLA, 26th Ann. Log. Symp. Trans.,
paper V.
REINECK, H.E., & SINGH, I.B. (1975, 1980). Depositional Sedimentary Environments. 1st and 2nd ed.
Springer, New York.
RIVEST, J.F., DELHOMME, J.P., & BEUCHER, S.
(1991). - Marker-Controlled Picture SegmentationApplied
to Electrical Logging Images. Proc. SPlE 1451: Non-linear
Image Processing II, p, 179-190.
ROGERS, W.F., & HEAD, W.B. (1961). - Relationships
between porosity, median size, and sorting coefficients of
synthetic sands. J. Sediment. Petrol., 31, p. 467-470.
RUKHIN, L.B. (1958). - Grundzuge des Lithologie.
Akademie-Verlag, Berlin.
RUSSELL, W.L. (1951). - Principles of Petroleum
Geology. McGraw-Hill Book Co., New York.
SARMA, V.V.J., & RAO, V.B. (1963). - Variation of
Electrical Resistivity of River Sands, Calcite and Quartz
Powders with Water Content. Geophysics, April.
Schlumberger (1979). - Well Evaluation Conference,
Serralog 0 2003
Texture
Algeria.
Schlumberger Middle East S.A. (1981). - Well
Evaluation Conference, United Arab Emirates / Qatar.
Schlumberger Technical Services, Inc. (1982). Essentials of Natural Gamma ray Spectrometry
Interpretation.
Schlumberger Technical Services, Inc. (1983). - Well
Evaluation Conference, India.
SELLEY, R.C. (1970, 1978, 1985). - Ancient
Sedimentary Environments. I s t , 2nd and 3rd ed. Chapman
& Hall, London.
SELLEY, R.C. (1976). - An lntroduction to
Sedimentology.Academic Press, London.
SEN, P.N. (1980). - Dielectric and conductivity response of sedimentary rocks. SPE of AIME, paper 9379.
SEN, P.N. (1981). - Relation of certain geometrical features to the dielectric anomaly of rocks. Geophysics, 46,
p. 1714-1720.
SERRA, 0. (1972). - Diagraphies & Stratigraphie. In:
Mem. B.R.G.M., 77, p. 775-832.
SERRA, 0. (1973). - Interpretation Geologique des
diagraphies en Series Carbonatees. Bull. Centre Rech.
Pau -SNPA, 7, 1, p. 265-284.
SERRA, 0. (1974). - Interpretation geologique des
Series deltai'ques a partir des diagraphies differees. Rev.
A.F.T.P., 227, OCt., p. 9-17.
SERRA, 0. (1989). - Formation Microscanner lmage
Interpretation. Schlumberger Educational Services, SMP7028.
SERRA, O., & ABBOTT, H. (1980). - The Contribution
of Logging data to Sedimentology and Stratigraphy. 55th
Ann. Fall Techn. conf. SPE ofAIME, paper SPE 9270, and
in SPE J., Feb. 1982.
SERRA, O., & SULPICE, L. (1975). - Sedimentological
Analysis of shale-sand series from well logs. SPWLA, 16th
Ann. Log. Symp. Trans., paper W
SERRA, O., & SULPICE, L. (1975). - Apports des diagraphies differees aux etudes sedimentologiques des
series argilo-sableuses traversees en sondage. 9th Cong.
Intern. Sediment., Nice, theme 3, p. 86-95.
SNEED, E.D., & FOLK, R.L. (1958). - Pebbles in the
lower Colorado River, Texas - a study in particle morphogenesis. Jour. Geol., 66, p. 114-150.
THEYS, P., LUTHI, S., & SERRA, 0. (1983). - Use of
dipmeter in Carbonates for detailed sedimentology and
reservoir engineering studies.
TRASK, P.D. (1932). - Origin and environment of source sediments of petroleum. Gulf Publ. Co., Houston.
VATAN, A. (1954). - Petrographie sedimentaire. Ed.
Technip, Paris.
VISHER, G.S. (1965). - Use of vertical profile in environmental reconstruction. Bull. Amer. Assoc. Petroleum
Geol., 49, p. 41-61.
WADELL (1932). - Volume shape and roundness of
rock particles. Jour. Geol., 40, p. 443-451.
WADELL (1933). - Sphericity and roundness of quartz
particles. Jour. Geol., 41, p. 310-331.
Van WAGONER, J.C., MITCHUM, R.M., CAMPION,
K.M. & RAHMANIAN, V.D. (1990). - Siliciclastic Sequence
Serralog 0 2003
I ChaRter 3 1
157
Stratigraphy in Well Logs, Cores, and Outcrops. AAPG
Methods in Exploration Series, no7.
ZINGG, T (1935). - Beitrag zur Schofferanalyse:
Schweiz. Min. u. Pet. Mitt., Bd. 15, p. 39-140.
157
Introduction
Review of general geological concepts
Definition
By definition a sedimentary structure (internal), or feature, is "a structure formed either contemporaneously with
deposition: a primary sedimentary feature, or by sedimentary processes subsequent to deposition: a secondary sedimentary structure" (Bates & Jackson, 1980).
According to Pettijohn & Potter (1964) the structure
is an inherent property of a rock and a guide to its origin.
Whereas the texture deals with the grain to grain relations
in a rock, structure has to do with discontinuities and
major inhomogeneities. The structure is concerned with
the organization of the deposit - the way in which it is put
together - . Hence structures are the larger features that,
in general, are best studied in the outcrop rather than in
the hand specimen or thin section
'I.
They explain the local variations of the composition or
texture.
A sedimentary structure refers to megascopic morphological features. These features have been studied for
some time, as they are often visible to the naked eye.
They include the thickness and the shape of beds, their
internal organization, the nature of their surfaces, joints,
concretions, cleavages, and fossil content.
Primary sedimentary structures are generated by
either current velocity, impeding the action of the gravity,
and its evolution, (scour and erosional marks, ripple
marks, cross-bedding, wavy-bedding, graded bedding, or
biogenic activity (tracks, trails, burrows, rootlets,...), or
action of climatic or physical agents (mud-cracks, pits,
load casts, dikes, convolute bedding, slump structures).
"Some structures are texture dependent. Ripples
marks and cross-bedding for example, characterize only
those sediments which have a grain size in the sand
range" (Pettijohn et a/., 1964).
Importance of sedimentary structures
The primary sedimentary structures are particularly
important because they will reflect the hydrodynamic
Serralog 0 2003
conditions at the time of deposition (e.9, energy, type of
current,...). They constitute an important element of the
facies of a sedimentary unit and will lead to a better definition of the depositional environment.
As mentioned by Selley (1970), structures "unlike
lithology and fossils are undoubtedly generated in place
and can never have been brought in from outside".
Consequently, it is essential to detect them by analysing open-hole logs and more specifically dipmeter and
image tool data.
Classification of sedimentary structures
There are several different types of sedimentary structures, all of which can be characterized by the shape of
their surfaces and their internal organization.
The classification of structures can be based on the
time of their formation. They are defined as:
- predepositional, (formed before the deposition of a
bed). They correspond to features observed on the surface of a preceeding bed such erosion or impressions;
- syngenetic or primary or syndepositional,(contemporaneous with deposits). These structures contain information on physical, chemical or biological conditions, existing in a depositional environment during sedimentation.
They are subdivided into inorganic and organic structures, depending on their origin; inorganic structures are a
result of physical agents; organic structures are formed in
connection with an animal or plant organic activity (burrow; impressions; root traces);
- epigenetic or secondary or postdepositional, (formed
after sedimentation). These are often of chemical origin
and their occurrence reflects diagenetic phenomena; or of
physical origin resulting from tectonic deformation.
A second classification can also be suggested which is
based on the agents or processes, which have created
the sedimentary structures:
- physical such as action of gravity, influences of current or stress (ripple marks, tool marks, convolute-bedding, slumping, mud-cracks,...),
- chemical such as dissolution, concretions;
- biological such as burrows, tracks, trails, foot impressions, root traces,...
A third classification of sedimentary structures based
159
on their location can be suggested:
- external structures, that cover the size and the shape
of beds, thus the nature of their boundaries and the shapes of the lower and upper bedding planes;
- internal structures, relative to internal organization of
the bed : i.e. massive, laminated, graded-bedding, growth
structures (stromatolitic limestones,...).
The classifications proposed by Krumbein & Sloss
(1963), by Selley (1976), essentially combine the time
period and the agent, whereas the classification of
Pettijohn & Potter (1964), and Blatt et al., (1979) are
based more on the location of features. We have retained
the classification of Table 4-1, derived from Pettijohn &
Potter, because it adapts more closely to log analysis and
illustration.
above and below." This general term includes bed and
lamina (McKee and Weir, 1953).
A bed is "the smallest formal unit in the hierarchy of
lithostratigraphy units. In a stratified sequence of rocks it
is distinguishable from layers above and below ... A bed
commonly ranges in thickness from a centimeter to a few
meters" (Bates & Jackson, 1980). A single bed is a sedimentary unit formed under essentially constant physical
conditions and constant delivery of the same material
during deposition. It corresponds to a volume with a certain geometry defined by surfaces. It may be internally
layered, consisting of smaller units or laminae with their
own characteristics (Fig. 4-1).
Lithology
Depositional
Junits
Table 4-1
Classification of sedimentary structures
(from Pettijohn & Potter, 1964).
Geological objects
BEDDING EXTERNAL FORM
1. Beds equal or subequal in thickness: beds laterally uniform in thick
ness: beds continuous
2. Beds unequal in thickness: beds laterally uniform in thickness: bed!
:ontinuous
3. Beds unequal in thickness: beds laterally variable in thickness: bed!
:ontinuous
1. Beds unequal in thickness: beds laterally variable in thickness: bed!
jiscontinuous
BEDDING INTERNAL ORGANIZATION AND STRUCTURE
1. Massive (structureless)
2. Laminated (horizontally-laminated; cross-laminated)
3. Graded
4. Imbricated and other oriented internal fabrics
5. Growth structures (stromatolites, etc.)
BEDDING PLANE MARKINGS AND IRREGULARITIES
1. On base of bed
(a) Load structures (load casts)
(b) Current structures (scour marks and tool marks)
(c) Organic markings (ichnofossils)
2. Within the bed
(a) Parting lineation
(b) Organic markings .
3. On top of the bed
(a) Ripple marks
(b) Erosional marks (rill marks: current crescents)
(c) Pits and small impressions (bubble and rain prints)
(d) Mud cracks, mud-crack casts, ice-crystal casts, salt-crys.
tal casts
(e) Organic markings (ichnofossils)
Figure 4-1 - Subdivisons of a bed in sedimentation units with their own
characteristics (adapted from Blatt et al., 1980).
A lamina is "the thinnest recognizable unit layer of original deposition in a sediment or sedimentary rock, differing from other layers in color, composition or particle
size, ... (commonly 0.05-1.00 mm thick). It may be parallel or oblique to the general stratification.... Several laminae may constitute a bed." A lamina is produced by minor
fluctuations in fairly constant physical conditions. Rarely,
a lamina can be up to a few centimeters thick (Reineck
and Singh, 1975)
A sedimentation unit is characterized by:
- its shap- the n'ature of its bobndaries,
- its internal organization.
BEDDING DEFORMED By PENECONTEMPORANEOUS PROCES
SES
1. Founder and load structures (ball-and-pillow structures, load casts)
2. Convolute bedding
3. Slump structures (folds, faults, and breccias)
4. Injection structures (sandstone dikes, etc)
5. Organic structures (burrows, "churned" beds, etc)
Primary sedimentary structures
Shape of sedimentation unit
One must stress that a bed or a sedimentation unit, is
that thickness of sediment which was deposited under
essentially constant physical conditions" (Otto, 1938),
separated from other under- and over-lying beds by physically and visually more or less well-defined bedding planes, "made evident because of the unlike texture or com"
Stratum. bed. lamina descrbtion
A stratum is a "tabular or sheet like body or layer of
sedimentary rock, visually separable from other layers
160
Serralog Q 2003
position" (Pettijohn & Potter, 1964).
The parameters defining the shape of a sedimentation
unit are:
- thickness;
- bedding planes;
- lateral dimensions.
reduction conditions at the sea floor explaining the absence of the benthonic fauna and related bioturbation.
The analysis of the vertical evolution of bed thicknesses within the same lithology can indicate a change in
rhythm of sedimentation. These changes can be used as
time-markers.
Bed thickness
Bed thickness can vary from a few millimeters to several meters (Campbell, 1967). A bed can be massive or
internally finely stratified, formed by a succession of finer
units (laminations). A lamination results from minor fluctuations in the physical conditions, which prevailed in the
depositional environment. They are particularly expressed
by changes in grain size (sand to silt) and sometimes in
composition (quartz to clay mineral). The thickness of a
lamination is measured in millimetres and generally does
not exceed several centimetres (Fig. 4-2).
0
s
Figure 4-3 -Approach of log normality of thickness of turbidite sandstone beds (from Scoff, 1966).
Surfaces and beddina Dlanes
v)
v)
E
Y
.-0
c
c
Figure 4-2 - Strata, laminae and beds as a function of their thickness
and vertical resolution of the logging tools for comparison.
Knowledge of the thickness of beds is important
because it is sometimes related to the granulometry or to
the depositional mode. For example, in sandy turbidites
and volcanic ashes the thickness of the beds and the size
of the grains are related and decrease in the direction of
flow (Scheidegger & Potter, 1971). This is therefore a
means to distinguish proximal from distal deposits. In another way, Scott (1966), after Schwarzacher (1953), points
out that the thickness of individual beds pertaining to the
same kind of deposit are distributed following a straight
line on a probability-logarithmicgrid (Fig. 4-3).
According to Pettijohn (1975) the existence of laminations in marine environments is either the indication of
very fast deposits, below the zone of wave action, or of
Semlog 0 2003
The form and attitude of strata and structures are defined by the shape of their surfaces. Subdividing strata into
beds requires the identification of bedding surfaces.
These surfaces should be distinguishable by textural or
compositional changes from one bed or lamina to another.
Bedding surfaces have no thickness, but have areal
extents equivalent to the beds with which they are associated. Consequently, the geometry of a bed depends on
the relative disposition of its two boundaries.
A surface is a "two-dimensional boundary between
geologic features such as formations or structures." Its
shape can vary from a plane to a very irregular surface. A
perfect plane is defined by three non aligned points in the
space. Irregular surfaces can be cylindrical, conical,
undulating or warped. They can be shaped like a spoon or
very irregular like a stylolitic boundary. Consequently,
layers can have different shapes according to the nature
of their bounding surfaces, as shown in Figure 4-4
(Campbell, 1967). Bed forms are determined by stream
power and grain size.
As Campbell (1967) has correctly analysed, the surface of a bed represents a surface of non deposition, or corresponds either to an abrupt change in the condition of
sedimentation (variation in the energy of environment) or
to a surface of erosion. Usually the upper surface of a bed
constitutes the lower surface of the following upper layer.
Hence the characterization of beds depends on the recognition of their surfaces.
161
Next Page
J
; 4-4 ~
- ~i~~~~~
~
~
~ different shapes that can be acquired by
beds and laminae, and the corresponding descriptive terms
(from Campbell, 1967).
F
Figure 4-5 - Relationship between the thickness and the Width Of sedimentary bodies, giving an external morphology of these deposits
(from Krynine, 1948).
Figure 4-6 - Geomorphology and characteristics of bed-load, mixed-load, and suspended-load stream channel segments and their deposits
(from Galloway, 1977, and in Galloway & Hobday, 1983, fig. 4-13).
Lateral dimension of beds
Taking into account the bed composition and its thickness, the nature of the bedding planes, and the bed frequency and depositional environment, it is possible to
estimate its lateral extent as suggested by Krynine (1948)
(Fig. 4-5). For instance, if several beds show planar and
parallel boundaries over a thick interval, one can assume
a large extent of each bed composing the interval.
If the environment is known, as in a fluvial system (Fig.
4-6) a better link can be done between the bed thickness
and its lateral extent.
162
Nature of bed boundaries
The transition from one layer to another can either be
abrupt or gradual. In the first case the boundary is well
defined and agrees with the bedding planes.
The boundary is conformable if it corresponds to a
short break in sedimentation without modification of the
depositional sequence or without erosion, and the beds
remain parallel.
The boundary will be unconformable if it corresponds
to a break in sedimentation, followed by a change in the
sequence of deposition underlined either by an erosional
or by a lateritized surface (if continental), by a truncation
and possibly by a change in dip magnitude and azimuth
Serralog 0 2003
Previous Page
Structure
I Chapter4 I
163
(Fig. 4-7).
Fininp-up
sequence
Composite
bedset
Simple
bedset
ONLAP
TRUNCATION
WWNLAP
WWNLAP
DOWNLAP
TRUNCATION
WWNLAP
TRUNCATION
J
TRUNCATION
WWNLAP
WWNLAP
TRUNCATION
Figure 4-9 - On the right, different bedding due to the change in composition or
in texture :shape, size and orientation of
the grains. These types of arrangement
affect the pore geometry and
consequently the petrophysical properties
of each unit of sedimentation
(adapted from Griffiths, 1961).
On the top-left, illustration of size change
through thin section.
TRUNCATION
TRUNCATION
HORIZONSOF BURROWS, ROOTS. ETC
GRAVEL
MUDSTONE
SANDSTONE
ORGANIC DEBRIS
BEDDING SURFACE
Figure 4-7 - Typical bedding surfaces and how to recognize them
(adapted from Van Wagoner et al., 1990).
The surface of the bed boundary can be planar, ondulated or irregular with load casts (Fig. 4-8). The latter
structures are generated by the differential loading of a
waterlogged sand on an unconsolidated plastic mud.
They are common feature of turbidite deposits, yet they
also occur in deltaic and fluvial sediments.
change of charateristics.
The group of beds is called simple if it consists of two
or more superimposed beds, having the same general
characteristics (mineralogy, texture, internal structure)
and surrounded by other beds of a different nature.
A sequence of beds is called composite if it consists of
a group of layers, having different composition, texture,
and internal structure (Fig. 4-10).
1
Figure 4-8 - Example of undeformed bed (the two beds at the top), bed
with load casts on its lower surface due to deformation of the
mud/sand interface, and segmentation of bed in pseudo nodules or
pillows.
The bedding surface can be associated to changes in
composition, in grain shape, grain orientation, grain size
and cementation (Fig. 4-9).
In the case of a gradational transition between beds
the boundary is not clearly defined and thus not visible. It
then agrees with a sequence, which is either granulometric (normal or reverse), or mineralogic, or both (sand to
shale, or shale to silt to sand).
Groups of beds (“bedset‘;)
McKee & Weir (1953) have defined a group of beds as
a succession of layers or laminae, essentially conformable, separated from adjacent sedimentary units by a surface of erosion or by a break in deposition, or a sudden
Serralog 0 2003
J
Figure 4-10 - Schematic illustration of bedding terminology: lamina,
bed, simple bed set, Composite bed set and bedding type
(from Reineck 8, Singh, 1975).
Internal oraanization of beds
Several types of internal organization can be recogni163
zed. They depend on the type of transport, the stream
power or the energy in the environment and the water
depth.
Massive bedding
-
A bed can be apparently homogeneous, (i.e. without
any visible variation due to textural changes or sedimentary features). This corresponds to a constant condition of
sedimentation, without apparent stratification either due to
the absence of current ripples or even to a massive sudden sedimentation. It can also correspond to an intensive
bioturbation, which has completely destroyed all traces of
stratification, or to diagenesis especially in carbonates.
Genuine depositional massive bedding is often seen in
fine-grained, low energy environment deposits.
Direction of transport current
Laminated bedding
As previously mentionned, a single bed can be internally finely layered, made up of smaller units laminae. A
lamina is produced as a result of some minor fluctuations
in rather constant physical conditions. A bed shows therefore stratifications either parallel, oblique or cross-bedded. These laminations are control by the stream power
(Fig. 4-11).
Figure 4-12 -At the top :terms used in ripple description. L = ripple
length, H = ripple height, I1 = horizontal projection of stoss side, 12 =
horizontal projection of lee side. In the middle :dynamic mechanism
for ripple formation. At the bottom :flow pattern over a lee face of a ripple with velocity distribution, flow separation and the three major zones
on the lee side (from Reineck & Singh, 1975).
Ripples and other bedforms are present as undulations on a non cohesive surface. They are the results of
the interaction of moving fluids (water and air) - currents
or waves, wind - on a sediment surface. Interaction of
these fluids with a non cohesive material composed of
grains produces small ripples, megaripples, plane beds or
antidunes as a function of the flow regime (Fig. 4-13).
L
3”
P
E
e
j;
K
6
MEDIAN FALL DIAMETER
Figure 4-11 - Types of laminations as a function of the stream power
and the grain size (data from Guy et al., 1966, in Allen, 1968, and in
Reineck and Singh, 1975).
Ripples. foresets. plane beds and antidunes
A ripple is described in terms of its size and shape. It
is composed of a crest, or summit, and a trough (Fig. 412).
164
Figure 4-13 - Relationship between stream power, median fall diameter, bed form and sedimentary structure (after Simons et al., 1965).
In a low flow regime resistance to flow is large and
sediment transport is nihil or relatively small depending on
the coherence of the material. Water surface undulations
are out-of-phase with the bed undulations. The bedform is
either small ripples or megaripples or a combination of
Serralog 0 2003
both. Individual grains move up the back of the ripples
and avalanche down the lee face (Fig. 4-12).
In an upper flow regime resistance to flow is small and
sediment transport is large. Water surface undulations
and bed undulations are in-phase, and segregation of bed
material is negligible. The usual bed forms are plane beds
and antidunes. Individual grains roll almost continuously
downstream in sheets a few centimeter thick.
As well known, the mechanics of particle motion are :
suspension, bouncing or saltation, and rolling (Fig. 4-14).
LAMINAR FLOW
Low Reynolds numbers
R < 1for particles moving in a fluid
TURBULENT FLOW
High Reynolds numbers
R 5 1 for particles moving in a fluid
A
Figure 4-15 - Illustration of laminar and turbulent flows.
In laminar flow stream lines remain distrinct. Flow direction at every
point remain unchanged with time. Current velocity is inferior to a few
tenthes of mm/s depending on the width and the depth of a channel.
In turbulent flow the flow lines are confused and heterogeneously
mixed, generating eddies and vortices.
Figure 4-14 - Particle motions : A = suspension; B = bouncing or
saltation; C = rolling. (adapted from Selley, 1976).
The particle motion through a fluid is expressed by two
numbers : the Reynolds number, R,and the Froude number, F:
where V is the velocity of the particle, d is the diameter
of the particle, p is the density of the particle, p is the viscosity of the fluid, U is the average current velocity, g is
the acceleration due to gravity and D is the depth of the
channel.
For low Reynolds numbers, the fluid flow is laminar, it
means flow lines are parallel to the bounding surface. For
high Reynolds numbers the flow is turbulent, generating
eddies and vortices (Fig.4-15).
A Froude number of 1 separates two distinct types of
fluid flow regime in open channel: lower flow regime for
F < 1, upper flow regime for F > 1.
The stream power is the product of the fluid velocity by
the shear stress zo which is equal to p S wherey is the
specific weight of the fluid and sediment, D is the stream
depth and S is the slope of the energy gradient. The relationship between current velocity and water depth is
reproduced in Figure 4-16.
The critical flow velocity required for transportation,
deposition and erosion as a function of grain size is illustrated by Figure 4-17 based on works realized by
Serralog B 2003
Depth (cm)
Figure 4-16 - Relationship between current velocity and water depth, in
relation to bedforms indicated by the Froude number
(from Simons & Richardson, 1962).
165
100.0
0
3
Figure 4- 17 - Relationship between fluid-flow velocity and sedimentary
processes such as transportation, deposition and erosion as a function
of grain size (from Hjulstroms in Sundborg, 1956).
Hjulstroms (in Sundborg, 1956). For transportation, as
one might expect, the critical velocity increases with grain
size. For erosion, one must take into account the cohesion of the material. For cohesive beds, such as shales, a
higher velocity is required to start erosion. This is related
to their resistance to friction. This effect is responsible for
the preservation of delicate clay laminae, termed flasers,
in tidal deposits. At the opposite, non-cohesive beds as
silt or sand, the erosion starts for a lower velocity.
The different sorts of fluid-flow - traction currents, density current, suspension - generate different types of sediments with their own types of sedimentary structures.
Traction currents, for instance, generate ripples, crossbedding, foreset bedding, dunes, herringbone. Sediments
linked to density currents, which are a combination of
traction and suspension, are characterized by mixtures of
sands, silt and clay, and typically show graded bedding
with generally lack of cross-bedding. Table 4-2, from
Selley (1976), links the sediment types to the sedimentary processes which have generated them.
Table 4-2
Relationship between sediment types and sedimentary
processes (modified from Selley, 1976).
I Sedimentary processes I
Traction current
Density current
Environment
Type of deposits
ISubaerial
Subaqueous
IPredominantlycross-bedded sands
Subaerial
Subaqueous
INubes ardentes
IGraded sands, silts and clays
I
Velocity repartition from center to the edges
Velocity repartition from the bottom of a channel
1
Subaerial
Subaqueous
Depthofthachannel
ILoess
INepheloid clays
Particles are transported following different modes as
a function of their size and the stream power (Fig. 4-18).
One must also take into account the repartition of the
fluid velocity in a channel as a function of the proximity of
the channel edges, the channel depth and the type of
flow: laminar or turbulent (Fig. 4-19).
166
Figure 4-18- Modes of transport ofparticles as a function of their size
(from von Engelhardt, 1977).
Generally unstratified poorly sorted
Mass gravity transport
Suspension
I
Transport power mtgp (cm)
Bottom of tha channel
Figure 4-19 - Influence of the proximity of the edges and bottom of a
channel on the fluid velocity.
The particles carried in suspension in a turbulent flow
have a distribution which is represented by Figure 4-20.
Serralog 0 2003
Chapter 4
I 167
Height
Y/h
asymmetrical wave ripples. Foreset
laminae with off-shoots.
Height a b o v e
stream bed (m)
straight-crested small-current ripples
Cross-bedding planar.
Relative concentration
Figure 4-20 - Relative concentration of suspended particles in a turbulent flow as a function of the grain size. v: settling velocity; n,,.' number
of particles at height y; na :number of particles at height a < y; y:
height above the bottom; h: water depth; p f fluid density, 70: maximum
shear stress (at the top from von Engelhardt, 1977; at the bottom from
Hjulstroms, 1939).
Ripple marks result from the interaction of waves or
currents with a non cohesive sediment surface. They
generally trend at right angles or obliquely to the direction
of fluid flow. Their shape and size vary considerably as a
function of the current velocity. Ripples are asymmetrical
wave-like beds occurring in fine sands subjected to gentle traction currents. Migration of ripples generates crosslaminated deposits. Several types of ripples exist. They
are reproduced in Figure 4-21.
Ripples do not form in clay nor in coarse sand or gravel. They are restricted to coarse silt and sand with a grain
size of less than about 0.6 mm in diameter (Reineck &
Singh, 1975). Their length ranges between 4 and 60 cm
and their height varies from 0.3 and 6 cm depending on
the grain size (higher if coarser). They occur in dunes,
rivers and deltas.
Megaripples range between 60 cm and 30 m in length,
and 6 cm to 1.5 m in height. The grain size is more than
0.6 mm.
Foreset laminae result from deposition on the lee side
of a ripple by the progradation process. Foreset laminae
may be angular, tangential or concave as a function of
velocity, bed shear stress, depth ratio and particles in
suspension (Fig. 4-22). At low velocities sediment
transport is by bedload movement. The grains move
Serralog 0 2003
undulatory small ripples. Cross-beding
weakly festoon-shaped.
lingoid small ripples. Cross-bedding
strongly festoon-shaped.
h a t e megaripples.
Figure 4-21 - Block diagrams illustrating different types of ripples
(from Reineck & Singh, 1975).
along the bed and are deposited on the upper part of the
lee face from where they move downslope under gravitational force producing a planar slip face. With increasing
167
velocities, a greater proportion of grains is maintained in
suspension, carried beyond the lee face, and deposited in
the form of bottomset and toeset. Angular contact is replaced by tangential contact and the angle of repose of the
lee face is reduced.
Figure 4-22 - Factors controlling the shape of the foreset laminae
(modified from Jopling, 1965).
Cross-beddinq
It is probably the commonest type of bedding in clastic
deposits.
A cross-bed is a single sedimentation unit consisting of
foreset laminae inclined to the principal surface of sedimentation. It is separated from the adjacent units by a surface of erosion, non deposition, or abrupt change in character. Its thickness may vary from a few millimeters to
tens of meters.
Three principal types of cross-bedding exist (Fig. 423).
- Planar or tabular cross-bedding corresponds to
cross-beds whose bounding surfaces form more or less
planar surfaces (Fig. 4-24).
I
Figure 4-24 - Typical tabular planar cross-beds
(adapted from Allen, 1963).
- Wedge-shaped cross-bedding corresponds to crossbeds that thin out.
- Trough cross-bedding corresponds to cross-beds
having a trough shape and Whose bounding surfaces are
Curved (Fig. 4-25).
Set
(Tabular)
Set
Figure 4-25 - Typical trough cross-beds (adapted from Allen, 1963).
Coset
set
(C)
Trough
Cross-Bedding
Figure 4-23 - Terminology for cross-bedding
(adapted from McKee & Weir, 7953).
168
Migration of small-current or wave ripples can generate small-scale cross-bedding. Individual units vary from a
few millimeters to 5 cm in thickness. They are usually
trough shaped (Reineck and Singh, 1975).
Large-scale cross-bedding correspond to individual
units usually more than 5 cm thick. They can be planar or
trough-shaped and are produced by mega ripples, sand
dunes, long shore bars, etc.
Their shape and slope depend on the velocity, bed
Serralog 0-2003
Structure
shear stress, depth ratio and sediment type. Figure 4-22
illustrates the shape of the foreset laminae as a function
of these factors (Jopling, 1965).
The type of internal structure of the sedimentation
units allows their identification (Fig. 4-26), and the determination of their origins.
Chapter 4 ' 169
Cross bedding with flarera
simple
bifurcated
Flaser bedding
wavy
bifurcated
wavy
Wavy bedding
with thick Ienrer
connected
with fiat lenses
Lenticular bedding
with thick iensea
single
with flat lenses
Figure 4-26 - Recognition of the sedimentation units based on their
internal structure (from Potter & Pettijohn, 1977).
Scour-and-fill cross bedding and trough cross-bedding
result from erosion of a sediment surface by the current
flowing over it and refilling of the trough or depression with
sediment when the current velocity decreases. During the
filling process, the sediment transported down the steep
upcurrent slope. This produces inclined laminae that gradually flatten during the latter stages of scour fill.
Figure 4-27 - Schematic classification of flaser, wavy and lenticular
bedding (from Reineck & Singh, 1975).
Chevron cross-beddina or herrinabone cross-bedding
This type of cross-bedding relates to cross-bedding
that dips in different or opposite direction in alternating or
superimposed beds, forming a chevron or herringbone
pattern (Fig. 4-28). This type of cross-bedding is typical of
tidal environments.
Flaser bedding
During high tides, mud is commonly deposited across
the ripple crest and into the troughs. Flaser bedding
results when this mud is preserved in the troughs, but little or no mud is preserved on the crests. Wavy or bifurcated flasers result when there is some mud preserved
across the crests, and simple flasers result when there is
no mud preserved on the crests (Fig. 4-27).
Wavy bedding forms when alternating mud and sand
layers occur as continuous layers (Fig. 4-27). Their genesis requires conditions where the deposition and preservation of both sand and mud are possible. Lower current
energies are necessary to avoid erosion of the previous
deposits.
Lenticular bedding forms when the ripples or sand lenses are discontinuous and isolated in all directions (Fig. 427). It is produced where there is meager sand supply and
under conditions more favorable for the deposition and
preservation of mud than of sand.
These structures are generally observed in tidal environments.
When the current velocity increases, it starts to truncate or erode the previous deposits. The surfaces that are
generated can have different shapes as a function of the
flow type. Consequently, the shape and organization of
beds and laminae strongly depend on the current or wave
activity. Terminology has been proposed to describe this
organization (Reineck and Singh, 1975).
Serralog 0 2003
J
Figure 4-28 - Example of chevron or herringbone cross-bedding. They
can be recognized only in 3-13 (from Reineck & Singh, 1975).
Graded bedding
Sedimentation units characterized by a gradation in
grain size are termed graded beds.
Pettijohn (1957) described two main types of grading.
In the first type there are no fines in the lowest part of the
graded bed (Fig. 4-29a). In the second type the fines are
distributed throughout (Fig. 4-29b).
169
Well logging and Geology
Coorse iaikgrading
a
b
Figure 4-29 - Different types of graded beding
(adapted from Pettijohn, 1957).
Dzulynski and Walton (1965) recognized 8 types of
graded bedding. They are illustrated by Figure 4-30.
Graded beds commonly occur in thick sequences of
flysch type of sediments. Commonly, graded beds are
separated from each other by parallel-bedded clayey
layers, which represent the product of normal sedimentation interrupted by turbidity currents. Bouma’s sequence
is an example of graded bed (Fig. 4-31).
In fact, there are two types of vertical evolution of grain
size in a bed : namely normal (fining-upward) or reversed
(coarsening-upward)graded beddings (Fig. 4-32). Normal
graded bedding is generally underlined by an abrupt
contact at the base.
Figure 4-31
- The various types of graded bedding (from sellex
1976).
Figure 4-32 - Typical graded bedding in the Bouma’s sequence and its
interpretation in terms of flow regime.
-I
Figure 4-30 - The 8 different types of graded beds
[from Dzulynski & Walton (1965).
170
Bioaenic activitv
Other typical sedimentary features are linked with biolgenic activity, either animal or vegetable. Tracks, trails,
burrows, and rootlets are the most important biogenic features. These trace fossils are defined and named by the
characteristics of their makings, rather than the organisms
responsible for making them. In some cases, the origin of
trace fossils is unknown (Fig. 4-33).
Serralog 0 2003
Next Page
Deep marine
weight of overlying material, accumulation of moisture,
earthquakes or regional tilting. These secondary structures appear as small deformations within the deposit, such
as slides, slumps and convolute bedding (Figs. 4-35 and
4-36).
*
Mwsmentdlrechon
Figure 4-33 - Illustration of biogenic activity, based on Seilacher, 1964,
1967; Rodriguez 8 Gutschick, 1970; Heckel, 1972 (from Selley, 1976).
l m
Figure 4-35 - Typical slump structure (from Selley, 1976).
Action of climatic or phvsical aaents
Current
c
Mud cracks, load casts and pits are typical features
generated by climatic (shrinkage) or physical forces (Fig.
4-34).
Figure 4-36 - Sketch of convolute bedding (from Selley, 1976).
Disconformities
Certain surfaces are related to sea-level changes.
Disconformities (breaks in sedimentation caused by erosion or non deposition) can result from transgression,
regression, the formation of hardgrounds, levels of
condensation or other diagenetic effects.
1
Figure 4-34 - Sketch of mud cracks generated by shrinkage of clay, silt
or mud, generally in the course of dfying (from Selley, 1976).
Dish structures
Dish structures consist of small meniscus-shaped lenses (4 to 50 cm long and one to a few centimeters thick)
generally found in sandstone. They are created by water
escaping during elutriation of clay soon after deposition of
the sand. These structures are common in turbidite deposits.
Flame structures
Flame structures correspond to "wave- or flame-shaped plumes of mud that have been squeezed irregularly
upward into an overlying layer. It is probably formed by
load casting accompanied by horizontal slip or drag"
(Bates & Jackson, 1980).
Secondary sedimentary structures
When shear stress is applied to a sloping surface, the
mass movement that results can create secondary sedimentary structures. This stress can be caused by the
Serralog 0 2003
-
Intrusions Dikes
They correspond to igneous intrusions that cut across
the bedding of sedimentary rocks or across other rock
types (cf; Fig. 2-1).
Diapiric structures
Salt or clay uplifts, sometimes igneous intrusions,
generate anticlinal structures called domes. The overlying
sedimentary strata are ruptured by the squeezing-out of
plastic material.
Determination of rock structure
Traditional approach
As the structures are megascopic features, in the
range of millimeters to a few meters, they are better described and studied on outcrops or by core observations.
For a correct description and analysis a 3-D observation
is absolutely necessary.
Due to the lack of vertical resolution of surface seismics, generally only diapiric structures can be recognized.
In wells, thanks to the vertical resolution of dipmeters
and imagery tools, most of the structures will be detected
171
Previous Page
Determination of rock structure
Well logging approach
and analyzed.
The interest of these well logging measurements will
now be illustrated by examples. Outcrop or core photographs of similar structures will be presented at the same
time to demonstrate the validity of this approach..
The link beheen structural geological features and
well logging parameters is indicated in Table 4-3.
Table 4-3
Link between the structural geological features and the dipmeter and imaging data.
Logs to use for detecting the phenomenon
Shape : lateral variation in thickness
External form of bed
?
On base
(lower face)
Bedding surface
features
fLoad marks
origin
Erosion
Truncations
Organic
origin
Trails
On top
(upper face)
4
f
Massive or homogeneous
Hetero!geneous
Ripples
Horizontal
Laminations
Oblique (foresets)
Internal organization
of beds
Graded bedding
Growth
Physical
origin
Post depositional
deformation
I
Physico-chemical
origin
Organic
I72
Normal
Inverse
Load marks
Convolute bedding
Injections
Fractures
Falling
Slumps
Folds
Faults
1{
svlOlites
Bird's eyes
Dish structure
Burrows
Serralog 0 2003
Structure
As previously indicated, the vertical resolution of the
dipmeter tools and more especially of imaging tools
allows recognition of features as thin as a few millimeters
thick (Table 4-4).
I Chapter 4 I
173
Consequently, most of the sedimentary features will be
detected by these two kinds of tools.
Table 4-4
Vertical resolution of the well logging measurements.
.Electrical
. . . . . .'. .images
.... . . I
Dipmeters - - . . - . - .w--1
Acoustic images
hicrolog w
b
Electromagnetic (GH) t+
Micro SFL W
Litho - density - - - W
Radioactivite - - - - Acoustic 1-b
Azimuthal resistivity
b
-I
Laferolog +
Spherical log t-w
Neutron - - - t,
Spectrometry +
Array induction
Dual ihduction - - SP
i
I
I
I
I
0.0005
0.001
I
I
I
I
0.005
0.01
I
I
0.05
I
I
0.1
I
I
0.5
I
I
1
-
b
VSP b-I
1
I
5
*
I
I
10
Scale (m)
O m
0 3 Fractures
a > * -----------------
0.0
0
&
5
r;/l
- - - Small sedimentary features - - - - - - - - Large sedimentary features - - - - - Sedimentary bodies - - Faults
b
. . . . . Higher sensitivity than resolution
Serralog Q 2003
- - - - - - Higher resolution techniques
173
Bed description
Bed shaue and beddina planes
The shape of a bed is determined by the attitude of its
limiting surfaces. Most of the individual surfaces penetrated by a well can be approximated by planes (excluding
trough cross bedding, stylolites and load casts) because
the size of the borehole is small compared to their radius
of curvature. But, it happens that the surface is not really
planar even at the scale of the borehole diameter.
The attitude of a structural plane is its position in three
dimensions, expressed quantitatively by its strike and dip.
The strike of a surface is the direction of its intersection
with a horizontal plane. The dip of a surface is the angle
that it makes with a horizontal plane, measured in the vertical plane perpendicular to the strike.
When the surfaces are parallel over a representative
thickness one can suppose even, parallel and continuous
surfaces and a relatively wide extent of the beds (Fig. 437).
GEODIP presentation they correspond to the computation
of four planes by combination of the resistivity curves
three by three (1-2-3, 2-3-4, 3-4-1, 4-1-2). In the LOCDIP
or SYNDIP program presentation they are emphasized by
a wavy symbol.
Figure 4-38 - Example of even, non parallel and continuous boundaries. a :shale bed; b :sandstone beds; c :cross-bedding; d :unparallel
boundaries; e :parallel boundaries. Observe the erosion at the base of
bed b.
Examples of discontinuous wavy boundaries are given
in Figure 4-39.
-I
Figure 4-37 - Example of even, parallel and continuous sutface boundaries. Observe the very good consistency of dips. Two higher dips are
related to wrong correlations.
Surfaces, which may be flat, but non parallel, at the
scale of the borehole, are shown in Fig. 4-38. They will
probably remain non parallel at the scale of the layers.
The upper and lower boundaries of each bed can show
different dip magnitudes and sometimes different azimuths suggesting non parallel boundaries. Examples of
wavy, non planar surfaces are given by Fig. 4-38. In the
174
J
Figure 4-39 - Example of discontinuous wavy beds orlenses. The
core description on the side confirms the HDT interpretation.
Serralog 0 2003
Structure
Bed thickness
DIPS
I
Chapter 4
I
175
CORRELATIONS
Standard tools do not allow a precise determination of
the bed thickness due to their lack of vertical resolution
and to the fact that the borehole deviation and the dip and
azimuth of the bedding surfaces are not known.
This is not the case with dipmeter and image tools.
As soon as laminae are thicker than 1 cm for dipmeter
tools and 5 mm for imagery tools, these tools represent
the best and most accurate approach for determination of
the bed or lamina thickness. This is due to the fact that the
borehole drift and the dip of bed can be taken into account
for the thickness computation. This determination is not
possible from core as its orientation and the borehole
deviation are generally not known precisely.
The bed thickness computation is explained on Figure
4-40.
North
Figure 4-40 - Graph explaining the bed thickness computation.
Figures 4-41 to 4-44 illustrate different bed thicknesses as they can be observed on dipmeter curve displays
and images.
Nature of bed boundaries
The standard logs allow the detection of surfaces limiting beds of different lithologies or textures, consequently
the apparent thickness of each bed, but they do not allow
the determination of the real thickness of the nature of the
surface.
On the contrary, dipmeters may allow the determination of the nature of the surfaces limiting beds as illustrated by previous figures.
But, the best way to determine the nature of the surfaces and their origin is from image observation. Surfaces
can be classified based on several parameters such as
resistivity contrasts and values, proximity between surfaces, consistency of dip magnitude and azimuth, etc.
Figure 4-45 illustrates several types of bed boundaries as
they can be observed on a Formation Microscanner
image. One can imagine the interest to recognize these
features for facies definition and reconstruction of the
environment.
Serralog 0 2003
Figure 4-4 I - Thin beds well detected from the HDT resistivity curves.
One can observe typical thickening-up sequences of resistive beds.
This suggests a distal turbidite deposit.
Erosions, truncations, load casts, hardgrounds, etc.
will be illustrated during dipmeter or images examples
showing internal organization of beds.
Internal oraanization of beds
Curve homogeneity, especially on dipmeter or image
data, may reflect a homogeneous massive bed as illustrated by Figure 4-46.
Remark
A homogeneous bed must not be confused with an interval without
curve activity due to poor choice of EMEX current. In this case the resistivity curve is saturated in the very high or the very low resistivity region
(compare the logs in Fig. 4 4 7 recorded on same interval, but with two
different EMEX). The EMEX current is the current emitted by the whole
sonde. It focusses the current emitted by the button electrodes. It is
generally chosen to have the best contrast either in low or in high resistivity environments. In the SHDT and image resistivity tools it can be
automatically adjusted.
175
Figure 4-42 - Very thin beds very well detected from the SHDT resistivitv curves. The bed thickness varies from less from 1 cm to 5 cm.
and ‘‘fedpatterns which suggest
shows a lot Of
This
fan progradation with direction of transport towards NE, and draping of
the previous fan by the following one.
Figure 4-44 - Recognition of thin beds by various log measurements,
On the left microSFL, then FMs images,, on the right EPT measurement (eaff and tpl).
Figure 4-43 - Segmentation of one meter FMS images in beds
(from Trouiller et al., 1989).
Laminations are generally well detected by dipmeter
resistivity curves (if the EMEX has been well selected) as
illustrated by Figures 4-41 and 4-42. GEODIP and LOCDIP processings allow the detection of each bed boundary and, consequently, one can recognize parallel laminations, foreset or cross-bedding (Fig. 4-48) which is difficult
to do from results of processing using correlation interval.
176
1
Figure 4-46 - Example of homogeneous bed seen on HDT
Logic based on the curve activity (FBR for frequency of
breaks) and the density of correlations (DCL) (see Fig. 449)aHows the classifications Of the laminae (Table 4-5).
Serralog 0 2003
Structure
FMS
IMAGE
NATURE OF
BOUNDARlES
HOLE
DMFf
mm
Chapter 4
177
OOllllEWlCIMY
WARP. PLANAR
SHARP, CONVEX
SHARP, CONCAVE
WARP. CONVEX
SHARP. WAW,
SYMMETRICAL
SHARP. WAW,
ASYMMETRICAL
SHARP. W A W
WAW. GRADATtONAL
SHARP. IRREGULAR
Figure 4-48 - Example of trough cross-bedding, foreset bedding, erosional surfaces and sand lens detected on HDT curves.
Table 4-5
Logic applied to the dipmeter curves for determination of
the type of internal organization.
ABRUPT
IRREGULAR
ABRUPT
PLANAR
Figure 4-45 - Several types of bed boundaries can be observed on
borehole wall images.
I
--I
Figure 4-47 - Example of change of curve appearence due to change
of the EMEX current. According to the choice of the EMEX value, the
curves are saturated in low or high resistivities.
Serralog Q 2003
This logic, explained by Figure 4-49, has been used in
the SYNDIP processing of which a result is reproduced in
Figure 4-50 (see further for explanation of the SYNDIP
program).
I77
178
Well logging and Geology
Figure 4-49 - Typical curve activity allowing the recognition of the internal organization of beds. No curve activity (flat curves) is interpreted as
indicating massive and homogeneous bed. High frequency of breaks
without correlations is interpreted as indicating a heterogeneous interval. High frequency of breaks with high density of correlations is interpreted as indicating either laminations or cross-bedding if the resistivity
contrast is low, or as succession of beds of two lithology types if the
resistivity contrast is high.
But the best tools for recognition of the internal organization of beds are imagery tools as for the first time we
“see” the formations. This will be illustrated by the following figures.
J
Figure 4-50 - Example of results of SYNDlP processing of resistivity
curves.
Figure 4-51 - FMS image illustrating cross-bedding with truncations.
On the left, raw images in dynamic presentation; on the right, same
images with superposed correlations for dip computation.
As known, processing of the recorded data provides
images which look like core photographs, with the following characteristics:
- high vertical resolution (0.2 in. [5mm])
178
- coverage of 80 % of the borehole wall in a 8 in. diameter (for FMI tool of Schlumberger)
- complete coverage of the borehole wall for ultrasonic
images
- very large dynamic range for electrical images - from
less than 0.1 to more than 10,000 ohm-m
- high sensitivity, allowing detection of very thin (fractures)
or small events (vugs, pyrite crystals) for electrical images
- high sampling rate - one sample each 0.1 in. [2.5mm]
for electrical images
- low sensitivity to heavy mud, borehole ovalization and
rugosity
- image enhancement allows detection of very minute
variations of conductivity.
Serralog Q 2003
Next Page
Structure
I Chapter4 1
179
J
Figure 4-53 - Other example of planar cross-bedding (foresets) with a
layer showing contorted bedding, with shale drape and shale protuberance towards the top of the image (courtesy of Halliburton).
Figure 4-52 - Example of cross-bedding showing different dips and azimuths indicating different transport current directions. The depth scale
is in meters. Several thin shale layers (more or less horizontal black
brown features) are intercalated. Observe small more or less vertical
features at 180" difference in azimuth. They correspond to small
drilling induced fractures (courtesy of Schlumberger).
Recommendations
Image interpretation is a multistep process of observation, description, interpretation by reference to previous
examples and experience, and implication or exploitation.
To be correctly interpreted images must have the
same horizontal and vertical scale. A 1/1Oe scale is generally the best scale for a good observation.
It is often necessary to use enhanced images but their
interpretation must be done with the static normalized
Serralog 0 2003
images as a reference to check the resistivity range of the
observations.
Images must be interpreted with the help of the other
openhole logs, preferably recorded at a high sampling
rate to relate events seen on the images to petrophysical
properties and vice versa.
When comparing images and core photographs, only
bed boundaries, laminations and fractures that are not
oblique to the borehole axis will be common to the two
sources of data (Fig. 4-54).
When cores exist one can orient them with the DIAMAG€ program written by ELF engineers and commercialized by Schlumberger. It is also possible to better locate the core losses. Another application is the precise calibration of resistivity with core porosity and permeability as
179
Previous Page
the plugs can be perfectly located on the images (see Fig.
4-54). In heterogeneous formations, what is seen in the
core cannot be on the borehole wall and perfect correspondence between core and image is unlikely.
Illustration of structures by images
Bed boundaries
Bed boundaries are marked by abrupt and relatively
high resistivity contrasts that cross all the images. These
lines are easily correlated from pad to pad and are visible
on static images. Careful examination of the shape of the
lines on each pad determines the nature of the bed boundary (cf. Fig. 4-45 and Fig. 4-56). These lines correspond
generally to surfaces or boundaries separating two beds
of different lithologies or porosity. Their length defines the
bed thickness and geometry.
When several consecutive bed boundaries, planar and
parallel over a relatively long interval, delineate consecutive conductive and resistive beds, they generally characterize parallel bedding. In sand-shale series, parallel bedding indicates a low energy environment, below the wave
or tide influence that was undisturbed by bioturbation.
1
Figure 4-54 - Depth match of image with photograph of the external
core surface using the DIAMAGE program of ELF. Observe the fracture on the image and its equivalent on the core. Its vertical extent is not
the same. This is due to the difference in diameter of the core and the
borehole as explained by Figure 4-55.
A
Figure 4-55 - Difference in size between core and borehole wall image
for an inclined surface.
180
X570
Figure 4-56 - Typical erosional surface marked by a sharp irregular
contrast with load cast.
Variations in thickness along the images are also observed. These features correspond to wavy bedding characteristic of a higher energy level than that reflected by
parallel bedding (Fig. 4-57). Laterally it can evolve either
towards parallel or lenticular bedding. The latter is characterized by lenses of more conductive (dark) or more
resistive (white) material (Fig. 4-58). Lithology of the lenses is not easy to determine. When dark lenses correspond to shales they indicate a higher energy than
wavy bedding; these lenses do not constitute permeability barriers. When white lenses correspond to sands, they
indicate a lower energy depositional environment.
Load casts are characterized by “ a slight bulge, a deep
or shallow rounded sack, a knobby excrescence or a
highly irregular protuberance”. They are produced “by the
exaggeration of the depression as a result of unequal
settling and compaction of the overlying material and by
the partial sinking of the such material into the depression
as during the onset of deposition of a turbidite on unconsolidated mud‘ (Bates & Jackson, 1980).
Serralog 0 2003
Structure
I Chapter 4 I
181
may be abrupt, tangential or concave. These features correspond to foreset bedding or cross-bedding. Jopling
(1965) gives a list of useful indices of foreset laminae
which may serve as a qualitative guide in determining the
current strength and the mean grain size.
Figure 4-57 - Non parallel wavy bedding as seen on Formation
Microscanner images (courtesy of Schlumberger).
Pad orientation
1
Figure 4-59 - On the lefi, example showing different types of bed boundary, very thin laminations in siltstone (in red) and in sand (in yellow),
with some thin shale laminae (flasers ?). A shale bed at the bottom and
in the center. Observe vertical red lines which can correspond to fractures (courtesy of Halliburton).
On the right, foreset and truncation in eolian sand with the core photograph on the side. One can recognize on image thin layers more
cemented which are not detected on core (courtesy of Schlumberger).
Figure 4-58 - Isolated sand lens in a shaly formation as seen on
Formation Microscanner images.
Internal oraanization
Features that are parallel, planar and oblique to bed
boundaries may be observed on a relatively thick interval
on all the pads. They are sometimes interrupted by other
parallel features inclined in different directions. Resistivity
contrasts, caused by changes in porosity (possibly related
to diagenetic effects) and/or grain size (alternating coarser laminae and thin fine-grained laminae), are often observed (Fig. 4-59). The shape of the contact with the base
Serralog 0 2003
The angle of foreset dips is about 30" at low velocity
and is less than 20" at high velocity. The foreset slope
becomes steeper when the sand grains are coarse and
angular, sorting is poorer and clay is absent (McKee,
1957).
With increasing velocity, the character of contact between foreset and bottomset changes from angular to tangential and even to a sigmoidal shape.
Two populations of laminae dipping abruptly and practically in opposite directions correspond to herringbone
cross-bedding.
Observation of foresets in carbonates indicates that
carbonates have a grainstone texture (Fig. 4-60 and 4-61)
since foresets can only be generated with progradation of
clastic material.
The dip azimuth of the foresets gives the direction of
181
the transport current. This also corresponds to the direction of the progradation of the body (Fig. 4-60).
Foresets introduce permeability anisotropy and, sometimes, permeability barriers when they are cemented or
truncated by other features. This must be taken into
account when defining the block size and the position of
injection wells during a secondary recovery phase.
-I
Figure 4-62 - Small scale cross-bedding organized in cosets, truncations and overturned feature (at the top) are well seen on these FMS
images. On the side a 30 view of the overturned feature is reproduced
(from Serra, 1989).
Figure 4-60 - Foresets observed in a carbonate. They indicate a
grainstone texture of this formation.
Fgure 4-61 - Photograph of the same formation seen on outcrop. One
can see in the Massangis quarw, France the foreset laminae in this
carbonate of the Dogger formation.
Very fine, hairline features with curved surfaces generate with other similar surfaces a festoon shape. They correspond to trough cross-bedding. Depending on their
size, they may not necessarily cross the borehole and are
not seen on all pads (Fig. 4-62).
182
d
Figure 4-63 - Succession of planar cross-bedding with shale drapes,
contorted bedding and non planar cross-beds with shale intrusion characterizing the upper section of a fluvial channel sand
(courtesy of Halliburton).
Serralog 0 2003
Structure
Chapter 4
183
Contorted and overturned structures can be well recognized on images as illustrated by Figures 4-62 and 4-63.
Flasers and water escaped features are also easily
observed on images without need of core (Fig. 4-64).
A
Figure 4-64 - Flasers, bifurcated flasers (dark “chips” features), and
water escape features are well observed on this FMI image over this
sand deposited in a intertidal environment with a marked increase in
water depth from the bottom to the top (courtesy of Schlumberger).
Deformation of layers or laminae are sometimes observed as previous illustrated (cf. Fig. 4-62 and 4-63). The
deformations can be over a longer interval. They correspond generally to slump structures (Figs. 4-65 and 466) or to microfolds (Fig. 4-67 next page).
1
Figure 4-66 - 3 example of slump structures. At the top, coming from a
2-pad FMS tool (Serra, 7989). At the bottom, image coming from a FMI
tool (courtesy of Schlumberger).
Irregular features either more conductive or resistive,
especially those vertical or oblique intersecting laminae or
bed boundaries, can sometimes be observed. These images frequently correspond to bioturbation by animal or
plant activity (Figs. 4-68 and 4-69). They indicate either an
oxidized environment with strong biologic activity or emersion when they correspond to root traces.
In some cases typical features correspond to fossils as
illustrated by Figure 4-70.
Graded beddinq
Ll
Figure 4-65 - Slump structure seen on FMS-2 images (internal figure)
confirmed by the Core Photograph (external View) (from Serf8 7989).
Serralog 0 2003
Progressive vertical evolution of the color (or gray) correspond to graded bedding. If they have sufficient vertical
evolution, the other logs such as gamma ray, thorium,
density, neutron and sonic should show similar trends.
183
Next Page
o
90
~0210360
Yettical burrow
Boturbated
zone
Root traces
Figure 4-69 - These FMI images show in a sandstone-mudstone
sequence vertical and horizontal burrows and zone of bioturbation
superposed to a period of emersion confirmed by the root traces seen
at the bottom of these images (courtesy of Schlumberger).
Core photograph
Static FMS image
Enhanced FMS image
832.4
Figure 4-67 - Microfold with steeply dipping limb
(courtsey of Schlumberger).
832.4
Core photograph
Enhanced FMS image
Burrow
Pyrite
-
Figure 4-68 The white vertical feature on the leff of the FMS image
corresponds to a vertical burrow. Other white dots correspond to
oblique or horizontal burrows. The dark feature with sharp and linear
edges corresponds to a pyrite crystal. The core photograph on the left
collected at the same depth shows similar features confirming the
interpretation (from Serra, 1989).
Figure 4-70 - The crescent shape of the white even observed below
the arrow on the enhanced FMS image (central track) suggests an
oyster shell. On the core photograph at the same depth a similar oyster is observed (from Serra, 1989).
When these vertical trends are observed over a long
interval they correspond either to a fining-up sequence in
a fluvial deposit (Fig. 4-71), or to a coarsening-up sequence due to progradation in a deltaic environment (Fig. 472).
When these trends are observed on short interval they
correspond to turbidite deposits. In such case one can
also observe thickening-up sequences (Fig. 4-73 nextpage). In such case the other logs cannot detect such trends
due to their lack of vertical resolution.
Other features such as conglomerates, grain or mudSerralog Q 2003
Previous Page
Structure
supported, and breccias can be recognized on images
Chapter4
j
185
(cf. Figs 3-56, 3-57, 3-62, 3-63, 3-84c and 3-85).
0
DEPTH
Figure 4-71 - Example of openhole logs recorded with a high sampling rate indicating a fining-up sequence with an abrupt erosional contact at X76
m and X53.5 m. The images on the right, reproduced at the same vertical scale, show the same trends. White yellow color corresponds to sandstone, dark brown color to shale, intermediate color to silt (from Serra, 1989).
140
200
260
320
20
80
140
In the same way, anhydritic or chert nodules can be
recognized. They are characterized by very resistive
peaks on few resistivity curves, or more or less irregular
white spots on images.
The determination of the mineralogic nature of the very
resistive features requires other log data to recognize the
nature of the deposit: detritic for conglomerates, carbonate for chert or anhydrite nodules the latter being frequent
in dolomitic deposits (Fig. 4-74 next page).
Automatic analysis of dipmeter data.
The SYNDIP program
Figure 4-72 - Typical coarsening-up sequence seen on FMS images
(from Serra, 1989).
Serralog 0 2003
As demonstrated in the previous chapter and in the
first part of this one, the interest of the dipmeter data to
extract information on the texture and the sedimentary
structure of rocks is obvious. This information is very
185
186
Well logging and Geology
PAD ORIENTATION
PAD ORIENTATION
Figure 4-73 - Typical grain size evolutions detected by FMS images.
They correspond to color trend between white which corresponds to
sand and black which correspond to shale. Observe also the thickening-up sequence. These two types of observation characterize
turbidite deposits (Sera, 1989).
important for a better and more accurate definition of the
electrofacies, consequently of the facies and the sedimentary environnements, and cannot be ignored.
But this information is qualitative by nature: character
of the dipmeter resistivity curves, type of the bed boundaries, dip evolution with depth...
Consequently, its utilization for the automatic determination of electrofacies is only possible if the data, extracted from the dipmeters and their processing by GEODIP
or LOCDIP programs, are quantified. On the other hand,
to be used those data must be assigned to the electrobed
or electrosequence.As known, fast channel dipmeter data
are sampled each 5 mm for HDT tool, or 2.5 mm for SHDT
or for imagery tools instead of each half-foot (15 cm) or
exceptionally each 1.2 inch for wireline tools.
To achieve these goals, dipmeter resistivity curves and
186
Figure 4-74 - Anhydrite growth (white spots) in a sandy dolomite.
Central images correspond to an isolation of the white spots. Right curves are the results in density of spots (in blue) and the area they cover
(in violet) coming from a quantitative analysis of the central images
(from Delhomme & Motet, 1993).
dip computation results were initially described by a series
of parameters :
- the variability or activity of the curves (VAR), which
reflects the homogeneity (very low variance), or the
heterogeneity (high variance) of the formation,
- the frequency of events (peaks or troughs) per curve
(FRE) over a given interval (6 in.) as recognized by the
GEODIP program (cf. Fig. 4-49),
- the average event thickness per curve (ALT) (P9, parameter of pattern vector computed during the GEODIP processing),
- the balance of positive to negative excursions of the
resistivity curves or, in other words, the ratio of the average thickness of the peaks over the average thickness of
Serralog 0 2003
the troughs in the interval (BAL). If these peaks and
troughs correspond to sand and shale beds, this parameter is used to compute a sand-shale ratio (cf. Fig. 4-50),
-the density of correlations found by the GEODIP or LOCDIP program (DEN), the frequency of events (FRE) can
be high and the density of correlation (DEN) can be low if
these events are not similar and consequently not easily
correlated (case of conglomerates, recifal boundstones,
zones of bioturbation),
- the sharpness of the curve events (averaging from the
four resistivity curves) (SHA),
- the average resistivity of the interval (SRES).
These parameters were computed for each 6-in. interval corresponding to 30 samples. They represented an
attempt to convert the information included in the curve
shape to a quantitative form useable in the definition of
electrofacies similar to a log. It is for that reason that the
name of "synthetic logs" was given to these computed
curves.
Since some of the HDT or SHDT-derived logs can
reflect same phenomena (i.e. FRE, DEN, ALT, VAR), and
consequently can be highly correlated (Fig. 4-75), a certain amount of redundancy is present. Some of these
synthetic logs exhibit a high degree of correlation with
openhole logs such as gamma ray, or spontaneous potential (Fig. 4-76). Comparison of the two shows that positive
excursions of the new curve indicates heterogeneity
(laminations, poor grain sorting, etc.) and the negative
side indicates homogeneity (well-sorted sands, etc.). This
curve can then be used in the processing as a textural
and structural indicator, instead of FRE, DEN, VAR ....
The SYNDIP Proaram
SYNDIP is a program developed by Schlumberger
(Delhomme & Serra, 1984), to replace the previous procedure. It generates dipmeter-derived (or synthetic) logs
from HDT or SHDT raw data and computation results. The
so-called synthetic logs are based on the features and
likeness of the microresistivity curves, and on the evolu-
J
Figure 4-75 - Correlations between some of the synthetic logs derived from dipmeter data (from Serra, 1989).
Serraiog 0 2003
187
tion versus depth of dips and planarity. SYNDIP can be
focused, according to different criteria (dip quality and planarity, size of curve events ...).
In normal use, SYNDIP synthetic logs are output with
a half-foot sampling rate; this rate was chosen in order to
be consistent with the sampling rate of most open-hole
logs. However, for detailed studies, another rate (e.9.
1.2", that is the usual sampling rate for EPT tool measurements) can be selected.
The frequency of inflexion points, correlated or not, on
a single dipmeter resistivity curve is first computed as an
indicator of curve activity. The sharpness of this synthetic
log is ensured by the tact that no window is used to compute frequency; rather, it is based on the thickness bet
ween consecutive inflexion points (ATBR).
maintained.
An idea of the bed internal organization is obtained by
combining curve activity with density of correlations. High
activity with no correlation will reflect a coarse heterogeneous bed organisation (conglomerates, flaser bedding,
reefs, vuggy limestones, bioturbation ...). Low or no activity with no correlations will reflect either a homogeneous
bed or a very finely heterogeneous bed. High activity with
high density of correlations will correspond to thinly larninated formations (cf. Table 4-5).
As for bed boundaries, a non-planarity flag is output
when the GEODIP or LOCDIP planarity coefficient is
below a certain threshold. This indicates non planar and
possibly erosive surface.
The parallelism between consecutive bed boundaries
can be appreciated by comparing the corresponding dipping planes; the angle between a layer's top and bottom
is computed and a non-parallelism flag is output if the
angle is higher than a certain value (e.9. 10'). Dip dispersion is also computed in a given sliding window (typically
3 m or 10 ft wide); it is an angular standard deviation computed on the unit hemisphere.
Intervals with low dip dispersion, thin beds, small
angles between bed tops and bottoms, and with a lithology that corresponds to low energy deposits (shaly-silty or
shaly-marly laminations) are those intervals where the
structural dip can be picked.
The nature of contacts and transitions is also identified. The high vertical resolution of the dipmeter and its
sampling rate allow separation of abrupt changes from
gradational ones. This last type corresponds to conductivity ramps, which generally reflect grain size or lithology
evolution (sequences). SYNDIP ramp analysis outputs
small and large-scale ramps.
Lastly, in SYNDIP, dipmeter data are corrected for
EMEX (i.e. for the value of the total current sent out into
the formation), rescaled and averaged in conductivity over
the output sampling interval since levels appear in parallel.
Description of a SYNDIP Display
20-
2
0-
Figure 4-76 - Correlations between HOT derived synthetic curve (DEN)
with SP curve (from Serra, 1989).
A bed, or a lamina, can be defined as the depth interval between two consecutive correlations found by GEODIP or LOCDIP programs. Correlation lines link upper or
lower inflexion points of similar events on several dipmeter curves. Layer thicknesses are thus computed between
consecutive correlation links (ATCL). When correlation
links are closer than the output sampling interval, an average thickness is simply computed, but the resolution is
188
An example of the SYNDIP graphic output is given in
Fig. 4-50. The first track reproduces, on a logarithmic
scale, 3 calibrated resistivity curves of dipmeter : the minimum and maximum values recorded by the 4 or 8 buttons
in thin lines, and the average value computed from these
4 or 8 resistivity measurements as heavy lines. This presentation enables detection of homogeneous or heterogeneous beds.
The second track displays the internal organization of
the beds : small dots represent the homogeneous or massive beds, bubble-like shading correspond to heterogeneous beds, and dark gray with horizontal lines the laminated beds. The limiting curve is the frequency of inflexion
points, an indicator of the total curve activity (I/ATBR),
and the frequency of correlations (I/ATCL).
On the side, in three columns, are flags indicating
detection of non planar surfaces, non parallelism between
Serralog 0 2003
Structure
consecutive bed boundaries, and correlated parallel planes. The following column indicates intervals containing a
minimum of five dips with angular spherical dispersion
lower then 5",with the computed average dip. This represents the structural dip if the interval corresponds to a low
energy environment (shale, silt or marl).
Sometimes, selected dip results are represented in a
third track. The dip, which is taken as representative of a
1.5 ft wide interval, is the one which minimizes the spherical distance to the (n/2)th closest dip result among the n
ones found in this interval. The solid curve to the right is
the dip spherical standard deviation, and the dashed line
is the angle between current layer's top and bottom.
The next track shows a conductivity curve, shaded
with a continuous grey or colour scale to reinforce the
sand/shale opposition in sandlshale series (cf. Fig. 4-50).
Colours are selected by using the histogram of dipmeter
resistivity measurements. Sometimes, two scales of
conductivity ramps are displayed on the left side.
The last track displays thicknesses of conductive or
resistive layers in two mirror logarithmic interpolation between the 8 resistivity curves, after equalization, following
the dips computed by the LOCDIP program.
Geologists must be convinced, if necessary, that the
measurements realized by dipmeters contain much more
than the dip information. The resistivity curves reflect in
detail the internal organization of the formations crossed
by the well. This information must be exploited in order to
have a return on investment. Oil companies paid to aquire this information. Consequently, why not exploit it?
"Squeeze the lemon until its last drop!"
Complementary remarks on features seen by dipmeters and imagery tools
The above mentioned examples demonstrate that dipmeter and imagery tools provide a mass of information,
covering the texture as well as the sedimentary structure
of rocks, the direction of transport and the thickness of the
beds. The advantages of their integration in all subsurface sedimentological studies can therefore be understood.
It is obvious that HDT; SHDT and imagery tools do not see
all sedimentary features: they detect only those presenting a minimum resistivity contrast. Thus, the figures visible to the naked eye due to a color change are not detected unless this change is accompanied by a resistivity
variation. The features appearing on the surface of a bed
without repercusion on a vertical section (tool marks, rain
imprints, ...), cannot be recognized because the tool only
analyses a cylindrical section.
In general, one can admit that all events detected by
dipmeter or imagery tools (corresponding to resistivity
variations) inevitably explain a change of geological parameters (mineralogy, texture, sedimentary structure, fluids
...) with the condition, of course, that the pad is properly
applied to the borehole wall and the tool is working correctly.
Serralog 0 2003
I Chapter 4 I
189
Rules for interpretation of dipmeter and imaae data
The interpretation of a dipmeter arrow-plot should
never be implemented without the integration of all other
available data, consequently the open-hole logs.
- The first step of the interpretation must be the compilation of a composite log combining at 11200 scale all available logs and the GEODIP or LOCDIP plot or images
after depth matching. The result of a processing describing the lithology or mineralogy can also help considerably if it is reproduced alongside.
- The second step consists of the observation and description of the following aspects, by refering if necessary
to the GEODIP or LOCDIP plot at 1/40 scale, or to the
image data at 1/10 or 1/5 scale. At this stage it is necessary to separate dips corresponding to tectonic structure
from sedimentary dips generated by current (Fig. 4-76).
- The investigated interval is subdivided into zones with
approximately constant characteristics.
- The mineralogical composition of each zone, and of
each event in the zone, is defined as precisely and carefully as possible. To achieve that, the data of other logs,
especially litho-density, neutron, sonic, natural gammaray spectrometry ..., or a LITHO display, must be used.
- The nature of the contact (abrupt or gradational, planar
or warped, conformable or not) is observed.
- The type of layer succession is described simple or composite, with or without parallel boundaries, continuous or
discontinuous.
-The thickness of each bed type and its evolution with the
depth must be noted.
- The existence of current features (flaser, wavy, or lenticular bedding), revealed by the thickness of the events
and the dip variations (magnitude and azimuth), or by the
images, must be extracted and analyzed.
- Internal structure of a bed including the curve shape
(massive and homogeneous, heterogeneous, resistivity
ramp revealing a graded bedding), and dip evolutions
(oblique bedding, cross-bedding, foreset,...) must be
added. The amplitude of dip variation is also analysed.
The absence of these variations indicate either a low
energy, or, on the contrary, a very high energy. The choice between the two hypotheses is derived from the mineralogical nature or the vertical position of this phenomenon in the granulometric sequence, and from the dips.
The important variations of the dip magnitude indicate
changes of energy in the environment.
- Finally, the sequential evolutions, the rhythms or cycles,
the evolution in thickness of each bed and sequence as
they are revealed by the dipmeter resistivity curves or the
images, are studied.
- The third step corresponds to the direct interpretation. It
involves the translation of the observed features into geologically meaningful interpretations. For instance a "blue
pattern" (increasing dip magnitude upward) will be interpreted as a foreset.
- The fourth step constitutes the deductive interpretation
189
190
Well logging and Geology
plldsrad w s
Local great
circle axes
Figure 4-77 - To restitute correctly the sedimentary dips, reproduced on the right, it is necessary to substract the structural dip. The left figure
reproduces the dips determined over a 6 meters interval. It is difficult to separate the structural dip from the sedimentary dips. Based on the determination of the local great circle axes (cf. Fig. 4-77 below), the structural dip can be determined and substracted (central figure)
(from Etchecopar & Bonnetain, 1992).
Figure 4-78 - Trough cross-bedding surfaces fit great CifC/eS With different axes. Those axes fit a great circle which allows the determination
of the structural dip (from Etchecopar & Bonnetain, 1992).
190
in terms of a depositional environment. Analysis of azimuth frequency plots established on selected intervals will
help to define the uni-, bi- or polymodal nature of the dips.
To achieve this, we have to integrate data obtained
from the observation of cuttings and cores, such as the
presence of glauconite, lignite fragments, phosphate,
shells; heavy minerals; grain size, granulometric sorting,
nature of cement, shale type, ...
One proceeds by the elimination of the environmental
hypotheses which do not fit with the observed features.
The final selected solution among the remaining hypotheses, is the one which fits the best with the geological knowledge we have of the formation, the basin and the main
tectonic features.
It is suggested to summarize all the observations in a
table (Table 4-6) or in several columns put alongside the
composite log at 1/200 scale, one column for lithology,
one for sedimentary features, one for sequential evolution, one for the direct interpretation and complementary
one for the finalinterpretation in terms of facies,
subenvironments and environment.
Serralog 0 2003
Structure
Quantitative analysis of image data
As done for dipmeter data, images must be analyzed
quantitatively in order to extract and classify the information they contain (Fig. 4-79 next page).
As previously seen in Chapter 3, heterogeneities,
represented by features more or less resistive than the
background, are analyzed thanks to a processing similar
to the BorTex program developped by Schlumberger.
Their frequency, their shape and their size can be determined.
Surfaces can be classified as a function of several criteria:
- amplitude of the resistivity contrast on each side of
the surface. Low amplitude will suggest the persistency of
the same lithology; high contrast will suggest an important
change in rock characteristics either due to lithology change or to porosity change.
- proximity of succesive surfaces. A very short depth
1
Chapter 4
~
191
ne bedding. Change of dip and azimuth with hogh resistivity contrast will probably suggest an unconformity.
Sedimentological applications of
sedimentary structure detection
The main application is the determination of the depositional environment.
Without minimizing the sedimentological interest of the
CLUSTER, MSD or BORDIP type programs - see the
works and papers of Gilreath et a/. (1964, 1969, 1971),
Campbell, (1968), Goetz et a/. (1977), Selley, (1979),
Serra (1985) - it should be underlined that these processings do not exploit the very detailed analysis of the dipmeter that is now possible with thelast dipmeter tool (e.g.
SHDT) or imagery tool and improved processing techniques. The curves were generally not shown and the precise events from which the dips are computed are still
unknown. These dips are determined with the help of a
Table 4-6
Summary of the internal structure information.
I
interval between consecutive surface with low reistivity
contrast will suggest laminations.
- planarity of the surfaces. Irregular non planar surface
will probably correspond to an erosional surface r to a presence of load cast.
- dip and azimuth consistency. Practically no change in
dip and azimuth combined with low resistivity contrast and
high proximity will correspond to parallel laminae or foresets from which the transport current direction can be
determined. Change in dip and azimuth with low resistivity contrast can correspond to cross-bedding or herringboSerralog 0 2003
correlogram established from a correlation of events in a
given interval, without distinction and selection of their origin. The obtained dip is consequently an average dip for
an interval which can cover several types of surfaces:
sedimentary units, each of them possibly having sedimentary features with different dips, fractures, and isolated events. For these reasons it is highly preferable to use
a GEODIP or LOCDIP presentation for sedimentological
interpretation of dipmeter data.
As previously stated the primary sedimentary structu191
Q U A N T I T A T I V E BOREHOLE I M A G E ANALYSIS
mw
QEOLOWCM OBJECTSDOWN TO A
UILUYIETERL) s m ARE IDENT~REDwmi BOREHOLE IMAGES
MATHEMATICAL
TOOLS ALLOW TO AUTOMATICALLY Ucriuct AND QUANTIFY THEM
A FAST ZONATION CAN BE EXTRACTED FROM
BlNARlZED IMAGES DERIVED WITH AN
ADAPTATWE BACKGROUND
RESISTIVITY FILTER
Landscape view of the
conductivity image.
Landscape view of the
gradient image.
FOR EACH ZONE, GEOLOGICAL OBJECTS ARE AUTOMATICALLY
EXTRACTED AN0 THEIR ATTRl9UTESARE COMPUTED
THIS METHOD INVOLVES A PRIMARY STEP OF INTERPRETATIONAND THE MACBE ANALYSIS
IS QUIDED BY THE QEOLOQlCAL KNOWLEDOE OF THE FORMAWOMS
DEPOSlTlONAL
SURFACE
ATTRIBUTES:
- Resistivitycontrast
- Dip and azimuth
- Planarity
- Proximity
LAYER
ATTRIBUTES:
-Medianconducthrity
h w
-T
- _
DlAQENETlC
HETEROGENEITY
TYPE AND DISTRIBUTION:
- Size, shape, area
- Frequency
- Connectedsless
- Median#Mducthllty
STRESS INDUCED
FRACTURE
ATTRIBUTES:
- Polarity ( r e s i s t i V 9 , w ~ )- Dlp, pzlmuth
- Lenom
- Density
- Electrlcalaperbwe
- Planarity
Figure 4-79 - Quantitative borehole image analysis and classification of the geological objects as a function of their origin
(from Delhomme & Motet, 1993).
192
Serralog 0 2003
Structure
res -which, can be detected on dipmeter logs or better on
images - are particularly important because they reflect
the hydro- (or aero-) dynamic conditions prevailing in the
environment at the time of deposition.
Another important application is in petrophysical interpretation. As soon as cross-bedding or foresets are recognized, the grain size can be evaluated (cf. Fig. 4-11) and
through the relationship between porosity and permeability as a function of the grain size (cf. Fig. 3-89 Chapter 3)
the permeability estimated. In addition, the permeability
anisotropy can be evaluated (Fig. 4-80).
Figure 4-80 - Permeability anisotropy as a function of the internal organization of the laminae.
Another petrophysical application of dipmeter and
imagery tools is for interpretation of thin bedded reservoirs. The lack of resolution of the standard logs does not
allow their recognition. Only dipmeter and imagery tools
have a vertical resolution allowing to detect succession of
two different types of depositional units with different
petrophysical properties and thicknesses (cf. Figs. 4-41,
4-42, 4-44, 4-52 and 4-53).
References and Bibliography
ALLEN, J.R.L. (1963). - The classification of crossstratified units, with notes on their origin. Sedimentology, 2, p.
93-114.
ALLEN, J.R.L. (1968). - Current Ripples. North
Holland, Amsterdam.
ALLEN, J.R.L. (1970). - Physical Processes of
Sedimentation. Elsevier, New York.
BATES, R.L. & JACKSON J.A., ed. (1980) - Glossary
of Geology. American Geol. Institute, Falls Church,
Virginia.
BLATT, H., MIDDLETON, G. & MURRAY, R. (1972,
1980). - Origin of Sedimentary Rocks. 1st and 2nd ed.
Serralog 0 2003
I
Chapter 4
I
193
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
BOYELDIEU, C. & JEFFREYS, P. (1988). FormationMicroScanner: New developments. SPWLA,
11th Europ. Eval. Symposium Trans., Oslo, paper G.
BOURKE, L.T. (1989). - Recognizing artifact images of
the Formation Microscanner. SPWLA, 30th Ann. Log.
Symposium Trans., Denver, paper WW.
BOURKE, L.T., DELFINER, P. FETT, T, GRACE, M.,
LUTHI, S. , SERRA, 0. & STANDEN, E. (1989). - Using
Formation Microscanner Images. The Technical Review,
37, 1, p. 16-40.
CAMPBELL, C.V. (1967). - Lamina, Laminaset, Bed
and Bedset. Sedimentology, 8, p. 7-26.
CAMPBELL, R.L. (1968). - Stratigraphic applications
of dipmeter data in Mid-Continent. Bull. Amer. Assoc.
Petroleum GeoL, 52, 9, p. 1700-1719.
CAROZZI, A.V. (Ed.) (1975). - sedimentary Rocks.
Benchmark Papers in Geology, 15, Dowden, Hutchinson
& Ross, Inc., Stroudsburg, Pennsylvania.
Chambre Syndicale de la Recherche et de la
Production du Petrole et du Gaz naturel (1966). - Essai de
nomenclature & caracterisation des principales structures
sedimentaires. Ed. Technip, Paris.
Chambre Syndicale de la Recherche et de la
Production du Petrole et du Gaz naturel (1974). Methodes modernes de geologie de terrain. 1 - Principes
d’analyses sedimentologiques. Ed. Technip, Paris.
COLLINSON J.D. & THOMPSON, D.B. (1982). Sedimentary Structures. George Allen & Unwin PubL Ltd.,
London.
CROSBY, E.J. (1972). - Classification of Sedimentary
Environments. In : “Recognition of Ancient Sedimentary
Environments”, edited by RIGBY J.K. & HAMBLIN, WK.,
SEPM, special publication 16.
DELFINER, P., PEYRET, 0. & SERRA, 0. (1984). Automatic determination of Lithology from Well Logs. 59th
Ann. Techn. Conf. SPE of AIME, Houston, Texas: paper
no SPE 13290.
DELHOMME, J.-P. (1992). - A quantitative characterization of formations heterogeneities based on borehole
image analysis.
DELHOMME, J.P., & MOTET, D. (1993). - Reservoir
Description and Characterization from Quantitative
Borehole lmage Analysis. Schlumberger, Africa and
Mediterranean Region, Reservoir Characterization, 5; p.
5-19.
DELHOMME, J.-P. & SERRA, 0. (1984). - Dipmeterderived Logs for SedimentologicalAnalysis. SPWLA, 9th
Europ. Intern. Format. Eval. Trans., paper 50.
DUNHAM, R.J. (1962). - Classification of Carbonate
Rocks according to Depositional Texture. Amer. Assoc.
Petroleum GeoL, Mem. 1, p. 108-121.
DZULINKI, S . & WALTON, E.K. (1965). - Sedimentary
features of flysch and greywackes. Elsevier, Amsterdam.
EKSTROM, M.P., CHEN, M.Y., ROSSI, D.J., LOCKE,
S. & ARON, J. (1986). - High Resolution Microelectrical
Borehole Wall Imaging. SPWLA, 10th Europ. Symp.
Trans., Aberdeen, paper K.
EKSTROM, M.P., DAHAN, C.A., CHEN, M.Y., LLOYD,
193
194
Well logging and Geology
P.M. & ROSSI, D.J. (1986). - Formation lmaging with
Microelectrical Scanning Arrays. SPWLA, 27th Ann. Log.
Symp. Trans., Houston, paper BB.
von ENGELHARDT, W., & GAIDA, K.H. (1963). Concentration changes of pore solutions during the compaction of clay sediments. J. Sediment. Petrology, 33, p.
919-930.
ETCHECOPARA.& BONNETAIN, J.-L. (1992) - Cross
Sections from Dipmeter Data. Arner. Assoc. Petrol. Geol.
Bull., 76, 5, p. 622-637.
FONS, L. Sr. (1969). - Geological application of well
logs. SPWLA, 10th Ann. Log. Symp. Trans.
FRIEDMAN, G.M. & SANDERS, J.E. (1978). Principles of Sedimentology. John Wiley & Sons, New
York.
GALLOWAY, W.E. (1977). - Catahoula Formation of
the Texas Coastal plain: depositional systems, composition, structural develoment, ground-water flow history, and
uranium distribution. Bur. Econ. Geol. Univ. Texas, Austin,
Rept. Invest. no 87.
GARY, M., McAFEE, R.Jr. & WOLF, C.L. (1972). Glossary of Geology. Amer. Geol. Institute, Washington,
D. C.
GILREATH, J.A. & MARICELLI, J.J. (1964). - Detailed
Stratigraphic Control through dip Computations. Bull.
Amer. Assoc. Petroleum Geo/., 48, 12, p. 1902-1910.
GILREATH, J.A., HEALY, J.S. & YELVERTON, J.N.
(1969). - Depositional Environments Defined by Dipmeter
Interpretation Gulf Coast Assoc. GeoL SOC.Trans., 19, p.
101-111.
GILREATH, J.A. & STEPHENS, R.W. (1971). Distributary Front Deposits lnterpreted from Dipmeter
Patterns. Gulf Coast Assoc. Geol. SOC.Trans., 21, p. 233243.
GILREATH, J.A. & STEPHENS, R.W. (1975). Interpretation of Log Responses in a Deltaic Environment.
Amer. Assoc. Petroleum Geol. Marine Geology
Workshop, Dallas, Texas.
GOETZ, J.I., PRINS, W.J. & LOGAR, J.F. (1977). Reservoir Delineation by Wireline Techniques.Paper presented at 6th Ann. Conv. Indonesia Petroleum Assoc.,
Jakarta, May 1977.
GOODALL, I., WESTLAKE, I., BLOCKVAGLE, G.,
MUNDAL, I., & MULCAHY, M. (2002). - Resistivity lmage
Data, Voring Basin, Offshore Norway. In: LOVELL, M., &
PARKINSON, N. (ed.) (2002). - Geological Applications of
Well Logs. AAPG Methods in Exploration, 604 p.
GUY, H.H., SIMONS, D.B., & RICHARDSON, E.V.
(1966). - Summary of alluvial channel data from flume
experiments 7956-67. U.S. Geol. Surv. Profess. Papers
462-1, 96 p.
HACKBARTH, C.J. & TEPPER, B.J. (1988). Examination of BHTV, FMS, and SHDT images in very
thinly bedded sands and shales. 63rd SPE Ann. Techn.
Conf. & Exhib., Houston, paper SPE 18118.
HARKER, S.D., McGANN, G.J., BOURKE, L.T. &
ADAMS J.T. (1988). - Formation Microscanner Applications in Integrated Reservoir Description in
Claymore and Scapa Fields (North Sea). Geological
194
Applications of Wireline Logs, London, June 1-2, 1988,
Geological Society of London (in press).
HARKER, S.D., McGANN, G.J., BOURKE, L.T. &
ADAMS J.T. (1988). - Methodology of Formation
Microscanner lmage Interpretation in Claymore and
Scapa Fields (North Sea). Journal of the Geological
Society of London (in press).
HECKEL, P.H. (1972). - Recognition of ancient shallow
marine environments. In: Rigby & Hamlin (eds):
Recognition of ancient sedimentary environments, Spec.
Publs. S.E.P.M., 16, p. 226-286.
HEPP, V. & DUMESTRE, A.C. (1975). - CLUSTER - A
method for selecting the most probable dip results from
dipmeter survey. SPE of AIME, 50th Ann. Fall Mtg.,
Dallas, Paper SPE 5543.
HOBSON, G.D. & TIRATSOO, E.N. (1975). lntroduction to Petroleum Geology. Scientific Press Ltd,
Beaconsfield, England.
HJULSTROM, F. (1939). - Transportationof detritus by
moving water. In: Trask (ed): Recent marine sediments.,
Amer. Ass. Petrol. Geol.
HUTIN, R. (1986). - Formation lmaging with Micro
Electrical Scanning Arrays. A New Generation of
Stratigraphic High Resolution Dipmeter Tool. SPWLA,
10th European Formation Evaluation Symposium
Transactions, Aberdeen, Scotland, April 22-25, 1986,
paper L.
INGRAM, R.L. (1954). - Terminology for the thickness
of stratification and parting units in sedimentary rocks.
Bull. Geol. SOC.Amer., 65, p. 937-938.
JOPLING, A.V. (1965). - Hydraulic factors Controlling
the Shape of Laminae in Laboratory Deltas. Journal of
Sedimentary Petrology, 35, 4 (December 1965), p. 777791.
KRUMBEIN, W.C. & SLOSS, L.L. (1963). -Stratigraphy
and Sedimentation. 2nd ed. W HFree. man & Co., San
Francisco.
KRYNINE, P.D. (1948). - The megascopic study and
field classification of sedimentary rocks. Jour. Geol., 56, p.
130-165.
KUENEN, Ph. H. (1953). - Significant features of graded bedding. Bull. Amer. Assoc. Petroleum Geol., 37, p.
1042-1066.
LANDES, K.K. (1951). - Petroleum Geology. John
Wiley & Sons, New York.
LEET, L.Don, JUDSON, S. & KAUFFMAN, M.E.
(1978). -Physical Geology. 5th ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
LINK, P.K. (1982). - Basic Petroleum Geology. OGCI
Publications, Tulsa.
LLOYD, P.M. (1986). - The formation Microscanner :
A New Generation High Resolution Logging Technique.
LLOYD, P.M., DAHAN, C.A. & HUTIN, R. (1986).
Formation lmaging from Microelectrical Scanning Arrays :
A New Generation of Stratigraphic High Resolution
Dipmeter Logging Tool. SPWLA, 10th Europ. Symp.
Trans., Aberdeen.
LLOYD, P.M., DAHAN, C., & LUTHI, S.M. & BANAVAR, J.R. (1988). - Application of Borehole lmages to
Serralog 0 2003
Structure
Three-Dimensional Geometric Modeling of Eolian
Sandstone Reservoirs, Permian Rotliegende, North Sea.
AAPG Bulletin, 72, 9, p. 1074-1089.
LOMBARD, A. (1956). - Geologie Sedimentaire. Les
series marines. Masson, Paris.
LOMBARD, A. (1972). - Series sedimentaires. Genese
-Evolution. Masson, Paris.
LOVELL, M., & PARKINSON, N. (ed.) (2002). Geological Applications of Well Logs. AAPG Methods in
Exploration, 604 p.
MATTHEWS, R.K. (1974). - Dynamic Stratigraphy. An
introduction to Sedimentation and Stratigraphy. PrenticeHall Inc., Englewood Cliffs, New Jersey.
McGOWEN, J.H. & GARNER, L.E. (1970). Physiographic features and stratification types of coarsegrained point bars : Modern and ancient examples.
Sedimentology, 14, p. 77-111.
McKEE, E.D. (1957). - flume experiments on the production of stratification and cross-stratification. J . sediment. Petrol., 27, p. 129-134.
McKEE, E.D. (1957). - Primary Structures in Some
Recent Sediments. AAPG Bulletin, 41 , 8, p.1704-1747.
McKEE, E.D. (1966). - Structures of dunes at White
Sands National Monument, New Mexico (and comparison
with structures of dunes from other selected areas).
Sedimentology, 7, p. 1-69.
McKEE, E.D. & WEIR, G.W.(1953). - Terminology for
stratification and cross-stratification in sedimentary rocks.
Bull. geol. SOC.Amer., 64, 4, p. 381-390.
MIDDLETON, G.V. (ed.) (1965). - Primary sedimentary structures and their hydrodynamic interpretation.
SEPM, Spec. Pub. 12.
MIDDLETON, G.V. (1976). - Hydraulic interpretation of
sand size distributions. J. Geology, 84, p. 405-426.
NURMI, R.D. (1984). - Geological evaluation of stratigraphic high resolution dipmeter data. SPWLA, 25th Ann.
Log. Symp. Trans., New Orleans.
NURMI, R., CHARRARA, M. & WATERHOUSE, M.
(1988). - Heterogeneities in Carbonate Reservoirs:
Detection and Analysis Using Borehole Electrical
Imagery. Geological Applications of Wireline Logs,
London, June 1-2, 1988, Geological Society of London.
OTTO, G.H. (1938). - The Sedimentation Unit and its
Use in Field Sampling. Journal of Geology, 46, 4, p. 569582.
PAYENBERG, Tobias H.D. (2002). - Paleocurrents
from FMS and Scribe-oriented Core - A comparison. In:
Lovell, M., & Parkinson, N. (ed.) (2002). - Geological
Applications of Well Logs. AAPG Methods in Exploration,
604 p.
PAYRE, X. & SERRA, 0. (1979). - A case study Turbidites recognized through dipmeter. SPWLA, 6th
Europ. Log. Syrnp. Trans., London, paper K.
PERRIN, G. (1975). - Comparaison entre des structures sedimentaires a l'affleurement et les pendagemefries
de sondages. Bull. Centre Rech. Pau, SNPA, 9, p. 147181.
PETTIJOHN, F.J. (1930). - Imbricate arrangement of
pebbles in a pre-Cambrian conglomerate. Jour. GeoL, 38,
Serralog 0 2003
Chapter 4
I
195
p. 568-573.
PETTIJOHN, F.J. (1975). - Sedimentary Rocks. 3rd
ed. Harper & Row, Publishers, New York.
PETTIJOHN, F.J. & POTTER, P.E. (1964). -Atlas and
Glossary of Primary Sedimentary Structures. Springer,
New York.
PETTIJOHN, F.J., POTTER, P.E. & SIEVER, R.
(1972). -Sand and Sandstone. Springer, New York.
PEZARD, P.A. & LUTHI, S.M. (1988). - Borehole
Electrical Images in the Basement of the Cajon Pass
Scientific Drillhole, California; f facture Identification and
Tectonic Implications. Geophysical Research Letters 15
(August supplement, 1988), p. 1017-1020.
PILENKO, T. (1986). - Formation Microscanner
Applications in Italy. Milano, Italy. Schlumberger Italiana,
S.p.A., 1988.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Houston.
PLUMB, R.A. & LUTHI, S.M. (1986). - Application of
Borehole Images to Geologic Modeling of an Eolian
Reservoir. 61st SPE Annual Technical Conference and
Exhibition, New Orleans, paper SPE 15487.
POTTER, P.E. & PETTIJOHN, F.J. (1977). Paleocurrents and Basin Analysis. 2nd ed. Springer, New
York.
PRESS, F. & SIEVER, R. (1982). - Earth. 3rd ed. W H.
Freeman & Co, San Francisco.
READING, H.G. (Ed.) (1978). - Sedimentary
Environments and Facies. Blackwell Scientific
Publications, Oxford.
REINECK, H.E. & SINGH, I.B. (1975, 1980). Depositional Sedimentary Environments. 1st and 2nd ed.
Springer-Verlag, New York.
REINECK, H.E. & WUNDERLICH, F. (1968). Classification and Origin of Flaser and Lenticular
Bedding. Sedimentology, 11, 1-2, p. 99-104.
RUSSELL, W.L. (1951). - Principles of Petroleum
Geology. McGraw-Hill Book Co., New York.
SCHEIDEGGER, A.E. & POTTER, P.E. (1971). Downcurrent decline of grain size and thickness of single
turbidite beds : a semi-quantitative analysis.
Sedimeniology, 17, p. 41-49.
Schlumberger Ltd (1970). - fundamentals of Dipmeter
Interpretation.
Schlumberger Ltd (1974). - Log Interpretation. Volume
I1 - Applications.
Schlumberger ( I 979). - Well Evaluation Conference.
Algeria.
Schlumberger Ltd (1981). - Dipmeter Interpretation.
Volume 1 - fundamentals.
Schlumberger Technical Services, Inc. (1986). Carbonate Close up. Middle East Well Evaluation Review
1, p. 22-35.
Schlumberger Technical Services, Inc. (1987). Surprising Sandstones. Middle East Well Evaluation
Review, 2, p. 6-19.
Schlumberger Technical Services, Inc. (1988). Sandstones - The Inside Story. Middle East Well
Evaluation Review, 5 p. 32-47.
195
Schlumberger Technical Services, Inc. (1989). - Peak
Viewing for Thin Beds. Middle East Well Evaluation
Review, 6 p. 46-55.
Schlumberger Technical Services, Inc. (1982). Essentials of Natural Gamma ray Spectrometry
Interpretation.
Schlumberger Technical Services, Inc. (1983). - Well
Evaluation Conference. India.
Schlumberger WTA Marketing Services (1992). - FMl*
Fullbore Formation Microlmager. SMP 9210.
SCHOLLE, P.A. & SPEARING, D. (Ed.) (1982). Sandstone Depositional Environments. Amer. Assoc.
Petroleum GeoL, Mem. 31.
SCHWARZACHER, W. (1953). - Cross-bedding and
grain size in the Lower Cretaceous sands of East Anglia.
Geol. Mag., 90, p. 322-330.
SCOTT, K.M. ( I 966). - Sedimentology and dispersal
pattern of a Cretaceous flysch sequence, Patagonian
Andes, southern Chile. Bull. Amer. Assoc. Petroleum
GeoL, 50, p. 72-107.
SEILACHER, A. (1964). - Biogenic Sedimentary
Structures. In: lmbrie & Newell (eds): Approaches to
Paleoecology. John Wiley & Sons, New York.
SELLEY, R.C. (1970, 1978, 1985). - Ancient
Sedimentary Environments. Is t, 2nd and 3rd ed.
Chapman & Hall, London.
SELLEY, R.C. (1976). - An lntroduction to
Sedimentology.Academic Press, London.
SERRA, 0. (1985, 1989). - Sedimentary Environments
from Wireline Logs. 1st & 2d ed. Schlumberger, M081030, SMP-7008.
SERRA, 0. (1989). - Formation Microscanner lmage
Interpretation. Schlumberger Educational Services, SMP7028.
SERRA, 0. & ABBOTT, H. (1980). - The Contribution
of Logging data to Sedimentology and Stratigraphy. 55th
Ann. Fall Techn. conf. SPE of AIME, paper SPE 9270, and
in SPE J., Feb. 1982.
SIMONS, D.B., RICHARDSON, E.V., & NORDIN, C.F.
(1965). - Sedimentary structures generated by flow in
alluvial channels. In: Middleton (ed): Primary Sedimentary
Structures and their Hydrodynamic interpretation. Spec.
Publs. S.E.P.M., 12.
SPWLA Reprint Volume (1990). - Boehole Imaging.
STEINMETZ, R. (1967). - Depositional history, primary sedimentary structures, cross-bed dips, and grain size
of an Arkansas river point bar at Wekiwa, Oklahoma. Rep.
F67-G-3 (In: REINECK & SINGH, 1975).
THEYS, P., LUTHI, S. & SERRA, 0. (1983). - Use of
dipmeter in Carbonates for detailed sedimentology and
reservoir engineering studies.
TROUILLER, J.-C., DELHOMME, J.-P., CARLIN, S. &
ANXIONNAZ, H. (1989). - Thin-Bed Reservoir Analysis
from Borehole Electrical Images. 64th SPE Ann. Techn.
Conf. & Exhib., paper SPE 19578.
VINCENT, P., GARTNER, J. & ATTALI, G. (1979). GEODlP - An approach to detailed dip determination
using correlation by pattern recognition. J. Petroleum
Technol., Feb. 1979, p. 232-240.
196
van WAGONER, J.C., MITCHUM, R.M., CAMPION,
K.M., & RAHMANIAN, V.D. (1990). - Siliciclastic
Sequence Stratigraphy in Well logs, Cores, and Outcrops.
A.A.P.G. Methods in Exploration Series, no 7.
WALKER, R.G. (Ed.) (1979, 1984). - Facies Models.
1st and 2nd ed. Geoscience Canada, reprint series 1,
published by Geol. Assoc. Canada.
ZEMANEK, J. GLEN, E.E., NORTON, L.J. & CALDWELL, R.L. (1970). - Formation inspection with the borehole televiewer. Geophysics, 35, p. 252-269.
Serralog Q 2003
WELL LOGGING, FACIES, SEQUENCE
AND ENVIRONMENT
Introduction
Review of general geological concepts
Definitions
Since its introduction by Gressly (1838), the term
facies has been used in many different ways and these
uses have been the centre of considerable debate.
Usages and definitions have been reviewed by Moore
(1949), Weller (1958), Teichert (1958), Krumbein & Sloss
(1963), and more recently by Selley (1970), Reading
(1978) Middleton (1978) and Walker (1984).
Without reopening the debate, the general, almost
identical, definitions proposed by the Glossary of Geology
(Bates & Jackson, 1980) and by several geologists are listed here below.
- "The aspect, appearance, and characteristics of a
rock unit, usually reflecting the conditions of its origin;
esp. as differentiating the unit from adjacent or associated
units" (Bates & Jackson, 1980).
- Haug (1907) : "the sum of the lithologic and palaeontologic characteristics of a [sedimentary] deposit at a
given place".
- Moore (1949) : "any areal/y restricted part of a designated stratigraphic unit which exhibits characters significantly different from those of other parts of the unit'.
- Selley (1970) : "a mass of sedimentary rock which
can be defined and distinguished from others by its geometry, lithology, sedimentary structures, palaeocurrent
pattern, and fossils".
It is obvious from such definitions that a facies has a
necessarily limited extension, both stratigraphic and geographic, even if it can be found at different levels in the
same stratigraphic unit.
Geologists have noticed that a facies observed in a
stratigraphic unit can show similar features and characteristics to those described in other units of different ages or
coming from other regions of the earth. This is related to
the fact that such facies, with the same aspects, were
deposited under identical physico-chemical conditions.
Consequently, the term facies can be used with an
extended meaning for designating sedimentary rocks with
the same aspect, fauna and flora, even if they are different in age, reflecting similar physico-chemical conditions
Serralog Q 2003
in their environment of deposition.
In the following, the term facies will cover the more
general meaning given above. But, it will always be descriptive, without any genetic or environmental connotation. It will correspond to the general aspect of a sedimentary rock as it results from the sum of lithological,
structural and organic characteristics which can be detected in the field, and which distinguish this rock from other
surrounding rocks.
These present characteristics, on one hand, are the
results of the physical, chemical and biological conditions
under which the sediment was deposited, and, on the
other hand, are derived from its evolution under diagenetic influences since the time of its deposition.
It is from its characteristics and the context in which it
is found - vertical and lateral sequential evolutions, timespace relationship with neighbouring facies, regional tectonic control in the period of deposition - that one will be
able to determine its origin, its depositional environment
and its geological history.
Selley (1970) states "that a facies has five defining
parameters, viz. geometry, lithology, palaeontology, sedimentary structures and palaeocurrent pattern".
Generally, a facies is surrounded by other facies which
are related to it. This means that in a given environment
the facies are not randomly distributed, but constitute a
predictable association or sequence.
As pointed out by Selley (1970) "there are a finite number of sedimentary facies which reoccur in time and space
in the geological record' as "there are on the earth's surface today a finite number of sedimentary environments".
As pointed out by Middleton (1978) "it is understood
that (facies) will ultimately be given an environmental
interpretation". "However, many, if not most, facies defined in the field have ambiguous interpretation... The key
to interpretation is to analyse all of the facies communally in context. The sequence in which they occur thus
contributes as much information as the facies themselves" (Walker, 1984).
As also stated by Selley (1976) the idea of facies analysis "can be usefully extended to consider not just a vertical sequence, but a whole body of rock; what has been
termed a genetic increment of strata (Busch, 1971)".
197
"A genetic increment of strata (Fig. 5-1) is a mass of
sedimentary rock in which the facies or subfacies are
genetically related to one anothef (Selley, 1976). An
example of genetic increment would be one Bouma's
sequence.
"A genetic sequence of strata includes more than
one increment of the same genetic type" (Selley, 1976).
An example of genetic sequence would be several successive Bouma's sequences.
The general meaning of a sequence is a "succession
of geologic events, processes, or rocks, arranged in chronologic order to show their relative position and age with
respect to geologic history as a whole" (Bates and
Jackson, 1980).
Nomarine coal suMacfes
(swamp subenvironment)
Carbonaceous
sand fecies
(deltaic environment)
Carbonaceous sand subfacies
(delta front subenvironment)
Marine shale subfacies
(otrshmsubenvironment)
Figure 5-1 - Illustration of genetic increment and genetic sequence
(adapted from Selley, 1985).
Lombard (1956) has introduced the concept of lithological sequence that he defines as "a series of two lithological units, at least, forming a natural succession, without
any other important break except for the joints of stratification. The thickness of the bed is not considered'. He
distinguishes three orders of sequence:
- thin microscopic sequences (i.e. varves);
- medium macroscopic sequences (i.e. cyclothem);
- large megascopic sequences (i.e. stage, system).
Other concepts must be added. A granulometric
sequence corresponds to a grain size evolution without
change in mineralogy (i.e. coarse, medium, fine, very fine
sands). It can be fining upward, or coarsening upward.
A facies-sequence corresponds to a series of facies
which gradually merge into each other. The sequence
may be bounded at top and bottom by a sharp or erosive
junction, or by a hiatus in deposition. An example is the
Bouma's sequence.
Following the order of succession of the facies A, B
and C, or terms of the sequence, we have:
- a rhythm (Fig. 5-2), which corresponds to ABC, ABC,
AB, ...; such succession characterizes a rhythmic sedimentation and the results are rhythmites (e.9. cyclothems,
turbidites, varves);
198
- a cycle (Fig. 5-2), which corresponds to the succession of two sequences with opposite evolution ABCBA;
such succession characterizes a cyclic sedimentation.
The first type of succession is more frequent than the
second.
A lateral evolution or association of related facies
deposited at the same time, in different places in the
same environment but forming a continuum, creates a
lateral sequence; a succession of superposed terms in
relation to time corresponds to a vertical sequence (Fig.
5-3).
A
Figure 5-2 - Illustration of rhythm and cycle. In X, the three genetic
, and 111) show symmetric cyclic motifs: DCD, DCD,
increments (II1
DCD. In Z,the same three genetic increments are superposed with
asymmetric rhythmic motifs: AB, AB, AB (from Selley, 1976).
But, as new terms have been introduced by Vail,
Mitchum, Posamentier and others, (most of the time
without taking too much into account previous acceptations, definitions and works made essentially by european
geologists but also by north american geologists), it
seems important to clarify the sequence analysis concept
by going back to the basics and review the definition of
the terms now used in geology and sequence stratigraphy
(cf. Table 5-1).
Seauence
- "An unconformity-bounded stratal unit' (Sloss, 1948).
- "A relatively conformable succession of genetically
related beds or bedsets bounded at its top and base by
unconformities and their correlative conformities"
(Mitchum, 1977; Vail et al., 1977). "It is composed of a
succession of systems tracts and is interpreted to be
deposited between eustatic-fall inflection points" (ibid.).
Svstem tract
"A linkage of contemporaneous depositional systems "
(Brown and Fisher, 1977).
Paraseauence set (cf. Fig. 5-3)
"A succession of genetically related parasequences
forming a distinctive stacking pattern and commonly
bounded by major marine-flooding surfaces and their cor
relative surfaces" (Van Wagoner et al., 1990).
Serralog 0 2003
I
I
Facies, Sequence and environment Chapitre 5 199
Table 5-1
Stratal unit hierarchy
(modified from Van Wagoner et al., 1990).
Stratal Units
Thickness range
Area range
Time range
(feet)
(Sq. miles)
(Yeam)
Paraseauence
"A relatively conformable succession of beds or bedsets bounded by marine-flooding surfaces and their correlative surfaces" (Van Wagoner, 1985).
Bed set
"A relatively conformable succession of genetically
related beds bounded by surfaces (called bed set surfaces) of erosion, non-deposition, or their correlative conformities" (Van Wagoner et a/., 1990).
For Reineck and Singh (1975) "a simple bed set
consists of two or more superimposed beds characterized
by similar composition, texture and internal structure... A
composite bed set denotes a group of beds and bedsets
differing in composition, texture, or internal structure, but
associated genetically, representing a common type of
depositional sequence".
Tool Resolution
Use
surfaces) of erosion, non-deposition or their correlative
conformities" (Van Wagoner et al., 1990).
Lamina
"The smallest megascopic layet" (Van Wagoner et al.,
1990). For Payne (1942), it corresponds to a stratum less
than 1 cm in thickness. For Campbell (1967), the thickness of a lamina is lower than 3 cm.
But, as previously seen, sequence is also defined in a
more general meaning.
In accordance with this more general definition one
can recognize and adopt the other stratal units introduced
by Busch (1971) and Selley (1976) which are the genetic
increment and the genetic sequence or elementary
sequence.
Other types of sequences must be introduced as well.
Litholoaic sequence or Iithoseauence
"A relatively conformable succession of genetically
related laminae or lamina-sets bounded by surfaces (called bedding surfaces) of erosion, non-deposition, or their
correlative conformities" (Van Wagoner et al., 1990). From
the Glossary of Geology, 1980, it is "the smallest formal
unit in the hierarchy of lithostratigraphic units".
Lamina set
"A succession of, at least, two lithologic units (genetically related) forming a continuous succession without
interruption other than the stratification surfaces"
(Lombard, 1956).
Facies tract ("faciesbezirk" de Walther)
"Conformable vertical sequence of genetically related
facies" (Walther, 1893, 1894) generated by a lateral
sequence of environments (in Selley, 1970).
"A relatively conformable succession of genetically
related laminae bounded by surfaces (called lamina set
Serralog 0 2003
199
Facies-seauence
I
I
“A series of facies which pass gradually from one into
the other. The sequence may be bounded at top and bottom by a sharp or erosive junction, or by a hiatus in deposition indicated by a rootlet bed, reworking or early diagenesis. A sequence may occur only once, or it may be
repeated (i.e. cyclic)“ (Reading, 1978).
Virtual series
“A succession of lithofacies which are superposed
without major discontinuities“ (Lombard, 1956). It would
correspond to the logical succession of lithofacies if the
sedimentary process could evolve as suggested by the
previous deposits and was not interrupted due to external
causes.
Textural seauence
A succession of, at least, two types of texture occurring
in a same lithologic unit (i.e.: bounstone, grainstone,
packstone, wackestone and mudstone textures in limestone; grain size evolution: coarsening-up or fining-up
sequences in sandstones; change in sorting in a sand
body, etc).
Seauence of structures
A succession of, at least, two types of sedimentary features occurring in a same lithologic unit (i.e.: massive,
laminated, cross-bedded, and laminated sands composing units A, B, C and D of the Bouma’s sequence).
Seauence of thickness
A succession of beds of same lithology, with decreasing or increasing thickness upward (i.e.: thickening-up or
thinning-up sequences observed in turbidity deposits)
(Mutti & Ghibaudo, 1972; Ricci-Lucchi, 1975 in Walker,
1984).
J
Figure 5-3 - Progressive development of genetic increment and genetic
sequence with parasequence boundaries. Stage one corresponds to a
progradation of parasequence A during a time when the rate of deposition exceeds the rate of water depth increase. Stage two corresponds
to a rapid water depth increase flooding the top of the parasequence A
creating a surface of non deposition with respect to siliciclastic sediment. Thin carbonates, glauconites, organic-rich marls, or volcanic
ashes may be depoited on the surface. Stage three corresponds to a
progradation of the parasequence B during a time when the rate of
deposition exceeds the rate of water depth increase. Bedsets in parasequence B downlap onto the boundary of parasequence A. There is
an abrupt deepening of facies across the boundary of parasequence A
(from Van Wagoneret al., 1990).
Walther introduced the term facies tract (faciesbezirk)
for a conformable vertical sequence of genetically related
facies and observed that the various deposits of the
same facies area and similarly, the sum of rocks of different facies areas, were formed beside each other in
space, but in a crustal profile we see them lying on top of
each other... it is a basic statement of far-reaching significance that only those facies and facies areas can be
superimposed, primarily, that can be observed beside
each other at the present time” (translation from Blatt et
a/., 1972). This principle is at the origin of the Walther’s
law which can be stated as “a conformable vertical
sequence of facies was generated by a lateral sequence
of environments” (Selley, 1976) (Fig. 5-4).
“
Environments A
Importance of facies and sequence analysis
The study of facies and their arrangement or association in lateral and vertical sequences is the only way to
establish the depositional environment, and thus to
reconstruct the palaeogeography at the time of deposition. The physical, chemical and biological conditions
existing in an environment, which define it, can only be
determined by its imprints on the sediments. Among these
imprints, the primary sedimentary features are more
important because they have been formed in-situ. The
sequences will reflect the modifications in the conditions
both in space and in time.
Figure 5-4 - Cross-section illustrating the Walther’s law. Environments
A, B and C deposited during the same time period the lateral sequence
of facies a’, b’and c‘respectively in the increment of sedimentation
designated I. The increments I, I1 and 111 together constitute a genetic
sequence of strata. The same facies a’, b’ and c’ are superposed in
position X (adapted from Selley, 1976).
As pointed out by Walker (1984) “careful application of
the Walther’s law, therefore, suggests that in a vertical
200
Serralog c.c3 2003
I
I
Facies, Sequence and Environment Chapter 5 201
sequence, a gradational transition from one facies to another implies that the two facies represent environments
that once were adjacent laterally. The dangers of applying
the law in a gross way to stratigraphic sequences with
cyclic repetitions of facies have been emphasized by
Middleton (1973)".
De Raaf et al; (1965) and Reading (1978) indicated the
importance to clearly define gradational facies boundaries
in vertical section as opposed to sharp or erosive boundaries. In case of sharp or erosional boundaries between
two facies there is no way of knowing whether the two vertically adjacent facies represent environments that once
were laterally adjacent. Sharp breaks between two different facies may signify fundamental changes in depositional environment and the beginnings of new cycles of sedimentation.
Based on these remarks, as one can imagine, it is very
important to determine the type of boundaries between
facies.
Walker (1976) made a comparison between the results
of facies sequence analysis and facies models (established on the basis of modern environments). This comparison allows, by analogy, the determination of the depositional environment by application of uniformitarianism or
actualism, theories introduced by Hutton (1785) and Lyell
(1830): "the present is the key to the past'.
Van Wagoner et a/. (1990) indicate the necessity of
facies determination and facies succession for the determination of the depositional environment and for sequence stratigraphy applications (Fig. 5-5).
Traditional approach of facies analysis
The determination of the facies, sequence and environment is based on the observation and description of
the traditional geological objects: outcrops and cores. The
results and synthesis of this study is perfectly illustrated
by Van Wagoner et al. (1990) as previously seen (cf. Fig.
5-5).
Many books have been written on the facies models
(Walker, ed., 1984) and sedimentary environments in
which the traditional approach for determination of facies,
sequences and environments is abundantly described
and illustrated (Krumbein and Sloss, 1951, 1963; Selley,
1970, 1978, 1985 and 1976; Schwarzacher, 1975;
Reineck and Singh, 1975; Reading, 1978; Friedman and
Sanders, 1978; Scholle and Spearing, ed. 1982; Scholle
et al., ed. 1983; Galloway and Hobday, 1983; Davis, ed.
1978, 1985; Selley, 1988...). Please refer to these books,
if needed!
Well logging approach
As demonstrated in the previous chapters, practically
all the facies attributes (Table 5-2) can be extracted from
well logging data, as soon as a certain methodology is
applied.
Serralog 0 2003
Table 5-2
Sedimentary rock characteristics. Comparison of the
geological approach and the well logging approach.
Geologicalapproach
Fades
Well logging approach
Electrofacies
I
Definition
I
Definition
Set of geobgiial attributes that charactefees ]Set of w ell lcg attributesthat characterizesan
a bed. abws its differentiationfromthe
electrobed and abws its differenfiationfromthe
surrounding beds and reflects the conditions surrounding electrobeds
(physical, chenicaland biological and so
clirretic) of its orun
Attributes
WOlcgY
- Conposition
- ninerakgical
Attributes
BecboWhofacies
Wo-density. hydrogenindex, natural and induced
radioacWty, sound sbw ness. resistivity data...
- elemental
- Texture
Natural and inducedradicWty ...
(grain size. sorting. pore size. pore type and
distribution. p cking... )
S.P.. reskW!. porosw took. borehole images.
dipmeter curve data and nuclear magnetic
resonance
hternalsbuctures: energy, transport current
direction, environmnt
Borehole images. dipmter data
FossiS : age. environmnt
Up data, correlabns between w e b
-Width
I
-apparent
- real
I
I
Standard bgs
Upmeter data, borehole m
e
s
It appears clearly that practically all the parameters or
attributes which determine a facies can be extracted from
well logs, if we except color and fossils. Consequently, it
seems quite reasonable to admit that a valuable facies
recognition can therefore be achieved with log data. We
call this facies determination electrofacies to indicate its
origin, and its definition is : " the set of well log responses
which allows differentiation of an electrobed from its surrounding electrobeds".
It is obvious that :
- a complete set of open hole logs, possibly calibrated
on core data, allows the determination of the lithology, and
even the principal minerals composing the rocks, with a
good accuracy (cf. Chapter 2);
-as previously explained and illustrated, electrical images provide information about texture, sedimentary features, the nature of surfaces, organic activities and thickness, as well as diagenetic activities. In addition, they
identify significant geological events or markers with limited vertical extent such as hardgrounds, concretions,
vugs, roots, burrows, clast lag, debris, breccias and
conglomerates (cf. Chapter 3 and 4).
Consequently, most of the attributes that characterize
a facies can be extracted from logs and, so, can be used
to define it. In addition, interpretation of the image data by
a program similar to BorTex provides a useful description
201
GRAIN
SIZE
TROUGH CROSS B€DS
TROUGH CROSS BEDS - SMALL SCALE
SlGMOlOAL CROSS BEDS
CURRENT - R
W SEEMW(0
CONTORTEO BEDS
PLANAR BEDS
HUMMOCLV BEDS
WAVE
wrnr~omffi
c u v CUSTS
AWNOANT
COMMON
RARE
ABSfNT
-
HST = HIGHSTAND SVSTEMS TRACT
TST TRANSGRESSIVE SVSTEMS TRACT
LST - LOWSTANO SYSTEMS TRACT
CHURNED BV BURROWS
Figure 5-5 - Typical type of formation description which includes grain-size data, facies characteristics, depositional environment and sequence
stratigraphy. All these steps can be achieved from well logging data, including obviously borehole image data, with a high vertical resolution
(from Van Wagoner, et al., 1990).
202
Serralog 0 2003
Facies. Seauence and Environment khaoter 51 203
of the internal organization and the thickness of each unit
of deposition composing the formation. This information
completes the lithology determination using the standard
logs.
However, as the lithology and sometimes the texture
observed in the present day result from two factors:
- the original dynamic and biological conditions
prevailing at the time of deposition
- the subsequent diagenetic effects on the sediment since its deposition,
the original facies cannot be accurately determined by
relying only on lithology and textural parameters from
porosity levels or SP/gamma ray vertical evolution. When
sedimentary features are identified in images, however,
the original facies can be determined because these features reflect the dynamic and biological conditions prevailing during the deposition. In the same way, conglomerates, roots, hard-ground or burrows signify certain conditions which cannot be extracted from the standard logs.
All this information is critical for accurate facies and
sequence recognition.
Most of the time diagenetic effects do not significantly
modify the internal structure or sedimentary features,
except in some carbonates. Many of the common diagenetic effects can be identified on images (cf. Chapter 6).
For instance, cemented nodules or layers, dissolution features, and stylolites are easily recognized. Diagenetic
sequences can also be determined from log data (cf.
Chapter 6). In any case, these diagenetic features characterize diagenetic environments which are usually linked to the depositional environments.
From the previous remarks it seems evident that
image data should be recorded at least once in a field.
Whenever possible, they should be calibrated on core
description and analysis. Finally, they should be run in
other wells or replaced by dipmeter data if typical curve
responses are demonstrated to reflect features seen on
the images.
As soon as the facies of each unit composing a formation is determined, the vertical facies succession, combined with the surface type (hardground, erosion, truncation) or features (roots, burrows) observed on images,
and completed by information on facies thickness, enables the determination of genetic increment units and
genetic sequences (see next topic).
The overall depositional environment can then be derived from these data. In addition, breaks of higher order
and wider extent can be detected, which is important for
sequence-stratigraphy analysis.
zing the SP curve, the type of contact (abrupt or progressive) between sands and shales, and the character of the
curve (smooth or serrated; concave, rectilinear or convex)
one can establish the classifications shown in Fig. 5-6.
UPPER BED CONTACT
Figure 5-6 - Classification of SP curve shapes as a function of the
curve vertical evolution (adapted from SHELL-PECTEN).
Pirson (1970, 1977) associates a facies and a depositional environment to each shape, and he interprets the
curvature of the curve as an indicator of the speed of
transgression or regression processes (Fig. 5-7).
Fig. 5-7 - Classification of SP curve shapes in terms of sedimentary
patterns (courtesy of Pirson, 1970, and Gulf Publishing Co., fig. 2-1).
Several geologists (Visher, 1965; Serra & Sulpice,
1975; Coleman & Pryor, 1980; Galloway & Hobday, 1983)
have used log patterns for several purposes:
- vertical grain size evolution,
- determination of vertical sequence,
- recognition and mapping of log facies
- interpretation of depositional environment.
Historical
It seems that the idea of using wireline logs as sedimentological tools, first came in 1956-1957, from engineers working for the SHELL-PECTEN Company in
U.S.A. Studying the Mississippi delta, they stated that the
spontaneous potential curve (SP) presented characteristic shapes. Each of these shapes corresponds to a facies
or facies succession of a particular sand body. By analySerralog 8 2003
Figures 5-8 and 5-9 (see next pages) summarizes
some of the log patterns and their interpretation.
Several geologists have used this very rapid and synthetic method of log pattern analysis to construct facies
maps (see Fig. 5-10 next pages).
203
1
3
2
4
Figure 5-8 - Examples of interpretation of log patterns (SP curve), grain size and sedimentary structures.
I - Case of a low-sinuosity braided channel. Sequence A dominated by migration of a gravely longitudinal bar. Sequence B records deposition of
successive transverse bar cross-bed sets upon a braid channel fill.
2 - Case of a laterally accreting (A) and symmetrically-filling channel segments (6) of an anastomosed channel system.
3 - Case of a meanderbelt sand body produced by a high-sinuosity channel. Sequence A illustrates a complete fining-upward sequence typical of
the mid- or downstream point bar. Section B illustrates the truncated vertical sequence commonly found in the upstream end of the bar.
4 - Case of a chute-modified point bar. Sequence A corresponds to upstream portions of the point bar capped by chute-channel deposits.
Sequence B corresponds to downstream, the channel and lower point bar deposits are capped by chute-bar sediments
(adapted from Galloway & Hobday, 1983).
204
Senalog 'B 2003
Next Page
Facies, Sequence and Environment (Chapter 5! 205
1
3
2
4
Figure 5-9 - Examples of interpretation of log patterns (GRcurve), grain size and sedimentary structures.
1 - Case of an upward-coarsening parasequence formed in a beach environment on a sandy, wave- or fluvial-dominated shoreline.
2 - Case of a similar upward sequence but formed in a deltaic environment on a sandy, wave- or fluvial-dominated shoreline.
3 - Case of stacked upward-coarsening parasequences interpreted as formed in a beach environment on a sandy, wave- or fluvial-dominated shoreline where the rate of deposition equals the rate of accomodation.
4 - Case of two upward-fining parasequence interpreted as formed in a tidal flat to subtidal environment on a muddy, tide-dominated shoreline.
FS = Foreshore. USF = Upper shoreface. LSF = Lower shoreface. D.LSF = Distal lower shoreface. SH = Shelf. OSMB = Outer steam-mouth bar.
DF = Delta front. PRO D = Pro delta. CP = Coastal plain. SBT = Subtidal. INT = Intertidal. SRT = Supratidal.
(adapted from Van Wagoner et at., 1990)
Serralog Q 2003
205
Previous Page
206
Well logging and Geology
Knowing the fundamental reasons for the choice of the
SP or GR curve (other logs often being unavailable, resistivity curves being heavily affected by the presence of
hydrocarbons, large number of wells to be studied, etc.)
and its real possibilities in detrital sand-shale series (in a
majority of cases it reflects shaliness and grain size evolution), we have to acknowledge that the use of such
curve alone is often insufficient to clearly determine a
facies and a depositional environment. It may lead to a
misreading of the events because of "parasite effects" on
SP deflection (influence of R,, invasion, contrast of RR
/, ,
compact zones, thickness of beds, etc.). Finally, this curve
becomes unusable if the contrast of R,/R,
is insufficient.
The GR response may be affected by the presence of
feldspars, micas, zircon, uranium salts, etc., which may
be confused with clay content or grain size evolution. This
is why the shape of a single curve may only exceptionally define the facies and the depositional environment, particularly if we intend to use this method to study other
types of sediments such as carbonates and evaporates.
covering their chemical and mineralogical composition,
their texture and their structure. The higher the number of
well logs used, the richer the spectrum will be, and the
better defined the rock characteristics. Hence there will be
less risk of ambiguity and error in their interpretation.
Figure 5-11 - Facies determination from GR curve
(from Serra & Sulpice, 1975).
Moreover, dipmeter data processed by GEODIP program for HDT tool, or LOCDIP program for SHDT tool, or
image data allow, in many cases, the detection of sedimentary structures and the determination of the palaeocurrent pattern and the direction of transport.
Finally, the geometry is defined by both analysis of the
true thickness of the beds (only dipmeter or image data
provide this information), and by the lateral extent of beds.
This can only be defined by correlations between several
wells and by the drawing of isopach maps based on thickness data (Fig. 5-12 and 5-13).
G
H
Fig. 5-10 - Distribution of facies based on the shape of the resistivity
curve (from Lennon, 1976).
The electrofacies concept
As we have seen in previous chapters, every well logging measurement gives, more or less, some information
about the mineralogical composition, the texture, and the
sedimentary structures, even if this information is sometimes implicit. In other words, each well log gives a particular spectral picture of the rock properties.
In certain cases one or two spectral pictures, therefore one or two logs as we have previously noted, are sufficient for the determination of rock characteristics (for
instance, the use of the shape of the SP curve in the
sand-shale series of the Gulf Coast, or the use of GR
curve in Algeria : cf. Fig. 5-11). But, it is preferable to use
all the available log data for interpretation. Their number,
their diversity and their complementarities allow, in fact,
the establishment of a spectrum of rock characteristics
206
Figure 5-12 - Log-correlation between wells in the South Glenrock
Oilfield, Wyoming. Bar, beach and channel fill are recognized
(from Curry & C u q , 1972).
Most of the parameters defining the facies, or at least
the lithofacies in Moore's definition (1949), are directly
accessible from logs. The latter, therefore, produce a picture of the present facies. This picture is certainly particular, incomplete, sometimes confused, but always permanent and objective. If the set of well logging measurements is diversified and rich enough for a better covering
Serralog 0 2003
Facies, Sequence and Environment 'Chapter -5
il
of geological parameters, the picture will be sufficiently
precise. In other words the spectrum will be sufficiently
rich and detailed to permit a new representation of the
facies by means of log data.
J
Figure 5-13 - lsopach map of the lower Muddy in the South Glenrock
Oilfield, Wyoming, showing two buried streeam channels, based on
Figure 5- 12
(from Curry & Curry, 1972).
This parafacies was named by Serra (1970) electrofacies, the definition of which is:
"the set of log responses which characterizes an electrobed and permits it to be distinguished from the others"
(Serra, 1972; also in Schlumberger Well Evaluation
Conference, Algeria, 1979).
All log responses (electric, nuclear, acoustic, dipmeter,
images, etc.), that indicate the quantitative (log values
and dip data) as well as qualitative aspects (curve characteristics, textural and structural data) represent, therefore, the component elements of the electrofacies.
Electrofacies constitutes more than one element of a
facies. It is, in fact, its equivalent since it includes in itself
the parameters which define the facies.
But we have to realize that there is a parameter which
has never been taken into consideration by sedimentologists in their definitions of facies. This parameter is the
fluid that occupies the porous space of the rocks. If in surface outcrops it is neglected because it is absent or
without significance, it is always present in the subsurface and it influences the response of most logging tools.
Therefore, we cannot eliminate it and, in fact, it enters into
the definition of electrofacies. Consequently, several electrofacies, depending on the nature of fluids present in the
rocks (gas, oil, fresh or salty water), may correspond to
the same geological facies. This situation may be considered, at first sight, as an important disadvantage of the
electrofacies concept and of its utilization. In fact, it is not
important because the purpose of the electrofacies analysis is, first of all, to describe the formations as they are
"seen" by logging tools. After all, we can possibly utilize
logging tools, and their data, less sensitive to fluids or to
Serralog 0 2003
207
i
"
Y
l"lil(_)
porosity (i.e. natural or induced gamma ray spectrometry,
photoelectric index), or we can correct the log response
for porosity and fluid influence. For all that, in the absence of hydrodynamism, the fluid may be an important factor for the recognition of a depositional environment: fresh
water in fluvial or lacustrine environment, briny water in
swamp or marsh, salty water in sands deposited in marine environments...
The objection has often been raised that the electrofacies is only an equivalent of the lithofacies because it
does not contain palaeontological information. Without
being argumentative, it is necessary to make the followings remarks.
1) If the fossils are utilized as indicators of the depositional environment, we have to remember that in many
cases the fauna and flora are almost nonexistent, especially in sandstone deposits, and, consequently, the facies
recognition is realized without this information;
- the fauna and flora are not always good indicators of
the environment (ubiquitous species, mixtures, allochtone
species, etc.);
- in many cases the presence of macrofossils (animal
or vegetal) is shown by their influence on log responses,
particularly on dipmeters and images. Moreover, it is
controlled by the physico-chemical conditions existing at
the time of deposition, which also determine the other
parameters, especially the sedimentary structures;
- other elements of facies (mineralogy, texture, sedimentary features, palaeoccurents and geometry) are
often sufficient for a precise definition of the facies, and
also for specifying the depositional environment, especially if we use the additional information on sequential
evolution.
2) If the fossils are used to define the geological age,
we may note that this information is implicitly included in
the subsurface data through the depth data and the position in relation to markers. Besides, at this stage, the logs
allow a more precise appraisal of the lapse of time than
that defined by fossils.
Hence, we have good reasons to assimilate the electrofacies into the facies. Moreover, the reconstitution of
the time-space repartition of the different facies, and the
definition of their mutual relations is the final goal of facies
and sequence analysis, regardless of the methods used
to achieve this goal (traditional method by examination of
rocks, or from log data).
Depositional environments will therefore be defined by
these analyses, and thus we may forecast more accurately the continuity of a reservoir, the presence, nature and
distribution of permeability barriers, and the location of
mineral resources in exploitable economic quantities.
/J
Electrofacies analysis from well logs
The goal of this analysis is to describe objectively the
formations penetrated during drilling, through their well
logging responses, and to recognize all the different fundamental electrofacies present, in such a way that ultimately it will be possible to study their association in vertical
207
sequences, and, consequently, deduce the lateral evolution by applying Walther's law. In other words, the first
goal is to reconstruct the electrofacies models which will
help to define the depositional environments.
An electrofacies analysis can be carried out manually
or automatically. In both cases the basic approach is
essentially the same.
DEPTH
Track 1
Track 2
Manual identification of electrofacies
Originally a sedimentological study from logs involved
examining the shapes of various curves for indications of
the type of sedimentation and the depositional environment. As previously seen, a ramp or a gradient on a log
could indicate an upward-fining or coarsening of grains
(cf. Figs.5-6, 5-8 and 5-9). Classification of electrofacies
by the shape of spontaneous potential or GR response is
well known and has been used for many years. But, as
drawbacks of SP or GR curves may occur (Fig. 5-14 and
5-15), it is highly recommended to use all the available
data.
Figure 5-14 - Based on the GR curve evolution the interpretation
would be coarsening-up sequence followed by fining-up sequences.
DEPTH
Figure 5-15 - The complete set of log allows the recognition of dolostone invaded episodically by arkosic deposits as perfectly illustrated by the
potassium content. The shaly layers are indicated by peak of thorium with a value above 8 ppm.
Feldspathic dolostone
Dolostone
One starts from a contrived document, the compositelog. This document gathers together and depth matches
all the log data recorded in a well, including dipmeter data
and dip computation results obtained by GEODIP- Or
LOCDIP-processing types and borehole image data. It
can be obtained on a workstation (Fig. 5-16).
208
Arkose
One divides the studying interval into electrobeds or
electrosequences. For this purpose the amplitude of the
variations on macrodevice curves is analyzed. According
to its importance and its shape, it is subjectively decided
- either it corresponds to an electrobed boundary
- or it indicates a "noise" that is inherent either to the
Serralog Q 2003
Facies, Sequence and Environment IChapter 51 209
Figure 5-16 - Composite-log combining results of a quantitative interpretation of standard logs, borehole images and dip data with cross-plot of
density vs neutron measurements (courtesy of Schlumberger).
measurement (statistical variations of nuclear measurements), or to hole conditions (borehole wall rugosity, presence of caves, etc.), or to minor changes in the geological parameters;
- or it reflects a gradual evolution (ramp) with minor
variations.
In the following step one determines the electrolithofacies for each electrobed presenting a thickness greater
than the average vertical resolution of most macrodevices
(about 2 to 3 feet, or 60 to 90 cm).
As the synthesis of different measurements, corresponding to the same electrobed (especially when it
concerns a large amount of data) is not easy, a representation using rosettes, spider's webs (Fig. 5-17a), or histograms has been proposed to visualize the electrofacies
(Serra, in Schlumberger Well Evaluation Conference,
Algeria, 1979). As histograms cannot easily be obtained
manually, and a spider's web varies in shape according
the number of involved measurements (hence the branches), a ladder diagram presentation (Serra, in
Schlumberger Well Evaluation Conference,Algeria, 1979)
seems to be more usefull because the absence of one log
Serralog Q 2003
data will not modify the general shape of the figure (Fig.
5-17b). In these representations, each branch of the spider's web, or each bar of the ladder, represents a scaled
log axis with is range of variation. One plots on these
basic documents the minimal, maximal and median
values of each log on the interval corresponding to the
electrobed .
When the points on each axis are joined, a characteristic shape is formed. For each electrofacies, there will be
an allowed band on each axis; hence, an allowed area is
created corresponding to one electrofacies.
Figure 5-18 shows the shapes of two electrofacies.
Two shapes need to differ in only one axis to establish the
difference between two electrofacies. Essentially, the
comparison of shapes allows an analyst to break down a
logged interval into some 10 to 15 electrofacies (Fig. 519). This last figure shows the progressive change in shapes from one electrofacies to the next. The interval covered is shown in Figure 5-20. The shallowest depth is at the
top left corner, and figures are arranged in columns with
depth increasing downward. Correlation with facies defined from core analysis is also made.
209
Next Page
210
Well logging and Geology
A correlation between electrofacies from different wells
can also be achieved by this technique (Fig. 5-21).
Figure 5-1 7 - After logs have been zoned and electrofacies established, the log parameters are plotted on rosette or spider's web diagrams (left image), or on superposed bars of a ladder (right image).
(from Serra, in Schlumberger Well Evaluation Conference, Algeria,
1979).
MSFL
MSFL
Figure 5-18 - Spider's web diagrams for (a) a limestone and (6) a
sandstone (from S e r a & Abbott, 1980).
this is a long and very tedious operation. Therefore, the
following empirical method may be preferable. One draws
the electrofacies representation of the thin bed and indicate, for each measurement, by an arrow the direction
toward which the correction should displace the representative point. By comparison with preliminary defined
electrofacies, and by using dipmeter or image resistivity
curves, it is possible to estimate the closest electrofacies
(Fig. 5-22 & 5-23).
In the case of ramps or electrosequences, one defines
the electrofacies of the starting and stopping depths or of
the surrounding electrobeds.
This manual analysis is often long and sometimes
tedious. Certain traditional geologists may consider it as
uninteresting. In this connection we remind these geologists that this method is used by some oil companies in
the world, perhaps not in the above mentionned way, but
at least in spirit. For example this method allowed the
ELF-Aquitaine group to study more than 1000 wells and
to have, thus, a synthetic idea of the facies and environments for each well. They were able to draw facies maps.
It was done at low cost and in a very short time.
Otherwise, it would have been impossible to obtain this
information, because cores were rarely cut and the core
analyses were often unavailable (exchange wells). This
method enabled a more accurate covering of the seismic
profiles and a more reliable interpretation in terms of seismofacies. This gave the best possible knowledge of the
basins and consequently a more judicious and justified
choice of prospects.
J
-
Figure 5-19 Comparison of 20 electrofacies (from Serra & Abbott, 1980).
Determination of the electrofacies in thin beds requires a
preliminary step : the correction of the different measurements for the influence of the surrounding beds. For this
purpose environmental
correction charts are used. But,
210
But this analysis is subjective, because the results
may partly depend on the analyst performing the study.
For this
specialists dreamed for an automatic process using computers. The computer processed method
Serralog Q 2003
Previous Page
~~~~~
Facies, Sequence and Environment [ C h a p t a 211
(Serra & Abbott, 1980) described hereafter, was developed by Schlumberger and commercialized under the mark
of FACIOLOG. Its description is destined to explain the
different steps of data processing. Other approaches
could be imagined, and will doubles be developed. They
will certainly be inspired by the same general philosophy.
Automatic Electrofacies Identification : FACIOLOG
Essentially the same steps are taken as when applying
GtOMP
CURVES
Figure 5-20 - Correlation of electrofacies wifh core facies (from Serra 8, Abboff, 1980).
Well B Fades 2
Well A Facies 2
SHALE
Figure 5-21 - Correlation of electrofacies (from Sera & Abboff, 1980).
Serralog 0 2003
the manual method. We simply try to translate the
approach of the analyst into mathematical functions or
statistical processing.
The set of n log responses that characterizes an electrobed, or level of reading, may be considered as defining
the coordinates of electrofacies (here represented by a
point) in a n-dimensional space. Since the same causes
produce the same effects, we may think that another bed
with same geological facies and containing the same
fluids will have the same electrofacies. Consequently, its
representative point in n-dimensional space has to be
very close to the previous point. Hence, an electrofacies
must correspond to a cluster, or to a cloud of points very
close to each another in this space. Contrarily, the distinct
electrofacies must correspond to different and separated
clusters. They can possibly overlap in one or several
dimensions of the n-space.
If we start with "raw", unzoned data, we will observe a
certain dispersion of points corresponding to one electrofacies, and it will be more dispersed with a higher number
of log measurements. This dispersion is related to the
"noise" due to the tool, to hole conditions and even to
weak variations of the geological parameters. A preliminary zoning and an analysis of the principal components
211
THEK
LIGNITIC
BED
Figure 5-22 - Composite-log associating dipmeter (HDT) data.
GR Pb -
'h4b-
At -
L,
RL c
Figure 5-23 - (a) :Electrofacies of the coal bed at 1241-1243 m. (b) :electrofacies of the thin bed at 1238-1238.5 m which corresppnds to a coal as
indicated by dipmeter resistivity curves.
will, on the one hand, decrease this dispersion, and on the
other hand, reduce the dimensionality of the space (cf.
Table 5-3). Only after this step, can an automatic clustering be carried out.
We can, in certain cases (the absence of ramps and of
thin beds), immediately start the analysis of the principal
components and carry out a clustering. The results of the
processing leads to an automatic zoning (Fig. 5-24).
The use of automatic zoning concentrates the cluster
by attempting to eliminate measurement errors. If an ndimensional histogram is created and the frequency of
each cell analyzed, we obtain results as in Figure 5-25a.
The results on the same interval after zoning are shown in
212
Figure 5-25b. Obviously, the data distribution shown in
Figure 5-25b is easier to handle. Figures 5-26a and 5-26b
show the corresponding frequency plots on two selected
axes of the interval.
Automatic seamentation of loas
The clustering technique does not allow the detection
of natural ramps (or electrosequences) on curves. It is the
reason why another approach has been developped for
an automatic segmentation of log data into electrobeds
and electrosequences (ramps).
Serralog 0 2003
Facies, Sequence and Environment
a
b
Figure 5-25 - Frequency plots of the log data corresponding to Table
5-3. (a) :unzoned cross-plot; (b) :zoned cross-plot of density vs neutron (from Serra & Abbott, 1980).
Table 5-3
Number of cells with their number of elements.
Variable
I
Mnirmm
11
Maximum
I
Steps
I
Figure 5-24 - Example of automatic zoning of logs obtained by clustering techniques and comparison with raw logs (from Serra, in
Schlumberger Well Evaluation Conference, India, 1983).
An electrobed can be defined as a succession of levels
with contiguous reading, according to the sampling rate
used (6 or 1.2 inches). These levels present essentially
the same values. This means that their response variations do not exceed certain limits (allowed variations).
These variations correspond to the tool error in the measurement, and to minor acceptable changes in geological
parameters (Fig. 5-26).
These changes may be expressed in terms of measurement error or may follow a more complex law.
An electrosequence can be defined as a succession of
contiguous readings. The level having an order n, shows
a value higher than that of level with order n - 1, but lower
than that of level with order n + 1, or vice-versa. This evolution must continue in a depth interval thicker than the
vertical resolution of the measuring tool, to be considered
as representative of a natural electrosequence.
Otherwise, it corresponds to an artificial ramp, due to the
lack of resolution of measuring tool (Fig. 5-26).
Serralog Q 2003
213
Allowed variations on the measurement
including statistics and mlnor geoiogical changes
This symbol wmsponds to a sample
(measurement )
A thin electrobed Inside a thicker one
tConsaquently,wrrectlons must be applied
to roach the roal value of this thin electrob4
-Mean value of the bed
:samp#ng rate:
oach half a foot In this case.
or each Inch In case of high resolution recording
ty of points). The task of the clustering program is then to
distinguish each cloud. For this purpose, we determine,
for each level or reading depth, the distance dk of the k'
closest neighbour. It corresponds to the radius of the
smallest circle with its center at a known level and containing k neighbouring points. If any other level, situated
below or above a certain depth window (BAND) does not
have a lower dk value, and if among these neighbouring
k none of them has a lower dk value, a level will be taken
as a cell, or as a local mode (Fig. 5-29). Each cell obtained in this manner corresponds to an elementary electrofacies.
Natural ramp'
or
electrosequence
0.G
c
Mean value of the bed.
A thldc electrobed
Real thickness
of the middle
electrobed
Artifk4ai ramp
A thick electrobed
Real thickness
of the lower
electrobed
~Verticairemlutlon of the measurement
,expressedin numbor of samples
Allowed variationsbn the measurement
nciuding statistics and minor geological changes
Loo i
Figure 5-26 - Principle of the automatic segmentation of logs into electrobeds and electrosequences.
Hence, it is necessary to define for each tool, the vertical resolution and the amplitude of the accepted variations, expressed as a percentage. We can state that the
effect is minor if we do not exceed a certain value.
Generally, we carry out the zoning from active logs, on
which we rely to determine the limits and the type of evolution. We impose these limits to other so-called passive
logs. The type of evolution between these limits defines,
for passive logs, either a bed or a sequence. Figure 5-27
(next page) gives an example of automatic log segmentation obtained with the help of three active logs gamma ray,
neutron and density.
Clusterina techniaues
The method described in the previous section is
essentially a way of trying to divide n-dimensional log
space (n corresponds to the number of different log data
taken into account) into definable volumes corresponding
to each electrofacies (Fig. 5-28).
Many of the mathematical techniques known as clustering are adaptable to such problems. In more than two
dimensions, we can think of the electrofacies points falling
into clusters or clouds. (A cluster may be described as a
continuous region of the n-dimensional space containing
a relatively high density of points, separated from other
such regions by regions containing a relatively low densi214
Figure 5-28 - Three-dimensional case of three electrofacies on a threelog axis (from Serra & Abbott, 1980).
The shape of the clusters is difficult to define, as one
cannot assume, automatically, that the distribution within
each facies, for each log, is normal. A cloud is dispersed
in any direction by the effects of log errors, as described,
or by changes in the facies itself. Consider, for example,
the grain size variation within a sandstone. This will
disperse the sandstone points, leading to a "cloud" with a
concentration of points at one end and a tail. In the case
of a gradual change from one facies to another, the actuation is even more complicated since the clouds are not
separated. In most cases at facies interfaces the differing
vertical resolution of the logs introduces some scatter in
the response. For these reasons statistical techniques are
needed to establish criteria to separate clusters.
In deciding what kind of clustering technique to use,
several things are important
- the number of levels to be examined at any time;
- the number of variables (logs) to be considered;
- the size and distribution of each cluster;
- the occurrence of electrofacies with only limited
representation;
- the effects of gradational changes from one cluster to
another.
It is desirable to introduce information provided by a
dipmeter or by borehole image to really determine the
Serralog 0 2003
Facies, Sequence and Environment IChapter 51 215
Figure 5-27 - Example of log-segmentation results (lower figure) compared to the composite-log of raw data before segmentation (upper figure).
Serralog (9 2003
215
electrofacies. In that case this information must be quantified. This is achieved, for dipmeter, through the dipmeter
derived synthetic logs previously described and obtained,
for instance, from the SYNDIP program for dipmeter and
through BorTex program results for image data.
For data where the significance of changes over the
range of a log is heavily dependent on the log value, some
nonlinear-scale changes are necessary. For example,
resistivity values that vary from 0.1 to 10,000 are transformed to logarithm base 10. This makes the processing
easier and allows the use of a metric standard.
points for a partitioning algorithm, working not on the histogram but on the original data.
Single-link clustering methods on the original data
have not been used because of the amount of data and
the effects of chaining. However, clustering directly on
constant values within zones after zoning gives a reduced
data set and works well. This is, in effect, a method of
segmentation followed by clustering.
Figure 5-30 - Schemes explaining the clustering of points info local
modes and terminal modes.
X
Figure 5-29 - Determination of local modes or elementary electrofacies.
Final electrofacies determination
The previous described method defines a certain number of local modes (or elementary electrofacies) that is
much smaller than the original number of levels. A list of
the local modes classified following the order of creation
is also produced. It indicates the depth of the most representative value, the dk value, and the link with other
modes. The coordinates of the principal components are
also given.
Even if they describe logging reality better and more
objectively (compare zoned logs with raw logs of Fig. 524), these local modes are often too numerous to be easily correlated to geological facies. The analysis of their proximity to each other can group some elementary modes to
reduce the number of final electrofacies to a value close
to the number of geological facies (Fig. 5-30). One
method is to search within the multidimensional histogram
for cells of the locally highest frequency (modes) and then
to construct a dendrogram of similarities between the
modes (Fig. 5-31).
The dendrogram is a graphic representation of the distance, in n-dimensional space, between each mode and
its closest neighbour. Values along the x-axis of Fig. 5-31
correspond to these distances. They give an idea of the
degree Of
between modes and can be used to
justify the grouping of modes into the final electrofacies.
The analyst then chooses a reduced list of modes as seed
216
15 10
Figure 5-37- Example of dendrogram and its use for defining the terminal modes.
Serralog Q 2003
Facies, Sequence and Environment /Chapter 51 217
Principal CornDonent Analvsis
In fact, the clustering methods are applied on the principal component logs, which are derived from a Principal
Component Analysis (PCA).
As we noted previously, each log is influenced, in different degrees, by the geological parameters of the rocks.
In combining several logs we have, automatically, a certain redundancy of geological information. This is highly
useful for mutual verification of the data quality .The goal
of the Principal Component Analysis is to study the correlations between log data, to reduce the number of n variables to a lower number rn, by eliminating the insignificant
components.
The Principal Component Analysis, or PCA, is a statistical study of the logging data over a given interval. From
the search of correlations between logging parameters
(Tables 5-4 and 5 4 , PCA replaces n measured parameters (or n curves or logs as IL, SP, FDC, CNL, GR, etc.)
with n other uncorrelated parameters (or n PC logs). In
fact, this results in a change of coordinate axes (Fig. 532). The readings of n logs at a given depth can be considered as coordinates of a point (corresponding to the
depth) in a n-dimensional space.
the largest variability. The second (or PC 2) is the second
largest but in a perpendicular direction. etc. There is no
correlation between PC 1, ... , PC n, so their use eliminates the redundancy between original logs and permits isolation of the elementary effects. The amount of original
information that each PC axis contains decreases from
PC 1 to PC n (Table 5-6). When the n of them are taken,
the total amount of information contained in the original
set of logs is restored. PC axes of high order contain little
information, which in some cases can be considered
noise. In this case, we can eliminate them in a further
treatment. This corresponds to a filtering and reduces the
number of parameters to be taken into account (it reduces
the dimensionality of the cloud from n to rn, rn being lower
than n).
Table 5-4
Statistical analysis of logging data.
GR
hRI
M E
DT
VAR
F?=E
BAL
SW
ALT
I
60,0135 21,4679 85,8293 24,9209 110,7502
0,3139
0,1726
0,4866
0,3261
0,0739
2,3779
0,456 2,1101
2,5661
0,0931
72.564
74.362 146.926
100.7467 18.3361
1,97711
1,0761 6.19611 0,10981 6,30591
1.503
8.25
3,784
0,25
8,5
9,6154 89,5028
47,0103 14,2632 79,8874
2,9793
1,217
0,2996
8,4376
8,7372
6,662
3,0032
21,918
2,8018 24,7199
2
15
1,6
1
I
0.35
4%
0.35
06
0,35
Figure 5-32 - Principal Component Analysis (PCA) corresponds to a
change of axes.
0,35
Table 5-5
Correlation matrix between wireline logs.
But, if variations observed on a PC of high order correspond to events that appear on one log and not on the
others, these events cannot be considered as noise (i.e.
organic material only detected by the uranium content,
measured by the NGT tool).
Table 5-6
Evaluation of the principal axes and rank by amount of
original information carried.
1
VAR. FRE. BAL, SHA, ALT refer to synthetic logs coming from the SYNDIP program applied to HDT data processed by GEODIP program.
Over a given interval, all measurements or levels (or
sets of log data) define a cloud in n-dimensional space.
This n-dimensional cloud can be described by a set of
axes. PCA defines the principal axes of inertia. If each
data point has the same weight, the first axis of inertia (or
PC 1) will be aligned according to the direction of maximum length. In other words, the first axis is the one having
Serralog 0 2003
AXE
1
_w_ _ _
m
;1I
pelce"tage[
--
cumtatad
.- -___
I
percentage
64,6171
64,6171
22.341I
86.9581
b.l s e rcb
]
I
11 0 , 6 6 8 8 W l I
21 0.2312 MII
31
l
0,5081 W
4,9091
91,8671
-7,457
I
41
51
0,3224WI
0.2992~1
0,1008 D+O
3,1151
2.891I
0,974
94,9821
97.8731
98,847
-0,208
-0.3871
-2,965
0,786
0,294
99,634
99.928
-3,605
3,223
6
7
8
I
0,8140D01
0.3045D01
91 0.7478 DO21
0.0721
1001
-1,003
-2.5521
14.321
217
The distribution of the information carried by a log between the PC logs, can be represented by histograms
(Fig. 5-33). At the same time correlation between logs and
PC 4s provided (Table 5-7). This can help to understand
what type of information is essentially represented by a
given PC log.
Active
inertia = 0.00866
Active
Inertia = 0.00546
PC
PC
As previously mentionned PC logs do not carry additional information. It is just another way to present the
essential information contained in the original logs. This
technique makes clustering somewhat easier, with little or
no loss of information. Hence, after the principal component analysis of the standard logs, a reduced set of principal component logs is chosen. PC logs can be drawn as
logs (Fig. 5-34).
The definition of the axis is based on the cloud distribution. The latter will depend on the weight given to each
data. Hence, if we want to give priority to mineralogy, we
can put more weight on those logs that are more sensitive to composition (NGT, LDT, GST, ACT tools). This situation will amplify the variations of parameters measured by
these tools, and will diminish the weight given to those
logs particularly sensitive to porosity and saturation (FDC,
CNL, LL, IL ... ). This will naturally compress the variations
of parameters obtained from the last mentioned tools.
The results of either of these approaches is an affectation of each log sampling level to one electrofacies.
Figure 5-20 shows the results obtained by the partitioning
technique. The electrofacies is coded arbitrarily as a number from 0 to 20. Evidently, we expect the n’ electrofacies
to reappear whenever the core description indicates a
repeat of the lithofacies type. The maps from electrofacies
to core lithofacies can be made easily by means of the
representation of Figure 5-20. For example, the no7 electrofacies is a clean sand, whereas no 5 is a limestone.
Electrofacies 12 repeats several times. It corresponds to
the shale intervals.
Figure 5-33 - Histograms illustrating the contribution of the principal
components to the reconstruction of the logs.
Table 5-7
Correlations between wireline logs and principal
components.
1450
Figure 5-34 - Example of PC logs.
Once the principal axes of inertia have been calculated
from a set of “active” logs, it is also possible to project
other logs or curves (i.e., “passive” logs) onto these axes.
Thus, one can determine which part of the passive logs
can be explained by the information contained in the set
of active ones and how much the passive logs are related
to each PC log.
218
The correspondence with core data is shown.The
“ladder” presentation for some of the electrofacies (no 5,
6, 7, 8 and 12) also is shown. Aglance at the “ladder” corresponding to facies no 12 shows a wide variation of several logs. Furthermore, the GEODIP results shown on the
right indicate different response types within electrofacies
Serralog 0 2003
Next Page
1
Facies, Sequence and Environment Chapter 51 219
no 12. It can be subdivided by establishing subfacies in
the shale zone.
Figure 5-35 illustrates the result of direct clustering on
the zone values. In this case, the criteria used in the automatic zoning program have been relaxed to give longer
and, hence, fewer zones than in Figure 5-20. The dendrogram shows the zone "families" grouped in terms of their
similarities and distinguished by use of a cutoff level.
Each of these families is assigned an electrofacies number, as shown in the appropriate column. We can note
immediately that this method gives four different shale
types and three shaly sand types. With a more detailed
definition of the core, we probably would see differences
in the composition of each of these shales or shaly sands.
The main layers of marl, sand, and limestone are seen in
Figure 5-20.
- It is not recommended to try to obtain, at any price,
the same number of electrofacies as well recognized in
the core by facies analysis made by a geologist. Even if
the logging tools, and especially image tools, "substitute
the geologist's eyes", it is impossible for them to "see" the
formations as a geologist would see them. The tools do
not react to the same parameters as the eyes and the
brain of geologist do. To define a facies, he will often "filter" the available information, and rely on a feeling that is
based on some of the information : composition, colour,
certain elements of texture, and more often sedimentary
features and faunistic associations. He pays little attention
to the precise proportion of each mineral or element,
because these are studied on chosen samples, and they
may vary appreciably from one point to another. He rarely estimates the porosity and never includes the fluids.
Figure 5-35 - Direct clustering on zone values (from Serra & Abbott, 1980).
An example of electrofacies analysis in carbonates is
given by Figure 5-36. A final selection of 9 terminal electrofacies was made. The superposition of the squared
logs obtained from these 9 terminal modes on the actual
logs (Fig. 5-37) gives a good quality-control indicator of
the selection of terminal electrofacies. The 9 selected
electrofacies allow an adequate description of the main
geological facies over this interval. At the same time, the
representative mean values of the parameters characterizing each final electrofacies are reproduced under the
form of a listing (Table 5-8 next pages).
At this stage it is necessary to make the following
remarks.
Semlog Q 2003
The majority of tools are sensitive to minor variations in
composition, and the latter is not easily detected in a core
except through a very detailed and expensive analysis.
- The purpose of electrofacies analysis is to describe
the formations from the tool responses. If we wish to be
precise and objective it is necessary that the retained
electrofacies reflect, first of all, log parameters. This verification is carried out by superposing the average values,
defined for each electrofacies, on the raw data (Fig. 5-38).
- It is important that the geologist gets to know this new
concept and learns to exploit the results of this new
method of formation analysis. It is necessary that he
benefits from its advantages; objectivity, rapidity, quantifi219
Previous Page
A = carbonaceous shale; B = nodular bedded shale; C = dolomitic mudstone; D = packstone to wackestone with abundant rounded algae;
E = packstone to wackestonerich in foraminifers; F = packstone to wackestonerich in large foraminifers; G = bioclastic mudstone to wackestone;
H = packstone to wackestone rich in bioclasts; I = upper shale, calcareous.
Figure 5-36 - Open-hole logs, synthetic logs, GEODIP and FACIOLOG results and GLOBAL evaluation of a well
(from Schlurnberger Well Evaluation Conference, India, 1983).
cation, accuracy, and finesse of analysis. However, while
maintaining the detail made possible by analysis of the
electrofacies, the geologist can cluster several electrofacies under the same shading, representing one lithofacies.
- As it was previously mentionned he can give more
weight to those log data particularly sensitive to lithological parameters, or to sedimentary features.
220
- If the beds are very thin, compared to the vertical
resolution of the tools, or if the series are fundamentally
composed of thin sequences (i.e. distal turbidites, tempestites), the majority of open-hole logging tools will not
recognise the proper electrofacies of each bed or each
term of the sequence. The computer processing will thus
define an electrofacies corresponding to an average
response that combines the influence of two or more
Serralog CJ2003
Facies, Sequence and Environment IChapter 51 221
beds, and therefore facies. This electrofacies will not give
an accurate idea of the real facies. In such a case, it is
necessary to use tools with a high vertical resolution (dipmeters, EPT, image tools, NMR tools), and a combination
of the results obtained by processing of the data with, for
instance, the LITHO and SYNDIP programs for dipmeter
tools or with BorTex for image tools. This allows conversion of the dipmeter or image resistivity curves in terms of
lithology and structure (Fig. 5-38), by analyzing dipmeter
or image resistivity histograms (Figs. 5-39 and 5-40)and
the nature of the events on the curves.
Translation of electrofacies to facies
An electrofacies is an abstract mathematical concept
that needs to be "translated" into geological words to be
easily understood and used by geologist. Even if the two
approaches are very similar, they differ in language : a set
of words and qualitative adjectives for the description of
the facies by a geologist, a set of log parameters for characterizing an electrofacies by a log analyst. To pass from
one language to the other it is necessary to "bring together the two vocabularies". The most accurate method is
to put alongside each other, at the same scale, the two
Table 5-8
Listing of the mean values of the parameters characterizing each terminal mode or electrofacies.
I
I
I
I
I
I
11
-3.4071 6.8871 33,7851 24.328)
21 -3.7041 1.5881 48.1121 18.5191
31 -2.1621 16.821I 45.7921 27.5261
2.4691 84.1751
2.5421 75.9251
2.461I 87,631I
2.7711
2.3081
1.7281
5.0671 41.6481
3.751 63.2891
3.231 50.6981
3.7591
3.6271
2.7221
5.2281
5.831
6.6781
41
-0.9921 11.2581
2.4911 87.9281
2.4181
4.3931
4.1071
4.9221
5
-1,156
0,933
1,747
2,308 93,075
2,423 101,593
2,429 100,284
1,616
2,627
3,581
4,194
2,338
3,825
4,936
6,802
5,831
5,523
6
7
29,801
7,02
2,649
67.671 27.2811
41,152
70,53
80,276
30,77
37,164
39,906
__
1
-
Figure 5-37 - Terminal modes or electrofacies :averaged squared logs
superposed on actual logs
(from Schlumberger Well Evaluation Conference, India, 1983).
Serralog 0 2003
4,094
54.531
43,311
37,255
5,5 41,979
types of analysis, and to attach to each electrofacies the
descriptive terms used by the geologist to describe facies.
One must attach to the set of log values which characterize an electrofacies (pb, I$,., Pe, At, GR, Th, K ...), the
set of words and adjectives describing the composition,
the texture and the structure of the corresponding bed
(e.g. feldspathic well sorted sandstone with dolomitic
cement, cross-bedded, having a porosity between 18 and
23 %, water saturated).
Sometimes the logs recognize an electrofacies that a
first core analysis has not differentiated. In this case, if it
is not linked to a fluid change, it is better to review and reanalyze the core in a more elaborate manner, and try to
explain the origin of this electrofacies (i.e. presence of
heavy radioactive minerals, of uranium linked to phosphates or organic matter ...).
If the interval was not cored, we will try to translate the
electrofacies into facies using interpretation techniques.
To achieve this objective the lithology must first be defined, and the features, extracted from the dipmeter or
image analysis, must be converted into qualifying adjectives.
The conversion of well log responses to lithology is
obtained by cross-plot analysis or by using an automatic
program based on the construction of a lithofacies data
base which itself is done by combining multi cross-plot
analysis and conversion of the composition of rock lithofacies into log responses. This last program is LITHO or
SQWIZLOG and has been described in Chapter 2.
The conversion of the dipmeter data in textural and
structural information is explained in Chapters 3 and 4
and can be achieved with the SYNDlP or BorTex programs described in the same chapters.
221
SvNTHETK=
LlTHOFAClES
DESCRIPTION
THICKNESSES
STRATK3RAPHK:
COLUMN
Z
SUBDEFlNlTlON
3
=Ill
n
Figure 5-38 - Example of the lithological interpretation of dipmeter resistivity curves by a combination of LlTHO and SYNDIP results.
-
Zoning has been defined
Figure 5-39 - Example of dipmeter resistivity histogram and its conversion in lithology.
222
Figure 5-40 - Example of resistivity histogram of image data.
Serralog 0 2003
Facies, Sequence and Environment‘Chapter 51 223
Result presentation
The presentation of the results of the interpretation by
the FACIOLOG program must include the original logs,
the dipmeter results provided by a processing of the HDT
or SHDT data using GEODIP or LOCDIP programs, and if
possible the lithological column obtained from LITHO or
SQWIZLOG as well as a display of the SYNDIP results.
Hence, we have the means of verifying the quality of the
electrofacies analysis. We may also add the results of a
quantitative interpretation provided by a program such as
ELAN. It is also possible to superpose on the raw logs the
zoning derived from the electrofacies analysis.
The electrosequence concept
We sometimes observe, on certain recordings, progressive evolutions of measured parameters (resistivity,
gamma ray, spontaneous potential, etc.) in relation to
depth (Fig. 5-41). These evolutions, having the shape of
ramps, were named electroseguences by Serra (1970).
The proposed definition of an electrosequence is :
“a depth interval thicker than the vertical resolution of
the measuring tool, presenting a progressive and continuous evolution between two extreme values of measured parameter, tracing a ramp”.
This variation may reflect:
- a progressive change in mineralogical composition
with depth, generally observed on Pe, density and radioactivity measurements: percentage evolution of shale in a
sand or in a limestone; enrichment of a limestone in dolomite, or of a sand in radioactive heavy minerals (Fig. 5-
Figure 5-42 - Example of ramps corresponding to grain size evolution
detected on well logs. Observe, between 210 and 182 ft, the progressive evolution of GR, thorium (Th), uranium (U)and density curves. They
suggest a progressive enrichment of the sand with heavy radioactive
minerals, thorium and uranium-bearing, which can be correlated with a
decrease in grain size, heavy minerals being more abundant in the
silty fraction than in the sandy. Observe the neutron curve (IHCNL)): it
does not show significant porosity variations in the same interval. The
potassium curve (K) is practically zero, indicating the absence of clay
minerals
U
Figure 5-41 - Electrosequences (ramps) clarly seen on well logs.
Serralog Q 2003
223
-the evolution of a textural parameter: grain size change reflecting fining or coarsening upward sequences; sorting decrease, etc.;
- a simultaneous variation in mineralogical composition
and in texture (conglomerate to sand to shale);
- a saturation evolution in the transition zone between
oil- and water-bearing reservoirs that appears particularly
on resistivity curves.
This electrosequence is not necessarily observed on
all curves and, moreover, if it has a small vertical extent,
it will be detectable only by microdevices (dipmeters,
borehole imagery tools, microlog Fig. 5-43).
medium to fine
:graMsadStone
n
wor wave
laminated
i
Imedium
%Ya%
%&
' toe
grad
medium grirned
s;ndslone m a l l y
graded, clean
Figure 5-44 - Thin microsequences or ramps within one macrosequence seen on HDT dipmeter resistivity curves
(from Sera &Abbott, 1980).
ters (hydrogen-index and density for instance) shows a
continuous form (Fig. 5-45), even if each separate curve
(Fig. 5-46) does not clearly show a smooth evolution.
COAL
Figure 5-43 - Example of dipmeter resistivity evolution suggesting thin
fining upward sequences contradicting the evolution suggested by the
gamma ray log (from Serra, in Schlumberger Well Evaluation
Conference, Algeria, 1979).
It corresponds, most often, to a first order sequence
(thin or microscopic according to the definition of Lombard
1956), and sometimes to a second order (medium or
macroscopic), if detailed study of the curves shows very
fine variations in the general trend (Fig. 5-44).
The "bell" and "funnel" shapes of the SP curve correspond to electrosequences (cf. Fig. 5-6), and not to
facies or environment.
It is possible to extend the electrosequence concept to
all depth intervals in which the cross-plot of two parame-
Figure 5-45a - Plot of one progradational sequence showing the crossplot trend (from Rider 8 Laurier, 1979).
The "boomerang" shape_lSometimes observed on density vs neutron cross-plots and well known by log analysts, is another example of an electrosequence of this
type (Fig. 5-47).
Sequence analysis from well logs
It consists of analyzing, on the one hand, the type of
transition from one electrofacies to another (gradational or
abrupt contact), and on the other hand, the arrangement
of electrofacies in vertical sequences, at different scales
(elementary, meso-and mega-sequences).
1
Facies, Sequence and Environment Chapter 51 225
ANALVIIm
correspond to a transition from water-bearing to hydrocarbon-bearing
reservoir, and thus may indicate a progressive change in saturation. A
look at other logs less or not sensitive to fluids (particularly gamma ray)
distinguishes this case from that of a sedimentary sequence.
OP ONE
PIOORAOATtONAL IEOUENCI
ON COC.CNL
caoem
PLOT
FREcuENcTPL0-l
-
Interval 4940 4680 R
Figure 5-456 - Interpretation of Figure 5-45a cross-plot in terms of
facies sequence (from Rider & Laurier, 1979).
Figure 5-47 - The %oomerang”shape observed on neutron vs density
and sonic vs density cross-plots reflects a grain size evolution in a
sand-shale deposit.
LEGEND
Figure 5-46 - Log response of an idealized prograding sedimentary
sequence from shale to sand, shale and coal
(from Rider & Laurier, 7979).
Gradational transition - Elementary seauence
A gradational transition corresponds to a ramp or an
elementary electrosequence, simultaneously detectable
on one, two or several curves : most often gamma ray,
spontaneous potential, resistivity, but sometimes sonic,
density, photoelectric index and neutron.
It may happen that, following the vertical evolution of
thicknesses of each elementary sequence, the macrodevices give a hypothesis for the sequential evolution that
does not agree with that given by the dipmeter- or imageresistivity-curveevolution (Fig. 5-43). In this case, only the
data derived from dipmeter or image are valid.
Rem
ark
It is necessary to remember that a ramp on resistivity curves may
Serrabg 8 2003
When the sedimentary sequence is not very thick (few
centimeters or decimeters), it is not detectable on macrodevices. In this case the electrosequences often appear
more clearly on dipmeter or image resistivity curves (Fig.
5-48 and 5-49).
Remark
Once more, we must be careful in the interpretation of curve evolution in the case of deviated wells. In fact, some sequences are purely
artificial and simply due to the influence of surrounding layers which,
because of the apparent low dip between the axis of the hole and the
beds, are situated behind the bed immediately in front of the measuring
devices (Fig. 5-50).
In a terrigenous detrital series, the elementary electrosequences (cf. Fig. 5-8) indicate evolution both in grain
size and in lithology : normal graded bedding (fining
upward) or reverse graded bedding (coarsening upward).
In a carbonate series, an elementary electrosequence
more often indicates a change in lithology (enrichment in
shale or in dolomite), a diagenetic influence, or, sometimes, a textural evolution (transition from grainstone or
packstone to wackstone or mudstone Fig. 5-51).
225
z
P
w F
DIPS
aP
88
8
n
CORRELATIONS
RESISTIVITY
CURVES
shale
sand
shale
sand
shale
Flning up
wquencs
sand
shale
sand
shale
sand
shale
sand
shale
sand
Figure 5-48 - Example of gradational evolutions or ramps very easily
seen on dipmeter resistivity curves (HOT tool) corresponding to electrosequences. Observe as well the thickness evolution of these electrosequences (between 1 foot and 3 feet). This reflects a thickening
upward sequence of fining upward sequences: With such thicknesses
these sequences cannot be correctly detected by other standard tools.
Figure 5-49 - Example of grain size (fining up) and thickness (thickening and thining up) sequences well seen on FMS images
(from Serra, 1989).
Hole
Axis
A b r u ~ tcontact
It corresponds to an abrupt and significant change of
reading observed at the same depth on one, and more
generally, on several logs simultaneously. Its interpretation needs the study of the relationship between electrofacies. From that analysis one should be able either to
determine sequences of electrofacies, or the significance
of this change : transition from one term of the sequence
to the following one, or break in the sedimentological
sequence, or fault, unconformity ... For instance, if the
abrupt change corresponds to the transition from anhydrite to halite, it only reflects a natural evolution inside the
evaporitic sequence from a less to a more resticted environment.
226
Figure 5-50 - Scheme explaining the different evolutions observed on
dipmeter resistivity curves, related to the apparent low angle between
the axis of the hole and the beds.
Inversely, the abrupt change from a shale to a sand, or
from a sand to a limestone, may indicate a fundamental
Serralog 6 2003
Next Page
Facies, Sequence and Environment Chapter 5 / 227
change in depositional sequence. It marks the beginning
of'a new cycle of sedimentation, after a break or an erosion even if it does not correspond on the dipmeter or
image to a non planar surface (wavy symbol on LOCDIP
display, 4 dip computation on GEODIP display).
Mudstone
But, undoubtedly the best and most accurate method
will be the analysis of their arrangement by probability
methods. Once the electrofacies have been defined, a
study of their vertical organisation can be made following
a method proposed by de Raaf ef a/. (1965). Applying a
procedure suggested by Selley (1970), the succession of
electrofacies has been represented in a diagram resembling a spider's web. An example of what is now referred
to as a "Facies Relationship Diagram" (FRO) is given in
Fig. 5-52. In this diagram, the probability of transition from
one facies to another, as observed in the well, is shown by
numbered arrows.
Observed facies relationship
-Sharp
..__
+Gradational
Grainstone
Figure 5-52 - Example of a Facies Relationship Diagram, showing the
observed number of sharp and gradational transitions between facies
(from Walker, 1979).
SS = scoured surface; A = poorly defined trough cross-bedding;
B = well defined cross-bedding; C = large planar tabular cross-bedding; D = small-scale planar tabular cross-bedding; E = isolated
scours; F = trough cross laminated fine sandstones and shales;
G = low angle stratification.
Boundstone
Figure 5-51 - Example of textural evolution in a limestone. Observe the
general trend on the averaged resistivity curve on the leftside.
Finally, an abrupt change may be linked to a fault or to
an unconformity that brings into contact two rocks having
different properties. The interpretation of the dipmeter and
image data generally allows recognition of these phenomena, and hence determination of the origin of this abrupt
change.
Seauences of electrofacies
The electrofacies, as well as the facies, do not superpose each another randomly, except if there are tectonic
accidents, unconformities, erosion, or periods of non
deposition.
As we have seen previously some cross-plots indicate
directly these sequences of electrofacies. Rider & Laurier
(1979) used this technique to characterize deltaic sequences. They give the log responses for an ideal sedimentary cycle (Fig. 5-46) that consists of a progradation from
shale to sand, with occasional coal deposits. Their
method uses the location of each facies on crossplots
(Fig. 5-45).
Serralog Q 2003
In order to construct a FRD, a computer program tabulates a difference matrix (observed minus random transition probabilities) as shown in Table 5-9. The observed
transition probability is the number of observed transitions
in the well from one electrofacies to another converted to
probabilities. The random transition probability is obtained
on the assumption that all facies transitions are random,
and depends only on the absolute abundance of the
various electrofacies. The difference between the two probabilities gives the difference matrix. It is to be noted that
some values are high-positive (transitions much more
common than if electrofacies were random) and some are
high-negative (transition much less common than random). The last row gives the total number of occurrences
of a particular facies (e.g. facies H is observed only once).
Remark
Walker (1984) mentionned that this method is statistically incorrect.
He suggests using a more complex methods of Markov chain analysis.
The interpretation of the FRD suggests a cycle of
deposition as represented by Fig. 5-53.
Applications of facies and sequence
analysis
Facies and sequence analysis is a fundamental step of
the geological study of a formation. Consequently, it is
normal that its applications cover a certain number of
fields related to sedimentology, and the quantitative inter227
Previous Page
pretation of reservoirs.
Table 5-9
Observed minus random transition probabilities
for well of Fig. 5-36
(from Serra, in Schlumberger Well Evaluation
Conference, India, 1983).
FACES
1
2
3
I
I
1 1 2 1 3 1 4 1
I 18 I 39 I -20 I
1-8
-19 -19
1-1 1-1 I
I 9 I
**L
1-1
-
I
I
5 1 6 1 7
-2 I -18 I -15
-20 -17
85
0 I 3 1-9
I
I
I
I
I
I
I
why we cannot be solely interested in the electrofacies of
sedimentary bodies. In fact, when we analyze the shape
of the SP curve we make this kind of interpretation unconsciously, the "bell" and "funnel" shapes corresponding to
electrosequences as previously noted. But, still we cannot
be conclusive, because such sequential evolution can
belong to several environments. The choice between the
several hypotheses will be done by taking into account
other information related to the thickness of each electrofacies and electrosequence, and its evolution with depth
(Figs. 5-54 and 5-55) and in space through well to well
correlations (Fig. 5-56).
Alluvian fan
(Steel eteL, 1977)
River dominated
delta
(Miall, 1979)
5,o
Wave dominated
delta
(Miall, 1979)
mud
silt
sand
conglomerate
coal
trough
crossbedding
herringbone
crossbedding
Iowan I
crossading
crossbed8ng
hummoc
Figure 5-53 - Spider's web diagram for well of Fig. 5-37 representing
the facies relationship as obtained from data of Table 5-7, and its FRD
interpretation (from Sera, in Schlumberger Well Evaluation
Conference, India, 1983).
.
Reconstitution of the depositional environment
As it was pointed out by Middleton (1978), the final
goal of facies and sequence analysis is the reconstitution
of the depositional environment. But, Walker (1979) stated "many, if not most, facies defined in the field have
ambiguous interpretations - a cross-bedded sandstone
facies, for example, could be formed in a meandering or
braided river, a tidal channel, an offshore area dominated
by alongshore currents, or on an open shelf dominated by
tidal currents". In such cases, "the key to interpretation is
to analyze all of the facies communally, in context. The
sequence in which they occur thus contributes as much
information as the facies themselves". Finally, the
sequence will usually allow recognition of the depositional
environment.
In the same way, we can only reconstruct the depositional environment, from logs, by replacing the electrofacies both in vertical and lateral electrosequences. This is
228
RftS
roots
shell debris
bioturbation
Barrier island
(Davies etel., 1971)
4
4
Prograding storry
dominated shoreline
Submanne fan
(Hamblin 8 Walker, 1979)
(Walker, 1979)
-
Figure 5-54 Lithology, grain size and thickness evolution allow the
determination of the depositional environment (in Miall, 1984).
Serralog @ 2003
Facies, Sequence and Environment IChapter 51 229
Gravel
Sand
Silt
Clay
Meandering
system
Braided
system
Figure 5-55 - Grain size evolution, type of bed boundary, thickness and
frequence of occurrence must be taken into account to determine the
depositional environment (adapted from Sellev, 1976).
An environment is a general term used by geomorphologists or oceanographers to characterize physiographic or morphologic units (mountain ranges, desert, deltas, continental shelves, abyssal plains ... A sedimentary
depositional environment is a geographically restricted
part of the earth's surface, which can be easily distinguisWELL
hed from its adjacent areas by the complex of physical,
chemical and biological conditions, influences or forces
under which a sediment accumulates. This complex largely characterizes the environment and determines the
properties of the sediments deposited within it (Krumbein
& Sloss, 1963; Selley, 1970; Reineck & Singh, 1975; Blatt
et al., 1980).
The physical conditions which act on and control an
environment are numerous. They include
- the climate: the weather and its components
. temperature variations (diurnal, nocturnal, seasonal);
. importance and frequency of rainfalls, snowfalls;
. humidity;
. wind regime (dominant direction, velocity, and their
variations);
. all these factors acting on the vegetal cover;
- the altitude and the topographic profile nature, size,
shape and slope of the mountains or of the receiving
basin, the energy of the flow, the water depth, which will
control the hydraulic regime; in marine environments : the
bathymetry, the amplitude of the tides, the waves, the current system of the water mass, the wind regime, the
Coriolis forces.
The chemical conditions which operate within an environment include the geochemistry of the rocks at the surface, and of the waters (river, lake, sea, ocean) : salinity
(nature and percentage of salts in solution), pH, Eh, gas
in solution.
The biological conditions comprise both fauna and
flora, terrestrial or aquatic, and bacteria present in the
environment.
These three conditions or factors are not independent
but, to the contrary, are strongly linked. Any change in one
of them has immediate repercussions on the others.
Following the medium (air, ice, water) and the relative
importance of each condition, factor and process, an environment can be depositional, erosional or non depositional (equilibrium). As a broad generalization, subaerial
environments are essentially erosional, while subaqueous environments are mostly depositional.
WELL
WELL
Figure 5-56 - Correlations between facies and sequences represented by postulated core and well-log (SPor GR) responses following the
Walther's law (from Van Wagoneret al., 1990).
Serralog 8 2003
229
In the study of ancient deposits, because we analyze
sedimentary rocks, we will always refer to depositional
environments. However, it must be kept in mind that
depositional sequences can be interrupted by periods of
non deposition or even erosion, which of course have no
associated sediments. It will be important to detect and
localize them, both in time and space, because they will
enable a better definition of the environment and the geological history of a sedimentary basin. They can also be
used for correlation purposes.
The determination of the depositional environment will
only be possible through the description of the imprints, or
responses, that the physical, chemical, biological and
geomorphological conditions characterizing the environment left in the deposit.
Those imprints define a facies which is, as previously
seen, the "sum of the physical, chemical and biological
characteristics which differentiate a sedimentary body
from another".
As pointed out by Krumbein & Sloss (1963), "the study
of any sedimentary environment includes considerations
of the four basic environmental elements" listed in Figure
5-57. To each element are attached several factors. "The
relative influence of environmental elements varies with
the nature of the environment. Similarly, the factors in any
element vary in their importance within different environments".
The material factors include the medium (air, fresh or
salty water, ice), the pH, the Eh, the temperature, the dissolved salts (Ca++, Na+, C03-- , SO4--, Cl-), and gases
(CO, O,, SH,), and the solids. "Each of the material factors may have some effect on sediments being deposited,
but the importance of the effect is controlled, in part, by
factors of other elements". For instance in a high energy
environment (zone of breaking waves), the solids (sand
and gravel) are dominant, salts and gases have very little
influence. On the contrary, in a low-energy environment
(quiet water), they become dominant in controlling organisms, and hence play a part in the deposition of carbonate.
"Boundary and energy factors include water depth,
distance from shore, topography of the bottom, and, in a
general way, the geography of the depositional area". The
energy controls the quantity of particles in suspension,
and the size of the settling grains. It also controls the current type (laminar or turbulent) and consequently the
nature of sedimentary features.
The energy and material factors control, in turn, the
biological factors. "In clear warm seas, well oxygenated
and mildly alkaline, organisms may thrive and produce
abundant carbonate sediment" (Krumbein & Sloss, 1963).
On the contrary, turbid medium, or reducing conditions,
may restrict living conditions for some organisms.
The area variations of the elements and factors within
an environment, called the environmental pattern by
Krumbein & Sloss, (1963), induce progressive evolution
of the nature and properties of the sediments creating a
lateral sequence of facies.
But, the accumulation of sediments in an environment
can modify its elements and factors sufficiently to create
new conditions and consequently new properties. "This
sort of interlocked relation between process and response is called feedback". These continuous modifications at
any given geographical point contribute to create a vertical sequence of facies.
Consequently, an environment will be characterized by
a typical sequence of facies both in space and time.
Following the Walther's law' these sequences are similar.
So, the knowledge of the vertical sequence, which can
be obtained from wireline logs, helps to predict the type of
ENVlRONMENTAL FACTORS
RESULTING FEATURES
- Size of the deposits
Geometry
Geometry of the environment
- Properties of the sediments
- composition
- texture
Energy of the environment
- sedimentary structure
- color
- fossils
-I
Lithology FACIE!
Paleocurrents
Paleontology
Matrials of the environment
3iological elements of the environment
- Lateral variations in the
sedimment properties
+
Lateral
sequence
V
Feedback on the environmental factors: prog
Figure 5-57 - A generalized sedimentary environment process-response model (from Krumbein & Sloss, 1963).
230
Serralog Q 2003
Facies, Sequence and E n v i r o n s 231
lateral sequence and the depositional environment.
By considering all the possible combinations of the
physical, chemical and biological processes or all elements and factors which characterize an environment,
one could guess that there should be an infinite number
of environments. In fact, due to the strong interdependence of all these processes and factors, a finite number of
environments has been recognized.This is also related to
the fact that there are a finite number of physiographic
types. It is for this reason that, even if two environments
or morphologic units are never totally identical, the number of major environments is reduced if one considers the
dominant factors. They are listed in Table 5-1 0. Of course,
especially when we study modern deposits, several subor even sub-sub-environments can be added to take into
account small variations. But, for the ancient sediments
the separation between them is not an easy task. It is for
that reason that we will limit the illustration of their recognition by wireline logs to the major environments.
Table 5-10
A classification of depositional sedimentary environment
(from Selley, 1970).
Fanglomerate
Fluviatile
Continental
Lacustrine
Eolian
Lobate (deltaic)
Shoreline
Linear (barrier)
Marine
I
Braided
Meandering
Terrigenous
Mixed carbonat
terrigenous
Shelf
Turbidite Carbonate
This table tabulates only those environments which have generated large volumes
of ancient sediments.
Interest of the reconstitution of the depositional
environment
The reconstitution of the depositional environment is
not a pure academic objective. It is a necessity in the
search for mineral natural resources such as hydrocarbons, coal, phosphates, potassium salts, uranium...
because these resources are, most commonly, strongly
associated to specific depositional environments.
In petroleum exploration, one of the essential objectives is the evaluation of the hydrocarbon potential of a
basin. This requests the determination of the quality, the
thickness and lateral extension (the volume) of the different facies which represent source rocks in which will be
generated hydrocarbons, reservoir rocks in which hydrocarbons will accumulate by migration from the previous
rocks, and cap rocks which will constitute impermeable
Serralog CJ2003
traps avoiding any dismigration of the hydrocarbons from
reservoirs.
In reservoir study, only the reconstitution of the depositional environment gives a correct idea of the lateral
evolution of the facies and consequently of the petrophysical properties of the reservoirs, and to enable prediction
of the existence, nature, importance and distribution of
permeability barriers. This information is of the utmost
importance to evaluate the production potential of a field
and to locate extraction and injection wells for a better oil
recovery.
Recoanition of the maior depositional environments
from loaaina data
In the study of ancient deposits, the traditional
approach of geologists consists of analyzing rock samples by defining their facies through the mineralogical,
textural and structural characteristics of the rocks.
The knowledge of the lithology and the mineralogy is
important, but it is necessary to control if diagenesis has
modified the original rock type.
The grain size and the textural maturity can help to
recognize the depositional environment as illustrated by
Figure 3-7 of Chapter 3.
As previously seen, sedimentary features constitute a
fundamental information for the recognition of the facies
and the environment.
But, the facies knowledge is not sufficient to identify an
environment. As pointed out by Walker (1979), “a crossbedded sandstone facies, for example, could be formed in
a meandering or braided river, a tidal channel, an offshore area dominated by alongshore currents, or on an open
shelf dominated by tidal currents.” Additional information
related to the thickness of each facies, or sequence of
facies, and to its evolution with time, and consequently
with depth, will often allow discrimination between two or
three possible environments (Figs. 5-54 and 5-55).
In a multi-well study the facies and the environments
will be more accurately recognized using correlation techniques (Fig. 5-12), and facies mapping (isoliths or isopachs, Fig. 5-13), the geometry of sedimentary bodies
being an important parameter for facies and environment
recognition. For example, the meandering channel fill system is easily recognized from the isopach map of the
lower Muddy in the South Glenrock Oilfield, Wyoming
(Fig. 5-13 from Curry & Curry, 1972).
In fact, the depositional environment will be better and
more accurately defined by integrating all the information
on facies and sequences of facies. “The key to interpretation is to analyze all of the facies communall~in context,.
The sequence in which they occur thus contributes as
much information as the facies themselves” (Walker,
1984).
These sequences can be described, as suggested by
Visher (1965), by a vertical profile (Fig. 5-58), or, in other
terms introduced by Walker (1976), by a facies model. As
illustrated by Figure 5-59 from Walker, the facies model is
obtained by “distilling and concentrating the important
231
~
features that' several similar environments "have in common".
As Walker explained "a facies model could be defined
as a general summary of a specific sedimentary environment, written in terms that make the summary useable in
at least four different ways".
For Walker a facies model must "act as
- a norm, for purpose of comparison;
- a framework and guide for future observations;
- a predictor in new geological situations;
- a basis for hydrodynamic interpretation of the environment or system that it represents".
In a subsurface study involving well logs, the equivalent of the facies model will be the concept of an electrofacies model. As seen in previous chapters, wireline logs
enable determination of all the components of facies
(mineralogy, texture, sedimentary features, geometry).
One can also define the organization of facies in sequences. Many authors (Fisher, Visher, Pirson, Coleman,
Galloway ...) have used the SP-resistivity curve shapes to
determine facies and environments. An illustration is
given by the small sketches of Figures 5-60 and 5-61 from
Fisher (1969) which quickly determine the different facies
and sub-environments of two delta systems.
But, as previously mentioned, SP and resistivity curves
are not always sufficient to correctly determine the facies.
The other logs, especially the dipmeter or images, bring
information which help to be more accurate and precise.
It is the reason why I suggest they be introduced, when
available, in any sedimentological study.
Local examples
A
B
C
D
E
distilled away
d
differ from norm
I1
A
@-
Model as framework
\'
@ Model as norm
@
54
, I
'y
Pure essence of
environmentalsummary
7G d f f
\
Model as basis for
hydrodynamic interpretation
Faciesmodel
\
-=-
0
Model as
c
.... ... a
pydictor
\L
[Evidence in local area X +
z:zE
= predictions in area
XI
~~~
Figure 5-59 - Distillation of a general facies model from various local
examples, and its use as a norm, framework for observation. predictor, and basis for interpretation of new example
(adapted from Walker, 1984).
SP
Resistlvity
SP ResbUvitySP Resistivity
SP
Resistivity
Figure 5-58 - Determination of facies from the grain-size evolution, the
sedimentary features and the geometry. Case of fluvial deposits
(from Visher, 1965).
Illustration of depositional environments from l o w
data
In the following pages, the main geological
ristics of several major environments, on a single vertical
profile, are described, through the geological faciesmodel concept.
232
Figure 5-60 - Examples of interpretation of vertical SP and resistivity
profiles in high-constructive lobate delta systems, Gulf Coast Basin
(from Fisher, 1969).
Serralog Q 2003
Facies, Sequence and Environment )Chapter 51 233
SP
SP
Resistivity
Resistivity
SP
Resistivity
SP
Resistivity
SP
Resistivity
SP
Resistivity SP Reslsthrity
A - Fluvial channel fades
-
E Chand -diannel mouthbar
facies (progradational)
stacked ~ 8 sbader
~ l
hies
D - Sbandplain-shelf
fades
E - Sbandpiainlagoonmarsh facies
F Lagoon-marshf!ad basin facies
~
G - shelf facies
Figure 5-61 - Example of interpretation of SP and resistivity profiles in high-destructive, wave dominated delta system, Delta Coast Basin
(from Fisher, 1969).
The geological parameters are then "translated" into
well-log responses, both openhole logs and dipmeter or
image data, the validity of which is in turn controlled and
illustrated by examples. In each case the most important
logging features which allow their recognition are indicated. As suggested by Walker (1976), such an electrofacies model will act as a norm to which any actual example will be compared, and as a guide for future observations and enrichment of the model.
Gamma ray
Calipar
(API)
Figures 5-62 and 5-63 illustrate other examples of log
interpretation in deltaic environment.
The type of environment corresponding to these
examples is illustrated by the sketch of Figure 5-64.
Neutron (%)
Dipmeter arrow plot
Figure 5-62 - Other example of interpretation of log profiles in terms of facies (from Serra & Sulpice, 1975).
Serralog Q 2003
233
DEPTH
Track 1
Track 2
Track 3
Track 4
fed
Figure 5-63 - Example of log response in a deltaic environment. The bottom sand corresponds to a mouth bar capped by a transgressive level
(resistive and compact level rich in shell debris). The second sand corresponds to a tidal channel deposit; the third sand to a beach or barrier bar
deposit; the top sand to a tidal channel deposit. The resistive peaks in the third sand corresponds to layers rich in shell debris.
Reconstruction of the geometry of the facies
After an electrofacies and electrosequence analysis
for each well of a field, or possibly of a licence or even a
basin, we can, through the correlations of electrofacies,
234
jointly with chronostratigraphic correlations between
wells, reconstitute the space-time distribution of the different electrofacies. The application of mapping techniques
will define the geometry of each facies or group of facies
(see Fig. 5-10 and 5-13).
Serralog Q 2003
Facies. Seauence and Environment ,ChaDter 51 235
Figure 5-64 - Sketch of environment corresponding to Figs. 5-62 and
5-63 (based on Weber 8, Daukoru, 1975).
Figure 5-66 - lsopach map of total sands in the Bisti Oilfield, New
Mexico (from Sabins, 1972).
The correlations between wells (Fig. 5-65) and the isopach map of total sands (Fig. 5-66) in the Bisti Oilfield
show clearly the barrier bar nature of the reservoir (from
Sabins, 1972).
application by the realization of isopach or percentage
maps of one typical electrofacies or group of electrofacies. One can also generate maps of sand-shale ratios
computed from FACIOLOG, LITHO or SYNDIP. For
Figure 5-65 - Log correlations between wells in the Bisti Oilfield, New Mexico (from Sabins, 1972).
Mapping of electrofacies
As seen previously, SP and resistivity curves can be
used for mapping purposes, to define the spatial repartition of typical facies. This helps to better define the environment (see Fig. 5-1 0 and 5-67). We can generalise this
Serralog Q 2003
instance, the ratio of the cumulative thickness of resistivity peaks to the cumulative thickness of conductive
troughs computed in a given interval, leads to an accurate sand-shale ratio map if the peaks correspond to sand
beds and the troughs to shale beds. These maps describe objectively, precisely and quantitatively, the lateral evolution of the facies.
235
Choice of a more judicious core sampling for
analysis
I
Figure 5-67
- Reconstruction of the depositional environment through interpretation
of the SP shapes.
This aspect is often neglected. In bringing together the
cores and the results of electrofacies analysis, or at least
the open-hole logs and more precisely the dipmeters and
borehole images, we are able to select in a more reasonable manner the sampling intervals and rates for detailed laboratory analysis. This way to sample allows both a
saving on expense and time, and a clarification of certain
log responses not clearly understood and interpreted.
Thus, if we refer to the example of Fig. 5-48, if the
sampling and the analysis were concentrated on one, two
or even three electrosequences with a sampling interval
of 5 cm, we should be able to calibrate the dipmeter resistivity curves (Fig. 5-68) in terms of grain size, shaliness
and permeability.
Constitution of a data base of electrofacies
By combining the results of electrofacies analysis processed on several wells of a field or a basin, we can constitute an electrofacies data base which can be used as
reference for the study of any new wells in the field or the
basin. Each set of log data will be cross checked with this
data base for attribution. If not attributed a new electrofacies must be introduced into the data base for its completeness.
This application is very important. It will later enable
definition of the depositional environment directly from
well logs using programs involving "expert systems".
Quantitative Interpretation
The electrofacies analysis leading to a reconstitution
of facies and depositional environment, we are able to
attribute to.each electrofacies, or group of electrofacies,
the most probable mineralogical model for quantitative
interpretation (i.e. quartz, potassium feldspar, plagioclase,
kaolinite as principal minerals entering into the composition of a sand). This limits the number of unknowns for
each electrofacies and optimises the interpretation by a
GLOBAL or ELAN type program as suggested by Suau et
a/. (1982).
Moreover, the texture and the sedimentary structure
indicated by dipmeters or borehole images enable definition of the types of distribution of shales and, consequently, a better choice of the response equations (for the
computation of shaliness and saturation), and the parameters a, m and n that relate porosity to the formation factor and saturation (textural and structural models).
One can also better specify the constraints and determine for each electrofacies the functions that permit an
approach to permeability. Finally, one can reduce the
computation time and thus carry out a higher number of
tests by working with local modes or electrofacies instead
of level by level.
236
Figure 5-68 - Calibration of the dipmeter resistivity curve in terms of
grain size, shaliness and permeability.
We should also be able to precise the facies and the
depositional environment (turbidite). In this way, all uncored intervals, and even all the wells of the same field,
representing the same characteristics, would be easily,
rapidly and economically interpreted.
References and Bibliography
BATES, R.L. & JACKSON, J.A. (1980). - Glossary of
Geology. Amer. Geol. Institute, Falls Chuch, Virginia.
BLATT, H., MIDDLETON, G., & MURRAY, R. (1972,
1980). - Origin of Sedimentary Rocks. 1st and 2nd ed.
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
BOUMA, A.H., & BROUWER, A. (eds) (1964). Turbidites. Elsevier, Amsterdam.
BUSCH, D.A. (1971). - Genetic Units in delta prospecting. Bull. Am. Ass. Petrol. Geol., 55, p. 1132-1154.
COLEMAN, J.M., & PRYOR, D.B. (1980). - Deltaic
sand bodies. Am. Ass. Petrol. Geol., Contin. Ed. Course
Note Ser. 15.
CURRY, W.H., & CURRY, W.H. Ill (1972). - South
Glenrock Oilfield, Wyoming : Prediscovery, Thinking and
Post-Discovery Description. In Stratigraphic Oil and Gas
Serralog Q 2003
Facies, Sequence and Environment /Chapter 51 237
-
Fields Classification, Exploration Methods, and Case
Histories. Arner. Assoc. Petroleum Geol., Memoir 16, p.
415-427.
DAVIES, D.K., ETHRIDGE, F.G. &BERG, R.R. (1971).
- Recognition of barrier environments. Bull. Amer. Assoc.
Petroleum Geol., 55, p. 3-26.
DELFINER, P., PEYRET, O., & SERRA, 0. (1984). Automatic determination of Lithology from Well Logs. 59th
Ann. Techn. Conf. SPE of AIME, Houston, Texas; paper no
SPE 13290.
DELHOMME, J.P., & SERRA, 0. (1984). - Dipmeterderived Logs for SedimentologicalAnalysis. SPWLA, 9th
Europ. Intern. Format. Eval. Trans., paper 50.
FISCHER, A.G. (1964). - The Lofer cyclothems of the
Alpine Triassic. In: Symposium on Cyclic Sedimentation.
(ed. by MERRIAM, D.F), p. 107-149. Bull. geol. Surv.
Kansas, 169.
FISHER, W.L. (1969). - Facies characterization of Gulf
Coast basin delta systems, with some Holocene analogues. Gulf Coast Assoc. Geol. SOC.Trans., 19, p. 239261.
FISHER, W.L., & McGOWEN, J.H. (1969). Depositional systems in Wilcox Group (Eocene) of Texas
and their Relation to Occurrence of Oil and Gas. Bull.
Amer. Assoc. Petroleum Geol., 53, p. 30-54.
FISHER, W.L., & BROWN, L.F. Jr. (1972). - Clastic
depositional systems -A genetic approach to facies analysis. University of Texas, Austin, Texas, Bureau of Econ.
Geol.
FISHER, W.L., BROWN, L.F., SCOTT, A.J. &
McGOWEN, J.H. (1969). - Delta systems in the exploration for oil and gas. - A research colloquium. University of
Texas, Austin, Texas Bureau of Econ. Geol.
FONS, L. Sr. (1969). - Geological application of well
logs. SPWLA, 10th Ann. Log. Symp. Trans.
FRIEDMAN, G.M., & SANDERS, J.E. (1978). Principles of Sedimentology. John Wiley & Sons, New
York.
GALLOWAY, W.E., & HOBDAY, D.K. (1983). Terrigenous Clastic Depositional Systems. Springer, New
York.
GILREATH, J.A., & STEPHENS, R.W. (1971). Distributary Front Deposits lnterpreted from Dipmeter
Patterns. Gulf Coast Assoc. Geol. SOC.Trans., 21, p. 233243.
GILREATH, J.A., & STEPHENS, R.W. (1975). lnterpretation of Log Responses in a Deltaic Environment.
Arner. Assoc. Petroleum Geol., Marine Geology
Workshop, Dallas, Texas.
GOETZ, J.I., PRINS, W.J., & LOGAR, J.F. (1977). Reservoir Delineation by Wireline Techniques. Paper presented at 6th Ann. Conv. Indonesia Petroleum Assoc.,
Jakarta, May 1977.
GRESSLY, A (1838). - Observations geologiques sur le
Jura Soleurois. Neue Denkschr. allg. Schweiz. Ges. ges.
Naturw., 2, p. 1-112.
GRIFFITHS, C.M. (1990).- The language of rocks: an
example of the use of suntactic analysis in the interpretation of sedimentary environments from wireline logs. In:
Serralog 0 2003
HURST, A., LOVELL, M.A., & MORTON, A.C. (eds),
Geological Applications of Wireline Logs, Geological
Society, Special Publication no 48, pp. 77-94.
HAUG, E. (1907). - Traite de geologie. Armand Colin,
Paris.
KRUMBEIN, W.C. & SLOSS, L.L. (1963). Stratigraphy and Sedimentation. 2nd ed. W.H. Freeman &
Co., San Francisco.
LENNON, R.B. (1976). - Geological factors in steamsoak projects on the west side of the San Joaquin Basin.
J. Petroleum Technol., July, p. 741-748.
LOMBARD, A. (1956). - Geologie Sedimentaire. Les
series marines. Masson, Paris.
LOMBARD, A. (1972). - Series sedimentaires. Genese
- Evolution. Masson, Paris.
LYELL, C. (1933). - Principles of Geology. John
Murray Ltd., London.
MIALL, A.D. (1977). - A review of the braided-river
depositional environment. Earth Sci. Rev., 13, p.1-62.
MIALL, A.D. (ed.) (1978). - Fluvial sedimentology.
Canad. SOC.Petrol. Geol., Mem. 5.
MIALL, A.D. (1979). - Deltas. In : Facies Models.
(Walker, R.G. ed.). Geoscience Canada, Reprint series 1.
MIALL, A.D. (1984). - Principles of Sedimentary Basin
Analysis. Springer, New York.
MIDDLETON, G.V. (1973). - Johannes Walther's Law
of the correlation of facies. Geol. SOC.America Bull., 84,
p. 979-988.
MIDDLETON, G.V. (1978). - Facies. In : FAIRBRIDGE,
R.W, & BOURGEOIS, J., eds. Encyclopedia of sedimentology. Stroudsburg, Pa., Dowden, Hutchinson & Ross, p.
323-325.
MOORE, R.C. (1949). - Meaning of facies. Geol. SOC.
Amer., Mem. 39, p. 1-34.
NORWOOD, E.M. Jr., & HOLLAND, D.S. (1974). Lithofacies mapping. A descriptive tool for ancient delta
systems of the Louisiana outer continental shelf. Trans.
Gulf Coast Assoc. Geol. SOC.,24.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Houston.
RAAF, J.F.M. de (1968). - Turbidites et associations
sedimentaires apparentees. I and /I. Kon. Ned. Akad.
Wetensch., Amsterdam, Verh., series B, 71, p. 1-23.
RAAF, J.F.M. de, READING, H.G., & WALKER, R.G.
(1965). - Cyclic sedimentation in the Lower Westphalian
of north Devon, England. Sedimentology, 4, p. 1-52.
READING, H.G. (ed.) (1978). - Sedimentary
Environments and Facies. Blackwell Scientific
Publications, Oxford.
RIDER, M.H. (1990).- Gamma-ray log shape used as
a facies indicator: critical analysis of an oversimplified
methodology. In: HURST, A., LOVELL, M.A., & MORTON,
A.C. (eds), Geological Applications of Wireline Logs,
Geological Society, Special Publication no 48, pp. 27-37.
RIDER, M.H., & LAURIER, D. (1979). - Sedimentology
using a computer treatment of well logs. SPWLA, 6th
Europ. Syrnp. Trans., paper J.
SABINS, F.F. (1972). - Comparison of Bisti and
Herseshoe Canyon stratigraphic traps, SanJuan Basin,
237
New Mexiso. In stratigraphic oil and gas field Classification, Exploration methods, and case histories.
Amer. Assoc. Petroleum Geol., Memoir 16, p. 610-622.
Schlumberger (1979). - Well Evaluation Conference.
Algeria.
Schlumberger Technical Services, Inc. (1983). - Well
Evaluation Conference. India.
Schlumberger Middle East S.A. (1984). - Well
Evaluation Conference. Egypt.
SELLEY, R.C. (1970). - Studies of sequence in sediments using a simple mathematical device. J. geol. SOC.
London, Quart. Jour., 125, p. 557-581.
SELLEY, R.C. (1970, 1978). - Ancient Sedimentary
Environments. 1st and 2nd ed. Chapman & Hall, London.
SELLEY, R.C. (1976). - An lntroduction to
Sedimentology. Academic Press, London.
SELLEY, R.C. (1978). - Concepts and methods of
subsurface facies analysis. Amer. Assoc. Petroleum
Geol., Continuing Education Course Note Series 9.
SELLEY, R.C. (1979). - Dipmeter and log motifs in
North Sea submarine-fan sands. Bull. Amer. Assoc.
Petroleum Geol., 63, 6, p. 905-917.
SERRA, 0. (1972). - Diagraphies & Stratigraphie. In:
Mem. B. R. G. M., 77, p. 775-832.
SERRA, 0. (1973). - Interpretation Geologique des
diagraphies en Series Carbonatees. Bull. Centre Rech.
Pau - SNPA, 7 , 1, p. 265-284.
SERRA, 0. (1974). - Interpretation geologique des
Series deltai'ques a partir des diagraphies differees. Rev.
AFTP, 227, Oct., p. 9-17.
SERRA, 0. (1977). - Methode rapide d'analyse faciologique par diagraphies differees. SPWLA, 5th Europ.
Symp. Trans., Paris, paper 9.
SERRA, 0. (1984). - Fundamentals of Well-Log
Interpretation. Volume 1 : The Acquisition of Logging
Data. Developments in Petroleum Science, 15A, 440 p.,
Elsevier, Amsterdam.
SERRA, 0. (1986). - fundamentals of Well-Log
Interpretation. Volume 2 : The lnterpretation of Logging
Data. Developments in Petroleum Science, 15B, 684 p.,
Elsevier, Amsterdam.
SERRA, O., & ABBOTT, H. (1980). - The Contribution
of Logging data to Sedimentology and Stratigraphy. 55th
Ann. Fall Techn. conf. SPE ofAIME, paper SPE 9270, and
in SPE J., Feb. 1982.
SERRA, O., MOTET, D., & VASQUEZ, J. (1993). Detailed Facies, Genetic Sequences and Depositional
Environment Analysis from Well Logs and Electrical
Images. A.A.P.G. Annual Convention, Poster session,
New Orleans, (Abstracts).
SERRA, O., & SULPICE, L. (1975). - Sedimentological
Analysis of shale-sand series from well logs. SPWLA, 16th
Ann. Log. Symp. Trans., paper W
SERRA, O., & SULPICE, L. (1975). - Apports des diagraphies differees aux etudes sedimentologiques des
series argilo-sableuses traversees en sondage. 9th Cong.
Intern. Sediment, Nice, theme 3, p. 86-95.
SHROCK, R.R. (1948). - Sequence in Layered Rocks.
McGraw-Hill Book Co., Inc., New York.
238
STEEL, R.J., MAEHLE, S., NILSEN, H, ROE, S.L., &
SPINNANGR, A. (1977). - Coarsening upward cycles in
the alluvium of Hornelen Basin (Devonian), Norway.
Sedimentary response to tectonic events. Bull. Amer.
Assoc. Petroleum Geol., 88, p. 1124-1134.
TEICHERT, C. (1958). - Concepts of facies. Bull. arner.
Assoc. Petroleum Geol., 42, p. 2718-2744.
VINCENT, P., GARTNER, J. & ATTALI, G. (1979). GEODIP - An approach to detailed dip determination
using correlation by pattern recognition. J. Petroleum
Technol., Feb. 1979, p. 232-240.
VISHER, G.S. (1965). - Use of vertical profile in environmental reconstruction. Bull. Amer. Assoc. Petroleum
Geol., 49, p. 41-61.
WALKER, R.G. (1975). - Generalized facies models
for resedimented conglomerates of turbidite association.
Bull. geol. SOC.Amer., 86, p. 737-748.
WALKER, R.G. (1976). - Facies Models. Turbidites
and associated coarse clastic deposits. Geoscience
Canada, 3, p. 25-36.
WALKER, R.G. (Ed.) (1979, 1984). - Facies Models.
1st and 2nd ed. Geoscience Canada, reprint series 1,
published by Geo% Assoc. Canada.
WALKER, R.G. & MUTTI, E. (1973). - Turbidite facies
and facies association. SEPM, Pacific Section, Short
Course, Anaheim.
WALTHER, J. (1893-4). - Einleitung in die Geologie als
historische Wisssenschaft. Fischer, Jena.
WEBER, K.J., & DAUKORU, E. (1975). - Petroleum
geology of the Niger delta. Trans. 9th World Petrol.
Congres, Tokyo, 2, p. 209-221.
WIDDICOMBE, R.E. & NOON, P. (1984). - Mulfiwell
FAClOLOG evaluation, Hartzog Draw field,Powder River
Basin, Wyoming. SPWLA, 25th Ann. Log. Symp. Trans.,
New Orleans.
WILLIAMS, P.F. & RUST, B.R. (1969). - The sedimentology of a braided river. J. sediment. Petrol., 39, 2, p.
649-679.
YAPAUDJIAN, L. (1972). - Une approche actualiste en
geologie sedimentaire (quelques donnees d'interpretation
des sequences de plateforme). In: Mem. BRGM, 77,
p.715-744.
Serralog Q 2003
WELL LOGGING AND DIAGENESIS
Introduction
Review of geological concepts
Definition
Diagenesis is defined as “all the chemical, physical,
and biologic changes undergone by a sediment after its
initial deposition, and during and after its lithification,
exclusive of surficial alteration (weathering) and metamorphism... It embraces those processes (such as compaction, cementation, reworking, authigenesis, replacement, crystallization, leaching, hydration, bacterial action,
and formation of concretions) that occur under conditions
of pressure (up to 7 kb) and temperature (maximum range
of 700°Cto 300°C) that are normal to the surficial or outer
part of the Earth’s crust; and it may include changes
occurring after lithification under the same conditions of
temperature and pressure.” (Bates & Jackson, 1980).
Several studies have been made of this phenomenon,
and the associated publications are listed in the bibliography. We now extract the some key points from these
publications.
‘Asdeposition continues, the lamina of sediment passes from the interface to successively lower positions and
enters a realm of greatur pressure, higher temperature,
and of changed chemical and biological conditions. These
new conditions promote the consolidation or lithification of
the sediment into a sedimentary rock” (ibid).
The diagenetic environment is characterized by the
original composition of the sediment and the original fluid
on which will act the temperature and pressure, therefore
the burial depth, and the possible circulation of fluids by
hydrodynamic or hydrothermic phenomena. Non-marine
environments are different from their marine analogues
due to the differences in fluids.
The original fluids occupying the pore space of the
rock at the moment of deposition are characterized by
their nature (water, air or gas), and possibly by their
content of salt and dissolved gases. The composition of
the fluids will be modified very rapidly by burial.
Interactions with the minerals and the action of bacteria
lead to a change in pH (hydrogen ion concentartion) and
in Eh (oxidation reduction potential) (Figs. 6-1 and 6-2).
The Diagenetic Environment
A “diagenetic environment is the environment of postdepositional change. It extends an indefinite distance
downward from the depositional interface” (Krumbein &
Sloss, 1963). In other words it may be defined as a certain volume of rock in which a series of transformations
takes place. “The nature of the depositional environment
and the rapidity of the postdepositional changes depend
upon the medium of deposition and the kind of sediment
being deposited‘ (ibid).
“The depositional interface represents an important
boundary condition that separates two different physiochemical regions. As a simple illustration, clay and silt
settle through sea water as a group of particles in aliquid
medium. When the Aparticles come to rest on the bottom,
they form a solid matrix with water-saturated pores. The
water has the same composition as the medium above,
but marked changes occur once it is sealed from free circulation by confinement in the pores” (ibid).
Serralog 0 2003
Figure 6-1- Diagram of changes in characteristics of marine bottom
mud below the depositional interface. Eh is in units of 0. I
(redrawn from data from Zobell; 1949,in Krumbein & Sloss, 1963).
239
Zone of intense dissolqtfi
Zone of Ininor dissolution
Zone of dolomitizatiau
Eh
Fresh water phreatic zone
PH
Zone of precipitation
Formation of meni=uS
and pendant calcite
cements
Formation of caliche crust
Intense solution near soil zone
Minor solution
Figure 6-2 - Diagram showing the limits on values of pH and Eh found
in natural environments, in particular in syndiagenetic, anadiagenetic
and epidiagenetic zones (based on Becking & Carrels)
(from Fairbridge, 1967).
Thereafter the diagenetic environment changes with
depth, but not necessarily in a uniform way, because
again grain size and composition, or the formation of new
(authigenetic) minerals, may create a permeability barrier,
preventing any further circulation of fluid. This is why the
diagenetic environment depends mainly on the original
facies, and consequently on the initial environment (Fig.
6-3).
a Zone
of active water
circulation and
significant precipitation
of rnanne cements
0 phdatic
Sta nant marine
zone with
Miqtization
and intr r a n k ?
little or no cementation cementzon
Factors Affecting Diagenesis
According to Wolf & Chilingarian (1976), there are
twelve factors which control the diagenetic process.
- geographic (climate --> humidity --> rainfall --> type
of terrestrial weathering --> surface water chemistry);
- geotectonism (rate of erosion and accumulation,
coastal morphology, emergence and subsidence, whether
eugeosynclinal or miogeosynclinal);
- geomorphologic position (basinal versus lagoonal
sediments, current velocity -. grain size, sorting, flushing
of sediments);
- geochemical factors in a regional sense (supersaline
versus marine water, volcanic fluids and gases);
- rate of sediment accumulation (halmyrolysis --> ion
transfer --> preservation of organic matter --> biochemical
zonation);
- initial composition of the sediments (aragonite versus
high-Mg and low-Mg calcite and isotopes and trace-elements content);
- grain size (content of organic matter --> number of
240
Zone of solution
near water table
0 Stagnant zone
8 Zone of active water
grain neomorphism
CtrculatiOn. Rapid
to calcite put little
neomorphism and
cementation by
cementation
equant calcite
Figure 6-3 - Schematic cross-sections showing diagenetic textures in
carbonates which have been developed in diagenetic environments
close to the surface (from Longman, 1980).
bacteria --> rates of diffusion);
- purity of the sediment (percentage of clay and organic matter --> basic exchange of clays altering interstitial
fluids);
- accessibility of rock framework to surface (cavity systems permit replacements);
- interstitial fluids and gases (composition, rate of flow,
and exchange of ions);
Serralog 0 2003
Diagenesis
- physico-chemical conditions (pH, Eh), partial pressures and C02 content);
- previous diagenetic history of the sediment (previous
expulsion of trace elements will determine subsequent
diagenesis).
These factors influence the local environments which,
in turn, influence the micro-environments. The factors
interact and overlap : the climate influences the geomorphology which in turn controls the grain size and hence
the type of bacteria, rate of diagenesis, pH and oxidation
reduction potentials, and finally the type of replacement
which takes place.
1 Chapter6 I
241
- Biochemical and Oraanic Processes
All organisms, and particularly bacteria, can modify the
pH and the oxidationheduction potential, Eh. The modification is particularly rapid within a few centimeters or tens
of centimeters of the surface (Zobell 1942), and produces
major changes within the sediments, such as reduction of
sulphates and conversion of organic matter (cf. Fig. 6-2).
In addition, there may be a reduction of grain size, mixing
of grains (bioturbation), corrosion of grains and creation of
gas bubbles.
Diagenetic Changes
The Fundamental Processes
There are three types of process involved.
- Phvsical Processes
The mechanical constraints of burial initially result in a
rearrangement of the grains of the sediment which tends
to produce compaction, and at a later stage may cause
fracturing.
Dessication will result in contraction, fissuring, penecontemporaneous internal deformation and internal
mechanical sedimentation.
Pressure and temperature will both increase with
depth and will modify the chemical equilibrium. They will
also cause rocks which are normally elastic to become
plastic or even viscous.
The combined effects of temperature and pressure will
also affect the solubility of the minerals, which may result
in either increased dissolution or increased precipitation
of salts in the pore space.
- Chemical Processes
The interactions between the solids and the fluids will
tend to establish a chemical equilibrium. This may result
in dissolution, alteration, oxidation or reduction. It may
also cause the growth of crystals (Fig. 6-4), or their replacement with new types.
The main factors controlling diagenesis are mineralogical composition, grain size, fluid content and organic
matter on the one hand, and temperature, pressure and
ambient chemical conditions on the other. As a result,
there are many different types of diagenetic changes.
Krumbein (1942) has listed some thirty of these. They
may be classified as summarized in Figure 6-5.
Lithification =d
-shale
includes
Sand
Sandstone
individual ~ ~ ~ , __+
d , , Conglomerate
limertone and
,
Processes sands, oozes
below
dolostone
Compaction
primarily
of muds
50-60% ofwater
Finely detailed
Structures
Recrystallization
of unstable
{I
minerals
ta
Dissolution of
more insoluble
minerals
Precipitation
of new minerals
or additions to
existing ones
New cryst@faces of
quartz preciptated on
rounded surface.,
New lcaolinite crystals
grown in pore space
1
_I
Figure 6-4 - On the left, SEM photograph showing quartz overgrowth.
On the right, thin section illustrating the quartz overgrowth.
Serralog 0 2003
Figure 6-5 - Changes in composition and texture resulting from diagenetic processes - Most of the changes tend to transform a soft and friable sediment into a hard, consolidated sedimentary rock
(from Press 8, Sieve6 1978).
241
- Rock components
According to Krumbein (1942), the raw materials of
diagenesis consist of organic or inorganic sediments and
of interstitial fluids. The sediments may be either allochthonous or autochthonous and will have been deposited
on the interface between the pre-existing formations and
the new depositional environment. The fluids will represent about 45 % of the volume of sands, 50 to 65 % of the
volume of silts, 80 to 90 % of the volume of argillaceous
muds, and 98 % of the volume of colloids at the moment
of deposition. Materials which are subsequently introduced or formed are included in the classification of "raw
materials".
The various minerals, materials and fluids making up
the sediment constitute a complex chemical mixture
which is generally not in an equilibrium state, and the
resulting interactions lead to a transformation of the sediment.
Both the size and composition of the minerals have a
bearing on the process. In fact, the percentage of certain
minerals depends on grain size (Fig. 6-6). Thus, phyllitic
minerals and organic matter are more abundant in the fine
than in the coarse fraction. Besides, some minerals are
more unstable or more soluble than others (aragonite,
pyroxenes, amphiboles, feldspars, micas), and will be
easily transformed or dissolved.
The size of the grains or crystals also controls both the
area of solid-fluid contact, and the permeability and
consequently the possibility of the fluids circulating. Grain
sorting will also control the location of the interactions.
The fluids will react according to their nature (water,
air, gas, ...) and their composition (salt and dissolved gas
content). We introduce the concept of diagenetic environment to describe how the fluids may evolve with time as a
result of hydrodynamic phenomena, or under the influence of temperature and pressure.
original void spaces, in other words, a reduction in the
initial porosity.
The amount of compaction depends on the initial
porosity and on the size, shape and sorting of the grains.
It also depends on the rate of sedimentation and time
period. Compaction is particularly marked in detrital sediments.
Compaction will be studied in detail in the following
chapter.
Cementation
This is one of the most common diagenetic phenomena. It is the deposition of minerals within the pore space.
The minerals may be derived from the sediment itself by
leaching and redeposition. They may also be derived from
salts dissolved in interstitial or circulating water, and may
be a single type or a mixture of several types.
Cementation may occur quickly or over a long period.
The most common cements are calcite, dolomite, silica,
clay minerals and less frequently anhydrite, halite, pyrite,
siderite or haematite (Fig. 6-7). Cementation results in a
reduction of porosity, and the quantity of cement cannot
exceed the initial porosity (in detrital deposits).
Figure 6-7 - On the left :thin section showing quartz cement (quartz
overgrowth) and calcite cement (in red). The pore space is in blue.
On the right :thin section showing early meniscus calcite cement
between rounded grains.
Recrystallization
This is the process by which the crystalline texture of
the rock changes as a result of the finer crystals dissolving and then growing again in different forms. The phenomenon is more common in chemical or biochemical
rocks than in detrital rocks and increases with burial.
Scale
Transformation
Figure 6-6 - Relationship between grain size and composition of the
detrital fraction in clastic silicate rocks (from Blatt et al., 7980).
Compaction
This is a mechanical rearrangement of grains under
the weight of sediments above during the burial process.
The result is a reduction in volume at the expense of the
242
This is the replacement of a mineral by its polymorph,
the commonest example being the transformation of aragonite to calcite.
Mineraloaical Redacement
In this case a new mineral replaces a previously existing one (Fig. 6-8). Included in this category are the pheSerralog 0 2003
is
nomena of dolomitization, phosphatization, pyritization
and silicification, along with the transformation of gypsum
into anhydrite and montmorillonite into illite. Dapples
(1962) gives a list of diagenetic reactions divided into
three stages (Table 6-1).
a) MUDSTONE
RHOMBOHEDRON
c) COMPOSITE CALCITE
RHOMBOHEDRON
(direct replacement origin)
d) PARTIALLY-LEACHED
DEDOLOMITE
RHOMBOHEDRON
e) RHOMBOHEDRALPORE
Redoxomorphic stage
(reversible reactions)
Locomorphic stage
Phyllomorphic stage
(replacement reactions) (unidirectional reactions)
Fe+2
aragonite by calcite
Fe+3
Hematite + calate
siderite
e
+ siderite + ferrodolomite
(cementation origin)
Figure 6-8 - Diagrams showing the diagenetic history of a limestone
(mudstone) (from Evamy, 1967).
carbonates by quartz
or chert
Hematite + clay minerals + quartz or chert by
silica +hlorite + greenalite carbonates
+ stilpnomelane (?)
Hematite +chlorite
chamosite (?)
Hematite + iliite
glauconite
+
+
"Bauxite" + silica
kaolinite
+
"clay minerals"-.chlorite
feldspars by carbonates
"clay minerals" + Fe*2 +
biotite
"clay minerals" +chert
chlorite
+
quartz, chert, clay minerals "clay minerals" +quartz +
by carbonates
sericite
opal by chert or quartz
silica solution
chert
kaolinite + glauconite or
illite + Mg+2+chlorite + K'
glauconite + feldspar
Diaspore or boehmite +
silica +clay minerals
illite or glauconite +
chlorite + muscovite
Iliaspore(?) + silica + K C
(glauconite)
kaolinite + illite +
glauconite + calcite + Mg+2
+ muscovite + biotite +
feldspar + dolomite +
chlorite
Bauxite + hematite +
silica (minor) =clay
minerals + pyrite
illite or glauconite +
calcite + Mg+2 +micas
+ feldspar + dolomite
f) RHOMBOHEDRALPORE
partially filled by calcite
g) COMPOSITE CALCITE
RHOMBOHEDRON
calcite by dolomite
"clay minerals"+ muscovite
or biotite
montmorillonite +chlorite
+ dolomite
+ clay minerals
druse
243
3
Table 6-1
Characteristic diagenetic reactions
(from Dapples, 1972).
Hematite + calcite + Mg +2
b) DOLOMITE
Chapter6
+
illite
Kaolinite + K +
Biotite
glauconite
Feldspar + clay minerals
+ chert
Glass G montmorillonite +
cheri
+
feldspar + sericite
plagioclase +chlorite + chert
Selective Leaching
Leaching is quite a widespread phenomenon, and is
often associated with other phenomena such as compaction (the solubility increases at the points of contact between grains with increasing pressure), recrystallization,
cementation and replacement. Selective leaching, however, must be considered separately. It affects only certain
constituents of the rock which have increased solubility
under certain conditions of temperature, pressure and
concentration of salt or dissolved gas in the interstitial
fluids. Of particular importance is the leaching associated
with acidic waters of meteoric origin which are rich in C02.
They can create vugs and even caverns in the rock which
can make a significant contribution to its porosity (Fig. 69).
Authiaenesis
This phenomenon results in the appearance of new
minerals, either by direct introduction or, more frequently,
by alteration of pre-existing minerals. It is very similar to
the process of cementation in terms of the end effect. An
example would be the formation of kaolinite by the alteration of feldspars.
Serralog 0 2003
Figure 6-9 - On the right :thin section showing oomoldic secondary
porosity. On the left : SEM photograph showing a vug.
Stvlolitization
This is a pressure-controlled solution phenomenon
which gives rise to stylolites (Fig. 6-10).
243
a
b
Figure 6-10 - a) Schematic of a stylolite. b) Schematic of a horizontal
stylolite with tension gash. The maximum principal strees axis is vertical. c) Photograph of a stylolitic surface in 30.d) Photograph of a stylolite seen on a plane surface.
The diagenetic phases
The numerous researchers in the field of diagenesis
have identified several phases in the process. Table 6-2,
drawn from a synthesis of the work of several authors by
Dunoyer de Segonzac (1968), attempts to unify the
various terminologies which have been applied to these
phases. We can confine our interests to the following
three main phases.
The first, which occurs early in the diagenetic process
is syndiagenesis (Bissell, 1959). This is characterized by
the presence of substantial amounts of interstitial water,
which is expelled very slowly, and by extreme variations
of pH and Eh. As Dunoyer de Segonzac pointed out, this
only affects a few tens of centimeters of sediment. It is
during this phase that the sediments are transformed into
rocks which are coherent and solid (lithification).
The second phase occurs late in the diagenetic cycle,
and is known as anadiagenesis (Fairbridge, 1967, Fig. 611). It is a phase of compaction and maturation which is
characterized by the expulsion and (usually) upward
migration of interstitial water and other fluids, such as oil
or gas, and by conditions of reduction. It takes effect over
several hundreds or thousands of meters of sediment
(Dunoyer de Segonzac), and generally results in a substantial reduction of porosity.
The third and final phase is epidiagenesis (Fairbridge,
1967). It is characterized by the modification of interstitial
waters as a result of the penetration and downward
migration of meteoric water, and by a reversion to conditions of oxidation. The phenomena of dissolution and of
the formation of "hard-ground'' mainly appear during this
phase. The zone affected is usually not very thick and
close to the surface.
These three phases more or less correspond to the
three periods and zones of diagenesis defined by
Choquette & Pray, 1970 (Figure 6-12)
Table 6-2
Various stages of diagenesis according to different authors (extracted from Dunoyer de Segonzac, 1968).
Stages
Authors
Van Hise
(1904)
Twenhofel
(1926,3930)
Krumbein
(1942,1947)
K. & Sloss
(1963)
Deposition
Deposition
of lithogenesis
Particles immobilized
in a sediment with a
Tectonic phenomena
place the sediments
V under conditions of
decomposition and
leaching or exposed
244
I
depositional
interface
katamophism weathering
(delithification)
Pettijohn
(1949,57)
I
Packham
Crook
(1960)
halmymlysis
I
weathering
Dapples
(1959)
Dapples
(1962)
I
I
halmymlysis
weathering
Williams
Turner
Gilbert
(1954)
weathering
Wolf
Conolly
(1965)
predepositional
I
I
halmymlysis initial stage
,
=depositional
1
Taylor
(1964)
preburial
syngenetic
Redoxo-
weathering
weathering
weathering
Serralog 0 2003
Next Page
Dianenesis
I
ChaDter 6
I
245
- The eogeneticperiod is the time interval between the
end of deposition and the start of effective burial. The
eogenetic zone is the interval of depth at or near the surface of sedimentation where the chemistry of the interstitial water is largely controlled by the surface environment
before any effective burial takes place (less than a few
tens of metres).
- The mesogenetic period is the post-depositional period from the start of burial to the start of the diagenetic
processes associated with erosion. The mesogenetic
zone is the depth interval within which the diagenetic processes associated with burial take place.
- The telogenefic period refers to the interval during
which rocks which have previously been buried are subjected to the diagenetic processes associated with subaquatic or subaerial erosion. The telogenetic zone is the
name applied to the corresponding depth interval.
Several diagenetic cycles incorporating some or all of
the above-mentioned phases may occur, each one leaving its marks and possibly obliterating those of preceding
cycles.
J
Figure 6-12- Schematic representation of the main areas and burial
zones in which porosity is created or modified
(from Choquette & Pray, 1970).
1
Figure 6-11- Idealized profile across a continental margin showing the
marine sedimentation sites and the three phases of diagenesis :(I)
diffusion during syndiagenesis; (2) movement of ascending liquid
during anadiagenesis; (3) movement of descending fluid during epidiagenesis (from Fairbridge, 1967).
Effects of diagenesis on the sediment properties
The sediment properties which were originally deposited will change as a result of physical, chemical, mineralogical, textural and structural changes associated with
the various diagenetic phenomena. The principal effects
are summarised in Table 6-3; among them are :
- Size and shape of particles : the grains may increase
in size and form crystals by growth or re-crystallization.
Alternatively, the grain size may decrease because of leaching.
- Mineralogical composition : transformation, replacement, cementation or authigenesis are the only means by
which the composition of the original sediment may change.
- Porosity: the majority of diagenetic processes, apart
from dissolution and occasionally replacement, involve a
reduction of porosity which may be quite substantial. This
reduction is referred to as poronecrosis. Conversely, the
Serralog 0 2003
phenomena of dolomitization and particularly of leaching
increase the porosity, sometimes quite visibly, by the formation of vugs and caverns. This effect is referred to as
porogenesis. However, the way in which the porosity
develops will depend on the type of sediment (detrital or
chemical : Table 6-4) and on the texture (Table 6-5).
- Permeability:precipitation of cement, recrystallization
and compaction will usually result in a substantial reduction of permeability along with a loss of porosity.
Dissolution and mineralogical replacement will only
increase the permeability if the newly-created pores are
connected to the existing pore system, and if the interconnecting channels are themselves enlarged.
- Sedimentary structure : the original sedimentary features can be obscured or deformed by certain of the diagenetic processes: recrystallization, compaction, the
action of organisms, leaching or expansion.
The need for diagenetic studies
Because of the ways in which diagenetic phenomena
modify the original sediment, it is important to be able to
evaluate the extent of these changes in order to construct
a reasonable model of the depositional environment and
to successfully identify the original facies.
In addition, by recognizing the various stages and
cycles of diagenesis, we are in a better position to reconstruct the geological history of a rock, and hence of the
sedimentary basin in which the rock has been deposited.
Diagenetic studies are generally performed on rock
samples by examination of thin sections and images
245
Previous Page
obtained by scanning electron microscopes.
Table 6-3
Relative effects of different diagenetic processes on the
properties of sediments (from Krumbein, 1942).
Property
Cornpaction
CemenRecrystation tallitation
Size
-
Shape and
roundness
-
Surfacetexture
-
Particleonentabon
7
-
Mineral
composition
-
ac
Porosity
Permeability
Color
x
X
x
xx
xx
m
xxx
X
xx
Replace- Differen- Authigenesis
rnent
tia!
solution
xxx
-
xx
xx
-
xx
xx
xx
X
XXX
xx
7
X
-
7
xxx
xx
xx
X
!a
X
x
xx
-
Aspect
xxx
Curhonute
mary porosity
in sediments
a%
Commonly 40-70%
Amount of
ultimate porosity in rocks
Commonly
half or more of
initial porosity;
I5-30% com-
Commonly none or only snuill
fraetton of initial porosity: 515% common in reservoir
facies
Typds) Of ptimary porosity
Almost exclu-
xx
xx
xx
Legend : x = small to moderate effect; x x = moderate to large effect; x x x = property most strongly affected by a given process; means a negligible effect; and
? indicates an unknown effect.
mon
Diagenetic studies are fundamentally based on :
tnterparticle commonly predominates. but intrapartidc
and other types are imponant
Almort exclusively primary
interparticle
Widely varied b u r of postdepositional modifications
S i z a of pores
Diameter and
throat sires
clovly related
to sedimentary
particle s i n
and sorting
Diameter and throat dzes cbmmonly show little relation to
sedimentary panicle uzc or
wrting
Shape of pores
Strong depcn&nce on partick s h a m
“negative” of
particles
Greatly varied. ranges from
strongly dependent “positive”
or “negative” of particks to
form completely independent
of s h a p a of depositional or
dmgenetic components
Uniformity of
size, shape. and
distribution
Commonly
fairly uniform
within homogeneour body
Variable. ranging from fairly
- analysis of rock-sample reaction to hydrochloric acid
(HCI);
- observations of polished surfaces using staining
techniques;
- analysis of thin sections;
- analysis of SEM photographs and Energy Dispersive
Analysis (EDAX);
- electron microprobe analysis;
- cathodoluminescence techniques.
Through these analyses the different phases of the
diagenetic history of a rock sample can be reconstructed
as previously illustrated by Figure 6-8 and by the following
figures.
Figure 6-13 (see next page) illustrates some of the diagenetic modifications that can occur into a sediment. As
previously indicated these modifications can occur either
in vadose (Fig. 6-14) or phreatic zones (Fig. 6-15) or induced by burial. Figure 6-16 indicates the influence of Mg
and Na on the morphology of carbonate crystals (Folk,
1974).
Staining techniques allow the distinction between carbonate minerals, aragonite from calcite by Feigel’s solution, dolomite from calcite by Alizarine Red-S in a 0.2%
HCI solution (Fig. 6-17), Mg calcite by use of Clayton yellow stain, and Fe calcite by use of a potassium ferricyanide stain in a weak HCI solution.
Figures 6-18 and 6-19 illustrate several diagenetic
phenomena that can be recognized in sedimentary rocks
by observation of thin sections.
Figures 6-4 (page 3) illustrate overgrowth of quartz
grain. Cathodoluminescence technique, allowing the
detection of secondary phase of cementation, is illustrated by Figure 6-20.
dvely interparticle
Typdr) of ultimate porosity
Diagenetic studies
Traditional methods
246
Sandstone
Commonly 25-
Amavnt of pri-
X
xx
X
xx
Figure 6-21 illustrates stylolitization seen on thin section. Secondary porosity development by selected leaching is illustrated by Figure 6-22 and geopetal fabric by
Figure 6-23.
Table 6-4
Comparison of porosities in sandstones
and carbonates (from Choquette & Pray, 1970).
uniform to extremely hetero-
geneous. even within body
made up of single rock type
Influence of
diagenesis
Minor; usually
minor reductionofprimary
porosity by
compaction
and cementatior
M a p r ; can create, obliterate.
or completely modify porosity;
cementation and solution important
Influence of
fracturing
Generally not
of m a p r imponance in
reservoir prop-
Of major importancc in m e r voir properties if present
erties
Visual evaluation of porosity
and perme
ability
Semiquantitative visual a t i -
mates commonly retatively eaay
Variable; rcmiquantiiative visual a t i m a t a ranse fromesay
to virtually impouible; inurument mcssuremcnts of masicy. permeability and capillary
pnwrrecommonly necded
Adequacy of
-re analysia
for nrcrvoir
evaluation
Core plugs of
I-in. diameter
commonly a&
quste for “matrix” porosity
Core plugs commonly inadequate;evenwhole cora(-3-in.
diameter) m a y k inadequate
for large p o r a
Permcabilityporosity interrelations
Relatively consiatent; commonly dcpendentpn panicle
dzeand wrting
Greatly varied; commonly in-
dependent of particle sire and
sorting
The traditional methods are the only ones that allow
the determination of the successive diagenetic phases
and periods which have modified the original properties of
Serralog 6 2003
Diagenesis
1
Chapter6
I
247
Table 6-5
Different types of porosity evolution during diagenesis (courtesy of Gulf Oil of Canada Ltd).
in matrix
dolostones
I
I
Preserved
sediment
porosity
No essentiel
changes in
sediment porosity
CompLction
pressure &
solution
I
Recrystallisation
& replacement
of matrix
I
I
I
leaching of
leaching of
leaching of
Arenite and rudite size carbonate particles
Fracture porosity can develop
in rocks
independent of composition
or texture
- with or without lime mud matrix
bioconstructed carbonate framework with
or without fine matrix
I
I
Cementation
I
I
Organic
cementation
I
Mineral
precipitation
I
Advanced.
dolomitization
-I
Porosity
Porosity
reduced or
eliminated
Porosity
eliminated
in matrix
mainly in
interfossil bridged
& intrafossil types
Porosity
reduced or
eliminated
I
Porosity reduces
reduced
or eliminated
mainly in
matrix
MUD-SUPPORTED SEDIMENTS
3
g
W
t
3f a
Intercrystalline
micropores & small
mesopores in
sucrose dolostone
Vugs in sucrose
or coarser
crystalline
dolostone
I
3
8
Dolomitization
2
n
I
-Iw
.v,
n ->
! &
I
Contemporaneous
solution of sand
or larger size
calcitic particles
in lime mud during
dolomitization
Vugs in dolostone
I
Leach ing of
incompletely or
non dolomitized
sand or larger
size carbonate
particles
Vuggy & moldic
porosity
in limestones
Leaching of more
soluble carbonate
particles
I
I
I
I
d
Serralog 0 2003
247
e
Figure 6-13 - The upper five figures illustrate schematically and successively from left to right: the original sediment, its cementation, its partial
dissolution, its dolomitization and its recrystallization (courtesy of B. Purser). The four lower figures illustrate how these situations can be detected
by observation of thin sections through microscope. The last figure explains the link between mold and vug.
A
Figure 6-14 - Diagrams illustrating the different types of cement generated in the vadose zone (courtesy of 6.Purser).
Figure 6-16 - Diagenetic realms and carbonate crystal morphology.
Three main realms exist: 1) Na+ and Mg++ both high as in marine
cement (beachrock, subtidal, reef cement...). 2) Na+ high and Mg++
low, as in the subsurface zone, where Mg++ has been removed
through trapping by clays and dolomite (mixing zone between connate
and fresh meteoric Waters). 3) Na+ and Mg++ both low, in the meteoric
zone (from Folk, 1974)
-l
Figure 6-15 - Type of cementation as a function of the depositional
environment (courtesy of B. Purser).
the sediment since its deposition. In addition, they also
allow a very precise sequence stratigraphy determination
that cannot be dtected by others techniques.
I
Figure 6-17 - Staining of a thin section. Calcite cement is in red.
240
Serralog 0 2003
Diagenesis
ChaDter6
1
249
I
Figure 6-18 - On the left, illustration of meniscus cement between each
oolitic grain. On the right, illustration of microstalactitic cement
developped in the vadose zone. Porosity is in black.
w
Figure 6-22 - Thin section illustrating oomoldic secondary porosity
(in purple).
I
Figure 6-19 - On this thin section one can recognize two phases of
cementation. The first one (in white) is composed of sparry calcite
crystals, the second is composed of ankeritic dolomite which has filled
the pore space. This later deposit is associated with an unconformity.
Fiigure 6-23 Illustration of
geopetal sediment or vadose
silt (micritic
mud) filling the
lower part of a
gastropod. The
upper part is
filled by sparry
calcite.
The calcite stability is a function of both salinity and
Mg/Ca ratio in water as illustrated by Figure 6-24.
3
Figure 6-20 - Cathodoluminescence technique (right image) allows the
detection of quartz overgrowth which was not detected through thin
section analysis alone (left image).
Y
Figure 6-21 - On the left, theoretical diagram of a stylolite. On the
right, a sty/olite is well detected on this thin section.
Figure 6-24 - Calcite and dolomite stability as a function of the
salinity and the Mg/Ca ratio (from Folk & Land, 1974)..
Serralog 0 2003
249
Well logging approach
If a detailed analysis of rock samples can provide a
reconstruction of the diagenetic changes which have
taken place and hence the stages of diagenesis, it is not
then necessary to make use of well logs to analyze the
diagenetic history of the rock. Well logs analyze the rock
in its present state, and hence both the properties which
were determined by the original depositional environment,
and those which were the result of diagenesis. At best,
thus, they will only give a picture of the final state of the
rock which is the end result of a chain of diagenetic processes.
Having said that, it is only necessary, during the detailed analysis of the logs, to bear this in mind and to keep
an eye out for signs which may indicate certain diagenetic phenomena. Clearly, from what has already been said,
this approach must be achieved taking into account the
type of rock studied.
Diagenesis in detrital sequences
The type and extent of diagenetic changes depend largely on the maturity of the rock, that is, on its content of
stable minerals and on its textural properties (Fig. 6-25).
Rounded
Mature
Figure 6-26 - Diagram showing changes in the relative mineralogical
maturity of sandstones (from Weller, 1960, in :Chilingarian & Wolf,
Compaction of Coarse-grained Sediments, Developments in
Sedimentology, Elsevier, Amsterdam).
Table 6-6
Some reactions involving clay minerals during
the diagenesis of sandstones
(from Pettijohn e t a / , 1972).
Feldspar
Components added to (+)
or subtracted to (-)
- (K+, S O 2 )
Kaolinite
+ (K+, S O 2 )
Mite
Kaolinite
+ K+, - H2O
Muscovite
Precursor
Clay mineral formed
Kaolinite
Montmorillonite
+ K+, - (Si02, H 2 0 , Na+, Ca++, M g ++,Fe++..) Mite
Montrnorillonite
+ (Fe++, Mg++),- ( 3 0 2 , H 2 0 , Na+, Ca++) Chlorite
Volcanic glass
- (Na+, K+, Ca++)
+ (Fe++,Fe+++), - (
K', A1203)
lllite
Porous space
+ H20,
+ (A1203, Si02, H 2 0 )
Montmorillonite
Glauconite
Kaolinite
Pressure-controlled solution
Figure 6-25 - Diagram illustrating Folk's textural maturity concept (from
Weller, 1960, in :Chilingarian & Wolf, Compaction of Coarse-grained
Sediments, Developments in Sedimentology, Elsevier, Amsterdam),
For example, the constituent minerals of a quartz-arenite will react differently to those of an arkose or a graywacke to burial and to the subsequent diagenetic processes (Figure 6-26 and Table 6-6). Consequently, the recognition of the mineral composition of a rock will inform on
the possible development of cement linked to unstable
mineral alteration.
Compaction
This is one of the major diagenetic phenomena in
detrital sequences, regardless of composition. It is generally accompanied by secondary effects. In view of its
importance, the following chapter is devoted entirely to
the study of compaction.
250
The underlying concept of this phenomenon is that the
high pressures developed between the points of contact
of the grains result in increased solubility. This leads to
preferential leaching and this is further encouraged by the
presence of a film of clay. Fuchtbauer (1967) suggested
that pressure-controlled solution appears early, and is the
reason for loss of porosity in sands. Sippel (1968), however, demonstrated by cathodoluminescence technique
that detrital grains are not involved, and that authigenic
crystal growths appear in sands at concavo-convex or
suture contacts (cf. Fig. 6-20).
Regardless of its importance, pressure-controlled
solution is not the only cause of cementation.Alteration of
feldspars and other silicates by meteoric water produces
dissolved silica which may be re-precipitated in the sand,
or may contribute to the growth of quartz crystals. The
water itself may be supersaturated in dissolved silica.
This phenomenon is not distinguishable from silica
cementation by well logs, because the only detectable
result of both phenomena is a reduction in porosity.
Serralog 0 2003
Diagenesis
Cementation
When analysing logs from the diagenetic point of view,
it is necessary to establish the type of cement involved
and the degree of cementation. Sands which are composed of quartz grains (orthoquartzite) are recognised by
their very low radioactivity, due to the almost complete
absence of potassium and generally low content of thorium and uranium, the exact amount depending on the
presence of heavy, stable minerals such as zircon or
monazite. These sands have a cement which may be
composed of quartz, calcite, haematite or, less frequently,
of authigenic clay minerals such as kaolinite or dickite.
Arkoses and graywackes, which have a relatively high
potassium content and hence moderate to strong radioactivity, usually have a cement of calcite or haematite, or
of authigenic clay such as kaolinite, chlorite, montmorillonite or illite. Siliceous cement is rare because alkaline and
alkaline-earth cations from the alteration of feldspars and
from mafic minerals or volcanic grains combine with silica
to form authigenic clays or zeolites.
One of the most reliable means of evaluating the type
of cement is to use the litho-density (LDT)-neutron combination and the corresponding crossplot (Figure 6-27).
Orthoquartzites with a siliceous cement are clustered
around the point Q (quartz). At the same time, the density vs. neutron-hydrogen index plot reveals the loss of
porosity with reference to the general compaction gradient of sands, and from that, the quantity of cement
(according to Pettijohn, 1963 and Pettijohn et a/., 1972).
Before concluding that the cement is siliceous, we must
be sure that the decrease in porosity is not due to poor
sorting.
Chapter 6
1 251
distance of the point from Q to the length of the line Q-C.
In the same way, the points for a dolomitic cement fall
around a line joining Q and D (dolomite). If the cement is
a mixture of calcite and dolomite, the points fall between
the lines Q-C and Q-D (Fig. 6-28).
Figure 6-28a - Crossplot RHO6 vs PHIN with PEF on the Z-axis. The
very porous sand (30% porosity) shows a decreasing porosity with
displacement of the points towards the dolomite pole (arrow). A development of calcite would be marked by an increase in Pe value.
n
6
.
Figure 6-28b - Crossplot RHOMA vs UMA with THOR on the Z-axis.
One can observe trends towards opal and microcline with low thorium
values and trends towards dolomite and clay, but not towards calcite.
tn
Y
E
Q
Uma ( b / c m 3 )
Figure 6-27 (pmaja vs (Uma)a crossplot for determining the mineralogy of rocks and the cement type.
In the case of a calcite cement, the points are grouped
around a line joining the points Q (quartz) and C (calcite).
The percentage of cement is the-, given by the ratio of the
Serralog 0 2003
Certain sands have a halitic cement. Thus, in theory,
the points should fall around a line joining Q and the halite point H. However, if the well has been drilled with a mud
which is not salt-saturated, the halitic cement will dissolve, and the logging tools will not detect it (Figure 6-29).
Halitic cements are usually found near salt formations.
The Pe, which is less sensitive to porosity, may show
values which are consistently higher than those of quartz
(1.8 b/e), and this distinguishes this case from a gas sand.
The Schlumberger TDT tool is useful in this-situation
because of its sensitivity to chlorine (Figure 6-30).
Pyrite or haematite cements will displace the points
towards high values of ,U
,
and pma. If the quantities are
high enough and the cement is more or less in continuous
25 1
contact, the resistivity will be very low since pyrite and
haematite are both conductive.
For arkoses or graywackes, the points fall in the area
AI-Q-An-C-F for a calcareous cement. These rocks may
be recognised by plotting the potassium percentage on
the Z-axis.
CPI WITHOUT TDT
CPI WITH TDT
-a
I
v)
s
-J
Figure 6-29 - pb vs @J crossplot showing an example of a sandstone
having a halitic cement (courtesy of Schlumberger).
Authiaenesis
The formation of authigenic minerals in terrigenous
detrital sequences will depend on the textural and chemical maturity of the rock, on the types of fluid and hence on
the hydrodynamic conditions and on the compaction
(simultaneous action of temperature and pressure).
The formation of authigenic minerals in orthoquartzites
is usually limited to the precipitation of "books" of kaolinite (Figure 6-31). These may be recognised on the same
crossplot. The Pe value of kaolinite is close to that of
quartz (1.8 b/e), but its content of hydrogen ions gives it a
value of (P,,,~)~ and (U,,,a)a which allow them to be distinguished (cf. Fig. 6-27). Thus, the points are grouped
around the Q-K line.
It is difficult to distinguish authigenic kaolinite from kaolinite of detrital origin. However, it is reasonable to assume that authigenic kaolinite is purer, and that, as a result,
it contains practically no potassium, and even less thorium and that it therefore has very low radioactivity.
In immature sequences (arkoses and graywackes),
authigenic minerals are common. The most common are
the clay minerals (illite, montmorillonite, chlorite, etc.),
less frequently potassium or plagioclase feldspars. It is
not usually possible to distinguish between these authigenic minerals and their detrital counterparts from well logs.
The presence of authigenic minerals will have a considerable effect on permeability, but this will vary according to
how the clays are distributed. The distribution in its turn
often depends on the type of clay mineral (Figure 6-32
and Table 6-7).
252
Figure 6-30 - Sandstone with a halitic cement. Comparison between
two CPls run with and without the TDT log (courtesy of Schlumberger).
Nodules
The analysis of the microresistivity curves from dipmeter logs, combined with the dips, allows identification of
resistive events which only appear on a few curves
(Figure 6-33), more rarely on all 4 or 8 curves. In this last
case the thickness of the resistive events varies from one
Serralog 0 2003
Diagenesis
curve to another, and the dips computed at the bottom
and at the top of the event will be different (Figure 6-35).
b - Pore lining by grain coating
-
c Pore bridging by fiber crystals
Schematic
SEM microphotographs
Chapter 6
253
When these events appear suddenly in a homogeneous
sandstone, they correspond either to cemented balls, formed by dissolution of shell fragments and reprecipitation
of the released calcite in the surrounding pore space by
radiating diffusion, or to isolated pebbles in a sandy
matrix, or to anhydritic nodules. The choice between
these hypotheses can be made by checking the nature of
the previous deposit or the type of the environment. If a
grain size evolution can be detected from the logs, or if
evidence of grain supported conglomeratesexists (cf. Fig.
3-56), or if we are in an alluvial fan, a channel bed load, a
mass flow deposit, or a glacial environment, the pebble
hypothesis is the most likely. The cemented balls hypothesis should be prefered in a deltaic or mid fan turbidite
environment because it is impossible to explain how one
pebble could be transported and deposited in a region
where the energy was such that it generated a homogeneous sand (Fig. 6-36).
Anhydritic nodules is the most likely hypothesis in the
interdune sabkha deposits of an aeolian environment.
They will be also characterized by a very high resistivity
(> 500 ohm-m).
Cemented or anhydrite nodules are well detected on
images of the borehole wall. They correspond to isolated
white spots in a more conductive background (Fig. 6-37).
Pyrite crystals
Pyrite crystals, formed by diagenesis by reduction of
sulfates, are easily detected by the Formation
Microscanner tool. They are characterized by very
conductive peaks (Figure 6-38) on the resistivity curves,
Table 6-7
Characteristics of authigenic clays
(from Wilson & Pittman, 1977).
Figure 6-31 - Distributionmode of authigenic clays
(from Neasham, 1977).
Relationship
to sand size
detrital grains
Thickness of coating
or long dimension of
aggregates (microns)
pore filling
2 - 2500 (generally 2-20)
vermicule
pore filling
10 - 2500 (generally 20-200
sheet
pore filling
Morphology of
individual flakes
Form of aggregates
Kaolinite
pseudohexagonal
staked plates (books)
Dickite
pseudohexagonal
pseudohexagonal
Clay mineral typt
Chlorite
pseudohexagonal
surled equidimensional with rounded
2dges
squidimensional
Nith angular or
lobate edges
'an-shaped fibrous
iundles
lllite
rregular with
?longatespines
Smectite
lot recognizable
Mixed-layer
smectite-illite
subsequant with
stubby spines
Serralog 0 2003
plates ( 2-dimensional
card house)
honeycomb
pore lining
pore lining
0.1 - 1
Special features
Flakes notched or embayed
(twinned?)
Flakes notched or embayed
(twinned?)
Flakes notched or embayed
(twinned?)
2-10
2-10
rosette or fan
pore lining &
pore filling
4-150
(generally 4 - 20)
cabbage head
pore lining &
pore filling
8 - 40
pore lining
0.1 - 10
bridging between sand grains
pore lining
2-12
bridging between sand grains
2-12
bridging between sand grains
sheet
wrinkled sheet or
honeycomb
imbricated sheet
to ragged honeycomb
pore lining
253
or by dark black spots on the image display, these spots
reflecting sometimes the shape of the crystals.
4a
Figure 6-34 - Resistive events appearing on a few curves.
They correspond to cemented balls or lenses.
Figure 6-32 - On the left: influences on porosity and permeability of the
distribution type of authigenic clay minerals. Data from Neasham’s
study of 14 sandstones (from Neasham, 1977).
On the right: effect of type and distribution of authigenic clay onpermeability. The porosity is unchanged, in spite of a changein permeability of 4 orders of magnitude in the extreme cases
(from Stadler, 1973; in Blatt et al., 1980).
LOCDIP results
Figure 6-35 - One cemented ball crossed by the well. Observe the
change in thickness from pad to pad. The core photograph confirms
the presence of the ball.
P d3
PPd 1
Figure 6-33 - Effects of detrital and authigenic clays on petrophysical
characteristics of reservoir.
EPidiaaenesis
As we have already seen, this corresponds to the
development of secondary porosity by leaching, and is
detected by comparing the porosity measurement of the
sonic tool with that obtained from the density-neutron
combination. This occurs in detrital sediments in which
were present carbonate fragments (bioclasts, oolites...).
Diagenesis in carbonate sequences
The diagenetic phenomena affecting carbonates are
appreciably different from those affecting detrital sequences. Thus, compaction is of minor importance because
other diagenetic phenomena may have already affected
the rock and made it more consolidated even before the
completion of the burial process.
254
Figure 6-36 - Calcite nodules (white or yellow spots) in a deltaic (left)
and turbiditic (right) sandstone reservoir. The volume of these nonporous nodules , calculated from the left FMS image, corresponds to
12% of the total sandstone. Thus, the porosity of the sandstone itself,
will be higher than the porosity value determined from the neutron-density measurement. The permeability estimation must be evaluated from
the porosity of the sandstone, not from the bulk volume porosity.
(courtesy of Schlumberger).
Serralog 0 2003
Next Page
Diagenesis
I Chapter6 1
255
The points will plot very close to the calcite point, C, on a
(P,.,,~)~vs (U,.,,Ja crossplot (Figure 6-39), except in the
case of aragonite, for which the points fall towards the
bottom of the crossplot.
Figure 6-37 - Anhydrite growth in a sandy dolostone and evaluation of
their density and the area covered by these anhydrite crystals
(from Delhomme & Motet, 7993).
LlTHO
SYNDIP
DESCRIPTION
TRACES
IMAGE
msisnvmINCREASE
Figure 6-39 - Several Z-plots (thorium on the Z axis) for mineral and
cement determination.
~
the superposition of points on the limesUpper left: pb vs I P Observe
tone and dolostone lines respectively with the change in porosity indicating development of cement of same type except at the left corner
where a dolomitic cement appears in the limestone.
Upper right: Pb vs Pe. Observe the grouping of points along the limestone and dolostone lines.
Lower left: Pe vs potassium. The points evolve from the dolomite position to the calcite position with few points moving to the right with a
thorium increase indicating the influence of clay mineral (illite).
Lower right: (n,,,J, vs (Umala. The points are grouped around the calcite and dolomite position with some influence of clay moving the
points towards a lower position.
Neomorphism
Figure 6-38 - Pyrite crystals in a shale. They are characterized on the
Formation Microscanner curves by very conductive peaks, and on the
images by very black spots showing the shape of the crystals
(courtesy of Schlumberger).
Crystallization
This is the phenomenon by which primary pore space
is filled with crystals of calcite (sparite) or aragonite by
precipitation or drusy growth. This is accompanied by a
loss of porosity which is easily detected from the porosity
logs, assuming the type of mineralogy does not change.
Serralog 0 2003
It corresponds to a "transformation between one mineral and itself or a polymorph whether the new crystals are
larger or smaller or simply differ in shape from the previous ones, or represent a new mineral species" (Bates &
Jackson, 1980).
A first type of transformation corresponds to the change of aragonite to calcite. It is accompanied by a reduction of density (2.94 g/cm3 for aragonite against 2.71 for
calcite) and a loss of porosity. In theory it should be very
easy to detect from the logs. This transformation always
occurs early in the cycle, and the majority of ancient
sequences have been subjected to it.
Dolomitization (replacement of calcite with dolomite)
belongs to the second type of transformation. It brings
about a 13% reduction in volume, and a corresponding
increase in porosity if other phenomena do not interfere.
The log matrix density increases to a value of 2.87 g/cm3.
It is this parameter above all which allows this diagenetic
phenomenon to be detected. Crossplots of density-neu255
Previous Page
tron, M vs N, or (p),
vs (U,),
provide the best means
of detection (Fig. 6-40). (p,.,,,), and (Urn,), vary with the
intensity of dolomitization, which may itself vary from one
point to another.
(SPl)rel= (+N-D - +s) 1+N-D
Figure 6-41 - Crossplot M vs N showing how points are shifted as a
result of secondary porosity (courtesy of Schlumberger).
A
Figure 6-40 - Other example of 2-plots (GR on the Z axis), density vs
on the upper right,
neutron on the upper left and (p,,,,), vs
sonic transit time vs neutron on the lower left and porosity from
neutron-density vs sonic transit time on the lower right, showing the
loss of porosity of limestone and dolostone and the trend towards
anhydrite. Limestones are slightly radioactive due to glauconitic pellets.
Recrystallization cannot be detected from the logs
unless accompanied by a change in mineralogy or porosity.
Dedolomitization is the inverse transformation of dolomite into calcite, and so leads to a reduction in porosity
and a decrease in density. This process cannot be distinguished from cementation by calcite using only the logs.
Selective Leaching
The method for detecting this phenomenon with well
logs is well known. It involves comparing the porosity derived from the density-neutron combination with that obtained from the sonic. The sonic tool detects the arrival of
the fastest acoustic wave which is the one corresponding
to the most direct path, that is, the one which avoids the
vugs and caverns which slow down the waves.
The crossplot of Figure 6-41 illustrates the method.
The degree of leaching is measured by the secondary
porosity index, SPI, given by the following relation:
SPI = +N-D - +s
or by the relative secondary porosity index, given by
the following relation:
256
The development of secondary porosity has an effect
on the m parameter of the Archie’s equation, especially if
the pores are not connected. This requires a continuous
computation of the m exponent. This is achieved taking
into account the measurement of the transit time of an
electromagnetic wave. In that case one can write the following equation:
tpl
= tpmf.($ND.Sxo+ ($ND(l-Sxo)tphc
+
(l-@ND)tpma
in which:
tp,= transit time measured by the electromagnetic propagation tool,
tpmf= transit time in mud filtrate, generally close to 25
or 30 nanos/m,
+ND = porosity computed from the neutron and density
tphc = transit time in hydrocarbon (3.3nanodm in gas,
4.7-5.2 in oil),
tpma= transit time in rock solid (matrix), close to 8.7 for
dolostone and between 9.1 and 10.2 in limestone.
S,, = water saturation in the invaded zone.
The later parameter is the only unknown. It can be
extracted from the previous equation and inserted in the
Archie’s equation of saturation in the invaded zone:
s,
=@7iFG
in which:
Rmf= resistivity of the mud filtrate,
R,, = resistivity measured by a microresistivity device,
a = a parameter generally assumed to be equal to 1.
The only unknown from this equation is the m exponent which can be computed continuously and reproduced and compared to the secondary porosity computed
from the sonic and neutron-density combination (Fig. 642).
Serralog 0 2003
Chapter6
Variable m
i
257
Constant m
Figure 6-43 - The saturation computation using a constant m value
indicates gas in the lower part of the reservoir. Using a variable m this
lower part is water bearing. This was confirmed by the test of the two
zones, dry gas was produced in zone A, water in zone B
(Schlumberger, Middle East Well Evaluation Review, 2, 1987).
Figure 6-42 - Results of a quantitative interpretation
which displays the secondary porosity and the value of the m exponent
of the Archie's equation. Interconnecting vuggy porosity is identified by
the divergence of secondary porosity computed from sonic measurement and the Archie's m exponent curves. Above the two curves run
together, indicating isolated vugs.
(Schlumberger, Middle East Well Evaluation Review, 10, 1991).
This variation of the rn exponent has a big influence on
the saturation computation in the virgin zone as illustrated
by Figure 6-43.
Selective leaching and vugs can also be recognized
on dipmeters and micro-resistivity image tools. Vugs correspond to conductive peaks with irregular shapes, which
are not easily correlated even from button to button and
all the more so from pad to pad (Fig. 6-44). On the microresistivity images vugs are represented by dark gray
spots (Fig. 6-45) of which the shape, the frequency and
the connectivity can be analyzed and estimated(Figs. 646).
Stylolitization
This is a pressure-controlled solution phenomenon.
Stylolites are characterized by very fine joints of irregular,
sawtooth form, of variable height and interpenetrating
(Fig. 6-47). These joints often contain a concentration of
insoluble materials (clays, iron oxides, organic matter,
coal, lignite fragments). There is a loss of porosity due to
reprecipitation of dissolved calcite in the surrounding pore
space.
Serralog 0 2003
Pad 1
Pad 2
Pad 3
Pad 4
Figure 6-44 - Typical dipmeter resistivity curves in a vuggy carbonate.
Vugs are characterized by very thin and n a m w conductive spikes.
Stylolitization can be detected through well logs by a
sudden large drop in porosity and a simultaneous increase in radioactivity (Figure 6-48) often due to enrichment in
uranium. This phenomenon is most frequently encountered in carbonate sequences, but may also be found in
compacted sequences of terrigenous clastics.
In favorable cases (sufficient thickness, or resistivity
contrast between stylolites and surrounding beds), stylolites are easily recognized on the micro-resistivity images
as illustrated by the examples of the Figure 6-27. By stu257
dying them it should be possible to determine their frequency, and the axis of maximum stress.
Orlentatlon
SIZE ( m d ) (lin)
Figure 6-466 - Other example of analysis of the mold population on
iamges in size, area and density (from Herron et al., 1994).
Figure 6-45 - On the left: example of Formation Micro Imager (FMI)
images showing oomoldic econdary porosity in limestone.
On the right: extraction of the vugs by thresholding
of the microresistivity FMS curves
(courtesy of Schlumberger).
Original FMS Curves of vug density Boundaries
size and conne&ty
ofthe vugs
image
c%2%
zone
I
Connectiilty
of the W S
Figure 6-47 - Examples of stylolites easily recognized on
microresistivity images (courtesy of Schlumberger).
Figure 6-46a - Example of image analysis showing the extraction
of conductive events, analysis of their connectivity and estimation of
their density and area
(Delhomme, 1992).
Cherts and anhvdritic nodules
When very resistive events appear in a homogeneous
and porous limestone (chalk) characterized by flat curves
without variations, chert nodules seem to be the most likely hypothesis. When these events are observed in a
heterogeneous dolomitic, or better, anhydritic limestone,
they may correspond to anhydritic nodules generated in
the supratidal subenvironment. The micro-resistivity
image tools enable recognition of their shape and their
fre&ency. They correspond to very resistive peaks (R
500 ohm-m), and are seen as white spots with irregular
shapes on the images (Figure 6-49).
’
258
Figure 6-48 - Example of stylolites detected by gamma ray and uranium measured on core lfrom Hassan et al.. 1976).
Serralog 0 2003
Diagenesis
Chapter6
I 259
1
Figure 6-49 - Example of anhydritic nodules detected on the
2-pads Formation Microscanner images, confirmed by core photograph (courtesy of Schlumberger).
Hard-Ground
-I
Figure 6-51 - Oysters fixed on hard-ground in the Dogger formation
(Massangis quarv, Paris Basin, France).
It corresponds to “ a zone at the sea bottom, usually a
few cm thick, the sediment of which is lithified to form a
hardened surface, often encrusted, discolored, case-hardened, bored, and solution-ridden” (Bates & Jackson,
1980). It is often rich in oxides of iron and manganese.
(Figs. 6-50 and 6-51). It indicates a pause in sedimentation. It is frequently overlaid with a fine intraformational
conglomerate, rich in phosphate debris and glauconite in
a marly matrix. It appears on the logs as a very narrow
peak of very high density and resistivity which is best
seen by the micro-devices (MLL, ML and HDT, SHDT or
images), with a radioactive peak just above (Fig. 6-52).
Figure 6-50 - A “hard-ground horizon in a chalk
(from Selley, 1976).
Diagenesis in volcaniclastic sequences
The basic diagenetic reactions observed in volcaniclastic sequences are the transformation of glass into
clays (montrnorillonite) or zeolites. In addition, plagioclases may be altered to give clays, and the zeolites themselves may be transformed into other zeolites or into albite. The reactions which give rise to these diagenetic minerals are hydration, carbonization and dehydration.
In the absence of in-depth studies of logs in sequences of this type, it is not possible to indicate methods of
detection.
Serralog 0 2003
Figure 6-52 - Example of hard-ground detection using well logs.
Obsefve the thin resistivity peak at 1740 m, and the small radioactive
peak just above it. It is the equivalent of the hardground illustrated by
Figure 6-51 (from Serra, 1973).
259
Diagenesis in shales
As previously indicated clay minerals can evolve into
other minerals due to interaction between minerals and
fluids by additon or substraction of ions and influence of
temperature and pressure (Table 6-6). The detection of
this type of diagenesis requires the determination of the
depositional environment in order to evaluate the original
clay mineral type. Continental clayey deposits are characterized by irregular log responses due to the depositional process. On the contrary, marine clayey deposits
show generally a homogeneous log response.
ce of organic matter uranium is irreversibly adsorbed from
uranyl solutions in the presence of bacteria and humic
fractions. Organic matter and acidic pH are favourable to
convert U022+into the insoluble quadrivalent ion U 0 2 . In
acidic pH environments, humic and fulvic acids, ethers,
alcohol, aldehydes, favour the precipitation of uranium by
the reduction of U6+ to U4+,forming urano-organic complexes or chelates. Clays also encourage the formation of
schoepite by hydrolysis of the uranyl ions (Fig. 6-54).
Diagenesis of organic material
Organic material present in sediments evolves as a
function of the conditions of oxido-reduction into the depositional environment. Oxidation conditions transform the
organic material into C02. In reducing conditions, the
organic material is transformed and can become a source
rock which can generate biogenic methane, oil, bitumen
or thermal gas as a function of its maturation under temperature and pressure influences (Fig. 6-53).
mDRocARBoN
OENaUTlOtl INTENSITY
J
Figure 6-54 - Diagramatic sketch showing possible association and
time of emplacement of uranium with common constituents of marine
black shales. Uranium is represented by black squares
(from Swanson, 1960).
Several authors, Beers & Goodman, (1944) (Fig. 6-
55),Russel (1945), Swanson (1960) (Fig. 6-56), Hassan
(1973), Supernaw et a/. (1 978) (Fig. 6-57) have observed
a strong correlation between uranium and organic material.
Figure 6-53 - Organic material maturation and formation of biogenic
methane, oil and dry gas (from Sellex 1988).
"Transition zone"
The evaluation of the potential of source rock can be
achieved through the well log analysis.
L
Carboh (%)
Reducing conditions generate transformation of sulphates into sulphides (i.e. pyrite). They are also favorable
to preservation of uranium which will precipitate or accumulate in sediments under certains conditions. In presen260
0
Figure 6-55 - Relation between uranium and organic carbon in sedimentary rocks (from Beers & Goodman, 1944).
Serralog 0 2003
Figure 6-56 - On the left: diagram showing possible relation of uranium
content to total organic matter as controlled by the proporfion of humic
and sapropelic material making up the organic matter. On the right: oil
yield of a marine black shale as a function of the total organic matter
(from Swanson, 1960).
So, after calibration with core data, it is possible to
evaluate the organic carbon content of a source rock from
its uranium content and from that its hydrocarbon potential if the nature of the organic material is known (Fig. 658).
Organic carbon (%)
Figure 6-59 (next page) illustrates a formation very rich
in uranium shown by the spectrometry of the natural
radioactivity. It may correspond to a potential source rock
which could correspond to a mixed layer illite-montmorillonite taking into account the potassium and thorium
content.
Figure 6-57 - Relation between uranium (expressed as a
uranium/potassium ratio) and organic carbon
(from Supernaw et al., 1978).
Figure 6-58 - Sketch showing theoretical distribution of humic and sapropelic materials in a shallow sea in which black muds are accumulating, and
the estimated uranium content and oil yield of the resulting black shale. Increase of total organic matter seaward due chiefly to seaward decrease
in amount of detrital sediment; proportional increase in sapropelic matter seaward due to decrease of humic /and-plant debris and predominance of
planktonic matter (from Swanson, 1960).
Other attempts to evaluate the source-rock potential
include the measurement of the carbon-oxygen ratio
(C/O) by spectrometry of the gamma rays induced by
inelastic collisions of fast neutrons (Fig. 6-60 next page).
This measurement requires either a stationary run or
several runs with very low recording speed.
Serralog 0 2003
261
NATURAL GAMMA RAY SPECTROSCOPY DATA
Figure 6-59 - Example of fromation very rich in uranium as shown by
the natural gamma ray spectrometry data. It may correspond to a
potential source rock (courtesy of Schlumberger).
A
5
5
0
Total Organic Carbon (%)
Figure 6-60 - Evaluation of the organic carbon content through the
measurement of the carbon oxygen ratio (from Herron et al., 1988).
262
References and Bibliography
BATES, R.L., & JACKSON, J.A., eds. (1980). Glossary of Geology. 2nd ed., Amer. Geol. Institute, Falls
Church, Virginia, U.S.A.
BATHURST, R.G.C. (1976). - Carbonate Sediments
and their Diagenesis. 2nd ed. Developments in
Sedimentology, 12, Elsevier, Amsterdam.
BEERS, R.F., & GOODMAN, C. (1944). - Distribution
of radioactivity in ancient sediments. Bull. Geol. SOC.
America, 55.
BISSEL, H.J. (1959). - Silica in sediments of the Upper
Paleozoic of the Cordilleran area. In IRELAND, H.A., ed.,
Silica in Sediments - A symposium. Tulsa, SEPM special
publication 7, p. 150-185.
BLATT, H. (1979). - Diagenetic processes in
Sandstones. SEPM, Special Paper 26.
BLATT, H., MIDDLETON, G., & MURRAY, R. (1980). Origin of Sedimentary Rocks. 2nd ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
CAROZZI, A.V. (Ed.) (1975). - Sedimentary Rocks.
Benchmark Papers in Geology, 15, Dowden, Hutchinson
& Ross, Inc., Stroudsburg, Pennsylvania.
CHILINGARIAN, G.V., & WOLF, K.H. (eds) (1975). Compaction
of
Coarse-Grained
Sediments.
Developments in Sedimentology, 18A & 188, Elsevier,
Amsterdam.
CHOQUETTE, P.W., & PRAY, L.C. (1970). - Geological
nomenclature and classification of porosity in sedimentary carbonates. Bull. amer. Assoc. Petroleum Geol., 54, p.
207-250.
CURTIS, D.M. (1976). - Sedimentary Processes
Diagenesis. SEPM, Reprint series number 1.
DAPPLES, E.C. (1967). - Silica as an Agent in
Diagenesis. In: LARSEN, G. & CHILINGAR, CV., eds. :
Diagenesis in Sediments, p. 91-125, Elsevier,
Amsterdam.
DAPPLES, E.C. (1972). - Some concepts of
Cementation and Lithification of Sandstones. Bull. amer.
Assoc. Petroleum Geol., 56, p. 3-25.
DELHOMME, J.P. (1992). - A quantitative characterization of formation heterogeneities based on borehole
image analysis.
DICKEY, P.A. (1979). - Petroleum Development
Geology. Petroleum Publishing Co., Tulsa.
DUNHAM, R.J. (1962). - Classification of Carbonate
Rocks according to Depositional Texture. Amer. Assoc.
Petroleum Geol., Mem., 1, p. 108-121.
DUNOYER de SEGONZAC, G.D. (1968). - The birth
and development of the concept of diagenesis (18661966). Earth-Science Reviews, 4, 3, p. 153-201.
EVAMY, B.D. (1967). - Dedolomitization and the development of rhombohedra/ pores in limestones. J. sediment. Petrol., 37, 4, p. 1204-1215.
FAIRBRIDGE, R.W. (1967). - Phases of Diagenesis
and Authigenesis. In : LARSEN, G. & CHILINGAR, CV.,
eds. : Diagenesis in Sediments, p. 91-125, Elsevier,
Amsterdam.
FERTL, W.H. (1976). - Abnormal Formation
Serralog 0 2003
Diagenesis
Pressures. Developments in Petroleum Science, 2,
Elsevier, Amsterdam.
FOLK, R.L. (1959). - Practical Petrographic
Classification of Limestones. Bull. amer. Assoc.
Petroleum Geol., 43, p. 1-38.
FOLK, R.L. (1962). - Spectral subdivision of Limestone
Types. Amer. Assoc. Petroleum Geol., Mem. 1, p. 62-84.
FOLK, R.L., & LAND, L.S. (1974). - MgKa ratio and
salinity: two controls over crystallization of dolomite. Bull.
Am. Assoc. Petrol. Geol., 59, p. 60-68.
FRIEDMAN, G.M. (1964). - Early diagenesis and lithification in carbonate sediments. J. sediment. Petrol, 34, p.
777-813.
FRIEDMAN, G.M., & ALI, S.A. (1981). - Diagenesis of
Carbonate Rocks : Cement-Porosity Relationships.
SEPM Reprint series 10.
FRIEDMAN, G.M., & SANDERS, J.E. (1978). Principles of Sedimentology. John Wiley & Sons, New
York.
FUCHTBAUER, H. (1967). - lnfluence of different
types of diagenesis on sandstone porosity. Proc. 7th Wld
Petrol. Cong., Mexico, p. 353-369.
CARRELS, R.M., & MACKENSIE, F.T. (1971). Evolution of Sedimentary rocks. Norton, W W & Co., New
York.
GARY, M., McAFEE, R.Jr., & WOLF, C.L. (1972). Glossary of Geology. Amer. Geol. Institute, Washington,
D. C.
GRIM, R.E. (1958). - Concept ofdiagenesis in argillaceous sediments. Bull. amer. Assoc. Petroleum Geol., 42,
p. 247-253.
HASSAN, M. (1973). - The use of radioelements in
diagenetic studies of shale and carbonates. Int. Symp.
Petrography of Organic Material in Sediments. Centre
Nat. Rech. Scientifique, Paris, September.
HERRON, S., Le TENDRE, L., & DUFOUR, M.
(1988)- Source rock evaluation using geochemical information from wireline logs and cores. Bull. amer. Assoc.
Petroleum Geol., 72, 8.
HERRON, S., PETRICOLA, M., SAPRU, A., DEVRAJAN, N.R., & IYER, R.S. (1994). - Through the reef barrier. Schlumberger, Middle East Well Evaluation Review,
15.
HOBSON, G.D., & TIRATSOO, E.N. (1975). lntroduction to Petroleum Geology. Scientific Press Ltd,
Beaconsfield, England.
KRUMBEIN, W.C. (1942). - Physical and Chemical
Changes in Sediments after Deposition. J. sediment.
Petrol., 12, 3, p. 111-117. Also in S.E.P.M. Reprint Series
number 1, Sedimentary Processes, p. 13-19.
KRUMBEIN, W.C., & SLOSS, L.L. (1963). Stratigraphy and Sedimentation. 2nd ed. W.H. Freeman &
Co., San Francisco.
KRYNINE, P.D. (1948). - The megascopic study and
field classification of sedimentary rocks. J. Geology, 56, p.
130-165.
LANDES, K.K. (1951). - Petroleum Geology. John
Wiley & Sons, New York.
LARSEN, G., & CHILINGAR, G.V. (1967). Serralog Q 2003
I
Chapter6
I
263
Diagenesis in Sediments. Developments in sedimentology, 8, Elsevier, Amsterdam, 551 p.
LEET, L. Don, JUDSON, S., & KAUFFMAN, M.E.
(1978). - Physical Geology. 5th ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
LINK, P.K. (1982). - Basic Petroleum Geology. OGCl
Publications, Tulsa.
LOMBARD, A. (1956). - Geologie Sedimentaire. Les
series marines. Masson, Paris.
LOMBARD, A. (1972). - Series sedimentaires. Genese
- Evolution. Masson, Paris.
LONGMAN (1980). - Carbonate diagenetic textures
from near-surface diagenetic environments. Bull. amer.
Assoc. Petroleum Geol., 64, p. 461-487.
NEASHAM, J.W. (1977). - The morphology of dispersed clays in sandstone reservoirs and its effect on sandstone shaliness, pore space and fluid flow properties. 52nd
Ann. Fall Meet., SPE of AIMS, paper SPE 6858.
PETTIJOHN, F.J. (1975). - Sedimentary Rocks. 3d ed.
Harper & Row, Publishers, New York.
PETTIJOHN, F.J., POTTER, P.E., & SIEVER, R.
(1972). - Sand and Sandstone. Springer, New York.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Book Division, Houston.
POWERS, M.C. (1967). - Fluid release mechanisms in
compacting marine mudrocks and their importance in oil
exploration. Bull. amer. Assoc. Petroleum Geol., 51, p.
1240-1253.
PRAY, L.C., & MURRAY, R.C. (eds) (1965). Dolomitization and Limestone Diagenesis. A symposium.
SEPM, Special Publication 13.
PRESS, F., & SIEVER, R. (1978). - Earth. 2nd ed. W.H.
Freeman & Co, San Francisco.
RIEKE, H.H. Ill, & CHILINGARIAN, G.V. (1974). Compaction of Argillaceous Sediments. Developments in
Sedimentology, 16, Elsevier, Amsterdam.
RUSSELL, W.L. (1945). - Relation of Radioactivity,
Organic content, and sedimentation. Am. Assoc. Pet.
Geol. Bull., 29, 10.
RUSSELL, W.L. (1951). - Principles of Petroleum
Geology. McGraw-Hill Book Co, New York.
SELLEY, R.C. (1976). - An lntroduction to
Sedimentology. Academic Press, London.
SERRA, 0. (1973). - Interpretation geologique des diagraphies differees en series carbonatees. Bull. Centre
Rech. Pau-SNPA, 7, 1, p. 265-284.
SIPPEL, R.F. (1968). - Sandstone petrology, evidence
from luminescence petrography. J. sediment. Petrol., 38,
p. 530-554.
SUPERNAW, I.R., McCoy, A.D., & LINK,A.J. (1978). Method for in-situ evaluation of the source rock potential
of each formation. U.S. patent 4,071,744, Jan. 31.
SWANSON, V.E. (1960). - Oil yield and Uranium
Content of black shales. Geol. Survey, Prof. paper 356-A.
TAYLOR, J.M. (1950). - Pore space reduction in
sandstones. Bull. amer. Assoc. Petroleum Geol., 34, P.
70 1-716.
TRASK, P.D. (1932). - Origin and environment of source sediments of petroleum. Houston.
263
WELLER, J.M. (1960). - Stratigraphic principles and
practice. Harped New York.
WILSON, J.L. (1975). - Carbonate Facies in Geologic
History. Springer, New York.
WOLF, K.L., & CHILINGARIAN, G.V. (1976). Diagenesis of sandstones and compaction. In :
CHILINGARIAN, G. V. & WOLF, K H. (eds): Compaction
of Coarse-grained Sediments, II; Developments in
Sedimentology, 18b, Elsevier, Amsterdam.
ZOBELL, C.E. (1942). - Changes produced by
Microorganisms in Sediments affer Deposition. J. sediment. Petrol., 12, 3, p. 127-136. Also in S.E.P.M. Reprint
Series no 1, Sedimentary Processes, p. 20-29.
264
Serralog Q 2003
K
W
I-
n
a
I
0
WELL LOGGING AND COMPACTION
Introduction
Review of geological concepts
Definition
Compaction is the “reduction in bulk volume or thickness of, or the pore space within, a body of fine-grained
sediments in response to the increasing weight of overlying material that is continually being deposited or to the
pressures resulting from earth movements within the
crust” (Bates & Jackson, 1980).
Generally, compaction is the result of the mechanical
rearrangement of grains. The result is a reduction in volume at the expense of the original void spaces, in other
words, a reduction in the initial porosity.
The principal stress is therefore vertical, and directed
downwards (Fig. 7-1). However, further compression may
occur because the ensuing forces may result from tectonic movements in various directions. This kind of compression only occurs after compaction.
Compaction study
Compaction is particularly marked in detrital or particle-composed sediments. The amount of compaction
depends on the initial porosity and on the size, shape and
sorting of the grains. It also depends on the rate of sedimentation and period of time.
As pointed out by Fertl (1976), the compaction is related to several parameters which are:
(T :
the stress on the system,
the density of the formation,
p:
the porosity of the formation,
@
the permeability of the formation,
k:
D:
the burial depth,
the time since the starting of the burial,
t:
c:
the compressibility relationship,
v:
the velocity parameter for solids and
interstitial fluids in the system,
V:
the volume relationships.
The stress caused by burial is a function of the force
per unit area, and is represented by (T. Its unit is the pascal (= newtonlmz) and its dimension is kg/m/s*.
(T
= force/surface
(7-11
The force (whose unit is the newton) generated by the
sediments is equal to the product of their mass, M, and
the acceleration due to gravity, g. Therefore, eq. (7-1)
becomes:
C
(T
= (M x g) / surface
(7-2)
The mass of the sediments is equal to the product of
their average density (usually expressed in glcm3) and
their volume (surface x height). Thus:
(T
Figure 7-1 - Classification of compaction stresses : (C) polyaxial stresses (pz > py > px) (A) hydrostatic stresses (px = py = pJ; (B) triaxial
uniaxial stresses (the four faces which are
stresses (px = py < pJ; (0)
parallel to stress pz are stationary); (E) biaxial stresses (px = py and
the faces parallel to these two stresses are maintained stationary)
(from Sawabini et al., 1974).
Serraiog Q 2003
= (&
g surface x h) / surface
(7-3)
Or simplifying, we get the total overburden pressure’
(7-4a)
(7-4b)
265
where:
0 = the overburden pressure in pascal2,
-Pb = total average density of the sediment in g/cm3,
g = acceleration due to gravity (m/s2),
h = height (or thickness or depth) in meters.
The hydrostatic component pp is also known as the
interstitial fluid pressure, or pore pressure, and is equal to
the hydrostatic pressure (i.e. the product of average fluid
density and the height of the fluid column), if the fluid is
pure water and the compaction is normal.
0-
For porous formations, P b can be expressed as:
2-
(7-5)
4-
where:
ir, = average density of the fluid in g/cmJ,
pma= average density of the matrix in glcm3, (matrix as
understood by the log analyst, i.e. the total non-fluid part
of the formation),
;b = average porosity of the formations.
Overburden pressure increases with depth, and, as a
first approximation, would be assumed to be uniformly
proportional to depth (Fig. 7-2). However, this is not the
case, since the density of rocks increases with depth.
Assuming a uniform increase, we define an overburden
pressure gradient which is assumed to be equal to 0.231
kg/cmn/m (= 1 psi/ft) for an average density of rocks equal
to 2.31 gkm3 (Fig. 7-3).
For porous formations, by using eqs.(7-2) and (7-3),
the overburden pressure can be broken down into two
components, hydrostatic, pp (pressure of fluid column)
and lithostatic, pe (pressure of sediment column or intergranular pressure). This is represented schematically in
Figure 7-4. Thus,
N.B. In the case of non-porous rocks, pp = 0 and
(J
= pe.
6-
z
8-
5
10-
0
X
v
'
5
o.
1214-
16 18
1
20
0.7
08
a9
I
I
19
to5
Overburden gradient (psilft)
Figure 7-3 - Variation of geostatic pressure gradient with depth.
(1) theoretical; (2) in Texas and Louisiana; (3) in California; (4) in the
North Sea (from Fertl, 7976).
h
(kglcm2 )
1Mx)
.pP
Zoo0
Figure 7-4 - Schematic representation of geostatic pressure and
its components.
3m
Figure 7-2 - Variation of geostatic pressure with depth.
This pressure may exceed the hydrostatic pressure if
the interstitial fluid is subjected to an excess of pressure
due, for example, to tectonic stresses or compaction (in
the case of an undercompacted formation). The hydrostatic pressure gradient is defined as the ratio pp/h, and is
equal to 0.1 kg/cmZ/m for pure water. The ratio of fluid
pressure, pp, to overburden pressure, 0,represented by
A,is minimal in a hydrostatic environment or in the case of
normal compaction.
The pressure is sometimes expressed in psi or kg/cm2.
266
Serralog 0 2003
Compaction
h=p,/o
h = 0.435 psilft (0,l kglcmzlm)
(7-7)
In an undercompacted environment, the fluid supports
some or all of the lithostatic pressure. pp increases while
pe decreases. pp tends towards the geostatic pressure o
and 31, tends towards a value of 1 psi/ft (Fig. 7-5).
Pressure (psi)
I
Chapter7
I
267
These will occupy the volume allowed by the arrangement of coarse grains, and will continue to do so until the
point at which the stress begins to crush the grains.
The composition of the sand also has a part to play.
The porosity of a "clean" sand, that is without shale or
mica, decreases less rapidly with depth than a shaly
sand. This is because, under stress, mica or shale grains,
which are less resistant and more plastic, lose their shape
and crumble, thereby invading the porous space.
Furthermore, the presence of such minerals as amphiboles, pyroxenes, plagioclases, feldspars, etc., encourages
the development of other diagenetic phenomena because
of their chemical instability. Such phenomena will quickly
take over from compaction.
According to measurements carried out by various
researchers (Maxwell & Verral, 1954; Borg & Maxwell,
1956; Maxwell, 1960), well-sorted quartz sands can have
very high porosities (25 to 30 %) at ambient temperatures,
for pressures corresponding to a depth of 13,000 m
(40,000 ft).
Pressure (kg/cm2)
Figure 7-5 - The concept of compaction in subsurface.
Sand compaction
The first stage of the consolidation is a mechanical
rearrangement of the grains 3. During this they roll or slide
over each other easily and rapidly depending on their
shape and sorting, because of the vertical stress exerted
by overlying sediments at the time of burial. This produces
a tighter or more compact arrangement and hence a
reduction of porosity, leading to an increase in the density. Furthermore, as shown by Taylor (1950), this rearrangement will cause the number of lower and total contacts
for each grain to increase (Fig. 7-6).
It can be demonstrated that for a stack of spheres of
the same size the number of lower contacts will increase
from 1 to 3 or 4 (Fig. 7-7). Taylor also established that the
type of contact changes with burial (Fig. 7-6 and 7-8).
From having a tangential shape (a point), the contact
becomes elongated (a ridge), then concavo-convex and
finally sutured (a surface). The change in porosity due to
compaction also depends on sorting. Initially, a poorly sorted sand is less porous than one which is well sorted.
However, the reduction in porosity with depth is less rapid
for a poorly sorted sand. Once the coarse grains have
been rearranged with tangential or elongated contacts by
the displacement of small grains, they will carry most of
the load and thus protect the smaller grains from further
stress.
See Allen 8 Chilingarian, 1975 for more details.
Serralog 0 2003
sdoo
Figure 7-6 - Effect of depth of burial on the type of grain-to-grain
contact in the Jurassic and Cretaceous sandstones in two Wyoming
wells (from Taylor, 1950).
12
4
Figure 7-7 - The six stacking schemes for spherical grains of the same
size which control the porosity and number of contacts
(adapted from Graton & Fraser, 1935).
According to Maxwell (1964), porosity decreases more
rapidly with depth when the temperature gradient increases (Fig. 7-9). For low temperature gradients (7°C per
1,000 ft) and young formations (Miocene), Maxwell recorded, in Louisiana and Texas, a porosity of 20 % at a depth
267
of about 6,000 m. It can be concluded that a high temperature gradient encourages the early development of
other diagenetic phenomena, particularly cementation, at
which time compaction becomes a secondary effect.
No
contact
Point
contact
cult to determine the diagenetic phenomena responsible
for the reduction in porosity on the basis of individual samples. For this, a detailed analysis should be carried out on
thin sections by cathodoluminescence or a scanning electron microscope. It is clear that beyond a certain temperature and pressure which depends on the composition
and texture of the rock, other phenomena begin to take
over from mechanical compaction. For very clean, mature and well-sorted sands, (i.e. free of clay), and for a low
temperature gradient, these secondary effects appear
later, hence at greater depths.
Line
contact
Surface
contact
Figure 7-8 - Variation of the number of lower contacts and of the porosity with depth of burial (adapted from Taylor, 1950).
h
2
s
t!
v
5
Q
5
0
8
8
Porosity (%)
Figure 7-9 - Variation of porosity with depth using various temperature
gradients and ages (from Maxwell, 1964).
Another factor in the reduction of porosity appears to
be the passage of time, which also encourages other diagenetic phenomena. Based on the results of 17,367 porosity measurements made on subsurface cores, Atwater &
Miller, 1965 (in Blatt, 1979) established that porosity
decreases linearly and continuously at depths below 350
m (Fig. 7-10). McCulloh (1967), on the other hand, defined a change of porosity with depth which is not linear.
This study was based on 4,000 porosity measurements
from cores (Fig. 7-11). The contradictory results of these
two researchers seems to be due to their analysis of porosity changes on a statistical basis only, rather than studying the origin of porosity reduction.
In fact, these curves include all the diagenetic effects
associated with burial and not just compaction. It is diffi268
Figure 7-10 - Relationship between depth of burial and porosity, based
on measurements on 17,367 core samples from sandstones of the late
Tertiary in Louisiana. The points represent the average values over
1000-foot intervals (from Atwater & Miller, 1965, unpublished).
The analysis of the effects of burial on sands should be
based on changes in maximum porosity which should be
related to the porosity of "clean", pure, stable quartz
sands (i.e. well-sorted, with no unstable minerals). These
sands will undoubtedly be less influenced by diagenetic
effects other than compaction. Furthermore, considering
the stress exerted on the grains, the porosity seems to
decrease in steps. Each step corresponds, for a given
temperature gradient, to a certain pressure linked with a
resistance to normal or shearing stresses, to collapsing,
or to a change of mechanical behavior. Figure 7-12, taken
from Roberts (1969), shows that there is no modification
of the void ratio4 and hence porosity, until a certain pressure, known as the point of collapse, is reached. In the
same way, the variation in maximum porosity with depth
is seen to be stepped, as measured by Maxwell (1964)
and plotted in Fig. 7-14, other diagenetic effects being
minimal. The following can be concluded:
- The study of compaction in sands must be based on
the maximum porosity.
Serralog 0 2003
Compaction
- The reduction of porosity due to burial does not
decrease proportionally, but in steps.
- Compaction is often accompanied by secondary diagenetic phenomena especially in immature sands.
1
Chapter7
1
269
APPLIED PRESSURE. P S I
Total porosity (%)
APPLIED PRESSURE. K6/CM-'
Figure 7-12 - Relationship between void ratio and applied pressure (1)
in a Rhode Island sand; (2) in a Plum Island sand; (3) in quartz; (4) in
feldspar (from Roberts, 1969).
a0
n
u)
L
0
z
Y
->
m
C
cu
Probable maximum
average porosity of
reservoir sandstones
5
Q
d
&
.-3;;
rn
C
Maximum probable
porosity for most
sedimentary rocks
a
Portion of solid grains (G)
Probable upper limit of
porosity for virtually all
sedimentary rocks
Figure 7-13 - Relationship between porosity, void ratio and percentage
of solids (from Robertson, 1967).
Range of porosities of most
sediments and sedimentary
rocks
h
",
5
P
Figure 7-11 - Relationships between the total porosity of sedimentary
rocks and depths of burial, based on laboratory measurements on over
4000 core samples frdm the basins of Los Angeles and Venture, and
other locations in the US and Italy (from McCulloh, 1967). The curve
through the crossed points is from the Niger Delta.
The void ratio, e, is defined as the ratio of pore volume to solid
volume
e=V+Vasor e = $/(I - $)
(7-8)
Robertson (1967) defined the percentage of solids G as the ratio of
the volume of solids to the total volume (Fig. 7-13)
G=Va/Vb,OUG=l - $
(7-9)
This gives
e = $lG
(7-10)
Serralog 0 2003
0"
Miocene sandstones
Frio (Oligocene)sandstones, Acadia, Allen
Figure 7-14 - Variation of maximum porosity with depth (from Maxwell,
1964). The variation can equally well be plotted in steps.
- At a given depth, a porosity which is below the com269
270
Well logging and Geology
paction line indicates either a decrease in sorting, or the
development of other diagenetic effects (cementation,
authigenesis, etc.), connected with a combined effect of
pressure and temperature and chemical instability of the
minerals present.
- Levels of higher porosity indicate undercompaction
and suggest the presence of zones of high pressure.
Such zones are detectable by an increase in mud pit
levels during drilling, in which case mud weight should be
increased. Higher porosity may also indicate secondary
porosity due to epidiagenesis if the reservoir pressure is
hydrostatic.
Carbonate compaction
One can consider, as Coogan & Manus (1975) did,
that the compaction of carbonates depends on three principal factors - which are also those which act in the compaction of other sediments.
- Inherited factors which are related to the original
composition of the carbonate (particle mineralogy), and to
the texture (grain and crystal size and shape, sorting, packing), consequently to the depositional environment.
- inhibitory factors which inhibit compaction and are
related to physically and biologically induced chemical
changes during preburial lithification or alteration such as
synsedimentary or early cementation and dolomitization.
- Dynamic factors which are related to the depositional, diagenetic and tectonic environments, and overburden pressure, subsurface temperature, duration of burial
stress, pore pressure and pore fluids.
Completely contrary to quartz and shales, the carbonate minerals (fundamentally calcite and dolomite), are
very soluble, their solubility depending greatly on their
purity, the pH and Eh conditions, and the temperature and
pressure.
The initial (depositional) porosity of carbonates is often
very high (between 40 and 80 %). This high porosity
favors fluid circulation if it is associated with a high permeability, which depends on the size of the crystals and
the throats between the pores. Consequently, the interstitial waters are not always in chemical equilibrium with the
surrounding rock. This can create ionic exchanges between the solutions and the minerals composing the rock.
Furthermore, a high dispersion of the size of the particles
and crystals leads to different amounts of dissolving and
recrystallization. All these factors result in early diagenesis which leads either to lithification of the rock by cementation and/or recrystallization, or to a weakening of the
rock due to dissolution. In the first case compaction by
burial will have little effect. In the second, it will become
important and will generate other diagenetic phenomena :
pressure solution, fracturing, stylolitization, brecciation,
etc.
So, carbonate rocks react to burial in different ways
depending on the original facies, and the importance and
type of diagenetic phenomena supported since the depo270
sition.
Other factors must be taken into account
- the rate of sedimentation and subsidence;
- the geothermal regional gradient (i.e proximity of a
volcanic activity, etc.), the temperature effect on the sohbility of calcium carbonate being important;
- the maximum effective stress or overburden pressure;
- length of time of burial;
-the fluid pressure, which, if it is high, will decrease the
effect of compaction;
- the formation fluids, the presence of oil or gas can
considerably decrease the ionic exchanges and consequently the diagenetic phenomena. Oil-wet particle surfaces could substantially reduce the potential for solution of
carbonate minerals and the redeposition of calcite as
cement in the pore space.
In the case of chalks, the initial porosity is very high
(between 70 and 80 %). The size (1 to 20 pm) of the organisms which compose them (coccoliths and foraminifers,
Figure 7-15), and the good sorting, give this type of rock
poor permeability, limiting fluid movement. Furthermore,
the magnesium content is often close to zero, giving them
a high degree of diagenetic stability. The decrease of
porosity with depth is, in its initial phase, fundamentally
linked to compaction. The porosity evolution with depth is
close to that found for sands (Fig. 7-16). But, the compaction is sometimes delayed either when the rate of
sedimentation is high, or when the hydrocarbons have
early occupied the pore space at an early stage, or when
the two phenomena are combined as in the North Sea
example (Fig. 7-16).
In the case of reefs (bioherms), the framework itself of
the recifal body is the reason that these deposits can support an important burial without a significant loss of porosity. Other diagenetic effects (cementation, dolomitization)
can, however, modify their behavior.
_I
Figure 7-15 - SEM photogrzaph of chalk showing the particle size.
In the case of carbonate sands (bioclastic, oolites) the
compaction will depend on the importance and the nature
of the other diagenetic phenomena which themselves will
depend on the depositional and diagenetic environments.
Serralog 0 2003
Compaction
From a general point of view, taking into account all
phenomena, a recent study on carbonates from the south
of Florida, would seem to indicate that the loss of porosity of limestones is more rapid than that of dolomites (Fig.
7-17).
ChaDter7
I
271
Shale compaction
Shales are well-suited for studying compaction because of their high initial porosity and the lower importance of
other diagenetic phenomena. So, a great deal of work has
been done on them, resulting in numerous publicationss.
Of particular interest are those of Rieke & Chilingarian
(1974), and Fertl (1976). The following paragraphs summarize the basic points necessary to understand compaction.
PrinciDal Staaes in Shale ComDaction
Like all other sediments, shales expel their interstitial
water when buried. Due to their significant permeability,
related to their high porosity, there is a considerable period during which shales can potentially expel this water.
North'Sea Coccolith 'Chalk
~
1
I L I
I
I
I
50
10
30
I
20
I
I
10
Porosiw, Dercent
-
Figure 7- 16 Porosity loss with depth in chalk. Gulf Coast Chalk lost
more porosity than normal. This is related to a high geothermal gradient. On the other hand, North Sea Chalk was deposited so quickly
that much more original porosity was preselved (courtesy of R. Nurmi).
Several models of compaction have been proposed.
- For Athy (1930) compaction is simply the process of
expelling interstitial fluids, resulting in a porosity decrease. However, after the deposition and burial of a sediment, its pore volume may be modified by:
- deformation of the grains;
- cementation;
- dissolution;
- recrystallization;
- grain compression.
- For Hedberg (1936) compaction has three stages
(Figs. 7-18 and 7-19).
Stages
Figure 7-18 - Hedberg's compaction model (from Hedberg, 1936).
Porosity, percent
-
Figure 7-17 Comparison of limestone and dolomite porosity evolution
with depth in a south Florida formation. Above 1500 m limestones are
generally more porous, below they are less porous than dolomites
(courtesy of R. Nurmi).
Serralog 0 2003
- Mechanical rearrangement and loss of water
from the shaly mass in an interval between 0 and 800 psi.
For a small pressure variation, the reduction in porosity is
high.
- Mechanical deformation of the particles together
with further expulsion of adsorbed water between 800 and
6,000 psi (0 e 35 %). Some recrystallization of shale particles may occur.
-
A detailed list of references is given for readers interested by the
subject.
271
- Recrystallization with porosities below 10 %.
The reduction of pore volume is slow and only happens
with a large increase in pressure. The larger crystals may
enlarge at the expense of the smaller ones.
- Weller's model (1959), which is very similar to that of
Hedberg, gives the main stages of compaction as follows:
- expulsion of interstitial water until the moment
when the grains come into contact with each other; porosity is reduced from 85 to 45 %;
- grain rearrangement;
- soft shaly minerals are squeezed between the
grains of the more resistant minerals;
- deformation, tending to eliminate all porosity.
1. The expulsion of interstitial water until the grains
come into contact with each other. The porosity in shales
falls from 70-85 % to around 45 %, a reduction which is
attained rapidly over a depth of a few tens of metres.
2. Mechanical rearrangement of grains and continued
expulsion of fluid. Porosity falls to about 25 %, a reduction
which is slower and occurs over several hundred metres.
3. Mechanical deformation of the particles and expulsion of the adsorbed water. Soft minerals sink into the
interstices between harder minerals. The porosity falls
from 35 to 10 %. This reduction is even slower and may
occur over several thousands of metres.
4. The development of major diagenetic phenomena
caused by modification of the physicalchemical equilibrium by the combined action of pressure and temperature, concentration of water and the presence of other
fluids, e.g. gas and oil. This development includes:
- recrystallization,
- cementation,
- dissolution.
ILLlTEAND KAOLlNlTECOMPACTION HSTMlY
Depth of burial (ft)
Figure 7-20 - Compaction history of various types of marine-deposited
shales and its probable relationship to the formation of hydrocarbons
(from Powers, 1967).
Figure 7-19 - Relationship between depth of burial and grain proportion
or porosity, showing Hedberg's compaction stages (from
Baldwin, 1971). (A) Oklahoma (Athy, 1930); ( 0 ) Venezuela (Dallmus,
1958); (DG) Gulf Coast (Dickinson, 1953); (ER1) Santa Barbara Basin
(Emery & Rittenberg, 1952); (ER2) California Coast (Emery &
Rittenberg, 1952); (ER3) Los Angeles Basin (Emery & Rittenberg,
1952); (G) Lake Mead (Gould, 1960); (H) Venezuela (Hedberg, 1936);
(J) Joides, Well 1 (Beall & Fisher, 1969); (K) 15 cores from the abyssal
plane (Kerrnabon et al., 1969) (KH) Venezuela (Kidwell & Hunt,1958);
(M) curves 3, 4, 7, 8, and 10 (Meade, 1966); (RK) New Scotland
(Richards & Kelley, 1962); (SH) compaction test by Skempton
(Hamilton, 1959); (SJ) composite curve by Skeels (Johnson, 1950);
compaction test on blue marine shales (Terzaghi,1925); (W) compaction test by Warner (Beall & Fisher, 1969);(B) average curve by
Baldwin (Baldwin, 1971).
Density = 1.32
Recent burial
Pore water
Interlayer water
Swelling day solids
Non-swelling clay solids
Non-day solids
- Powers (1967) considers that the mineralogical type
of the shale causes compaction to have different effects,
as shown in Figure 7-20.
-Finally, Burst (1969) proposed a compaction model
based on three stages of dehydration (Fig. 7-21).
From these, the stages of compaction can be summarized as follows:
272
Figure 7-21 - Global composition of a marine shale during its dehydration (from Burst, 1969).
Serralog 0 2003
Compaction
ComDaction Mechanisms
Note: The following pages will consider vertical compaction along a
single axis caused by the weight of overlying sediments on shales.
The mechanism of compaction was described by
Terzaghi & Peck (1948) and later by Hottman & Johnson
(1965). To explain this mechanism they created a model
like the one in Figure 7-22, consisting of perforated metal
plates, separated by metal springs and water, all within a
cylinder. The springs simulate the grain-to-grain contacts
between the shaly particles and the plates simulate these
particles. Manometers record the fluid pressure.
Overpressure
STAGEA
Hydrostalic pressure
STAGE B
At stage A of the experiment the valves are closed and
no fluid can be expelled. When pressure is exerted (o),
the springs are not subjected to any pressure (p, = 0), and
the pressure applied is totally counterbalanced by the
equal and opposite pressure of the water (p,) occupying
the volume Vi, and pp = o.
A practical way of recording this pressure is to note the
ratio of fluid pressure, pp and the pressure exerted o
(7-7)
At stage A, h = 1, and the system is overcompressed.
At stage B, the valve is partially open, the plates are
moved downwards (compaction), the volume V1 decreases and the springs transmit part of the applied pressure.
0 = Pp
+
Pe
(7-11)
and h = I .
At stage C, the valve is completely open and enough
water is expelled to allow the springs to reach compaction
equilibrium. From this moment the system is in equilibrium
and no more water is expelled. The water occupies a
volume V, smaller than V, and Vi (compaction). Now, the
applied stress is supported by the springs and the water
which is at the hydrostatic pressure. In such a case, h
approximately equals 0.433.
This model simulates well what actually happens:
- o represents the overburden pressure exerted by the
overlying sediments;
Serralog 0 2003
Chapter7
I 273
- pp is the interstitial fluid pressure at depth Z (formation or hydrostatic pressure);
- p, is the compaction pressure to which the shaly
matrix is subjected at depth Z, or the lithostatic pressure
to which it is subjected, if the proposals of Chapman
(1972) are accepted.
Thus, if the pore fluids can escape freely during subsidence, either towards the surface or by continuous drainage, the volume of original water (porosity) decreases
proportionally with Z, pe is maximum and pp tends to be
hydrostatic. In these circumstances, compaction is said to
be normal. However, if, given the burial conditions, fluids
can only escape with difficulty, the following conditions will
be found:
- slight reduction in the volume of original water, therefore a small porosity decrease with depth;
- p, will be abnormally low;
- pp will be close to the overburden pressure.
Here the compaction is defined as abnormal or undercompaction can be assumed.
The Hubbert & Rubev Law
STAGE C
Figure 7-22 - Schematic representation of shale compaction
(from Terzaghi & Peck, 1948).
h=pp/o
~
It has been established by these two authors that the
effective stress p, exerted on porous shale (or on the
springs mentioned in the experiment) depends on the
degree of shale compaction, and that p, increases proportionally with compaction. A practical guide to the degree of shale compaction is its porosity @,defined as the
ratio of pore volume to the total volume. From this it can
be deduced that, for a given shale, there is for each porosity value 4 a certain maximum value of effective compression stress p, which the shale can support without
further compaction. This is expressed by:
@sh =
(7-12)
with:
Pe = 0 Pp
(7-6b)
where:
@shO = shale porosity at zero burial depth (Z = O),
@sh = shale porosity at burial depth Z,
pe = compaction pressure exerted on the solid matrix
at depth Z,
k = constant,
(T = total pressure exerted on a porous shaly element
at depth Z (i.e. overburden pressure),
pp = fluid pressure.
This equation shows that shale porosity at a given
depth is a function of fluid pressure. If this pressure is
abnormally high the shale porosity will also be abnormally high, as in the case of overpressured shale. In hydrostatic conditions:
o = pbw.g.Z
(7-13)
and
Pp = Pw4.Z
(7-14)
273
where:
pbw = average density of the overlying sediments,
impregnated with water of a density pw in g/cm3.
pW= average density of the fluids in overlying formations in g/cmJ,
g = acceleration due to gravity,
Z = burial depth in meters.
Substituting values for (T and pp in eq. (7-14)
$sh =
(7-15)
with:
(7-16)
c = k (Bbw - Pw)4
c = compaction factor, with dimension L-1 (inverse of
length).
Eq. (7-16) shows that shale porosity varies exponentially with burial depth. So, plotting @sh logarithmically
against Z on an arithmetic scale, eq. (7-16) is shown by a
straight line, characterising normal compaction.
Using measurements from Athy (1930) on shales and
mudstone of the Paleozoic of North Oklahoma, Hubbert &
Rubey established the following coefficients:
These values are close to those formulated by Masse
(1971):
c = 1,14.10-3( m-' )
(7-18)
Os-,O = 48 %
which were established from measurements on 113 shale
samples in 18 wells drilled in 8 basins (Paris, Aquitaine,
Illizi, Gassi-Touil, Gabon and Majunga). They were made
at different burial depths varying between 800 and 5,500
m and ranging from the Gothlandian to the Upper
Cretaceous (Fig. 7-23). The differing values of c can be
explained by the fact that Masse's line is an average,
which integrates data from different basins with different
ages of rocks, while that of Hubbert & Rubey applies only
to the Permian formations of North Oklahoma.
Compaction Effects
Effects on shale porosity
As already seen, the mechanical effect of compaction
is a reduction in volume and therefore in porosity as well
as an increase in density. The connate water is expelled
towards less compact, more permeable areas of drainage.
The highest variations in density and porosity should
occur between 300 and 800 m and the gradient of specific mass of the shales should be between 0.02 and 0.05
g/cm3/100 m.
From Figure 7-23 it appears that during deposition
(mud with almost 85 % water, buried at 800 m, @sh = 20
%) porosity should decrease by 65 %, corresponding to
an expulsion of around 650 litres of water per m3 of sediments.
The change in shale porosity with depth has been studied by various researchers, and their findings are shown
in Figure 7-24. The curves reveal a wide range of differences, possibly due to the very varied composition of the
shale, different amounts of quartz, mica and calcite,
various shaly minerals or a variety of organic matter, and
various geological ages. The analysis of the graph shows
that, in many cases, curves can be more or less superposed by a simple translation. This seems to indicate a more
complex geological history, such as the effects of tectonic
stress, ages, or of a significant erosion of the overlying
sediments, bringing deeply buried rocks closer to the surface. However, this conclusion cannot be reached in the
absence of detailed analyses.
N
Porosity (%)
Porosity (%)
Figure 7-23 - Variation of shale porosity with depth
(from Masse, 1971).
274
Figure 7-24 - Relationship between porosity and maximum depth of
burial of shales and shaley sediments (from various authors:
(1) Proshlyakhov (1980); (2) Meade (1988); (3) Athy (1930); (4) Hosoi
(1983); (5) Hedberg (1936); (8) Dickinson (1953): (7) Magara (1968):
(8) Weller (1959): (9) Ham (1966): (70) Foster & Whalen (1986).
In: Rieke & Chilingarian (1974).
Serralog 0 2003
Next Page
Compaction p h a p t e r 7
Systematic studies (Chilingar & Knight, 1960) and
(Chilingar et a/., 1963) have been made of void ratio in
relation to pressure for pure clays (montmorillonite, illite,
kaolinite, dickite and halloysite), and the results are
shown in Figure 7-25. They indicate a different behavior
for each type of clay mineral.
I 275
0
l0OQ
2000
n 3000
E
W
01
100
10000
Pressure (kglcm’l
5
8 4coo
n
Figure 7-25 - Variations in the porosity of various types of shale with
pressure (from Chilingar & Knight, 1980).
5000
The curves do not seem to follow an exponential law,
but can be taken as such, this having been validated by
an analysis of log parameters over a considerable interval
of depth.
Effects on shale density
The general density of a rock depends on the percentage per unit volume of each component, and its respective density. This is given by:
where:
Pbsh = density of the shale recorded on a density log,
pw = density of the liquid in the shale pores,
pma= density of the shaly matrix or clay minerals,
@ = shale porosity.
Since the density of shale matrix varies between 2.05
(montmorillonite) and 3 (ferriferous chlorite), and that of
the liquid is close to 1, any porosity reduction is indicated
by an increase in total density and vice versa.
If density is proportional to porosity, in normal compaction the density of shale varies exponentially with
burial depth. So, plotting the density of the shales on a
logarithmic scale against depth on a linear scale, the line
given by P b against depth will be straight (at least where
the shale porosity varies exponentially with depth). This is
shown in Fig. 7-26. This is achieved at the wellsite by
measuring the density of shale cuttings and recording
how this density varies with depth. Similarly, as we will
see, the compaction of shales can be studied by following
the changes in density obtained from the density log as a
function of depth.
Serralog 0 2003
6000
5
Density, (g/cni3)
‘igure 7-26 - Variations of shale bulk densities with depth in sedimentary basins. 2 = mudstone of Po Valley Basin (after Storer, 1959);
3 = average Gulf Coast shale densities; values derived from geophysical data (after Dickinson, 1953);
4 = average Gulf Coast shale densities; values derived from density
logs and formation samples (after Eaton, 1969);
5 = Motatan-I - Maracaibo Basin, Venezuela (after Dallmus, 1958);
6 = Gorgeteg no?- Hungary; calculated wet density values
(after Skeels, 1943);
7 = Pennsylvanian and Permian dry shale density values, Oklahoma
and Texas, Athy’s adjusted curve (after Dallmus, 1958);
8 = Las Ollas-I - eastern Venezuela (after Dallmus, 1958).
Effects on the chemistry of interstitial fluids
The following facts are based on various observations:
- the salinity of interstitial waters in normally compacted shales increases with depth,
- the water salinity in undercompacted formations is
generally lower at a given depth, unless there are salt formations nearby.
The work of Engelhardt & Gaida (1963) (Figs. 7-27
and 7-28) shows that the concentration of salt in interstitial water in shales decreases initially with compaction,
but starts to increase once a certain degree of compaction
has been achieved. This phenomenon is explained by the
clay membrane effect. Once a certain degree of compaction has been reached, the clays allow water to pass, but
retain ions selectively (see, for example, the work of
Hitchon, 1964).
275
Previous Page
Porosity
sphere compactedzone
..._
lrovlerrn line8
-Rwmal Rux lines
Qmund surface
"Void ratio"
Figure 7-29- Distribution of thermal flux and temperature profile in the
case of a thermal conductor (from Lewis 8 Rose, 1870).
Figure 7-27- Variation in the concentration of NaCl in the interstitial
water of pores of montmorillonite (from Engelhardt & Gaida, 1983).
Ground surface
Lens insulator (high porosity)
- - _Isotherm lines
Thermal flux lines
Temperature
gradient
-
Temperature
Temperature gradient +
Q)
s
Figure 7-30-Distribution of thermal flux and temperature profile in the
case of a thermal insulator (from Lewis & Rose, 1970).
"Void ratio"
Figure 7-28- Enlarged section of the preceding figure.
Effect on geothermal gradient
The flow of heat from the depths of the earth crosses
the layers of formations within a basin and eventually dissipates at the surface at a rate which depends on climatic
conditions. For a given flow rate, the geothermal gradient
is a function of the thermal conductivities of the intervening formations (Lewis & Rose, 1970).
Any localised bodies of high conductivity, such as a
salt dome, cause a concentration of thermal flux (Fig. 729a) and the geothermal gradient decreases (Fig.7-29b),
while the opposite occurs with bodies of low conductivity
(Fig. 7-30).
In reservoirs which are thick and permeable, convection currents can be established, and this implies good
thermal conductivity. Thus the hydrodynamic renewal of
formation waters dissipates the thermal flux and tends to
diminish the geothermal gradient.
Conversely, the overlying non permeable formations
will appear as insulators, resulting in an increase in thermal gradient. Consequently, in sequences in which the
sandkhale ratio is low (probably indicating undercompacted shales), there will be few thermal conductors and the
geothermal gradient will be high.
276
Effects on mineralogical transformations
Chemical equilibria are affected by the effects of temperature and pressure associated with burial. Elements
and compounds will be subjected to solution, precipitation, cementation and mineralogical changes. We will only
list the well-known transformations
- carbonates : aragonite -+ calcite -. dolomite
- sulfates : gypsum -+ anhydrite
- clays: montmorillonite K illite, montmorillonite -. chlorite, kaolinite -. chlorite, kaolinite -+ illite.
The influence of time
The experiments of Terzaghi & Peck indicate little
influence due to time, the equilibrium position being reached almost immediately due to the high permeability of
the disks (Fig. 7-22). In the case of shales, however, the
permeability can often be very low and the loss of water
can be much slower, hence the time taken to reach equilibrium is longer. The influence of the time factor is illustrated by the diagram in Figure 7-31.
The pressure profile as a function of time will depend
on the rate of sedimentation and type of substratum, as
illustrated in Figure 7-32. It is possible that, on a geological time scale, a similar influence could exist, which would
explain the spread observed in the densities of shales
according to their ages (Dallmus, Fig.7-33).
On the basis of these observations, then, it can be
concluded that the trend of normal compaction will vary
from one basin to another according to the characteristics
of each basin, that is:
- the type of substratum under the shales,
- the rates of sedimentation and subsidence, and
hence the tectonic framework,
Serralog 0 2003
e n n u -
_-1_
Compaction E Chapter7
277
undercompaction (excess porosity) or to formations which
have been deeply buried being raised by extensive erosion, or to the consequences of tectonic movements.
0
1
SP*U
3 - Pliocene to Lower Oligocene
Sand
Figure 7-31 - Schematic representation of the compaction of a shale
and of the influence of time (from Katz & Ibrahim, 1971).
- Recent to Miocene
2 - Pliocene to Lower Miocene
WXK)
4 - Eocene
5 - Readjusted curve of Paleozoic
Figure 7-33 - Normal compaction curves for shales of different ages
(from Dallmus, 1955, in Weeks, L.G.(ed), Habitat of Oil).
Separation line for flow at t 2
Separation line for flow at t,
Permeable substratum
,Impervioussubstratum
Figure 7-32 - Diagrams showing the influence of time and of the type
of substratum on the compaction of shales.
- the composition of the sediments, their porosity and
permeability.
These factors will determine the thickness of sediment
which the water expelled from the shales by compaction
will have to cross to reach porous and permeable rocks.
They also determine the time taken for this migration to
occur, and hence the degree of compaction at a given
instant.
The time taken in fact depends on the permeability
which is a function of the type and grain size of the
sediments. Thus the trend of normal compaction must be
established for each basin and for each period of sedimentation. Every deviation from this relationship between
porosity and depth will correspond either to a state of
Serralog Q 2003
Compaction of organic sediments (peats)
From Weller (1959) the compaction processes of
peats are close to those of shales: water expulsion, followed by a physico-chemical transformation of the sediments. The variation of thickness can reach ratios of 30 to
1 (Renault, 1899; Rukhin, 1953; mentioned by Ryer &
Langer, 1980).
Compaction anomalies
Definition
A compaction anomaly or undercompaction is the
state of a sediment which has been unable to expel its
interstitial water during burial.
Oriain of compaction anomalies
Compaction anomalies are to be found in sequences
which lack porous and permeable escape routes for the
interstitial water of the sediments which should be expelled during compaction. The water remains trapped and
takes on part of the lithostatic overburden. The fluid pressure pp therefore increases, while the sediment itself is
undercompacted.
277
Zones Exhibitina Undercompaction
The phenomenon of undercompaction is found in
basins of rapid detrital sedimentation, poor in permeable
deposits or rich in clays, such as outer deltaic deposits. It
is also observed in basins of mixed detrital and evaporitic
sediments. The map in Figure 7-34 from Fertl (1972)
shows the regions throughout the world in which the phenomenon of undercompaction, or more generally of overpressured formations6 are found.
. increased sonic travel time, At,
. reduced density, Pb,
. reduced resistivity, R,and formation factor, F,
. increased neutron-hydrogen index, QN,
. increased capture cross-section, X
- formation pressure exceeding hydrostatic pressure,
- reduced formation water salinity,
- increased temperature,
- increased geothermal gradient.
If these parameters are plotted against depth, then,
undercompacted formations will exhibit anomalies as
illustrated in Fig. 7-35.
IHN X
-1
figure 7-34 - Countries in which overpressured formations have been
encountered (from Fertl, 1872).
These phenomena are encountered from Cambrian to
Pleistocene, and from depths of a few hundred meters to
about 6,000 m.
Traditional geological approach
Study of compaction is fundamentally based on measurement of the shale cuttings parameters (i.e. density),
collected during the drilling. Drilling parameters are used
to predict penetration into undercompacted or over-pressured formations in order to adjust mud parameters and
avoid blowout (Fertl, 1976).
Well logging approach
Compaction is one of the major diagenetic phenomena in detrital sequences, regardless of composition. It is
generally accompanied by secondary effects which can
be well studied using well logging data.
Associated Phenomena
Undercompacted formations will exhibit the following
features:
- increased porosity, and corresponding effects on
those parameters which depend on it:
~____
Overpressure is not always associated with undercompaction,but
may also be due to artesian pressure, osmotic phenomena, excessive
overburden or tectonic constraints.
278
Figure 7-35 - Schematic responses of the various logs on entry into an
undercompactedshale (completed from Fed/& Timko, 1977).
From the foregoing it is clear that the study of compaction can be undertaken using well logs, in particular
those affected by porosity variations. In broad terms, it
involves examining the variation in log parameters with
depth. The logs of prime concern are the resistivity, density and sonic travel time, and the formations of most
concern are fundamentally shales, but also to some
extent silts and sands, and very occasionally carbonates
(chalk in particular).
Construction of the comDaction Drofile
Manual plot
This method consists of plotting the resistivity, density,
sonic travel time and neutron-hydrogen index of each
pure, homogeneous shale formation on a logarithmic
scale against depth on a linear scale.
Construction of the profile at the wellsite
The wellsite computers present in the truck allows the
realization of a composite log which can be used to select
the shale zones and thereby to obtain the log parameters
of these zones (Fig. 7-36). A compaction profile can also
be generated that plots the sonic travel time on a logarithmic scale (Fig. 7-37).
DENSON automatic program
The DENSON program, written in 1969 by ELF
(Chiarelli et a/., 1973), automatically provides a shale or
even a sand compaction profile on a plotter using log
measurements which have been previously digitized or
recorded on magnetic tape.
Serralog 0 2003
Compaction
The following parameters are plotted against depth, as
shown in Fig. 7-38 in the left-hand track, on linear scales:
- the shale content, or the total natural gamma ray
curve corrected for mud effects,
- optionally, the neutron curve; in the right-hand track,
on logarithmic scales
- the sonic travel time and/or the bulk density of the
shale intervals,
- their resistivity,
- the salinity computed in clean zones from the SP or
in the shale zones by the Pickett's method (1960).
SP
SONIC
I
Chapter7
I 279
- by comparing the At and P b readings with those predicted for a normally-compacted shale at the same depth
based on the trend of normal compaction established for
a given basin or region. Only points having At values
greater than or equal to that of the normal trend, and Pb
below or equal to the normal are plotted (method 2).
Conductivity
Figure 7-37- Compaction profile created automatically at the wellsite
(courtesy of Schlumberger).
Figure 7-39 (next page) shows the profiles obtained by
both methods in the same well.
The logic of the program is illustrated by the simplified
flow-chart in Figure 7-40. Data are entered in the program
from magnetic tape which has either been recorded
directly or which is derived from digitizing optical logs.
Conduct of a study
Figure 7-36- Selection of shale zones and of shale parameters on a
composite log made at the wellsite, and construction of the trend of
normal compaction for detection entry into undercompactedzones
(around 2300 m) (courtesy of Schlumberger).
At the beginning of a regional study, the first method is
applied to several wells in order to establish the trend of
normal compaction. It then becomes possible to apply the
second method to individual wells7.
The shale levels chosen for plotting are selected
according to two criteria:
- by determining shale content using gamma ray , thorium and potassium data, and/or SP, retaining only those
zones showing a shale content (Vs,,) above 80 % (method
11,
In practice, the trend is displaced slightly upward, enabling normally-compacted shales, or values displaced by statistical or lithological
variations which would otherwise have been undetectable, to appear.
Serralog 0 2003
279
Particular care must be exercised in the definition of
the mean radioactive response in the shales, and the
intervals chosen must be representative of a pure shale,
and not of other radioactive formations. If necessary, the
neutron log can be used. It is often necessary to adjust
the response level several times in the same well to take
account of variations in hole diameter and mud type, as
well as changes in shale type. The use of LITHO results
(Delfiner et a/., 1984) can help to select shale intervals
more accurately.
might not be so apparent on a manual plot which would
naturally have a lower density of points.
Figure 7-39 - comparison of compaction profiles obtained by the two
methods applied to the same well logging data set
(Chiarelli et al., 1973).
J
Figure 7-38 - Example of a Compaction profile obtained with ELFAquitaine's DENSON program. At and Pb are used to detect entry into
undercompacted zones (marked by an asterisk).
The main advantages are:
- the speed of obtaining a profile, and the accuracy of
the plot,
- the possibility of correcting responses for borehole
and mud effects by introducing correction charts into the
program,
-the easy detection of levels which are badly caved or
very radioactive thanks to the symbols used,
-the possibility of using only the levels which are most
representative of a pure shale, marked by the symbol X,
and defined from the neutron log, in the computation of
the trend of normal compaction,
- the high density of points (one every six inches),
which allows poor quality points to be eliminated, and
which clearly shows certain geological phenomena which
280
Figure 7-40 - Logic of the DENSON program
(from Chiarelli et al., 1973).
Serralog 0 2003
Compaction
Construction of the normal comDaction trend
Surface seismic
and overlaid VSP trace
Chapter 7
281
Acoustic impedance
For a single well
In the case of a shale, the trend is constructed by
connecting a line through the points corresponding, on the
one hand, to a pure, homogeneous shale, and on the
other, to the hydrostatic pressures of the beds, determined either from pressure measurements in surrounding
reservoirs, or from the density of the mud, used for drilling
the formations, which must be the closest to 1 gkm3 (Fig.
7-41).
In the case of constructing the trend for a sand, "clean"
sands which are at hydrostatic pressure are chosen.
However, several different trends can usually be constructed, each one corresponding to a certain degree of
sorting, since this is what controls the initial porosity.
For several wells
The points for each well satisfying the above-mentioned conditions, that is falling on the normal compaction
line, are plotted together on a semi-logarithmic scale. A
regression line is then estabished for all of these points,
and the equation of this line is used in the second method.
m
11o
msonic transit time (pdft)
Figure 7-42 - Overpressure prediction from a VSP recording
(from Schlumberger, Well Evaluation Conference, Nigeria, 1985).
Figure 7-4 1 - Detection of over pressure using the sonic
(from Schlumberger, Well Evaluation Conference, Algeria, 7979).
Detection of undercompacted zones bv VSP
As previously mentioned, the VSP recording allows us
to see below the bottom of a well. An adapted interpretation of the recorded data, using inversion techniques to
model the boundaries below total depth, enables the
determination of the top of undercompacted or overpressured zones which correspond to a rupture in acoustic
impedance gradient.
Serralog 0 2003
This is possible by the fact that VSP allows the accurate separation between downgoing and upgoing wave
trains.
A layered model below the recording point, with up to
fifty boundaries for a length of trace of about 1 second, is
built such that the incident downgoing wavelet produces
an upgoing synthetic trace as close as possible to the real
one. The choice of these boundaries is done by stepwise
regression. This process insures that we deal mainly with
important contrasts and that we stay relatively immune to
noise. The successive iterative process keeps the initial
position of the boundaries, but computes the contrast
across these boundaries to obtain a synthetic trace which
nearly coincides with the real upgoing trace. The absolute value of the acoustic impedance is known at the recording point from the density and sonic logs, so the series
of impedance contrasts will start from that reference
value. When the change between two successive iterations is less than a pre-determined level the solution is
output (Fig. 7-42). Geological knowledge (presence of
fault, dips) can be introduced at some point of the trace
with a constraint related to the degree of uncertainty of
this knowledge (Fig. 7-43). An important break in the
acoustic impedance gradient can suggest the entry in an
undercompacted or overpressured formation.
Study of sand compaction using well logs
Since burial results in reduced porosity, it is easy to
study this from logs, given the wide range of measurements they provide and the great number of wells drilled
281
282
Well logging and Geology
throughout the world in which density and neutron-hydrogen logs have been recorded. Well logs also provide other
opportunities for analysis, when combining natural radioactivity data (total or selective) with the Pe index measured by the LDT tool, giving an accurate picture of the composition of the sand, without long and costly laboratory
analyses.
radioactive minerals, such as feldspars, mica or shale.
Figure 7-44 shows, as a function of depth, porosity
measurements for clean sands (radioactivity below a certain limit) determined from density logs recorded by ELF
in 26 wells in Nigeria and corresponding to 1875 measurements carried out at depths of between 2,000 and
14,000 ft.
Figure 7-43 - Overpressure prediction from VSP
The upper figure corresponds to a slightly dipping transition to massive
shales. The lower figure corresponds to a case of highly dipping sharp
transition through a fault surface
(from Schlumberger, Well Evaluation Conference, Nigeria, 1985).
Any tool sensitive to porosity (density, sonic transit
time or neutron-hydrogen index) may be used for this purpose. Either the raw values, or those corrected for the
"parasite" effects of the borehole influence, can be plotted
as a function of depth. With density or sonic logs, the
porosity value derived from the raw measurements (Pb or
At) is used, choosing those of quartz as the standard
values of density and sonic travel time (pma and Atma)of
the matrix, as understood by log analysts. In this case, the
calculation is made automatically from the following equations :
2000
$s = (Atf - At) / (A& - Attma)
(7-21)
An upper limit for radioactivity can also be set, excluding from the calculation
and crossplot
high levels
which may correspond to chemically and texturally h m a ture reservoirs. These may include chemically unstable
282
Figure 7-44 - Plot of porosity measurements, based on logs run by
ELF-Aquitaine in 26 wells in Nigeria, as a function of depth. There is a
clear general trend, while the variation in maximum porosity appears to
progress in steps.
Serralog 0 2003
_ _ _ _ _ ~
Compaction F h a p t e r 7 1 283
The porosity is plotted logarithmically. The average
change, given by the middle line, and corresponding to
the line of regression seems to be exponential, and represents the change of porosity with burial, all diagenetic
phenomena being considered. The same graph shows
the change in maximum porosity as well as minimum, the
maximum being defined in clear steps as previously mentioned.
Study of carbonate compaction using well logs
Thus, if the sonic travel time in shales is plotted logarithmically against depth on a linear scale, the points
representing the shales should fall on a straight line. This
is verified by the experimental results shown in Figures. 746 and 7-47.
The sonic, then, provides a means of studying the
compaction of shales as a function of depth. Because it is
largely unaffected by caving, the borehole compensated
sonic measurement provides a better measurement than
the density.
The carbonate compaction can be studied through the
porosity evolution with depth. But, one must remember
that the porosity changes can be related to other phenomena than compaction: cementation, dissolution, pressure-solution (see further).
Figure 7-45 gives an example of compaction study in
a chalk from North Sea.
Figure 7-46 - Variation of At., with depth for Oligocene and Neocene
shales of the Gulf Coast (from Hoffman & Johnson, 1965).
Figure 7-45 - Example of compaction profile in chalk. Observe the two
types of trends. In water bearing formations (lower resistivity zones
with low radioactivity) the average porosity is much lower than in oil
bearing zones (high resistivity).
Study of shale compaction using well logs
Sonic velocitv in shales
It has been widely accepted following the work of
Wyllie et a/. (1956)B that the sonic travel time is more or
less linearly dependent on porosity, and that therefore the
travel time in shales is an exponential function of depth in
the case of normal compaction:
Atsh
= @A&+ ( I -
(7-22)
where:
AtShis the sonic travel time in shales in pseclft,
Atf is the sound travel time in fluids,
Atrnathe sound travel time in the shale matrix, @ is the
porosity of the shales.
In spite of the errors pointed out by Raymer et a/. in the Wyllie
equation in certain conditions, we will continue to use it because it has
been more or less verified for porosities in the range of 10 to 45 %.
Serralog 0 2003
Resistivitv of shales
The resistivity of a rock depends on:
- the conductivity of the constituent minerals (metallic
sulfurs, haematite and graphite are conductive, for example),
- the porosity and the salinities and saturations of the
pore fluids,
- the temperature.
All other factors being equal, decreased porosity will
give increased resistivity. The salinity, however, often
increases with depth and may reduce or even reverse the
trend due to porosity.
Formation factor
Foster & Whalen (1985) proposed a shale formation
factor based on the ratio of the shale resistivity to the
resistivity of the water in the surrounding sands to overcome this difficulty
(7-23)
This formation factor then increases with cornpaction
and depth (Fig. 7-48). This method requires a knowledge
of R,, which is not always possible due to inaccuracies in
determining Rwfrom the SP, or indeed because the SP is
not available. It also assumes, controversially, that the
water in the shales is of the same type as that in the
283
Next Page
284
Well logging and Geology
sands. This method is little used.
I
I
10
I
I
1
b
I
I
.
1
20 3040 6080100
2
Figure 7-48 - Example of variations of Fst, as a function of depth in the
Guff Coast (from Fertl, 1976).
Hvdroaen index of shales
The presence of hydrogen is mainly due to molecules
of water or hydrocarbons. The shales contain molecular
water on the one hand, but also water or hydrocarbons
associated with porosity. It is possible, then, at least theoretically, to observe the reduction in porosity with depth by
studying the behavior of the hydrogen index of the shales.
However, most neutron tools saturate rapidly at porosities
above 50 to 55 %, and in addition they are adversely
affected by caves, which are common in shales. The compensated neutron tools allow this parameter to be used in
the study of compaction because of their increased range
and corrections for the effects of caves. Also, since these
tools are sensitive to gas, which is often present in shales,
the shales can appear to be more compacted than they
really are. For all these reasons, this method is little used.
CaDture cross-section of shales
Timko & Fertl (1970) have suggested this measurement for the study of compaction. The dependence of this
parameter on porosity suggests its usefulness for observing the effects of compaction on porosity, but it must be
remembered that the measurement is also very sensitive
to chlorine, and hence to salinity. This generally varies in
a different way to porosity, and its effect can mask the
effect of porosity.
Figure 7-47 - Trend of normal compaction of shales of a Nigeria well
over an interval of 14000 feet.
284
Natural aamma ray sDectroscoDv loq
Clays are transformed mineralogically under the cornbined effects of temperature and pressure, for example
Serralog 0 2003
Previous Page
Compaction
montmorillonite and kaolinite are changed into illite. This
results in a change in the potassium content and in the
thorium to potassium ratio. The latter should decrease
with depth, being around 10 for montmorillonite and 20 for
kaolinite, but only 4 for illite (Hassan & Hossin, 1975). This
method is not yet widely supported, partly due to a lack of
measurements, but also because this effect is slow to
appear and because illite can appear detritally at any
depth and because the ratio can be affected by other
radioactive minerals such as feldspars, micas or zircon.
1
~~
Chapter7
1
285
marks the entry into sequences where the chances of finding continuous reservoirs is considerably reduced.
Moreover, the problems, and hence the cost of drilling
increase considerably as soon as undercompacted (overpressured) zones are penetrated. It follows that isobath
maps of the upper limit of undercompaction provides useful information about the thickness of the zones which are
most attractive in the search for oil.
Applications of compaction studies
Hydrodynamic applications
Detection of entrv into an undercompacted formation
Note: Obviously, with the exception of VSP interpretation below the
bottom of the well, this analysis can only be applied during the drilling
through LWD measurement, or retrospectively since the wireline logs
are not available until the formation has been drilled.
Entry into an undercompacted formation will be marked by a shift of points relative to the trend of normal compaction already established. The displacements depend
partially on the logs used (see the diagrams of Fig. 7-35),
and partially on the degree of undercompaction (compare
the five profiles in Fig. 7-49).
J
Figure 7-50- lsobath map of the entry into undercompactedformations
(from Chiarelli et al., 1973).
The detection of the entry into undercompacted zones
using the logs allows certain interpretations related to the
existence of high-pressure formations, detected by well
site measurements and wrongly attributed to undercompaction, to be corrected (Fig. 7-51).
Evaluation of formation Dressures
Top of undrrumpldlon
Figure 7-49- Identification of undercompactedzones. Studying the
profiles enables the point of entry into the undercompactedshales to
be pinpointed, and gives an idea of the degree of undercompaction
(from Chiarelli et al., 1973).
The boundary at which the undercompaction appears
can be located very precisely. lsobath maps showing the
entry into undercompacted formations can be constructed
by applying this method to several wells in a region (Figs.
7-49 and 7-50). Such maps can be useful in designing
drilling programmes (casing points, selection of casing
and mud types) in future wells to be drilled in the region.
The appearance of undercompacted zones usually
Serralog 0 2003
Theoretically, the study of the degree of undercompaction of the shale should help in the evaluation of the
pressures of the surrounding reservoirs.
Let A be a point in the undercompacted zone at a
depth ZA (Fig. 7-52). The shale here suffers the same
compaction as the point E, at depth ZE.
According to eq. (7-6),
in A we have:
GA = (Pp)A -k @e)A
(7-24)
in E we have:
(7-25)
OE = (Pp)E
(Pe)E
Since the shales at A and E have suffered the same
degree of compaction, ( P ~=)(p,JE9.
~
+
In practice we assume that for equal porosities, the pressure supported by the matrix is the same. This is the basis of the model of
Terzaghi & Peck used here.
285
R
At E the shales are normally-compacted, and the
hydrostatic pressure ( P ~is) written:
~
where pp is in kg/cm3, Z, is in meters and pw is the
density of the fluid in g/cm3. We also havelo
<TE = PbwE
(zE/lo)
(7-27)
where :
PbwE is the global mean density of the sediments in the
interval surface-Z,.
Likewise, we can write:
OA
=
PbwA
(7-28)
(zA/l O)
If the type of shale does not vary, we can take PbwA as
equal to &,WE as a first approximation, since the sediment
porosity is the same at Z, and ZA. We can then write:
OA
Figure 7-51 - Example of a pressure anomaly due to gas in a zone of
normal compaction (from S e r a et al., 1975).
= PbwE (zA/lo)
(7-29)
) ~ their values in eq. (7-24)
Replacing bAand ( P ~with
gives:
or again
Nota: If pw and Pbw are not known, the following values can be used
pw = I,O g/crn3 et Pbw = 2,31 g/crn3.
ZE
The real case, illustrated in Fig. 7-53, shows that such
an evaluation is practicable.
In any case, in practice a very good approximation of
the trend of normal compaction tied to the shaly facies
studied is necessary in order to be exact. This poses a
serious difficulty.
lo
More precisely, we have on the one hand
(7-27b)
-1
and on the other
(7-28b)
or again
21
.A=&(
DEPTH
Figure 7-52- Estimating formation pressure from the compaction profile
[theoretical example).
286
cJZ)dZ+
~'~JZ)dZ)
(7-28~)
ZE
which gives the following expression by replacing OA and ( p e ) ~by their value
(7-31b)
Serralog 0 2003
~-~
Compaction
~
~~
Chapter 7
287
Certain authors (Hottman & Johnson, 1965; Ham,
1966), propose charts (Fig. 7-55 cf. next page) or nomograms (Fig. 7-56), which allow an empirical determination
of the pressure from the comparison of the shale resistivity (Rs&,bs, or the sonic travel time (Afsh),,bs, observed in
undercompacted shale intervals with the resistivity (Rsh)",
or the sonic travel time
defined at the same depth
for normally-compacted shales. These charts or nomograms are only valid within the region for which they have
been established.
Evaluation of Dressure aradients in massive. undercompacted shales
As soon as we are capable of estimating the interstitial
water pressure within a massive shale, the most probable
direction and value of the pressure gradient can be determined. Figure 7-57 represents a profile in an undercompacted zone where the points corresponding to the same
facies are distributed about a line which is practically vertical. This signifies that the interstitial water is subjected to
a geostatic gradient directed from bottom to top. This phenomenon is also detected on the profile in Figure 7-54.
It goes without saying that such a considerable gradient can influence the migration and hence the distribution of hydrocarbons. Such a distribulion of pressure reduces the value of any reservoir whose source rock is situated on top of it. This seems to be the case for the well
represented by Fig. 7-58, which, having penetrated a
sequence very rich in hydrocarbon shows, then entered a
water-bearing reservoir.
Figure 7-53 - Evaluation of formation pressure. The computed pressure is 470 kdcm' at 2925 m, while the pressure measured
during a test on the reservoir above is 455.1 kg/cm' at 2800 m
(from Chiarelli et al., 1973).
The three plots of Figure 7-54 (see next page) show,
nonetheless, that it is always possible to evaluate the
amplitude of the anomalies, which represents, in spite of
all, considerable progress.
Serralog 0 2003
Figure 7-58 - Compaction profile showing a steep pressure gradient,
but also showing variations in shale composition
(from Chiarelli et al., 1973).
287
288
Well logging and Geology
slight undercompaction
average undercompaction
high undercompaction
Figure 7-54 - Indication of the degree of undercompaction by a shift in the trend of normal compaction (from Chiarelli et al., 1973).
Figure. 7-55- Charts relating the fluid pressure gradient to R,. and At,,,. The fluid pressure is equal to p,r - FPG . ZAN where F f G is the pressure
gradient of the fluid from the chart (from Hottman & Johnson, 1985).
X
c.
al
P
Shale resistivity (ohm-m)
Acwsli impedance (gem Ip%)
Figure 7-56 - Nomograms for estimating formation pressure from resistivity or sonic travel time, established for the Gulf Coast
(from Hottman & Johnson, 1965).
288
Serralog 0 2003
Compaction
Figure 7-57 - Evaluation of pressure gradients. The alignment of the
shale points along a vertical indicates that the interstitial fluid is subjected to a geostatic pressure gradient (from Chiarelli et al., 1973).
Geological Applications
Detection of unconformities
Unconformities sometimes show up clearly on compaction profiles. In fact, they are marked, generally, by a
sudden shift in the trend of normal compaction (Fig. 7-59).
This shift undoubtedly indicates several phenomena:
- change of shale type,
-the strong influence of time elapsed since deposition,
- probable erosion.
[ Chapter7 I 289
I
I
Reconstruction of maximum burial deDths
If the log responses of normally-compacted shales of
the same geological period from different wells are plotted
together, it is frequently observed that the points from the
various well do not fall exactly on the same line, but rather
on a set of parallel lines, one for each well (Fig. 7-60).
This generally indicates a structural uplift of the whole
interval, but by different amounts for the different wells, so
long as we accept that compaction is an irreversible phenomenon.
Using the trend of well (1) for which the actual depth is
the greatest for a given value of At, the maximum apparent depth reached by each of the different shales (A, B,
C or D) in the interval can be determined. Thus, the shales C of well (4), currently between 2100 and 2250 m have
been buried to at least 3100-3250 m, according to the
trend for well (1). They have therefore been uplifted again
by about-I000 m.
This application thus provides a relatively precise
palaeogeographic reconstruction in basins which have
been subjected to deep erosion. The reader is referred to
the article by Lang (1978) which provides a good example.
These studies can also have geochemical applications
by providing an indication of the degree of diagenetic
change and the maturation of organic matter in relation to
maximum depths of sediment burial (Fig. 7-61). Through
this type of analysis, potential gas or oil provinces can be
evaluated.
However, these methods can only be applied to basins
which have not suffered major tectonic constraints.
S
F
Figure 7-59 - Three examples of unconformities detected using compaction profiles
(S = Serra, 1972; C = Chiarelli et al., 1973; F =Feel, 1978).
C
Serralog 0 2003
289
Porosity %
oil-filled sandstones
o water-filled sandstones
\ present depth
maximum depth of burial
/ maximum depth of burial
before oil accumulation
-
Figure 7-60 - Determination of the maximum depth of burial. (a) method of determination; (b) application example (from Philipp et al., 1963).
Correlations between wells
As the recent work of Lang (1984) pointed out, these
compaction profiles (Fig. 7-63) can be useful for establishing correlations between wells. They have the advantage of presenting the log data in a compressed form (due
to the scale used), thus emphasizing the major phenomena such as erosion, condensation, stable periods, etc.
Erosion
Deposition
Differential
compaction
Figure 7-61 - Maturation of the organic matter as a function of the
depth of burial and the temperature.
Differential comDaction
Since the loss of porosity with burial differs between
sands and shales, the change in volume will not be the
same. This leads to changes in the shapes of bodies as
illustrated by the diagrams in Fig. 7-62.
290
Figure 7-62 - Schematic showing the effects of differential compaction
on the shapes of sedimentary bodies.
Sedimentoloaical amlications
The strong influence of lithology on sonic velocity and
resistivity presented certain difficulties, and indeed cauSerralog 0 2003
Compaction
I Chapter 7 I 291
f 0 P STRATIORAPWC SECTION
WfPALtO DATA
Figure 7-63 - Correlations based on compaction profiles (from Lang, 1984).
sed some errors in the first interpretations of compaction
profiles. The measured variations are, in fact, more frequently related to changes in the type of lithology rather
than porosity variations. The variations can be quite
abrupt, as the example in Figure 7-58 of three different
shale facies illustrates. The slowest facies in particular
corresponds to a sediment rich in organic matter, which
explains its acoustic properties. Equally, the variations
can be progressive, due to a gradual enrichment in silts,
carbonates, etc., as in Fig. 7-64, rather than a change
from a state of normal compaction to one of undercompaction.
votl
v.h
Figure 7-64b - Other example of sequential variations from shale to
carbonate detected from compaction profiles
(from Chiarelli et al., 1973).
Figure 7-64a - Detection of sequential variations from compaction profiles (from Chiarelli et al., 1973).
Serralog 0 2003
Furthermore, the variation can sometimes be so gradual and extend over such a long interval (Fig. 7-65) as to
render it practically undetectable by a simple examination
of the logs.hi^ can lead to
of interpretationboth in
determining the trend of normal compaction and in
291
(7-32)
Assuming pma= 2.65 gkm3 and pf = 1 glcm3, the density is converted to porosity, and the following relationship
is obtained:
(7-33)
if Z is expressed in feet, and
(7-34)
with Z in meters.
Note : This statistical law is only valid for normally-compacted
sands (Fig.7-66) which have not suffered diagenetic effects (other than
those associated with compaction), and only in the interval 2000 - 14000
feet in Nigeria.
J
Figure 7-65- Sequential variations over 1700 m clearly shown by the
compaction profile (from Serra et al., 1975).
detecting undercompacted zones. Compaction trends will
therefore be used as a sequential tool providing a first
sketch or compensating for the deficiencies of the gamma
ray, and in certain cases correcting the trend it provides.
Application to Reservoir Studies
Variation of porositv with deDth
Several authors (i.e. Teodorovich & Chernov, 1968)
have proposed formula to express the variation of porosity and permeability with depth. For instance, for sandstones we have:
and
Z being the depth in meters.
A study of the density variations of sands using the
DENSON program in twenty-six wells, corresponding to
1875 density-depth pairs distributed between 2000 and
14000 feet, yielded the following relationship:
292
J
Figure 7-66- Laws of variation of density, sonic travel time and
porosity of sands and shales as a function of depth.
Geophysical Applications
A knowledge of the variations in density and sonic
velocity with depth of both shales and sands or silts can
be of major interest to the geophysicist. It can result in a
better interpretation of seismic profiles in the transformation of isochrones to isobaths, in the computation of
reflection coefficients and in the estimation of sand-shale
percentages. This is how data from numerous wells in
Nigeria were used by the DENSON program to provide
the curves in Fig. 7-66, which are represented by the following equations:
( pb) sd=1.95e
5.2 ~x10-5i3m)
(7-32)
Serralog 0 2003
Compaction
(7-36)
(7-37)
(At)
1.87x104Z(m)
=200e-
(7-38)
Sh
Note : The above relationships have been derived statistically from
measurements in a depth interval of 2000 - 14000 feet, and have only
been verified within this interval.
Chapter7
, 293
The second point concerns the formation of hydrocarbons in undercompacted sequences. Is the expulsion of
hydrocarbons from undercompacted source rocks always
sufficient for the creation of accumulations of commercial
interest? To what extent can a lack of expulsion inhibit the
normal functioning of a source rock? What exactly is the
effect of overpressure on the maturation process of organic matter? Such are the questions to which it would be
interesting to have answers; but these answers - some of
which are only now becoming clear -, will only have credibility if they are based on a great number of observations.
This is why the use of well logs must provide the basis of
the answers.
Decompaction
Geochemical Applications
This subject has been touched upon in the preceding
paragraphs in the discussions of detailed studies of shaley sediments, detection of zones rich in organic matter as
well as the reconstruction of maximum burial depths.
These are all questions which are of some interest to the
geochemist.
However, certain other points have not yet been mentioned. The first concerns the detection of gas zones in
massive shales.
Figure 7-67 speaks for itself. It shows without any
ambiguity the strong effect of gaseous hydrocarbons on
the resistivity. This property, which is a nuisance in the
study of shale compaction using only resistivity logs,
becomes of interest in qualitative studies of the distribution of gas shows within shales when other logs are available.
Obiectives
A decompaction procedure is necessary if a reconstruction of the arrangement of the different sedimentary
bodies as it should be at the time of deposition is required.
In that case it is important to apply to each body an appropriate decompaction coefficient. For that, the compaction
rate supported by each individual sedimentary body must
be taken into account. This compaction rate takes into
account the mineralogical composition, the texture, the
thickness, the environment of deposition and the maximal
depth of burial of each body. Figure 7-68 shows the chart
for estimating the compaction rate of shales as a function
of depth of burial and thickness. The law of decompaction
for a given lithology is generally deduced from the compaction law experimentally established for the same lithology from porosity measurements on rock samples or
from log data.
+
Figure 7-67 - Detection of gas zones in a massive shale
(from Chiarelli et al., 1973).
Serralog o 2003
Shale thkkness In m
Figure 7-68 - Chad giving the coefficient of decompaction, z, for pure
shale as a function of depth and thickness.
293
Procedures
They are defined by Brown, (1975) and illustrated by
Figure 7-69. They consist, firstly of correcting the apparent thickness of each sedimentary body, fundamentally
shales and peats, for overburden compaction; secondly,
of fitting the geometry to a subjacent palaeosurface or
assumed reference datum.
I
U
n
Figure 7-69 - Decompaction procedures (from Brown, 1975).
ADDlications
This decompaction technique allows reconstruction of
the palaeotopography at the end of deposition. It helps to
establish both chronostratigraphicand facies correlations.
References and Bibliography
ALLEN, D.R., & CHILINGARIAN, G.V. (1975). Mechanics of Sand Compaction. In : Compaction of
Coarse-Grained Sediments, I, (edited by Chilingarian,
G.V, & Wolf, K.H.), Developments in Sedimentology, 18A,
Elsevier, Amsterdam.
ASQUITH, G.B. (1982). - Basic Well Log Analysis for
Geologists. Amer. Assoc. Petroleum GeoL, Methods in
Exploration Series.
ATHY, L.F. (1930). - Density, porosity and compaction
ofsedimentary rocks. Bull. Amer. Assoc. Petroleum Geol.,
14, p. 1-24.
ATWATER, G.I., & MILLER, E.E. (1965). - The effect of
decrease in porosity with depth on future developemnt of
oil and gas reserves in South Louisiana. Amer. Assoc.
Petroleum Geol., Progr. Ann. Meet., New Orleans, p. 48.
BALDWIN, B. (1971). - Ways of deciphering compacted sediments. J. sediment. Petrol., 41, p. 293-301.
BATES, R.L. & JACKSON, J.A. (1980). - Glossary of
Geology. Amer. Geol. Institute, Falls Church, Virginia.
BLATT, H. (1979). - Diagenetic processes in sandstones. in: Aspects of Diagenesis. (edited by Scholle, P.A., &
Schluger, P.R.), Spec. Publs SEPM, 26, p. 141-157.
294
BLATT, H., MIDDLETON, G., & MURRAY, R. (1972,
1980). - Origin of Sedimentary Rocks. 1st and 2nd ed.
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
BOATMANN, W.A. (1967). - Measuring and using
shale density to aid in drilling wells in high pressured
areas. J. Petr. Technology, Nov., 19, 11, p. 1423-1430.
BOND, L.O.,ALGER, R.P., & SCHMIDT, A.W. (1971).
-Well log Applications in Coal Mining and Rock
Mechanics. Trans. SME, 250.
BOLT, D.B. (1972). - Detection of overpressured formations during drilling operations in worldwide environments. Ann. Sw. Petr., Short Course Ass. Mtg Proc.,
Lubbock, Texas, 19-20-21 May.
BOREL, W.J., & LEQIS, R.L. (1969). - Ways to detect
abnormal formation pressures. Part I geopressure prediction. Part 11: geopressure detection whil edrilling. Part 111 :
surface shale resistivity. Part IV : geopressure detection
procedures. Petrol. Engineer, July to Nov.
BORG, L.Y., & MAXWELL, J.C. (1956). - Interpretation
of fabrics of experimentally deformed sands. Amer. J. Sci.,
254, p. 71-81.
BROWN, L.F.Jr. (1975). - Role of sediment compaction in determining geometry and distribution of fluvial and
deltaic sandstones. In : Compaction of Coarse-grained
Sediments, I, (edited by Chilingarian, G.V., & Wolf, K.H.),
Developments in Sedimentology, 18A, Elsevier,
Amsterdam.
BURST, J.F. (1969). - Diagenesis of Gulf Coast clayey
sediments and its possible relation to petroleum migration. Bull. Amer. Assoc. Petroleum Geol., 52, 3.
CHAPMAN, R.E. (1972). - Clays with abnormal interstitial fluid pressures. Bull. Amer. Assoc. Petroleum Geol.,
56, 4, p. 7910-795.
CHIARELLI, A., SERRA, O., GRAS, C., MASSE, P., &
TISON, J. (1973). - Etude Automatique de la sous-compaction des argiles par diagraphies differees. Rev. Inst.
franp. PeGtrole, 28, 1, p. 19-36.
CHILINGAR, G.V., & KNIGHT, L. (1960). Relationship between pressure and moisture content of
kaolinite, illite and montmorillonite clays. Bull. Amer.
Assoc. Petroleum Geol., 44, p. 101-1016.
CHILINGAR, G.V., RIEKE, Ill, H.H., & ROBERTSON,
J.O. Jr. (1963). - Relationship between high overburden
pressure and moisture content of halloysite and dickite
clays. Bull. Geol. SOC.Amer., 74, p. 1041-1048.
CHILINGARIAN, G.V., & WOLF, K.H. (Eds) (1975). Compaction
of
Coarse-Grained
Sediments.
Developments in Sedimentology, 18A & 18B , Elsevier,
Amsterdam.
COOGAN, A.H., & MANUS, R.W. (1975). Compaction and Diagenesis of Carbonate Sands. In :
Compaction of Coarse-grained Sediments, 1, (edited by
Chilingarian, G.V, & Wolf, K.H.), Developments in
Sedimentology, 18A, Elsevier, Amsterdam.
COSTLEY, R.D. (1964). - Abnormal pressure determination. LCCOS, 2, 217, 14 Jan.
CURTIS, D.M. (1976). - Sedimentary Processes
Diagenesis. SEPM, Reprint series 1.
DALMUSS, K.F. (1958). - Mechanics of basin evoluSerralog 0 2003
I 295
~
Compaction
tion and its relation to the habitat of oil in the basin. In:
Habitat of Oil (edited by Weeks, L.G.), Amer Assoc.
Petroleum Geot, Mem. 36, p. 2071-2124.
DELFINER, P., PEYRET, O., & SERRA, 0. (1984). Automatic determination of Lithology from Well Logs. 59th
Ann. Techn. Conf. SPE of AIME, Houston, Texas; paper
SPE 13290.
DELHOMME, J.P., & SERRA, 0. (1984). - Dipmeterderived Logs for Sedimentological Analysis. SPWLA, 9th
Europ. Intern. Format. Eval. Trans., paper 50.
DICKINSON, G. (1951). - Geological aspects of
abnormal reservoir pressures in the Gulf Coast region of
Louisiana, U.S.A. Proc. 3th Wld Petrol. Cong., section 1,
p. 1-16.
DICKINSON, G. (1951). - Geological aspects of abnormal reservoir pressures in Gulf Coast Louisiana. Bull.
Amer. Assoc. Petroleum Geol., 37, p. 410-432.
DICKINSON, G. (1953). - Reservoir pressures in Gulf
Coast Louisiana. Bul. Amer. Assoc. Petroleum Geol., 37,
2, p. 410-432.
EATON, B.A. (1972). - The effect of overburden stress
on geopressure prediction from well logs. J. Petrol.
Technol., 24, p. 929-930.
ENGELHARDT, W.V. von, & GAIDA, K.H. (1963). Concentration Changes of Pore Solution During
Compaction of Clay Sediments. J. sedim. Petrology, 33,
4.
FAIT, I. (1957). - Effect of overburden and reservoir
pressure on electric logging formation factor. Bull. Amer.
Assoc. Petroleum Geol., 41, 11, p. 2456-2466.
FERTL, W.H. (1972). - World wide occurrence of
abnormal formation pressure. Part 1. SPE of AIME
Abnormal Subsurface Pressure Symp., Baton Rouge, La.,
15-16 mat 1972, prepint, paper SPE 3844, p. 29-42.
FERTL, W.H. (1976). - Abnormal Formation
Pressures. Developments in Petroleum Science, 2,
Elsevier, Amsterdam.
FERTL, W.H., & TIMKO, D.J. (1970). - Occurrence and
significance of abnormal pressure formations. Oil and
Gas J., 5 Jan. 1970, 68, 1.
FERTL, W.H., & TIMKO, D.J. (1970). - How abnormal
pressure detection techniques are applied. Oil and Gas J.,
12 Jan. 1970, 68, 2.
FERTL, W.H., & TIMKO, D.J. (1970). - Association of
salinity variations and geopressures in soft and hard
rocks. SPWLA, IlthAnn. Log. Symp. Trans., paper C.
FERTL, W.H., & TIMKO, D.J. (1970). - Occurrence of
cemented reservoir roof rock and geopressure cap rock
and its implication in Petroleum Geology and
Geohydrology. SPE of AIME Fall meeting, Houston, oct
1970, paper SPE 3085.
FERTL, W.H., & TIMKO, D.J. (1971). - Parameters for
identification of overpressured formations. Conf. on
drilling and rock mechanics, Austin, Texas, 5-6 Jan.,
paper SPE 3223.
FERTL, W.H., & TIMKO, D.J. (1971). - Utilisation des
diagraphies en recherche, forage et production d’hydrocarbures pour Etude des problemes de pressions geostatiques. Rev. Inst. franp. Petrole, 26, 9, p. 687-714.
Serralog 0 2003
Chapter7
_
FERTL, W.H., & TIMKO, D.J. (1971). - Salinity plots
evaluate subsurface formations. World Oil, 173, 1, p. 93100.
FERTL, W.H., & TIMKO, D.J. (1972). - How downhole
temperatures, pressures affect drilling. World Oil, Jun.Sept., part. 1 to 4.
FINCH, W.C. (1969). - Abnormal pressure in the
Antelope field, North Dakota. J. Petrol. Technol., 21, p.
821-826.
FONS, L. Sr. (1969). - Geological application of well
logs. SPWLA, 10thAnn. Log. Symp. Trans.
FOSTER, J.B. & WHALEN, H.E. (1965). - Estimation
of formation pressures from electrical surveys, offshore
Louisiana. 40th Ann. Fall Meet. SPE of AIME, and in J.
Petrol. Technol., Feb. 1966,18, p. 165-171.
FOWLER, W.A. (1968). - Pressures, hydrocarbon
accumulation and salinities, Chocolate Bayou fields,
Brazoria County, Texas. 43rd Ann. fall Meet. SPE of AIME,
Houston, paper SPE 2226, and in J. Petrol. Technol., 22,
p. 411-423.
GARY, M., McAFEE, R.Jr., & WOLF, C.L. (1972). Glossary of Geology. Amer. Geol. Institute, Washington,
D.C.
GRATON, L.C., & FRASER, H.J. (1935). - Systematic
packing of spheres with particular relation to porosity and
permeability. J. Geol., 43, p. 785-909.
GRETENER, P.E. (1969). - fluid pressure in porous
media, its importance in geology. A review. Bull. Canad.
Petroleum Geol., 17, p. 255-295.
HAM, H.H. (1966). - A method of estimating formation
pressure from Gulf Coast well logs. Gulf Coast Assoc.
geol. SOC.,16, p. 185-197.
HASSAN, M., & HOSSIN, A. (1975). - Contribution a
Etude des comportements du thorium et du potassium
dans les roches sedimentaires. C. R. Acad. Sci. (Paris), D
280, p. 533-535.
HEDBERG, H.D. (1936). - Gravitational compaction of
clays and shales. Amer. J. Sci., 5th series, 31, 184, p. 241287.
HILL, G.A., COLBURN, W.A., & KNIGHT, J.W. (1961).
- Reducing oil-finding costs by use of hydrodynamic evaluation. In: Economics of Petroleum Exploration,
Development and Property Evaluation. Prentice-Hall,
Englewood Cliffs, New Jersey.
HITCHON, B. (1964). - Formation fluids. In: Geological
History of Western Canada.Alberta SOC.Petroleum Geot,
Calgary, p. 201-217.
HOSOI, H. (1963). - First migration of petroleum in
Akita and Yamagata Prefectures. Jap. Assoc. Miner.,
Petrol., Econ. Geol., 49 , 2, and 3, p. 43-55 and 101-114.
HOTTMANN, C.E., & JOHNSON, R.K. (1965). Estimation of formation pressures from log derived shale
properties. J. Petrol. TechnoL,I 7, 6, p. 718-722.
Houston Geol. SOC. (1971). - Abnormal subsurface
pressure. A study group report 7969-7977. Houston
Geo% SOC.,p. 1-92.
HUBBERT, M.K., & RUBEY, W.W. (1959). - Role of
fluid pressure in mechanics of overthrust faulting. Bull.
geol. SOC.Amer., 70, 2, p. 115-166.
295
_
JONES, P.H. (1969). - Hydrodynamics of geopressure
in the northern Gul fof Mexico Basin. J. Petrol. Technol.,
21, p. 803-810.
KATZ, D.L. & IBRAHIM, M.A. (1971). - Threshold
displacement pressure considerations for caprocks of
abnormal-pressure reservoirs. 5th Conf. on Drilling and
Rock Mechanics, Austin, Texas, SPE 3222.
KERR, P.E., & BARRINGTON, J. (1961). - Clays of
deep shale zone, Caillo ulsland, La. Bull. Amer. Assoc.
Petroleum Geol., 45, p. 1698-1712.
KRUMBEIN, W.C. (1942). - Physical and Chemical
Changes in Sediments after Deposition. J. sediment.
Petrol., 12, p. 111-117.
LANG, W.H.Jr. (1978). - The Determination of Prior
Depth of Burial (Uplift and Erosion) Using Interval Transit
Time. SPWLA, 19thAnn. Log. Symp. Trans., paper B.
LANG, W.H.Jr. (1984). - Conductivity and interval
transit time as correlation tools. The Log Analyst, 25, 3, p.
21-33.
LEWIS, C.R., & ROSE, S.C. (1970). - A theory relating
high temperatures and overpressures. J. Petrol. Technol.,
22, p. 11-16.
MAGARA, K. (1968). - Compaction and migration of
fluids in Miocene mudstone, Nagaoka Plain, Japan. Bull.
Amer. Assoc. Petroleum Geol., 52, 12, p. 2466-2501.
MAGARA, K. (1969). - Porosity-permeability relationship of shale. J. Canad. Well Log. SOC.,2, 1, p. 48-74.
MAGARA, K. (1971). - Permeability considerations in
generation of abnormal pressures. SOC.Petrol. Eng. J.,
11, 3, p. 236-242, and paper SPE 28916.
MARTIN, G.B. (1972). - Abnormal high pressure and
environment of deposition. 3rd Symp. on Abnormal
Subsurface Pore Pressure, SPE of AIME, Baton Rouge,
La., preprint paper SPE 3846.
MASSE, P. (1971). - Relations entre les caracteristiques des couches d’argilites et leur profondeur maximale d’enfouissement. C.R. Acad. Sc. Paris, 272, 11, p.
1481-1484.
MATTHEWS, W.R. (1971). - How well logs indicate
whatpressures to expect. Oil and Gas J., 13 Dec., 69, 50.
MATTHEWS, W.R., McCLENDON, R.T., & SOUCEK,
C.R. (1972). - How to predict formation pressures of cretaceous-jurassicage sediments, Mississippi. 3rd Symp. on
Abnormal Subsurface Pore Pressure, SPE of AIME,
Baton Rouge, La.; preprint paper SPE 3893
MAXWELL, J.C. (1960). - Experiments on compaction
and cementation of sand. In : Rock Deformation (edited
by Griggs, D. T, & Handin, J.), Geol. SOC.Amer., Mem. 79,
p. 105-132.
MAXWELL, J.C. (1964). - lnfluence of depth, temperature, and geologic age on porosity of quartzose sandstone. Bull. Amer. Assoc. Petroleum Geot, 48, p. 698-709.
MAXWELL, J.C., & VERRALL, P. (1954). - Low porosity may limit oil in deep sands. World Oil, 138, (5) : 106113; (6) 102-104.
McCULLOH, T.H. (1963). - Facteurs contrdlant les
variations de densite dans les roches sedimentaires
Oligocenes. Abstr. G.S.A. and Ass. SOC.Joint Mtg, New
York.
296
McCULLOH, T.H. (1967). - Mass properties of sedimentary rocks and gravimetric effects of petroleum and
natural-gas reservoirs. U S . Geol. Surv. Prof. Paper,
528A, 50 p.
McGREGOR, J.R. (1964). - How abnormal pressures
behave in coastal and offshore Louisiana. Oil and Gas J.,
82, 37, p. 128-129.
McGREGOR, J.R. (1965). - Quantitative determination
of reservoir pressures from conductivity log. Bull. Amer.
Assoc. Petroleum Geol., 49, 9, p. 1502-1511.
MEADE, R.H. (1966). - Factors influencing the early
stages of compaction of clays and sands -review. J. sediment. Petrol., 35, p. 1085-1101.
MYERS, R.L., & Van SICLEN, D.C. (1964). - Dynamic
phenomena of sediment compaction in Matagorda
County, Texas. Gulf Coast Assoc. Geol. SOC.Trans., 14,
p. 241-252.
OTVOS, E.G. (1970). - High pressure shales and their
depositional facies, southern Louisiana Cenozoic. J. sediment. Petrol., 40, 1.
PHILIPP, W., DRONG, H.J., FUCHTBAUER, H., HADDENHORST, H.J., & JANKOWSKY, W. (1963). -The history of migration in the Gifhorn trough (NW-Germany).
Proc. 6th Wld Petrol. Cong., Frankfurt, section 1, pap. 19,
p. 457-481.
PICKETT, G.R. (1960). - The use of acoustic logs in
the evaluation of sandstone reservoirs. Geophysics, 25,
1, p. 250-274.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Houston.
POWERS, M.C. (1967). - Fluid-release mechanisms in
compacting marine mudrocks and their importance in oil
exploration. Bull. Amer. Assoc. Petroleum GeoL, 51, p.
1240-1253.
PROSHLYAKOV, B.K. (1960). - Reservoirproperties of
rocks as a function of their depth and lithology. Geol. Neftii
Gaz, 1960 (12), p. 2429.
RATHBUN, P.C. (1968). - Abnormal pressures and
conductivity anomaly. Northern Green River Basin,
Wyoming. 43rd ann. fall Mtg. SPE of AIME, Houston,
Texas.
RAYMER, L.L., HUNT, E.R., & GARDNER, J.S.
(1980). - An improved sonic transit time-toporosity transform. SPWLA, 21stAnn. Log. Symp. Trans.
REHM, B. (1972). - World wide occurrence of abnormal pressures. Part 2. 3rd symp. on Abnormal Subsurface
Pore Pressure, SPE of AIME, Baton Rouge, La., preprint
paper SPE 3845.
RIDER, M.H. (1986). - The Geological lnterpretation of
Well Logs. Blackie Halsted Press, Glasgow.
RIEKE, H.H. (1970). - Compaction of argillaceous
sediments (20-5OO;OOO PSI). Southern Calif. Univ. Ph.D
thesis, 720 p.; Abstr. no 72-21, 698, Diss. Abstr. Int. Sect.
B., 33, 2, p. 755-8-756-B.
RIEKE, H.H. Ill, & CHILINGARIAN, G.V. (1974). Compaction of Argillaceous Sediments. Developments in
Sedimentology, 16, Elsevier, Amsterdam.
ROBERTS, J.E. (1969). - Sand compression as a factor in oil-field subsidence. In : Land Subsidence, 1.A.S.H.Serralog Q 2003
Compaction
~~
Unesco Publ., 89, AIHS, 2, p. 368-376.
ROBERTSON, E.C. (1967). - Laboratory consolidation
of carbonate sediments. In: Marine Geotechnique (edited
by Richards, A.F.), Univ. of Illinois Press, Urbana, Illinois,
326 p.
RYER, T.A., & LANGER, A.W. (1980). - Thickness
change involved in the peat-to-coal transformation for a
bituminous coal of Cretaceous age in central Utah. J.
sediment. Petrol., 50, 3, p. 987-992.
SAHAY, B. (1972). - Abnormal subsurface pressures,
their origin and methods employed for prediction in India.
3rd symp. on Abnormal Subsurface Pore Pressure, SPE of
AIME Baton Rouge, La., preprint paper SPE 39100.
SAWABINI, C.T., CHILINGAR, G.V., & ALLEN, D.R.
(1974). - Compressibility of unconsolidated, arkosic oil
sands. SOC.Petrol. Eng. J., 14, p. 132-138.
Schlumberger (1979). - Well Evaluation Conference.
Algeria.
Schlumberger Ltd (1981). - Data Processing Services
Catalogue.
Schlumberger (1985). - Well Evaluation Conference.
Nigeria.
SERRA, 0. (1972). - Diagraphies et Stratigraphie. In:
Mem. B.R.G.M., 77, p. 775-832.
SERRA, 0. (1974). - Interpretation geologique des
Series deltaiques a partir des diagraphies differees. Rev.
A.F.T.P., 227, OCt., p. 9-17.
SERRA, 0. (1984). - Fundamentals of Well-Log
Interpretation. Volume I : The Acquisition of Logging
Data. Developments in Petroleum Science, 15A, 440 p.,
Elsevier, Amsterdam.
SERRA, 0. & SERRA, L. (2000). - Diagraphies.
Acquisition & Applications. Editions Serralog, Herouville,
France.
SERRA, O., CHIARELLI, A., & GRAS, C. (1975). Automatic study of clay compaction : the DENSON program. The Log Analyst, Jul.-Aug., 16, 4, p. 3-12.
SMITH, J.E. (1971). - Shale compaction. 46th Ann. fall
Mtg SPE of AIME, New Orleans, preprint paper SPE
3633.
SMITH, J.E. (1971). - The dynamics of shale compaction and evolution of pore fluidpressures. J. Int Ass. Math.
Geol., 3, 3, p. 239-263.
SMITH, J.E., & THOMAS, H.G. (1970). - Origins of
abnormal fluid pressures. 20th ann. AAPG Guff Coast
Sect. and SEPM Gulf Coast Sect. Conv., Shreveport,
paper 1970. Abstr., Bull. Amer. Assoc. Petroleum Geol,
54, 9, p. 1791.
SPURLOCK, W.T. (1965). - The relationship between
compaction of sedimentary rocks and high formation fluid
pressure. Tenneco Oil Co.
TAYLOR, J.M. (1950). - Pore space reduction in
sandstones. Bull. amer. Assoc. Petroleum Geol., 34, p.
701-716.
TERZAGHI, K., & PECK, R.B. (1948). - Soil mechanics in engineering practice. 2nd ed. John Wiley & Sons,
Inc., New York.
TIMKO, D.J. (1965). - Practical applications of log calculated pressures as an aid to solving drilling problems.
Serralog 0 2003
1
Chapter7
I
297
~~
The Log Analyst, Sept. -Oct., p. 37-54.
TIMKO, D.J., & FERTL, W.H. (1970). - Hydrocarbon
accumulation and geopressure relationship and prediction of well economics with log-calculated geopressure.
SPE of AIME, 45th ann. fall Mtg., Houston, Texas; et in J.
Petroleum Technol., 23, p. 923-933, 1971.
WALLACE, W.E. (1965). - Abnormal subsurface pressures measured from conductivity and resistivity logs. The
Log Analyst, 5, 4, and Oil and Gas J., July, 63, 27.
WEEKS, L.G. (ed) (1958). - Habitat of oil. Amer. Assoc.
Petroleum Geol,
WELLER, J.M. (1959). - Compaction of sediments.
Bull. Amer. Assoc. Petroleum Geol., 43, 2, p. 273-310.
WILSON, M.D., & PITTMAN, E.D. (1977). - Authigenic
clays in sandstones: recognition and influence on reservoir properties and palaeoenvironment analysis. J. sediment Petrol., 47, 1, p. 3-31.
WYLLIE, M.R.J., GREGORY,A.R., & GARDNER, L.W.
(1956). - Elastic wave velocities in heterogeneous and
porous media. Geophysics, 21, 1.
297
K
W
F
L
a
I
0
18
WELL LOGGING AND TECTONICS
Introduction
After formation, hydrocarbons migrate under the
action of compaction, from the source rock towards their
ultimate reservoirs. This mechanism is called primary
migration. Once in the reservoir rock they continue to
move, a process called secondary migration, settling out
above the denser water occupying the porous spaces of
the rock. The rate at which this process takes place
depends on hydrodynamic gradient within the reservoir
which is controled by the permeability. As the hydrocarbons rise, displacing the formation water, they may
encounter permeability barriers that constitute traps,
under which they accumulate and form hydrocarbon-bearing strata.
Traps are usually classified into three categories (Fig.
8-1 next page):
- structural traps;
- stratigraphic traps;
- mixed or diverse traps.
Among them, structural traps play the major role in
accumulating hydrocarbons. According to Halbouty
(1976), 78 % of the 306 major hydrocarbon fields in the
United States (more than 14 million tons of oil or gas equivalent, for each one) are related to the existence of structural traps. According to Perrodon (1980), this type of trap
represents 89 % of the 266 biggest fields in the world
(more than 70 million tons of oil or equivalent of gas).
These traps are formed by the action of forces that deform
the rocks. Rocks with very low permeability or cap rocks
serve as the closure of these traps.
Stratigraphic traps, representing 10 % of the biggest
fields in the world, correspond to the permeability barriers,
formed by lateral variations of facies, stratigraphic pinchouts of the rocks, or unconformities.
The third category, mixed or diverse traps, are linked
to diagenesis or related to differential pressure.
The drilling of exploration, appraisal and development
wells and the analysis of well logs recorded in them, in
particular dipmeter or image data, leads to the establishment of correlations and the study of pressure gradients
throughout the reservoir. The data may be also used to
complement high resolution seismic. All of these methods
Serralog 0 2003
allow the picture of the subsurface structure to be refined
to a high level of precision.
Analysis of the correlations between 3 wells, X, Y and
Z (Fig. 8-2) results in the definition of the structural form
represented in Figure 8-2a. In this case the information
obtained from dipmeter logs modified, appreciably, the
final structural features, particularly the position of the
anticline crest (Fig. 8-2b).
Certain structures in the North Sea (such as Piper, Fig.
8-3), originally interpreted as anticlines or as pinching out
layers under an unconformity, appeared as more a complex series of faulted blocks as a result of the study of dipmeter data.
It is necessary to emphasize that the interpretation of
dipmeter or image data cannot be carried out without knowledge of lithology, environment and the tectonic features
of the basins in which the drilling is carried out. Moreover,
it seems that knowledge of some of the essential
concepts concerning the deformation mechanisms, the
mechanical behavior of the rocks, and the relation between the types of deformation and the sedimentary
basins is necessary. These concepts allow a better
understanding of the problems confronted. Some of these
concepts will be reviewed briefly in Chapter 9. Others are
reviewed hereafter prior to discussing the actual interpretation of dipmeter or image data.
Review of geological concepts
Definition
Structural geology is "the branch of geology that deals
with the form, arrangement, and internal structure of the
rocks, and especially with the description, representation,
and analysis of structures, chiefly on a moderate to small
scale" (Bates & Jackson, 1980).
Tectonics is "a branch of geology dealing with the
broad architecture of the outer part of the Earth, that is,
the regional assembling of structural or deformationalfeatures, a study of their mutual relations, origin, and historical evolution. It is closely related to structural geology, ...,
but tectonics generally deals with larger features" (Bates
& Jackson, 1980).
299
Figure 8-1 - The various types of trap (adapted from Penn Well).
w
WELLX
Figure 8-2 - (a) Structure deduced from the correlations between wells. (b) Structural cross-section after introducing dipmeter data
(from Schlumberger, Well Evaluation Conference, Venezuela, 1980).
PH
A
Figure 8-3 - (a) Map showing the position of the cross-section across the Piper field. (b) Cross-section across the Piper Field, North Sea.
300
Serralog Q 2003
Tectonics
A structure, in the tectonic sense of the term, is "the
general disposition, attitude, arrangement, or relative
positions of the rock masses of a region or area" (ibid.).
A fold is a continuous deformation, a "curve or bend of
a planar structure such as rock strata, bedding planes"
(ibid.).
A fault is "a fracture or a zone of fractures along which
there has been displacement of the sides relative to one
another parallel to the fracture" (ibid.). This is a discontinuous deformation that acts at the occasion of surfaces of
weakness, irrespective of any deformation of the formations on either side of the fault.
Concepts of stress1
Every element of a rock is subject to a series of forces.
These forces are of two types.
- The first type corresponds to the forces that are
applied to the whole body of the rock. These are called
body forces, and are proportional to the mass of the substances, e.g. gravity, centrifugal forces, magnetic forces.
They are measured in force unit per unit volume (dimension : mLT-2)'.
- The second type are known as surface forces. They
act on the surface of a body and, because of this, are
measured in force units per unit of surface area (dimension: mLT-2/L2 = mL-IT-2). "ln a solid, the force per unit
area, acting on any surface within it", is termed stress
(Bates & Jackson, 1980). Stress is equivalent to a pressure, in which the SI unit is pascal. Taking into consideration all the elements of a rock or bed (Fig. 8-4), the surface forces acting on any imaginary surface are represented by:
- the weight of the above sediments, or the geostatic
pressure, S, and the reaction of the material below;
the fluid pressure pp; if the fluid is in equilibrium (no
movement) the fluid pressure is equal to the hydrostatic
pressure;
- the tectonic forces, T.
1
ChaDter8
I
301
pe
Figure 8-4 - Surface forces acting on a body.
Torsion is "the state of stress produced by two force
couples of opposite moment acting in different but parallel planes about a common axis" (Fig. 8-5d).
Tension
Distortion
-
One must distinguish between the external forces that
act on a body, and the resulting internal actions and reactions that constitute the stress. If the forces acting on a
body are equal on all sides, the body is in equilibrium. The
all-sided pressure is called the confining pressure, C.
In many cases the forces acting on a body are not
equal in all sides. This will cause deformation. If the external forces tend to pull a body apart, the body is said to be
under tension. If it is subjected to external forces that tend
to compress it, it is said to be under compression. If two
equal forces act in opposite directions in the same plane,
but not along the same line, we have a couple, and the
body is said to be under distortion (Fig. 11-4).
This part of the chapter has been reviewed originally by Prof. F.
Proust and Dr. A. Etchecopar, I want to extend my thanks to their unvaluable contributionwhich helped me to issue a text without errors or mistakes.
m - mass, L - Length, T - Time.
*
Serralog 0 2003
Torsion
with variation in rotation
with regards with a same axis
figure 8-5 The four types of forces which can act on a body.
Let us take A as a point in a rock (Fig. 8-6), and C as a
small plane surface, defined by thgntegection of a plane
P passing throughA+A pressure, p = AF/hc will act on C.
We can break the p down into two components: ( 0 ) normal to X, called the normal stress, and (T),parallel to C,
called shear stress.
+
Generally, the pressure p as well as o and T, vary in
magnitude and direction depending on the orientation of
the surface on which they are applied. The set of all the
pressures exerted on point A on all planes that pass
through this point is called the state of stress.
301
the principal stress axes. On these three mutually perpendicular axes, the three principal stresses are as follows (Fig. 8-8):
- greatest or maximum principal stress, o1
- intermediate principal stress, 02;
- least or minimum principal stress, 03;
with 01>02>03.
Figure 8-6 - The resolution of pressure, p, acting on a point, A, of a
surface, 2, into two components: 0,normal to Z,called the normal
stress, and 7,parallel to Z,called shear stress.
The state of stress at any point may be described in
terms of nine stress components of which only six are
independent if the body is in equilibrium. The stresses on
each face of a cube (Fig. 8-7) can be resolved into three
parts, one normal stress, and a shearing stress which
itself can be resolved into two components parallel to the
direction of two of the coordinates.
A
B
D
C
Figure 8-8 - Principal stress axes and the stress ellipsoid.
When the normal stresses are equal no shearing
stresses exist in the material. This state of stress is known
as hydrostatic stress. When they are different, shearing
stresses appear. The geometric representation of the
state of stress at a point is known as the stress ellipsoid
(Fig. 8-8). One can demonstrate that six planes of maximum shearing stresses exist associated in pairs each pair
countaining one of the principal axis, and forming between them an angle of 90” (Fig. 8-9).
Figure 8-9 - Planes of maximum shearing stress.
Figure 8-7 - The stress components acting on the faces of a cube
(from Ramsay, 1967).
There is no direct way to measure the stresses in a
body, but they may be calculated if the external forces are
known.
But it is possible to calculate all the stresses at any
point of the body if the applied stresses at this point on
three mutually perpendicular planes are known. It is also
possible to demonstrate that at each point A, there exist
three orthogonal planes, called principal planes of stress,
for which the shear stress z = 0, and therefore the stress
is perpendicular to them. They constitute symmetry planes for the state of stress.
The three normal vectors to these planes are called
302
4
Figure 8-10 - Planes of maximum shearing stress (S1and SZ, and planes of rupture (Fland FZ, forming an angle 0, close to 30’ to the
maximum principal stress.
Serralog 0 2003
Tectonics
The greatest shearing stress always occurs on the planes which contain o2 axis (2 is maximum the stress difference, o1 - o3being maximum), and make an angle of 45"
to the principal stresses o1and o3irrespective of the signs
or values of the principal stresses (ruptures and slippages
are produced more or less along these planes, Figs. 8-8
and 8-10). In fact, fractures form an angle 8 less than 45"
and close to 30" with the principal axis. By reference to
Coulomb's work, this can be related to the concept of
internal friction which suggests that, at failure, the relationship between the magnitude of shear stress 121 and
normal stress o is :
where T~ is the cohesive strength (sometimes expressed as c for cohesive);
p being the coefficient of internal friction of the material which is related to the angle of internal friction @ by :
1
Chanter 8
I
303
tangents defines the angle of internal friction @,for each
state of stress.
Strain is the deformation caused by stress. This deformation may correspond to a change in volume which is
called dilation Or ComPresSion. It may also ~ ~ uin l at
change in shape: distofiion.
01
@ being related to 8 by the following equation:
The relation between stress and rupture may be determined graphically by the Mohr stress circle (Fig. 8-11)
which is a graphic representation of the state of stress.
(a)
(b)
(C)
Figure 8-12 - Marble cylinders deformed in a laboratory by compression. (a) :undeformed; (b) :20% strain, 270 atm. confining pressure;
(c) :20% strain,445 atm. confining pressure. 01 indicates the direction
of the maximum principal stress (adapted from Press 8, Sieve6 1978).
Mechanical behavior of rocks
Every stress field imposes a strain field, but the resulting deformation also depends on the nature and the
mechanical behavior of the deformed medium.
There are three principle mechanical behaviors.
Figure 8-11 - Mohr stress circle.
To determine the cohesive strength and angle of internal friction, a series of experiments with different values of
the confining pressure must be run on cylinders submitted
to compression tests (Fig. 8-12), and the results reproduced as a Mohr stress circle (Fig. 8-13). The lines drawn
tangent to the successive circles define the Mohr stress
envelope. Their intersection with the vertical axis define
,
corthe cohesive, or shear, strength of the rock T ~ which
responds to the "inherent strength of a material when normal stress across the prospective surface of failure is
zero" (Bates & Jackson, 1980). The slope of each of these
Serralog 0 2003
- Elastic behavior
This behavior is characterized by a possible return to
the initial state. Deformation appears immediately after
the force is applied and strain does not build up. The
deformation obeys Hooke's law, which states that strain is
proportional to stress. The solid regains its dimensions
and its shape when the stress is removed (Fig. 8-14a).
However, this return to the initial shape is not necessarily
immediate, and may indeed take some time. An elastic
solid stands up until a certain limiting stress, called the
elastic limit. If this is exceeded, the solid does not return
to its original shape. When the stress exceeds the elastic
limit, the deformation is plastic. It means that the solid
only partially returns to its original shape (Fig. 8-14e).
When the stress increases, at a certain value the solid
303
304
Well Logging and Geology
fractures. We reach the rupture point The relation existing
between stress and strain is expressed by a stress-strain
diagram (Fig. 8-15).
E
f
(d
Norm01strou in kilotun
Figure 8-15 - Stress-strain diagrams for different rock behaviors.
A :elastic; B :elasto-plastic; C :elasto-plastic with strength hardening;
D :actual elasto-plastic (from Billings, 1972).
tensile strength = 0
tensile strength at low stress,
ductility at high stress
b
5
Figure 8-13 - (a) :Mohr stress envelope (adapted from Billings, 1972).
(b) :Different types of Mohr stress envelopes in relation with the rock
type: (A) :wet clay; (6) :dry sand; (C) :rock materials
(adapted from Ramsay, 1967).
Visco-elastic behaviour
10
Figure 8-16 - Differential stress (0, - od versus strain diagrams explaining the transition from brittle to ductile behavior when the confining
pressure increases (od
Elasto-plastic behaviour
Figure 8-14 - Ideal mechanical models and stress-strain or strain-time curves for material behaviour
(adapted from Ramsay, 1967; Billings, 1972; and Hatcher, 1990).
304
Serralog 0 2003
Chanter 8
The resistance of a material to elastic deformation is
defined as the stress-strain ratio. This ratio is the Young's
modulus E:
E = CJ,/E~
fO
(8-4)
Figure 8-16a - Young's modulus.
with
CJ, = uniaxial compressional stress
E~ = strain. E~ is equal to the ratio of the change in
length, AI, to the original length, lo.
I 305
- Plastic behavior
As previously explained, deformation is permanent
only above a certain threshold. Before this point is reached the substance behaves elastically (Fig. 8-14e).
Plastic deformations result from processes such as intergranular movements, dislocation gIide (intragranular
movements), and recrystallization (including diffusion).
The rheologic model is a mass moving with friction.
Movement will only take place above a certain value of
traction (Fig. 8-14c).
- Viscous behavior
Rigidity measures the
resistance to change in
shape:
(8-6)
G=z/y
F
Figure 8-16b - Shear modulus.
where G is the rigidity or shear modulus, z (= F/A) the
shear stress, and y (= u/l) the shear strain.
G may also be expressed in another way, in terms of
Young's modulus and Poisson's ratio:
G = E / [2(1 + v)]
In viscous material deformation appears immediately
and the strain is unrecoverable (Fig. 8-14b). The shear
modulus is zero in fluids as they do not support a shearing force. Also there is no shear wave propagation and
the Young's modulus has no meaning because stretching
implies shearing.
(8-7)
Fr3
where v is the Poisson's
rat;o equal to the ratio of Figure 8-16c - Poisson's ratio.
transverse strain to axial strain in elastic deformation by
uniaxial stress. In an elastic solid, v is a constant. For isotropic materials, v is between 0 and 0.5.
Viscosity, q, is the property that has a substance to
offer internal resistance to flow. It is equal to the ratio of
the shearing stress, z, to the rate of shear strain, y, per
unit of time, or dy/dt. The rate of shear strain, y, is measured by the change in angle y~ per unit of time t (Fig. 818):
Y=tgW
(8-10)
W
where Ad (= w) is the change
in diameter (equal to E ~ ) .
The bulk modulus or incompressibility K is given by :
K = C J /~ E = Ah I (AVNO) (8-9)
P
where Ah is the change in
hydrostatic pressure, p, and AV Figure 8-16d- '"Ik
the change in volume compared to the original volume Vo.
The rheologic model of an elastic body is a perfect
spring without mass (Fig. 8-14a).
Stress
b
7gure 8- 17 - In a viscous
material its strain is a
function of time and the
rapidity of its strain is a
function of its viscosity.
Deformation speed
Serralog 0 2003
Figure 8-18 - The rate of shear
strain y is measured by the
angular shear variation yJ (from
Billings, 1972).
The viscosity unit is called poise. Viscosity is very high
for rocks but decreases when temperature increases
(Table 8-1). Viscosity is an important property in geological processes. It determines, for example, the flow of
magma or lava during intrusive or volcanic activity, and
the velocity of displacement in plate tectonics.
Water at 100°C and one atmosphere
Water at 30°C and one atmosphere
Water at 0°C and one atmosphere
Corn syrup, room temperature and pressure
Roofing tar, ready to apply
Lava, Mt. Vesuvius, 1400°C
Lava, Mt. Vesuvius, 11 00°C
Rock salt, near surface
Rocks in general
Mantle of earth
0.00284
0.00801
0.01792
7 x 102
3 x 107
2.58 x lo2
2.83 x 104
1017
1017 to 1022
1023
The rheologic model for viscous behavior is a damper,
a perforated piston moving without friction in a fluid (Fig.
8-14b).
305
Factors controlling rock behavior
In addition to their inherent properties (mineralogy, texture, structure), the mechanical behavior of rocks is
controlled by several factors such as confining pressure,
temperature and time.
jected to very short duration stresses, becoming plastic if
these stresses are applied over a long time. This effect is
observed in creep experiments, where a small load
applied for a sufficiently long time produces a strain that
may continue and eventually cause rupture. The same
stress in instantaneous tests would not cause any measurable strain. Figure 8-21 illustrates an ideal creep curve.
Confinina pressure
The strength of a rock increases with the confining
pressure. Figure 8-19 illustrates the effect of confining
pressure on the breaking strength of several standard
rocks. At low confining pressure, all the rocks deform only
a few percent before fracturing. Under a high confinin
pressure, we observe a different behavior for the rocks.
I
25
I
I
I
Shale.
W.ln
~
Shondng in pwernt
Figure 8-20 - Effect of temperature on deformation of marble
(from Griggs, 1939).
tA
Figure 8-19 - Effect of confining pressure on the ductility of several
common rocks (after Donath, 1970).
When fractures appear at less than 3-5 % plastic
deformation, the rocks are said to be brittle. When rocks
are able to sustain, under a given set of conditions, 5-10
% plastic deformation before fracturing, they are ductile.
Ductility is "a measure of the degree to which a rock exhibits ductile behavior under given conditions, commonly
expressed by the strain at which fracture commences"
(Bates & Jackson, 1980). As a consequence, when the
confining pressure increases a brittle rock becomes ductile (i.e. limestone).
TemDerature
The elastic limit decreases when the temperature
increases. Moreover, less stress is necessary to produce
a given strain when the temperature increases (Fig. 8-20).
Time
Time plays a very important part in the behavior of the
rocks. Rocks may exhibit elastic behavior if they are sub306
S = A + B log t + Ct
+D
I
Tim.
-.
-rgure 8-21 - Ideal creep cuwe. A :instantaneous deformation. B :primary creep. C :secondary creep. D :tertiary creep
(from Billings, 1972).
The actual behavior of rocks
In nature, rocks have a complex behavior of all three
types of response visco-elasto-plastic. One of these components may dominate according to physical conditions
(temperature and pressure) and the way the stress is
applied.
At low temperature the elastic deformation of the crystal of quartz shows an almost perfect reversibility.
Rocks which show a good reversibility and admit the
greatest elastic deformation are:
- quartzite, plutonic rocks;
- slates.
Such rocks are brittle.
Semlog 0 2003
Next Page
Tectonics
Some other rocks are more or less ductile, or show an
elastoplastic behavior. Few rocks, such as halite and
undercompacted shales, may have a plastic to viscous
behavior.
According to the previous factors, it is possible to
determine the different kinds of strain following the depth:
- an upper zone, where most of the rocks have an
elastic (brittle) behavior and consequently will fracture
under a certain value of stress;
- an intermediate or middle zone, where the rocks
have an elastoplastic to elastoviscous behavior (ductile);
- a deep zone, where rocks will show a plastic behavior. This zone is characterized by the appearance of
schistosity, and then of foliation. It corresponds to anchimetamorphism and to metamorphism. This type of rock
has no interest in oil exploration, since porosity and permeability disappear.
Tvpes of state of stress
These are three types of state of stress:
- tension or traction: stretches the material and may
increase its volume;
- compressional: leads to a decrease in the volume of
the material;
- pure shear stress: produces a change in shape, but
not in volume (Fig. 8-22).
1
Chapter 8
I 307
Table 8-2
Compressive, tensile, and shearing strengths of
some rocks in kg/cm*
(from Billings, 1942).
I
Rock
Compressive
Sandstone
Limestone
Granite
Diorite
Gabbro
Basalt
Felsite
Marble
Slate
100 to 1400
1 000 to 2 800
1 000 to 2 600
1 000 to 1 900
2 000 to 3 600
2 000 to 2 900
800 to 1 600
700
I
Tensile
~
Shearing
~~
10 to 30
30 to 80
30 to 60
..................
..................
.................
..................
30 to 80
260
60 to 160
100 to 200
150 to 300
....................
....................
....................
....................
100 to 300
160 to 260
General orientation of stresses
The diagrams in Figure 8-23 represent the various
stress states which may be encountered and the main
structures that will be formed by those stresses, as well as
the type of sedimentary basins associated with them.
DIRECTION OF STESSES
01
RESULTS
TENSION
Compression
COMPRESSION
Sup.rficiJm(btirtkroeb1
(minimalhorimul stress)
thrust fwb
Fold axis
Figure 8-22 - Deformation may produce change in volume without
change in shape or change in shape without change in volume, or a
combination of the two (adapted from Leet et al., 7982).
Rock strenath
Rocks are more or less resistant to stresses. The
strength of a rock corresponds to the stress at which the
rock starts a permanent deformation.
Rocks show different types of strength, because they
respond differently to various stresses. Hence, there is,
for each rock, a compressive, tensile and shear strength.
The compressive strength for a brittle rock is sometimes 10 to 30 times more than its tensile strength (Table
8-2).
Serralog 0 2003
Figure 8-23 - Relations between structures, the forces acting upon
them and the type of basin.
307
Previous Page
h = (b/4)2 = h2 = (1 + e)2
The results of stresses : strains
As previously seen, stress deforms the bodies, causing variations in dimensions and the angles between
their different lines or planes (Figs. 8-24 and 8-25).
(8-13)
- the logarithmic, natural or true strain, E, is the logarithm of h:
(8-14)
&=
h
-the angular shear, u,, and the shear strain, y, that express the angular variations, are related by:
y = tangy
A. Unstrained
state
8. Homog.mour
strain
C. lnhomogmour
strain
Figure 8-24 - Various types of two-dimensional deformation
(from Ramsay, 1967).
Undeformed
Deformed
(8-10)
With strain acting on three dimensions (Fig. 8-25), one
considers the coordinate variations of a point (x, y, z)
before and after strain (xi, y,, z,).
Normally, strains are classified in two types, according
to the following geometrical criteria
- Homogeneous strains (Fig. 8-24B), characterized by
the following properties:
. straight lines remain straight after strain;
. parallel lines remain parallel after strain (u,cons
tant);
. all lines in the same direction have the same
values of e.
- lnhomogeneous strains (Fig. 8-24C). In this case we
observe that:
. straight lines become curved after strain;
. parallel line lose their parallelism;
. the values of e and u, vary in any given direction
of the strained bodies.
Strain results can be classified in two important categories.
Figure 8-25 - Deformation due to a finite force in three dimensions
(from Ramsay, 1987).
The reaction of rocks to stress falls into two categories
- continuous strains : fold and flow;
- discontinuous strains which are fractures (studied in
Chapter 9),faults (studied in this chapter), and pressuresolution (stylolites) studied in Chapter 6.
The measurement of these changes enables determination of the type of strain.
When the strain acts on two dimensions (Fig. 8-24),
we have:
- the extension, e, that is defined as the change in
length;
If we express the ratio
e=h-I
b/&as equal to h we have:
(8-12)
- the quadratic elongation, 1,defined as the square of
the ratio of the lengths before and after the strain;
308
Continuous strain : fold
A fold is a flexible strain of materials submitted to compressive or tensile tectonic stresses.
Tensile stress
A normal fault in a basement may be reflected in the
cover by flexure of the overlying beds (Fig. 8-26a).
Differential compaction (Fig. 8-26b), or vertical movements: diapiric uplifts (diapiric shales, mud lump, or halokinesis), and magmatic intrusions, present the same
situation.
ComDressive stress
All types of folds are possible, according to the mechanism of their formation (flexure by compression or by
action of a torque or folding by shearing), the mechanical
behavior of rocks, and their competency. The latter is a
relative property that indicates the capacity of a rock to
withstand compression without variation in thickness. A
competent rock will transmit a compressive force much
farther than a weak, incompetent rock. Hence, more competent beds will bend under plastic strain and slip over
other beds, under the effect of compression. The incomSerralog 0 2003
Tectonics
petent layers that are more passive may be subject to
complex strains. In the most simple case, the fold will be
an isopach type (Fig. 8-27b see the definition, later). We
will have similar folds (Fig. 8-27c) with developing schistosity in the most complex case, or in connection with a
viscous flow.
A fold usually has a general cylindrical symmetry (Fig.
8-28). However, regardless of the extent of a fold - it cannot be infinite - each extremity must develop into a conical structure (Fig. 8-29). Domes (diapiric uplift) generally
have conical symmetry. The apical angle is the characteristic element of a conical fold.
1 Chapter 8 I 309
_I
Figure 8 - 2 7 ~ Example of similar (isogonal) fold [from Ramsay, 7967).
Hinge point
Trough
over a fault
Figure 8-28- The geometrical elements of a cylindrical fold.
over a sand lens
Photograph of a drapping
of a sand lens
Figure 8-26 - Example of tension folds.
Fold4 Cbss'2
Similar
Figure 8-29 - ,The geometrical elements of a conical fold.
-
DescriDtive element of a fold (Fig. 8-30)
- The crest is the highest point of a given stratum in
Fold 5 Clssdi3
Figure 8-27a - Basic fold types (from Ramsay, 1967).
East
Figure 8-27b - Example of
parallel [isopach) fold. The
thickness of the compact beds
is Constant not the anbe oftheir
surface
(from Ramsay, 1967).
L
Serralog Q 2003
any vertical section through a fold and from which the surface slopes downward in opposite directions.
- The trough is the lowest point of a given layer in any
section through a fold.
- The hinge corresponds to the locus of the fold's surface, where the curvature is at a maximum. A box fold has
two hinges, while a perfectly circular fold has no hinge.
- The hinge line is a line joining the points of flexure or
maximum curvature of the bedding planes in a fold.
- The crest line is "a line joining the crest points in a
given stratum" (Bates & Jackson, 1980).
- The crestal surface is "a surface that connects the
crest line of the beds of an anticline" (Bates & Jackson,
1980).
- The axial surface is a surface that joins the hinge
lines of the various beds in a fold. In a vertical fold, the
axial and the crestal surfaces are indistinguishable; in an
asymmetric fold those surfaces can be distinguished from
one another.
- The fold limbs or flanks are the sides of a fold. They
309
Plunge of hinge line
Figure 8-30 - Descriptive elements of a fold (from Ramsay, 1967).
correspond to that area of a fold between two successives
hinges. They generally have a greater radius of curvature
than the hinge region. They may be planar.
- The plunge is the angle between the fold axis and the
horizontal plane measured in the vertical plane. It may be
as great as 90". This is the case of folds with a vertical
axis, which are observed in vertical series deformed by
the passage of strike slip (Fig. 8-31).
Synfom
Figure 8-31 - A fold with a vertical Figure 8-32 - The different forms of
axis.
fold.
- The angle of folding or interlimb angle is the dihedral
angle formed by the two planes that are tangent to the two
limbs of a fold at their inflection points.
- The bisecting plane is the plane that divides the angle
of foldina into two equal parts.
Fold shaDe (Fig. 8-32)
We can distinguish two main shapes of folds
- antiform folds that present a convex curvature in an
upward direction : the limbs close upward; an anticline is
a fold, generally convex upward, whose core contains the
310
stratigraphically older rocks.
- synform folds that present a concave curvature
upward : the limbs close downward. A syncline is a fold,
generally concave upward, of which the core contains the
stratigraphically younger rocks.
If the folded layers retain their correct depositional
sequence in the structure, the same folds are termed antiformal anticline or synformal syncline. If they are in an
inverted sequence, they are termed synformal anticline or
antiformal syncline.
Fold t w e
We can observe, either a homogeneity in the thickness
of the beds (elastic, brittle rocks), or a variation in thickness owing to creep or flowage (plastic rocks), depending
on the mechanical behavior of rocks during folding.
- Parallel or isopach folding (Fig. 8-27a, fold 2 and Fig.
8-27b), in which the normal thickness of a bed, i.e. the
thickness measured perpendicularly from top to bottom,
remains constant. The radius of curvature decreases with
depth. This is the most frequent type of fold formed at
shallow or medium depths.
- Similar or isogonal folding (Fig. 8-27a, fold 4 and Fig.
8-27c) "in which the orthogonal thickness of the folded
strata is greater in the hinge than in the limbs, but the distance between any two folded surfaces is constant when
measured parallel to the axial surface" (Bates & Jackson,
1980). The form of the fold does not vary with depth (Fig.
8-33). These folds are formed at great depth and in the
zone of schistosity.
Nomenclature of folds
In addition to the above-mentioned classifications (by
curvature and by type), folds are also classified as a function of their geometrical shape through a transversal section. Figure 8-34 illustrates the main folds of this classifiSerralog 0 2003
Tectonics
cation with their nomenclature and with the theoretical
representationof dips in a plane perpendicular to the axis.
Figure 8-33 - In the similar fold
the thickness is greater in the
hinge than in the limbs, but the
distance between any two folded surfaces is constant when
measured parallel to the axial
surface which is vertical in this
figure.
AP
I
Chapter 8
I
311
Movements alona a fault
We have seen that faults are fractures with displacements. The displacement or throw of the opposite parts of
the fault ranges from a few centimetres to several hundred kilometres, while its length may be of the same
orders.
The relative movement along a fault may be translational or rotational (Fig. 8-35). When the movement is
translational all straight lines in each block that were
parallel before faulting remain so afterwards. The movement may be vertical (dip slip fault), horizontal (strike slip
fault), or mixed ( diagonal slip fault). The dip of beds on
either side of the fault remains unchanged.
With rotational movement, certain elements in each
block that were parallel before dislocation are no longer
so. The dip of the beds in each block is different.
Figure 8-34 - Nomenclature of folds
We can distinguish the following folds
- A symmetrical or upright fold is a fold with a vertical
axial surface.
- An asymmetrical fold has an inclined axial surface
with limbs that dip in opposite directions and with different
angles.
- An overturned fold has an axial surface, and its two
limbs, inclined in the same direction, usually at different
angles.
- A recumbent fold has a nearly horizontal axial surface.
- A fan fold has a broad hinge region and limbs that are
overturned and converge away from the hinge.
A fold is called isoclinal when its limbs are parallel to
one another.
Discontinuous strain : fractures and faults
These deformations represent the surfaces of a distinct discontinuity, along which cohesion has been lost.
The materials between these surfaces may not be deformed.
- Fracture is a general term that indicates all breaks or
ruptures in a rock, with or without displacement.
- Fault is "a fracture or a zone of fractures along which
there has been displacement of the sides relative to one
another parallel to the fracture" (Bates & Jackson, 1980).
Calling a fault a fracture depends on the scale of observation. Due to the importance of fractures on reservoir
properties (especially permeability), a complete chapter is
devoted (cf. Chapter 9) to the study of fractured reservoirs. Hereafter, our discussion will be limited to faults.
Serralog 0 2003
Figure 8-35 - Different types of faults as a function of the movement of
each block.
Fault description
Afault is described (Fig. 8-36) by defining the following
elements
- its orientation or strike which is the direction or trend
of the line of intersection of the fault "plane" with a horizontal plane, with respect to North;
- its net slip which defines the amount of displacement.
It is divided into:
. dip slip, " c b - "ad", the component of the movement or the net slip measured parallel to the dip of the
fault plane. Further subdivided into
. throw, "ae", corresponding to the displacement
measured along a vertical direction (vertical slip),
. heave, "ad", equal to the displacement in a horizontal direction (horizontal dip slip or horizontal throw);
. strike slip, "ac", which corresponds to horizontal
displacement measured in a direction parallel to the strike
of the fault;
311
- rotation, that is the angle formed by both parts of the
same bed after separation, and measured on a plane perpendicular to the axis of rotation;
- dip, that is the angle between a horizontal surface
and the plane of the fault.
Other elements may be used to describe a fault. Pitch,
for instance, is the angle, measured in some specified
plane, that a line on this plane makes with the horizontal
of the same plane. In Figure 8-36 the angle bac, measured on the fault plane, is the pitch of the net slip.
The two parts separated by the fault plane are the
blocks. The upper side is the hanging wall, the lower side
the footwall. The faulted surfaces of the two blocks parallel to the fault plane are the lips. They are sometimes
polished or striated in the direction of displacement. This
polished surface worn away by erosion is then known as
slickenside. The upthrow of a fault is the upthrown side of
a fault.
ab : net slip
ad = cb : dip slip
ae : vertical slip (a throw s)
ed : horizontal dip slip (a heave s)
ac : strike slip
Figure 8-37 - Classification of fault movements.
- longitudinal faults strike parallel to the strikes of the
regional structures;
- transverse fauts strike perpendicularly or diagonally
to the strikes of regional structures.
Another classification is based on the attitude of adjacent beds relative to the fault plane
- a conformable fault dips in the same direction as the
affected beds;
- an unconformable fault dips in the opposite direction
to the affected beds;
- a bedding fault has its plane parallel to the bedding.
Faults are also classified by shape, being plane or
warped. Warped faults include listric faults, which are
concave spoon shaped, usually pointing upwards. A
growth fault (Fig. 8-38) "forms contemporaneously and
continuously with deposition, so that the throw increases
with depth and the strata of the downthrown side are thicker than the correlative strata of the upthrown side" (Bates
& Jackson, 1980).
Figure 8-36 - The descriptive elements of a fault.
Fault classification
There are several methods of classifying faults. They
can, for example, be classified according to their relative
movement along the fault (Fig. 8-19).
Thus we have:
- gravity or normal faults, sometimes called direct
faults, in which the hanging wall appears to have moved
downwards relative to the footwall;
- thrust or reverse faults in which the hanging wall has
moved upward relative to the footwall. This situation leads
to repeating series;
- strike-slip faults. The movement is predominantly
horizontal and the terms dextral or sinistral are used if the
displacement of each block as seen from the other block
is to the right or to the left respectively;
- oblique or diagonal faults are those that strike obliquely or diagonally to the strike of the dominant structure;
312
1
Figure 8-38 - Example of growth fault and explanation of its formation.
On the side photograph of an experiment.
Serralog 0 2003
Finally, faults can be classified according to their layout
pattern caused by the stress system invoked. Parallel,
radial, en echelon and peripheral faults are shown in
Figure 8-39.
\
Figure 8-39 - Geometrical classification of faults
-
Determination of structure Traditional geological approach
On outcrops, in favorable cases the determination of
tectonic structures can be achieved by direct observation
(Figs. 8-40 to 8-42). But, due to the size of the structure
or the fact that the observation is on a plane or sometimes
partly masked by vegetation, this determination is only
possible through dip determination requiring measurement with compass in two perpendicular vertical planes
and the use of stereographic projection.
I
Figure 8-41 - Photograph of a normal fault filled by calcitic cement. The
drag is very small.
Figure 8-40 - Typical anticline at Durbuy, Belgium.
In a well, the structure determination requires dip measurement from core which is only possible if the core is
oriented and the borehole deviation well known. In the
other cases the measurement provides only an apparent
non-oriented dip which does not correspond to the real
structural dip.
It is the reason why, in subsurface, tectonic information
is generally extracted either from seismic data or from
measurements recorded in wells using dipmeters or
image tools or nuclear measurements in some cases.
Serralog 0 2003
Figure 8-42 - Other example of several small normal faults.
Determination of tectonic structures
from seismic data
This determination is well known and provides the first
information needed to locate the exploratory well.
Unfortunately, the subsurface images provided by seismic
data have not the necessary resolution allowing the
detection of all the faults or deformations which can exist
at the vertical of the well (Figs. 8-43 and 8-44).
313
Next Page
Determination of tectonic structures from
gamma ray
In some special cases the study of the gamma ray
curve can show either overturned features (Fig. 8-45 and
8-46) or reverse faults (8-47). These features can be
confirmed by dipmeter data but the correlations of gamma
ray depth intervals are sufficient to detect them. The
gamma ray response is not affected by the stresses
applied to the formations as the other curves may be.
-
Figure 8-43 - Faults detected on seismic cross-section.
W+E
s
Figure 8-45 - Example of overturned fold detected from the gamma
ray curve. The small section on
the top-right corresponds to the
lower section of the main curve on
the left after its reversal. As it can
be observed the pics correlate
perfectly
(courtesy of Schlumberger).
Figure 8-46 - Other example of
double overturned fold better
detected on gamma ray curve
than on sonic curve. The dipmeter
confirms the upper fold but does
not clearly reflect the lower fold
(courtesy of Schlumberger).
Figure 8-44 - The dipmeter allows the detection of more faults than the
ones recognized on the seismic cross-section shown above.
As illustrated by Figure 8-44, coming from the dipmeter recorded in the well indicated on the seismic section,
many other faults exist in this well. The seismic section
allows the detection of the major faults with a sufficient
throw.
As this book is fundamentally devoted to well logging
applications we refer the readers to the books on seismic
interpretation.
314
J
Serralog 0 2003
Previous Page
~~
Tectonics
Laterolog deep
Dips
Chapter8
I 315
muth of any surface crossing the well (Fig. 8-48).
Combining these two informations the real dip of any surface can be computed. The accuracy of the measurement
depends on the caliper, the apparent dip angle and deviation, the magnetic declination, intensity and inclination
and the processing. This accuracy is close to f 2" for azimuth and 0.2"for dip magnitude for dip higher than 5".
*
Azimuth of hole deviation
Angle of
hole deviation
Figure 8-48 - Principle of dip measurement. Non-alined three points of
the space determine a plane. These three points correspond to the
intersection of a plane with three generatrices of the cylinder corresponding to the well.
Figure 8-47 - A reverse or thrust fault. The slip (80 m) can be determined by defining the depth interval of the gamma ray cutve
that is repeated
(from Schlumberger, Well Evaluation Conference, Iran, 1976).
Determination of tectonic structures
from dipmeter or image data
Tectonic applications were the first geological application developed when dipmeters were introduced. The goal
was to analyze dips and dip evolution with depth to recognize possible rock deformation under stresses.
Ultimately, this could be used to identify tectonic or stratigraphic traps.
The tectonic application is fundamentally based on the
evolution with depth of the dips measured in low-energy
environments. In such environments, the sediments are
deposited by slow settling of fine particles in suspension
only by the force of gravity. This usually occurs at relatively great depths, beneath the action of waves, currents
or tides. The fine particles settle down on surfaces that
are approximately planar and subhorizontal. Consequently, any non-horizontal dip measured in such deposits is assumed to be the result of stresses (tectonic, differential compaction, subsidence, etc.) which have acted
on the deposits.
In wells, tectonic structures are easily detected and
analyzed thanks to dipmeter or image tools. These tools
allow the measurement firstly of the borehole deviation
and azimuth and secondly of the dip magnitude and aziSerralog 0 2003
As well known, dipmeter and image tools (Figs. 8-49 to
8-53) measure changes of the same parameter (conductivity, resistivity, ultra-sonic travel time and amplitude, density, radioactivity, magnetic susceptibility). They do not
directly measure the dip of bed boundaries. But, as the
petrophysical characteristics (lithology, texture, etc.)
generally vary from one bed or lamina to another, they are
reflected in the measured parameter, and they enable the
detection of parallel or oblique laminations, bed boundaries or faults as illustrated by Figures 8-54 to 8-56.
-I
Figure 8-49 - On the left: photograph of the SHDT tool (Stratigraphic
High resolution Dipmeter Tool) (couriesy of Schlumberger). On the
right photograph of the six -arms tool (courtesy of Halliburion).
315
Figure 8-52 - Sketch of the ARI tool (Azimuthal Resistivity Imager)
(courtesy of Schlumberger).
J
Figure 8-50 - On the top: photograph of the FMI (Fullbore Formation
Micmlmager tool (courtesy of Schlumberger). Below: photograph of the
STAR tool (courtesy of Baker-Atlas).
J
Rotating seal
Transducer
interchangeable
rotating sub
Figure 8-51 - On the left: the UBI
(Ultrasonic Borehole Imager) tool. On
the right: the RAB tool (Resistivity At
the Bit) (courtesy of Schlumberger).
Each electrode or button of the dipmeter and image
tools detects resistivity changes because of their sensitivity and high vertical resolution (approximately 1 cm for
dipmeters and 0.5 cm for image tools). If the boundary is
not perpendicular to the tool axis, this resistivity change
occurs at different depths on each recorded curve (cf. Fig.
8-48). The computation of the displacements in depth between buttons and pads, for a given resistivity contrast
(assumed to correspond to a lamina or a bed boundary),
is made by correlation techniques. It allows determination
of the dip of this boundary. Images tools have the advantage to allow the separation of sedimentary features from
tectonic or other features (stylolites, fractures) which is
not possible from the dip data provided by dipmeters tools
processed by interval correlation techniques.
316
b
a
Figure 8-53 - (a) On the left: photograph of the OBMI* (Oil-Base Mud
Imager). (b) On the right: sketch of the tool
(courtesy of Schlumberger).
ADN Tool
LlNC Tool
Figure 8-53c - The ADN
(Azimuthal-Density-Neutron) tool
(courtesy of Schlumberger).
But, because of the
shallow depth of investigation, each dip measurement has an individual
significance which can only
be local if it corresponds to
phenomena of small lateral
extent and with minimal
vertical effect (small scale
sedimentary features).
On the other hand, the vertical persistence of a nearly
parallel dip with slowly varying amplitude or azimuth, over
a certain vertical thickness will indicate the presence of
large structural entities.
Serralog Q 2003
Tectonics
I Chapter 8 I 317
Dipmeter and image tools thus determine the attitude
of the surfaces they traverse.
data. Thus, the absence of correlations between curves
and of dips on an arrow plot does not necessarely mean
that the recording is bad. Over intervals including such
deposits, the number of true structural dips calculated
from dipmeter data will be very low. In these cases a great
deal of skill and experience is required to identify these
few dips. Previous determination of the environment will
give useful clues to identify such cases. Furthermore, it
will also be useful in determining the environment for the
choice of the tectonic model (e.g. growth faults with rollover will be preferred choice in deltaic environments, etc.).
Figure 8-54 - Three types of image of the borehole wall (courtesy of
Schlumberger).
Figure 8-56 - Example of density and photoelectric (PEF) images
obtained by ADN tool in a nearly horizontal well
(courtesy of Schlumberger).
J
Figure 8-55 - Comparison of FMI images, on the left, with RAB images
on the right, and dip computation (courtesy of Schlumberger).
Other tools, such as the RAB tool (Fig. 8-55) or the
nuclear ADN (Fig 8-56) tool, can provide images during
the drilling, but with a much lower vertical resolution.
Should no such surfaces exist, due to the geological
environments (e.9. alluvial fan, conglomerates, breccia,
reefs, bioturbated beds, etc.), it will obviously be impossible to determine structural dips from dipmeter or image
Serralog 0 2003
In any case, in a structural (tectonic) interpretation of
dip data, it is necessary to choose, among all these surfaces, those that will show bed deformations caused by
stress, the other surfaces corresponding to sedimentary
features, stylolites, fractures, erosion, unconformities, etc.
This selection is only possible if the dip data is generated
by a processing based on break correlation (i.e. GEODIP
or LOCDIP for Schlumberger) and is related to a reconstructed lithological column and to the environment.
Hence, only dips related to the boundaries of the beds
that correspond to a low energy sediment, (e.9. to the
nearly horizontal surfaces at the time of sedimentation),
without any current related features, are retained. This
situation is best observed in the well-laminated shaly or
shaly-silty-sandy series, deposited by gravity action, or in
series formed by chemical precipitation (alternance of calcareous series, mudstone type, with shale deposits, etc.).
The type of processing of the dipmeter measurements
must consequently be taken into account. In any case, it
is highly recommended to always interpret dip data kno317
wing the lithology in which dips occur.
The interpretation of the dip or image data requires
also the knowledge of the nature of the well: vertical, inclined or horizontal (Fig. 8-57).
W i l Wll
Ima in
Hori?ontai Hole
Bed Dipping Away from
Kickoff Point
Shale layers
True dip
1.73190
Figure 8-57 - Influence of the well inclination on the images
(courtesy of Schlumberger).
Structural analysis of dip data
By reviewing the stresses and strains, and the structures that result from these forces, it is evident that analysis
of the dips, measured either on outcrop or in a well by dipmeter or image tools, makes a geometrical reconstruction
of the structure possible.
But, to be efficient and correct the interpretation of dipmeter or image data must be done by taking into account
other information and external data.
- Firstly, all the information related to the tool and the
processing used for the dip computation must be known.
1- The type of tool and principle of measurement:
three, four or more arms, number of electrodes per arm,
presence or not of a three-axes accelerometer, presence
or not of "speed buttons" allowing better corrections for
318
downhole movement of the tool;
2- The type of processing used in order to extract
dip data from dipmeter tools (i.e. MARK IV, CLUSTER
(Hepp et a/., 1975), or GEODIP (Vincent et a/., 1979) for
the HDT tool, MSD, CSB, or LOCDIP for the SHDT tool,
BorDip for any tool, OMNIDIP, etc.). Programs based on
correlation interval provide less data and compute a mean
dip which may not correspond to the structural dip but to
a meaningless dip, if the correlation interval contains features related to sedimentary processes, to erosion or fractures. The probability of such features increases with the
length of the correlation interval. Consequently, it is highly recommended to use a correlation interval between 2
and 4 feet. This also reduces the possibility of excessive
tool rotation over the correlation interval, which would, of
course, cause an erroneous dip computation. The quality
of the dips obtained from a correlation-interval processing
can be estimated by the sharpness of the correlogram
and the planarity : the sharper the correlogram and the
better planarity of the computed dip the higher is the probability of a representative structural dip, especially if, at
the same time, it corresponds to a low energy environment. In such environments, we can reasonably assume
that a succession of planar and parallel boundaries
occurs, sedimentary features due to current, tide or wave
action being generally absent. In such cases, the crosscorrelation technique used in programs favours structural
dips, all events on the resistivity curves that are due to
beds or laminations with planar and parallel boundaries
will be in phase for the same displacement across the
borehole. Consequently, this creates a sharp maximum in
the correlogram.
GEODIP or LOCDIP-type programs provide much
more dip data, the quality of which can be checked by looking at the correlation links, which must be selected to
extract dips computed at the boundaries of beds corresponding to low energy sediments. This selection can
be achieved by processing the results with the SYNDIP
program (Fig. 8-58 see next page) combined with a
LITHO-type results.
The dip determination from images can be obtained on
a work station using points of correlation or by fitting a
sine wave over the surface seen on image. Dips can be
computed either using a ScanDip-type program (Fig. 859) or BorDip. The advantage of images is that dips can
be classified as corresponding to sedimentary structures,
bed boundaries, fractures, faults, stylolites or unconformities. This classification can be achieved taking into
account resistivity contrast, depth interval between dips,
dip and azimuth consistency, planarity, etc.
3- Parameters selected for the computation: correlation interval, step distance, search angle, standard or
California option, for correlation interval processings;
number of samples for the computation of the derivative,
values selected for the definition of the size of the events
(small, medium, large), weights applied on each of the 9
Serralog 0 2003
Tectonics
parameters used for the computation of the pattern vector, likeness and discernability thresholds, search angle
and option, for GEODIP processing; number of samples
for the derivative computation, derivative threshold, decimation factor, search angle, for LOCDIP processing, etc..
These parameters should be printed on the heading of the
arrow-plot.
1 ChaDter 8 I 319
OFUENTATION
Figure 8-59 - Example of dip data extracted from FMS images using
the ScanDip program (courtesy of Schlumberger).
A quality control of the recording and dip computation
must also be made by looking carefully at the field monitor log (deviation, tool rotation, orientation, azimuth, relative bearing, resistivity curves, calipers, EMEX current),
and by controlling the dip measurements. A look at the
activity of the curves and the thickness of the events
allows a better determination of the parameters for dip
computation, especially for GEODIP and LOCDIP-type
processings.
- The arrow plot(s).
- The corresponding listing of the computation results
with the quality rating if any has been computed during
the processing.
Figure 8-58 - Example of interpretation of the dipmeter measurements
by the SYNDlP program. This processing ana/yzes curves in terms of
texture (homogeneous, heterogeneous and laminated), recognizes the
green patterns and displays the resistivity curves as a function of the
resistivity values seen by the electrodes. Blue spots correspond to calcitic nodules, yellow to sandstone and red to shale.
Serralog 0 2003
- Secondly, a composite-log of the open-hole logs with
the arrow plot at the same scale, is required. It will help to
give a rough idea of the lithology.
Any dipmeter or image interpretation must be done in
combination with the knowledge of the lithology.
Of course, it is highly recommended to have the geologist's mud-log, or the lithology column obtained by a
processing of the open-hole logs with a program such as
LITHO. The results Of a quantitative interpretation by
GLOBAL or ELAN-type programs can also be useful.
If the sonic waveform has been recorded with the
Long-Spacing Sonic tool or the Array Sonic Service, or
better with a dipole shear sonic measurement, the computation of the elastic properties is possible using the
319
320
Well Logging and Geology
MECHPRO-type program. These data define the mechanical properties of the rocks, information which is helpful
to determine the type of behavior under stress.
- Thirdly, a magnetic tape of the results is required, if
complementary processing of the data is requested .
Polar Plots
These plots are based on:
- either the WULFF stereonet (Fig. 8-61);
- or the SCHMIDT stereonet (Fig. 8-62).
Remark: All this information is not always available, especially for
exchanged wells.
Statistical study of dip data
When a large number of measurements are available
it is preferable to process them statistically. In this case
we will use graphical representations such as dip and azimuth histograms, density stereograms, etc., in order to
identify a preferential grouping or orientation, etc., and to
visualize the results.
DiD and azimuth histoarams
The dip values are plotted on the ordinate and the
number of samples on the abscissa (Fig. 8-60).
The dip values are generally subdivided into 5" or 10"
intervals while their azimuths are grouped into 10" intervals.
W
Figure 8-61 - The Wulff stereonet (see further for explanation of its
construction).
a
0
Figure 8-62 - The Schmidt stereonet (see further for explanation of its
construction).
b
Figure 8-60 - Example of dip (a) and azimuth (b) histograms allowing
the determination of the most probable structural dip.
320
On these projections the planes are represented by
their poles, and dip values are marked from 0" to 90" from
the center towards the circumference of the stereonet.
On a modified SCHMIDT polar plot (Fig. 8-63) the dip
graduations are equidistant, concentric circles increasing
from 0" to 90" at the center. The azimuth graduations are
at 10" intervals.
When the dip values are low this type of plot makes it
easier to plot and study the data.
This plot, which is not a stereographic projection, is
known as polar plot (Fig. 8-64).
Azimuthal freauencv plot
Here a polar plot is also used. Dip azimuths are plotSerralog @ 2003
Tectonics
ted on a circle in 5" or 10" categories, for example. The
lengths of the sector radii (Fig. 8-65) are proportional to
the frequency of each azimuth category. Polar plots and
azimuth frequency plots are commonly plotted together
and known as a polar frequency plot (Fig. 8-66).
1
Chapter 8
I 321
Densitv Stereoarams
As we can see in diagrams or stereograms (Fig. 8-67),
the zones of clustered points are easily perceived. In certain cases of very dispersed measurements it is impossible to find any group of points or trends. The density
stereogram is an aid that transforms the qualitative appreciation of the eye to a quantitative presentation which can
be interpreted more easily.
C
I
SOUTH
Figure 8-63 - A modified SCHMIDT net.
b
Figure 8-67 - Construction of a density stereogram using the
Dimitrijevic counting net. (a): Polar plot of the 150 dips. (b): Counting
the points of (a). (c): Corresponding percentages. (d): Lines of iso-density of the same stereogram (adapted from Henry, 1976).
Figure 8-64 - On the left: a typical polar plot
(courtesy of Schlumberger).
Figure 8-65 - On the right:: an azimuth frequency plot
(courtesy of Schlumberger).
Histograms, polar diagrams and azimuth frequency
plots are all computed automatically by zone in standard
computer processing of dip data.
The principle consists of counting the points (poles of
dips) inside each division, of a fixed area, distributed as a
uniform grid over the plot or stereogram. For this purpose
a counting overlay is used : the one introduced by Pronin
for a conformable projection (WULFF stereonet), or the
one created by Kalsbeek (1963, Fig. 8-68), or Dimitrijevic
(Fig. 8-69) for an equal-area projection (SCHMIDT stereonet).
a
140'
Figure 8-66 - Example of an polar frequency plot
(courtesy of Schlumberger).
Serralog 0 2003
b
Figure 8-68 - (b): Counting diagram obtained by drawing iso-density
points, using a counting net (a) defined by Kalsbeek, 1963
(from Ragan, 1973).
Construction of densitv stereoaram
As the SCHMIDT (azimuth equal-area) stereogram is
generally used, we will use the Dimitrijevic counting net.
The number of points in each small ellipse of the coun32I
ting overlay is counted. This number (or its percentage of
the total) is recorded on the overlay in the center of each
ellipse. Contours of iso-density (or iso-percentage) are
then traced. The azimuth and average dip values of one
or several areas of higher point density can then be read
from the base stereogram (Fig. 8-67). They can correspond to the structural dips if all the sedimentary dips
related to current activities have been eliminated from the
stereogram.
Stereographic plotting techniques
Plotting structures can be done using stereonets,
sometimes called stereographs.
However before discussing plotting techniques it is
useful to review the principal stereographic projections,
their construction and their use.
StereoaraDhic Drojections
"Stereographic projection is a presentation and
abstract geometrical construction which enables analysis
of the orientation of tectonic elements in space independently of their geographical position" ( J . Henry, 1976).
While this is not the place for a detailed account of the
theory and practice of stereographic projection (it is well
treated in texts on structural geology or on this technique
alone), it is very useful to summarize the principles before applying them.
Principle
Stereographic projection involves moving various observed structural elements (lines and planes), while keeping them parallel to themselves, until they touch or cut
the top half of a reference sphere resting on a flat horizontal plane (Fig. 8-71).
Figure 8-69 - The Dimitrijevic counting net.
Interpretation of these diaarams
When the polar diagram shows a distribution of dips
such as that in Figure 8-70, we can conclude that the
structural dip is low and that the azimuth dispersion is due
to these low values. Other dip values that are grouped in
a triangular form, with the apex oriented toward the center, may correspond to either tectonic deformations, or
sedimentary features.
When a diagram (such as the one shown in Fig.8-67)
illustrates a more concentrated distribution, the sectors
having high density and low dip values probably define
the structural dip, sedimentary features, or folds being
added to this.
North
Figure 8-70 -.
Example of an interpretation of a
SCHMIDT diagram
(courtesy of
Schlumberger).
322
Nadir
a
Nadir
b
Figure 8-71 - Stereographic image of a plane (a) :polar projection;
(b) :cyclographic projection.
This plane contains the North-South axis. The intersections of the lines and planes with the hemisphere can
be projected stereographically onto the horizontal plane
with reference to the lower pole or nadir of the whole
sphere. This inversion - which is a geometric transformation (Fig. 8-72) - conserves angles, but looses all information relating to the geographical situation of the sturctural elements. For example, two parallel fault planes
1000 m apart would, when translated to the reference
hemisphere, touch it tangentially at the same point, and
be indistinguishable from each other on the stereogram.
This procedure is useful in that the attitude (strike and
dip) of a plane or line can be represented as a single
point. It is thus possible to analyze the angular relation
between a large number of planes.
Serralog 0 2003
Tectonics
1 Chapter8 I 323
KMxKrn = KZxKO = Constant
C
Diredion ande
Figure 8-74 - Cyclographic projection of a plane.
K
(nadir)
KMxKm = KBxKb = KZxKO = Constant
I
7gure 8-72 - Stereographic projection is a particular case of inversion.
a): inversion of a curve; b): inversion of a circle;
c): inversion of an upper hemisphere.
Planes and lines on a stereogram
As Figure 8-71 indicates, there are two ways of representing a plane.
. Polar plot : this is the stereographic projection of the
point of tangency of the plane and the hemisphere (Fig. 873), or of the intersecting point of the diameter perpendicular to the plane, with the hemisphere. The projection of
this point onto the horizontal reference plane is the pole of
the plotted plane.
If the dip of the plane is higher, its polar representation
will be farther from the center of the sphere, but its cyclographic representation will be closer to the center.
Plotting a line
There is only one method of plotting a line. That is to
project the point of intersection of the straight line passing
through the center of the sphere with the upper hemisphere. Its projection onto the horizontal reference plane
is a point.
Take, for example, the straight line N 160" - 40" NW.
Turn the tracing paper in an anticlockwise direction so as
to align its North with the 160" division of the stereonet
(Fig. 8-75a). Then count 40" southward of the North on
the principal diameter of the stereonet and mark a point.
A tail is usually added, pointing towards the center, to distinguish the representation of a point from polar representation of a plane (Fig. 8-75b). This point now represents
the dip and strike of the straight line.
Dip azimuth
Figure 8-73 - Polar projection of a plane
. Cyclographic plot: this is the stereographic projection
of the intersection of the hemisphere with a plane passing
through its center (Fig. 8-74). This line represents half of
a great circle. Its projection onto the horizontal reference
plane is a curve, part of a great circle.
Serralog 0 2003
Figure 8-75 - (a) fmage of a straight fine N 160" - 40" NW
(b) Stereogram of this line; and (c) its representation in space.
323
The greater the plunge of a straight line, the closer to
the center the point its projection will be. The plunge is the
inclination of a line measured in the vertical plane. Its
pitch is the angle of the line with the horizontal measured
in the plane containing the linear feature (Fig. 8-76).
Vertical plane
Figure 8-76 - Representation of the pitch and the plunge.
Construction of the WULFF stereonet
The WULFF stereonet represents the stereographic
projection of parallels and meridians of the hemisphere of
reference onto the horizontal plane; its North-South axis
has the lower pole or nadir as the projection pole.
Diagrams of Figure 8-77 explain how the meridians
are represented by the arcs of great circles, named on
projection as the great circles, and the parallels by the
arcs of the small circles.
This projection retains the angles : the great and small
circles are orthogonal to one another. The surfaces are
delimited by two parallels and two successive meridians.
They vary according to their position.
This projection is in common usage. In fact, it serves
to determine the following elements : the true dip from
apparent dips, the angle between two planes and the
intersection of the beds with a fault (see later).
However, this projection has disadvantages. It does
not preserve surfaces, due to inversion, and it is thus difficult to draw a density contouring (Figs. 8-67 and 8-68)
from this stereogram. In these cases, it is preferable to
use a SCHMIDT equal-area stereonet.
Construction of the equal-area or SCHMIDT stereonet
The SCHMIDT stereonet corresponds to a Lambert
azimuthal equal-area projection. Relative to the center of
a sphere parallels and meridians are represented on a
SCHMIDT stereonet by the arcs of an ellipse. Any area
bound by two meridians has the same area, regardless of
its position on the projection (Fig. 8-79).
ParalleL
Nadir
Figure 8-79 - The equal-area, or SCHMIDT stereonet construction.
Figure 8-77 - Construction of a WULFF stereonet.
Figure 8-78 represents the isogonal WULFF stereonet
for meridians drawn at 2" intervals and for the parallels
crossing the North-South meridians also, at 2" intervals.
Its uses are much the same as the WULFF stereonet.
It has the advantage of high resolution when used to
represent planes of low dip, and is, hence, the preferred
stereonet for the study of gently folded structures.
Angles are not well preserved but this projection is
especially useful for statistical studies. As J. Henry stated:
with a large number of measurements we can study "the
dip and strike of structural elements from a statistical point
of view". In other words it is possible to consider variations
due to accessory and complex phenomena or measurement errors as negligible and use statistical averages as
the basis for analysis and interpretation.
Stereographic representation of structural elements
Straight lines, planes and folds are the structural elements that will be represented in this exercise.
Small circle
Figure 8-78 - The WULFF stereonet.
The external circle of the stereonet is called the fundamental or primitive circle.
324
A stereonet (WULFF or SCHMIDT) on a firm base is
necessary. Tracing paper should be placed over the
stereonet, and held in position b its center. The center,
major circle and the points of the compass should be marked on the tracing Paper (Fig. 8-80).
Serralog 0 2003
Tectonics
Stereogram
(transparentpaper)
Stereonet
Figure 8-80 - Board for the stereographic plotting of dip data.
- Stereographic representation of a plane
Take, as an example, a plane with a dip of 60" in a
direction N 25" (Fig. 8-81). As noted previously this plane
can be represented in two ways:
. Polar representation of a plane
Turn the tracing paper in an anti-clockwise direction,
so as to align its North with the 25" division on the stereonet, corresponding to the azimuth of the dip to be plotted
(Fig. 8-82). At this point the plane to be plotted should,
conventionally, pass through the E-W axis of the stereonet. The pole of the plane is plotted on the 0" line of the
N-S diameters at the 60" graduation.
Plane direction
or dip strike
60"
-Horizontalplane
Dip
magnitude
I
Chapter8
I
325
The first method of plotting a plane is used when manipulating dipmeter data as it is the azimuth of dips that is
calculated during the processing of the data.
. Cyclographic representation of a plane
The pole plotted previously (25" N 60")is transposed
to the E-W diameter and marked either: 90" from the pole
towards the center of the stereonet; or 60" from the edge
of the stereonet along the line opposite the one on which
the pole lies (Fig. 8-82 right).
The cyclographic trace is drawn as the corresponding
portion of a great circle on the stereonet.
- Stereographic representation of a straight line on a
plane
This is common in structural geology, when, for example, the apparent dips of a plane seen in cross-section
need to be plotted.
Firstly, it is necessary to know the direction and dip of
the plane containing the line, then the azimuth and pitch
of the straight line are required.
Consider a plane with a direction N 60" inclined 50"
NW, containing a line inclined 50" to the West. Construct
the cyclographic trace of the plane using the method illustrated in Figure 8-83. Count along the trace 50" starting
from the N-E edge of the stereonet and following the great
circle. Rotate the tracing paper to bring the great circle to
the principal N-S diameter and read the dip and azimuth
of the line in the plane defined by the point marked on the
great circle.
Vertical plane
-
Figure 8-81 Descriptive elements of a plane dipping at 60" with an
azimuth of N 25".
Figure 8-82 - On the left: polar image of the plane in Fig. 8-81; on the
right: its cyclographic representation.
Figure 8-83 - Stereographic image of a straight line with a "pitch"of
50" towards the west, in a plane N 60" - 50" NW [from Henv, 1978).
Planes with small dips will plot with their poles near the
center of the diagram, while those with greater dips plot
towards the edges of the stereonet. A vertical plane has
its pole on the fundamental circle.
- Stereographic representation of folds
As we have seen, folds are curved surfaces, with
either a generally cylindrical or conical symmetry. The
axis of the fold may be horizontal or inclined and its axial
surface vertical or inclined and planar or warped.
We have also seen that stereographic projection involves the parallel translation of geometrical elements onto a
Note: In certain cases a plane is defined by its direction (strike)
rather than its azimuth. Its direction is 90"from its azimuth. In this case
its pole is plotted on the 270" line of the E-W diameter (Fig. 8-83).
Serralog 0 2003
325
Next Page
reference hemisphere. As the translation of planes is
easier than that of curved surfaces, we will represent
these folds as an infinite number of planes tangent to their
surfaces. This is close to reality, because we are only describing curved surfaces in terms of the attitude of the
plane within the fold that has been subject to same stresses as the fold and intersects the borehole.
Firstly we will describe the theoretical stereographic
representation of a cylinder and a cone. We will then
assume that the features of the fold are known and see
how it can be reconstructed from the stereographic projection.
. Stereographic representation of a horizontal
cylinder (Fig. 8-84)
sentative point can be plotted; from this point rotate the
transparent overlay on the W-E diameter and draw the
great circle, 90" from this point.
Cylindrical fold
Axial plane of the fold
Inflexion points
fold limb
Fold angle
Pole of the axial
plane
Fold axis
Point of tangency
Its reDreSentatlonon
a'stereogram
Figure 8-85 - Image of a cylindrical fold, with inclined axial planes and
a plunging axis (adapted rom Henv, 1978).
Cylinder simulated by
planes tangent to its
upper surface
Cylinder
axis
Point of tangency
Translation of planes
to make them tangent
to the upper hemisphere
Cylinder axis
Cylinder axis
Stereonet representation
of the cylinder
Poles fall on a great diameter
Figure 8-84 - Stereographic representation of a horizontal cylinder (a) :
planes tangent to the cylinder; (b) :translation of these planes onto the
upper hemisphere; (c) :stereogram of the cylinder
(adapted rom Henv, 1978).
As explained, a cylindrical surface may be decomposed into an infinite number of planes tangent to that surface. The planes form a network in space around a common axis. Each plane is represented in its projection by its
pole. The poles representing this network lie in a zone on
a great diameter of the stereonet. The fold axis is situated
on the fundamental circle and is perpendicular to the
great diameter.
. Stereographic representation of a cylinder with
an inclined axis (Fig. 8-85)
In this case the poles of the tangent planes gather in a
zone on a great circle situated at 90" to the axis of tilt.
Knowing this great circle the axis of inclination is defined
by counting 90" on the W-E diameter perpendicular to it.
Having now defined the axis of the cylinder its repre326
Axial
plane
. Stereographic representation of a cylinder with a
vertical axis
In this case the poles of the tangent planes lie in a
zone on the fundamental circle and the axis of the fold is
in the center of the stereogram.
. Stereographic representation of a cylindrical fold
with a straight, inclined axis.
A cylinder, being symmetrical about its axis is not
represented by any one plane. However, in nature a fold
only represents part of a cylinder, this part represented by
the folding angle (the dihedral angle between planes at
tangents to the limbs of a fold). Thus the axis of a fold may
be horizontal (Fig. 8-86) or plunging (Fig. 8-85). If the
stereogram contains points representing the dip at the
points of inflection, A and 6 ,of the two limbs of the fold (a
and b on the stereograms), the folding angle is defined as
the angle between the two planes tangent at these points.
The axial plane is represented by the intersection, on a
great circle, of the axis of the fold, and the bisector of the
two outer beds, assuming that the axial and the bisecting
planes are close (case of isopach or similar folds). It is
easy to construct the pole of the axial plane, by shifting
the axial plane on a great circle and by counting an angular distance of 90" from this great circle on the E-W diameter.
Axial Dlane
Fold axis
Pole
'Axis
%$I?
plane
Figure 8-86 - Image of a cylindrical fold, with inclined axial plane but a
horizontal axis.
Serralog 0 2003
Previous Page
Tectonics
1
Chapter8
I 327
where the poles are aligned.
The inclination of the axial plane in relation to the vertical is determined by counting the angular distance between the center of the stereogram and the great circle
that represents this axial plane. Its azimuth is defined by
turning the transparent overlay to bring the pole onto the
northern half of the N-S diameter and then counting on
the fundamental circle, in an anticlockwise direction, the
number of divisions between the N of the stereonet and
that of the stereogram.
. Stereographic representation of a cone with an
inclined axis (Fig. 8-88)
In this case the planes tangent to the cone form distinct lines once translated onto the hemisphere of reference: they describe the arcs of two oblique circles that do
not lie directly on the small circles of a stereogram.
However, after rotation of the WULFF stereonet by a
value corresponding to the plunge of the axis of the cone
(40" in the example) they fall in a line on two small circles.
. Stereographic representation of a cone with a
horizontal axis (Fig. 8-87)
In this example we will plot a cone with a horizontal
axis oriented to N 150" with a 60" apical angle. The
stereographic projection of all planes tangent to this cone
gives a symmetrical figure the planes are, in fact, tangent
to two small circles of the hemisphere and their poles lie
along two small circles on the stereogram. The horizontal
axis is situated on the fundamental circle on a diameter
perpendicular to the small circles.
Point of tangency
Cons simulated by planes
tangent to ik sur(ace
Nadir
Figure 8-88 - Stereographic image of a cone with an inclined axis
(from Henry, 1976).
Upper hemisphere
Translationof planes
. Stereographic representation of a cone with a
vertical axis (Fig. 8-89)
to make them tangent
lo the *per hemsphere
Cone with vartical axis
end an apical angle of 90'
Stereogram of a ume
with verfical angle down
P d o d tho plane
Stereonet reprerentatbo
d the cone.
Poks fdl on a Smell drcle
a, b. c. d and e are polesd the plan63
Figure 8-87 - Stereographic image of a cone with a horizontal axis (a) :
planes tangent to the cone; (b) :translation of the planes onto the
upper hemisphere; (c) :stereogram of the cone
(adapted from Henv, 1978).
The position of the small circles is a function of the
apical angle. It may be determined by counting the angular distance between the center of the stereogram and
one of the small circles (here 30") and by multiplying it by
2, or directly, by measuring the angular distance between
the two small circles. One of the small circles corresponds
to a synform fold, the other to an antiform fold. The direction of the apex of the cone representing the fold is the
same as the direction of the concavity of the small circle,
Serralog 0 2003
TransJationbf the ume
on UIE upper hamisphere
Figure 8-89 - Stereographic image of a cone with a vertical axis
(adapted from Henv, 1978).
In this case the poles of the tangent planes are gathered in a zone on a circle, the center of which coincides
with that of the stereonet. The apical half-angle is defined
by the angular distance between this circle and the fundamental circle. The axis of the cone is in the center of the
stereonet.
Plotting of dip data
When plotting dip data (magnitude and azimuth) it is
actually the intersection of a plane (or surface) with the
327
borehole that is being represented.
As the borehole axis does not usually follow the axis of
the fold we will be making a cross-section through the
borehole showing the various strata at varying angles,
and naturally at different depths. Stereographic plots do
not take into account the geographical position of features
(in this case their depth) but do record the relative angles
of features. Thus we will have a plot representing planes
tangent to the fold surface and passing through the borehole, and from this information will be able to make a full
description of the fold. Plotting the relation in space between the dips of the representative planes enables us to
determine if the variations define, for example, a cylindrical or conical fold, and to determine the axis and plunge
of the fold as well as the inclination of the axial plane and
the angle of folding.
This technique can also be used to define the intersecting line of any two vertical planes, the apparent dip on
a cross-section of a plane, remove regional dip, make
rotations, etc.
Except for particular applications we will use the polar
representation of planes as it usually simplifies both the
reading and interpretation of stereograms.
Procedure
Normal plotting of dipmeter data is extremely easy.
Quite simply the values of dip, read from logs, or preferably listings, of processed dipmeter or image data are plotted on a WULFF or SCHMIDT stereonet using the previously described techniques.
As prospective oil-bearing structures usually only
involve low dips it may be easier to use a large scale version covering the central part of the stereonet only (Fig. 890). This leaves more space between plotted points
making it easier to analyze their relation. However manual
plotting of several hundred or even thousands of dips that
a dipmeter log will give is a laborious and tedious task.
The use of data, taken directly from the magnetic tape of
the computed dipmeter results, in a simple plotting program can produce stereograms on a video monitor or in
hard copy. The first stage in analyzing dip meter data is to
zone the arrow-plots into coherent intervals. This zoning
is partly done using the lithology, as indicated by other
well logs, but also using the arrow plots themselves, the
arrangement, evolution and patterns of dips influencing
the choice of zones.
H
Figure 8-90 - Enlargement of the center of a WULFF net.
328
H'
For tectonic applications dips computed by the interval
of correlations programs, such as CLUSTER for HDT
data, or MSD or BorDip program for SHDT data, are most
commonly used. If dips computed by the GEODIP or
LOCDIP-type programs are used, intervals corresponding
to low energy beds showing a series of dips with the same
dip and azimuth (green patterns) should be chosen. A
SYNDIP processing (cf. Fig. 8-46) of the GEODIP or
LOCDIP results enables "green" patterns to be selected
(Fig. 8-91). The addition of the lithology information provided by a processing of the open-hole logs by a program
similar to LITHO allows "green" patterns, corresponding
to low energy (shaly) environments, to be identified.
Solid angle = 5"
Figure 8-91 - Determination of "green patterns" on arrow-plot and
stereonet. A "greenpattern" corresponds to a minimum 5 successive
dips falling in a 5" solid angle.
Now, using the coherent intervals chosen by the above
methods, stereograms can be plotted, either manually or,
more easily, using computer programs.
Use of stereograms
At this stage of the analysis the objective is to reduce
the geometric information contained in the stereograms of
each interval to a more synthetic variable.
Several cases may occur.
- The simplest case is that of an interval of approximately constant dip and azimuth : in this case the stereographic projection enables two variables to be defined : the
mean dip over the interval and the angular dispersion of
the dips about this mean. To define the angular dispersion
a circle enclosing the majority of the plotted points and
centered close to their highest density must be drawn. A
Dimitrijevic counter (see Fig. 8-67) may also be used. The
dispersion is interesting for several reasons. Associated
with the percentage of values within an interval it gives an
estimate of the confidence of the mean dip. On the other
hand the value of the dispersion may be characteristic of
certain formations as it is a good indicator of the amount
of stratification.
- The studied interval corresponds to a regular change
of dip and/or azimuth on the arrow-plot. Such patterns
Serralog 0 2003
Tectonics
usually correspond to a cloud of points on a stereonet,
stretching along a line, such that defining an average dip
or azimuth is impossible. This is where stereographic projection can provide important information. It is, in fact, a
precise geometrical translation which can be viewed as a
succession of segments of cylinders or cones inside each
other, each dip corresponding to a plane tangent to these
successive cylinders or cones.
. If the dips are grouped in a line following a great
diameter of the stereonet we can conclude that the fold's
axis is horizontal.
. If the dips lie on a great circle they represent a
cylindrical fold with a plunging axis (orange pattern introduced by Etchecopar). Determination of the azimuth and
the plunge of the axis is carried out in the following manner:
- Draw the great circle passing through
the poles of the planes.
- Count 90" from this great circle on the
E-W diameter towards the center. This point represents
the axis of the cylindrical fold, while its plunge is defined
by counting angular distance from the edge to this point.
- To find its azimuth, turn the transparent
overlay to bring the axis onto the N half of the N-S diameter.
- Then count, in an anticlockwise direction, the angle between the N marked on the stereonet
and that on the stereogram.
. If the dips are grouped in a zone on a small circle we have a conical fold with a horizontal axis. The axis
is on the fundamental circle and on the N-S diameter. The
apical angle is defined by measuring the angular distance
between the center and the small circle and multiplying it
by two. The azimuth is determined by rotating the axis to
align it with the N of the stereonet, and then reading the
anticlockwise angle between the two North (stereonet and
stereogram).
. If the plotted points lie along a curve that does
not correspond to any great or small circles then the axis
of the fold represented may be inclined. By shifting the
stereogram around the stereonet it may be possible to
align the plotted points along a small circle (Fig. 8-92), the
shift corresponding to the pitch of the axis of the cone.
Chapter8
1
329
- In certain cases the arrow-plots may show no apparent organization (Fig. 8-93). However after plotting, the
stereogram shows a coherent pattern, representative of
stratification planes A and 6, indicating a fold, overturned
to the South. The other points (a, b, a and b) also fall perfectly on two small circles of the stereogram on the right,
obtained after rotation to bring the axis of the fold horizontal. These points correspond to conjugated orthogonal
joints. The plunge of the axis is 10". By convention these
patterns are called dispersed axes.
Armw olot
Figure 8-93 - Example of an incoherent arrow plot. When plotted on a
stereonet certain points (A and B), representative of bedding planes,
show a coherent grouping along the line of a great circle. The other
points (a and b, a and p) also represent a coherent feature on the
stereogram plotted on the right. This has been obtained by rotating the
projection to bring the fold axis horizontal. They correspond to the
conjugated orthogonal joints.
Another example of an apparently incoherent arrow
plot is shown in Figure 8-94.
I
Figure 8- 94 - Another example of a random arrow-plot corresponding
to a group of points along a great circle (contributed by J. Henry).
- The opposite case can also occur; an arrow plot
showing clear red and blue patterns giving a dispersed
stereogram with no apparent axes (Fig. 8-95).
Figure 8-92 - Stereogram o f a non-cylindrical fold with an inclined axis
and its plot on a shifted stereonet (adapted from ffenv, 1976).
Serralog 0 2003
- Certain folds have a strongly curved axis. The measurements carried out on this axis at different points are
grouped in a zone on a great circle of the stereogram.
This great circle represents the axial plane of fold hence,
its pole, pitch, and azimuth can be defined (Fig. 8-96a).
329
softwares developped by service companies for work stations.
"Quick-Look" Techniaue : Dip Patterns
Arrow plots or tadpole plots where the results of dip
calculations are presented versus depth, often show dip
arrangements (Fig. 8-97), called dip patterns that can be
organized into four groups and coded by color (Gilreath &
Maricelli (1964).
Dips
Figure 8-95 - The arrow-plot seems to show a "mega-blue pattern".
The stereogram clearly illustrates the dispersion of the plotted dips
suggesting a more complex structure (contributed by J. Henry).
Come otids fimdsmcutally to
di if observed on a
long h t e J with low energy
deposits
On a short interval, nlay
correspond to laminations
s b u c d
Green
Red
Yellow
Figure 8-96 - a) Defining the axial plane by the dispersion of the axes.
b) Summary of dipmeter data illustrating the average dips, the general
trend and the major and minor axes (contributed by J. Henry).
Methods ofh dip data interpretation
Generally, three methods of dip data interpretation are
used.
The first one, which can be qualified as a "Quick-Look"
technique, uses dip patterns observed on arrow plots.
The second, more sophisticated, uses techniques
introduced by Bengtson, an inventor formerly working for
Chevron (1980) and Sohio (1981).
The third, more scientifically based, uses work stations
and softwares in which the stereographic projection is the
base for dip analysis. As previously illustrated the stereographic projection allows a more detailed analysis of dip
data giving a complete and precise reconstruction of
structures crossed by a well.
Visual arrow-plot analysis
A dip, or tadpole, represents a surface cutting the
borehole. The dip and azimuth of this surface may have
two components, firstly a stratigraphic or sedimentary
component caused by current or wave activity, and
secondly a structural component caused by later, largescale tectonic movements.
The first step in the tectonic application of dip data is
to determine the structural dip. The arrow plot interpretation can be achieved either using a "quick-look technique
called dip patterns, or automatically using stereographic
projection. Dip data extracted from dipmeters or imagetool measurements can also be interpretated thanks to
330
Interpretation
Blue
FoUowiiq its length it comesponds
to :
a fault
a draping of a reef or a bar
a rilhg of a chsrmel
an unconformity
--
Emtic events which may
correspond to cross bcddin if
observed on a short intarf.
its length may
:
;:l;Yato:
a halt
foreset laminations
an uuconfonuity
--
Figure 8-97 - Typical colored dip patterns and their interpretation
(from Gilreath & Maricelli, 1964).
- Green patterns correspond to a set of dips with nearly constant magnitudes and azimuths.
- Blue patterns correspond to dip angles of similar azimuth that decrease with increasing depth.
- Red patterns are the inverse, the dip angle decreases with decreasing depth, the azimuth remaining essentially constant.
- Yellow patterns correspond to adjacent dips with random magnitudes and azimuths.
Green patterns, when observed in low energy deposits, usually correspond to a structural dip, a monocline or
a limb of a fold (Fig. 8-98, zone 7500-7580 ft).
Blue and red patterns, when also observed in a low
energy deposits, indicate the structural deformations of
the beds or layers, that may be related to folds, faults with
their associated phenomena (rollover, drag), or to an
unconformity.
The diagrams in Figure 8-99 show the theoretical
arrow plot responses for different cases.
The rapid analysis generally starts by determining the
structural dip from the arrow plots. The dip of the beds is
defined, assuming they were deposited horizontally and
parallel to each other. By extrapolating this dip the architecture of the beds and the shape of the structure they
Serralog 0 2003
Tectonics
constitute can be inferred. Structural dip is generally defined as a predominant dip, over a thick interval, of constant azimuth and magnitude. Therefore, a “base line”, or
succession of green patterns with the same angle and
ostensibly the same azimuth, can be determined on the
arrow plots if the dip is steep . Structural dip is thus determined by a visual, or graphic statistical analysis. In the
case of graphic statistical analysis, dip histograms and
azimuth frequency plots or rosettes are used. The determination of the azimuth of the structural dip may be very
precise as soon as the dip exceeds a few degrees.
Gamma ray
1 Chapter8 1 331
tern (Figs. 8-100 to 8-102) and the extent of the deformation can generally be related to the amount of slip.
Asymmetric foM -Anticline
Ownturned anwine
Dips
Asymmetric fold - syncline
Drag along the fault
plane upper block
Rollover in upper block
of a growth faun
Recumbent antidine
Drag along the fault plane
upper and lower blocks
Beds overturned by drag in
the upper blodc of a thrust fault
Figure 8-99 - Theoretical arrow plots, the four upper diagrams for
anticlines, the four lower diagrams for faults.
Figure 8-98 - Example of structural dip determination on a dipmeter
arrow plot. It corresponds to “green” patterns
(courtesy of Schlumberger).
Determinina structural diD
Structural dips must be determined in rocks that represent low energy environments. These rock types generally correspond to shales, silty shales or mark (Fig. 8-98)
and are identify by:- shaly zones from logs (high gamma
ray, high thorium, uranium and potassium [usually], neutron-density shale separation, high transit time values,
high attenuation of electromagnetic waves, etc.)
- consistent dips (green patterns) over relative long
intervals;
- the lowest dips (usually).
Detectina faults
Faults can be detected on dip arrow plots if drag or rollover was generated during the faulting. This type of
deformation corresponds generally to a red or blue patSerralog 0 2003
In the case of a reverse fault, a repeat section is present and the slip can be determined using the gamma ray
curve; i.e., part of the gamma ray curve will be duplicated
(cf. Figs. 8-46 and 8-47).
Dip
Figure 8-100 - A normal fault is suggested by the color paffems
associated with the drag.
331
Dips
Figure 8-101 - A reverse fault is suggested by the color patterns associated with the drag.
the crestal plane (CP), as well as the type of drag of a
fault (Fig. 8-105).
Figure 8-102 - Example of SCATplots. From the top left to the bottom
right are successively reproduced: the contour map, the dip versus azimuth plot, the croos-section, the DAPSA plot (aeimuth and dip separated but each versus depth, the transverse dip component and the longitudinal dip component (from Bengtson, 1981).
Faults that have no associated deformation can be
detected only with images (see later).
Detectina folds
Folds are characterized by a progressive change of
dip angle, and commonly dip azimuth, with depth over a
long interval (Fig. 8-47 from 2525 to 2675 m). The dips
generally plot on a great circle on a stereonet.
Bengtson Technique
This technique is a method of analyzing dip data, the
apparent advantages of being:
- more analytical than the colored pattern method,
- less time-consuming than the stereographic projections.
This technique uses several graphs or plots to geometrically describe most of the parameters that are used to
define a fold or a fault. An example of such graphs is
shown in Figure 8-102. They are known under the acronym SCAT (Statistical Curvature Analysis Techniques)
plots. The dip-azimuth histogram combined with the azimuth versus depth plot and dip versus depth plot - these
two lastplotsbeing known under the acronym of DAPSA
plots (Dip and Azimuth Plots for Structural Analysis of
dipmeter data) -, allow determination of the structural
transverse direction. This direction will be used to obtain
stick-plots which themselves will be used to draw a vertical cross-section through the well. A dip versus dip-azimuth scatter diagram (Fig. 8-103) enables determination
of the fold or fault type, the drag pattern and the "local"
transverse and longitudinal dip component. The DAPSA
plots combined with the "local" transverse dip component
determine the bearing, the crestal plane (CP), the axial
plane (AP) and the inflection plane (IP) as well as the
plunge for an axial plunging fold (Fig. 8-1O4), or the drag
zone, the precise depth of the fault, the trough plane (TP),
332
Figure 8-103 - Example of dip versus azimuth plot
(from Bengtson, 1981).
-1
Figure 8-104 - Example of analysis of SCATplots in the case of a fold
with plunging axial plane (from Bengtson, 1981).
J
Figure 8-105 - Example of analysis of SCATplots in the case of normal
faults (from Bengtson, 1981).
Serralog 0 2003
Tectonics
DIP AZIMUTH
DIP
S
0
40
D.vi.tion
W
N
I
I
E
S
I
Figure 8-106 - Example of SODA-Seria plot
(courtesy of Schlurnberger).
The DAPSA plots do approximately the same thing but
the presentation is slightly different (Fig. 8-102), the apparent dips being displayed versus depth in the transverse
and longitudinal directions which are determined from the
analysis of the azimuth versus depth plot or the azimuth
frequency plot. Countouring of the data concentration can
be done, the DAPSA apparent dip plots being continuous
equal-area presentations. As pointed out by Wise &
McCrory (1982) this technique can suppress noise and
reveal broader patterns and major trends.
SCAT plots correspond to the plots introduced by
Bengtson (1980, 1981). They include the dip versus azimuth plot (Fig. 8-103) and the azimuth versus depth and
dip versus depth plots associated to the DAPSA plots.
Automatic and interactive programs
The methodology for determining structural dip, faults
and folds has been implemented in two different programs
(Etchecopar & Dubas, 1992). The first program automatically extracts large-scale structural information from dip
data. The second program is interactive, allowing the user
to accurately compute the local value of the structural dip
using stereographic projection. Hereafter are briefly described the different programs
DiDTrend Droaram
The program that operates automatically is called
DipTrend. The processing consists of the following
steps.
Serralog @ 2003
I Chapter8 I 333
1 - Filter the dip data, taking into account the quality
and structural continuity of the data. The result of the filtering will depend on the sequence length selected by the
user (three or five tadpoles).
2 - Sequence the data into green, “orange”, and “yellow” patterns, and determine the axis of each orange pattern.
- A green pattern suggests structural dip, as discussed
previously. Tilted green sequences are differentiated from
the horizontal sequences because their tilt is necessarily
caused by a tectonic event.
-An “orange”pattern corresponds to dips fitting a great
circle on a stereoplot and is characterized by the pole of
this circle. Figure 8-107 (see next page) illustrates the
results of these two steps.
- A “yellow” pattern corresponds to dips fitting a small
circle and it has an axis and a conicity (do not confuse
with the yellow pattern of Gilreath & Maricelli)..
3 - Merge the sequences with consistent axes into
megasequences, assuming that each megasequencewill
characterize a single tectonic event. During this step, an
orange sequence can be merged with other orange or
green sequences into a megaorange, but never into a
megagreen.
4 - Smooth the data according to the previously determined structural axes. Smoothing avoids small discontinuities when building cross-sections or three-dimensional
models at seismic scale. This allows each structure to be
described in three dimensions, as if made of perfectly planar, cylindrical or conical sections.
- Megagreen sequences are smoothed by using either
a median or an averaging filter of the data. The length of
the window used for smoothing, is usually twice the minimum sequence length (expressed in number of dips) of
the initial sequencing, minus one dip.
- Megaorange sequences are smoothed in two steps.
The first step consists in an orthogonal projection of each
pole on the circle characterizing the structure (Fig. 8-108).
The second step consists of smoothing the dips along the
great circle, using a median filter. This smoothing is justified by the assumption that the small angle between each
dip and the circle is caused by microstructures, sedirnentary features or compaction.
5 - Remove the smoothed structural trend from the filtered dips. For each filtered dip, there is now an associated “structural” trend which can be removed. For a green
sequence, the removal is a single rotation around the strike (horizontal line) of the smoothed dip. The value of the
rotation is opposite the value of smoothed dip. For an
orange sequence, the trend removal is divided into two
rotations. The first one removes the dip of the axis of the
great or small circle and is applied to the filtered and
smoothed dips. The second one is, as for a green
sequence, a rotation that removes the smoothed dip to
the horizontal.
The trend removal described above is severe, because the removed structural dip is smoothed considerably to
allow comparison with seismic dip. For accurate sedimentary studies using images or using dipmeters inter333
Figure 8-107 - On the left part is presented the initial sequencing of the arrow plot. On the right part of the figure from left to right are displayed the
original data with the initial sequencing of the upper interval, the filtered dips, dips after structure trend removal, and local curvature axes
(from Etchecopar & Dubas, 1992).
preted by a CSB, GEODIP or LOCDIP - type processing,
a new interactive method for structural dip determination
was developed.
Automatic structural dip determination
The program for automatic structural dip determination
is called, in Schlumberger, SediView. Its application
allows the computation of the structural dip even in intervals showing a lot of sedimentary dips (cross bedding,
foreset bedding, etc.).
SediView derives accurate sedimentological information from dip and lithological results. The module helps
reservoir delineation by inferring geometrical information
from depositional features. The module also helps predict
locations and directions of permeability anisotropy.
SediView consists of four separate steps.
1. Data loading. Precomputed dip data are loaded
from the GeoFrame database. Lithological information
can also be loaded from the same database, but may be
manually input or computed from application of simple
thresholds to the openhole logs.
2. Data filtering. Avariety of filtering techniques may be
applied to focus the interpretation on meaningful dips. A
typical filter would be the dip quality index. In addition,
lithological zoning may be specified.
334
3. Structural dip evaluation. This phase evaluates and
removes the effects of post depositional deformation or tilting. An innovative and accurate approach, based on
Local Curvature Axis techniques, is used to derive the
structural dip which can then be removed in the standard
manner (Fig. 8-109).
The structural dip is traditionally determined from
zones where the dips are constant in magnitude and azimuth (green patterns) and that correspond to low energy
deposits (essentially shales or mark). The value of these
dips provide evidence of structural tilt.
However, in thick sandstone intervals with cross-bedding or foresets the determination of the structural dip is
more difficult. SediView makes use of the observation that
the cylindrical or conical bed surface axes within each
bedset are often co-planar. The dip and orientation of this
plane reflects the effect of the post-depositional deformation. The program for automatic structural dip determination in such cases is based on the two following observations, and one assumption (Beaudoin et a/., 1983). The
first observation is that usually a few poles of consecutive
dips that are close in depth appear aligned on a great circle on a SCHMIDT stereonet (Fig. 8-110). This may be
because of sedimentary structures, compaction or microstructures. Whatever their origin, these "local" great cirSerralog 0 2003
Chapter8
Tectonics
I 335
Local curvature axis
(Schmidt projeclion
-
Upper hemisphere)
Figure 8-108 - Stereographic Schmidt projection - upper hemisphere o f : - (a) raw dip data, - (b) dips after initial sequencing and filtering, - (c) dip
after local smoothing and structure delineation - and (g) determination of the local curvature axes, for the interval 142.0 m to 854.3 m,
radial extent = 90"
(adapted from Etchecopar & Dubas, 1992).
cles can be characterized through the orientation of their
poles (Local Curvature Axis).
The second observation concerns the poles of the
local great circles. Most of the consecutive poles of local
great circles, also called local curvature axes, are themselves aligned on a great circle (cf. Fig. 8-110).
The basic assumption is that the plane that corresponds to the local great circle poles, or local curvature
axes, was nearly horizontal during sedimentation and now
represents the structural dip. In other words, the axes of
microstructures or sedimentary structures, which are
recognized as locally cylindrical, are assumed to have
been nearly horizontal during sedimentation.
The first pass of this module computes the local curvature axis of each bedset or laminaset. These axes are
plotted on the stereonet and analyzed interactively to define structural dip common to a group of beds. This value
is assigned to the corresponding interval. The technique
detects the subtle changes in structural dip that can occur
Serralog 0 2003
between depositional episodes.
4. Sedimentary dip determination.After substraction of
the structural dip for each depositional unit, a dip dispersion analysis is performed to organize the bedsets into
groups. The thickness and the azimuth of each depositional unit are then computed. These values are integrated
by the interpreter with other data from images, logs and
cores as well as with general knowledge of the depositional environment in order to infer reservoir elongation and
transport current direction. Results are then displayed
(Fig. 8-109 right).
Figure 8-109 shows a Formation Microscanner dip file
(on the left) that has been filtered, but not smoothed, by
DipTrend processing. This dip file corresponds to a 6 m
zone where a fairly constant structural dip is assumed.
The Schmidt plot shows that the azimuth of the layers
varies significantly (Fig. 8-111 left). Figure 8-111 right
shows wide variation in the orientations of local great circle axes. As can be observed on the Schmidt projection of
these axes the latter fit a great circle, this great circle
335
Next Page
dip
Local great
circle axes
Dips after
Filtered
dip substraction of
structural dip
Cross-section
Dips measured in the well
Figure 8-109 - On the left: the original arrow plot and the determination of the local great circle axes.. In the middle filtered dips and removed-dip
file. On the right: cross-section of the sedimentary dips. The truncation surfaces can be readily determined. The direction of transport current and
its change with depth can be determined from this display. (from Etchecopar & Dubas, 1992).
representing the structural dip. This structural dip can be
removed, as shown in Figure 8-109 middle, after which a
sedimentary analysis is possible. Finally, a cross -section
of these sedimentary features is constructed (Fig. 8-109
right). The changes in the transport current direction were
almost impossible to detect without this dip removal.
(4)
J
-
Figure 8-110 - Sedimentary features. Their dips correspond to orange sequences of small extent fitting great circles that have clearly different axis orientations. The axes of these Sedimentary features fit a great
circle as they correspond to the same plane that characterizes the original depositional surface. This plane is the structural dip (SO)
(from Etchecopar & Dubas, 1992).
336
Figure 8-111 On the left: stereographic projection of dips of Figure 8109. From this plot the precise determination of the structural dip is not
easy. On the right: Schmidt plot of these axes. The fit of the local axes
with a great circle that represents the structural dip plane is precisely
determined in this case
(from Etchecopar & Dubas, 1992).
Serralog 0 2003
Previous Page
Tectonics
1
Chapter 8
I 337
Analysis of borehole images
-
Figure 8-112 - Possible classification
of surfaces seen on images. This
classification can be based on the
m No type
following attributes:
Cross bed
- resistivity contrast
rn Cemented f r a c t u r e - surface proximity
- planarity
Open f r a c t u r e
- dip magnitude
m Bed boundary
- dip azimuth
Unconformi t y
- lithology
m Others
Surfaces with low resistivity contrast,
high proximitx same dip in a sandstone or limestone corresDonds to
sedimentary features (foresets or cross-bedding). Low planarity with
high resistivity contrast may correspond either to erosional surface if a
lithology change is observed or to stylolites in a limestone Isolated
conductive or very resistive surfaces with a dip angle close to 30"or
60"may correspond either to a fracture or to a fault if a change in texture is observed from one side of the surface to the other.
1
Figure 8-114 -Another fault with small drag (close to 1 m). The fault is
confirmed by change in image texture on each side of the surface, and
by dip evolution with depth ("red pattern").
(from Schlumberger, Middle East Well Evaluation Review, no 9, 1990).
Combined with the lithology determination it is possible to determine the nature of the material filling the fault
(Fig. 8-116).
Fault with very small slip (Fig. 8-117), as well as reverse fault (Fig 8-118), are detectable on resistivity images.
X253
Figure 8-113 - On the left - Complex microfold with steeply dipping
limbs very well detected on the FMI images.
Such features are impossible to detect from dipmeter data
- On the right -Another example of normal fault with a small drag.
(courtesy of Schlumberged.
Serralog 0 2003
Figure 8-115 - A normal fault with no deformation is readily detected on
the images because of the change in texture on each side of the fault
surface (courtesy of Schlumberger).
337
Figure 8-116 - An anhydrite cemented fault (from Schlumberger, Well Evaluation Conference, Angola, 1991).
ORIENTATION
- X277
- X278
- x279
Figure 8-118 - Small reverse fault well detected on the images. On the
right a 3 D view of the borehole wall surface
(courtesy of Schlumberger).
Figure 8-117 - Normal fault with very small slip
(courtesy of Schlumberger).
Fault can be detected as well during the drilling with
the RAB tool as illustrated by image of Figure 8-119.
Images also allow the recognition of en echelon faults
(Fig. 8-120).
338
J
Figure 8-119 - Fault detected by the RAB tool during drilling. An apparent fault is seen in both the RAB (right) and FMI (left) images
(courtesy of Schlumberger).
Serralog 0 2003
Tectonics
I Chapter 8 I 339
ted (Fig. 8-122) the orientation of the minor axes clearly
coincides with those of the larger scale trends. The rotation of the major axes with depth (indicated by the arrow
on the frequency plot) is also significant.
Figure 8-120 - Example of en echelon faults detected on FMS images
(from Schlumberger, Well Evaluation Conference, Angola, 1991).
As illustrated by the previous examples it is obvious
that the image tools are much better than dipmeter tools
to achieve a precise description of the tectonic structure.
This is fundamental to precisely put back the reservoir in
its tectonic setting.
Structural dip versus regional or seismic dip
Usually, the dips determined in a well correspond to
the structural dip, and not regional dip, with the latter better defined by seismic surveys. However, where the structural dip remains generally constant over a large depth
interval, it can be assimilated into the regional dip.
Figure 8-121 - (a) Defining the axial plane by the dispersion of the
axes. (b) Summary of dipmeter data illustrating the average dips, the
general trend and the major and minor axes (contribution of J. Henry).
Naturally the plotting of stereogram summaries should
be adapted to the objective of the study and the geological characteristics of the area in which the well was
drilled. For example, it is evident that data plotted from
formations separated by an unconformity should be plotted on separate stereograms or with different marks on
the same stereogram. It is also useful to label or distinguish the plotted values corresponding to principal formations crossed by the borehole, or to label successive average values to illustrate the evolution of dips intersecting
the well.
Reconstruction o f the tectonic features o f a well
The use of stereograms by chosen intervals gives
information such as the average dip that can be used in
the construction of geological cross-sections for example.
But the geometrical analysis of such data can also provide information about the complete structure. Here, once
again, the stereogram will be a useful summary of the
information available. On the stereogram we can plot the
representative points from each interval (obtained from
the interval stereograms). If the well has been drilled
through a monoclinal structure the resulting cloud of
points plotted will give the average regional dip. In the
case of Figure 8-121b these points form a linear cloud,
enabling the axis of the general structure to be defined. In
other cases intervals will show an evolution of points
which define the major axes characterizing the structure
over an interval of several hundred meters. The orientation of such axes can vary with depth, and such information may also be relevant to the regional geometry. Finally
all the axes identified on the interval stereograms can be
plotted on a summary stereogram.
Another method provides a clear comparison between
the various orientations of the axes plotted. The major
axes and the general trend are added to an azimuth frequency plot of the minor axes, eventually taking into
account their direction of plunge. In the example presenSerralog Q 2003
46 minor axes
Figure 8-122 - Azimuth frequency plot of the minor axes
(contribution of J. Henry).
Dip analysis, containing the same information, but
conserving the relative depths of the structural elements
can be summarized in a table (Table 8-3). Some columns
require comment. In the aspect column the blank sections
represent zones of consistent dip, hatched sections indicate intervals showing an organized evolution of dips that
339
enable axes to be defined, and black corresponds to
zones where the dips were too dispersed to give an average dip or indicate an axis. The average dip column lists
the values read from each interval stereogram. In the
axial direction column all of the minor axes constructed
during analysis of the interval stereograms are listed.
Under the title % of good measurements is the proportion
of good dips to total dips on the arrow plots or listings. The
column dispersion factor in o the angular size of the circles containing the clouds of points are listed. Finally in
the last column the numbers attributed to each level that
have been used in the analytical stereograms are given.
Such tables are used as the basic reference document
when analyzing the geometry of different wells drilled on
the same structure.
Table 8-3
Example of a dip analysis table
(contribution of J.Henry).
Cross-section construction
Arrow plots and stereograms constitute a kind of presentation of the dip data. However, they are not enough
descriptive. Structural geologists are more familiar with 2
D (cross-section) or 3 D (block diagram) representations
of a structure. Consequently, to better visualize the tectonic features it is important to establish the cross-section
from the dips recorded in a well.
Cross-sections can be drawn manually either by adopting a parallel model, in which each of the layers keeps a
constant thickness, or the similar model, in which each
layer can be derived from the others by a simple translation. Figure 8-123 illustrates a cross-section drawn
manually assuming a similar fold.
NNE-
OBSERVATIONS
Figure 8-123 - Cross-section corresponding to the Figure 8-47.
The slip (80 m) can be determined by defining the depth interval of the
gamma ray curve that is repeated
(from Schlumberger, Well Evaluation Conference, Iran, 1976).
With work station and adapted softwares it is now possible to draw automatically the cross-sections.
Automatic cross-section construction
To describe structures in a basin, two main types of
information can be used: seismic sections and dipmeter
measurements or imaging techniques. Geologists are
familiar with seismic sections, which are, increasingly
accurate. However, there are several reasons why significant structures may be missed on seismic sections:
- the vertical resolution of seismics is at the best 20 rn,
- structures may be too small (less than 50 meters),
- average dip may be too high,
- velocity contrast may be too low.
In each of those cases, seismic techniques are insuffi340
Serralog 0 2003
Tectonics
cient to achieve an accurate structural description of a
basin. However, dip measurements are infrequently used
by geologists and geophysicists, even though great
improvements have been made in data acquisition with
the development of imaging techniques. The conventional
display of results (''arrow plots") may not be intuitive to
structural geologists, which may be one reason for their
infrequent use.
The final goal of a geologist is to transform the tadpole information into cross-sections and even block diagrams. Unfortunately, the manual construction of crosssections from large amounts of data is tedious and typically inaccurate. A great number of dipmeter measurements remains poorly interpreted because of the lack of
an appropriate software tool to facilitate this process.
To answer this need, special programs, such as
StrucView (Etchecopar & Bonnetain 1992), were developed for the construction of structural cross-sections from
dipmeter or image data. The StrucView program runs on
a SUN workstation using the dip files generated by
DipTrend processing.
An important constraint imposed on the cross-sections
built with the StrucView program is that they have to be
strictly consistent with the dips measured at the well.
Even when meeting this constraint, different cross-sections are possible for a single data set, depending on fundamental choices under the user's control. The main choices concern the zoning of the well and, for each zone, an
assumption about the type of structure. These decisions
are not always clear-cut and may have to be changed
during the course of the interpretation. Therefore, the software was designed to help the user by providing a completely automatic stereoplot analysis of the data, and it
warns when unusual situations are encountered.
Moreover, all important choices are easily modified, and
the consequences of those changes are quickly taken into
account.
I Chapter8 I
341
lysis.
The parallelism implies that in the plane perpendicular
to the fold axis, each layer is strictly parallel to its neighbours. The similarity implies that the position of a layer
boundary in space can be derived from the others by a
simple translation (Figs. 8-124 to 8-126).
Whatever the orientation of the cross-section plane
relative to the structure, it contains a direction called the
"translation direction" along which a layer boundary can
be derived from the others by a translation. For a given
structure the different directions of translation define a
"translation plane". This plane corresponds to the axial
plane for a fold, the fault plane for a fault and the detachment plane for a rollover.
cs2
Figure 8-124 - Structural axis and translation parameters for a fold.
This figure illustrates the determination of the translation direction for
any section across a similar fold. This fold is cut by two cross-section
planes normal (CS,)and oblique (CSd to the fold. In each of these
planes, the different layer boundaries are obtained from each other by
a translation in a particular direction (A or 6 ) which is the intersection
between the axial plane and the cross-sectionplane. SAX = structural
axis (Etchecopar and Bonnetain, 1992).
Hwotheses for cross-section construction
Cross-sections can be described in terms of folds and
monoclines that may be separated by faults. As already
seen, to describe a fold, geologists classically consider
two models (Ramsay, 1967):
- the parallel fold or isopach fold model - each layer
has a constant thickness within the fold,
- the similar fold or isogonal fold model, - each layer
can be derived from the others by a simple translation.
Similar folds versus parallel folds
Since 1929 (Busk, 1929), all the proposed methods for
cross-section construction, such as the kink method
(Suppe, 1985), have been based on the parallel fold
model. The StrucView program allows this method to be
used when the geologist determines that the dip evolution
corresponds to a fold. However, because this model is
appropriate only for fold analysis, and not for structures in
which thickness varies (such as rollover or drag), the similar fold model was also developed for the program. As a
first approximation, this model may be used for fold anaSerralog 0 2003
SAX
Figure 8-125 - Structural axis and translation parameters for normal
fault IF The movement along the fault plane has developed a drag fold
for which the axis is the structural axis (SAX).By contrast with folds,
the translation plane is not the axial plane of the drag, but the fault
plane itself The intersection of the cross section planes (CS1) and
(CSd with the fault plane defines the translation directions A and 6.
The fault plane, which has pole p, is assumed dipping at 65"
(Etchecopar & Bonnetain, 1992).
341
0 10 20
plane, which contains the structural axis (i.e. the fold axis)
and is the bisector of the two limbs (Fig. 3.2-17). Some
potential polarity changes can be detected between two
consecutive dip sequences, and the user can always validate or invalidate these automatic detections. Once the
decisions have been made about possible overturned
limbs, the axial plane is computed using classical stereoplot techniques.
Faults. Admittedly, the fault plane is the translation
plane (Figs. 8-125 & 8-126). Dipmeter data alone cannot
determine a fault plane, so other information (tectonic
style of the basin, tension or compression forces) is needed to specify the fault type (normal or reverse) and the
fault plane orientation. On images, however, the fault
plane can usually be easily detected and located (Fig. 8127) and the strike and dip easily determined.
Figure 8- 126 - Structural axis and translation parameters for reverse
fault F. The movement along the fault plane has developed a drag fold
for which the axis is the structural axis (SAX).By contrast with folds,
the translation plane is not the axial plane of the drag, but the fault
plane itself The intersection of the cross section planes (CS,)and
(CS2) with the fault plane defines the translation directions A and B.
The fault plane, which has pole p, is assumed dipping at 25" in an
azimuth opposite to that of the layers in the drag zone
(Etchecopar & Bonnetain, 1992).
Constructina cross-sections
To generate a StrucView cross-section, take the following four steps:
1. select a zone
2. determine the structural axis and possibly the translation plane
3. determine the cross-section plane and possibly the
translation direction
4. draw the cross-section.
Selecting a zone
Few wells exist where only a single structure is present. Therefore, the first task for the user is to split the well
into subzones, each one corresponding to a single structure. Correctly identifying each structure is not always
obvious, but the user can change the boundaries of the
zone at any time during the analysis. Furthermore, these
boundaries can vary depending on how much detail the
user needs in the cross section. The user can also choose the zoning determined by the DipTrend processing.
Determining the structural axis and translation plan
Both folds and drags are assumed to be cylindrical
structures, characterized on a stereoplot by the bed poles
aligning along a great circle. The pole of this great circle
is the structural axis. For a cross-section based on similarity, a translation plane must be defined that necessarily
contains the structural axis. Because a straight line is not
enough to determine a plane, some extra information is
needed, which is different for folds and faults.
folds. The translation plane for a fold is the axial
342
s 27 w
Figure 8-127 - Example of a fault that is well detected on FMS 2 pads
images. The change in texture on each side of the sine wave confirms
the slip between the two parts of the figure and the nature of the fault.
Its strike and dip are determined by the sine wave.
It is impossible to determine the fault type without
regional knowledge and the study of missing or repeat
sections (only repeat sections can be detected in a single
well, using a gamma ray for example). To approximate the
fault dip from dipmeter data, an assumption based on the
classical observation that the angle between the main
stress and the failure plane is usually about 25" is introduced. Therefore, where a fault appears, it is assumed to
dip about 65" for a normal fault (Fig. 8-128) and about 25"
for a reverse fault (Fig. 8-129). The fault plane can then
be computed from the stereonet, because it is the only
plane that satisfies the following two constraints:
- it contains the axis of the drag
- it dips 25" in the azimuth opposite to the drag for a
reverse fault, or 65" in the same azimuth as the drag for a
normal fault.
In practice, the angle between the main stress and the
failure plane varies with respect to different parameters,
Serralog 0 2003
Tectonics
such as the confining pressure. The user may modify this
angle if regional knowledge indicates that the angle is different from 25” (or 65”). If structural tilting of the bedding
exists, the user is asked to indicate whether it occurred
before or after the faulting. Fortunately, it is easy to compute and draw the cross-sections in both cases (Figs. 8130 and 8-131).
a
C
a
b
C
d
I ChaPter 8 I
343
b
d
Figure 8-130 - (a) Polar projection of the dips showing a “red pattern”
from the “green pattern” (G). (b) Determination of the structural axis
(SAX). (c) Construction of the polar projection of the fault plane by
counting 65”from the great diameter of the stereogram, point P which
must be on the great circle determined previously, or by counting 65”
from the primitive circle for the cyclographic projection which must go
through SAX.(d) Return to the original position
(adapted from Etchecopar,private communication).
Figure 8-128 - (a) Polar projection of the dips showing a “red pattern”
from the “green pattern” (G). (b) Determination of the structural axis
(SAX)and prolongation along the great circle in order to have 65”from
G (point P). (c) Construction of the cyclographic projection of the fault
plane with P a s its polar projection. The cyclographic projection of the
fault plane must go through the structural axis SAX. (d) Return to the
original position (adapted from Etchecopar, private communication).
a
b
C
d
Figure 8-131 - (a) Polar projection of the dips showing a “red pattern”
from the “green pattern” (G).(b) Determination of the structural axis
(SAX). (c) Construction of the polar projection of the fault plane by
counting 25” from thegreat diameter of the stereogram, point P which
must be on the great circle determined previously, or by counting 25”
from the primitive circle for the cyclographicprojection which must go
through SAX. (d) Return to the original position
(adapted from Etchecopar, private communication).
Figure 8-129- (a) Polar projection of the dips showing a “red pattern”
from the “green pattern” (G). (b) Determination of the structural axis
(SAX)and prolongation along the great circle by counting 25” from G in
the opposite direction (point P). (c) Construction of the cyclographic
projection of the fault plane with P a s its polar projection. The cyclographic projection of the fault plane must go through the structural axis
SAX (d) Return to the original position
(adapted from Etchecopar, private communication).
Serralog 0 2003
On images, the fault dip is easily determined either
manually or with workstations similar to FLIP or
GeoFrame for Schlumberger.
Of course, dipmeter data do not provide information
about the amount of slip. Exceptions include viewing the
343
Next Page
size of the drag, which is qualitatively related to the fault
movement, or identifying reverse fault with repetition of
formations (cf. Figs. 8-46 and 8-47). Extra information
from repeat or missing sections is required.
Rollovers. A direct modeling approach has been developed (Etchecopar and Bonnetain, 1992) on the basis of
recent work that shows rollovers as parts of similar structures (White eta/., 1986; Faure and Chermette, 1989). In
this model, the depths of the top and bottom of the subzone where dips are perturbated by the rollover must be
determined. The bottom is characterized by a clear break
in the dip value. The top, which is more subtle, corresponds to the zone where the fault has no further effect.
The distance between these two depths is taken into
account to determine the radius of the listric part of the
fault. The shear angle is about 60". After the depths have
been selected, the angle of the arc starting from the maximum dip variation in the rollover zone is computed. Figure
8-132 is a real example of rollover with dip data, crosssection and image of the listric fault itself. Figure 8-133 is
an example of rollover cross-section construction.
tion and the direction chosen for the cross section.
Because the dip information is acquired along the borehole, the cross-section plane should stay as close as possible to the borehole trajectory. In the StrucView program,
this plane is constrained to go through the first and last
points of the selected zone. The plane is vertical only if
computed along the azimuth of the borehole deviation. If
computed in the direction perpendicular to the azimuth of
the borehole deviation, the cross-section plane is deviated, as is the well itself, as shown in Figure 8-134.
a
R
Figure 8-133 - On the top, the
reconstruction method for rollover.
First, the intersection of each measured dip with the cross-section
plane is computed. Then each dip
trace is translated parallel to the
direction of translation until it is in angle of arc = 75". shear angle = 601
continuity with the others. On the b
right theoretical geometty of layers
infilling the half graben related to a
listric fault. (a) Well crossing the
detachement plane. (b) Well crossing the curved part of the listric
fault. The dipmeter response
depends only on 8'. (from maximum dip value of layer 37"
Etchecopar & Bonnetain, 1992).
W
Figure 8-132 - Arrow-plot from F M dips,
~
section across the rollover
zone and FMS image of the fault plane. The small deformation feature
on the top of the rollover is an assumed slumped zone
(from Etchecopar & Bonnetain, 1992).
Determining the cross-section plane.
The CrOSS-SeCtiOn plane iS a function Of the Well devia344
Figure 8-134 - CroSS-SeCtiOn Plane in Case Of a deviated Well. The dip
of the Ct-OSS-SectionPlanes D1 and D2 depends on their orientation
relative to the well deviation d. (from Etchecopar & Bonnetain, 1992).
In principle, the user can request a cross-section in
any direction, except for rollovers which cannot be consi&red cylindrical structures. Obviously, the most interesSerralog 0 2003
Previous Page
Tectonics
ting cross-section direction is the direction perpendicular
to the structural axis. The direction of translation is defined as the intersection of the cross-section plane with the
translation plane.
Constructing the cross-section
Once the cross-section direction has been selected,
the intersection of each dip plane with the cross-section
plane is computed (this is equivalent to building a "stick
plot" Fig. 8-135). The cross section is deduced from this
stick plot by shifting each dip trace parallel to the direction
of translation as described in the following.
I Chapter 8 I 345
ging the value of a parameter, it is possible to cancel such
an "anomaly". However, because these anomalies commonly indicate secondary features, exercise caution in
canceling them.
Structural
cross'sec tion
Figure 8-135 - Stick-plots in planes of various strikes of the dips shown
on the left
(from Schlumberger, Well Evaluation Conference, Iran, 1976).
Starting from a given dip trace, such as at the top of
the interval under study, the next trace is translated until it
appears aligned with the first. These two traces are then
linked together to form a piece of curve. A third dip trace
is then shifted and aligned with this new piece of curve
and so on for all dip traces down to the bottom of the interval. Finally, the curve obtained is replicated by translating
it down the well to fit all observed dip traces in the interval
(Fig. 8-136).
The curves under construction are interrupted if the
distance between consecutive dips becomes greater than
a user-defined distance, or if too sharp a break appears in
the average value of the dips (e.g., where the well crosses a major fault). In these cases, the structures are divided into different parts. If an anomalous dip is present, the
corresponding dip trace is shifted into place but not linked
to the others, causing a little break in the curves. By chanSerralog 0 2003
Figure 8-136 - Example of structural cross-section corresponding to
Fig. 8-107. The well was split into three main zones. From the bottom
to 460 m the main anticline. From 460 to 290 m a monocline over an
unconformity. The dip pattern at 325 m is interpreted as a reverse fault.
From 290 to the top a ramp anticline
(adapted from Etchecopar & Bonnetain, 1992).
Exploitation of dip data
Knowledge of the structural dip means that the relative
position of the well can be determined by reference to the
top of the structure. It also helps to predict, establish, and
confirm correlations between wells drilled near each other
(cf. Fig 8-2 and Fig. 8-137). To achieve these kinds of correlations, stick plots may be very helpful. Finally, the struc345
tural dips, determined in several wells of a field, are also
useful in tracing contours of given horizons around the
field (Fig. 8-138).
top of a feature that may be a sand lens or reef (Figs. 8139 and 8-140).
Well A
Figure 8-137 - Correlations between wells done with the help of the
dipmeter data.Red lines indicate extrapolation of structural dip from
wells (from Schlumberger, Well Evaluation Conference, Nigeria, 1974).
(Faun distorsion)
Figure 8-139 - Example of a “redpattern”above a sand bed.
Figure 8-138- Example of isobath map done with the use of the dipmeter data. In black from 5 wells. In blue the final structure (from
Schlumberger, Well Evaluation Conference, Nigeria, 1974).
The second stage of dip data analysis involves detecting the deformations of the beds that took place under
tectonic stress. Faults for example can be identified and
located in depth, but only if their formation was accompanied by associated phenomena such as tilting, stretching,
rotation, brecciation, etc.
Fortunately, this is often the case, and these phenomena manifest themselves as red, blue or yellow patterns. Faults are not, however, the sole cause of such patterns as they may in fact correspond to sedimentary features (foreset beds, large scale cross-bedding, etc.), a
fold, a draping or compaction fold, or an unconformity.
The lithology, the facies, the depositional environment,
the mechanical properties of the formations, fluid pressure measurements, the pattern itself and its position in the
stratigraphic series, all have to be taken into account.
when correlating patterns with such phenomena. Hence,
a red pattern at the top of a sandy or carbonate formation
generally corresponds to a draping or compaction fold on
346
In the case of a convex shape, we can, however, define the direction of elongation of that form that is perpendicular to the azimuth of the beds covering it. Its crest is
in the opposite direction to the dip.
A pattern without significant azimuth variation (& 5”)
may correspond to a cylindrical fold with a horizontal axis,
while one with a progressive change in azimuth could
represent a fold with a plunging axis. A succession of
green patterns can draw a “mega red” or “mega blue” pattern. Such “mega” patterns are often related to folds. For
a simple red or blue pattern, the reality of the hypothesis
establishing a fold is proportional to the length of the
depth interval along which this effect is observed : when
the depth interval is thicker, the hypothesis of a fold is
more valid and it can be established with greater certainty.
The drawing of the cross-section of a well may suggest
a wrong position of the well due to an erroneous interpretation of seismic section. In that case the drilling of a sidetrack well may be decided and may allow the crossing of
the reservoir in a better position, and the discovery of
hydrocarbon (Fig. 8-141).
Determination of the type of fold (cylindrical or conical), the plunge of its axis, etc., can only be done by
stereographic projection of the dips corresponding to the
Serralog Q 2003
Tectonics
patterns (Fig. 8-142). Knowledge of the fold type is important to better evaluate the volume of the reservoir and the
potential hydrocarbon reserve (Figs. 8-143 to 8-145).
The probability of the presence of a fault is higher
when the interval along which the patterns are observed
is short, but may still be ambiguous.Associating such patterns with a fold or a fault can be done:
- either by drawing cross-sections with the help of
stick-plots (Fig. 8-146). These plots give a three dimensional view.
- or using a program similar to StrucView (Fig. 8-147).
An extrapolation of the structural dip to the area adjacent
to the well by applying rules of isogone or isopach conservation of beds following a given direction, can help to
represent the tectonic feature of the well and suggest the
presence of faults (Fig. 8-148);
I Chapter 8 I 347
Side track
well
Main
well
Figure 8-141 - The cross-section drawn for the main well clearly show
that the well was positionned on the side of the main structure. A sidetrack well was decided and the cross-section drawn at the same scale.
It confirms the anticline and the reservoir was found oil-bearing
(from Etchecopar 8, Dubas, 1992).
Figure 8-140 - Example of “red patterns” above a reef
- by finding in the well, by correlation technique using
most of the time gamma ray (cf. Figs. 8-45 to 8-47), a
repeating series (reverse fault); or a gap in series (normal
fault). The latter can be detected when well to well correlations are available (Fig. 8-151 see next pages);
- by knowledge of the tectonic style of the basin;
- or by using stereogram plotting techniques, in which
dip data is analyzed in detail. This technique is described
in a following section.
Very often, in the case of tension that usually creates
normal faults, faulted zones, rather than single faults, are
observed. Thus major faults are generally accompanied,
within a short distance, by several, more or less parallel,
faults. The fault blocks between the minor faults played a
significant role during their respective movements. Thus,
a vertical section ’how’ a
Of closered and
blue patterns, usually accompanied by a change in azimuth from one block to another.
Serralog 0 2003
2
3
Figure 8-142 - The influence of fold axis and fold type on the volume
and the oil-water contact
(from Etchecopar).
347
identified as the series repeat, this phenomenon being
detected on gamma-ray logs as previously seen (cf. Figs.
8-45 to 8-47). In such cases it is also possible to determine the net slip. Image data allow the direct determination
of the fault angle and azimuth.
F.UT
a
Gaslwater
contact
Figure 8-143 - Block diagram corresponding to the lower part of Figure
8-136. It is obtained by interpolation of cross-sections from various
directions using a standard countouring package
(from Etchecopar & Bonnetain, 1992).
FAIAT 1
Figure 8-146 - Stick-plot can help to better visualize faults not so
obvious on the arrow-plot presented on the right
(from Schlumberger, Well Evaluation Conference, Egypt, 1984).
J
Figure 8-144 - Contour map of one layer
(from Etchecopar & Bonnetain, 1992).
Figure 8-145 - Stereogram of the
lower part (465 - 840 m) confirming the conical shape of the
fold.
(from Etchecopar & Bonnetain,
1992).
J
A fault’s direction can be determined from the dips it
has created, being perpendicular to the dip azimuth.
However, its inclination can rarely be determined unless
the fault plane itself causes correlatable resistivity features. Such features identify the fault as an isolated dip with
either the same azimuth as neighboring dips (Fig. 8-149),
or 180” out of phase. The first case corresponds to a
conformable normal fault, while the second case corresponds to an unconformable normal fault if there is no
repetition of series. If series are repeated the contrary
applies, the first case corresponds to a conformable
reverse fault and the second to an unconformable reverse fault (Fig. 8-150). Reverse faults are generally quickly
348
Figure 8-147 - On the left: typical Dip
Trend output. The dips are colored
according to their initial structural
classification (“green”, “orange”, ‘ye/low’y. Next: groups of tadpoles belonging to the same structural trend are
linked by colored bands.
On the upper right: StrucView based
on the Dip Trend results.
Stereogram indicates conical folds
(from Schlumberger,Middle East Well Evaluation Review, no 13, 1992).
Serralog 0 2003
Tectonics
Figure 8-148 - Extrapolation of the structural dip
to the area adjacent to the well by applying rules
of isogone conservation following a given direction. From this reconstructed geometty of the
beds it can be concluded that the crossing of a
fault with drag has a high probability (from
Etchecopar,private communication).
I
Chapter 8
I
349
of a well deviated at 10". The arrow-plot shows, from the
bottom up, a "green pattern" dipping to the north, a "blue
pattern" dipping to the south and a "blue pattern" dipping
to the north. The stereogram clearly shows that this corresponds to a cylinder with axis nearly horizontal and
lying ENE-WSW. This geometry could correspond to
several geological models. Four of these possibilities are
suggested in Figure 8-152:
- in a, two reverse faults,
- in b, a normal fault,
- in c, a disharmonic fold,
- in d, the fill of a paleomorphologicalfeature (a channel, an erosional surface on top of a regional disconformitY).
r
L!
Any fault that is not reversed is normal. But its nature
(tension fault, growth fault, etc.), as well as its net slip, can
only be deduced from complementary information such as
local geological knowledge (tectonic style of basin, depositional environment, etc.), and correlation with other
wells.
The methods discussed previously have given precise
and quantitative information on the geometry of formations that intersect the borehole. The goal in interpreting
this information is to describe geologically this geometry.
-I
Figure 8- 149 - Constructing cross-section using dip-data and stereographic projection for normal fault
(from Schlumberger, Middle East Well Evaluation Review, no 8, 1990).
Figure 8-150 - Constructing cross-section using dip-data and stereographic projection for reverse fault
(from Schlumberger, Middle East Well Evaluation Review, no 8, 1990).
Unfortunately a geometric configuration alone is not
enough, often indicating several possible geological
interpretations. To remove this ambiguity it is always
necessary to have additional information (such as stratigraphy, sedimentology, well log correlations, geophysics
and the regional geology). This is illustrated in the follOwing example.
Figure 8-152 shows an arrow-plot, its stereographic
representation and a stick-plot section normal to the axis
Serralog 0 2003
Figure 8-151 - well to well correlations a//owthe detection of normal
faults positively identified while some are confirmed by dipmeter data
(see Fig. 8-146 corresponding to well C)
(from Schlumberger, Well Evaluation Conference. Egypt, 1984).
349
compact brittle beds. One can however adopt an isogonal
fold-type and allow variations in thickness in plastic beds
(shale, sands, halite, etc.) at hinges or at inflection points.
In drawing the structures, based on measured values,
it is also necessary to consider the tectonic nature of the
region.
Finally for the representation of complex structures
one can use block diagrams. However drawing block diagrams is a long process and requires a certain skill at perspective drawing.
Work stations
C
d
Figure 8-152- Arrow-plot, stereogram and a stick-plot of part of a dipmeter log. Geological models compatible with the geometry
(contributed by J. Henry).
Well log correlation could confirm the solution, with the
exception of evidence indicating repeating series.
The structural style or geophysical (seismic) information, or correlation with nearby wells could indicate a missing series, tending to favor b as the probable solution.
Sedimentological knowledge, lithology, facies, seismic
sections or the regional context (embankment at the edge
of a platform for example) could imply that solution d was
the answer.
Although the solution c was selected, as the series
was a well stratified turbidite sequence that had been
compressed and folded in a N-S direction, the presence
of a normal fault seems to be correct as the extrapolation
of the structure based on the hypothesis of an isogonal
folding (translation of the dip along a direction parallel to
the general axis) seems to suggest (cf. Fig. 8-148).
As illustrated by the previous examples, cross-sections present the arrangement and the geometry of the
structures, deduced from dip data interpretation on the
whole well or a part of it.
The position of the cross-section must be chosen carefully, so as to change the picture of the structure as little
as possible. The represented dips will be the apparent
dips on the chosen plane (most often vertical).
One can start with the results of a stick plot, for which
the plane of cross-section must first be defined. It can be
perpendicular to the vertical plane that passes through the
borehole axis. But if this axis is inclined, this section will
alter the geometric picture (for example, over-estimate the
radius of curvature). In this case it will be better to have
an inclined cross-section with an orthogonal axis. The
position of the cross-section can vary along the same well
depending on the orientation of the structures traversed.
When drawing the beds an isopach fold can be preferred. These are the commonest type of fold, particularly in
350
The emergence of work stations in early 1980’s years
has considerably simplified the work of geoscientists. This
is due to several factors:
- easy importation of data from different origins in the
same environment
- set of softwares for processing and interpretation of
raw and computed data, and display and storage of
resuIts
- interactivity
- rapidity
Work stations have relieved geoscientists of all the
repetitive, tedious, sometimes difficult, and time-consuming part of the analytical aspect of their work. They simplify interpretation as they allow users to easily manipulate any kind of data and display results of processing. The
interpretations use softwares containing rules defined by
“experts”.
Integrating all well data (logs and images, cores, tests,
raw and computed data, well seismics) in the workstation
environment aids geologists, petrophysicists and reservoir engineers in their evaluation of formations.
Geologists can improve the structural, sedimentological,
diagenetic and stratigraphic analysis of logs calibrated on
core data. They can put back the reservoirs in their geological setting. Petrophysicists can refine log interpretation and minimize uncertainty taking into account the geological information provided by the logs. Reservoir engineers can better understand the dynamic potential of a
reservoir through recognition of petrophysical volume properties and surface transmissibility.
One of its most attractive features is the ease with
which all of the tedious dip-data processing techniques
can be performed (stereograms, arrow plots, azimuth frequency plots, cross-sections, etc.), freeing the user from
the onerous tasks of plotting the data and enabling him to
concentrate on the objective of all the previously described manipulations : an interpretation.
On the screen open-hole logs, dip results or resistivity
curves, images, and any kind of plots (azimuth, polar,
stereogram, etc.) can be displayed at any scale over a
chosen interval. Zooming in on a given feature or pattern,
and vertical scrolling is also possible. Elimination of
doubtful results, or addition of dips computed from correlations between dipmeter or image resistivity curves made
by hand on the screen, can be made.
Serralog 0 2003
Tectonics
Other applications of
stereographic projection
The stereographic projection is the main mean of
determining certain angular relations in space. Their
applications are explained in the following sections.
Find the line of intersection of two Dlanes
The line of intersection of two planes will be represented on the stereogram by the intersection point of the two
cyclographic representations of the planes. It is, therefore, sufficient to trace two great circles that are the cyclographic representations of the two planes (Fig. 8-153).
The intersection point P is the projection of this line. To
find its dip and azimuth we rotate the transparent overlay
to place P on the N branch of the N-S diameter. First read
from the edge to point P the angular distance that corresponds to the dip, then count, in an anticlockwise direction, the angle between the N of the stereonet and that of
the stereogram.
N2OVi
I Chapter 8 1
351
nes along two lines that define the apparent dip. It
contains, therefore, these two lines, represented on the
stereogram by two points.
The required plane is represented by the great circle
that passes through these two points (Fig. 8-155). The
method consists in tracing the cyclographic representation of the planes, the orientation of which is known (in this
case the vertical planes in Fig. 8-155, which are N 30" E
and N 40" W). In this case the planes are represented by
the diameters on which the apparent angles (pitches) are
plotted, measuring up from the south of the transparent
overlay because they are oriented towards the north. The
plane is represented by the great circle passing through
these two points. The construction of this great circle and
its pole enables its azimuth and dip to be plotted : N 3" E
28" in this case.
N 40' E
N40'E
N 20" W
Figure 8-153 - Definition of the line at the intersection of two planes
and the angles between them (courtesy of Schlumberger).
Determine the anale between two Dlanes
The intersection of two planes defines four dihedral
angles, opposite pairs of which are equal, the total making
360".
One of the angles will be measured on the plane which
is perpendicular to the intersection line and contains the
poles of the planes, hence it will be measured on the great
circle that passes through these poles .The angular distance between the poles P, and P, is measured along this
great circle.
The complementary dihedral angle (for control) is
determined by measuring the angular distance between
the cyclographic traces, also along the great circle (Fig. 8154).
~
?I
Figure 8-154 - Measurement of the
angle between two planes. Two dihedral angles exist. Their addition gives
180" (courtesy of Schlumberger).
Figure 8-155 - Determining the real dip of a bed from 'rte apparent dips
measured in two vertical planes.
Determination of the aDparent dip of a bed in anv azimuthal direction, when its true dip is known
This is the inverse problem of the previous case. It is
sufficient to plot the cyclographic trace of the bed and to
trace the diameter, the cyclographic representation of the
azimuth of the apparent dip, we wish to measure. The
point of intersection represents the stereographic projection of the straight line at the intersection of the bed and
the plane. Move this point to the N-S diameter and read
the value of apparent dip, the angular distance between
the circumference and the point (Fig. 8-156 : the apparent
dip of the previous bed in a direction N 70" W is 9"). This
type of operation is done automatically when constructing
stick-plots.
N70'W
Determination of the true diD of a bed from amarent
dip measured on two vertical planes
The plane that represents the bed cuts the vertical plaSerralog 0 2003
Figure 8-156 - Determining the apparent dip of a bed in any azimuthal
direction, when its true dip is known
(courtesy of Schlumberger)
351
DiD Removal
In determining the original dip of sedimentary formations, that have been subjected to deformation by folding,
it is important to remove the structural dip of the fold from
the measured dip. The problem consists in applying a
rotation of a fixed value and in a fixed direction to a plane,
the amount of this rotation being determined by another
plane. In this case the cyclographic representation of the
structural dip as well as the polar projection of the measured plane must both be plotted (Fig. 8-157).
N20"E
s40"w
Figure 8-157 -An example of structural dip removal. A dip of 30" N 20"
E is computed for a sedimentary feature found in a formation whose
general inclination is 15" S 40" W.Finding the sedimentary dip of the
feature at the time of deposition is equivalent to rotating the major system back to the horizontal. One finds 46" N 25" E
(courtesy of Schlumberger).
To bring the structural plane horizontal it is sufficient to
align its cyclographic trace with a great circle on the
stereonet. Then, by moving the stereograph by an angle
equal to the structural dip, bring this trace to the circumference. Displace all other poles by the same angle along
the small circles. It may occur that this displacement
along the small circles passes through a horizontal position and intersects the fundamental circle of the stereonet.
It is, in this case, necessary, to shift it diametrically on the
fundamental circle and complete the rotation on the opposite side of the stereogram. This type of operation, dip
removal, is also performed by computers.
References and Bibliography
ASQUITH, G.B. (1982). - Basic Well Log Analysis for
Geologists. Amer. Assoc. Petroleum Geol., Methods in
Exploration Series.
Association des Geologues du Sud-Ouest (1974). Divers aspects de la tectonique. Journees de Toulouse,
28-29 novembre.
BATES, R.L., & JACKSON, J.A. (1980). - Glossary of
Geology. 2nd ed. Amen Geol. InsL, Falls Church, Virginia.
BEAUDOIN, B, PINOTEAU, B., & DELHOMME, J.P.
(1983). - Geological aanalysis of logs and dipmeter data
for well-to-well correlation.
BENGTSON, C.A. (1980). - Statistical Curvature
Analysis methods for interpretation of dipmeter data. Oil &
Gas J., June 23, p. 172-190.
BENGTSON, C.A. (1980). - Structural use of tangent
diagrams. Geology, 8, p. 599-602.
352
BENGTSON, C.A. (1981). - Statistical Curvature
Analysis Techniques for structural interpretation of dipmeter data. Amer. Assoc. Petroleum Geol. Bull., 65, p. 312333.
BIGELOW, E.L. (1985). - Making more intelligent use
of log derived dip information. 5 Parts. The Log Analyst,
26, 1, 2, 3, 4, and 5.
BILLINGS, M.P. (1942, 1972). - Structural Geology. 1st
& 3rd ed. Prentice-Haft Inc., Englewood Cliffs, New
Jersey.
BUSK, H.G. (1929). - Earth flexures. Cambridge
University Press.
BUSCH, D.A. (1974). - Stratigraphic Traps in
Sandstones - Exploration Techniques. Amer. Assoc.
Petroleum Geol., Mem. 21.
Chambre Syndicale de la Recherche et de la
Production du Petrole et du Gaz Naturel. Comite des
Techniciens. (1976). - Methodes modernes de geologie
de terrain. Manuel d'analyse structurale. 2b. Traitement
des donnees. Ed. Technip, Paris.
Chambre Syndicale de la Recherche et de la
Production du Petrole et du Gaz Naturel. Comite des
Techniciens. (1983). - Methodes modernes de geologie
de terrain. Manuel d'analyse structurale. 2a. Methodes
d'observation, de mesure et de notation. Ed. Technip,
Paris.
DELFINER, P., PEYRET, O., & SERRA, 0. (1984).
Automatic determination of Lithology from Well Logs. 59th
Ann. fall Mtg. and Techn. Conf. of SPE of AIME, Houston,
Texas, paper SPE 132900.
DELHOMME, J.P., & SERRA, 0. (1984). - Dipmeterderived Logs for Sedimentological Analysis. SPWLA, 9th
Europ. Intern. Format. Eval. Trans., paper 50.
DICKEY, P.A. (1979). - Petroleum Development
Geology. Petroleum Publishing Co., Tulsa.
DICKINSON, W.R. (ed.) (1974). - Tectonics and
Sedimentation. SEPM, special publication 22.
DODGE, C.F., HOLLER, D.P., & MEYER, R.L. (1971).
- Reservoir heterogeneities in some cretaceous sandstones. Amer. Assoc. Petroleum Geol. Bull., 55, p. 18141828.
DONATH, F.A. (1970). - Some information squeezed
out of rock. American Scientist, 58, p. 54-72.
ETCHECOPAR,A. (1992). - Looking at dips from a different angle. Charismatique, 1, Schlumberger Data
Services, p. 16-29.
ETCHECOPAR, A., & BONNETAIN, J.L. (1989). Cross Sections Construction from Dipmeter Data.
SAID/SPWLA Logging Symposium, Paris.
ETCHECOPAR,A., & DUBAS, M.O. (1992). - Methods
for geological interpretation of dips. SPWLA, 33rd Ann.
Logg. Symp. Transactions, paper J.
ETCHECOPAR, A., & BONNETAIN, J.L. (1992). Cross Sections from Dipmeter Data. Amer. Assoc.
Petroleum Geol. Bull., 76, 5, p. 621-637.
ETCHECOPAR, A., & MOTET, D. (1992). - Structural
interpretation of dip data using the Dip Trend method.
Schlumberger, Reservoir Characterization in the Africa
and Mediterranean Region, no 3, p. 24-31.
Serralog 0 2003
Tectonics
I Chapter 8 I 353
~~
FAURE, J.L., & CHERMETTE, J.C. (1989). Deformation of tilted blocks, consequences on block geometry and extension measurements. Bull. SOC. Geol.
France, 5, 3, p. 451-476.
GARY, M., McAFEE, R.Jr., & WOLF, C.L. (1972).
Glossary of Geology. Amer. Geol. Institute, Washington,
D. C.
GILREATH, J.A., & MARICELLI, J.J. (1964). - Detailed
Stratigraphic Control through dip Computations. Amer.
Assoc. Petroleum Geol. Bull., 48, 12, p. 1902-1910.
GOGUEL, J. (1952, 1965). - Traite de Tectonique.
Masson, Paris.
HALBOUTY, M.T., et a/. (1970). - World's Giant Oil and
Gas Fields, Geological factors affecting their formation,
and basin classification. Amer. Assoc. Petroleum Geol.,
Mem. 14.
HATCHER, R.D.Jr. (1990). - Structural Geology.
Principles, Concepts, and Problems. Merril Publishing
Co., Columbus, Toronto, London, Melbourne.
HENRY, J. (1976). - see Chambre Syndical ... (1976).
HEPP, V., & DUMESTRE, A.C. (1975). - CLUST€R - A
method for selecting the most probable dip results from
dipmeter survey. 50th Ann. Fall Mtg. and Techn. Conf. of
SPE of AIME, Dallas, paper SPE 5543.
HOBBS, B.E., MEANS, W.D., & WILLIAMS, P.F.
(1976). - An outline of Structural Geology. John Wiley &
Sons, New York.
HOBSON, G.D., & TIRATSOO, E.N. (1975). lntroduction to Petroleum Geology. Scientific Press Ltd,
Beaconsfield, England.
JEAN, F. & MASSE, P. (1974). - Methodes d'analyse
structurale. Elf-R.E. - Departement Developpement
Formation.
KERZNER, M.G. (1983). - Formation dip determination - An arfificial intelligence approach. The Log Analyst,
24, 5, p. 10-30.
LANDES, K.K. (1951). - Petroleum Geology. John
Wiley & Sons, New York.
LEET, L.Don, JUDSON, S., & KAUFFMAN, M.E.
(1978). - Physical Geology. 5th ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
LINK, P.K. (1982). - Basic Petroleum Geology. OGCl
Publications, Tulsa.
PERRODON, A. (1980). - Geodynamique petroliere.
Genese et repartition des gisements d'hydrocarbures.
Masson, Paris.
PHILLIPS, C.F. (1955). - The use of the stereographic
projection in structural geology. Arnold edit. Ltd, London,
73 p.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Houston.
PRESS, F., & SIEVER, R. (1982). - Earth. 3rd ed. W.H
Freeman & Co., San Francisco.
RAGAN, D.M. (1973). - Structural Geology. John Wiley
& Sons, New York.
RAMSAY, J.G. (1967). - Folding and Fracturing of
Rocks. McGraw-Hill Book Co., Inc., New York.
RIDER, M.H. (1978). - Dipmeter Log Analysis - An
Assay. SPWLA, 19thAnn. Log. Symp. Trans., paper G.
Serralog 0 2003
RUSSELL, W.L. (1951). - Principles of Petroleum
Geology. McGraw-Hill Book Co., New York.
Schlumberger (1979). - Well Evaluation Conference.
Algeria.
Schlumberger (1983). - Well Evaluation Conference.
Afrique de I'Ouest.
Schlumberger (1985). - Well Evaluation conference.
Nigeria.
Schlumberger Ltd (1970). - Fundamentals of Dipmeter
Interpretation.
Schlumberger Ltd (1981, 1986). - Dipmeter
Interpretation. Volume 1 - Fundamentals.
Schlumberger Middle East S.A. (1981). - Well
Evaluation Conference. United Arab Emirates/ Qatar.
Schlumberger Middle East S.A. (1984). - Well
Evaluation Conference. Egypt.
SELLEY, R.C. (1985). - Elements of Petroleum
Geology. Freeman & Co., New York.
SERRA, 0. (1984). - Fundamentals of Well-Log
Interpretation. Volume I : The Acquisition of Logging
Data. Developments in Petroleum Science, 15A, 440 p.,
Elsevier, Amsterdam.
SERRA, 0. (1986). - Fundamentals of Well-Log
Interpretation. Volume 2 : The lnterpretation of Logging
Data. Developments in Petroleum Science, 15B, 684 p.
Elsevier, Amsterdam.
Services Techniques Schlumberger (1970). - Well
Evaluation Conference. Libya.
SMITH, R.G., & BAKER, J.D. (1983). - The Dipmeter
Advisor System: a case study in commercial expert system development. Proc. 38th Intern. Joint Conf. on
Artificial Intelligence, p. 122-129.
SMITH, R.G., & YOUNG, R.L. (1984). - The design of
the Dipmeter Advisor System. Proc. ACM ann Conf.,
ACM, New York, p. 15-23.
SUPPE, J. (1985). - Principles of Structural Geology.
Prentice-Hall, Inc., Englewood Cliffs, New Jersey.
VINCENT, P., GARTNER, J., & ATTALI, G. (1979).
GEODIP - An approach to detailed dip determination
using correlation by pattern recognition. J. Petroleum
Technol., Feb., p. 232-240.
WHITE, N.J., JACKSON, J.A., & McKENZIE, D.P.
(1986). - The relationships between the geometry of normal faults and that the sedimentary layers in their hanging
walls. Jour. Struct. Geology, 8, p. 897-909.
WILLIS, B., & WILLIS, R. (1934). - Geologic
Structures. McGraw-Hill, New York.
WISE, D.V., & Mc CRORY, T.A. (1982). - A new
method of fracture analysis :azimuth versus traverse distance plots. Geol. SOC.Amer. Bull., 93, p. 889-897.
353
WELL LOGGING AND FRACTURES
Introduction
Fracture is a general term that indicates all breaks or
ruptures in a rock, whether accompanied by a displacement or not. It corresponds to a surface along which there
is a loss of cohesion.
These ruptures are caused by tectonic forces (tension,
compression or torsion), by expansion of rocks upon
release of overburden load (i.e. erosion), unequal pressure, or excessive pressure and compression, or by changes of temperature (i.e. dessication), by leaching in the
plane of stratification or schistosity, and also by drilling.
Generally grouped in the category of fractures are:
- crack is a partial or incomplete fracture;
- fissure is “ a surface of fracture or a crack along which
there is a distinct separation. It is often filled with mineral
bearing material‘ (Bates & Jackson, 1980);
-joint is “ a surface of fracture without displacement;
the surface is usually plane and occurs with parallel joints
to form part of a joint set“ (Bates & Jackson, 1980);
- gash is a small-scale tension fissure of several centimeters to a few decimeters in length, and several millimeters to a few centimeters in width. It may be gaped or,
most often, filled with crystals. Several gashes are most
frequently arranged in en echelon (Fig. 9-1). They are
produced by simple shear;
Figure 9-1 - En echelon tension gashes produced by simple shear.
Top : Theory :deformation of a cube under couple.
Bottom : Photograph of an actual case (from Ramsay, 1967).
Serralog 0 2003
- fault is “ a fracture or a zone of fractures along which
there has been displacement of the sides relative to one
another parallel to the fracture” (Bates & Jackson, 1980).
Calling a joint or fault a fracture depends on the scale of
observation.
The fractures may be cemented (filled with crystalline
material: calcite, anhydrite, pyrite or by clay material) or
open. Clearly it is the open fractures which are of interest
for production, because they create substantial permeability, and a preferred flow path for the fluids. The latter are
largely caused by tension or torsion, while closed fractures are generally associated with compression.
Fractures are “generally” transverse to the plane of
stratification, and are usually more or less planar. The two
lips of the fracture plane are locally not separated.
Moreover, the occurrence of fractures is not random (Fig.
9-2). In a constrained formation, the fractures appear as
interconnected systems, each system consisting of a
group of more or less parallel fractures. They result in the
rock being broken up into small volumes or parallelepipeds which can be broken off by the drill-bit or the rotating
drill-pipe.
Figure 9-2 - Fracture systems related to folding. Outer layer in tension,
inner layer in compression, separated by a neutral layer.
L Longitudinal fractures, normal fractures and tension fissures in the
outer layer, inverse fractures in the inner layer.
T : Transverse tension fissures related to stylolitic peaks S’ in the inner
layer. Ds :Sinistral diagonal fractures. Dd :Dextral diagonal fractures.
S :Stylolites with vertical peaks (often more numerous in the outer
layer) and associated with small vertical fissures S’ :Stylolites with
peaks parallel to the bedding planes (subhorizontal).
355
The average gap of a fracture, or fracture aperture, is
often less than 0.1 mm, and so the porosity of fractures is
generally negligible (less than 2%). Boyeldieu et a/.
(1982) have estimated that, if the fracture system breaks
the rock into cubes with 10 cm edges, a gap of 1 mm
Would be necessary to create a Porosity of 1 (Fig. 9-3).
Borehole axis
Figure 9-3 - SuPPOsing a fracture
aperture equal to t, in a rectangular
parallelepiped, its volume is equal
to: t x (a / cosa ) x b , which must
be compared with the parallelepiped
volume V = a x b x h.
If one takes into account the intersection of the fracture plane with the
borehole, the fracture volume is
equal to :t xn x ( r / cosa) x r
and the borehole volume is equal
to: x ?h. For a horizontal fracture in
a cube (a=b=h=l m), the porosity of
fracture having an aperture of 1 mm
is equal to 0.1%. For a fracture with
an angle of 60" the porosity of fracture is equal to 0.2%. Comparing
with the borehole volume, the borehole having a radius of 10 cm, for a
fracture with a dip of 60" the porosity of fracture is equal to 0.2%. For a
vertical fracture crossing the borehole along a diameter the porosity
of fracture is equal to 0.636%
Fractures appear predominantly in brittle rocks, hence
in consolidated formations. Q~~~~~~~and dolomites are
the rocks in which fractures occur more frequently (Fig. 94). Very often fractures disappear on entering formations
which are more plastic (clays or halite), or friable (sands).
Review of general concepts
Relation of fracture to stress orientation
It has frequently been observed that the fracture system, or network, in a given region tends to have the same
orientation as the fault system and is related to the stress
orientation(Fig. 9-5a).
08
Figure 9-5a - The fracture orientation is linked to the principal stresse
orientation which control the fault type.
This is confirmed by the observations made on outcrops and in wells (Fig. 9-5b).
a
!
3
-
b
C
c
0
a
t
c
0
Figure 9-5b - Fractures and faults have almost identical strike patterns
as illustrated by the mapped faults (a) made on Buchan field, and the
strikes of measured fractures and veins in cores of wells 21/1-2 (b),
and 21/1-3 (c) (from Butler et al., 1976).
Figure 9-4 - Frequence of fractures as a function of rock composition.
Metamorphiques quartzites and dolostones fracture more easily than
quartz cemented sandstones and even limestones.
356
However, although the orientation may be statistically
significant, it must be remembered that there can be
considerable dispersion. This is a function of the fracture
types (Fig. 9-6).
Serralog 0 2003
Fractures
Chapter 9
357
strength varies continuously from a minimum in that direction to a maximum in planes perpendicular to it (Fig. 9-8).
The density of fractures (number of fractures per unit
area) is inversely proportional to the thickness of the rock
unit.
Maximum stress axis
F,
/
Angle of internal friction
I
Figure 9-6 - Schematic illustrations of the most common fractures
associated to fold. Type I fractures develop in gently folding anticline.
Type I1 fractures only appear on the anticline when the folding has progressed far enough to deform the rock elastically. Type 111 fractures
develop as a result of shear stresses. Type llla corresponds to tensional stress. Type lllb corresponds to compressional stress
(courtesy of Schlumberger).
Figure 9-8 - Relation of fractures to stress orientations. Planes of maximum shearing stress (S1 and Sd and planes of rupture (F, and FA
which form respectively an angle of 45" and an angle close to 25"-30"
to the maximum principal stress.
0.1
Ql
4
?I
As previously seen (cf. Chapter 8), in homogeneous
isotropic materials under compression, laboratory experiments have shown that compressive stress can be
expressed in terms of a set of three mutually perpendicular axes (Fig. 9-7).
Figure 9-9 - Occurrence of fractures under stress. Cases of single fractures :(a) and (d). Cases of symmetrically inclined sets (b) and (c)
(courtesy of Schlumberger).
B
C
D
Figure 9-7 - Principal stress axes, the stress ellipsoid (A) and conjugated shear fractures (B)or faults (C and 0).
When brittle rocks are stressed beyond their strength,
fracture results and takes the form of two sets of shear
plane. Their directions are bisected by the major stress
axis which forms an angle of approximately 25"-30" with
each of the shears (Fig. 9-8).
Strong anisotropy affects the rupture process. There is
probably a single plane of weakness and the shear
Serralog 0 2003
(a)
(b)
Figure 9-10 - Case of an anisotropic material. S :plane of weakness;
F :rupture plane oblique to S
(courtesy of Schlumberger).
357
Importance of fractures
In formations of low porosity and permeability, the production potential relies on an extensive system of open
fractures. The productivity will vary greatly according to
the number, extent and opening of the fractures and to the
porosity and permeability of the matrix.
As already mentioned, the porosity of fractures is insignificant in all but a few exceptional cases (highly compacted rocks), and makes no significant contribution to
the reserves. However, the presence of fractures may
significantly enhance the drainage surface, and thereby
the contribution of the matrix porosity to the production.
Open fractures considerably increase the permeability
but may cut the potential output of a reservoir if they are
not taken into account during the secondary recovery
phase.
A subvertical fracture system may be fed by an underlying reservoir.
Finally, in the case of fluid injection to maintain pressure, they act as preferred paths for the injected fluids
with the risk of isolating formation blocks which are still
hydrocarbon-saturated, and of having early production of
injected fluids.
Mechanical properties evaluation
Knowledge of the mechanical properties of a rock is
required in several domains.
where Pb, is the overburden pressure, assumed to be
equal to oz.In the simplified Terzaghi and "hard rock"
options a is assumed equal to unity.
Only elastic constrains determine o, = P&. The laws of
elasticity associate to this vertical stress a minimum horizontal stress o h , and the tectonic stresses are estimated
through the value of oHwhich can vary between o h (in a
non tectonic regime), and 0,.
J
The fracture initiation pressure Pb is a function of several parameters. It is expressed by the following relation:
where o h and oHare the minimum and maximum horizontal stresses respectively. oHis usually defined in terms
of the tectonic imbalance factor q / o h . Existence of tectonic imbalance can be inferred from borehole deformation
tests, or from break-out identification with the aid of multiple diameter caliper logs or, better, from borehole wall
images. Pore pressure is obtained from measurements
with the tester tools in new wells, or from pressure buildup tests in producing wells. T~ is the tensile strength. In
Terzaghi or "hard rock" options, a is assumed to be equal
to unity.
To compute the fracture re-opening pressure Pfr the
tensile strength is set equal to zero. So we obtain:
Mechanical behavior of the reservoirs - Stress comTIc)utations
To know if reservoirs require tubing or gravel packing,
or if they can be produced in open-hole conditions, or if
they will collapse, it is necessary to estimate the critical
wellbore pressure P., It can be demonstrated that P, is
expressed by the following relation using the MohrCoulomb failure criterion:
where a = 1 - c&b, C, and Cb being respectively the
rock compressibility (at zero porosity) and the bulk compressibility (with porosity), Ppis the pore pressure, zi is the
initial shear strength ( = zo), and v the Poisson's ratio. o h
is the minimum horizontal stress. It can be obtained assuming a horizontally constrained elastic model and is
expressed, following the Griffith and Mohr-Coulomb failure criteria, by:
7500
7600
Figure 9- 1I - Formation strength analysis (Schlurnberger's courtesy).
358
Serralog 0 2003
Fractures
The previous parameters are computed and displayed
in the MECHPRO program of Schlumberger (Fig. 9-12
two left tracks).
Chapter9
1
359
Table 9-1
Dynamic elastic parameters and how they can be
computed from wireline log data.
I
I
v
Poisson's Ratio (PR)
Longitudinal strain
G
shear Modulus
I
E Young's Modulus (YME)
Kb
cb
Bulk compressibility
C,
I
I
1/2(Ats/Atc)*
Volumetric deformation
Hydrostatic pressure
Rock compressibility
Change in matrix volume
(zero porosity)
Hydrostatic pressure
Biot Elastic Constant
Pore pressure
proportionality
-1
(Ats/Atc)2 - 1
I
Applied stress
Shear strain
( pb/ AtS2) X a
I
I
2G(1 + v)
Amlied uni-axial stress
Normal strain
Hydrostatic pressure
Volurnetric strain
Bulk Modulus
(with porosity)
YME
I
pb(l/AtC2 -4/3Ats2)a
I
PR
G
a
1 - (c&))
CB
Figure 9-12 - Display of mechanicalproperties (from Edwards, 1985).
Dvnamic elastic proDerties
Computation of some of the previous factors require
the knowledge of the dynamic elastic properties. If a sonic
waveform recording has been made using a Long
Spacing Sonic tool (LSS) or a dipole shear sonic tool, Atc
and Ats can be obtained. By combining these two data
with the corrected bulk density, it is possible to compute
the dynamic elastic parameters at each sampling level
(Table 9-1). This is achieved for instance by the MECHPRO program of Schlumberger.An example of the display
of the results is given in Figure 9-11 (right track).
Inherent strenath computations
The inherent rock strengths are computed, for instance, by the MECHPRO program of Schlumberger. They
are related to one another by simple functions expressed
below.
Initial shear strength zi
This parameter is derived by an empirical model based
on Deere & Miller's work (1969) and elaborated by Coates
& Denoo (1981).
'Ti
= 0.025E[
0.008Vclay
Tensile strength z,
The tensile strength is set at one-twelfth of Co as the
average value (Table 9-2): T' , = C0/12
In addition to these applications mechanical properties
evaluation can be used for:
- mud weight control to avoid hydraulic fracturing and
loss of circulation;
- drillability of the formation : adaptation of drilling
parameters, choice of rock bit, of the rotation speed,
weight on the rock bit ...;
- dip data interpretation by enabling a choice between
the faulting or folding of rocks.
Table 9-2
Uniaxial compressional and tensile strengths for rocks
Quartzite, Cheshire
Granite, Westerly
Diabase, Frederick
Sansdtone, Gosford
Marble, Carrara
CO
MPa
TO
461
229
486
50
90
28
21
40
3.6
6.9
(9-4)
- sin$)]
with $ = 30"
(9-5)
$ is the angle of friction in the Mohr-Coulomb failure
model. It is set at 30".
Serralog 0 2003
90
16.5
10.9
12.2
13.9
13.0
+ 0.0045(1 - Vclay)]/cb.106
I
Uniaxial compressive strength C,
Co = zi [2cos$/(l
CO
MPa
Granite
Sandstone
Marble
0.84
0.51
0.75
MPa
0.31
0.28
1.1
Detection of fractures on cores
This detection is generally quite difficult except if the
fractures are cemented. This difficulty is linked to the fact
that when open fractures exist into a formation, generally
the formation is more or less broken and, due to that, the
core barrel is quite often empty, the rock pieces falling
down in the hole.
Detection of fractures from well logs
With the exception of image tools, which can, in most
of the cases, see fractures directly, the responses of the
logging tools are affected only indirectly by the presence
of fractures, due to the “anomalies” on the normal sensor
response produced by fractures. It is only by these indirect effects that the fractures can be detected.
With this in mind, we will now examine, tool by tool, the
effects of fractures on their responses, and so get an idea
of the capacity of each tool for detecting them.
Natural Gamma Radioactivity
To the extent that the circulation of fluids may have
contributed to the precipitation of uranium in the fracture
system, the standard gamma ray tool, or the spectrometry of the natural gamma ray, will show increased activity
levels or increased uranium content in front of fractured
zones (Fig. 9-13).
NATURAL GAMMA RAY
SPECTROMFTRY
Similarly, a comparison between two successive
gamma ray measurements, the first with a non-radioactive mud, and the second over the same section after a
radioactive tracer with short lifetime has been circulated
briefly in the mud, may show up fractured zones. The tracers invade the permeable zones and cause the open
fractures to exhibit increased radioactivity. A further measurement made some time later, or after the start of production, should show decreased radioactivity over the
fractured zones.
NOTE: In cases of deep invasion, the start of production may cause
a temporary increase in activity by bringing the radioactive mud closer to
the borehole wall.
Caliper
Fractured zones may appear on the caliper curve as:
- either a reduction in hole diameter in compacted
zones which are in gauge, most probably due to a deposit of mud cake, especially if lost-circulation material has
been used (cf. Fig. 9-43);
- or an increase in hole diameter due to crumbling of
the fractured zone during drilling resulting in chunks of
various sizes falling away.
These phenomena can best be seen by a four arm
caliper or multi-fingers tool or dipmeters and image tools
rather than the standard two-arm calipers.
An increase on only one of the diameters is due to the
presence of fractures and follows their orientation (Fig. 914). The orientation can be obtained from the inclinometry measurement. The direction of elongation is often that
of a major system of faults and fissures, as has been
shown by various researchers (Babcock, 1978 (Fig. 9-15);
Bell & Gough, 1979; Cox, 1983).
Single steeply
Closely spaced steeply
. _
dipping fractures
Intersecting fractures
dipping fracture
Drill encounten vug
D
Figure 9-13 - Possible fractured zones in this Ordovician formation of
Algeria identified by uranium peaks not associated to thorium and
potassium peaks.
Those uranium peaks may also correspond to stylolites
(from Schlumberger, Well Evaluation Conference, Algeria, 1979).
360
Change in deviation
of hole
E
Figure 9-14 - Several reasons for hole ovalization in fractured zones,
most likely caused by the chipping of the formation by the drilling, and
the fall of the blocks of various sizes comprised between fractures
(courtesy of Schlumberger).
Another example (Fig. 9-16) coming from Timimoun
Serralog 0 2003
Fractures
I Chapter9
361
R 26
Figure 9-15 - On the left :remarkable consistency of hole ovalization over a large area.
On the right :relashionship between hole ovalization and direction of surface outcrop joints (Cretaceous to Devonian sandstones, Canada)
(from Babcock, 1978).
conductivity
Caliper
Figure 9-16 - On the left :dipmeter measurements allowing the analysis of the well geometv. Observe the borehole ovalization and also
the slowing down of the tool rotation. On the right :diagrams of tectonic ovalization in wells of the Timimoun Basin, Algeria
(from Schlumberger, Well Evaluation Conference, Algeria, 1995).
Basin in Algeria, illustrates this high consistency of hole
ovalization and their link with the direction of the current
major horizontal stress direction (N 135"-145").
Serralog 0 2003
Temperature
In a well the temperature gradient in the mud is affected by the presence of open fractures due to the invasion
361
of the fracture system by the drilling mud which has the
effect of cooling the formations. This phenomenon must
.not be confused with gas production which also causes a
drop in temperature.
The circulation of mud disrupts the normal distribution
of heat which depends partly on the difference in temperature between the mud and the formations, and partly on
the thermal conductivity of the rocks. The latter varies
considerably as each type of rock has its own thermal
conductivity (Table 9-3 and Fig. 9-17a). For this reason, a
thermometer log recorded immediately after drilling and
measured on the run-in can be a good indicator of the
types of rock encountered.
The mud at the bottom of the well is usually cooler
than the formations, while near the surface it is hotter (Fig.
9-17b). When circulation has been stopped for some time,
the mud temperature tends to homogenize by thermal
exchange, horizontally by conduction, and sometimes
vertically too, by convection. Thus, temperature changes
at all depths are slow, and some time is required before
the temperatures revert to their original values. Thus, the
mud becomes heated in the deeper part of the well. This
means that the temperature gradient of the mud intersects
the geothermal gradient at a certain depth (Fig. 9-17b).
Above this point of intersection the mud is hotter than the
formations, while below it is cooler. Consequently, mud
invasion in the upper zone increases the formation temperatures, while in the lower zones they are decreased.
Clearly, the interpretation of temperature logs must take
account of the position of this point of intersection.
Ternpemtdncreases
where there has been a partial or total loss of circulation.
Table 9-3
Thermal conductivity of the principal rocks.
Thermal Conductivity in 10-3 Calories/sec/cm/"C
2.8 - 5.6
Shale
3.5 - 7.7
Sand
Porous Lm. 4 - 7
Dense Lm.
6-8
0.9-13
Dolostone
13
Quartzite
1.2 - 1.4
Water
0.065
Gas
Gypsum
An hydrite
Halite
Sulphur
Steel
Cement
Air
Oil
3.1
13
12.75
0.6
110
0.7
0.06
0.35
b
Figure 9-18 - (a) : Temperature log showing
zones of cooling characterized by local
decrease in temperature compared to the
temperature gradient. (b) :Schematic diagram and logs after circulation of a cold fluid
(courtesy of Schlumberger).
Formation density
Figure 9-1 7 - (a) :normal earth temperature gradient. (b): temperature
profile after drilling of a well. Equilibrium is re-established after many
hours or even days (courtesy of Schlumberger).
When a cold fluid such as the drilling mud penetrates
the formation it displaces the formation fluid. The time
taken for the formation to revert to its normal temperature
will depend on the duration of circulation and on the degree of invasion (Fig. 9-18).
Zones which have been more deeply invaded will thus
appear as cooler zones on the temperature log. This will
be particularly noticeable in zones with open fractures
362
In the case of the compensated formation density tool,
two measurements may be considered : the density measurement itself, and the density correction.
- Being a skid-mounted device, the density tool may
face in different directions on two successive runs over a
fractured interval. One would then expect a drop in density if on one of these runs the skid was facing an open fracture. However, the dense, compact formations in which
fractures usually occur will produce low count rates on the
detectors, and hence a high level of statistical variations.
The resulting poor repeatability between successive runs,
which is a feature of high-density formations, whether
they are fractured or not, makes it impractical to look for a
variation in density as an indication of the presence of
fractures across one axis of the hole.
The fact that the tool is unidirectional and not free to
rotate does not simplify matters. However, it may be assumed that, if the hole is eccentric, the long axis will have
the same orientation as the vertical fractures, as long as
Serralog 0 2003
Fractures
these are more or less unidirectional.
- The readings of skid-mounted tools will be affected
by small depressions in the borehole wall which are the
result of small pieces of rock falling away. The short-spacing detector is more influenced by the mud filling these
small cavities than is the long-spacing detector.
- In zones where the caliper indicates a smooth borehole wall, the Ap curve will show a higher correction than
normal in the case of baryte muds (Fig. 9-19). This is
often accompanied by a very low density reading, but may
be localized, blurred or even hidden by the time constant
of the measurement circuit.
-The caliper may indicate sudden changes of hole diameter. When these changes are due to scaling of the formation wall, they can be "seen" by the short-spacing
detector.
I Chapter9 I 363
very high values (Figs. 9-19 and 9-20). This is due to the
high atomic number of barium compared to those of the
elements making up the majority of sedimentary rocks.
This property can be useful for estimating the porosity of
the fractures.
Caliper
Pb
Example 1
Figure 9-20 - Another example of fracture influence on density (Ap)
and photoelectric index measurement (Pe) (courtesy of Schlumberger).
Neutron Hydrogen Index
Figure 9-19 - Effect of baryte on the density correction Ap and on the
photoelectric capture cross-section index. Observe as well the caliper
(courtesy of Schlumberger).
Photoelectric capture cross-section
This measurement, which is made with the litho-density tool (LDT for Schlumberger), is more or less independent of porosity. Consequently it is of no use for detecting
fractures in normal muds.
However, the measurement is very sensitive to baryte,
and so can detect fractures which have been invaded by
baryte muds. When the pad of the tool passes a fractured
zone, the photoelectric capture cross-section will show
Serralog 0 2003
This measurement responds essentially to formation
content in hydrogen, and so it is a measurement of total
porosity. Since the porosity of fractures is usually small
compared to that of the matrix (e. g. in chalk or compacted clays), it is difficult to identify fractures because the
small variation in porosity is masked by statistical variations. In any case, because it is not a directional measurement, the neutron tools will give a more stable measurement. This is especially true in dense, compact formations because of higher count rates and lower statistical
variations.
Sonic measurements
Sonic measurements are known to be sensitive to
fractures in different ways, especially noticeable on full
waveform data. The effects of fractures on body wave
components (compressional and shear), and on Stoneley
waves are different.
Effects on bodv waves
The sensitivity of body waves to fractures depends on
the medium homogeneity on sound propagation. Any
heterogeneity whose dimension is not negligible with
363
respect to the wavelength will have an effect on its propagation, and therefore on the measurement. This is linked
to the large contrast between the elastic properties of the
fluid filling the fracture and the ones of the formation.
Fractures act as major discontinuities.
The effects on body waves include:
- time delay causing slowness differences between
receiver and transmitter modes realized with modern tools
with multiple receivers and transmitters including dipole
(Fig. 9-21);
- amplitude reduction (attenuation) of compressional
and shear waves;
- reflections causing chevron-patterns;
- mode conversions causing criss-cross patterns;
- borehole coupling enhancements causing amplitude
spikes.
Transmitter mode
time path. This is the case with subvertical fractures, or
more correctly fractures which are parallel to the tool axis,
and these are generally not detected by the sonic tool.
Common
receiver
position
Receiver mode
Interval
easured
ki
Fig 9-21 - Receiver and transmitter modes for computation of the
slowness of the sound waves and the Normalized Differential Energy
(courtesy of Schlumberger).
The Variable Density Log (VDL) plot on Figure 9-22
illustrates some of these effects indicated by arrows:
- a step in the arrival time of the compressional, representing a delay of about 20 ps on a depth interval of 3 m
which represents the transmitter-receiver spacing;
- a reduction of the shear amplitude;
- some distortion of the shear arrival caused by mode
conversion;
- a reduction of the Stoneley amplitude after 2500 ps;
- some interference patterns in the Stoneley wave caused by reflections.
Arrival time
In theory, the arrival time of the compressional wave is
unaffected by fractures which do not cross the shortest
364
Figure 9-22 - Effects of fractures on a Variable Density Log (VDL).
Observe the step in the arrival time of compressional wave, the shear
amplitude reduction and distortion, the Stoneley amplitude reduction
(courtesy of Schlumberger).
Whenever the fracture system is more complex, diffraction and reflection will attenuate the compressional
wave to such a degree that detection may not occur until
the second or third peak in the wave train, resulting in
erratic increases in the apparent travel time (so-called
"cycle -skips", Figs. 9-23 and 9-24). This phenomenon is
detected more easily with the older tools. New tools (i.e.
Schlumberger's Array Sonic Service, SDT, or the Dipole
Shear Sonic Imager, DSI) with multiple transmitters, including dipole technique, and multiple receivers are capable
of detecting cycle-skip conditions and may automatically
take steps necessary to avoid cycle skipping that may be
due to presence of fracture.
The shear wave velocity, on the other hand, is more
affected by fractures than that of the compressional wave.
It is seen to decrease while the compressional velocity
remains constant. Thus, by comparing Ats with Atc possible fractured zones can be identified when Ats increases
while Atc remains constant.
Attenuation of body waves
In general, the amplitude of an acoustic wave is
decreased when it crosses a fracture. This is the result of
a transfer of energy. The coefficient of transmission is a
function of the apparent dip of the fracture relative to the
direction of propagation (Fig. 9-25 left). Energy transmission across a fracture depends to a large extent on the
efficiency of mode conversions at the fracture interface.
Serralog 0 2003
Fractures
For acoustic energy to cross a fracture, a propagating
compressional or shear wave must be converted to a fluid
wave at the first fracture interface and then converted
back again at the second.
Transit time (pslft)
90
40
Figure 9-23 - Spikes and cycle skips on the sonic transit time measurement indicating a fractured zone.
Attenuated signal
TIME
Figure 9-24- Explanation of the cycle skipping development.
Fracture dip
Fracture dip
Figure 9-25 - On the left : variations of the coefficients of transmission
as a function of the apparent dip angle of a fracture plane with regard
to the propagation direction (blue and red lines from Knopoff et al.,
1957; green line from Mom's et al., 1964).
On the right :attenuation across a fracture as a function of the angle of
dip of the fracture for waves with a vertical trajectory
(from Moms et al., 1964).
Obviously, the inclination of the fracture is crucial here.
Figure 9-25 (right), from Morris et. al. (1963), is based on
Serralog 0 2003
1
Chapter9
I
365
experimental results and shows that compressional
waves suffer little attenuation on crossing fractures which
are parallel or perpendicular to the tool axis. The attenuation is high when the angle is between 35" and 80". Shear
waves on the other hand, are strongly attenuated by fractures at low angles. According to J. Gartner (in a personal
memorandum) this contrasting behavior could suggest a
conversion from one mode to the other (compressional to
shear) for certain values of inclination of the fractures. The
attenuation decreases with increasing dip. It becomes
very small when the dip of the fracture is above 65" (25"
to the axis of the tool or borehole).
Normalized Differential Energies (NDE)
Loss of acoustic energy, especially of shear and
Stoneley propagations, is also a fracture indicator as illustrated by Figure 9-26. This loss, expressed in dB/ft, is calculated as the difference, known as the Normalized
Differential Energy (NDE), between the 1 foot BHC NDE
and a baseline calculated as the average over 50 ft. The
computation of the difference between energies measured at two distinct receivers, divided by the distance between them is done, in practice, for 1 foot spacing between
the different receiver locations for a single position of the
transmitter (AERx).Similar computations can also be performed in transmitter mode (AERx) (cf. Fig. 9-21).
Receiver and transmitter mode NDE measurements are
affected in opposite direction by borehole enlargements
and formation coupling changes. Their average yields a
BHC Normalized Differencial Energies measurements,
AEBHc,which is practically free from effects of hole size
changes and lithology variations. These three curves are
displayed with a shading between receiver and transmitter curve to indicate the difference magnitude. Fractures
are detected as sharp negative peaks. The NDE computation with ompressional and shear are more noisy than
with Stoneley.
Amplitude spike analysis
Right at a fracture (Fig. 9-27) there is far better acoustic coupling between the borehole and and the formation
than elsewhere. In principle this causes two momentary
increases in amplitude as the tool passes a fracture. The
two peaks should be separated by the tool spacing and,
as previously discussed, the amplitude between them
should be reduced. The analysis starts by computing for
each amplitude a spike curve in order to pinpoint the
depth of local enhencements. For each amplitude curve,
a slow-moving average is computed and then substracted
from the amplitude curve itself, leaving positive and negative departures from the norm. Only positive departures
are kept. As seen on Figure 9-28, pair of spikes on each
curve are centered around the same depth (2655 m). The
spacing between spikes increases with transmitter-receiver spacing. The spikes at the top correspond to the receiver spikes, the spikes at the bottom to the transmitter spikes. The fracture depth is at the center of the two receiver-transmitter spikes.
This analysis has a sharp vertical resolution.
365
Next Page
Figure 9-26 - Energy losses of compressional, shear and Stoneley waves, by computation of the Normalized Differential Energies (NDE), reflection
coefficient and fracture aperture computation from the waveform reproduced on the right (courtesy of Schlumberger).
Figure 9-27- The acoustic coupling between the formation and the borehole is
improved when transmitter or receiver
passes a fracture. This generates an
amplitude spike
(courtesy od Schlumberger).
Crisscross patterns
They are the direct result of mode conversions at the
fracture interface, each set corresponding to a given pair
of propagation modes (Fig. 9-29).
The distinguishing characteristic of any crisscross pattern is the Slope of the two symmetrically juxtaposed heS.
This is a simple function of the velocities of the two propagating modes under consideration:
306
J
Figure 9-28 -Amplitude spike analysis. Compressional amplitude curves are displayed on the left track, spike curves on the middle track,
and spike analysis indicating fractured zones
(courtesy of Schlumberger).
Serralog 0 2003
Previous Page
Fractures
1
Chapter9
I 367
so it is more straightforward.
- The Stoneley wave, being mainly influenced by borehole fluid, does not react much to changes in lithology.
Thus, a strong Stoneley reflection most likely indicates an
open fracture, not a bed boundary.
Variable Density Display
Transmitterreceiver
spacing
Compressional
Shear
StoneIey
Figure 9-29 - Mode conversions at a fracture generate crisscross patterns in the VDL display, with length equal to transmitter-receiver spacing. Note interval transit time lengthening on compressional and
shear. Compressional/Stineley crisscross is not shown as it would
superimpose itself on top of what is already illustrated
(courtesy of Schlumberger).
Focussing on compressionallshear crisscross, which
is less cluttered by direct arrivals than later pairs, the first
step is to determine the compressional and shear speeds
using conventional waveform analysis. This enables the
computation of the crisscross slope. Then, for each given
depth, a window, equal to the transmitter-receiver spacing, is defined with that depth as its midpoint. The measurement of the crisscross energy contained in this windowed set of waveforms is achieved, first by normalization of the waveforms, second by applying a median filter
in order to remove the bulk of the direct arrivals, third, by
summing up the total energy following each crisscross
line knowing the crisscross slopes and a suitably chosen
time window. As the window moves up the well, a log of
crisscross energy is plotted. The vertical resolution of this
crisscross analysis is about two feet.
Effects on Stonelev waves
The Stoneley wave, and especially its low frequency
component known as the tube wave, is a borehole fluid
mode that propagates as a pressure wave along the borehole.
The way fractures affect the Stoneley wave is quite different compared to the way they affect compressional and
shear waves. Acoustic energy is not lost through inefficient mode conversions, but more as a result of moving
the fluid in the fracture system, resulting in a pressure
drop in the borehole. As a result, the direct Stoneley wave
is attenuated, and a reflected Stoneley is generated (Fig.
9-30).
Three advantages of the Stoneley wave analysis can
be considered.
- In fast formations, where we generally look for fractures, Stoneley wave amplitude is much higher than the
other two arrivals (compressional and shear, Fig. 9-31),
Serralog 0 2003
Borehde
Figure 9-30 - The Stoneley energy
attenuation is a function of the fluid
movement in the fracture, irrespective of dipping angle. The fluid is
pumped in and out of the fracture,
thus dissipating energy. The attenuation is a function of the fracture
permeability, therefore its width or
aperture, and the fluid viscosity
(generally mud).
T for transmitter, R for receiver
(courtesy of Schlumberger).
- The roughly constant Stoneley velocity eases the
signal processing task of measuring the reflected signal.
Figure 9-31 -Amplitude and frequency content of the sonic waves.
The amplitude of the Stoneley wave is much higher in the low frequen.
cy region allowing its detection
(courtesy of Schlumberger).
Stoneley reflections
A wave sent by the transmitter is capted by the receiver, but propagates further and can be reflected down by
a fracture and detected a second time later (Fig. 9-32).
Slope =
2
v
Time
Figure 9-32 - Effect of reflections on waveforms
(courtesy of Schlumberger).
367
368
Well Logging and Geology
The time difference between the direct and the reflected arrivals depends on the distance between the receiver
and the fracture, and on the slowness of the non fractured
formation. It changes linearly as the tool moves up closer
to the fracture, and becomes nil when the receiver faces
the fracture. In this case a fraction of the down going
signal can be reflected up and received later. Both effects
result in the chevron patterns visible on the VDL plots
(Fig. 9-33). They are not necessarily visible on compressional and shear waves because the reflection has a
smaller energy and may be masked by the direct arrivals.
The reflection coefficient, r, which is linked to the
transmission coefficient, t, is an important parameter related to the fracture aperture:
r+t=l
600
Frequency (Hz)
Figure 9-34 - Theoretical influence of fracture aperture and dip on the
reflection coefficient r (fcourtesy of Schlumberger).
Resistivities
The electrical system consisting of the formation, the
borehole and the fracture network is represented by the
diagram in Figure 9-35. The fractures are assumed to be
subparallel to the borehole axis and invaded by a conductive fluid.
Figure 9-35 - Fracture identification
from resistivity tools. The induction
“sees”the fractures in series, the
laterolog tools “see”the fractures in
parallel
(courtesy of Schlumberger).
700
800
Figure 9-33 - Effects of reflections on VDL plot. The chevron patterns
are well detected on Stoneley arrivals, exceptionally on shear. The
analysis of the waveform allows the computation of the reflection
coefficient r. The higher values of r mean more permeable fractures
(courtesy of Schlumberger).
The theoretical prediction of r for horizontal and dipping fractures is shown on Figure 9-34.
The reflected Stoneley arrival is separated from the
direct arrival thanks to,a velocity filter. Once the direct and
reflected Stoneley arrivals have been separated, they can
be stacked to obtain the direct and reflected wavelets at
this level. Finally, the reflection wavelet on a certain number of levels, below the fracture, are extrapolated to
obtain the reflection wavelet where the fracture intercepts
the well. The maximum value of the envelope at this location is the final reflection coefficient.
368
Taking into account the current distribution for each
type of device, it will be observed that, in the case of fractures which are subparallel to the borehole axis:
-the induction is unaffected by the fractures which only
constitute a negligible part of the whole circuit since they
are in series for the Foucault currents;
- the electrode tools will be strongly affected because
the fracture network presents paths of lowered resistance
which act as shunt resistances to the current.
In the case of fractures which are subperpendicular to
the borehole axis:
-the induction will be strongly influenced because now
the fractures are in parallel rather than in series, and their
conductivity is very high compared with that of the surrounding formations;
- for the other tools, these fractures continue to offer
paths of lowered resistance.
Thus, a comparison of resistivity values from induction
and electrode tools in zones containing subparallel open
fractures will show substantially lower resistivities on the
laterologs than on the induction (Fig. 9-36). However, we
must bear in mind that the induction measurement is not
recommended in resistive, compact formations because
of low signal level. The analysis will therefore rely on the
relative behavior of the two laterologs (deep and shallow)
and of the microdevices.
Serralog 0 2003
Fractures
~
Resistivity (ohm-m)
+
=igure 9-36 - Comparison between
!he responses of the induction and
laterolog in a fractured zone
(courtesy of Schlumberger).
)) Suau's model
3gure 9-37 - Current distribution in
he case of a fracture which is subparallel to the borehole axis. a) :
Clavier's model; b) Suau's model
(courtesy of Schlumberger).
Chapter 9
369
fluid (gas, oil or fresh water), the resistivity of the LLs will
be substantially less than that of the LLd (Fig. 9-37).
If the mud is less conductive than the original fluids in
the fractures, the separation of LLs and LLd is much less
and may even be inverted.
In compact zones of low porosity which are not fractured, and therefore with little invasion, the two measurements will read about the same resistivity (Fig. 9-38, top
interval).
Because they are pad-mounted, the microdevices only
respond to fractures in front of the pad. But because the
borehole wall tends to crumble near the fractures, it
becomes ovalised, and the pad tends to ride the low side
of the major axis. Hence, the probability of following the
fracture network is increased. Clearly the presence of
fractures will strongly influence these devices because of
their small volume of investigation. Moreover, this part of
the fracture system will be invaded by mud or mud filtrate, and so the resistivities will be much lower (Fig. 9-38,
bottom interval). In addition, crumbling of the borehole
wall will create zones of current leakage.All this enhances
the difference in the resistivity readings
Dipmeter
All dipmeter tools consist of arms, extending from the
tool in different azimuths, so that the pads attached to the
ends of the arms are in contact with the formation (Figs.
9-39 and 9-40). High-resolution sensors on the pads provide the logs necessary for heterogeneity or surface
detection.
HDT 1969
J
Figure 9-38 - Example showing the responses of the laterologs and the
MSFL in a fractured zone (courtesy of Schlumberger).
When the fratures are subparallel to the borehole axis,
the apparent drop in resistivity becomes more pronounced with decreasing depth of investigation although it
remains constant within a fracture. Consequently the deeper-reading device is less affected by the fracture than the
shallow-reading device. A ratio of 1.5 to 2 is commonly
observed between RLLd and RLLs. Moreover, if the
drilling mud is more conductive than the original formation
Serralog 0 2003
Figure 9-39 - The dipmeter tools
(courtesy of Schlumberger
The circuitry for the electrode output is arranged so
that the curve deflections are proportional to the button
current. This button current varies widely according to the
contrast between the resistivity of the formation in front of
the button and the formation surrounding the tool.
Because the voltage is ignored, the curves are recorded
with a "floating zero" on a nonlinear scale designed to
accommodate large variations in local resistivity.one pad
Several parameters must be analyzed with this type of
tool.
369
Figure 9-40 -Another type of dipmeter
tool :the Six Arm Dipmeter of
Halliburton
(courtesy of Halliburton)
Figure 9-42 - FIL presentation for two different orientations of the HDT
sonde. Observe also the caliper difference and the variation of the tool
rotation
(courtesy of Schlumberger).
Resistivitv Curves
As with all the pad-mounted microdevices, only the
pads which are in front of the fractures will be affected and
show a drop in resistivity (Fig. 9-41).
Figure 9-41 - Diagram explaining the detection of an open fracture by
one pad of the HDT (on the left) and the SHDT (on the right).
If the hole is ovalized because of fractures, the usual
orientation of the tool will be with two of the four arms
across the major axis, the other two being perpendicular.
Thus in compact, fractured formations, the two opposite
pads, which "see" the fractures, will show a drop in resistivity, while the other pair, which does not see them, show
a high resistivity value with little or no curve activity (Fig.
9-42), assuming that a low EMEX value has been used.
Superimposing the resistivity curves of two adjacent
(i.e. 90" apart) pads will reveal fractured zones whenever
there is a separation between the two curves. A visual
representation of the presence of fractures is obtained by
shading between the two pairs of adjacent curves (Fig. 943).
This technique was known as Fracture ldentification
Log (FIL), and this presentation could be obtained at the
well-site using the CSU system.
370
Figure 9-43 - Reduction in hole diameter and slowing down in the rate
of rotation of the tool in fractured zones (courtesy of Schlumberger).
Unfortunately, the FIL is often confused by sedimentary features such as laminations, flasers or pebbles, and
the majority of the shaded areas correspond to beds with
an apparent dip rather than to fractures (Fig. 9-44).
This problem has been eliminated with the introduction
by Schlumberger of a new program known as DCA
(Detection of Conductive Anomalies). Conductive events
which cannot be correlated are searched for, and only
these can be interpreted as possible fractures. The events
are defined during GEODIP processing. The conductive
anomaly is then reproduced only if the following conditions are satisfied
- the conductivity exceeds a certain value;
- there is a sufficient difference between the conductivity values;
- the anomaly is detected on a minimum number of
successive intervals.
Serralog 0 2003
Fractures
1
Chapter9
~
371
malies are then indicated along the corresponding azimuth curve. The available fracture indicators with this
presentation include:
- the conductive anomalies revealed by the DCA program;
- borehole rugosity and the axis of ovalisation;
- changes in the speed of rotation of the tool.
J
Figure 9-44 - Influence of apparent dip of beds on the FIL presentation
(courtesy of Schlumberger).
The two electrodes by pad existing in the SHDT tool
gives even better detection of fractures by comparing the
measurements of the two buttons on the same pad (Fig.
9-46). In certain favourable cases, the dip of the fracture
can even be determined (Fig. 9-47).
The three thresholds can be set by the log analyst and
so adapted to local conditions. The results are presented
in the form of a log. The azimuths of pads 1 and 2 are
displayed against depth in the left-hand track (Fig. 9-45).
The shaded areas indicate a difference between the
nominal hole diameter and the readings of the two calipers.
AZ=60"
Figure 9-46 - Display of the SHDT curves showing their interest to
detect fractures. Observe the very good repeatability of the curves
recorded by electrode 1 and the speed electrode. The fractures, indicated by f., correspond to intervals showing a local increase in conductivity of one electrode compared to the other of the same pad
(adapted from Schlumberger).
A polar frequency plot of the conductive anomalies is
also provided (Fig. 9-48). It is used to determine the direction of the fracture network or networks. This direction is
related to the axis of maximum constraint and to the general orientation of the faults in the region.
n
Figure 9-45 - Example of the DCA presentation
(courtesy of Schlumberger).
The azh-~uthsof Pads 1, 2, 3 and 4 are displayed
against depth in the right-hand track. The conductive anoSerralog 0 2003
Azimuth curve of Dad 1
When the hole is not very ovalised, the tool will rotate
because of the torque in the logging cable. The fractures
are then seen successively by the different pads (Fig. 9-49).
As we have seen, the tool normally rotates as it travels
uphole. Any slowing down, stopping or change of direction
in the rotation usually indicates the presence of fractures.
This phenomenon is the result of the pad following a sort
of subvertical or oblique pathway created by crumbling of
the fractured zone for a certain distance (Fig. 9-50). The
tool then resumes its normal rotation, usually after a brief
period of more rapid rotation to release the torsion which
has built up in the cable.
371
t
Stsaogrrun of fractuns
Of conductive
Figure 9-47 - (a):
malies detected by the SHDT tool.
J
Figure 9-49 - A fracture seen successively by two different pads. Note
that the azimuth is constant (courtesy of Schlumberger).
(b): They can be correlated to determine the
dip and the azimuth of the fractures
(courtesy of Schlumberger).
Neutron (pa.)
+
Scale 1% brbivision
Figure 9-48 - Example of a polar frequency plot which provides a
means of orienting the fracture network (courtesy of Schlumberger).
Caliper
Since the dipmeter tool has two measurements of diameter 90” apart, comparison between them will reveal
any hole ovalisation, sudden variations in diameter, or
restrictions due to deposits of mud cake or lost circulation
material in the fractured zones (cf. Figs. 9-16 and 9-50),
Figure 9-50 - Example of reversed tool rotation in fractured zones.
Observe as well the caliper and the Ap curves
(courtesy of Schlumberger).
When the GEODIP program is used for the HDT tool,
or the LOCDIP program for the SHDT tool, there is a noticeable absence of four-pad dips. There may, however, be
some dips which are erratic in dip angle and azimuth
which are due to three-pad correlations. In certain favourable cases (e. g. a single fracture), the conductive peaks
can be correlated to give the dip of the fracture (Fig. 9-47).
Spontaneous Potential
DiDs
In compact fractured formations, the fractured zones
can be identified from the CLUSTER program for the HDT
tool, or the MSD program for the SHDT tool by examining
the values of erratic dips or dips of poor quality.
Correlations which are due to conductivity peaks have no
reason to produce dips which are consistent in either dip
angle or azimuth.
372
Negative “anomalies”, especially if they are important,
are sometimes observed in fractured zones. This is often
explained by the development of an electrofiltration potential when the formation has been drilled with a fresh mud
(salinity of less than 5,000 ppm).
Pyrite filling a fracture could also generate SP deflections.
Serralog 0 2003
Fractures
I
Chapter 9
I
373
Ultrasonic image tools
BoreHole Televiewer (BHTV)
This tool (Zemanek et a/., 1969) provides an acoustic
image of the borehole wall (Fig. 9-51). It is obtained by
measuring part of the acoustic energy reflected from the
borehole wall. The same piezo-electric transducer acts as
both transmitter and receiver.
The formation is more reflective when the rock is
smooth and compact. When it is rugose, fractured or
vuggy, the acoustic energy is more dispersed and these
irregularities then appear as darkened areas on the film.
In the BHTV tool two parameters can be used for fracture detection, the amplitude of the received signal and its
transit time. The amplitude of the signal is reduced due to
the dispersion of energy at the edges of the fracture, while
the transit time will be increased (Fig. 9-52). This tool provides, then, not only a detection of all the open fractures,
but also their orientation and dip. The only requirement is
to minimize the amount of material in suspension in the
mud to avoid having a speckled image due to dispersion
of the energy. Other adverse conditions to be avoided are
excessive mud-cake, excessive hole ovalisation or gascut mud.
Amplitude
Borehole
"Tnsducer
Receiver
Figure 9-52 - Ovalization and breakouts generate reflections and diffractions which reduce the signal
amplitude received at the receiver
and increase the transit time as
shown by the multiple reflections
of a signal not perpendicular to the
borehole wall before its am'val at
the receiver.
An interval of 5 feet, recorded in 1 minute, contains
180 scan-levels, each level corresponding to 250 measured points. This corresponds to a measurement each 2.7
mrn horizontally and 8.38 mm vertically in a 8.5 in. holediameter.
The BHTV and its more recent version (Acoustic
TeleScanner de Schlumberger) has been replaced by
modern tools such as Ultra-sonic Borehole Imager (UBI)
of Schlumberger, the Circumferential Acoustic Scanning
Tool (CAST) of Halliburton, and the Circumferential
Borehole Imaging Log (CBIL) and Simultaneous Acoustic
and Resistivity Imager (STAR) of Baker Atlas (Fig. 9-53).
Filtered transit time
Compensating device
Motor assembly
Gear box assembly
Rotatin electrical
m n J m
Centralizer
,
&i=R?S#2
Rotating seal
Figure 9-53 - On the left
the UBI* tool
(courtesy of Schlumberger).
On the top right the STAR tool
(courtesy of Baker Atlas).
.Transducer
Interchangeable
mtating sub
-7.5 rps
-
Figure 9-51 Fractures can be detected by both the amplitude and the
filtered transit time recorded by the borehole televiewer
(courtesy of Schlumberger).
Serralog 0 2003
These tools are fundamentally sensitive to the borehole micro-rugosities. These micro-rugosities may either
reflect actual geological events (vugs, stratifications, fractures, etc.), or scratches generated by the rock-bit or the
turbine. In that case other information is necessary.
Figure 9-54 illustrates fractures seen by the UBI tool
compared to the resistivity images provided by the resistivity images tools.
373
FMI
ARI
UBI
The borehole coverage is a function of the borehole diameter (Fig. 9-56). In a 8.5 in. diameter the FMi covers
80% of the borehole surface and the EM1 60%.
Description of the
EM1 pad
Photograph of the
OBMl
Figure 9-54 - Three types of borehole Wall images. Observe the good
agreement between the FMI and UBI images. The ARI images
confirms the most important features (courtesy of Schlumberger).
With these tools and the resistivity image tools, for the
first time it was possible to “see” the formation and consequently better describe it. They provide information allowing a much better detection of fractures.
Figure 9-55 - The resistivity imaging tools. At the top, the Formation
Microlmager (FMI) tool of Schlumberger. Below, photograph of the
Halliburton’s half EM1 tool and of a pad. On the right, the photograph of
the Oil Base Mud Imager (OBMI) tool of Schlumberger.
Figure 9-56 - Evaluation of the
borehole coverage of the FMI
tool as a function of the borehole
diameter
(courtesy of Schlumberger).
2
Resistivity image tools
The first resistivity image tool, the Formation
Microscanner (FMS-2 pads), was introduced by
Schlumberger in 1986. It was followed by the FMS-4 pads
in 1988. In 1991 Schlumberger introduced the Formation
Micro-Imager tool (FMI), with 4 pads and 4 flaps. This tool
has 192 electrodes of 5 mm diameter, 24 in 2 rows on
each pad and flap (Fig. 9-55). More recently (2002),
Schlumberger has introduced on the market the Oil Base
Mud Imager (OBMI). Halliburton has developed the EM1
tool (Fig. 9-55), and BakerAtlas the STAR tool (Fig. 9-53).
The EM1 and STAR tools have150 electrodes divided in
two rows on 6 pads.
When one of the button electrodes, on these pads or
flaps, of these tool passes an open fracture in the formation, the current it emits will take the least resistive path.
This will be reflected on the corresponding conductivity
curve as a sharp increase, while the images will represent
open fractures as one or several dark irregular lines (Fig.
9-54).
One of the major advantages of these tools is the lateral coverage it provides due to the large number of electrodes and the number of pads with the shift of each row
compared to the other row by half the electrode diameter.
374
Figures 9-57 and 9-58 illustrate how an open fracture
can be identified by resistivity image tools.
Pad
Formation
Average surrounding
bed resistilty
R*
Areaofthepeak
Electrode’
Figure 9-57 - When the
electrode crosses an ooen
fracture the current will
preferably flow through the
fracture which is less resistive than the surrounding
non-fractured formation.
This creates a conductive
peak with a size similar to
the electrode diameter and
a deflection function of the
mud conductivity and the
fracture aperture.
Healed cemented fractures can also be detected, if the
resistivity contrast with the surrounding rock is sufficient.
These appear as white irregular lines on the images (Fig.
9-59).
Serralog 0 2003
Fractures
Cms-section
Chapter 9
375
Front view
Figure 9-60 - Two other resistivity
tools providing images of the
borehole wall (cf Fig. 9-54)
(courtesy of Schlumbergetj.
RAB
The images provided by these tools have not the vertical resolution of the previous tools but have a higher
depth of investigation giving a better idea of the fracture
prolongation inside the formation.
Borehole wall
Figure 9-58 - Diagram explaining how long an oblique open fracture
will be detected by the two rows of a pad.
Figure 9-59 - (a) Two cemented conjugated white features intersect on
the left strip with an angle of 130". They form an angle of 25" with the
vertical to the bedding. This set of observations indicates that they correspond to conjugated shear fractures with the greatest principal stress
axis perpendicular to the bedding (practically vertical). The strike of the
fracture planes is about N 160" which is also the direction of the mean
principal stress axis. The apparent dips of the fracture planes are 60"
N 170" and 70" N 230". (b) The core photograph confirms the
presence of cemented fractures (courtesy of Schlumberger).
Other resistivity measurements allow the detection of
fracture either during the drilling itself - Resistivity At the
Bit or RAB tool - or by wireline technique - Azimuthal
Resistivity Imager ARI and High-resolution Azimuthal
Laterolog Sonde (HALS) of Schlumberger (Fig. 9-60).
Serralog 0 2003
Other examples of fractures are given in Figures 9-61
to 9-63. They come from FMS, FMI, RAB or STAR tools.
Imager (Pai (L Flap)
Figure 9-61 - On the left :
conjugated shear open fractures in a granite. Few induced
fractures are also visible in
some places {courtesy of
Schlumberger). On the right :
cemented fractures in laminated formations(top :courtesy of
Baker Atlas. Bottom: courtesy
of Schlumbergetj.
As it can be easily understood, the image interpretation must take into account the well orientation (vertical,
deviated or horizontal). As illustrated by Figure 9-64 recorded in a horizontal well a fracture may appear practically
horizontal when in fact it is vertical, and shale layers with
high dip when practically horizontal. To be complete mention must be made of images provided by Azimuthal
Density Neutron (ADN) tool which measures density and
photoelectric index in four quadrans.
375
Next Page
' Figure 9-64 - The appa-
'
shale
layers
rent dip of the conductive
dark beds is in fact equal
to 1.7" N 90" and the
fracture is practically vertical in this horizontal
well. The actual orientation of the well is illustrated by the lower figure
(courtesy of
Schlumberger).
fracture
Figure 9-62 - On the left :open fractures in a dolostone with computation of their mean aperture. On the right :calcite-cemented fractures in
a limestone with determination of their dip. Histograms of the fracture
strikes are shown at the top. Fracture density is reproduced on the
right of each image (from Delhomme and Motet, 1993).
They rarely allow the detection of individual fractures,
only indicating the presence of fractured zones. But the
variations in tool response due to fractures could also be
caused by other phenomena. The following procedure is
recommended to be sure of the origin of these variations.
- It is necessary first of all to look for these variations
in intervals which are likely to be fractured. These may be
zones in which there has been a loss of circulation or an
inflow of fluids, or consolidated formations such as compact chalks, limestones or dolostones, quartzites, anhydrites, plutonic or metamorphic rocks. In general terms, it
is zones of high resistivity which are of interest, and not
porous, unconsolidated sands or plastic clays.
- The next step is to note all possible occurrences by
identifying on each available measurement all the phenomena which could be attributed to fractures (Fig. 9-65).
Tight bed
Open fracture
Figure 9-63 - Open fractures detected by the RAE tool, confirmed by
the FMI. Observe on the FMI images a local small shift between the
pad and the flap images (6 & 7 strips)
(courtesy of Schlumberger).
As illustrated by these numerous figures, image tools
allow a direct detection and analysis of the fractures.
Other logging tools are not capable of detecting fractures
themselves'
but by the effect that the fractures have On
the log measurements.
376
Figure 9-65 - Several effects on different measurements confirm the
presence of fractures (courtesy of Schlumberger).
Serralog 0 2003
Previous Page
Fractures
I Chapter 9 I 377
~~~
The probability of fractures is in fact much greater than
the phenomena observed on the logs may indicate. Thus,
if several of the phenomena already described are detected, it is reasonable to conclude that fractures are present.
Schlumberger had developped a program for the
detection and evaluation of fractures. It was commercially available under the name of DETFRA. This program
(Boyeldieu & Martin, 1984)grouped all the known fracture indicators into five categories : electrical, acoustic,
radioactive, electromagnetic and multi-pad.
Each log was analyzed, and a fracture probability was
estimated using certain criteria (threshold, median and
maximum probability (Fig. 9-66).The probabilities were
then combined using bayesian logic. Thus, two criteria
with individual probabilities PI and P, would have a combined probability which was given by:
P = 1 - (1 - P,)(1
- P2)
(9-7)
Fracture evaluation
The evaluation of fractured zones requires the following information
- depths of the fractured zones;
- types of fractures : natural (open or cemented) or
induced;
- orientation (dip and azimuth) of fractures;
- vertical and lateral extent of fractures;
- total fracture length per unit volume;
- fracture density: number of fractures per unit length;
- fracture aperture;
- fracture porosity;
- fracture permeability.
Well logs do not provide all of this information, only the
following being obtainable.
Depths of fractured zones
This is the simplest information to obtain from the logs,
especially from the image tools. So there is no need to
elaborate.
L
P ~ ~ Y D
Ovality 1-3 2-4 in inch
Type of fracture
The image tools can usually differentiate between
open fractures, fractures induced by the drilling process,
and healed cemented fractures (Figs. 9-54,9-59,9-61,962,9-63and 9-68and 9-69).
Figure 9-66 - Fracture probability in the case of a caliper measurement
(from Boyeldieu et al., 1984).
This rule was associative, and could be extended to an
unlimited number of probabilities. The results were presented in the form of a log (Fig. 9-67).
c
o
y
:&
v
*
"
Figure 9-68
Figure 9-69
Figure 9-68 - Open fracture and stylolites. Figure 9-69 - Few conjugated shear fractures and induced fracture in the middle and on the sides
of the images (courtesy of Schlumberger).
Figure 9-67 - Example of results obtained with the DETFRA program
(from Boyeldieu et al., 1984).
Serralog 0 2003
For the other tools, only open fractures will affect the
log responses and be detected. In any case, it is only
open fractures which are of interest for production. Hence
every fracture which is detected as a conductive anomaly
377
is by definition open. However, not every conductivity
peak is a fracture.
Construction
of the fracture
network
and
Fracture orientation
3D visualization
There are two parameters to be determined: dip and
azimuth. The image tools are the only tools which allow
the determination of both the orientation and the dip of
fractures (cf. Figs. 9-59, 9-61 & 9-62).
The dip cannot be determined with any certainty from
the other measurements, because even if a correlation is
made between conductivity peaks, there is no guarantee
that they all belong to the same fracture.
If one now considers the size of an event detected by
a pad (Fig. 9-70), one can attempt to define two possible
dips and select the ones which show the most constant
values. These data must also be plotted as a function of
pad azimuth.
First network
’
Second network’
Figure 9-71 - The fracture network can be constructed. This allows the
determination of the fracture and stress orientation
(from Delhomme & Motet, 1993).
The fracture trace length (F,,,,) is the cumulated fracture trace length seen per square foot of borehole wall. It
is computed by the following equation:
(9-8)
~~
with :
R = borehole radius
C = borehole wall coverage by pads as a fraction
H = height of the sliding window
z = depth
JzH(i)is equal to 1 if the fracture i lies within H centered at depth z, otherwise is equal to 0.
Figure 9-72 explains how fracture length can be detected and computed. Figure 9-73 illustrates a case of interupted fractures.
Figure 9-70 - Determination of fracture dip from the length of conductive anomalies on a single pad. Several explanations are possible
(courtesy of Schlumberger).
The azimuth can be determined if fracturing is accompanied by hole ovalisation, or from a polar frequency plot
of conductive anomalies detected by the DCA program.
Figures 9-15 and 9-16 show the consistency of results,
and their correlation with the predominant fracture or fault
directions.
The two buttons on each pad of the SHDT provide a
means of determining the apparent dip of the planes of
the fractures picked up by each pad. The dip and azimuth
of the fractures can then be defined if we assume that the
two anomalies correspond to the same fracture, or at
least to the same system of parallel fractures (cf. Figs. 947 & 9-48).
With image tools a fracture network can be realized
(Fig. 9-71) and the paleo-stress orientation determined.
Fracture length
As previously illustrated individual fractures can be
identified with the image tools. This allows the determination of the number of fractures in a given window, and of
their length (Fig. 9-72).
378
Figure 9-72 - Diagram
explaining how the fracture
length can be computed.
Fracture density
The corrected fracture density (FD,) is the number of
fractures observed per foot inside an interval of height H
corrected for the orientation bias. It is given by the following equation:
(9-9)
Serralog Q 2003
Chapter9
in which :
Bi = apparent dip of fracture i
IzH(i)is equal to 1 if the fracture i lies within H centered
at depth z, otherwise is equal to 0.
cubes summed over an interval of height H:
I-CJJ.h:
y
A F & ~
(9-11)
JZHOI,
The area corresponding to the conductive peak linked
to the fracture (cf. Fig. 9-57) allows the estimation of the
fracture aperture and through that the evaluation of the
fracture porosity (Fig. 9-75).
Figure 9-73 - Example of fractures interrupted when joining a plastic
(shaly) bed. They are detected and their aperture determined all along
their trace. Conductive layers are also isolated in the processing
(courtesy of Schlumberger).
Figure 9-75 - The additional current (A) created by an open fracture
filled by the conductive mud (R,) divided by the fracture aperture
(Win mm) is function of the ratio of R, (or RJ over R, .
Hole axis
Apparent
interval
between
real
interval
between
fractures
9-74 - Diagram explaining how the fracture density can be evaluated
from the images.
With the other tools this can be evaluated from the frequency at which the fracture indicators occur, notably on
the dipmeter and on the FIL (Fig. 9-42) and DCA (Fig. 945) presentations, and from the porosity of the fractures.
This can be evaluated by various means.
Fracture aperture
The mean fracture aperture (AF,) is the mean of fracture trace aperture averaged over an interval of height H:
Sibbit & Faivre (1985) related the opening (in pm) of
vertical and horizontal fractures to the conductivity measured by the Dual Laterolog (DLL) tool and the difference
between the deep (LLd) and shallow (LLs) resistivities.
They also showed a relation between their lateral extent
(depth into the formation) and the same Dual Laterolog
measurements. In the case of vertical fractures (parallel to
the tool axis) the two measurement curves separate (LLd
> LLs) and their difference is proportional to the product of
the fracture aperture, E , and the conductivity of the invading fluid, C
., For horizontal fractures (perpendicular to
the tool axis) the two curves show a resistivity decrease
over approximately 0.8 m (Fig. 9-76). Again the separation is proportional to the product of the fracture opening
and the invading fluid conductivity. The image tools enable to determine if the fractures are vertical or horizontal.
As previously seen an evaluation of the fracture aperture can be obtained from the Stoneley wave analysis (cf.
Fig. 9-26).
Fracture porosity from images
(9-10)
with :
lj = length of segment j
aj = local aperture of segment j
One determines the mean hydraulic fracture aperture
AF, which is the cubic root of the fracture trace aperture
Serralog 0 2003
The apparent fracture porosity (Qf) is estimated as the
ratio of two areas: the apparent area of fractures seen on
the borehole wall (Fig. 9-75), and the area of the borehole wall covered by the image:
(9-12)
379
380
Well Logging and Geology
The first term is always very small and can be ignored.
The matrix porosity of compact fractured rocks is also low
(usually less than 10 %) while Pef is also very small
(0.358 for water, 0.48 for oil and 0.807 b/e for salt water).
We can therefore write as a first approximation:
The porosity QXp is derived from the density-neutron
combination, and includes both matrix porosity and fracture porosity. This gives:
(9-16)
Figure 9-76 - Relationship between the fracture aperture e in mm (a) :
for vertical fracture and the conductivity; (b) :for horizontal fractures
and the resistivity (from Sibbit & Faivre, 1985).
Now, we can show that:
The results of the previous analyses are reproduced
as logs versus depth (Fig. 9-77).
PeBa
(Pe)Ba = 070
(9-17)
and further, as a first approximation, we can take:
which gives:
Note: The last equation only holds if the borehole wall is smooth,
so that the pad fits closely to the formation. Otherwise there may be a
cave due to crumbling of the borehole wall filled with barite mud. It is
necessary, therefore to examine the caliper and the density correction
before applying this formula. We must also bear in mind that, being a
unidirectional tool, it will only analyze the part of the formation in front of
the pad, and so it will not necessarily measure the total fracture porosity. In any case, if the hole is ovalized due to the presence of fractures,
the pad will usually ride the major axis of the hole, and so face the fractures. The measurement will thus be representative of the fracture porosity since it is unlikely that there is another fracture network at 90" to the
first when the hole is ovalised.
Figure 9-77- Example of display of resistivity image analysis. From the
left to the right are successively reproduced: fracture density, fracture
length, fracture aperture and fracture porosity
(courtesy of Schlumberger).
Fracture porosity from photoelectric index
We have already seen that the photoelectric capture
cross-section is strongly influenced by baryte muds, and
this feature can be used to evaluate fracture porosity.
The following equation introduces the electronic density
Pe pe = C Pei pei
(9-13)
Or, for the case of fractured rocks invaded by baryte
muds:
380
Fracture porosity from dual laterolog (DLL)
Boyeldieu et a/. (1982) proposed the following equation for fracture porosity after studying the effects of fractures on the deep and shallow laterologs, and making certain assumptions
where Qfr is the fracture porosity, (@fr)c
is the computed
fracture porosity and CLLsand CLLd are the conductivities
in mhos of the LLs and LLd; m is between 1.3 and 1.5.
The assumptions made by the authors are as follows
- The fracture system is seen by both laterologs as a
system of resistivities in parallel with the compact, nonfractured, formation (a perfectly reasonable assumption).
- There is no invasion of the non-fractured part of the
Serralog 0 2003
Next Page
Fractures
formation (the blocks contained within the fracture system), but only of the fracture system. This assumption is
justified by the very high permeability of the fracture system compared with that of the rock itself, so that the overpressure of the mud column will act preferentially on the
fracture network.
- The invasion of the fracture system is not too deep,
but sufficient to ensure that the LLd reads the virgin formation while the LLs reads the flushed zone. The validity
of this assumption will depend on the type of mud and on
the degree of opening of the fractures. If the losses observed during the drilling are low, it can be assumed that
the openings are small and that a mud-cake was able to
develop and limit the invasion. In this case the assumption is valid. If the losses were considerable, the invasion
will be deep, and we can no longer assume that the LLd
reads the virgin zone.
- The water saturation of the uninvaded fracture system is almost zero. This is a reasonable assumption given
the permeability of the fractures.
- The filtrate saturation of the invaded fracture system
is 100 %. Again, due to the high permeability of the fractures, we can assume that all the hydrocarbons have
been flushed.
The authors then derived the following inequalities:
I
Chapter9
I
381
Formation characteristics
(9-20)
and
(9-21)
where $ma is the matrix porosity, $fr is the fracture porosity, S,,, is the water saturation of the non-fractured,
uncontaminated formation. Subtracting equ. 9-13 from
equ. 9-14 gives:
(9-22)
The above hypotheses assume that SXofr= 1 and Sdr
= 0. This then gives equ. 9-16 by substituting conductivities for resistivities.
However, as the authors themselves pointed out, the
best results are obtained when the mud resistivity is about
equal to that of the formation water, and when the formation contains hydrocarbons.
In water-bearing sequences, on the other hand, the
two salinities (mud and formation water) should be very
different. In this case the authors proposed the following
equation:
(9-23)
Figure 9-78 shows an example of results from an
interpretation of very compact, fractured formations.
Serralog
2003
Figure 9-78 - Example of the computation of fracture porosity from the
dual laterolog in granite (courtesy of Schlumberger).
Determination of fracture permeability
Mathieu et a/. (1984) have estimated that fracture permeability can be determined from an analysis of Stoneley
wave detected by a tool which records the complete
acoustic wave train. The results they obtained in a solid
crystalline formation seem encouraging.
In any case, the fracture permeability is controlled by
the minimum aperture of the fracture.
Fracture effects on reservoir interpretation
Through their influence on log measurement it is
obvious that fractures affect the reservoir evaluation.
Tortuosity factor m
In a reservoir the influence of the pore geometry on
electrical measurements has been recognized for many
years (Fig. 9-79). The well known Archie’s equation links
381
Previous Page
the resistivity in the virgin reservoir, R,, to the porosity, @,
the saturation, S,, in formation water, R
,, and to the changes in pore geometry in the form of the tortuosity factor,
also known as the cementation or Archie's factor, rn :
R, I R, = a I @mS,n
(9-24)
Figure 9-79 - On the top left, the pore geometry is modelled as a
set of water-filled tubes within a cube of rock of length L, the
average length of each tube being La
Below, the flow path length (in red) and pore radius are dependent on
the variations in grain size.
The cross-sectional variations can be modelled as step changes.
On the right the diagram shows the relationships between path length,
porosity and Archie's factor m. The longer the path length, the greater
the value of LJL
(adapted from Waffa, in Schlumberger Middle East Well Evaluation
Review, no 2, p. 49-57, 1987).
The effects of pore geometry on current flow path can
be modelled as a set of tubes showing variations in their
cross section area due to variations in grain size. The
length of the current flow path, La,can be compared to the
length of the reservoir cube, L. For intergranular porosity
the ratio La/L is close to 3.5. An open fracture, filled by
mud, will create a short circuit, since the open fractures
are more or less rectilinear planes. Consequently, one
would expect for an horizontal fracture a ratio La/L= 1 and
a tortuosity factor to be close to 1, at least when the porosity is due to the fractures, and the current lines are parallel to the plane of the fractures. In fact, even if the fractures have not been healed, there will be crystals in the fractures which are not evenly distributed, and these will
increase the tortuosity. In addition, the fractures are not
always planar or indeed open, and they are frequently at
an angle to the borehole axis. Finally, there are often
several criss-crossing fracture systems. As a result, the
tortuosity factor, rn, is always greater than 1, but usually
well below 2 or 2.3, the values observed in compact formations, and more usually around 1.4. It depends on the
fracture porosity compared to the formation porosity (Fig.
9-80).
382
vugs
Figure 9-80 - The effects of fractures and non-connecting vugs on the
Archie's factor m (from Waffa, 1987).
Saturation evaluation
In fractured reservoirs, electrical current travels along
two paths, one through fractures and another through
intergranular porosity. The entire system can be modelled
as a parallel resistance network (Fig. 9-81).
Circuit in fracture
Circuit of equivalent resistances
Figure 9-81 - Modelling of the electrical current travel in a fractured
reservoir (adapted from Waffa, 1987).
In the fracture path one can write:
R,=%
(9-25)
4fI
and in the intergranular pore system:
Serralog 0 2003
Fractures
(9-26)
where index ma is for matrix with intergranular porosity, and fr for fracture.
The measured resistivity can be defined in two ways:
(9-27)
or
Rma+Rfr
Rt=RmaXRfr
and neutron hydrogen index combination. These different
values are introduced in the law of parallel resistances
which controls the current path into the formation. This
law is given by the following equation :
- Vt 4Sxo-Vf
Rin Rw
Rrnr
-+
46x0
(9-31)
From this relationship the value Rin is computed and
introduced in the Archie’s equation for the invaded zone
assuming n =2 :
(9-28)
This simplifies to:
I Chapter9 I 383
(9-32)
The only unknown is Sxo which can be computed
through the following equation:
(9-29)
(9-33)
It often happens that vugs are associated to fractures.
The rn factor being affected both by fractures and vugs
(Fig. 9-81), it is necessary to compute rn continuously.
This can be realized thanks to the electromagnetic measurement. As generally shale is absent or in very poor
amount in this type of reservoir, one can write the transit
time of the electromagnetic wave as follows :
MSFL
In this relationship the electromagnetic wave transit
times are generally known or obtained from tables or
charts:
- in filtrat,,,t(
is close to 25-32 ns/m as a function of
salinity),
- in hydrocarbon (tphcis between 4 , 7 4 2 ns/m for oil
and 3,3 ns/m for gas),
- in solid fraction (tpmavaries between 8,7-9,l nslm).
The porosity is evaluated by the density-neutron combination.
The only unknowns are finally the resistivity of the fluid
in the flushed zone (Rin) and S,, .
The electromagnetic wave travels in a zone very close
to the borehole wall (Fig. 9-82). Its travel time is sensitive
to the water content in this zone. In fact, in this zone fluids
are composed by original irreducible formation water,
retained by capillary forces, by mud filtrate and, possibly,
by hydrocarbon. The mixture formatio- water-filtrate has a
resistivity, Ri,, which depends on the salinity of the two
fluids and of their respective percentage: V, and $Sxo.The
estimation of V, creates a problem. The quantity of formation water retained by capillary forces is generally unknown, this quantity varying with the nature of the grain or
crystal surface (smooth or rugous). In the case of a drilling
with water-base mud one can estimate this volume as, for
instance, 10% of the filtrate volume in the invaded zone
(V, = O,l$Sxo). In the case of a drilling with oil-base mud
V, corresponds to @SWirr.
$Sxoor @Swirr
is computed from
Archie’s equation assuming rn = 2 at the beginning.
Moreover, the total porosity is deduced from the density
Serralog 0 2003
Figure 9-82 - Comparison of the formation volume investigated by
MSFL and EPT measurements. The
MSFL has generally a deeper depth
of investigation than the EPT.
Consequently, it can be slightly
affected by the virgin zone.
EPT
Saturation in the invaded zone is given by the previous
equation:
sx0=pd+ RXO
(9-34)
From this relationship, the rn value can be computed
and introduced in the Archie’s equation for the virgin zone
in order to evaluate the water saturation :
m=
loga+logRn-nlogsxa
~Og4,+10gRxa
(9-35)
and
(9-36)
$, V, and Rin are computed using the new value of rn
until a convergence on Sxo.Figure 9-83 illustrates results
383
of quantitative evaluation in a fractured formation.
ble. The lithology determination must use neutron-sonicthorium-potassium-photoelectric index combination.
Figure 9-84 - Example of cross-plot combining the formation factor and
the porosity derived from the neutron-density combination
(from Suau et al., 1978).
Figure 9-83 - Example of log interpretation in a low porosity formation.
Four zones were tested. Zone A had better production then the more
porous zone 0.Zone A appears to be fractured while zone D seems to
be made up of vuggy porosity. Fractures near zones B and C may
have contributed to the higher production though with lower vuggy
porosity. Observe the drop of m in front of fractured zones
(from Waffa, 1987).
100
1w
Porosity (%)
POmSity (%)
-
Formation factor Porosity relationship
As the fractures influence resistivity measurement and
have a very low effect on density-neutron measurements,
the relationship between the resistivity formation factor
and the porosity will be affected. If the porosity is plotted
on a logarithmic scale as a function of the formation facfractured zones will appear as zones
tor (FR = RJR,),
having the lowest values of FR for a given value of porosity in a low-porosity zone (Fig. 9-84). This is due to the
drop in resistivity associated with fractures.
Similar plots can also be made by replacing FR by R,
or RLLd (Fig. 9-85) and (I by At (Fig. 9-86). These plots
allow the computation of the rn factor
Figure 9-85 - Examples of crossplots of Rt vs porosity
(from Aguilera, 1980).
Lithology determination
In compact, non-fractured formations, the lithology and
the mineralogy of the reservoirs is easily determined from
the various logging measurements using Z-plot techniques (cf. Chapter 2). In the fractured zones, the readings of the density are frequently affected by caves or
borehole rugosities. Consequently, they are often unusa384
R,
Figure 9-86 - Plot of Rt vs At showing a low value for the cementation,
or tortuosity, factor m (from Aguilera, 1980).
Core orientation
Images of the borehole wall compared to photograph
Serralog 0 2003
Chapter9
I
385
of the external surface of a core allow the core orientation
as illustrated by Figure 9-87.
Closing
Figure 9-89 - Illustration of possible slippages
(courtesy of Schlumberger).
horizontal (q,)stresses (Fig. 9-90) and to pore-fluid pressure (Po).
The vertical stress corresponds to the weight of the
rock lying above. It is called the overburden. It is calculated by integrating the overlying rock's density with respect
to depth.
I
Figure 9-87 - Core orientation by comparison of the photograph of the
external surface of the core by the AUTOCAR apparatus with the FMS
image of the borehole wall (courtesy of ELF and Schlumberger).
Hole slippage
A careful examination of borehole shapes imaged on
UBI survey (Fig. 9-88) may reveal slippage or shear
displacement along fault. In few cases, the slippage may
be so large that the borehole can collapse and the tubing
or casing damaged (Fig. 9-89). The well is lost!
1
Figure 9-90 - Stresses and pressures undergone by a piece of rock at
great depth assuming a continuous fluid path from surface to
subsurface (adapted from Brie et al., 1985).
Figure 9-88 - A cross-section of the borehole shape may reveal slippage along the fault plane. The right figure corresponds to UBI data in
downward perspective (courtesy of Schlumberger).
Detection of borehole damaae in deviated wells
In subsurface, rocks are submitted to vertical (OV) and
Serralog 0 2003
The horizontal stresses, generally anisotropic, correspond to:
- elastic and plastic deformations caused by the overburden:
- tectonic stresses linked to the deformation of the
earth crust:
- stresses stored in the rock resulting from events linked to geological history.
The pore-fluid pressure may be equal to the hydrostatic pressure when exists a continuous fluid path from the
surface to the subsurface. When there is no continuous
path the pore-pressure, P, either exceeds the hydrostatic head in overpressured formation, or is lower in a deple385
ted or underpressured reservoir.
In vertical holes, the minimum horizontal stress direction (q,) is the only parameter that can be determined
directly from the breakout analysis, as the vertical stress
(6,)is parallel to the borehole axis and has no influence
on the breakout orientation.
In deviated holes, breakouts depend on the three main
stresses - vertical (o,,),
maximum horizontal (q,)and minimum horizontal (oh).
As previously mentioned, stress modifications caused
by drilling can lead to borehole stability problems. It is the
reason why it is important to clearly understand the tectonic forces at work before setting a well plan, assessing
fracture efficiency and formulating injection/production
optimization schemes. This can be achieved through the
stress orientation.
In deviated holes, the ultrasonic images of the borehole wall are providing the opportunity to identify the exact
nature and origin of the borehole damage (stress-induced
or drilling-induced breakouts).
Most of the time (80%) the ovalization is due to drilling
damage.
To determine the nature, geometry and orientation of
other types of damage than breakout, an analysis of the
borehole curvature, based on ultrasonic transit time data,
is performed at each level. This analysis is carried out
using either the HOSANA (Hole Shape ANAlysis) software or an interactive program called HSView (Hole Shape
View) developed by Schlumberger. Figure 9-91 is an
example of hole shape analysis results.
The curvature analysis is performed in three steps:
- at each level and at each azimuth, the radius and
center of curvature are computed in a sliding window
approximately 40" wide;
- if consecutive curvature computations match, the
corresponding angular zones are merged into a single arc
whose center is computed;
- according to the distribution of the arcs and their centers, the borehole damage is automatically identified as
breakouts (Fig. 9-91), shearing at the borehole wall (Fig.
9-92), keyseat (Fig. 9-93) or reaming (Fig. 9-94).
Bcnhcla radiw (In.)
Figure 9-91b- Borehole crosssection showing a breakout. Dots
are data ooints. solid lines reDresent the part of the borehole wall
that have a constant curvature.
The broken lines are the continuation of the largest solid
(courtesv of Schlumberaer).
"
Figure 9-91c - Perspective view.
(courtesy of Schlumberger).
I
~
X430
x460
X431
X432
Figure 9-91a - Example of Hole Shape Analysis results over 2.5
m. Track 3 from left presents the image of the borehole radius, the left
edge being the top of the hole. The value of the radius corresponding
to each color is reproduced in track 2 with the median, the lower 25
percentile and the upper 25 percentile radius values. Tracks 1, 4 and 5
show respectively the well deviation, the orientation of damage relative
to top of borehole, the extent of the feature relative to the hole size
(courtesy of Schlumberger).
386
Borehole cross-section
Figure 9-92 - Example of shear at a
pre-existing fracture and asymmetric
reaming. Observe the sudden enlargement of the hole at the shear
plane (top image). Bottom image
illustrates cross-section of a shear at
a pre-existing fracture
(see caption Figures 9-91 for notation)
(courtesy of Schlumberger).
Serralog 0 2003
Chapter9
f
387
to faulting on a pre-existing plane. The projection onto the
shear plane of the two vectors joining the centers of the
two arcs corresponds to the slip vector (Fig. 9-96).
-
x406
Figure 9-93 - Example of zone of
keyseat as illustrated by the HOSANA program, and cross-section of
the borehole
(courtesy of Schlumberger).
Brehole radius (in.)
Bonhols radius (in.)
Figure 9-96 - On the left :borehole cross-section showing a shear at a
pre-existing fracture. On the right :a perspective view of this shear.
Observe the displacement of the center of the two arcs
(courtesy of Schlumberger).
To avoid confusing this to reaming, a plane crossing
the level must be observed on the image or in a perspective view of this part of the well (cf. Fig. 9-96 right). The
obvious slip shown in Figure 9-96 is partly masked by
subsequent reaming creating a sudden borehole enlargement at the slip plane itself, followed by a smooth decrease in the average hole radius (Figs. 9-92 and 9-97).
Slip occurring
during drilling
Figure 9-94 - Example of zone of
reaming as illustrated by the
HOSANA program and cross-section of the borehole. Observe the
enlargement of the borehole and
the displacement of the centers of
the two arcs (green and red)
(courtesy of Schlumberger).
Preserved
plane
,
-
mdlm(In.)
Breakouts correspond to two undeformed arcs (i.e.
whose radii are close to the radius of the bit) separated by
two damaged zones 180" apart. If the centers of the two
arcs do not coincide this is due to a closing movement in
the direction of the main stress (Fig.9-95).
Figure 9-95 - Other example of
breakout well illustrated by the
borehole cross-section. Observe
the two damaged zones 180"
apart. Observe the displacement of
the center of thr two arcs and the
1.5 cm closure (courtesy of
Schlumberger).
-
.E
1
Reamed
volume
Figure 9-97 - Sketch of a shear at a pre-existing fracture followed by a
reaming. Observe the asymmetric enlargement at the shear plane and
the smooth upwards decrease in the average borehole radius
(courtesy of Schlumberger).
Keyseat can be assumed when a radius of curvature
smaller than the bit size is detected at the bottom of the
borehole (cf. Fig. 9-93).
Reaming can be recognized when two arcs, whose
radii correspond to the bit size, exhibit centers clearly
separated. An enlargement of the well must be observed
at the same depth (cf. Fig. 9-94).
RelationshiD between stress-induced features and
stress axis
Bwehoi8 radius (In.)
Shearing corresponds to two undeformed arcs displaced relative to each other. Such movement corresponds
Serralog 0 2003
The previous described stress-induced features can
be related to the full stress tensor: o,,oHand oh.
387
lnduced tensile fractures
In a vertical well they are usually parallel to the borehole axis and are generated along planes perpendicular
to the minimum stress.
In deviated wells they are en echelon as they are inclined relative to each of the principal stress axes;
Breakouts
A breakout is generated when there is a maximum tangential stress at the borehole wall.
In vertical wells they indicate the b h azimuth.
In deviated wells, the well axis is oblique to the three
main stresses which influence the maximum tangential
stress.. The breakout orientation depends essentially on
the following parameters (Fig. 9-98):
-the direction of the minimum horizontal stress (sh)
- the stress regime -which principal stress (maximum,
intermediate or minimum) is vertical
- the ratio R of the stress magnitude, given by:
Stress determination
The complete state of stress parameters cannot be
determined from a single breakout or fault movement.
Fortunately, the number of features is generally much
greater. If breakouts are observed at several borehole
orientation, the observations can be inverted to give the
main stress parameters. A method for breakout inversion
is proposed by Etchecopar eta/. (1981).
The unknown are o h and Q.
The solution lies on:
- the analysis of breakout orientation data relative to
the well,
- the well orientation.
The method consists to determine the parameters
which give the lowest value for the misfit M (Mmin).This
misfit is defined as the sum of the angular misfit between
actual and "computed" breakouts. Determination of the
minimum misfit value (Mmin)for all the data involves investigating a range of solutions using a plot of q, versus Q
(Fig. 9-99).
The two last parameters can be combined into a single
parameter Q in the range O< = Q < 3 such that:
O < = Q < 1 impliesov>oH>oh,R = Q
1 < = Q < 2 implies oH> (s, > b h , R = 2 - Q
2 < = Q < 3 implies oH> b h > a, R = Q - 2
Figure 9-98 - Theoretical sketch
showing the change in breakout
orientation due to change in intermediate stress magnitude. At the
top the intermediate stress is equal
to the minimum; at the bottom it is
equal to the maximum
(here vertical)
(courtesy of Schlumberger).
The parameters <Th and Q are the same as those constraining the slip direction of microfaults, or the geometry of
nodal planes of focal mechanisms.
Shear movements
During or after drilling some fracture or fault planes
can act as slipping surfaces. Those planes normally lie at
an oblique angle to the current stress axes. If the fracture
or fault plane is reactivated a reltive displacement of Zwo
borehole sections may occur (Fig. 9-96).
The shear-vector orientation depends on the natural
state of stress but not on the well trajectory, as is the case
for the breakout orientation. So, the amount of breakouts
is a function of the well trajectory, but the occurrence of
slippage at pre-existing planes is independent of the well
trajectory.
388
Figure 9-99 - The cross-plot, on the left, shows the mean absolute
deviation to all data points as a function of the stress parameters oh
and Q. Each value of the cross-plot represents a stress state. The
stress state giving the best fit to the data point is marked by a red
square. The digits 1, 2...9 mark the stress state where the mean absolute deviation is 10, 20...90% above the best-fit value. The stress states with no marking have an aboslute deviation less than 10% in the
area around the square and more than 100% in the other area.
The histogram on the right shows the absolute deviation delta of the
individual data points for the best-fit stress state. The mean absolute
deviation is 35" (courtesy of Schlumberger).
The numbers N around the best solution indicate, for
the corresponding tensors, the misfit Mf as a function of
Mmin
1 means the fit Mf is within 10% of the minimum fit
Mmin,etc. If M, < 2 * Mmin,N is not displayed.
Many incorrect data, due to misinterpretation of the
borehole damage, may disturb the interpretation of the
cross-plot. To improve the interpretation, the method miniSerralog 0 2003
Fractures
mizes for the best N%, where N is generally fixed to 80%.
In order to determine the minimum misfit, each possible
tensor is examined as follows:
- for each stress tensor solution, the misfit for each
individual data point is computed
- the individual misfits are sorted in increasing order
- the total misfit Mf is taken to be the sum of the N%
smallest individual misfits
- the best solution is the tensor for which Mf is minimum.
Figure 9-100 reproduces the results of this filtering of
the data using these criteria. This new processing is constrained in orientation (N 150), but less in Q (0.6<Q<0.9).
This Q range corresponds to 0.6<R<0.9 with 6, being
Chapter 9
389
Q
omax.
Figure 9-101 - Stress states compatible with shearing of the borehole
at pre-existing fracture. Th star corresponds to the best-fit stress state
obtained from breakout analysis (courtesy of Schlumberger).
Q
N
Figure 9-100 - The cross-plot, on the left, shows the trimmed (best
80%) mean absolute deviation as a function of the stress parameters
oh and Q. Each value of the cross-plot represents a stress state. The
stress state giving the best fit to the data point is marked by a red
square. The digits I, 2...9 mark the stress state where the mean absolute deviation is 10, 20...90% above the best-fit value. The fit is computed for the best 80% of points at each stress state. Different points
may be rejected at different stress states. A much precise best-fit
stress state is obtained. The histogram on the right is also narrower if
we compare to Figure 9-99. The trimmed mean absolute deviation is
17” for the best-fit stress state (courtesy of Schlumberger).
oHis much closer to ovthan to oh. Using the misfit histogram associated to this new processing, it is possible to
locate zones where the orientation of the stress varies.
These can often be confirmed by changes in the orientation of induced fractures.
Interpreting the last solution, two type of microstructures support it:
- tensile fractures oriented N 150 as the maximum
horizontal stress determined,
- single slip movement in which it has been possible to
measure a pre-existing fracture.
Figure 9-101 shows that the stress is compatible with
both types.
Using stress information, stress components tangential to the borehole wall can be computed for all well orientation (Fig. 9-102).
Serralog 0 2003
Figure 9-102- Schmidt diagram showing the tangential stress state at
the borehole wall for different well deviations and azimuths. The center
of this diagram corresponds to a vertical well. A point on the outer circle corresponds to a horizontal well. The longuest length of the crosses indicates the maximum and the shortest the minimum tangential
stresses and their orientation. The numbers correspond to the ratio
between the maximum tangential stress and the vertical one (the ratio
a/av is assumed to be 0.6 and the fluid pressure is not taken into
account). The squares and circles indicate respectively the well orientation minimizing or maximizing the tangential stress at fhe borehole
wall. When the maximum tangential stress reaches its minimum value
it is also identical all around the borehole wall. This also indicates the
direction where drilling is safer (courtesy of Schlumberger).
This information of this diagram allows the determination of the well direction in order to limit the breakout
extension, and reduce the risks of hydraulic fracturing and
slip along pre-existing fractures. In addition, the mud
weight can be reduced as well, and also the drilling cost.
389
Prediction of the fracture efficiency
RECAP
As previously seen, the knowledge of the compressional and shear slownesses, and of the bulk and grain densities allows the determination of the mechanical rock properties. Through these parameters it is possible to calculate the vertical fracture migration. At each level up and
down the perforated interval, the additional pressure
required above extension pressure is calculated. With
delta pressure plotted versus depth, the relative breakdown pressures of the different zones may be determined
(Fig. 9-103).
In a certain extent, the pore-fluid pressure conteracts
the stresses imposed on the rock. Consequently, the rock
sustains an effective stress, 0,which is given by:
OV,H
= sV,H -
(9-39)
pp
Knowing this in-situ stress in and around the hydrocarbon bearing zones, the fluid density and volume can
be determined to predict the fracture migration behavior.
Then, The total fracture height for a given injection pressure can be calculated (Fig. 9-103).
Depth
Qualitative frac geometry
a 2 D representation
Before and after
frat gamma ray
One can conclude that the image tools are the best
technique for detection and a complete analysis of the
fractures.
In the absence of these measurements, the existence
of fractures can be concluded if several of the following
phenomena are observed simultaneously at about the
same depth:
- a change in temperature gradient;
- a change in hole diameter;
- a localised decrease in density, accompanied by a
variation in Ap while Pe, At and @N, remain steady, but not
if there is a cave, or the mud contains baryte;
- a very slight increase in porosity;
- secondary porosity;
- a reduction in the value of the rn factor;
- a change in the ratio LLd/LLs;
- sudden drops in resistivity on the microdevices;
- high Pe values when the mud contains baryte;
- conductivity peaks on the FIL;
- DCA showing conductive anomalies;
- a pause in tool rotation;
- strong attenuation of acoustic waves;
- a blurred zone on the VDL, or a lack of vertical coherence on the wave train;
- radioactivity peaks or uranium peaks;
- strong negative SP deflections.
8100
Figure 9-103 - Example of Fracture prediction log. In this example
each color represents the fracture height in 200 psi increments of frac
pressure. The maximum frac pressure must not exceed 800 psi. The
upper zone is weaker than the lower zone and will have a tendency to
take more of the treatment. Therefore, fluid entry must be limited to
this zone by increasing perforation density in the lower zone. The frac
job is crontrolled by the recording of the gamma ray before and after
the frac with radioactive tracers. One can observe that the predicted
frac height is within two feet of the actual height
(courtesy of Schlumberger).
390
Serralog 0 2003
References and Bibliography
ADAMS, J., BOURKE, L., & BUCK, S. (1990). Integrating Formation Microscanner images and cores.
Oilfield Review, 2, 1, p. 52-65.
AGUILERA, R. (1976). - Analysis of naturally fractured
reservoirs from conventional well logs. Journ. Petroleum
Technol., 28, July, p.764-772.
AGUILERA, R., & KIVAU POOLLEN, M., (1977). Current status on the study of naturally fractured reservoirs. The Log Analyst, 18, 3, May-June.
AGUILERA, R. (1980). - Naturally Fractured
Reservoirs. Petroleum Publishing Co., Tulsa, Oklahoma.
ANDERSON, T., & WALKER, T. (1972). - Log derived
rock properties for use in well stimulation design. SPE of
AIME, SPE paper 4095.
ATKINSON, A. (1977). - Fracture pressure gradients
from acoustic and density logs : an updated approach.
SPWLA, 18th Ann. Log. Symp. Trans.
BABCOCK, E.A. (1978). - Measurements of subsurface fractures from dipmeter logs. Bull. Amer. Assoc.
Petroleum Geol, 62, p. 1119-1126.
BARNES, R.P., DEE, S.J., & SANDERSON, D.J.
(2002). - Characterization of Fracture Networks Using
Data from Boreholes: An Example from the Concealed
Borrowdale Volcanic Group in West Cumbria, U.K. In:
LOVELL, M., & PARKINSON, N. (ed.) (2002). - Geological
Applications of Well Logs. AAPG Methods in Exploration,
604 p.
BARTON, C.A., & ZOBACK, M.D. (2002). - Wellbore
Imaging
Technologies
Applied
to
Reservoir
Geomechanics and Environmental Engineering. In:
LOVELL, M., & PARKINSON, N. (ed.) (2002). - Geological
Applications of Well Logs. AAPG Methods in Exploration,
604 p.
BATES, R.L., & JACKSON, J.A. (1980). - Glossary of
Geology. 2nd ed. Amer. Geol. Inst., Falls Church, Virginia.
BECK, J., SCHULTZ, A., & FITZGERALD, D. (1977). Reservoir evaluation of fractured cretaceous carbonates
in South Texas. SPWLA, 18th Ann. Log. Symp. Trans.
BELL, J.S., & GOUGH, D.I. (1979). - Northeast-southwest compressive stress in Alberta : evidence from oil
wells. Earth Planet. Sci. Let., 45, p. 475-482.
BEYDOUN, W.B., CHENG, C.H., & TOKSOZ, M.N.
(1983). - Detection of subsurface fractures and permeable
zones by the analysis of tube waves. 8th Europ. Form.
Eval. Trans., London, paper I.
BIGELOW, E.L. (1985). - Making more intelligent use
of log derived dip information. 5 Parts. The Log Analyst,
26, 1, 2, 3, 4, and 5.
BILLINGS, M.P. (1942). - Structural Geology. 1st ed.
Prentice-Hall Inc. Englewood Cliffs, New Jersey.
BILLINGS, M.P. (1972). - Structural Geology. 3rd ed.
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
BOND, L.O., ALGER, R.P., & SCHMIDT, A.W. (1971).
- Well log Applications in Coal Mining and Rock
Mechanics. Trans. SME 250.
BOURKE, L., DELFINER, P., FETT, T.H., GRACE,
L.M., LUTHI, S., SERRA, O., STANDEN, E., &
Serralog 0 2003
TROUILLER, J.-C. (1989). - Using Formation
Microscanner images. The Technical Review, 37, 1, p.
16-40.
BOYELDIEU, C., & MARTIN, C. (1984). - Fracture
detection and evaluation. SAID-SPWLA, 9th Europ. Form.
Eval. Trans., Paris, paper 21.
BOYELDIEU, C., &WINCHESTER, A. (1982). - Use of
the Dual Laterolog for the Evaluation of the Fracture
Porosity in Hard Carbonate Formations. Offshore South
East Asia 1982 Conference, 9-12 Feb., Singapore.
BRIE, A., HSU, K., & ECKERSLEY, C. (1988). - Using
the Stoneley normalized differential energies for fractured
reservoir evaluation. SPWLA, 29th Ann. Log. Symp.
Trans., paper XX.
BUTLER, M., PHELAN, M.J., & WIGHT, A.W.R.
(1976). - Buchan field :evaluation of a fractured sandstone reservoir. SPWLA, 4th Europ. Symp. Trans., London,
paper A.
BURKE, J.A., CAMPBELL, Jr. R.L, & SCHMIDT, A.W.
(1969). - The Litho-Porosity Cross Plot. SPWLA, 10thAnn.
Log. Symp. Trans., paper Y.
CHEUNG, Ph. (1984). - Fracture detection using the
sonic tool .SAID-SPWLA, 9th Europ. Intern. Form. Eval.
Trans., Paris, paper 42.
CLAVIER, C., & RUST, D.H. (1976). - MID-PLOT a
new Lithology Technique. The Log Analyst, 17, 6.
COATES, G.R., & DUMANOIR, J.L. (1974). - A new
approach to log-derived permeability. SPWLA, 14th Ann.
Log. Symp. Trans.
COATES, G.R., & DENOO, S.A. (1980). - Log Derived
Mechanical Properties and Rock Stress. SPWLA, 21s'
Ann. Log. Symp. Trans.
COATES, G.R., & DENOO, S.A. (1981). -Mechanical
properties programs using borehole analysis and Mohr's
circle. SPWLA, 22nd Ann. Log. Symp. Trans., paper DD.
COX, J.W. (1970). - The high resolution dipmeter
reveals dip-related borehole and formation characteristics. SPWLA, 11th Ann. Log. Symp. Trans.
COX, J.W. (1983). - Long axis orientation in elongated
boreholes and its correlation with rock stress data.
SPWLA, 24th Ann. Log. Symp. Trans., paper J.
CRARY, S. et a/. (1987) - Fracture Detection with Logs.
The Technical Review, 35, 1.
CURRIE, J.B., & NWACHUKWU, S.O. (1974). Evidence on incipient fracture porosity in reservoir rocks
at depth. Bull. Canad. Petrol. Geol., 22, 1.
DEERE, D.U., & MILLER, R.P. (1969). - Engineering
Classification and lndex Properties for lntact Rock. US.
Air Force Systems Command Air Force Weapons Lab.,
Kirtland Air Force Base, New Mexico, Tech. Rep.
DELFINER, P., PEYRET, O., & SERRA, 0. (1984). Automatic determination of Lithology from Well Logs. 59th
Ann. Techn. Conf. SPE of AIME, Houston, Texas; paper
SPE 13290.
DELHOMME, J.P., PILENKO, T., CHERUVIER, E., &
CULL, R. (1988). - Reservoir applications of dipmeter
logs. Journ. Petroleum Technol., 40, 2, p. 180-186.
DELHOMME, J.P., and MOTET, D. (1993). -Reservoir
description and characterization from quantitative boreho391
le
image
analysis.
Schlumberger
Reservoir
Characterization. Africa and Mediterranean Region, no 5.
DENNIS, B., STANDEN, E., GEORGI, D.T., &
CALOW, G.O. (1987). - ldentification and productivity predictions in a carbonate reef complex. 9th Europ. Form.
Eval. Trans., Paris, paper 21.
DICKEY, P.A. (1979). - Petroleum Development
Geology. Petroleum Publishing Co., Tulsa.
DONATH, F.A. (1970). - Some information squeezed
out of rocks. Amer. Sci., 58, p. 54-72.
DRUMMOND, J.M. (1964). - An appraisal of fracture
porosity. Bull. Canad. Petrol. Geol., 12 , 2.
EDWARDS, D. (1985). - Mechanical Properties
Evaluation and its applications. Schlumberger Technical
Services Inc., Europe Unit.
EKSTROM, M.P., CHEN, M.Y., ROSSI, D.J., LOCKE,
S., & ARON, J. (1986). -High Resolution Microelectrical
Borehole Wall Imaging. SPWLA, 10th Europ. Symp.
Trans., Aberdeen.
EKSTROM, M.P., DAHAN, C.A., CHEN; M.Y., LLOYD,
P.M., & ROSSI, D.J. (1986). - Formation Imaging with
Micro-Electrical Scanning Arrays. SPWLA, 27th Ann. Log.
Symp. Trans., Houston.
FAIVRE, 0. (1993). - Fracture evaluation from quantitative azimuthal resistivities. SPE of AIME, 68th Ann.
Techn. Conf. & Exhib., SPE paper 26434.
FAIVRE, O., & SIBBIT, A. (1984). - Dual Laterolog
Response in Fractures.
FOCKE, J.W., & MUNN, D. (1987). - Cementation
exponents in Middle Eastern carbonate reservoirs. SPE
Formation Evaluation, p. 155-167.
FONS, L. Sr. (1969). - Geological application of well
logs. SPWLA, 10th Ann. Log. Symp. Trans.
FRISINGER, M.R., & GYLLENSTEN, A. (1986). Fracture detection in North Sea reservoirs. SPWLA, 10th
Europ. Form. Evaluation Symp.
GARY, M., McAFEE, R.Jr., & WOLF, C.L. (1972). Glossary of Geology. Amer. Geol. Institute, Washington,
D.C.
GAUER, P., RUHLAND, M., BEAUCOURT, F. de, &
JANOT, P. (1984). - Ovalisation des trous de forage evaluee a partir des pendagemetres 4-patins; caracteres
geologiques de ces variations; leur utilisation pour la
determination de /'orientation des contraintes horizontales
actuelles dans le Fosse Rhenan. SAID-SPWLA, 9th
Europ. Intern. Form. Eval. Trans., Paris, paper 31.
GEORGI, D.T. (1985). - Geometrical aspects of
BoreHole Televiewer images. SPWLA, 26th Ann. Log.
Symp. Trans., paper 0.
GEORGI, D.T. (1986). - Guides for the interpretation of
dipmeter fracture logs. SPWLA, 27th Ann. Log. Symp.
Trans., paper TT.
GOGUEL, J. (1952). - Traite de Tectonique. Masson,
Paris.
GRIFFITH, A.A. (1920). - The phenomena of rupture
and flow in solids. Phil. Trans. Roy. SOC.London, ser. A
221, p. 163-197.
GRIFFITH, A.A. (1924). - The theory of rupture. Proc.
(1st) Intern. Congr. Appl. Mech., p. 55-63.
392
GRIGGS, D.T. (1936). - Deformation of rocks under
high confining pressure. J. Geol., 44, p. 541-577.
GRIGGS, D.T. (1939). - Creep in rocks. J. Geol., 47, p.
225-251.
GRIGGS, D.T. et al. (1951). - Deformation of Yule
Marble: part lV, effects at 150°C. Geol. SOC.Amer. Bull.,
62, p. 1385-1406.
GRIGGS, D.T., & HANDIN, J. (eds) (1960). - Rock
deformation. Geol. SOC.Amer. Mem., 79 , 382 p.
GRIGGS, D.T., TURNER, F.J., & HEARD, N.C. (1960).
- Deformation of rocks at 500"to 800".Geol. SOC.Amer.
Mem., 79, p. 39-104.
HATCHER, R.D. Jr. (1990). - Structural Geology.
Principles, Concepts, and Problems. Merril Publishing
Co., Columbus.
HEPP, V., & DUMESTRE, A.C. (1975). - CLUSTER -A
method for selecting the most probable dip results from
dipmeter survey. SPE of AIME 50th Ann. Fall Mtg., Dallas,
SPE
paper
5543.
HOBBS, B.E., MEANS, W.D., & WILLIAMS, P.F.
(1976). - An outline of Structural Geology. John Wiley &
Sons, New York.
HORNBY, B.E., JOHNSON, D.L., WINKLER, K.W., &
PLUMB, R.A. (1989). - Fracture evaluation using reflected
Stoneley wave arrivals. Geophysics, 54, 10.
HORNBY, B.E., LUTHI, S.M., & PLUMB, R.A. (1990).Comparison of fracture apertures computed from electrical borehole scans and reflected Stoneley waves: an integrated interpretation. SPWLA, 31s' Ann. Log. Symp.
Trans., paper L.
HUANG, C.F., HUNTER, J.A. (1981). - The correlation
of "tube wave" events with open fractures in fluid-filled
boreholes: current research, Part A. Geol. Survey of
Canada, paper 81-1A, p. 361-376.
INGRAM, D.S., ANDERSON, R.A., & ZANIER, A.M.
(1972). - Fracture pressure Gradient Determination from
well Logs (SPE paper 4135).
KLEINBERG, R.L., CHOW, E.Y., PLONA, T.J.,
ORTON, M., & CANADY, W.J. (1982). - Sensitivity and
reliability of Fracture Detection Techniques for Borehole
Application. SPE of AIME, 57th Ann. Fall Mtg., New
Orleans, SPE paper 11036.
KOERPERICH, E.A. (1977). - lnvestigation of acoustic
boundary waves and interference patterns as techniques
for detecting fractures. SPE of AIME, SPE paper 6820.
LEET, L.Don, JUDSON, S., & KAUFFMAN, M.E.
(1978). - Physical Geology. 5th ed. Prentice-HOY Inc.,
Englewood Cliffs, New Jersey.
LESLIE, H.D., & MONS, F. (1982). - Sonic Waveform
Analysis: Applications. SPWLA, 23rd Ann. Log. Symp.
Trans., paper GG.
LINK, P.K. (1982). - Basic Petroleum Geology. OGCl
Publications, Tulsa.
LIU, O.Y. (1985). - Fracture evaluation using borehole
sonic velocity measurements. SPE of AIME, 60th Ann.
Techn. Conf. & Exhib., SPE paper 14399.
LLOYD, P.M. (1986). - The Formation Microscanner :
A New Generation High Resolution Logging Technique.
LLOYD, P.M., DAHAN, C.A., & HUTIN, R. (1986). Serralog 0 2003
Fractures
Formation Imaging from Micro-Electrical Scanning Arrays
: A New Generation of Stratigraphic High Resolution
Dipmeter Logging Tool. SPWLA, 10th Europ. Syrnp.
Trans., Aberdeen.
LOVELL, M., & PARKINSON, N. (ed.) (2002). Geological Applications of Well Logs. AAPG Methods in
Exploration, 604 p.
LUTHI, S., & SOUHAITE, P. (1990). - Fracture apertures from electrical borehole scans. Geophysics, 55, 7, p.
821-833.
MATHIEU, F., & TOKSOZ, M.N. (1984). Determination of fracture permeability using acoustic
logs. SAID-SPWLA, 9th Europ. Intern. Form. Eval. Trans.,
Paris, paper 47.
MAURY, V., & ETCHECOPAR,A. (1992). - Shear induced by oil and gas wells drilling and production along
faults and discontinuities. 29th Intern. Geol. Congress,
Kyoto, Japan.
MAXWELL- GARNETT, J.C. (1904). - Transactions of
the Royal Society, 203 , p. 385, London.
MAYER, C., & SIBBIT, A. (1980). - GLOBAL, a new
Approach to Computer-processed Log Interpretation.
SPE of AIME, 55th ann. Fall Mtg., Dallas, SPE 9341.
MINNE, J.C., & GARTNER, J. (1979). - Fracture
Detection in the Middle East. Middle East Oil Techn. Conf
of SPE, Bahrain, 25-29 Mar.
MORRIS, R.L., GRINE, D.R., & A R K F E L ~ ,T.E.
(1963). - The use of compressional and shear acoustic
amplitudes for the location of fractures. SPE of AIME, 38th
ann. fall Mtg., New Orleans, Louisiana, SPE paper 723.
Also in J. Petrol. Techn., June 1964, pp. 623-635.
NOREL, G. (1984). - Detection des fractures et des
pendages par ultra-sons. SAID-SPWLA, 9th Europ. Form.
Eval. Trans., Paris, paper 16.
NURMI, R., CHARARA, M., WATERHOUSE, M., &
PARK, R. (1990). - Heterogeneities in carbonate reservoirs: detection and analysis using borehole electrical
images. In: Hurst, A, Lovell, M.A., & Morton, A.C. (eds) :
Geological Applications of Wireline Logs, Geological
Society, special publication, 48, p. 95-111, London.
PAILLET, F.L. (1985). - Problems in fractured-reservoir
evaluation and possible routes to their solutions. The Log
Analyst, 24, 1, p.26-41.
PERRIER, P., & QUIBLIER, L. (1973). - Orientations
des fractures observees sur carottes a I’aide de la pendagemetrie. Revue IFP, XXVIII, 6.
PEZARD, P.A., & LUTHI, S. (1988). - Borehole electrical images in the basement of the Cajon Pass scientific
drillhole, California: fracture identification and tectonic
implications. Geophysical Research Letters 15 (August
supplement), p. 1017-1020.
PICKETT, G.R. (1963). - Acoustic Character Logs and
their Applications in Formation Evaluation. J . Per.
Technol., 15, 6.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Houston.
PRESS, F., & SIEVER, R. (1982). - Earth. 3rd ed. W H.
Freeman & Co., San Francisco.
RAGAN, D.M. (1973). - Structural Geology. John Wiley
Serralog 0 2003
I
Chapter9
I 393
& Sons, New York.
RAMBOW, F.H.K. (1984). - The BoreHole Televiewer:
some field examples. SPWLA, 25th Ann. Log. Symp.
Trans., paper C.
RAMSAY, J.G. (1967). - Folding and Fracturing of
Rocks. McGraw-Hill Book Co., Inc., New York.
RAMSAY, J.G., & HUBER, M.I. (1987). - Modern structural Geology vol 2 : Folds and Fractures. Academic
Press, New York.
RASMUS, J.C. (1982). - Determining the type of fluid
contained in the fractures of the Twin Creek Limestone by
using the Dual Laterolog tool. J. Petrol. Techn., Nov.
RASMUS, J.C. (1983). - A variable cementation exponent, m, for fractured carbonates. The Log Analyst, 24, 6,
p. 13-23.
RIDER, M.H. (1978). - Dipmeter Log Analysis - An
Assay. SPWLA, 19thAnn. Log. Syrnp. Trans., paper G.
RODGERS, S.F., & EVANS, C.J. (2002). - Stressdependent Flow in Fractured Rocks at Sellafield, United
Kingdom. In: LOVELL, M., & PARKINSON, N. (ed.)
(2002). - Geological Applications of Well Logs. AAPG
Methods in Exploration, 604 p.
RUSSELL, W.L. (1951). - Principles of Petroleum
Geology. McGraw-Hill Book Co., New York.
Schlurnberger Ltd (1970). - Fundamentals of Dipmeter
Interpretation.
Schlurnberger (1979). - Well Evaluation Conference.
Algeria.
Schlurnberger Ltd (1981). - Dipmeter Interpretation.
Volume 1 -Fundamentals.
Schlurnberger Middle East S.A. (1981). - Well
Evaluation Conference. United Arab Emirates/ Qatar.
Schlurnberger Middle East S.A. (1984). - Well
Evaluation Conference. Egypt.
Schlurnberger & Sonatrach (1995). - Well Evaluation
Conference, Algeria .
Schlurnberger and Agip (2000). - Italy 2000 - Valueadded reservoir characterization.
SERRA, 0. (1989). - Formation Microscanner Image
Interpretation. Schlurnberger, SMP-7028.
Services Techniques Schlurnberger (1970). - Well
Evaluation Conference. Libya.
SHANKS, R.T., KWON, B.S., VRIES, M.R. de, &
WICHMAN, P.A. (1976). - A review of fracture detection
with well logs. SPE of AIME, paper SPE 6159.
SIBBIT, A.M., & FAIVRE, 0. (1985). - The Dual
Laterolog Response in Fractured Rocks. SPWLA, 26th
Ann. Log. Syrnp. Trans., paper T.
SMEKHOV, E.M. (1969). - Etude de la fracturation des
roches et des reservoirs fractures d’hydrocarbures.
Travaux du VNIGRI. Traduction en franpais d’ELF R.E.
SNYDER, R.H., & CRAFT, M. (1977). - Evaluation of
Austin and Buda formations from core and fracture analysis. Gulf Coast Assoc. Geol. SOC.,27.
SUAU, J., BOYELDIEU, C., ROCCABIANCA, R.,
CIGNI, M., & SPILA, M. (1978). - Evaluation of very low
porosity Carbonates (Malossa, Italy). SPWLA., 19th Ann.
Log. Syrnp. Trans.
SUAU, J., & GARTNER, J. (1979). -Fracture detection
393
394
Well Logging and Geology
from the logs. SPWLA, 6th Europ. Symp. Trans., paper L.
SUPPE, J. (1985). - Principles of Structural Geology.
Prentice-Hall Inc., Englewood Cliffs, New Jersey.
THEYS, P., LUTHI, S., & SERRA, 0. (1983). - Use of
dipmeter in Carbonates for detailed sedimentology and
reservoir engineering studies.
TIMER, M.P., LOVELESS, G.W., 81ANDERSON, R.A.
(1973). - Estimation of Formation Strength from the
Mechanical Properties Log. 48th Ann. Meeting of SPE of
AIME, paper SPE 4532
VINCENT, P., GARTNER, J., & ATTALI, G. (1979). GEODIP - An approach to detailed dip determination
using correlation by pattern recognition. J. Petroleum
Technol., Feb., p. 232-240.
WALDSCHMIDT, W.A., FITZGERALD, P.E., & LUNSFORD, C.L. (1956). - Classification of porosity and fractures in reservoirrocks. Bull. Amer. assoc. Petroleum Geol.,
40, 5.
WALKER, T., & KESSLER, C. - Defection of natural
fractures with well logs. Welex, L 38.
WATFA, M. (1987). - Current affairs in saturation.
Schlurnberger Middle East Well Evaluation Review, no 2.
WATFA, M. (1987). Seeking the saturation solution.
Schlumberger Middle East Well Evaluation Review, no 3.
ZEMANEK, J. et al. (1969). - The Borehole Televiewer
: A New Logging Concept for Fracture Location and other
types of Borehole Inspection. SPWLA, 9th Ann. Log.
Syrnp. Trans
-
394
Serralog 0 2003
WELL LOGGING AND SOURCE ROCK
Introduction
The interest of a sedimentary basin for petroleum
exploration requires the conjunction of four characteristic
properties:
- oil or gas potential of the sedimentary basin because
if the basin has no suitable source rock there can be no
hydrocarbon accumulation and consequently no interest
to pursue the exploration of this basin (Fig. 10-1);
- presence of reservoir rocks capable of accumulating
hydrocarbon during their migration from source rocks;
- existence of traps in which hydrocarbon can accumulate;
- existence of seal rocks avoiding any later dismigration of hydrocarbon.
The geological description of reservoir rocks and of
their transformation by diagenesis, and their deformation
under tectonic stresses has been done in the previous
chapters
The estimation of the hydrocarbon potential of a sedimentary basin is now analyzed.
This analysis starts by the location of source rocks and
the evaluation of their geochemical characteristics - organic content and potential to generate oil or gas.
Figure 10-1 - Potential source rocks depend on the organic carbon
content and their potential to produce hydrocarbon
(from Perrodon, 1980) .
Traditional geological approach
The source rock evaluation of a sedimentary basin is
carried out under laboratory conditions by direct geochemical analysis of rock samples collected on outcrops or in
wells in shaly intervals (Fig. 10-2). Unfortunately, in wells
the samples are rare as generally shales are not cored.
Sidewall cores give representative results but are discrete points of data. Cuttings collected in zones appearing
shaly may be contaminated by material falling from layers
further up the well or by the mud if the latter is oil-based.
The original organic matter - humic, sapropelic or
planktonic - present within a sediment evolves as a function of the conditions of oxydo-reduction into the depositional environment. Oxygen transforms the organic material into C02. The reducing conditions that exist fundamentally in these environments - swamps, lagoons, lakes,
fjords, deep seas - preserve it from this transformation in
Serralog 0 2003
J
Figure 10-2 - Comparison of the chromatograms of source rocks with
the chromatograms of oil samples. The two upper source rocks of
well 2 have not generated the oils produced in well 2. The lower source rock corresponds to the lower oil.
395
C02. The bacteria living in anaerobic conditions act on
the organic matter. Under their action it is transformed
and can become a source rock which can generate biogenic methane, oil, bitumen or thermal gas as a function
of its maturation under temperature and pressure influence.
Analysis of the organic matter and comparison with oil
collected in a well allows the determination of theorigin of
the source rock
Several authors, Beers & Goodman (1944) (Fig. 10-4),
Russel (1945), Swanson (1960) (Fig 1 0 4 , Hassan
(1973), Supernaw et a/. (1978) (Fig. 10-6), have observed
a strong correlation between uranium and organic matter.
Well logging approach
The evaluation of the source rock potential can be
achieved through well log analysis. Different approaches
will be studied.
The Carbolog method
This technique, proposed by Carpentier et a/. ( I 989),
is a quick, quantitative and qualitative method that uses
measurements which have been recorded in old wells. It
requires a calibration for each sedimentary basin. There
are also some difficulties in defining the 100% point for
organic material.
The uranium concentration
Figure 10-4 - Relation between
uranium and organic carbon in
sedimentary rocks
(from Beers 8, Goodman, 1944).
So, after calibration with core data, it is possible to
evaluate the organic carbon content of a source rock from
its uranium content and from that its hydrocarbon potential if the nature of the organic material is known (Fig 10-
7).
The reducing conditions which are favorable for the
transformation of organic matter are also favorable to the
preservation of uranium which will precipitate or accumulate into sediments under certain conditions. In presence
of organic matter uranium is irreversibly adsorbed from
uranyl solutions in the presence of bacteria and humic
fractions. Organic matter and acidic pH are favorable to
convert U022+ into the insoluble quadrivalent ion U02.
The same reducing conditions generate the transformation of sulfates in sulfides (pyrite for example). In acidic
pH environments, humic and fulvic acids, ethers, alcohols, aldehydes, favour the precipitation of uranium by the
reduction of US+to U4+,forming urano-organic complexes
or chelates. Clays also encourage the formation of schoepite by hydrolysis of the uranyl ions (Fig. 10-3).
Figure 10-5 - On the left: diagram showing possible relation of uranium
content to total organic matter as controlled by the proportion of humic
and sapropelic material making up the organic matter. On the right: oil
yield of a marine black shale as a function of the total organic matter
(from Swanson, 1960).
Consequently, the uranium measured through the
spectrometry of the natural radioactivity is a good indicator of the source rock potential (Fig.10-8).
The A log R method
J
Figure 10-3 - Diagram showing possible association and time of
emplacement of uranium with common constituents of marine black
shales. Uranium is represented by black squares
(from Swanson, 1960).
396
This method was introduced by Exxon geologists. It is
based on a A log R which is obtained by superimposing
sonic and resistivity logs on the correct scales and using
the value of the level of organic metamorphism units
(LOM).
A log R represents the area of separation between the
sonic transit time and the resistivity curves. The sonic
curve is superimposed on the resistivity using a logarithmic scale for resistivity and a linear scale for sonic (Fig.
Serralog 0 2003
Source rock
10-8). A base line is determined for resistivity and sonicby
overlaying the two curves in a non-source rock interval.
A log R is given by the following relation:
1 Chapter 10
The LOM values vary as follows:
- 0 to 7.5 : immature
- 7.5 to 11.5 : oil phase
- > 11.5 : gas phase.
NATURAL GAMMA RAY SPECTROSCOPYDATA
Figure 10-8 Example of formation very rich in uranium as shown by
the natural gamma
ray spectrometry
data. It may correspond to a potential source rock
(courtesy of
Schlumberger).
0
0
1
2
3
4
5
6
Organic carbon (Oh)
Figure 10-6 - Relation between uranium (expressed as a
uranium/potassium ratio) and organic carbon
(from Supernaw et al., 1978).
To determine the Total Organic Carbon (TOC) from the
A log R, the level of maturation (LOM) of the organic
material, must be determined. The LOM values vary from
0 to 12. It is evaluated from the pyrolysis maturation parameters (TmaX)by optical methods: determination of the
Vitrinite Reflectance (VR or PRV, Table 10-I), and
Thermal Alteration Index (TAI), and by the nuclear fission
traces or the estimation of the paleotemperatures or the
subsurface temperatures (Fig. 10-10).
a
Organic matter, percent
30
Humic, percent
15
Sapropelic. percent
85
Uranium, percent 0.0030
Oil, gallons per ton
26.6
6
25
25
75
0.0041
20.3
Knowing A log R and LO the TOC value can i e
determined from the chart (Fig. 10-11). If TOC measurements have been carried out from core samples in laboratory, the values can be plotted to confirm the exact
value of LOM.
A second parameter is required. It correspond to the
hydrocarbon generating potential S2. This parameter is
deduced from the TOC and LOM values as illustrated by
the chart of Figure 10-12.
C1
C2
c3
c4
20
50
17
50
50
15
10
50
50
0.0031
6.2
50
0.0063
12.5
10
50
50
0.0048
0.0054
10.6
9.4
0
I l I l I
d
15
75
25
0.0070
6.6
I
f
5
95
5
0.0051 0.0031
3.6
1.4
20 Miles
10
I
e
10
85
15
I
J
Figure 10-7 - Sketch showing theoretical distribution of humic and sapropelic materials in a shallow sea in which black muds are accumulating, and
the estimated uranium content and oil yield of the resulting black shale. Increase in total organic matter seaward due chiefly to seaward decrease
in amount of detrital sediment; proportional increase in sapropelic matter seaward due to decrease of humic land-plant debris and predominance of
planktonic matter (from Swanson, 1960).
Serralog 0 2003
397
DJOUI
W.1
Figure 10-11- Chart to allow the
determination of TOC from A log
R separation and LOM
(from Malla & Baci, 1995).
Figure 10-12 - Determination of S2
from TOC and LOM for type I1
(oil-prone) kerogen
(from Malla & Baci, 1995).
Methods based on cross-plots
Figure 10-9 - Two example Of A log R, TOC and S2curves determined
on 2 wells of Algeria (from Malla & Baci, 1995).
The validity
Of
this approach is controlled by plotting
the
deduced from logs
sured on cores (Fig. 10-13).
the
values
mea-
Table 10-1
Vitrinite reflectance (PRV) values for hydrocarbon
and coal
These methods are based on the theoretical influence
of organic matter on the log responses.
Organic matter is considered to be fundamentally nonconductive. Therefore, increases in resistivity may indicate the presence of organic matter in the shaley beds.
Organic matter has a specific density in the range of
0.95 to 1.O5g/cm3. This is comparable to the density of oil
or water. If one considers that organic matter replaces
part of the matrix, reduction in bulk density can be expected. Unfortunately, at the same moment influence of pyrite, often present in reducing conditions favorable to the
preservation of organic matter, will increase the apparent
bulk density. One has also to remember that shales generally break and cave and, consequently, the density measurement will be affected.
Wl1
Figure 10-13 - Comparison of the TOC measured on cores and TOC
log, at the same depth for the two wells (from Malla & Baci, 1995).
Oil phase
.Gas phase
Figure 10-10- The value of LOM, determined from measured PRV
equivalents for well DJD-1, was calculated at 11.5
(from Malla & Baci, 1995).
398
The average sonic transit time of organic matter is
assumed (Mendelson & Toksoz, 1985) to be close to 180
@ft, so comparable to that of water.
The neutron response of organic matter is estimated to
average 67 p.u. (Mendelson & Toksoz) whereas that of
clay minerals composing essentially shales is generally
lower than 50 p.u. Consequently, one can expect an
increase of neutron index in organic-rich rocks.
As previously seen one can expect an increase of
gamma ray response and uranium content in organic rich
-
Serralog 0 2003
Source rock
rocks. The proportion of uranium showed the best relationship with TOC (Mann et a/., 1986) confirming the
works mentionned previously.
Autric & Dumesnil (1985) suggested the use of sonic
or gamma ray versus resistivity, these measurements
being less affected by borehole rugosity (Fig. 10-14).
I Chapter 10 I 399
Mendelson & Toksoz (1985) modelled expected log
responses in source rocks assuming a simple formation
model of matrix, organic matter and porosity and comparing the calculated total organic carbon (TOC) value with
the measurements made on cores. They used multiple
regression analysis to define the relationship between
TOC and the various responses.
The methods based on cross-plots allow a qualitative
detection of possible source rocks but they do not allow
the precise evaluation of the organic carbon concentration.
Method based on nuclear measurements
Rl
Figure 10-14 - Characterization of source rocks by cross-plots combining sonic transit time and resistivity (left diagram), or gamma ray and
resistivity (right diagram) (from Autric & Dumesnil, 1985).
Meyer & Nederlof (1984) suggested same type of
cross-plots but they corrected resistivities to standard
temperature (Fig. 10-15).
Source Rocks
Other attemps to evaluate directly the source-rock
potential are based on the measurement of the carbonoxygen ratio (C/O) by spectrometry of the gamma rays
induced by inelastic collisions of fast neutrons (Figs. 1016 and 10-17). This measurement requires either a stationary run or several runs with very low recording speed.
This method was proposed by S. Herron (1986).
Knowing the C/O ratio, the determination of the organic
carbon content needs two determinations:
- a measurement of the oxygen concentration in rock.
This evaluation is relatively easy if the lithology is well
known. In any case the average percentage of oxygen in
the earth’s crust is 46% in weigth. Knowing the ratio and
the 0 content it is easy to compute the total C content;
- a correction for inorganic carbon influence on the
measurement. The later is possible if the lithology is
known and especially the percentage of calcite and dolomite.
Inelastic
collision of
neutrons
Source Rocks
Figure 70-16 - Example of interactions between high energy neutrons
and C and 0 atoms (courtesy of Schlumberger).
Resistivity at 75°F (omh-m)
Figure 10-15 - Bulk density-resistivity cross-plot and sonic-resistivity
cross-plots allowing the identification of source rocks
(from Meyer & Nederlof, 1984).
Serralog 8 2003
After corrections for these two parameters, the remaining carbon corresponds to the organic carbon present
within the rock.
As illustrated by Figure 10-17 a very good match exists
between the core measurements and the log measurements.
399
Reference and Bibliography
Energy (MeV)
Figure 10-17 - Typical Inelastic collision spectrum. Observe the peaks
of Oxygen and Carbon. They are more characteristic and so can be
easily detected by window energy detection collimated around these
peaks (courtesy of Schlumberger).
This method does not require a calibration on core
measurement. Unfortunately, this technique is expensive
and due to that not often used.
Figure 10-18 Evaluation of the organic carbon content
through the measurevent of the carbon-oxygen ratio
(from Herron et al.,
1988).
Total Organic Carbon (%)
400
AUTRIC, A., & DUMESNIL, P. (1985). - Resistivity,
radioactivity and sonic transit time logs to evaluate the
organic content of low permeability rocks. The Log
Analyst, 26, 3 , p. 36
BATES, R.L., & JACKSON, J.A., eds (1980). Glossary of Geology. 2nd ed., Amer. Geol. Institute, Falls
Church, Virginia, U.S.A.
BEERS, R.F., & GOODMAN, C. (1944). - Distribution
of radioactivity in ancient sediments. Bull. Geol. SOC.
America, 55.
CARPENTIER, B. et a/. (1989). - Diagraphies et
roches-mere : Estimation des teneurs en carbone organique par la methode du Carbolog. Revue I.F.P., 44, p.
600-719.
DELLENBACH, J., ESPITALIE, J., & LEBRETON, F.
(1983) - Source rock logging. SPWLA 8th Eur. Form.
Evaluation Symp. Trans., paper D.
HASSAN, M. (1973). - The use of radioelements in
diagenetic studies of shale and carbonates. Int. Symp.
Petrography of Organic Material in Sediments. Centre
Nat. Rech. Scientifique, Paris, September.
HERRON, S.L. (1986). - Derivation of the total organic
carbon log for source rock evaluation. SPWLA, 27th Ann.
Log. Symp. Trans., paper HH.
HERRON, S., Le TENDRE, L., & DUFOUR, M. (1988).
- Rock evaluation using geochemical information from
wireline logs and cores. Bull. amer. Assoc. Petroleum
Geol., 72, 8.
MALLA, M.S., & BACI, S. (1995). - Organic
Geochemistry and Well Logging. In: Schlumberger &
Sonatrach, Well Evaluation Conference, Algeria.
MANN, U., LEYTHAEUSER, D., & MULLER, P.J.
(1986). - Relation between source rock properties and
wireline log parameters: An example from Lower Jurassic
Posidonia shale, N. W. Germany. Advances in Organic
Geochemistry, 10, p. 1105-1112.
MEISSNER, F.F. (1978). - Petroleum geology of the
Bakken Formation, Williston, North Dakota and Montana.
In: The Economic of the Williston Basin. Montana Geol.
SOC.,1978 Williston Basin Symp. p. 207-227.
MENDELSON, J.D., & TOKSOZ, M.H. (1985). Source rock characterisation using multivariate analysis
of log data. SPWLA 26th ann. Logging Symp. Trans.,
paper UU.
MEYER, B.I., & NEDERLOF, M.H. (1984). ldentification of source rocks on wireli,?e logs by density/
resistivity and sonic transit time/resistivity crossplot. Bull.
amer. Assoc. Petroleum Geol., 68, 1, p. 121-129.
PASSEY, Q.R., et a/. (1990). - A practical model for
organic richness from porosity and resistivity logs. Bull.
amer. Assoc. Petroleum Geol., 74, 12, p. 1777-1794.
PERRODON, A. (1980). - Geodynamique petroliere.
Genese et repartition des gisements d’hydrocarbures.
Masson, Paris.
RUSSELL, W.L., (1945). - Relation of radioactivity,
organic content, and sedimentation. Bull. amer. Assoc.
Petroleum Geol., 29, 10.
Serralog 0 2003
Chapter 10 I 401
SCHMOCKER, J.W. (1979). - Determination of organic
content of Appalachian Devonian Shales from FormationDensity iogs. Bull. arner. Assoc. Petroleum Geol., 63.
SCHMOCKER, J.W. (1981). - Determination of organic
matter content of Appalachian Devonian Shales from
gamma ray logs. Bull. arner. Assoc. Petroleum Geol., 65,
p. 1285-1298.
SMAGALA, T.M.C., et a/. (1984). - Log-derived indicator of thermal maturity Niobrara Formation, Denver basin,
Colorado, Nebraska, Wyoming. In WOODWARD, J. et a/.
(eds), Hydrocarbon Souce Rocks of the greater Rocky
Mountain Region. Rocky Mountain Assoc. geologists, p.
355-363.
STOCKS, A.E., & LAWRENCE, S.R. (1990). ldentification of source rocks from wireline logs. In: Hurst,
A., Lovell, M.A., & Morton, A.C. (eds), Geological
Applications of Wireline Logs, Geological Society, Special
Publication, no 48, London.
SUPERNAW, I.R., McCOY, A.D., & LINK, A.J. (1978).
- Method for in-situ evaluation of the source rock potential
of each formation. U.S. patent 4,071,744, Jan. 31.
SWANSON, V.E. (1960). - Oil yield and Uranium
content of black shales. Geol. Survey, Prof. paper 356-A.
WILLIAMS, P.F.V. (1986). - Petroleum geochemistry of
the Kimmeridge Clay of Onshore Southern and Eastern
England. Marine and Petroleum Geology, 3, p. 258-281.
Serralog 0 2003
401
WELL LOGGING AND
SEQUENCE STRATIGRAPHY
Introduction
As defined by Gignoux (1936), stratigraphy is that
branch of geology which “studies the layers of the earth‘s
crust and rocks from the point of view of their chronological succession and their geographical distribution”.
For american geologists, “stratigraphy is the science
of rock strata. It is concerned not only with the original
succession and age relations of rock strata but also with
their form, distribution, lithologic composition, fossil
content, geophysical and geochemical properties indeed, with all characters and attributes of rocks as strata; and their interpretation in terms of environment or
mode of origin, and geologic histor)/‘ (Bates 8,Jackson,
1980).
Stratigraphic sequence is “a chronologic succession of
sedimentary rocks from older below to younger above,
essentially without interruption; e.g. a sequence of bedded rocks of interregional scope, bounded by unconformities” (Bates & Jackson, 1980).
Sequence stratigraphy “is the study of genetically related facies within a framework of chronostratigraphically
significant surfaces’’ (van Wagoner et a/., 1990).
than from any other method (Figs. 11-1 and 11-2).
Furthermore, logs provide much greater precision in
determining the stratigraphic distribution. Because of the
degree of detail they provide, largely due to the very good
vertical resolution of microdevices and images of the
borehole wall, the logs can provide relative dating on a
scale of the order of decimeters or even centimeters. This
cannot be achieved by any other stratigraphic technique,
Darticularlv in uniform deposits.
Figure 11-1 - Example of overturned fold
detected from the gamma ray curve. The
small section on the top-right corresponds
to the lower section of the main curve on
the left after its reversal. As it can be observed the pics correlate perfectly
(courtesy of Schlumberger).
Figure 11-2 - Other example of double
overturned fold better detected on
gamma ray curve than on sonic curve.
The dipmeter confirms the upper fold but
does not clearly reflect the lower fold
(courtesy of Schlumberger).
The modern suite of logs allows us to determine the
succession and relative age of rock strata as well as their
characters and attributes. As well admitted, well logs provide a continuous image of the sequence of formations
encountered during a logging run and, from correlations
between wells, the logs can also give an idea of geographical distribution. Given this, it is clear that well logs can
be a rich source of stratigraphic information, a fact which
has long been recognized by geologists.
The stratigraphic information which can be derived
from well logging data can be classified into several categories.
Relative dating
Because the logs show, from top to bottom, the
sequence of formations in reverse chronologic order, they
are generally useful for providing a relative dating. This is
true even in the case of folded or overturned sequences,
which can in fact be identified much more easily from logs
Serralog 0 2003
J
403
Definition of parastratigraphic units
The recorded formations can be subdivided into units
which may be termed parastratigraphic (Busson, 1972).
These units can be directly dated if characteristic flora or
fauna can be detected in core samples or cuttings, or indirectly, from surrounding sequences which have themselves been dated. These units are delimited by marker
levels above and below. Of substantial extent, these markers are approximately parallel, suggesting a continuous
sequence of deposition.
Depending on the refinement of the division into stratigraphic units, it may be possible to identify units which
correspond to very specific periods in the geological history of a basin (e.9. a volcanic episode, a coal bed, a
transgressive level), thus providing stratigraphic markers
of considerable significance.
Stratigraphic phenomena
Stratigraphic phenomena such as cessation of sedimentation, erosion, unconformities or disconformities are
usually indicated by abrupt changes in the values and
appearance of the curves on at least one log, usually
several. The use of the logs to identify these phenomena
will thus involve analyzing each discontinuity on a curve in
order to establish the most likely cause for its occurrence
and its hierarchy.
boundary, an erosional surface or a fault.
- Surfaces related to a textural change reflect either an
abrupt grain-size change in siliciclastic deposits or an
abrupt textural change in carbonates.
- Surfaces generated by currents can correspond to
planar parallel laminations, cross bedding, foreset bedding, etc.
- Surfaces can relate to erosion, hardgrounds, paraconformities (horizons of roots, burrows, mud cracks, or
flooding surface), disconformities, angular unconformities
or transgression.
- Surfaces can be caused by tectonic stresses, such
as faults or fractures.
- Surfaces can correspond to compaction or diagenetic influences, such as change in the degree of compaction, pressure solution (stylolite), cementation (calcrete or
dolocrete).
- Surfaces can be linked to a fluid change, such as a
gas-water contact.
For bedding surfaces, the criteria proposed by van
Wagoner et a/., (1990) can be extended to the borehole
images, because most of the features are easily recognized on images (Fig. 11-3).
Break detection
Break definition
A curve discontinuity or break is any significant
response change occurring over a depth interval not
exceeding the vertical resolution of the tool. This break
will appear sharper when the depth scale is more compressed and the resolution of the tool is good. Thus, it will
be easier to identify it on a 1/1000 log scale than on a
1/200 or 1/40 log scale, and if the measurement is by a
microdevice than a macrodevice.
Break sianificance
Any break is an indication of a major change in at least
one of the factors affecting the response of the tool. This
is why a break is so significant and why one must try to
determine the reason for it. However, any major change in
one of the geological parameters will provoke a response
change, and thus a discontinuity, but only on those logs
which measure parameters which are susceptible to such
a change.
Breaks on curves correspond generally to surfaces.
The best tools for detection of these surfaces are image
tools. Effectively, images of the borehole wall can detect
nearly all the surfaces crossing a well and they allow their
classification.
These surfaces can have different origins, as listed
below.
- Surfaces related to an abrupt lithologic change,
which can be marked by a plane, can correspond to a bed
404
GRAVEL
MUDSTONE
SANDSTONE
ORGANIC DEBRIS
BEDDING SURFACE
Figure 11-3 - Typical bedding surfaces and how to recognize them
(adapted from van Wagoner et al., 1990).
Break classification
These surfaces are classified by their attributes and
lithology, and those that correspond to breaks in sedimentation are extracted and placed into a hierarchy.
Breaks without maior litholoaical chanae
When the lithology, reflected by the log responses essentially GR, thorium, potassium and Pe measurements - does not change but certain curves show existence of breaks, the origin of these breaks must be estaSerralog Q 2003
Sequence stratigraphy
I Chapter 11
405
blished from among the following possible causes.
S p o n l a m s potential
- Change in the type of fluid - In this case, the break
appears mainly on the resistivity curves and possibly on
the density, neutron and sonic curves if there is gas-oil or
gas-water contact.
- Textural change - A change in sorting or in the percentage of cement will affect the porosity and consequently all the measurements which depend on it, such as
density, neutron, sonic slowness and resistivity. An example is given in Figure 11-4. These textural changes which
occur suddenly may indicate either an erosion or a transgression.
Figure 11-5 - Examples of unconformities identified by the displacement of the shale base line of the SP curve.
(a) and (b) from Doll, 1948; (c) from Serra, 1972).
Figure 11-4 - Example of textural change in a sand, detected on the
density, neutron and resistivity curves, and, to a lesser degree, on the
gamma ray. Sand (e) is more coarse and poorly sorted than the lower
sand (d). This is interpreted as an erosional surface linked to the distributary channel cutting a mouth bar. Observe also the transgressive
levels at the top of the sand bodies. They correspond to a cemented
sandstone rich in shell debris partly dissolved, the calcite or dolomite
being reprecipitated due to diagenetic phenomenum
(from Serra & Sulpice, 1975).
- Diagenesis - Diagenetic phenomena usually result in
a variation in porosity. They can correspond to local
cementation due to drying periods in a channel (calcrete
or dolocrete) or to transgression marqued by an accumulation of shell fragments associated to the sand (Fig. 114). Later, due to their partial dissolution and, ultimately,
the precipitation of the dissolved calcite linked to change
in temperature and pression, the sand is cemented.
- Tectonic accident - It may bring into contact two identical lithologies with different petrophysical properties.
Diprneter or image analysis, possibly supplemented by
correlations between wells, should identify such an occurrence.
- Unconformitv - For instance a change in the shale
base line (Fig. l l - i ) , possibly associated with a change in
radioactivity or thorium/potassium ratio (Fig. 11-6) ITIay
indicate an unconformity. A sudden change in the Sonic
slowness trend with depth is also a good indicator of
unconformity (Fig. 11-7).
Serralog 0 2003
Figure 11-6 - Detection of an unconformity using the natural gamma
ray spectrometry measurement. Note the very marked change in the
thorium/potassiumratio on the two sides of the unconformity (from
Schlumberger, Well Evaluation Conference, Venezuela, 1980).
405
C
Figure 11-7 - Three examples of the detection of unconformities from compaction profiles using slowness measurements
(F from Fertl, 1976; S from Serra, 1972; C from Chiarelli et al., 1973).
Breaks correspondina to major litholoaical chanae
A change in lithology represents a major change in
sedimentological conditions and may or may not be part
of a sequential pattern.
1) In the first case, the lithological change only represents the passage from one element of the sequence to
the next one; e.g. dolomite-,
anhydrite + halite +
potassium salt.
The nature of the lithological change will shed light on
the polarity of the sedimentary sequence. Figure 11-8
gives an example of a sequence generated in an increasingly restricted basin, while Figure 11-9 provides an
example of the converse.
2) In the second case, several reasons may be advanced to explain such a change. The choice between them
will depend on:
- a detailed analysis of the various logs, covering lithology, facies and sequential analysis, as well as interpretation of dipmeter or image data and correlations with seismic measurements in the vicinity;
- complementary information, such as analysis of samples or cuttings and knowledge of the geology of the
region.
The possible reasons for a lithological break are now
studied. For each reason the controlling elements which
support the hypothesis will also be analyzed.
Diagenesis together with a mineralogical change
This phenomenon occurs frequently in
sequences. There is a response change in the measurements which are sensitive to mineralogy, photoelectric
index, density, neutron index, sonic slowness (Fig. 11-10).
406
caliper
Figure 11-8 - Breaks observed on the curves indicate major lithological
changes in sequences characterizing an increasingly restricted basin:
limestone - do/ostone - anhydrite - halite - polyhalite (radioactive zone)
(from Serra, 1980).
Serralog 0 2003
Sequence stratigraphy
Density
w-7
2.5
I Chapter 11 I 407
Erosion
The analysis of dipmeter resistivity curve (Fig. 11-11)
or better image of the borehole wall may reveal either a
reduction in the thickness of a bed from one curve to another, or a variation in dip angle or azimuth between the
upper and lower boundaries, or sometimes both. The analysis may also reveal an absence of planarity. Correlations
with nearby wells can reveal also a gap in the sequence
(Figs. 11-12 and 11-17)
Figure 11-9 - Breaks on curves indicating lithological changes which
characterize an opening of the basin and increasingly marked marine
influences: from the bottom to the top halite anhydrite - dolostone sandstone shale (from Sera, 1980).
-
-
Figure 11-11 - Evidence of erosional surface as shown by this old HDT
dipmeter processed by GEODIP. Observe the abrupt contact below the
upper sand, the erosion of the lower shale (pad 2 resistivity curve thinner than the other curves) and the filling of the trough by the sand
(curves 1 and 4 thinner than 2 and 3) (courtesy of Schlumberger).
Figure 11-70 - Example of lithological changes due to diagenesis.
Observe from the bottom to the top the succession of limestone - shale
- dolostone - limestone - shale - dolostone. The dolostone at the origin
was a limstone. Circulation of less salty water has transformed partly
the limestone in dolostone. The part of the limestone not transformed
at the top has been protected from diagenetic change by the formation
of an anticline before the migration of oil.
Serralog 0 2003
Unconformity
The unconformity may be due to a prolonged interruption in the process of sedimentation, that is, the upper
sequence being deposited on the eroded upper surface of
the existing beds which have been exposed to various
types of erosion.
When two successive beds have been deposited horizontally, the change in the surface of erosion is the only
indication of an unconformity. In this case, the surface of
erosion usually shows up as either (a) an interval without
dip or with no dip values of coherent azimuth which indicates a heterogeneous formation whose original stratification has been disrupted by alteration and bioturbation,
or (b) by a grouping of "blue" dip pattern (Fig. 11-13).
When the unconformity is angular it is generally much
to detect Since the beds do not have the Same dip
on either side of the UnConformitY: the upper sequence
rests on the upset beds of the lower sequences affected
by tectonic movement. This condition is shown on an
arrow plot by a fairly abrupt and significant change of dip
and azimuth (Fig. 11-14).
407
Figure 11-12 - Example of unconformity identified by correlations between wells (from Lardenois & Serra, 1967).
E m
DIP ANGLE L
MAECWN
UNCONfORYITY
WEATHERE0 ZONE
Figure 11-13 - Example of an unconformity characterized by a “blue”
pattern. Observe also the change in dip magnitude and azimuth
(courtesy of Schlumberger).
Figure 11-14 - Angular unconformity identified by a change of dip
magnitude and azimuth
(from Schlumberger, Well Evaluation Conference, Iran, 1976).
408
The erosion connected with the unconformity may
have resulted in a degree of a topographical relief whose
hollows may have been the first to be filled when sedimentation restarted. This is usually indicated by dips
incressing with depth, showing up on the arrow-plot as a
“red” pattern right up to the boundary of the unconformity.
As previously seen, an unconformity may also be
revealed by a change in the shale base line on a SP curve
(cf. Fig. 11-5). Another indication of an unconformity may
be a change in the average radioactivity level of the clays,
or a change in the thorium/potassium ratio (cf Fig. 11-6).
These phenomena may be explained by a combination of
several factors:
- a change in the mineralogical composition of the
clays (e.g. transition to an illite or montmorillonite) and of
the associated elements (e.g. organic matter and heavy
minerals). This may be due either to a change in the
depositional environment, or to different conditions on
either side of the unconformity (e.9. deep burial of the
lower sequence);
- variation in the salinity of the water bound to the
clays.
As also previously seen, a sudden change in the compaction gradient may be detected by analyzing the variations of the acoustic slowness in the shales with dept.
This will represent a change in the conditions of burial or
sedimentation at the unconformity, and so indicate its presence (cf. Fig. 11-7)
Ashes, lava flows or volcanic intrusions
A bed of volcanic ashes may be characterized by a
radioactive peak due to an increase in the thorium
concentration. Lava flows or volcanic intrusions (andesite
or basalt) can be more closely identified by a detailed
lithological analysis. Basalt is dense, sometimes vesicular, poor in radioactive elements.
Emersion
A period of emersion can be detected on images which
show traces of rootlets (Fig 11-15).
Transgression
This may appear in various forms:
- as a thin calcareous bed, rich in shell debris, at the
head of a prograding sequence, topped by fine prograding
sediments (cf. Fig. 11-4);
- as a condensation level rich in phosphate debris,
glauconite and in organic matter rich in uranium characterized by a radioactive peak (Fig. 11-16);
- as deposits over an erosional surface well recognized by correlations between wells (Fig. 11-17). This figure shows the transgression of the Cretaceous over the
Upper Jurassic in the Bays de Cau, Paris Basin, France.
The quality of the correlations reveals the gradual east-towest disapprearance of the Portlandian, as well as almost
the whole of the Kimmeridgian, with only the extreme
base of the limestones with Astartes remaining at
Preuseville 16.
Serralog 0 2003
Sequence stratigraphy
1 Chapter 11 I 409
Sequence stratigraphy from well logs
Nomenclature issues
Bioturbated
ZOM
Root trams
Figure 11-15 - Example of a period of emersion detected by the observation of root traces on borehole wall image
(courtesy of Schlumberger).
Sequence stratigraphy is a method that subdivides the
geological record in terms of transgressive-regressive
facies cycles bounded by unconformities or their correlative conformities. This approach is fundamental to the
subdivision, correlation and mapping of Sedimentary
rocks.
Seismic sequence stratigraphy is a powerful application of the sequence stratigraphic method that is most
applicable to basin exploration. One can determine:
- lines or surfaces (seismic reflectors) that are considered by geophysicists as chronostratigraphic markers,
which is debated by many geologists,
- basic stratigraphic units that represent depositional
sequences, systems tracts and periodic parasequences
(Vail eta/., 1977).
However, "seismic stratigraphy [does] not offer the
necessary precision to analyze sedimentary strata at the
reservoir scale" (van Wagoner et a/., 1990). This is becau-
Figure 11-16 - Example of unconformities marked by a peak of radioactivity on the resumption of sedimentation (Serra, 1972).
Preuseville 14
Pr 15
Pr 16
Pr 10 bis
Pr 9
Pr 8 bis
Eawy 101
Ea 2
CRETACEOUS
Portalandian
Turonian
Cenomanian
CRETACEOUS
-
Sequanian
Sequanian
Figure 11-1 7 - Example of erosion and an unconformity in the Paris Basin, which can be followed from the resistivity logs. The lower Cretaceous
transgression; which is difficult to differentiate, and the Gault shales can also be followed easily (from Lardenois 8, Serra, 1967).
Serralog 0 2003
409
se of the present-day resolution of seismic surveys. Even
with improved resolution, it would not be sufficient to
detect all surfaces. Consequently, the use of seismic data
must be restricted to determining the overall geometry
and geological setting of the unit that includes the reservoir. Van Wagoner et a/., (1990) also mentioned that
"using well logs and cores, a very high resolution chronostratigraphic framework of sequence and parasequence boundaries... could be constructed to analyze stratigraphy and facies at the reservoir scale".
It is essential to calibrate the seismic sequence stratigraphy with a higher-resolution technique that is better
able to detect and interpret each sedimentation break.
This way, the surfaces that bound rock masses at different
scales can be arranged in a hierarchy.
If we refer to the example of facies, depositional environment and sequence stratigraphy analysis (Fig. 11-18)
given by Van Wagoner et a/. (1990), a modern logging set
and their interpretation can provide practically all the information needed to achieve such an analysis of formations.
As indicated in Figure 11-18, lithology, bedding type, biologic activities, partly grain size, facies, and depositional
environment can be extracted from a complete set of
modern logs including images of borehole wall recorded
in a well. This is the reason why sequence stratigraphy
and correlations between wells are possible and must be
achieved from well logs.
The rock masses can be classified from the thinnest to
the thickest (Table 11-1) as:
10) Lamina: the thinnest strata unit of uniform composition and texture, which is never internally layered
(Campbell, 1967)
9) Laminaset: a conformable succession of laminae
(Campbell, 1967)
8) Bed: a succession of laminae or lamina set bounded by surfaces (called bedding surfaces) of erosion or
nondeposition (not all beds contain laminasets)
(Campbell, 1967)
7) Genetic increment of strata (Fig. 11-19): a succession of strata (or beds) with no interruption of sedimentary processes (e.g., Bouma's sequence, varves) (Busch,
1958, 1971)
6) Bed set: a relatively conformable succession of
beds, each one separated from the others by small
breaks in sedimentary processes, and bounded by surfaces of truncation, erosion or nondeposition (Campbell,
1967). A bed set can be classified as a simple bed set if
composed of several superimposed beds of
same lithology separated by truncations or erosional surfaces, or as a composite bed set if composed of several
beds of differing lithologies (Reineck & Singh, 1975)
5) Genetic sequence of strata: a succession of genetic
increments of strata (e.g., several turbidite deposits)
(Busch, 1958, 1971; Selley, 1976)
4) Parasequence: a "succession of genetic sequences
or "relatively conformable succession of genetically related bedsets, bounded by marine flooding surfaces or their
correlative surfaces" (van Wagoner et a/., 1990). The floo410
ding surfaces indicate an abrupt increase in water depth
generally caused by external events (tectonic, climatic,
etc.)
3) Parasequenceset: a "succession of genetically related parasequences forming a distinctive stacking pattern
bounded by major marine flooding surfaces and their correlative surfaces" (van Wagoner et al., 1990);
2) System tract: a linkage of contemporaneous depositional systems (Brown & Fisher, 1977) or three-dimensional assemblages of lithofacies (Fisher & McGowen,
1967). The four recognized system tracts are: lowstand
system tract, shelf margin system tract, transgressive system tract and highstand system tract.
1) Sequence: a "relatively conformable succession of
genetically related strata bounded by unconformities or
their correlative conformities" (Mitchum, 1977). Two types
of sequences are recognized (van Wagoner eta/., 1988).
A Type-I sequence comprises lowstand, transgressive
and highstand systems tracts bounded below by Type-I
unconformities and their correlative conformities. Type-2
sequence comprises shelf-margin, transgressive and
highstand systems tracts bounded below by Type-2
unconformities and their correlative conformities.
Nonmafirmcoal subfades
(swamp subenvironment)
Carbonaceous
sand facies
(deltaic environment)
Carbonaceoussand subfacies
(delta front subenvironment)
Mnrine shale subfadeg
(otfshon, subenvironment)
Figure 11-19 - Illustration of genetic increment and genetic sequence
(adapted from Selley, 1985).
The numbers of this classification, adapted from van
Wagoner et a/., (1990), indicate the relative hierarchy
from the thickest volumes, corresponding to the longest
time interval (more than 106 yr), down to the thinnest volumes, which may be deposited in a very short period (few
days, hours or even minutes). Two intermediate hierarchies are introduced to incorporate the geological phenomena that generate intermediate stratal unit.
Most hydrocarbon bearing reservoirs fall in the category of beds to parasequences, exceptionally to parasequencesets. Consequently, they usually represent relatively short time periods, and their limited thickness cannot
be precisely studied by seismic methods. In addition,
paleontological data, where available, are generally not
precise enough to provide reliable time markers.
Because time markers are essential for determining
lateral facies evolution, they must be determined from well
logs.
Serralog Ci 2003
Sequence stratigraphy
I Chapter 11 I 41 1
TROUGH CROSS REDS
TROUGH CROSS BEDS - SMALL SCALE
SIGMOIDAL CROSS BEDS
CURRENT RIPRE OEOOING
CONTORTED BEDS
PLANAR BEDS
HUMMOCKY BEDS
WAVP RIPPLE BEDDING
CLAY CLASTS
ABUNDANT
COMMON
RARE
ABSENT
CHURNED BY eunmws
Figure 11-18 - Example of facies and sequence stratigraphy of formations (from van Wagoner et al., 1990).
Serralog 0 2003
411
Next Page
Table 11-1
Scales of geological events from laminations to macroscale cyclic sequences. Most reservoirs correspond to parasequences or parasequence sets, as their thickness ranges from few meters to hundred of meters
(modified from Van Wagoner et a/., 1990).
Stratal Units
Thickness range
Area range
Time range
(feet)
(Sq. miles)
(Years)
Using log patterns
Well logs have long been us d for facie recognition
and sequence stratigraphic analysis. Most of the time,
users base their interpretation on patterns in standard
logs such as SP, gamma ray, resistivity and sonic. In a
typical basin, this approach is valid when the patterns are
sufficiently calibrated on core data. It is not valid when
used without control, because the same patterns may be
found within different environments or reflect different
phenomena.
A reduced logging set is not considered accurate
enough for precisely determine lithology, facies, or the
nature of surfaces or correlations, as will be demonstrated
later. The first step for well log-seismic sequence stratigraphic analysis is "to interpret lithology from log character. Any error in this step can lead to misinterpretations in
the later stages of the procedure, thus confirmation with
cores and cuttings, where possible, is very important"
(Vail & Wornardt 1990). Consequently, and especially
where no core data are available, it is desirable to use the
complete logging set, including dipmeter and borehole
images. Then, one can determine lithology, facies and the
nature of their correlative surfaces, the vertical and lateral
organization in genetic sequences, and parasequences at
the reservoir scale without compromising the interpretation. One can also better detect and interpret the breaks
observed on logs and images (e.g., rootlets or burrows)
for a more accurate sequence stratigraphic analysis. The
analysis of dipmeter and borehole images allows the user
to define hierarchical relationship between surfaces,
412
Tool Resolution
Use
using the same concepts that are applied to reflector terminations in the seismic sequence analysis (Fig. 11-20).
Of course, the interpretation of these surfaces must be
guided by the facies of the two surrounding beds.
The integration of:
- seismic data to resolve sequences to parasequence
sets,
- standard logs and dipmeter to resolve parasequence
sets to beds,
- high-resolution logs with borehole imaging to resolve
bed sets to laminae and to precisely determine the nature and origin of surfaces and their lateral extent defines
the whole range of sequence units (van Wagoner et a/.,
1990) and even some units in the range of beds to parasequences (Flint & O'Byrne,l992; Busch,l971).
The proposed sequence units were inherited from
seismic interpretation and later modified by information
from well logs, cores and outcrops (van Wagoner et a/.,
(1990). Seismic attributes, lithologies, log properties,
facies, and surfaces can be traced between rock masses.
Each stratum boundary, with a minimum of two contacts
on each surface, may record evidence of the processes
active during the diastem between beds (sedimentary
impact structures, traces of fossils, burrows, rootlets,
etc.). This concept has been clearly addressed by several
workers trying to interpret different contacts (Doglioni et
a/., 1990). In summary, a hierarchy of surfaces intersects
a hierarchy of rock masses, as defined using the reflector
terminations in seismic, bedding arrangement in outcrops,
Serralog Q 2003
Previous Page
Sequence stratigraphy
lo's km
L
32 My reflector (dated)
1's km
0.1's km
1 Chapter 111 413
0.001's km
Bomhole
reflector
I O.l'sm
37My
40 to 44 My
reflector
t
Paleornagnetism
Calibrated
Paleomagnetic Seismic
Detailed
Seismic Tracing
Dipmeter Tracing
Images
(Tomography )
Basin Geometry
External Reservoir
geometry
internal Reservoir
geometry
Figure 11-20 - The nesting concept of geometry from basin to internal reservoir geometry and from seismic scale to electric imaging.
dipmeter angular relationship or borehole images.
Table 11-2
Drawbacks of simplistic use of log patterns
Simplistic use of log patterns
Facies analysis is often attempted by interpreting log
patterns from a reduced set of logs (e.g., SP, gamma ray
and resistivity). This kind of approach involves "the matching of a series of unique log patterns between wells. It
works well in marine areas where there are small
contrasts in bathymetric relief... However, in nonmarine
and marginally marine areas and where there are large
differences in bathymetric relief, as occurs along a basin
margin, correlations are difficult at besf" (Vail et al., 1990
or Vail & Wornardt, 1990).
In many cases, even in the favorable cases mentioned
by Vail, these reduced logging sets are not reliable
enough to correctly determine facies (Serra & Sulpice,
1975). There are several drawbacks associated with the
simplistic use of these tools (Table 11-2). These include
the lack of vertical resolution for SP and resistivity, the
influence of salinity contrasts between mud and formation
water on the SP deflection, radioactivity that does not
necessarily reflect the shaliness of the formation, the
influence of diagenetic effects on sonic logs, their lack of
significant response in carbonates and the inability of
these tools to detect faults, emersion surfaces, etc.. For
these reasons, using a complete set of logs is recommended (Serra & Sulpice,l975). Many examples of misinterpretationof a reduced set of logs may be shown. This
will be illustrated by two examples.
Serralog 0 2003
SP :
Lack of resolution
Flat if R,f = R,
Poor in carbonates
GR :
Influence of radioactive minerals other than clays
such as: feldspars, micas, zircon, monazite,
phosphates, organic matter...
R:
Lack of resolution
Influence of fluids
Influence of pore geometry and distribution
Influence of conductive minerals other than clays
Sonic :
Influence of compaction
Influence of pore geometry and distribution
Influence of fluids
In the first example (Fig. 11-21), the evolution of the
gamma ray, resistivity and sonic curves might be interpreted as a fining-upward sequence. However, by incorporating the photoelectric index and potassium and neutrondensity curves, the correct lithologic interpretation is a
dolostone to feldspathic sandstone with practically no
shale. This sequence corresponds to an upward increase
in energy, the opposite of the interpretation based on the
reduced set of logs. This lithologic sequence results from
episodic discharges of detrital immature material into a
brackish lake.
The second example (Fig. 11-22) illustrates a similar
case, but this could be interpreted as a coarseningupward sequence. Instead, it corresponds to an arkosic to
subarkosic sandstone grading to a dolostone, as illustra413
414
DEPTH
Well Logging and Geology
Track 1
Track 2
Track 3
Track 4
Figure 11-21 - On the basis of the evolution of the GR, resistivity and sonic, this interval would have been interpreted as a succession of fining- and
coarsening-upward sequences. The complete logging suite, especially the natural gamma ray spectrometv, photoelectric index and neutron-density curves, allows a more precise determination of the lithology and indicates that the patterns correspond to the evolution of dolostones to arkosetype sandstones with practically no shale except at the lower part at 1026 m. The sandstones are interpreted as episodic discharges of detrital
material in a brackish lake.
interpretation and even identify anhydrite nodules and
vugs.
Figure 11-22a - The upward evolution of the gamma ray suggests a
coarsening-upwardsequence. In fact, the complete logging suite clearIy indicates a dolostone abruptly superposed by an arkosic to subarkosic sand (flash flood) and a progressive return to a chemical deposit.
The Formation Microscanner images on the right (Fig. 11-22b) illustrate the bed boundary types with heterogeneous internal bed organization. Conductive (dark) patches are vuggy porosity (Vu). Resistive
(white) patches are anhydrite cement growth in the sand
(from Schlumberger Well Evaluation Conference,Angola, 1991).
ted by the lithologic reconstruction from the complete logging set. Formation Microscanner images complete the
414
J
Figure 11-22b - Images show boundaries with heterogeneous internal
bed organization. Conductive patches are vuggy pores (Vu), resistive
patches are anhydrite cement growth
(from Schlumberger, Well Evaluation Conference,Angola, 1991).
Serralog 0 2003
Sequence stratigraphy
Other misinterpretations exist. For instance, in carbonates the gamma ray curve cannot be used as a shale
indicator, because the radioactivity is essentially linked to
uranium content, or to phosphates or even to glauconite.
Natural gamma ray spectrometry is much more accurate
for carbonates, enabling the recognition of more shaly
layers which will be associated with at hence high thorium
and potassium values.
Thinly layered formations cannot be correctly interpreted without high-resolution logs such as electromagnetic
measurement type EPT or ADEPT, dipmeters or image
tools.
Surface hierarchy from well logs
The identification criteria of surfaces too small to be
resolved by seismic methods (covering the range from
parasequence sets to laminae) are discussed in this section. Because the sequence analysis of rock masses is
well defined, an analysis based upon a hierarchy of surfaces should also be established. This analysis should
take into account facies distributions and the geometric
relationships between the rock units (baselap, toplap,
truncation, concordance). Table 11-3 proposes a hierarchy of some of the surfaces and facies that are described
in the literature.
Table 11-3
Proposed hierarchy of surfaces and facies.
Surfaces
Unconformities
Paraconformities
Aggradation surfaces
Progradation surfaces
Flooding surfaces
Erosional surfaces
Hardgrounds
Burrowed surfaces
Root surfaces
Reactivation surfaces
Bedding surfaces
Amalgamation surfaces
Avalanche surfaces
Truncation surfaces
Bedding surfaces
Facies
Basin Succession
Facies Systems
Facies Cycles
recognition is obvious in outcrops, where the bedding is
visible and needs no interpretation. Because of the significant loss of geometry in the subsurface, especially below
the scale of seismic resolution, this type of identification is
critical for obtaining a detailed reservoir description.
Fortunately, the vertical arrangement of surfaces, combined with facies determination and thickness data, allows
the geometry of the rock units to be inferred.
For higher-resolution sequence stratigraphy, well logs
must be used to determine surfaces and geometries.
However, surfaces defined by log pattern correlations are
not adequate time markers, in that they correspond to
facies correlations that are generally diachronous.
Accurate time markers must be defined by analyzing the
facies succession to determine genetic increments, bed
sets, genetic sequences and parasequences. Through
this analysis, the major breaks in sedimentary processes
are determined and understood and their correlative
conformities inferred at the reservoir scale.
Building a conceptual geological model
Conceptual geological models for seismic data processing can be generated using the logging approach.
This is fundamental to obtain accurate seismic sections
that are correctly interpreted in terms of sequence stratigraphy, even in continental or carbonate deposits. It is
well known that in continental deposits, the breaks in sedimentation are not necessary linked to changes in sea
level and not easily detected on seismic sections. It is also
understood that above a certain value of dip, beds are no
longer detected vertically below the source.
Some type of scaling-up can be applied using surface
hierarchy and angular relationships between the different
hierarchies and assuming log properties like density and
sonic in consistent geometric and facies intervals to create an appropriate seismic model. Automatically detecting
geometric breaks helps to define the lateral extent of sedimentary bodies by using the angular relationships between different sequence units in a given area.
Methodology
Facies Sequences
Facies
The hierarchy of a surface may change, but the related rock mass retains its definition. The play of surfaces
and rock masses defines the geometry of subsurface
bodies, allowing us to better predict the development of
sedimentary units at the reservoir scale. Of course, this
Semlog 0 2003
I Chapter 11I 415
The methodology presented in Figure 11-23 is proposed to maximize the use of well logs in sequence stratigraphy. By applying these techniques, all the recommended (van Wagoner et a/., 1990) data are used, because
the complete set of logs provides most of the information
regarding lithology, grain size, bedding types and organic
activities (cf. Fig. 11-18). No other data source, including
cores in some instances, provides this information. These
data are essential for sequence stratigraphic studies at
the reservoir scale and also for an accurate and detailed
basin history description.
This methodology must be applied to a network of key
wells in which a complete logging set has been recorded.
This "calibration" enables a correct interpretation of log
responses in other wells with a reduced logging set.
415
METHODOLOGY
Images
Dipmeter data
Openholelogs
MINERALOGY
POROSITY
DIAGENETIC EFFECTS
SURFACE TYPES 8 ORIGIN
REALTYKNESS
,
.Facies succession
.
-
FACIES OFEACH BED
Genetic sequence
Break hierarchy
I)
w
Depositionalenvironment
Surfaces of higher rank
Imagedata
Tectonic events
(folds, faults, fractures)
Succession of surfaces in terms of reflection coefficients
and convolution with a wavelet
w
SYNTHETIC SEISMOGRAM
.Comparison with seismic section
Figure 11-23 - Diagram explaining the methodology proposed for a
complete and accurate sequence stratigraphy analysis.
Correlations from well logs
At the origin correlations were realized from core analysis using paleontological data. Since their introduction in
1927, well logs have always been used for correlation
purpose. It is the reason why well log was first called electrical coring because the resistivity logs replaced cores.
Electrical coring was replaced by the american term log,
but should be returned to common use with the emergence of images which complete the description of the formation providing a view of the borehole wall surface similar to a picture of the external core surface.
The first stratigraphic applications of well logs started
with geologists, who observed similar resistivity features
in different wells at approximately the same depth. This
suggested that these features corresponded to the same
formation. Subsequently, experience has shown that correlations established with well log features or log patterns
correspond to facies correlations (more precisely, electrofacies).
If the resistivity was historically the first log type used
for correlations, and is still in use, gamma ray should probably be preferred as this measurement is less sensitive
to fluid. Figure 11-24 is a good example of correlations
between wells using this measurement converted in color
as a function of the radioactivity level. The color image
has a better impact and allows a higher perception of the
vertical variations and lateral continuity of the electrofacies.
Modern log data and processing allow geologists to
make good facies correlations between wells. However,
only under certain conditions do the correlations have a
chronostratigraphic value.
In order to establish chronostratigraphic correlations, it
is more appropriate to correlate geologic events of higher
rank than facies. Chronostratigraphic events are reflected
by important breaks in sedimentation that are marked on
the logs by abrupt changes in lithology. These changes
Figure 11-24 - Correlations of the Permian-Cretaceous section of western Kansas based on gamma-ray logs converted to a false-color spectrum
scaled to gamma-ray values as indicated by the colorred bar (from Collins et al., 1992).
416
Serralog 0 2003
Sequence stratigraphy
indicate sudden sea-level rise (flooding, marked by overlying deeper-water facies) or fall (receding, marked by
overlying shallow-water facies or emergence features
[e.g., rootlets]). Consequently, making chronostratigraphic
correlations require the considerations described in the
following sections.
In order to correlate a discontinuous sequence of surface outcrops, we generally look for certain characteristic
patterns such as type of lithology, color, texture, sedimentary features, and the sequence of flora and fauna.
Chronostratigraphic correlations between outcrops are
generally based on characteristic fossil species or faunal
assemblage.
In subsurface datations are esentially based on microfauna or microflora collected from cores. In their absence
correlations between wells are based on logs and electrofacies. The equivalent characteristics shown by events or
groups of events on the logs are used. This technique has
been known for a long time and the early log analysts
recognized that similar characteristics were often observed between wells, sometimes over wide areas. Well logs
have the added advantage of giving a continuous, objective and quantitative evaluation of the formations. In addition, they have very good vertical resolution which makes
it possible to detect fine detail, as well as very small changes which might be missed on cores. Furthermore, some
parameters are virtually unobtainable by any other
means. Some of these parameters have a considerable
bearing on the geological history, as well as spanning
great distances and variations in facies (for example,
radioactive levels associated with cinerite or volcanic
ashes).
However, most electrofacies are diachronous because
they correspond to deposits that migrated laterally during
appreciable periods of geologic time. Consequently, electrofacies correlations differ often from the chronostratigraphic framework.
Recent introduction of the measurement of the magnetic susceptibility and the natural remanent magnetization
in a well (Lalanne eta/.,1991; Pozzi eta/., 1993; Bouisset
& Augustin, 1993; Etchecopar et a/., 1993) allows the
datation based on inversion of geomagnetic polarity (Fig.
11-25).
To detect the reversals (polarity inversion), the problem is to find the zones where the TOTal-Field signal
(TOTF) variations are higher than the SUS ones (normal
zones), or smaller (reverse zones). SUS is the
Susceptibility Signal (converted in equivalent magnetic
field).
To that end, a cross-plot (Fig. 11-26) is made for each
zone, and the slope of the best straight line is computed.
The zone can slide over the whole log, and a C curve can
be plotted (Fig. 11-27). The length of the sliding zone is
set to a compromise between a good resolution and a
high frequency noise filtering effect. The C computation
algorithm is the classical least mean square one
C=
ZTOTFi x XSUSi - M x C(T0TFi x SUSi)
( CSUSi)’ - M x CSUSi’
Serralog 0 2003
(1
-,,
)
IChapter 11 I 417
Cae data
p
3
Figure 11-25 - Comparison between polarity inversion measured on
cores and polarity inversion deduced from Natural Remanent
Magnetization (NRMT) and Magnetic Susceptibility (SUMT) measurements realized in a well (from Etchecopar et al., 1993).
where:
TOTFi is the total field signal at level i,
SUS, is the magnetic susceptibility signal (converted in
equivalent magnetic field) at level i,
M is the number of levels in the sliding-zone length.
Imem
polarlty
Figure 11-26 - The cross-plot of the corrected total field (TOTF) in
nanoteslas versus the induced field (SUS) in nanoteslas allows the
determination of the polarity of the inversions of the earth’s magnetic
field (courtesy of Schlumberger).
Figure 11-27 - Logs of the TOTF and SUS and of the C coefficient for
the research of the reversals of the magnetic Earth’s field
(courtesy of Schlumberger).
417
Total oil company used these measurements to make
correlations between wells in the Paris Basin (Fig. 11-28).
The magnetostratigraphic time scale is calibrated on the
scale published by Haq et a/. (1987). As it can be observed the correlations based on electrofacies (in fact
gamma ray) compared to the ones realized with the
magnetism logs show the diachronism of the electrofacies
correlations.
duce the same log responses just as they produce the
same lithology and facies.
Sedimentatiar
rate
Sedimenlalb
rate
Figure 11-29 - Sedimentation rates deduced from the magnetostratigraphic scales allowing the verification of the calibrations. A wrong
interpretation of the inversion sequences (right part of the figure) generates erratic sedimentation rates, especially at the reversals
(from Etchecoparet al., 7993).
J
Figure 11-28 - Correlations based on gamma ray (top) compared to
correlations based on magnetism (bottom). Induced magnetic intensity
(in nanoteslas) and the Well Magnetostratigraphic Sequence (WMS) of
polarities are shown together with interpreted absolute age time lines
based on the Haq et al. time scale. The base of limestone marked by
depths A, B and C on gamma-ray curves are indicated by same letters
on the magnetostratigraphic curves .
Observe that the induced magnetism log (in nanoteslas) is an approximate inverse image of the gamma-ray log. Both tools are sensitive to
shale content. Radioactive isotopes and magnetic minerals are both
more abundant in shale than in limestone.
(from Bouisset & Augustin, 1993; Etchecopar et al., 7993).
The determination of the inversions of polarity can be
used for the estimation of the sedimentation rate (Fig. 1129). However, wrong determination of the inversion
sequences generates over estimation of the sedimentation rate.
Correlations based on magnetostratigraphic measurements must be made taking into account other log data,
especially dipmeter or image data, as any fault may eliminate inversion sequences making the correlations more
complex.
Principle of causality
The principle of causality states that the same causes
produce the same effects. Thus, the same set of depositional conditions in a given geological period should pro418
The application of this principle allows us to assert that
the persistence of certain criteria from one point of observation (a well) to another is proof that the original causes
were the same at both points. Thus if we observe similar
log features in different wells we may conclude that:
- the depositional conditions were the same at both
locations,
- it is probably the same formation, unless the phenomenon is not isolated in time (vertically on the log), or is
repeated in the same stratigraphic interval.
Geological phenomena of considerable importance
such as periods of burial, erosion, transgression or tectonic movement will all leave their mark on the log measurements, just as they do on rocks and formations, regardless of facies and environment. These features will therefore indicate the presence of these geological phenomena. It is so true that in many places well logs provide the
basis for lithostratigraphic formations, and each lithological unit in a basin can be designated by a type well which
is used as a reference for both lithology and log characteristics.
Concepts of log correlation
There are three fundamental concepts used in the
process of log correlation.
ConceDt of similarity
This is the first concept to apply simply because it is
the most obvious and the most intuitive. It is essentially
based on the shape of the curves, that is, the frequency,
amplitude and position of log events in vertical sequences. Clearly, for each event the value of all the log parameters must be considered, otherwise there can be incorrect correlations (Fig. 11-30). This concept will be used for
correlations of fine detail, for very close and precise studies, e.g., of a field. It may be useful to consult logs with
very good vertical resolution, such as the microlaterolog
and the dipmeter or images.
Care must be taken when applying this concept for
Serralog 0 2003
Sequence stratigraphy
Sonic
Sonic
Chapter 11
I 419
Sonic
Sonic
Figure 11-30 - Example of erroneous correlations (in red) established without taking into account the measurement values. Following the initial correlations, a condensation and unconformity were proposed for the well on the right. Without rejecting this hypothesis out of hand, the dashed correlations seem to point clearly to the presence of faults (from Serra, 1972).
correlations over a considerable distance, especially
when the similarity of shape is not perfect. If, on the other
hand, the similarity is perfect, then it can be safely
concluded that the correlation is valid and chronostratigraphic.
ConceDt of rhythmicity
Sedimentation takes place in sequences, rhythms or
cycles related to geological phenomena of some importance, and will thus be characterized regionally regardless
of the type of deposition (rhythmostratigraphy as defined
in Pomerol et a/., 1980) : "Sequences are separated from
each other by discontinuity surfaces or limit-surfaces
which reveal a break in sedimentation before a return to
the conditions of deposition similar to those which formed
the basis of the preceding sequence").
This often produces similar general evolution or, put in
another way, "electrosequences" which are comparable.
This is a very important concept, since it enables us to
identify geological phenomena showing close synchronization, even allowing for a certain "delay" between one
part of the basin and another. Such phenomena include
breaks in sedimentation due to tectonic movement, transgressive periods or eustatic cycles, erosion, gaps in
deposition and "hard-ground".
Concept of lateral variability
This concept is based on two sets of evidence:
-the lateral linking of facies which is not random since,
according to the Walther's law, there is a relationship between the juxtaposed and superposed elementary
sequences over the scale of the sedimentary sequence.
In other words, at the same instant in a given basin, for
example, there will be deposition of sands, silts, clays and
coals;
- the thickness of the deposits during the same time
period depends on:
. the type of lithology and depositional environment as well as on its compaction capacity,
. the phenomenon of subsidence,
Serralog 0 2003
. a combination of the two previous phenomena.
Lateral variability will be preferred in certain types of
deposit and especially in basins (deltaic, evaporitic), all
the more so if chronostratigraphic correlations are to be
established.
Intrinsic value of correlations.
Concept of reliability
From the quantitative information given by the well
logs we can evaluate the quality of the correlations by
introducing the concepts of correlation coefficients and
reliability.
A correlation coefficient may be computed for any correlation between two curves of the same type, just as dip
is calculated using the correlogram established from the
cross-correlation technique of the resistivity curves recorded by the dipmeter tools. In a given window, that is, over
a certain interval, the greater the similarity between the
curves in terms of shape, parameter values and frequency of events, the higher will be the coefficient of correlation and the greater the reliability of the correlation.
The measure of similarity between similar curves of
two wells is computed by the standard (Pearson) correlation coefficient, r:
c o w 1 L2)
r=
SLI -SL2
(11-2)
where Cov and s are the covariance and the standard
deviation of logs L, and L, . A segment of the curve of the
first well is then moved by small increments past the similar curve of the second well. At each step a correlation
coefficient is computed and plotted as a function of lag in
a correlogram (Fig. 11-31). The lag is the depth shift between the two curves at each position of comparison.
Programs for automatic correlations using this
approach have been written. The parametersthat must be
set are similar to those used in dipmeter processing:
depth interval for correlations, step length, and search
419
length.
Well A
I
Well B
Correlogramm
Well A vs Well B
negatlve R '- positive R
Figure 11-31 - Correlogram produced by cross-correlation of a curvesegment from well I , moved by increment over similar curve from well
2. The best correlation between the two curves is selected as the peak
of the correlogram with the maximum value and the smallest lag
(from Poelchau, 1987).
Though this technique has been successfully tested by
certain companies, it is difficult to apply since changes in
the type of tool, as well as possible variations in thickness
or in the amplitude of different events from one well to the
next all have to be taken into account. Hence, a subjective judgment as to the quality of the correlation is generally based on:
- the degree of similarity between the shape of each
log and the corresponding or equivalent one in the other
well,
-the number of logs showing this characteristic similarity,
- the interval over which this similarity is observed.
Thus the correlation coefficient will be high if the degree of similarity is high on each log over a sufficiently long
interval, say several tens of meters. By applying these
concepts to log correlations between wells a specific chronostratigraphic value for the correlations can be established.
Stratigraphic value of log correlations
With a few exceptions, such as cinerite or tuff levels, or
radioactive markers, log correlations are lithological or
facies correlations. Thus, the chronostratigraphic value of
these correlations lies in the answer to the following question : do the lithological correlations or facies correlations
follow the time-lines, that is, are they synchronous, or do
they cut across these lines ?
The answer lies on both the intrinsic value of the correlations, that is, their reliability, and, in the area covered,
on the type of basin, the type of lithology, and the sedimentation model. If, in a given interval there is nearly
420
exact repetition of the shapes of the curves and of the
measured parameter values between wells, thereby providing a maximum correlation coefficient, then there is
such a low probability that there will be another such period and within a similar time span another such sedimentary cycle showing the same exact repetition of characteristics that it can be assumed that the correlation reliability is maximal and the synchronism is confirmed.
When considering the type of basin, it is clear that in a
deltaic basin, a certain large-scale similarity of sequences
can be observed. There are successions of sequences of
similar type due to the fact that the depositional process
remained more or less constant over several epochs.
There is not, however, any exact similarity, except possibly in the homogeneous pro delta shales. In the direction
of progradation of the delta there is every chance of finding diachronous facies correlations. Moving along isopic
aureoles perpendicular to the direction of progradation
should give a more or less synchronous path. Some peat
or limestone beds can constitute good time markers especially in a limited extended area (a field for instance).
In the case of a calm intracratonic basin, such as the
Paris Basin at the Kimmeridgian stage, with marly limestone type deposits, an overall similarity of sequences and
often a peak-to-peak similarity of curves can be observed
over thicknesses of more than 100 m between wells separated by several tens of kilometres. The correlations are
easily established regardless of the direction taken.
Furthermore, since the correlations are surrounded by
excellent and well-dated markers, there can be no doubt
about their chronostratigraphic value.
Whereas in some detrital sequences it is almost
impossible to establish valid correlations in massive
sands, this is not the case in shaly groups where very
good correlations are frequently found (Fig. 11-32 see
next page). In the example given (Serra, 1972). the correlations show this kind of similarity and continuity, and it
is quite reasonable to give them a chronostratigraphic
value. It follows that:
- the lateral variation of the facies and the variations in
the thickness of shale and especially sand deposits can
be traced;
- in the absence of any fauna, the space-time development of the Cretaceous transgression can be traced
very precisely, and relative dating can be established for
its lower boundary at each point;
- the chronostratigraphic non-validity of each break
can be established on the basis of lithology changes.
Correlations based on the concepts of rhytmicity and
lateral variability are by definition more chronostratigraphic. This is because they correlate events of greater
amplitude, such as major discontinuities in sedimentation
cycles, and therefore represent a higher probability of
having a certain degree of synchronism.
Search for time-markers
Chronostratigraphic correlations must be based on
time constant horizons which are rare, because they
Serralog C 2003
Sequence stratigraphy
I Chapter 11 1 421
Figure 11-32 - Correlations in shale-sand sequences. The existence of shaley episodes in the center of this group of “green sands” in the AlboAptian of the Paris Basin provides excellent correlations which more than likely have very good chronostratigraphic value. In addition, they enable
space-time variations in the Creataceous transgression to be dated accurately, in a relative way, for each correlation point. Finally, as this figure
shows, any chronostratigraphic limit based on the lithology reference is worthless (from Serra, 1972).
generally correspond to more or less instantaneous phenomena, in terms of geological time, and they must be
independent of environment. To be usefull, they also must
be of wide geographical extent. The most common timemarkers which fulfill these criteria are those corresponding to cinerite or volcanic ashes. They are often difficult
to detect from cuttings, except with a microscope, but they
are often easily seen on well logs. Other time-markers
correspond to deposits in deep-water environments related to chemical changes in water masses due to changes
in ocean water-circulation patterns. Finally, good timemarkers are those corresponding to important breaks in
sedimentary and stratigraphy records. Such breaks are
usually easily detected on logs. They correspond to either
dip changes, or change of compaction, and/or “anomalous” peaks such as radioactive peaks or very dense and
compact levels (hardground), or even breaks of sequential evolution.
Extension to seismic sections
P
Figure 11-34 - Synthetic seismogram realized taking into account density and sonic travel time from which are computed the reflection coefficient. The convolution with a wavelet of this coefficient generates the
synthetic seismogram on which the seismic section can be correlated
and calibrated
(courtesy of Schlumberger).
The high resolution of sequence stratigraphy realized
from well logs must be transfered to seismic sections in
order to extend the markers detected on well logs to the
field or the basin. This extension is achieved through vertical seismic profile (VSP) recorded in wells and calibrated
on well log data (Fig. 11-33 see next page). In absence of
VSP a synthetic seismogram can be generated and compared to the seismic sections (Fig. 11-34).
Serralog 0 2003
421
SEISMIC
SECTION
VSP
SEISMIC
SECTION
Figure 11-33 - The left image is a calibration of VSP data with formation analysis from well logs. In the center the Vertical Seismic Section recorded
in the well. On the right the comparison of the VSP with the seismic section. This succession of processing allows a better interpretation of the
seismic section
(courtesy of Schlumberger).
References and Bibliography
BATES, R.L. & JACKSON, J.A. (1980).- Glossary of
Geology. 2nd ed. Amer. Geol. Inst., Falls Church, Virginia.
BOUISSET, P.M., & AUGUSTIN, A.M. (1993). Borehole magnetostratigraphx absolute age dating, and
correlation of sedimentary rocks, with examples from the
Paris Basin. Bull. amer. Assoc. Petrol. Geologists, 77,4,
pp. 569-587.
BROWN, L.F., & FISHER, W.L. (1977).- Seismi-stratigraphic interpretation of depositional systems: examples
from Brazil rift and pull-apart basins. In C.E. Payton, (ed.),
Seismic Stratigraphy - applications to hydrocarbon exploration. AAPG Memoir, 26,p. 213-248.
BUSCN D.A. (1971).- Genetic units in delta prospecting. Bull. amer. Assoc. Petroleum GeoL, 55, p. 1137-
1154.
BUSCH, D.A. (1974). - Stratigraphic Traps in
Sandstones-Exploration Techniques. Amer. Assoc.
Petroleum Geol., Mem. 21,164 p.
BUSSON,. G. (1972).- Principes, methodes et resultats d'une etude stratigraphique du Mesozoi'que saharien.
Mem. Museum Nat. Hist Natur., nouvelle serie, serie C,
Sci. de la Terre, 26.
CAMPBELL, C.V. (1967).- Lamina, laminaset, bed
and bedset. Sedimentology, 8,p. 7-26.
CAMPBELL, R.L. (1968). - Stratigraphic applications
of dipmeter data in Mid-Continent. Bull. amer. Assoc.
Petroleum GeoL, 52,p. 1700-1719.
422
CHIARELLI, A., SERRA, O., GRAS, C., MASS, P. &
TISON, J. (1973).- Etude automatique de la sous-compaction des argiles par diagraphies differees. Rev. Inst.
Franq. Petrole, 28,1, p. 19-36.
COLLINS, D.R., DOVETON, J.H., & MacFARLANE,
P.A. (1992). - Regional gamma-ray gray-tone intensity
images of the Permian-Cretaceous sequence of western
Kansas [Abstract]. SEPM 1992, Theme Meeting, Fort
Collins, CO, p. 19.
DESVIGNES, G., BARTHES, V., & TABBAGH, A.
(1992).- Direct determination of the natural remanent
magnetization effect in a hole drilled in layered ground
from magnetic field and susceptibility logs. Geophysics,
57, 7,pp. 872-884.
DICKEY, P.A. (1979). - Petroleum Development
Geology. Petroleum Publishing Co., Tulsa.
DICKINSON, W.R. (ed) (1974). - Tectonics and
Sedimentation. SEPM, special publication 22.
DOLL, H.G. (1948). - The S.P. Log: Theoretical
Analysis and Principles of Interpretation. AIME, Technical
Publication No. 2463.
DOVETON, J.H. (1994). - Geologic Log Analysis
Using Computer Methods. AAPG Computer Applications
in Geology, no 2.
DUFF, P.Mc L.D., HALLAM, A. & WALTON, E.X.
(1967). - Cyclic Sedimentation. Developments in
Sedimentology, 10,Elsevier, Amsterdam.
EINSELE, G., & SEILACHER, A, eds. (1982).-Cyclic
and event stratification. Springer-Verlag, New York, 536 p.
ETCHECOPAR, A., MOULIN, P., PAGES, G., &
Serralog 0 2003
Sequence stratigraphy
POCACHARD, J. (1993). - Harnessing paleomagnetics
for logging. Oilfield Review, 5, 4, pp. 4-13.
FERTL, W.H. (1976). - Abnormal formation pressures.
Development in Petroleum Sciences, 2, Elsevier
Amsterdam
FISHER, W.L., & McGOWEN, J.H. (1967). Depositional systems in the Wilcox Group of Texas and
their relationship to occurrence of oil and gas. Trans. Gulf
Coast Assoc. Geol. SOC.,17, p. 105-125.
FRIEDMAN, G.M. & SANDERS, J.E. (1978). - Principles of Sedimentology. John Wiley & Sons, New York.
GARY, M., Mc AFEE, R.Jr. & WOLF, C.L. (1972). Glossary of Geology. Amer. Geol. Inst., Washington, D.C.
GIGNOUX, M. (1950). - Geologie stratigraphique.
Masson, Paris.
GILREATH, J.A. & MARICELLI, J.J. (1964). - Detailed
Stratigraphic Control through dip Computations. Bull.
amer. Assoc. Petroleum Geol., 48, 12, p. 1902-1910.
HOBSON, G.D. & TIRATSOO, E.N. (1975). lntroduction to Petroleum Geology. Scientific Press Ltd,
Beaconsfield, England.
KRUMBEIN, W.C. & SLOSS, L.L. (1963). Stratigraphy and Sedimentation. 2nd ed. W H Freeman &
Co., San Francisco.
LALANNES, B., BOUISSET, P., PAGES, G., & POCACHARD, J. (1991). - Magnetic Logging: borehole magnetostratigraphy and absolute datation in sedimentary rocks.
paper SPE 21437, presented at the SPE Middle East Oil
Show, Bahrain, Nov. 16-19.
LANDES, K.K. (1951). - Petroleum Geology. John
Wiley & Sons, New York.
LARDENOIS, J. & SERRA, 0. (1967). - Series aux
multiples visages. Rev. Inst. Franq. Petrole, 22, 12, p.
1793-1817.
LEET, L. Don, JUDSON, S. & KAUFFMAN, M.E.
(1978). - Physical Geology. 5th ed. Prentice-Hall Inc.,
Englewood Cliffs, New Jersey.
LINK, P.K. (1982). - Basic Petroleum Geology. OGCl
Publications, Tulsa.
LOCK, G.A. & HOYER, W.A. (1971). - Natural Gamma
Ray Spectral Logging. The Log Analyst, 12, 5, p. 3-9.
LOVELL, M., & PARKINSON, N. (ed.) (2002). Geological Applications of Well Logs. AAPG Methods in
Exploration, 604 p.
MACQUAKER, J.H.S., & JONES, C.R. (2002). - A
Sequence-stratigraphic
Study
of
Mudstone
Heterogeneity: A Combined PetrographidWireline Log
lnvestigation of Upper Jurassic Mudstones from The
North Sea (U.K.). In: LOVELL, M., & PARKINSON, N.
(ed.) (2002). - Geological Applications of Well Logs.
AAPG Methods in Exploration, 604 p.
MATTHEWS, R.K. (1974, 1984). - Dynamic
Stratigraphy. 1st and 2nd ed. Prentice-Hall Inc., Englewood
Cliffs, New Jersey.
MITCHUM, R.M. (1977). - Seismic stratigraphy and
global changes of sea level, Part 1: Glossary of terms
used in seismic stratigraphy. In C.E. Payton, ed., Seismic
Stratigraphy - applications to hydrocarbon exploration.
AAPG Memoir, 26, p. 205-212.
Serralog 0 2003
1 Chapter 11 1 423-
PERRODON, A. (1980). - Geodynamique petroliere.
Genese et repartition des gisements d'hydrocarbures. ElfAquitaine et Masson, Pau et Paris.
PIRSON, S.J. (1977). - Geologic Well Log Analysis.
2nd ed. Gulf Publishing Co., Houston.
POELCHAU, H.S. (1987).- Coherence mapping - an
automated approach to display goodness-of-correlation
between wells in a field. Mathematical Geology, v. 19, no
8, pp. 833-850.
POZZI, J.P., MARTIN, J.P., POCACHARD, J., FEINBERG, J., & GALDEANO, H. (1988). - In situ magnetostratigraphy: interpretation of magnetic logging in sediments. Earth & Planetary Science Letters, 88, pp. 357373.
POZZI, J.P., BARTHES, V., THIBAL, J., POCACHARD, J., LIM, M, THOMAS, Th., & PAGES, G. (1993).
- Downhole magnetostratigraphyin sediments: comparison with paleomagnetism of a core. J. Geophysical
Research, 98, pp. 1939-1957.
PRESS, F. & SIEVER, R. (1978). - Earth. 2nd ed. W H.
Freeman & Co., San Francisco.
REINECK, H.E., & SINGH, I.B. (1975). - Depositional
Sedimentary Environments. 2nd ed., Springer-Verlag,
Berlin., 439 p.
RUSSELL, W.L. (1951). - Principles of Petroleum
Geology. McGraw-Hill Book Co., New York.
Services Techniques Schlumberger (1976). - Well
Evaluation Conference, Iran.
Schlumberger Surenco S.A. (1980). - Evaluacionde
Formaciones en Venezuela.
Schlumberger (1991). - Avalia@o de forma@oes de
Angola, Well Evaluation Conference.
Schlumberger Ltd (1981). - Dipmeter Interpretation.
Volume l- Fundamentals.
Schlumberger Technical Services, Inc. (1982). Essentials of NGS Interpretation.
SELLEY, R.C. (1976). - An introduction to
Sedimentology. Academic Press, London, 408 p.
SERRA, 0. (1972). - Diagraphies et Stratigraphie. In :
Mem. B.R.G.M, 77, p. 775-832.
SERRA, 0. & SULPICE, L. (1975). - Sedimentological
analysis of shale-sand series from well logs. SPWLA, 16th
Ann. Log. Symp. Trans., paper W
SERRA, 0. (1980). - Aspects diagraphiques des evaporites. Bull. Centres Rech. SNEA (P), 4, 1, p. 411-431.
VAIL, P.R., MITCHUM, R.M., & THOMPSON, S.
(1977). - Seismic stratigraphy and global changes of sea
level, Part 3 :Relative changes of sea level from coastal
onlap. In C.E. Payton, (ed.), Seismic Stratigraphy Applications to hydrocarbon exploration. AAPG Memoir,
26, p. 63-97.
WAGONER, J.C. van, MITCHUM, R.M., CAMPION,
K.M., & RAHMANIAN, V.D. (1990). - Siliciclastic
Sequence Stratigraphy in Well Logs, Cores, and
Outcrops. AAPG Methods in Exploration Series, no 7.
423
Acknowledgements
We wish to express our gratitude to the companies which have encouraged us to publish this book. Thus, we are
deeply indebted to TOTAL, Schlumberger, ENSPM (IFP), and Georex which have made a pre-order of this book allowing thereby its printing.
Thanks also to Schlumberger, Halliburton and Baker Atlas to have authorized the reproduction of many of their
figures, most of the time after redrawing, in order to support and illustrate the text.
Finally, for use of previously published illustrations, after redrawing and coloring, we are grateful to all the cited
authors and to the following : Academic Press, The American Association of Petroleum Geologists, American
Geological Institute, Applied Science Publishers, Blackwell Scientific Publications, The Canadian Association of
Petroleum Geologists, Chapman and Hall, Elsevier Publishing Company, W.H. Freeman and Company, The
Geologists Association of London, The Geological Society of London, Geosciences Canada, Gulf Coast Association
of Geological Societies, The Institute of Petroleum, the Journal of Geochemical Exploration, the Journal of Geology,
the Journal of Petroleum Geology, the Journal of Petroleum Technology, McGraw-Hill, Masson, The Offshore
Technology Conference, Oil & Gas Journal, Petroleum Publishing Company, Prentice-Hall Inc., The Society of
Economic Paleontologists and Mineralogists, The Society of Petroleum Engineers, The Society of Professional Well
Log Analysts, Springer, John Wiley & Sons, Wilson and World Oil.
Oberto & Lorenzo SERRA
ii
Serralog 0 2003
INDEX OF REFERENCED AUTHORS
Index Terms
Links
A
Abbott
8
35
210
211
214
219
224
238
56
114
115
Aguilera
384
391
Alger
133
155
Allen
122
125
Adams
155
213
164
168
193
Anderson
82
115
Athy
271
274
Atwater
268
294
Augustin
417
418
Autric
399
400
Babcock
360
361
Baci
398
400
Baker
145
156
Baldwin
112
113
115
272
294
1
34
119
121
155
159
171
180
193
197
236
239
255
259
262
265
294
299
301
306
309
310
312
352
355
391
400
403
422
123
124
155
Beaudoin
33
34
334
352
Beers
57
115
260
262
352
294
422
B
Bates
Beard
391
396
400
Bell
360
391
Bengtson
330
332
333
81
82
117
304
305
306
Biggs
Billings
391
Bissell
244
262
This page has been reformatted by Knovel to provide easier navigation.
307
352
Index Terms
Blatt
Links
48
115
160
193
200
242
254
262
268
294
82
83
116
Bonnetain
190
193
Borg
267
294
Bourke
13
34
Bousset
417
418
422
Boyeldieu
356
377
380
391
Bram
85
115
Brie
145
146
155
385
391
Brown
198
410
422
Burst
272
294
Busch
197
199
236
410
412
191
193
199
Blunden
422
Busk
341
352
Busson
404
422
Butler
356
391
161
162
410
422
Carpentier
396
400
Chermette
344
353
Cheung
27
34
Chiarelli
278
280
285
287
288
289
291
293
294
406
262
263
271
155
244
245
45
115
C
Campbell
422
Chilingar
275
294
Chilingarian
240
250
274
294
127
128
246
262
Christensen
67
85
115
Clarke
38
39
42
Coleman
203
232
237
Collins
416
422
Coogan
270
294
Cox
360
391
Crain
82
115
Curry
206
207
Choquette
231
This page has been reformatted by Knovel to provide easier navigation.
236
Index Terms
Links
D
Daly
39
44
63
66
115
Dallmus
276
294
Dapples
243
262
Daukoru
235
238
Davis
201
Deer
49
115
Delhomme
19
34
147
148
149
151
153
155
186
192
193
237
255
258
376
378
391
56
70
74
115
Dodge
123
155
Doll
405
422
Donath
306
352
Dragoset
81
115
Draxler
85
115
Dubas
333
335
336
347
352
Dumesnil
399
400
Dunoyer de Segonsac
244
262
Dzulinski
170
193
359
392
85
116
124
155
166
167
275
276
295
190
194
333
335
336
341
342
343
344
345
347
348
352
417
418
284
289
Desbrandes
E
Edwards
Emmermann
van Engelhard
Etchecopar
422
Evamy
243
262
Fairbridge
240
244
295
Faivre
380
392
393
Faure
344
353
Fertl
265
271
278
295
406
423
F
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Fisher
198
232
422
423
Folk
157
Ford
233
237
248
249
263
79
80
116
Foster
283
295
Fraser
122
123
125
Frayssinet
112
114
116
Friedman
201
237
Fuchtbauer
250
263
Gaida
154
275
276
295
Galloway
162
194
201
203
204
232
237
Garrels
39
116
Gasparini
56
114
Ghibaudo
200
Gignoux
403
423
Gilreath
191
194
237
330
353
Goetz
191
194
237
4
34
57
115
260
262
396
267
295
267
410
295
G
Golyer
Goodman
400
Gough
360
391
Graton
122
123
85
116
Gressly
197
237
Griggs
306
Gutschick
171
Guy
164
194
82
83
Halbouty
299
353
Haq
418
Grau
H
Haile
Harker
15
34
Hassan
285
295
116
258
400
Hatcher
304
353
Haug
197
237
This page has been reformatted by Knovel to provide easier navigation.
260
396
Index Terms
Links
Head
124
125
Heckel
171
194
Hedberg
271
295
Heliot
27
34
Henkel
76
77
116
Henry
322
324
325
326
327
329
330
339
340
350
400
203
204
353
Herron
262
263
Hitchon
275
295
Hjulstroms
166
167
194
Hobday
162
194
201
237
Holmes
39
116
Hossin
285
295
Hottman
283
288
1
201
277
296
Illing
4
35
Islam
4
35
1
34
119
121
155
159
171
180
193
197
236
239
255
259
262
265
294
299
301
306
309
310
312
352
355
391
400
403
422
Johnson
283
288
295
Jopling
168
169
194
Hutton
295
I
Ibrahim
J
Jackson
K
Karker
139
Katz
277
296
Kenyon
145
156
Knight
275
294
Kohonen
112
113
116
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Krumbein
Krynine
Links
47
48
116
119
120
121
122
136
154
160
194
197
201
230
237
239
241
242
246
263
45
46
48
116
162
296
194
L
Lahee
4
35
Lalanne
417
423
Land
249
263
Lang
289
290
291
Lardenois
408
409
423
Lauger
277
297
Laurier
225
227
Lee
124
156
Leet
3
52
Leith
45
117
Lennon
206
237
Lewis
276
296
Lombard
198
199
Longman
240
263
Luthi
27
35
Lyell
201
237
Malla
398
400
Mann
399
400
Manus
270
294
Maricelli
330
353
Martin
377
384
Mathieu
381
393
Maxwell
267
268
99
117
117
200
237
294
296
M
Mayer
McClelland
113
McCulloh
268
269
McGowen
410
423
McKee
160
163
McKenzie
39
116
Mead
39
45
296
168
117
This page has been reformatted by Knovel to provide easier navigation.
181
194
Index Terms
Links
Mendelson
398
400
Meyer
399
400
Miall
228
237
Middleton
197
201
Miller
268
294
Mitchum
198
Monk
136
154
Moore
197
206
Morris
365
393
Motet
19
34
192
228
236
237
151
155
186
193
255
376
378
156
253
254
237
391
Mutti
200
Neasham
126
127
Nederlof
399
400
129
155
160
195
N
Nurmi
156
O
Otto
P
Payne
199
Peck
273
276
285
297
Perrodon
299
353
395
400
Pettijohn
39
40
41
46
117
121
159
160
161
164
170
195
250
251
263
Philipp
290
296
Pickett
66
67
117
Pilenko
152
Pirson
201
232
237
Pittman
253
Poelchau
420
423
Poster
5
35
Potter
123
156
160
161
169
195
272
296
Powers
159
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Pozzi
417
423
Pray
127
128
246
262
40
155
244
245
41
42
117
241
263
303
353
124
125
156
de Raaf
201
227
237
Ragan
321
353
393
Ramsay
302
304
309
353
355
393
Rao
133
156
Rasmus
145
156
393
Raymer
81
82
117
Reading
197
200
201
237
Reineck
163
164
167
168
169
195
199
201
410
423
Press
Pryor
203
R
Renault
277
Ricci-Lucchi
200
Richardson
165
167
Rider
225
227
237
Rieke
271
274
294
Rivest
153
156
Roberts
268
269
Robertson
269
297
Rodriguez
171
Rogers
114
117
Rose
276
296
Rukhin
277
Rumelhart
113
Russel
260
263
Ryer
277
297
296
124
125
396
400
S
Sabins
235
Sanders
201
Sanyal
85
86
Sarma
133
156
Sawabini
265
297
310
117
This page has been reformatted by Knovel to provide easier navigation.
341
Index Terms
Links
Scheidegger
161
Scholle
201
Schwarzacher
161
195
Scott
161
195
Seilacher
171
195
Selley
125
157
159
160
165
166
170
171
191
194
197
198
199
200
227
229
231
237
238
260
263
410
423
140
157
7
8
13
35
58
69
79
117
131
132
139
140
147
148
155
182
183
184
185
186
187
188
191
195
203
207
210
211
213
214
219
223
224
226
228
233
238
259
263
286
289
292
297
405
406
407
409
412
413
419
420
421
423
Sibbit
99
117
380
392
393
Siever
40
41
42
117
241
263
303
353
Simons
164
165
196
Singh
163
164
167
168
169
195
199
201
410
423
Sippel
250
263
Sloss
48
116
119
120
121
160
194
197
198
201
230
237
239
263
122
157
Souhaitié
27
35
Spearing
201
Spurlin
100
Stadler
254
Suau
100
118
236
384
393
Sulpice
132
139
149
157
203
233
238
405
413
423
Sen
Serra
Sneed
195
201
118
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Sundborg
166
Supernaw
57
118
260
396
397
401
396
397
401
Taylor
267
268
297
Teichert
197
238
Terzaghi
273
276
Theys
141
157
Timko
278
284
Toksoz
398
400
Trask
121
157
Vail
198
409
Verral
267
296
Visher
231
232
122
157
16
35
199
Swanson
261
263
T
285
297
295
V
412
413
423
163
196
198
200
201
202
205
229
238
403
404
409
410
411
412
415
423
197
200
201
227
228
231
232
238
394
238
W
Wadell
van Wagoner
Walker
Walther
199
Walton
170
Washington
42
Watfa
382
384
Weber
235
238
39
46
118
Weir
160
163
168
194
Weller
197
250
263
272
155
Wedepohl
297
Weyl
123
124
Whalen
283
295
39
118
Wickman
This page has been reformatted by Knovel to provide easier navigation.
277
Index Terms
Links
Wilson
253
263
Wolf
240
250
Wyllie
283
297
Zemanek
373
394
Zingg
122
157
Zobell
239
241
262
Z
264
This page has been reformatted by Knovel to provide easier navigation.
263
INDEX – GLOSSARY
Here below are listed terms or expressions used in geology and well logging, followed by a short definition or explanation. When they
are explained more fully in this book, only the relevant chapter and paragraph is referred to. See also the Index and Glossary of
Volume 1 for all the usual well logging terms. The figure refers to the chapter and the pages.
Index Terms
Links
A
Absorption
- the process of taking up by capillary, osmotic, chemical or solvent
action.
- the process by which energy, such as that of electromagnetic or
acoustic waves, is converted into other forms of energy.
Abyssal plain : flat regions at the bottom of major ocean basins with
a water depth greater than 4 000 m.
Accessory minerals : minerals present in such small amount (i.e.
less than 1 %) that their presence or absence is not significant when
considering the mineral composition for classification purposes, but
which can affect some logging measurements if their inherent properties
are considerably different from those of the principal minerals (i.e. zircon
and monazite which have a so high content in thorium and uranium that
they affect the total radioactivity even with a percentage less than 1 %).
Accretion : a gradual increase in size of an inorganic body by the
external addition of new particles deposited by a stream.
Acidic : a descriptive term applied to those igneous rocks that
contain more than 60 % SiO2
Acoustic : of or pertaining to sound.
Acoustic impedance : the product of acoustic velocity and density.
Activation : technique in which the rocks are irradiated with neutrons
that transmute some nuclei into radioisotopes which are characterized
by the energy of the induced gamma rays and by their decay time
schemes.
Actualism :see Uniformitarianism.
Adsorption : adherence of gas molecules, or of ions or molecules
in solution, to the surface of solids with which they are in contact.
Aeolian : pertaining to the wind.
Agglomerate : a pyroclastic rock composed mostly of bombs.
Aggradation :the building-up of the Earth's surface by deposition of
detrital material by a stream.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Aggregate : a mass or body of rock particles or mineral grains or
both.
Albite : pure sodium-feldspar end member in the plagioclase series.
Allochthonous : formed or produced elsewhere than in its present
place.
Allogenic : formed or generated elsewhere, usually at a distant
place.
Alluvial : pertaining to or composed of alluvium, or deposited by a
stream or running water.
Alluvial fan : mass of loose rock material, shaped like an open fan,
deposited by a stream.
Alluvium : a general term for detrital material deposited by a stream
or running water in the bed of the stream or on its flood plain or
delta ,or as a cone at the base of a mountain slope.
Alteration : any change in the chemical or mineralogical composition
of a rock produced by weathering or by the action of hydrothermal
solutions.
Amphibole : a group of dark rock-forming ferromagnesian inosilicates
having the general formula :A2-3 B5 (Si,AI)8 O22 (OH)2, where A =
Mg, Fe2+, Ca, or Na, and B = Mg, Fe2+, Fe3+, Al orTi.
Amphibolite : a metamorphic rock consisting mainly of amphibole
and plagioclase with little or no quartz.
Amplitude: half the height of the crest of a wave above the adjacent
troughs. The maximum value of the displacement in an oscillatory
motion.
Anadiagenesis
Chapter 6, p. 244
Anaerobic : said of organisms that can live in the absence of free
oxygen, or of conditions that exist only in the absence of free oxygen.
Anastomosis [streams] : a product of braiding; esp. an interlacing
network of branching and reuniting channels.
Andesite : a dark-colored, fine-grained extrusive (volcanic) rock.
Anhydrite : anhydrous calcium sulfate of the evaporite group.
Anion : a negatively charged ion.
Anisotropy : the condition of having different properties in different
directions.
Anorthite : pure calcium-feldspar end member of the plagioclase
series.
Anorthoclase : sodi-potassic alkali feldspar.
Anticline : a convex upward fold.
Chapter 8, p. 310
Antiform: a fold whose limbs close upward
Chapter 8, p. 310
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Antithetic : pertaining to minor faults that are oriented opposite to
the major fault with which they are associated.
Aragonite : orthorhombic calcium carbonate with greater density
and hardness and less stability than calcite.
Arcuate delta : a curved or bowed delta with its convex outer
margin facing the sea or lake.
Arenaceous : said of a sediment or sedimentary rock consisting
wholly or partly of sand-size fragments.
Arenite : a general name used for consolidated sedimentary rocks
of sand-size fragments irrespective of composition.
Argillaceous : pertaining to, largely composed of, or containing
clay-size particles or clay minerals.
ARI*: Schlumberger's tool. Acronym for Azimuthal Resistivity
Imager.
Chapter 3, p. 132
Arkose : a feldspar rich, coarse-grained sandstone, pink or reddish.
Arrow plot : a display of dipmeter data.
Chapter 8, p. 330
Ash :fine (< 2 mm in diameter) pyroclastic material.
Asthenosphere : the layer or shell of the Earth below the
lithosphere.
Authigenesis : the process by which new minerals form in place
within a sediment or sedimentary rock.
Chapter 6, p. 252
Authigenic : formed or generated in place.
Autochthonous : formed or produced in the place where now
found.
Avulsion : an abrupt abandonment of a segment of a river channel.
Axial surface
Chapter 8, p. 309
Axis : the line which, moved parallel to itself, generates the form of
a fold.
Azimuth : direction of a horizontal line as measured clockwise from
North on an imaginary horizontal circle.
Azimuth frequency plot
Chapter 8, p. 320
Azoic : said of an environment that is devoid of life.
B
Back roof : the landward side of a reef.
Bar : a generic term for any of various elongate offshore ridges,
banks or mounds of sand, gravel, or other unconsolidated material.
submerged at least at high tide.
Barchan : an isolated crescent-shaped sand dune lying transverse
to the direction of the prevailing wind.
This page has been reformatted by Knovel to provide easier navigation.
Chapter 9, p. 375
Index Terms
Links
Barite : sulfate of barium.
Barrier : an elongate offshore ridge or mass rising above the hightide
level, generally extending parallel to, and at a some distance from,
the shore ,and built up by the action of waves or currents, or by organisms.
Basal-conglomerate : a conglomerate that forms the bottom stratigraphic
unit of a sedimentary series and that rests on a surface of
erosion, thereby marking an unconformity.
Basalt : a general term for dark-colored basic and mafic igneous
rocks, commonly extrusive but locally intrusive.
Basement: the undifferentiated complex of rocks that underlies the
rocks of interest in an area.
Basic : said of an igneous rock having a relatively low silica content,
relatively rich in iron, magnesium and/or calcium, and thus includes most
mafic minerals.
Basin : a low area in the Earth's crust, of tectonic origin, in which
sediments have accumulated.
Bathyal : pertaining to the ocean environment or depth zone
between 200 and 2 000 metres.
Bauxite : a rock composed of a mixture of various amorphous or
crystalline hydrous aluminium oxides and hydroxides. A common
residual of clay deposits in tropical and subtropical regions.
Bay : a wide, curving open indentation, recess, or inlet of a sea into
the land.
Beach : a shore of a body of water, formed and washed by waves
or tides, usually covered by sandy or pebbly material.
Bed : the smallest formal unit in the hierarchy of lithostratigraphic
units, distinguishable from layers above and below.
Chapter 4, p. 160
Bedding : the arrangement of a sedimentary rock in beds or layers
of varying thickness and character.
Chapter 4, p. 161
161
168
169
Bed load : the part of the total stream load that is moved on or
immediately above the stream bed, such as the larger or heavier
particles transported by traction or saltation along the bottom.
Bedset : a group of strata bounded by stratification surfaces.
Chapter 4, p. 163
Bell shape : an evolution of a curve (i.e. SP or resistivity) with depth
drawing the shape of a bell.
Chapter 5, p. 203
Benthic : pertaining to the benthos.
Benthos : those aquatic organisms that live on or within the
sediment at the bottom of a body of water.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Bentonite : a soft, plastic, porous, light-colored rock composed
essentially of clay minerals of the montmorillonite group plus colloid silica,
produced by devitrification and accompagnying chemical alteration
of a glassy igneous material, usually a tuff or volcanic ash.
Biochemical : characterized by, or resulting directly or indirectly
from, the chemical processes and activities of living organisms.
Bioclastic : consisting primarily of fragments of organisms.
Biogenic : produced directly by the physiological activities of
organisms.
Bioherm : a moundlike, domelike, lenslike, or reeflike mass of rock
built up by sedentary organisms, composed almost exclusively of their
calcareous remains.
Biostrome : a distinctly bedded and widely extensive blanketlike
mass of rock built by and composed mainly of the remains of sedentary
organisms.
Biotite : a dark and dense mineral of the mica group.
Bird's-eye fabric : a common pattern in supratidal carbonates in
which former gas bubbles become preserved as open or calcite-filled
cavities. These cavities are typically 2 to 5 mm in diameter and may
constitute 50 % of the rock.
Block
- [part. size] a large, angular rock fragment having a diameter
greater than 256 mm; it may be nearly in place or transported
by gravity or ice.
- [volc] a pyroclastic particle larger than 64 mm ejected from a
volcano in a solid state.
"Blue pattern ": a convention used in dipmeter interpretation. It
corresponds to an increasing dip magnitude with decreasing
depth with nearly uniform azimuth.
Body Force
Chapter 8, p. 330
Chapter 8, p. 301
Bog : waterlogged, spongy ground, consisting primarily of mosses,
containing acidic, decaying vegetation that may develop into peat.
Bomb : a pyroclastic particle larger than 64 mm ejected from a volcano
while viscous but solidified and received its more or less rounded
shape while in flight.
Bone bed : a sedimentary layer characterized by a high proportion
of fossil bones, scales, teeth, coprolites (phosphatic deposits).
Bottomset : a nearly horizontal layer of sediment deposited in front
of the advancing foreset beds.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Boulder : a rock fragment or particle having a diameter greater than
256 mm.
Boundstone : (Dunham's classification] a term used for a sedimentary
carbonate rock whose original components were bound together
during deposition and remained substantially in the position of growth.
Bound water
- water which has become adsorbed to the surfaces of solid particles
or grains. Under natural conditions this water tends to be viscous and
immobile but might not have lost its electrolytic properties.
- water which is chemically bound by becoming part of a crystal lattice.
This water cannot be removed without changing the structure or
composition of the material. It has lost its electrolytic properties.
Break : syn.: discontinuity.
Chapter 11, p. 404
Breccia : a general term for a coarse-grained clastic rock consisting
of angular, broken rock fragments held together by a mineral cement or
in a fine-grained matrix. This implies a minimum transport of fragments.
Brine : a term used for highly saline waters present in restricted
basins.
Brittle : "said of a rock that fractures at less than 3-5% deformation
or strain" (Bates & Jackson, 1980).
Bulk density : the weight of a material divided by its volume
including the volume of its pore spaces.
Bulk modulus
Chapter 8, p. 305
Burrow : a tubular or cylindrical hole made by a mud-eating animal.
Chapter 4, p. 170
Button : a small disc-shaped, button-like electrode used in microresistivity pads (ML, MLL, SHDT, FMS/FMI, EMI, STAR).
C
Cable : a wireline.
Calcarenite : [Grabau's classification] a limestone consisting predominantly
(more than 50 % ) of detrital calcite particles of sand size.
Calcareous : said of a substance that contains more than 10 % and
less than 50 % calcium carbonate.
Calcilutite : [Grabau's classification] a limestone consisting predominantly
(more than 50 %) of detrital calcite particles of silt and/or clay
size.
Calcirudite : [Grabau's classification] a limestone consisting predominantly
(more than 50 % ) of detrital calcite particles larger than sand
size.
Calcite : a calcium carbonate.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Calibration : the process wherein the scale and sensitivity of the
measuring circuit is adjusted to meaningful units.
Chapter 1
Caliper : a wireline logging tool which measures hole diameter.
Capillary forces : .
Cap rock : an impervious rock overlying a reservoir.
Carbonate : a sediment formed by the organic or inorganic precipitation
from aqueous solution of carbonates of calcium, magnesium, or
iron.
Carnallite : an evaporite mineral.
Cast : a sedimentary structure representing the infilling of an original
mark or depression made on top of a soft bed, and preserved as a
solid form on the underside of the overlying stratum.
CAST-V™ : Halliburton's tool. Acronym for Circumferential Acoustic
Scanning Tool.
Chapter 3, p. 132
Cation : a positively charged ion.
Cation exchange : the displacement of a cation bound to a site on
the surface of a solid, as in clay-minerals, by a cation in solution.
Cation Exchange Capacity : symbol CEC: a measure of the extent
to which a substance will supply exchange cations.
Cave : part of a borehole where the hole diameter becomes larger
than the drill bit diameter.
CBIL : Baker Atlas's tool. Acronym for Circumferential Borehole
Imaging Log.
Chapter 3, p. 132
Celestite : sulfate of strontium occurring in deposits of salt, gypsum,
and associated dolomite and shale, and in residual clays.
Cement : mineral material usually chemically precipitated in the spaces
between the individual grains or crystals (pores), thereby binding
them together as a rigid, coherent mass.
Cementation
Chapter 6, p. 242
Cementation factor : the porosity exponent m in Archie's formula.
Syn.: tortuosity factor.
Chalk : a soft, friable, pure, earthy, fine-textured limestone of marine
origin consisting almost wholly (90-99 %) of calcite, formed mainly by
shallow-water accumulation of calcareous tests of floating
microorganisms.
Chamosite : an hydro-alumino-silicate of the chlorite group, rich in
iron. An important constituant of many oolitic and other bedded iron ores.
Channel : an elongate depression where a natural body of water
flows; an abandoned or buried watercourse represented by stream
deposits of gravel and sand.
This page has been reformatted by Knovel to provide easier navigation.
251
Index Terms
Links
Channel lag : a deposit consisting of the coarsest material that settles
out and accumulates along the deepest part of a river channel.
Chemical rock : a sedimentary rock composed primarily of material
formed directly by precipitation from solution or colloidal suspension.
Chert : a hard, extremely dense or compact, dull to semivitreous,
microcrystalline or cryptocrystalline sedimentary rock, consisting
dominantly of interlocking crystals of quartz less than about 30 gm
indiameter. It may contain amorphous silica (opal). It occurs principally
as nodules, less commonly as areally extensive layers.
Chlorite : an hydrous-alumino-silicate of iron and magnesium.
Clastic : pertaining to a rock composed principally of broken fragments
that are derived from preexisting rocks or minerals and that have
been transported some distance from their place of origin.
Clay
- a rock or mineral fragment or a detrital particle of any composition
having a diameter less than 1/256 mm.
- a loose, earthy, extremely fine-grained, natural sediment or soft
rock composed primarily of clay-size or colloidal particles and characterized
by high plasticity and by a considerable content of clay minerals.
- a clay mineral.
Clean : containing no appreciable amount of clay or shale.
Cluster analysis : a procedure for arranging a number of objects in
homogeneous subgroups based on their mutual similarities and
hierarchical relationships.
CLUSTER* : a Schlumberger trade mark for a dip computation
technique.
Coal: a readily combustible rock containing more than 50 % by
weight and more than 70 % by volume of carbonaceous material, formed
from compaction of altered plant remains similar to those in peat.
Cobble : a rock fragment or sediment particle having a diameter in
the range of 64-256 mm.
Chapter 3
Cohesiveness : a mass property of unconsolidated, fine-grained
sediments by which like or unlike particles (having diameters less than
0.01 mm )cohere or stick together by surface forces.
Cohesive strength
Chapter 8, p 303
Compaction
Chapter 6, p. 242
Competent : said of a layer which, in contrast to adjacent layers,
has formed more nearly parallel folds (the adjacent layers being more
nearly in similar folds).
This page has been reformatted by Knovel to provide easier navigation.
Chapter 7
Index Terms
Links
Component : one of a set of chemical compositions the relative
masses of which may be varied to describe all compositions within it.
Composite log
Chapter 2
Chapter 1, p. 21
Composition
Chapter 2
Compressibility :the reciprocal of bulk modulus.
Chapter 8, p. 305
Compression : "a system of forces or stresses that tends to decrease
the volume of a subsatnee" (Blates & jackson", 1980).
Compressive strength
Chapter 9, p. 359
Compressive stress : a normal stress.
Chapter 8, p. 301
305
308
Cone : a structure shaped like a cone.
Chapter 8, p. 327
Confining pressure
Chapter 8, p. 301
Conformable : said of strata or stratification characterized by an
unbroken sequence in which the layers are formed one above the other
in parallel order by regular, uninterrupted deposition under the same
general conditions.
Conglomerate : a coarse-grained clastic sedimentary rock, composed
of rounded to subangular fragments larger than 2 mm in diameter.
Conical fold : a fold model that can be described geometrically by
the rotation of a line about one of its ends, which is fixed.
Chapter 8, p
Consolidated : pertains to a rock framework provided with a degree
of cohesiveness or rigidity by cementation or other binding means.
Core
- a cylindrical section of rock.
Chapter 1, p. 5
- the central zone or nucleus of the Earth's interior.
Correlations
Chapter 11, p. 418
CORIBAND*: a Schlumberger mark for a program of interpretation
for complex lithologies.
Coset: a sedimentary unit composed of two or more sets.
Chapter 4, p. 168
Crack
Chapter 9, p. 355
Craton : a part of the Earth's crust that has attained stability, and has
been little deformed for a prolonged period.
Creep : continuously increasing strain resulting from a small
constant stress acting over a long period of time.
Crest: the highest point or line of a landform.
Chapter 8, p. 309
Crestal line
Chapter 8, p. 309
Crestal surface
Chapter 8, p. 309
Crevasse :
- a wide breach or crack in the bank of a river.
- cracks in the top of a glacier.
This page has been reformatted by Knovel to provide easier navigation.
327
Index Terms
Links
Cross-bedding : cross-stratification in which the cross-beds are
more than 1 cm in thickness.
Crossplot: a graphic plot of one parameter versus another.
Chapter 4, p. 168
Chapter 2, p. 67
Crust : the outermost layer or shell of the Earth.
Crystal : a homogeneous, solid body of a chemical element,
compound, or mixture, having a regularly repeating atomic
arrangement that may be outwardly expressed by plane faces.
Cuttings : rock chips cut by a bit in the process of well drilling.
Cycle
Chapter 1, p. 6
Chapter 5, p. 198
Cyclic sedimentation
Chapter 5
Cyclographic projection
Chapter 8, p. 323
Cylinder
Chapter 8, p. 326
Cylindrical fold : a fold model that can be described geometrically
by the rotation of a line through space parallel to itself.
Chapter 8, p. 326
D
DAPSA plot
Chapter 8, p. 332
DCA* : a Schlumberger mark for a program of Detection of
Conductive Anomalies.
Chapter 9, p. 370
Debris : any surficial accumulation of rock fragments, soil material,
and sometimes organic matter detached from rock masses by chemical
and mechanical means.
Debris flow : a moving mass of rock fragments, soil and mud, more
than half of the particles being larger than sand size.
Decompaction
Chapter 7, p. 293
Deep-sea fan : a submarine equivalent of an alluvial fan. Syn.
turbidite.
Deformation : syn. strain.
Chapter 8, p. 303
Deltaic : pertaining to or characterized by a delta.
Dendrogram : a treelike diagram depicting the mutual relationships
of a group of items sharing a common set of variables.
Chapter 5, p.216
Density current : any current that flows downslope because of it is
denser than the fluid around.
Density stereogram
Chapter 8, p. 321
Depositional environment : a natural geographic entity in which
sediments accumulate.
Depth of investigation : the radial distance from the measure point
of a sonde within which material contributes significantly to the readings
from the sonde.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
DETFRA* : a Schlumberger mark for a program of Detection of
Fractures.
Chapter 9, p. 377
Deviation : departure of a borehole from vertical. Syn. : drift.
Detrital : pertaining to or formed from detritus.
Dextral
Chapter 8, p. 312
Diagenesis
Chapter 6
Diapir : a dome or anticlinal fold in which the overlying rocks have
been ruptured by the squeezing-out of plastic core material (salt, shale,
or igneous intrusions). Syn. : salt dome.
Diapirism :the process by which a diapir is formed.
Diatomite : a siliceous sedimentary rock consisting chiefly of opaline
frustules of the diatom, a unicellular aquatic plant related to the algae.
Dike : a tabular igneous intrusion that cuts across the bedding or
foliation of the country rock.
Chapter 4, p. 171
Dilation : deformation by a change in volume but not shape.
Diorite : a group of plutonic rocks intermediate in composition
between acidic and basic.
Dip : the angle that a structural surface makes with the horizontal,
measured perpendicular to the strike of the structure and in the vertical
plane.
Chapter 8, p. 315
Dipmeter
Chapter 8, p. 315
Dip pattern
Chapter 8, p. 330
Dip slip
Chapter 8, p. 312
Discontinuity : any interruption in sedimentation, whatever its
cause or length, usually a manifestation of nondeposition and
accompanying erosion.
Discordance : lack of parallelism between adjacent strata. Angular
unconformity.
Dispersed : a term used to refer to particles (clays) distributed
within the interstices of the rock framework.
Distal: said of a sedimentary deposit consisting of fine clastics and
formed farthest from the source area.
Distributary : a divergent stream flowing away from the main
stream and not returning to it, as in a delta or on an alluvial plain.
Dolomite : a carbonate of calcium and magnesium.
Dolomitic : said of a rock that contains 10-50 % the mineral
dolomite in the form of a cement and/or grains or crystals.
Dolomitization : the process by which limestone is wholly or partly
converted to dolomite rock or dolomitic limestone.
Chapter 6, p. 255
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Dome : an uplift or anticline structure, either circular or elliptical in
outline, in which the beds dip gently away in all directions.
Drag : the bending of strata on either side of a fault, caused by the
friction of the moving blocks along the fault surface.
Chapter 8
Drift
- [drilling ] the attitude of the borehole; the drift angle or deviation is
the angle between the borehole axis and the vertical.
-(glacial geol.].;
- [geophysics] .
Drumlin : a low, smoothly rounded, elongate oval hill, mound or
ridge of compact glacial till.
DUALDIP*: a Schlumberger mark for a program of dip computation
for SHDT dipmeter tool.
Ductile : said of a rock that is able to sustain 5-10% of strain
before fracturing.
Chapter 8, p. 304
Ductility : a measure of the degree to which a rock is ductile.
Chapter 8, p. 306
E
Effusive : see extrusive.
Eh : oxidation-reduction potential.
Elastic : said of a body in which strains are instantly and totally
recoverable and in which deformation is independent of time.
Elastic behavior
Chapter 8
Chapter 8, p 303
Elastic limit:" the greatest stress that can be developed in a material
without permanent deformation remaining when the stress is
released" (Bates & Jackson, 1980).
Chapter 8, p. 304
Electrobed : corresponds to an interval of depth in which log
responses are nearly constant.
Electrofacies
Chapter 5, p. 206
Electrosequence
Chapter 5, p. 223
Eluvium :fine soil or sand moved and deposited by wind, as in a
sand dune.
EMI™ : Halliburton's trade mark for Electrical Micro Imaging.
End member : one of the two or more pure components of a mixtue.
Chapter 3, p. 132-133
Chapter 2, p. 48
Endogenetic : "derived from within; said of a geological process, or
of its resultant feature or rock, that originates within the Earth. The term
is also applied to chemical precipitates (evaporites) that originate within
the rocks that contain them" (Blates & Jackson, 1980).
Chapter 6
Endogenous : endogenetic.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Entropy : a measure of the degree of mixing of the different kinds of
rock components in a stratigraphic unit.
Environment : "a geographically restricted complex where a sediment
accumulates, described in geomorphic terms and characterized by
physical, chemical and biological conditions, influences or forces "
(Blates & Jackson, 1980).
Chpater 5, p. 228
Environmental effects : effects related to the influence of the
borehole on the measurements made by wireline tools.
Chapter 1
Eogenetic : a term proposed by Choquette & Pray (1970) for the
period of time between final deposition of a sediment and its burial below
the depth to which surface or near surface processes are effective.
Chapter 6, p. 245
Eolian : see aeolian.
Epidiagenesis
Chapter 6, p. 244
Epigenetic : said of a sedimentary mineral, texture, or structure
formed after the deposition of the sediment.
Chapter 6
Epsomite : a hydrous sulfate of magnesium.
Eruptive : said of a rock formed by the solidification of magma.
Esker : a long, narrow, sinuous, steep-sided ridge composed of irregularly
stratified sand and gravel that was deposited by a subglacial or
englacial stream flowing between ice walls or in an ice tunnel of a
stagnant or retreating glacier.
Estuary : the seaward end or the widened funnel-shaped tidal
mouth of a river valley where fresh water comes into contact with
seawater and where tidal effects are evident.
Eustasy : the worldwide sea-level regime and its fluctuations,
caused by absolute changes in the quantity of seawater.
Eustatism : syn. : eustasy.
Euxinic : pertaining to an environment of restricted circulation and
stagnant or anaerobic conditions.
Evaporite : a non clastic sedimentary rock composed primarily of
minerals produced from a saline solution.
Excavation effect : a decrease in the neutron log apparent porosity
reading below that expected on the basis of the hydrogen indices of
the formation components.
Exogenetic : said of processes originating at or near the surface of
the Earth, such as weathering and denudation, and to rocks and
landforms that owe their origin to such processes.
Chapter 6
Exogenous : exogenetic.
Extrusive : said of igneous rocks that has been erupted onto the
surface of the Earth.
This page has been reformatted by Knovel to provide easier navigation.
254
Index Terms
Links
F
Fabric : the orientation in space of the elements composing a
sedimentary rock.
Chapter 3
Facies
Chapter 5
Facies model :
FACIOLOG* : a Schlumberger mark for a program of facies
analysis.
Chapter 5, p. 211
Factor :
- cementation- : the porosity exponent m in Archie's formula.
- formation- : symbol F;
Failure : fracture or rupture of a rock that has been stressed beyond
its ultimate strength.
Chapter 9
FAST plot : a contraction for Formation Anomaly Simulation Trace,
a plot versus depth obtained by the intersection of dip planes with the
borehole considered as a cylinder in space. Dip presentation introduced
by Schlumberger.
Fault
Chapter 8, p. 301
311
Feldspar: a group of abundant rock-forming minerals of general formula
MAI(Al,Si)3O8 where M = K, Na, Ca, Be, Rb, Sr, and Fe. Feldspars
are the most widespread of any mineral group and constitute 60 % of the
Earth's crust. On decomposition, they yield a large part of the clays.
Feldspathic : said of a rock containing feldspar.
Felsic : a mnemonic adjective derived from feldspar + /inad (feldspathoid) + silica + c, and applied to an igneous rock having abundant
fight-colored minerals in its mode; also, applied to those minerals
(quartz, feldspars, feldspathoids, muscovite) as a group.
Ferromagnesian : containing iron and magnesium.
Ferruginous : pertaining to or containing iron.
FIL* : a Schlumberger mark for the Fracture Identification Log.
Chapter 9, p. 370
Fissure
Chapter 9, p. 355
Flank : limb.
Chapter 8, p. 309
Flaser : ripple cross-lamination in which mud streaks are preserved
in the troughs but incompletely or not at all on the crests.
Flexure : syn.: hinge.
Chapter 4, p.169
Chapter 8, p. 309
Fluvial : of or pertaining to a river.
FMI* : acronym for Formation Micro Imager
FMS* : acronyme for Formation MicroScanner tool.
Fold
Chapter 8, p. 301
Footwall : the underlying side of a fault or the wall rock beneath an
inclined fault.
Chapter 8, p. 312
This page has been reformatted by Knovel to provide easier navigation.
308
Index Terms
Links
Fore roof : the seaward side of a reef.
Foreset : pertaining to or forming a steep and advancing frontal
slope, or the sediments deposited on such a slope.
Chapter 4, p. 164
Formation : a general term applied in well logging to the external
environment of the drilled well bore without stratigraphic connotation.
Formation factor: expresses the relationship between the porosity
and the resistivity of the water formation: FR = R0/Rw = a/Φm.
Formation Micro Imager* (FMI): Schlumberger's tool.
Formation MicroScanner* tool (FMS) :Schlumberger's tool.
Chapter 3, p. 132-133
Chapter 3, p. 132
Formation water : water present in the virgin formation under natural
conditions, as opposed to introduced fluids such as mud filtrate.
Fossil : any remains, trace, or imprint of a plant or animal that has
been preserved in the Earth's crust since some past geologic time.
Fracture
Chapter 9
Funnel shape : an evolution of a curve (i.e. SP or resistivity )with
depth drawing the shape of a funnel.
Chapter 5, p. 203
G
Gabbro : a group of dark-colored basic intrusive igneous rocks.
Gash : a small scale tension fracture.
Chapter 9, p. 355
Geochemistry : the study of the distribution and amounts of the
chemical elements in minerals, rocks, ores, soils, waters, and the
atmosphere.
GeoColumn* : a Schlumberger mark for a program of automatic
lithofacies determination. Syn. : LITHO*.
Chapter 2, p. 94
GEODIP*: a Schlumberger mark for a dip computation program by
pattern recognition technique written for the HDT dipmeter tool.
GEOGRAM* : a Schlumberger mark for a program of synthetic
seismogram.
Geology : the study of the planet Earth.
Geometry : the three dimensional (length, width and thickness)
shape of a sedimentary body.
Geophysics : study of the Earth by quantitative physical methods.
Geostatic pressure : the vertical pressure at a point equal to the
pressure caused by the weight of the column of overlying rock.
Chapter 8, p. 301
Glacial : of or relating to the presence and activities of ice or
glaciers.
Glacier: a large mass of ice.
Glass : an amorphous product of the rapid cooling of a magma.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Glauconite : a dull-green earthy or granular mineral of the mica
group. It occurs abundantly in greensand, and seems to be forming in
the marine environment. It is an indicator of very slow sedimentation.
Gneiss : a foliated metamorphic rock.
Chapter 2, p. 82
Goethte : an hydrous oxide of iron, the commonest constituent of limonite.
Graben : an elongate, relatively depressed crustal unit or block that
is bounded by faults on its long sides.
Graded bedding : a gradual and progressive change in grain size.
Chapter 4, p. 169
Grain : a mineral or rock particle of all sizes more or less rounded.
Chapter 3, p. 120
Grain size : the general dimensions of grains or particles in a
sediment or rock.
Chapter 3, p. 120
Grainstone : [Dunham's classification] a term used for a mud-free,
grain-supported, carbonate sedimentary rock.
Chapter 3, p. 128
Granite : a plutonic acidic rock in which quartz constitutes 10 to 50
% of the felsic components.
Granulite : a metamorphic rock.
Chapter 2, p. 172
Chapter 2, p. 82
Granulometry : the measurement of grain size.
Gravel : a particle having a diameter in the range of 2-20 mm. Syn.
pebble.
Graywacke
Chapter 2, p. 38
Green pattern : a convention in dipmeter interpretation. It represents
a succession of dips of relative constant azimuth and magnitude.
Chapter 8, p. 332
Greensand : a sand having a greenish colour, consisting largely of
dark greenish grains of glauconite.
Growth fault : a fault in sedimentary rock that forms
contemporaneously and continuously with deposition.
Chapter 8, p. 314
Gypsum : hydrous calcium sulfate of the evaporite group.
H
Haematite: iron oxide: Fe2O3.
Halite : sodium chloride of the evaporite group. Syn. : salt.
Halmyrolysis : the geochemical reaction of sea water and
sediments in an area of little or no sedimentation.
Chapter 6, p. 244
Halokinesis : a general term for the study of the mechanism of salt
movement and related structures.
Hanging wall : the overlying side of a fault or the wall rock above a
fault.
Chapter 8, p. 312
Hard-ground : a zone at the sea bottom, usually a few cm thick, the
sediment of which is lithified to form a hardened surface, often
encrusted, bored, and solution-ridden. It implies a gap in sedimentation.
This page has been reformatted by Knovel to provide easier navigation.
141
Index Terms
Links
Hardness : the resistance of a mineral to scratching.
Heave : "in a fault, the horizontal component of deparation or
displacement" (Bates & Jackson, 1980).
Chapter 8, p. 312
Heavy mineral : a detrital mineral from a sedimentary rock, having
a specific gravity higher than a standard (usually 2.85), and commonly
forming as an accessory mineral (less than 1 % ).
Hemipelagic : deep-sea sediment in which more than 25 % the fraction
coarser than 5 μm is of terrigenous, volcanogenic, and/or neritic origin.
Heterogeneous : said of a bed showing numerous uncorrected
events on dipmeter or FMS resistivity curves.
Hinge : syn. : flexure.
Chapter 8, p. 310
Homogeneous : said of a bed without any resistivity variations on
the dipmeter or FMS resistivity curves.
Horizontal slip : in a fault, the horizontal component of the net slip.
Syn. : heave.
Chapter 8, p. 312
Hornblende : the commonest mineral of the amphibole group.
Horst : an elongate, relatively uplifted crustal unit or block that is
bounded by faults on its long sides.
Humic : pertaining to or derived from humus.
Humus : the generally dark, more or less stable part of the organic
matter of the soil.
Hydrostatic pressure : stress that is uniform in all directions. The
pressure exerted by the water at any given point in a body of water at
rest.
Chapter 7, p. 266
I
Iceberg : a large, massive piece of floating or stranded glacier ice of
any shape, detached from the front of a glacier into a body of water.
Igneous : said of a rock or mineral that solidified from molten or partly
molten material, i.e. from a magma.
Chapter 2, p. 37
Illte : a clay mineral containing less potassium and more water than
true mica.
Imbibition : the absorption of a fluid by a granular rock or any other
porous material under the force of capillary attraction in the absence of
any pressure.
Immature : said of a clastic sediment characterized by unstable
minerals (i.e. feldspars and plagioclases), abundance of mobile oxides
(i.e. alumina), presence of weatherable material (such as clay), and
poorly sorted and angular grains, indicating processes (i.e. transport and
weathering) acting over a short time and/or with a low intensity.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Injection : the forcing, under abnormal pressure, of sedimentary
material (downward, upward, or laterally) into a pre-existing deposit or
rock.
Intertidal : pertaining to the benthic ocean environment or depth
zone between high and low tide. Syn. : littoral.
Intrusion : the process of emplacement of magma in pre-existing
rock.
Intrusive : of or pertaining to intrusion, both the processes and the
rock so formed.
Invasion : the process by which the mud filtrate penetrates in a
porous rock.
Irreducible saturation : it corresponds to the minimum saturation of
a fluid when the fluid is displaced from a porous medium by another fluid
immiscible with the first.
Isobar : a line on a map connecting points of equal pressure.
Isobath : a line on a map connecting points of equal depth.
Isochore : a line on a map connecting points of equal drilled thickness.
Isochrone : a line on a map connecting points of equal travel time.
Isohypse : a line on a map connecting points of equal elevation.
Isolith : a line on a map connecting points of equal aggregate thickness of a given lithologic facies within a formation.
Isomorphic : having identical or similar form.
Isopach : a line on a map connecting points of equal true thickness.
Isopic : said of sedimentary rocks of the same facies.
Isopycnic : a line on a map connecting points of equal density.
Isorad : a line on a map connecting points of equal radioactivity.
Isostasy : the condition of equilibrium, comparable to floating, of the
units of the lithosphere above the asthenosphere.
Isotope : one of the two or more species of the same element
having the same number of protons in the nucleus but differing from one
another by the number of neutrons.
Isotropy : the condition of having properties that are uniform in all
directions.
J
Joint : a surface of fracture in a rock without displacement.
Chapter 9, p. 355
K
Kainite : a mineral of the evaporite group.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Kame : a low mound, knob, hummock, or short irregular ridge,
composed of stratified sand and gravel deposited by a subglacial
stream as a fan or delta at the margin of a melting glacier.
Kaolinite : a common clay mineral of the kaolin group, generally
derived from alteration of alkali feldspars and micas.
Karst : a type of topography formed on carbonate or gypsum rocks
by dissolution.
Kettle : a steep-sided, usually basin- or bowl-shaped hole or
depression in glacial-drift deposits.
Kieserite : a hydrous sulfate of magnesium of the evaporite group.
L
Labile : said of rocks and minerals that are mechanically and
chemically unstable.
Lacustrine : pertaining to, produced by, or formed in a lake.
Lamina : the thinnest recognizable unit layer.
Chapter 4, p. 160
Laminated : said of a rock that consists of laminae.
Lamination : the finest stratification or bedding.
Langbeinite : a sulfate of potassium and magnesium of the
evaporate group.
Layer : a general term for any tabular body of rock.
Leaching : selective removal of soluble minerals by throughgoing
water.
Chapter 6, p. 243
Lens : a geologic deposit bounded by converging surfaces.
Lenticular bedding : a form of interbedded mud and ripple
crosslaminated sand, in which the ripples or lenses are
discontinuous vertically and horizontally.
Chapter 4, p. 163
Lignite : a brownish-black organic rock that is intermediate between
peat and coal.
Limb : that area of a fold between adjacent fold hinges.
Limestone : a sedimentary rock consisting of more than 50 %
calcium carbonate.
Limnic : said of coal deposits formed inland in freshwater basins,
peat bogs, or swamps.
Limonite : a general term for a group of brown, amorphous naturally
occurring hydrous ferric oxides.
Listric : a curvilinear, concave spoon-shaped, usually pointing
upward, surface of fracture or fault, which becomes less steep as one
goes deeper, becoming nearly horizontal at some depth.
Chapter 8
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Lithification : the conversion of a newly deposited unconsolidated
sediment into a coherent solid rock.
LITHO* : see GeoColumn.
Lithofacies : a facies characterized by particular lithologic features.
Lithology : the description of rocks on the basis of colour,
mineralogic composition and texture (grain size).
Lithosphere : a layer of strength relative to the underlying asthenosphere.
It corresponds to the relatively rigid outer shell of the Earth
comprising the crust and upper mantle.
Lithostatic pressure
Chapter 7, p. 266
Littoral : syn. : intertidal.
Load : the material that is moved or carried by a stream, a glacier,
the wind, or waves, tides and currents.
Lobate delta : syn. : arcuate.
LOCDIP*: a Schlumberger mark for a dip computation program by
derivative technique written for the SHDT dipmeter tool.
Log : a continuous record of a parameter as a function of depth.
Longshore bar : a low, elongate sand ridge, built chiefly by wave
action, occurring at some distance from, and extending generally
parallel with, the shoreline.
Lutite : a general term used for consolidated rocks composed of silt
and/or clay.
M
Mafic : a mnemonic term derived from magnesium + ferric + is to
denote ferromagnesian minerals.
Magma : a naturally molten mass, formed within the crust or upper
mantle, which may solidify to form an igneous rock.
Mantle : this portion of the Earth's interior lying between the crust
and the core.
Marble : a metamorphosed carbonate (chiefly limestone).
Marker
Chapter 2, p. 82
Chapter 11, p. 420
Marl : an argillaceous limestone.
Marsh : a water-saturated, poorly drained area, intermittently or
permanently water-covered, having aquatic and grasslike
vegetation, without the formation of peat.
Massive : said of a rock that occurs in very thick homogeneous
beds.
Chapter 4, p. 164
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Matrix
- for a log analyst the solid framework of rock, except shale, which
surrounds pore volume.
- for a geologist the smaller or finer-grained material filling the
interstices between the coarser grains or particles of a sediment
or sedimentary rock.
Chapter 3
Mature : said of a clastic sediment characterized by stable minerals
(i.e. quartz), deficiency of the more mobile oxides (such as soda),
absence of weatherable minerals (such as clay), and well sorted but
subangular to angular grains, indicating processes acting over a long
time and with a high intensity.
Meander : one of a series of regular freely developing sinuous
curves, bends, loops, turns, or windings in the course of a stream.
Meandering stream : a stream having a pattern of successive
meanders.
Mechanical behavior
Chapter 8, p. 301
Mesogenetic
Chapter 6, p. 245
Metamorphism :the processes by which changes in solid rocks
under influence of heat, pressure and chemically active fluids.
Mica : a group of minerals pertaining to phyllosilicates of general
formula (K,Na,Ca) (Mg,Fe,Li,AI)2-3(Al,Si)4O10(OH,F)2
Micrite : a descriptive term for carbonate mud with crystals less than
4 microns in diameter.
Microcline : the triclinic form of potassium feldspar.
Migration : the movement of liquid and gaseous hydrocarbons fromtheir source rocks through permeable formations into reservoir rocks.
Mineral : a naturally occurring inorganic element or compound
having an orderly internal structure and characteristic chemical
composition, crystal form, and physical properties.
Chapter 2, p. 45
Mineralogy : the study of minerals.
Mixed-layer mineral : a mineral whose structure consists of
alternating layers of clays minerals and/or mica minerals.
Mode : the mode is the percentage (by weight) of the individual
minerals which make up a rock.
Modulus of elasticity : the ratio of stress to its corresponding strain
under given conditions of load for materials that deform elastically.
Mohr stress circle
Chapter 8, p. 303
Mohr stress envelope
Chapter 8, p. 303
Monogenetic : said of a conglomerate composed of a single type of
rock.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Montmorillonite : a group of expanding-lattice clay minerals. Syn.
smectite.
Moraine : an accumulation of material which has been transported
or deposited by ice.
Mouth : the place of discharge of a stream.
Mud cake : the residue deposited on the borehole wall as the mud
looses filtrate into porous, permeable formations.
Mud filtrate : the effluent of the continuous phase liquid of drilling
mud which penetrates (invades) porous and permeable formations.
Mudflow : a general term for a mass-movement landform and a process
characterized by a flowing mass of predominantly fine-grained
earth material possessing a high degree of fluidity during movement.
Mudstone
- an indurated mud having the texture and composition of shale.
- [Dunham' s classification] a term used for a mud-supported
carbonate sedimentary rock containing less than 10 % grains.
Chapter 3, p. 128
Muscovite : a white mineral of the mica group.
N
Nadir : the point on the celestial sphere that is directly beneath the
observer and directly opposite the zenith.
Chapter 8, p. 322
Natron : hydrous sodium carbonate occurring mainly in solution in
soda lakes or in saline residues.
Natural levee : a long broad low ridge or embankment of sand and
coarse silt, built by a stream on its flood plain and along both banks of
its channel.
Neogenesis : the formation of new minerals by diagenesis or
metamorphism.
Neomorphism : term suggested by Folk (1965) for all transformations
between one mineral and itself or a polymorph.
Chapter 6, p. 255
Neritic : Pertaining to the ocean environment or depth zone between
low-tide level and approximately the edge of the continental shelf.
Net : a stereographic or an equal-area projection of a sphere in
which the network of meridians and parallels forms a coordinate system.
Net slip : the distance between two formerly adjacent points on
either side of a fault, measured on the fault surface or parallel to it.
Chapter 8, p. 311
Nodule : a small, irregularly rounded knot, mass, or lump of a mineral
or mineral aggregate (i.e. pyritic nodules in a coal bed, chert nodules
in limestone, anhydritic nodules in limestone or dolomite, phosphatic
nodules in marine strata).
Chapter 6, p. 252
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Norm
- the theoretical mineral composition of a rock expressed in terms of
normative mineral molecules that have been determined by specific
chemical analyses for the purpose of classification and comparison.
- recognized type as reference.
Normal stress : that component of stress which is perpendicular to
a given plane.
Chapter 8, p. 301
Normative mineral : a mineral whose presence in a rock is
theoretically possible on the basis of certain chemical analysis.
O
Offlap :the progressive offshore regression of the updip terminations
of the sedimentary units within a conformable sequence of rocks.
Oligomictic : said of a clastic sedimentary rock composed of a
single rock type.
Olistostrome : a sedimentary deposit accumulated as a semifluid
body by submarine gravity sliding or slumping.
Olivine : a group of common rock forming minerals of basic, ultrabasic, low silica igneous rocks (gabbro, basalt) of formula
(Mg,Fe,Mn,Ca)2SiO4
Onlap : an overlap characterized by the regular and progressive pinching
out, toward the margins or shores of a depositional basin, of the
sedimentary units within a conformable sequence of rocks.
Oolite : a sedimentary rock, usually a limestone, made up chiefly of
ooliths.
Oolith : one of the small round or ovate accretionary bodies in a
sedimentary rock, having the size of a sand.
Opal : a mineral or mineral gel of the silica group, having a varying
proportion of water.
Orogeny : the process by which structures within fold-belt
mountainous areas were formed.
Orthoquartzite
Chapter 2
Orthose : the monoclinic form of the potassium feldspar. Syn.
orthoclase.
Outcrop : that part of a geologic formation or structure that appears
at the surface of the Earth.
Outwash : stratified detritus removed or a washed outs from a glacier
by meltwater streams and deposited in front of or beyond the end
moraine or the margin of an active glacier.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Overburden : the upper part of a sedimentary deposit compressing
and consolidating the material below.
Overburden pressure : syn. geostatic pressure.
Chapter 7, p. 267
Overlap : a general term referring to the extension of marine,
lacustrine, or terrestrial strata beyond underlying rocks.
Overlay : graphic data on a transparent sheet to be superimposed
on another sheet.
Overpressure : pressure in excess of lithostatic pressure, from
tectonic stress.
Oxbow : the abandoned crescent- or bow-shaped channel of a
former meander.
P
Packing :the manner of arrangement of the solid clastic particles in
a sediment or sedimentary rock.
Chapter 3, p. 122
Packstone : [Dunham's classification] a term used for a sedimentary
carbonate rock whose granular material is arranged in a
selfsupporting framework.
Chapter 3, p. 128
Paralic : said of coal deposits formed along the margin of the sea.
Particle : a general term for a separate or distinct unit in a rock
without restriction as size, shape, composition, or internal structure.
Peat: an unconsolidated deposit of semicarbonized plant remains in
a watersaturated environment such as bog or fen, and of persistently
high moisture content.
Pebble : a general term for a small roundish rock fragment having a
diameter in the range of 4-64 mm.
Pelagic : said of marine organisms whose environment is the open
ocean, rather than the bottom or shore areas.
Pelite : syn.: lutite.
Pellet: a silt or sand-sized aggregation of carbonate mud, generally
fecal in origin.
Permeability : the property of a porous rock for transmitting fluid.
pH : the negative logo of the hydrogen-ion activity in a solution; a
measure of the acidity or basicity of a solution.
Pitch : the angle between the horizontal and any linear feature.
Chapter 8, p. 312
Plagioclase : a group of triclinic feldspars of general formula
(Na,Ca)Al(Si,Al)Si2O8, which form a complete solid-solution series from
albite (pure Na) to anorthite (pure Ca) .They are among the commonest
rock-forming minerals of igneous rocks.
Planktonic : said of that type of pelagic organism which floats.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Plastic behavior
Chapter 8, p.305
Platform :that part of a continent that is covered by flat-lying or gently
tilted strata, mainly sedimentary.
Playa lake : a shallow, intermittent lake in an arid or semiarid region.
Plunge : the inclination of a fold axis or other linear structure,
measured in the vertical plane.
Plutonic : pertaining to igneous rocks formed at great depth.
Chapter 8, p. 310
Chapter 2, p. 37
Point bar : one of a series of low, arcuate ridges of sand and gravel
developed on the inside of a growing meander by the slow addition of
individual accretions accompagnying migration of the channel toward
the outer bank.
Poisson's ratio
Chapter 8, p. 305
Polar plot
Chapter 8, p. 320
Polyhalite : a mineral of the evaporite group.
Polymictic : said of a clastic sedimentary rock composed of
manyrock types such as arkose, graywacke, conglomerate.
Pore : a small to minute opening or passageway in a rock. Syn.
interstice.
Pore-bridging
Chapter 3, p. 126
Pore-fillingChapter 3, p. 126
Pore-lining
Chapter 3, p. 126
Porogenesis
Chapter 6, p. 245
Poronecrosis
Chapter 6, p. 245
Porosity : the percentage of the bulk volume of a rock that is
occupied by interstices, whether isolated or connected.
Chapters 1
3
Pressure : the force exerted across a real or imaginary surface
divided by the area of that surface.
Chapter 8
Principal axis of stress
Chapter 8, p. 302
Principal plane of stress
Chapter 8, p. 302
Proximal : said of a sedimentary deposit formed nearest the source
area.
Pyroclastic : pertaining to clastic rock material formed by volcanic
explosion.
Pyroxene : a group of dark rock-forming minerals having the general
formula ABSixOe, where A = Ca, Na, Mg, or Fe2+, and B = Mg, Fe2+,
Fe3+, Fe, Cr, Mn, or Al. They constitute a common constituent of
igneous rocks.
This page has been reformatted by Knovel to provide easier navigation.
2
Index Terms
Links
Q
Quartz : crystalline silica, an important rock forming mineral. It is,
next to feldspar, the commonest mineral and has a widespread
distribution in igneous, metamorphic and sedimentary rocks.
Quartzarenite : a sandstone that is composed primarily of quartz
(more than 95 %).
Chapter 2
Quartzite : a very hard but not metamorphosed sandstone, consisting
chiefly of quartz grains that have been completely cemented.
Quartzose : containing quartz as a principal constituent.
Quartz wacke : a moderately well-sorted, commonly fine-grained
sandstone containing up to 90 % quartz and chert, and with more than
10 % argillaceous matrix, less than 10 % feldspar, and less than 10 %
rock fragments.
Quick Look : a general term for a rapid survey of logs.
R
Radiolarite : a comparatively hard fine-grained chertlike homogeneous
consolidated rock composed predominantly of the remains of
Radiolaria.
Recrystallization : the formation, essentially in the solid state, of
new crystalline mineral grains in a rock.
Recumbent fold : an overturn fold.
Chapter 6, p. 242
Chapter 8, p. 311
Redoxomorphism : a diagenetic phenomenum characterized by
mineral changes primarily due to oxidation and reduction. It is typical of
unlithified sediments.
"Red pattern" : a convention used in dipmeter interpretation to
denote decreasing formation dip with decreasing depth with a near
constant azimuth.
Chapter 8, p. 330
Reef : a ridgelike or moundlike structure, layered or massive, built
by sedentary calcareous organisms.
Regression : the retreat or contraction of the sea from land areas.
Repeat section : a short section of a log that is recorded in addition
to the main survey section in order to provide an inter-run comparison of
log similarity, and, therefore, instrument stability and repeatability.
Chapter 1, p. 18
Reservoir rock : a porous and permeable rock.
Rhyolite : a group of extrusive igneous rocks.
Rhythm
Chapter 5, p. 198
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Rhythmic sedimentation : the consistent repetition of a regular
sequence of two or more sedimentation units organized in a particular
order and indicating a frequent and predictable recurrence or pattern of
the same sequence of conditions.
Chapter 5, p. 198
Rift : a long, narrow continental trough that is bounded by normal
faults.
Rigidity : the property of a material to resist applied stress that
would tend to distort it.
Chapter 8, p. 305
Ripple mark : an undulatory surface consisting of alternating subparallel small-scale ridges and hollows formed at the interface between
a fluid and incoherent sedimentary material.
Chapter 4, p. 164-168
Rotation : angle formed by both parts of the same bed after separation
and measured on a plane perpendicular to the axis of rotation.
Chapter 8, p. 312
Roundness : the degree of abrasion of a clastic particle as shown
by the sharpness of its edges and corners.
Chapter 3, p. 122
Rudite : a general term used for consolidated sedimentary rocks
composed of rounded and angular fragments coarser than sand.
Rupture point
Chapter 9
Rupture strength : the differential stress that a material sustains at
the instant of rupture.
S
Sabkha : a supratidal environment of sedimentation, formed under
arid or semiarid conditions.
Salt : syn. : halite.
Saltation : a mode of sediment transport in which the particles are
moved forward in a series of short intermittent jumps.
Salt dome : syn.: diapir.
Sand
- a rock fragment or detrital particle having a diameter in the range
of 1/16 to 2 mm.
Chapter 3
- a loose aggregate of unlithified mineral or rock particle of sand
size.
Sandstone : a lithified, consolidated sand.
Sapropel : an unconsolidated, jellylike ooze or sludge composed of
plant remains, mostly algae, macerating and putrefying in an anaerobic
environment on the shallow bottom of lakes and seas. It may be a
source material for hydrocarbons.
Sapropelic : pertaining to or derived from sapropel, indicating a high
sulfate or reducing environment.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
SARABAND* : a Schlumberger mark for a program of interpretation
of shaly sand.
Saturation : the percentage of the pore volume occupied by a
specific fluid.
SCAT plot : acronym for Statistical Curvature Analysis Techniques
intorduce by Bengtson, 1981.
Chapter 8, p. 332
Schist : a strongly foliated crystalline rock, formed by dynamic
metamorphism.
Schmidt stereonet
Chapter 8, p. 324
Scour and fill : a process of aternate excavation and refilling of a
channel, as by a stream or the tides.
Sedimentary rock
Chapter 4
Chapter 2, p. 38
Sedimentary structure : a structure in a sedimentary rock.
Sedimentation unit :
Chapter 4
Chapters 1
4
Chapter 5, p. 198
Chapter 11, p. 410
Seif : a very large, sharp-crested tapering longitudinal dune.
Seismogram : a seismic record.
Sequence
Shale : a fine-grained detrital sedimentary rock, formed by the
consolidation of clay, silt, or mud. It is characterized by finely laminated
structure, and by an appreciable content of clay minerals and detrital
quartz.
Shear modulus :syn. modulus of rigidity.
Chapter 8, p. 305
Shear strength : the internal resistance of a body to shear stress.
Chapter 9, p. 359
Shear stress : that component of stress which acts tangential to a
plane through any given point in a body.
Chapter 8, p. 301
Shelf : a stable cratonic area that was periodically flooded by
shallow marine waters and received a relatively thin,
well-winnowed cover of sediment.
Siderite : carbonate of iron.
Sill: a tabular igneous intrusion that parallels the planar structure of
the surrounding rock.
Silt
- a rock fragment or detrital particle having a diameter in the range
of 1/256 to 1/16 mm.
- a loose aggregate of unlithified mineral or rock particles of silt size.
Siltstone : an indurated silt.
Sinistral: pertaining to the left.
Slate : a metamorphic rock.
Chapter 9, p. 355
Chapter 2, p. 38
Slickenside : a polished and smoothly striated surface that results
from friction along a fault plane.
Chapter 8, p. 312
This page has been reformatted by Knovel to provide easier navigation.
82
Index Terms
Links
Slip : the relative displacement of formerly adjacent points on
opposite sides of a fault.
Chapter 8, p. 312
Slump : a landslide characterized by a shearing and rotary
movement of a generally independent mass of rock.
Chapter 4, p. 171
SODA plot: acronym for Separation Of Dip and Azimuth.
Chapter 8, p. 333
Solution : a process of chemical weathering by which mineral and
rock material passes into solution.
Sorting : the spread or range of particle-size distribution.
Source rock
Chapter 3 p. 121
Chapter 10
- the parent rock from which other sediments or rocks are derived.
- sedimentary rock in which organic material was transformed in
hydrocarbons under pressure, heat and time influences.
Sparite : a descriptive term for clean, relatively coarse-grained
calcite accumulated during deposition or introduced later as a cement.
Sphericity : the relation to each other of the various diameters of a
particle.
Chapter 3, p. 122
Stack : the sum of several seismic traces or log runs.
State of stress
Chapter 8, p. 302
Stereogram :a graphic diagram on a plane surface, giving a threedimensional representation.
Chapter 8, p. 321-329
Stereographic projection
Chapter 8, p. 322
Stereonet
Chapter 8, p . 320
Stick plot
Chapter 8, p. 345
Stoneley wave : a type of guided wave.
Chapter 9, p. 367
Strain : change in the shape or volume of a body as a result of
stress.
Chapter 8, p. 303
308
Stratification : the formation, accumulation, or deposition of
material in layers.
Chapter 4
Stratigraphy: the science of rock strata. It is concerned with all
characters and attributes of rocks (succession, age, form, distribution,
composition, fossil content, geophysical and geochemical properties.
STRATIM* : a presentation of the borehole image obtained from the
SHDT dipmeter data.
Chapter 8
Stratum : a tabular or sheetlike body or layer of sedimentary rock.
Strength : the ability to withstand differential stress.
Chapter 8, p. 303
Chapter 9, p. 359
Stress : the force per unit area acting on any surface within it.
Chapter 8, p. 301
Stress components
Chapter 8, p. 302
Stress difference : the difference between the greatest and the
least of the three principal stresses.
Chapter 9, p. 357
This page has been reformatted by Knovel to provide easier navigation.
307
Index Terms
Links
Stress ellipsoid : a geometric representation of the state of stress
at a point.
Chapter 8, p. 302
Stress field : the state of stress.
Chapter 8, p. 302
Stress-strain diagram
Chapter 8, p. 304
Stress-strain ratio
Chapter 8, p. 305
Chapter 9, p. 357
Stretch : the measure of the change in length of a line.
Strike : the direction or trend taken by a structural surface (bedding
or fault plane) as it intersects the horizontal.
Strike slip
Chapter 8
Chapter 8, p. 311
Structural geology : the branch of geology that deals with the form,
arrangement, and internal structure of the rocks.
Chapter 8
Structure
- a megascopic feature of a rock mass or rock unit.
Chapter 4
- the general disposition, attitude, arrangement, or relative positions
of the rock masses of a region or area.
Chapter 8
Stylolite : a surface or contact marked by an irregular and
interlocking penetration of the two sides.
Stylolitization : a pressure-controlled solution phenomenon.
Chapter 6, p. 243
Chapter 6, p. 243
257
Chapter 8, p. 301
Chapter 9
Subsidence : the sudden sinking or gradual downward settling of
the Earth's surface with little or no horizontal motion.
Subtidal : below low tide level.
Supermature : said of a mature clastic sediment whose well-sorted
grains have become subrounded to well-rounded.
Supratidal: above high tide level.
Surface force
Suspension : a mode of sediment transport in which the upward
currents in eddies of turbulent flow are capable of supporting the weight
of sediment particles and keeping them indefinitely held in the
surrounding fluid.
Swamp : an area intermittently or permanently covered with water,
having shrubs and trees but essentially without the accumulation of
peat.
Sylvite : potassium chloride of the evaporite group.
Syncline : a concave upward fold.
Chapter 8, p. 310
Syndiagenesis
Chapter 6, p. 244
SYNDIP*
Chapters 3 and 4
Synform : a fold whose limbs close downward.
Chapter 8, p. 310
Syngenetic : said of a primary sedimentary structure formed
contemporaneously with the deposition of the sediment.
Chapter 4, p.159
Synsedimentary : accompanying sedimentation.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
T
Tectonics : a branch of geology dealing with the broad architecture
of the outer part of the Earth.
Chapter 8
Telogenetic
Chapter 6, p. 245
Tensil strength
Chapter 9, p. 359
Tension : a state of stress in which tensile stresses predominate.
Chapter 8, p. 301
Terrigenous : derived from the land or continent.
Texture : the general physical appearance or character of a rock.
Chapter 2
Throw : the amount of vertical displacement on a fault.
Chapter 8, p. 311
Thrust : an overriding movement of one crustal unit over another.
Chapter 8, p. 312
Tidal flat : an extensive, nearly horizontal, marshy or barren tract of
land that is alternately covered and uncovered by the tide.
Till : dominantly unsorted and unstratified drift.
Toe : the lowest part of a slope.
Toeset : the forward part of a tangential foreset bed.
Topset : one of the nearly horizontal layers of sediments deposited
on the top surface of an advancing delta and continuous with the
landward alluvial plain.
Torsion
Chapter 8, p. 301
Tortuosity factor : syn. : cementation factor.
Traction
- the stress vector acting across a particular plane in a body.
- a mode of sediment transport in which the particles are swept
along and parallel to a bottom surface by rolling, sliding, dragging,
pushing or saltation.
Transgression : the spread or extension of the sea over land areas.
Transportation : a phase of sedimentation that includes the movement
by natural agents of sediment, either as solid particles or in solution,
from one place to another.
Trap : any barrier to the upward movement of hydrocarbons
allowing them to accumulate.
Chapter 8
Trona : a bicarbonate of sodium ourring in saline residues.
Trough
Chapter 8
Trough cross-bedding : cross-bedding in which the lower
bounding surfaces are curved surfaces of erosion; it results
from local scour and subsequent deposition.
Chapter 4, p. 168
Tuff : a general term for all consolidated pyroclastic rocks.
Turbidite : a sediment or rock deposited from a turbidity current.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
U
Unconformable : said of strata or stratification exhibiting the
relation of unconformity to the older underlying rocks.
Unconformity : a substantial break or gap in the geologic record.
Chapter 11, p. 405
Uniformitarianism : the fundamental principle or doctrine that geologic
processes and natural laws now operating to modify the Earth's
crust have acted in the same regular manner and with essentially the
same intensity throughout geologic time, and that past geologic events
can be explained by phenomena and forces observable today.
Introduction.
Unstable : said of a constituent of a sedimentary rock or mineral that
does not resist further mineralogic change under weathering.
Upthrow :
- the amount of upward vertical displacement of a fault.
- the upthrown side of a fault.
Chapter 8, p. 312
V
Vadose zone : zone of aeration.
Chapter 6
Varve : a glaciolacustrine layer seasonally deposited in a glacial
lake.
Vertical resolution : the minimum thickness of formation that can
be distinguished and fully characterized by a tool under operating
conditions.
Viscosity : the property of a substance to offer internal resistance to
flow.
Viscous behaviour
Chapter 8, p. 305
VOLAN* : a Schlumberger mark for a program of interpretation of
shaly sand.
Volcanic : pertaining to the activities, structures, or rock types of a
volcano.
Vug : a small cavity in a rock.
Chapter 6, p. 243
W
Wacke : a durty sandstone containing more than 10 % argillaceous
matrix.
Wackestone : [Dunham's classification] a term used for a mudsupported
carbonate sedimentary rock containing more than 10 % grains.
Wadi: the channel in an arid region that is usually dry except during
the rainy season.
This page has been reformatted by Knovel to provide easier navigation.
Index Terms
Links
Washing : the selective sorting, and removal, of fine-grained
sediments by water currents.
Wavy bedding : bedding characterized by undulatory bounding
surfaces.
Chapter 4, p. 169
Weathering : the destructive processes (physical disintegration and
chemical decomposition) which transform earthy and rock materials on
exposure to atmospheric agents, and prepare sediments for transportation.
Wettability : the ability of a liquid to form a coherent film on a surface
owing to the dominance of molecular attraction between the liquid
and the surface over the cohesive force of the liquid itself.
Winnowing : the selective sorting, or removal, of fine particles by
wind action, leaving the coarser grains behind.
Wireline : a general term for any flexible steel line or cable connecting
a surface winch to a tool assembly lowered in a well bore.
Wulff stereonet
Chapter 8, p. 320
Y
Young's modulus : a modulus of elasticity in tension or compression,
involving a change of length.
Chapter 8, p. 305
Z
Zenith : the point on the celestial sphere that is directly above the
observer and directly opposite to the nadir.
Zeolite : a generic term for a large group of hydrous aluminosilicates.
Zircon : the silicate of zirconium. A common accessory mineral.
References
BATES, R.L., & JACKSON, J.A. (1980). - Glossary of Geology.
Amer. Geol. Inst., Falls Church, Virginia.
SHERIFF, R.E. (1973). - Encyclopedic Dictionary of Exploration
Geophysics. Soc. Explor. Geophysicists, Tulsa, Oklahoma, U.S.A.
Society of Professional Well Log Analysts (1975). - Glossary of
terms and expressions used in well logging.
WHITTEN, D.G.A., & BROOKS, J.R.V. (1972). – Dictionary of
Geology. Penguin Books Ltd., Harmondworth, Middlesex, England.
This page has been reformatted by Knovel to provide easier navigation.
324
