Stainless Steel Refining

Published on December 2017 | Categories: Documents | Downloads: 51 | Comments: 0 | Views: 542
of 19
Download PDF   Embed   Report

Comments

Content


1
Stainless Steel Refining
Richard J . Choulet
Consultant to Praxair
(Presented at AISE Seminar on June 5, 1997, in Detroit
Copyright © 1999, Praxair Technology, Inc.
All rights reserved
Introduction
Stainless steels were invented by Krupp Stahl in
1912. Stainless steels typically contain between 9
and 30 percent chromium. Varying amounts of
nickel, molybdenum, copper, sulfur, titanium,
niobium, etc., may be added to obtain the desired
mechanical properties and service life.
All stainless steels are corrosion resistant and
provide wide ranges of strength, formability, and
high or low temperature service. Stainless steels
are primarily classified as austenitic, ferritic,
martensitic, duplex, or precipitation hardening
grades. Typical wrought alloy AISI series
designations include 200 (high manganese
austenitic), 300 (austenitic), and 400 (ferritic or
martensitic). Martensitic and ferritic steels are
magnetic. Martensitic steels are typically hardened
by heat treatment and are not easily formable.
Austenitic steels harden when cold worked.
Duplex grades (austenitic/ferritic) are more resistant
to stress corrosion cracking than austenitic and are
tougher than ferritic grades. Precipitation hardened
grades (martensitic or austenitic) are strengthened
during heat treatment by precipitation hardening.
Growth In Stainless Steel Production
Annual Western World production of stainless
steels did not reach one million tons until the early
l950s. However, since that time, annual production
has increased dramatically as indicated in Figure 1.
In recent years, stainless production has sustained
an average annual growth rate of about 5 percent
and should continue at this rate through the year
2000. While the annual growth rate in North
America and Europe is projected at only 3 percent,
the rate in developed countries in Asia (Japan,
Korea, and Taiwan) should continue at 5 percent,
and the rate in developing countries (including China
and other Asian countries, as well as South Africa
and Brazil) could be as much as 15 percent.
1950 ‘55 ‘60 ‘65 ‘70 ‘75 ‘80 ‘85 ‘90 ‘95 2000
0
2
4
6
8
10
12
14
16
18
20
Year
Stainless
Production
(millions of tons)
- Europe
- N. America
- Asia
- Dev. Countries
Figure 1: Growth of Stainless Steel Production in the
Western World
Table I indicates the l996 estimated Western
World stainless production figures in thousands of
metric tons slab or ingot. While reliable figures are
very difficult to obtain, Table II shows rough
estimates of production figures for Eastern Europe
and the CIS during l996, and for comparison,
1990. Clearly, the political and economic
disruption in the early 1990s has resulted in
significant decreases in production.
2
Table I
1996 Estimate of Western World Stainless
Production
(Thousands of Metric Tons)
Country Slab or Ingot Tons
Africa South Africa 315
Americas Brazil 200
Canada 200
United States 1,920
Total 2,320
Asia China 320
India 455
Japan 3,870
South Korea 890
Taiwan 650
Total 6,185
Europe Belgium 580
Finland 450
France 985
Germany 1,290
Italy 890
Spain 840
Sweden 585
United Kingdom 560
Total 6,180
Western World Total 15,000
Table II
Estimated Stainless Production in Eastern Europe
and CIS
(Thousands of Metric Tons)
Eastern Europe CIS
1990 220 1,800
1996 120 390
The growth rate of ferritic steels (particularly
AISI 409L and 439L) used in the manufacture of
automotive exhaust systems was particularly healthy
in the early 1990s. Table III indicates the estimated
1995 production for ultra-low carbon and nitrogen
ferritic grades. Demand for these grades has
probably leveled off in the developed world;
however, demand will continue to grow rapidly in
the developing world. Despite the growth in ferritic
grades, overall production of tonnage stainless
steels remains roughly split at 70 percent austenitic
and 30 percent ferritic grades.
Table III
Estimated 1995 Production of ULC and ULN
Ferritic Grades
(Metric Tons Slab)
AISI 409L AISI 439L
Americas
Canada Atlas Tracy 6,000 ---
United
States
Alleghany
Ludlum
86,000 4,500
Armco (all
plates)
276,000 92,000
Jones &
Laughlin
25,400 1,500
Nucor 16,000 ---
Asia
Japan Kawasaki
Chiba
31,000 12,000
Nippon Steel
Yawata
67,000 28,000
Nisshin
Shunan
28,000 7,000
Sumitomo
Wakayama
30,000 7,000
Korea Posco 3,000 ---
Europe
Finland Outokumpu 5,000 ---
France Ugine
Isbergues
20,000 6,000
Ugine
L’Ardoise
55,000 59,000
Germany Krupp
Bochum
60,000 1,500
Italy AST Terni 60,000 6,000
Spain Acerinox 28,000 2,500
United
Kingdom
Avesta
Sheffield
17,000 5,000
Total 821,400 226,000
* Includes similar grades (e.g. 436L). Tonnage under
409L does not include 3CR12 estimate at 65,000 tons
total (of which Columbus makes 30,000). Japanese
companies make significant amounts of ULC and ULN
ferritic grades for non-automotive use (e.g., Nisshin,
74,000; Kawasaki, 60,000; Nippon Metals, 25,000;
Nippon Steel, 10,000 tons).
3
Demand for stainless steel appears to be
growing at an average worldwide rate of 5 to 7
percent. It is growing much faster in Asia (aside
from Japan) and slower in Europe and the United
States. Perceived continued growth in demand and
favorable investment returns have encouraged the
modification and expansion of existing facilities
and/or the installation of new facilities. Typically,
new melting capacity exceeds demand near term.
For example, estimated worldwide production in
l996 was 15 million tons and approximately equal
to demand. Estimated worldwide melting capacity
at the end of l996 was about 19.7 million tons.
Process Development
Prior to the late l940s, stainless steel was made
in the electric arc furnace by melting carbon steel
scrap, iron ore and burnt lime; slagging off; adding
burnt lime, ferrosilicon and fluorspar; increasing the
temperature; and adding low carbon ferrochrome to
attain the specified chromium. In this process
(known as the "Rustless Process"), most stainless
grades were made to an aim carbon specification of
0.08 percent. During this period, the only practical
and economic way to use stainless scrap was to
remelt it in induction furnaces. With the introduction
of tonnage oxygen in the steel industry in the late
l940s, a new arc furnace practice evolved in which
stainless scrap, high carbon ferrochrome, nickel,
and lime were melted, the bath blown with oxygen
to temperatures between 1850-1950°C, reduced
with silicon and aluminum, and trimmed with low
carbon ferrochrome. Relative to the “Rustless
Process,” this practice resulted in significant
decreases in power consumption, process time, and
the use of expensive low carbon ferrochrome as
well as improved quality (decreased hydrogen
contents) and increased chromium recoveries.
In l954, W. Krivsky of Union Carbide was
studying carbon-chrome-temperature relationships
in ferrochromium melting in the laboratory. The
experiments involved blowing oxygen onto the
surface of molten chromium alloy baths under
isothermal conditions. In order to control
temperature, he added argon and found that he
could decarburize the melt to low carbon levels
without excessive chromium oxidation. Krivsky's
observation led to plant-scale experiments injecting
argon/oxygen mixtures by lance into the arc furnace
between l958 and l962. Eventually, it was
concluded that a separate refining vessel (duplex
process) was necessary to develop a commercial
process, and the first successful AOD (argon
oxygen decarburization) heat was made in October
1967 in a modified 15-ton ladle. These trials
resulted in the first commercial AOD installation at
Joslyn (now Slater Steel) in July l968.
During the late l950s, vacuum degassing
processes (DH, RH, ASEA-SKF, Finkl VAD,
etc.) were developed for carbon steel production,
and led to the development of vacuum
decarburization processes for stainless steels in the
mid-1960s. The VOD (vacuum oxygen
decarburization) process was developed by Witten
(now Thyssen) in Germany between 1962 and
1967. Between 1959 and 1962, Witten had
produced stainless steels in their LD converter
(decarburization of pig iron, addition of low carbon
ferrochrome, and reduction with silicon and
aluminum). With the installation of their ladle
vacuum degassing unit in 1962, they began trials
duplexing premelt from the LD to the vacuum
degasser. Initially, iron ore was used for
decarburization. Later, premelting took place in the
arc furnace, and oxygen was top blown onto the
bath surface in the degassing ladle.
The AVR process, in which oxygen was
injected below the surface by top lance in the
degassing ladle, was developed by Allegheny
Ludlum in the mid-1960s. In recent years,
Allegheny has used their TMBI (top mixed bottom
inert) process metal supplied from their induction
furnaces when stainless demand exceeded AOD
capacity.
The development of the AOD and VOD
processes between 1954 and l968 revolutionized
stainless steelmaking and was the primary impetus
for the dramatic growth in production between
1970 and the present. The use of these or similar
processes resulted in significant decreases in raw
4
material costs, increases in productivity, and
improved quality.
Figure 2 is a schematic that illustrates the
various types of secondary steelmaking processes
used to produce stainless steel. Figure 3 illustrates
the breakdown of stainless steel production by
process for 1983 and l996. The AOD is the
dominant process, accounting for 68.7 percent of
Western World production in l996. Details of
various processes are discussed in the following
sections.
(a) AOD
(b) RH-OB/KTB (c) VOD (d) AOD-VCR
(e) K-BOP/K-OBM-S (f) CLU (g) ASM/MRP
Side or
Bottom
Blow
Plug or Tuyeres
Figure 2: Secondary Steelmaking Processes for
Stainless Production
AOD 68.7%
Other
Converter 5%
Converter/VOD
19.5%
AOD 71.8%
Converter/VOD
10.4%
VOD
9.2%
Other Converter 4.4%
EAF 4.2%
1983 Total: 6,550,000 tons
1996 Total: 15,000,000 tons
VOD
6.8%
Figure 3: Stainless Production by Process
Secondary Steelmaking Processes
All secondary steelmaking processes share
similar characteristics including: theoretical
concepts, raw material savings, productivity
increases and quality improvements, and specialized
equipment and vessel operation. An increasing
number of converter process operators use post-
converter treatment facilities to minimize converter
operating costs and tap-to-tap times. Minimizing
converter tap-to-tap time is critical for optimization
of sequence casting.
Theory
It is generally accepted that when oxygen is
injected into a stainless steel bath, chromium and
iron is oxidized. Decarburization occurs when
dissolved carbon reduces the chromium and iron
oxides form. The overall reaction and its
equilibrium constant is:
CO(g) Cr C (s) O Cr
4
3
4 3
4
1
+ = +
) aC )( O aCr (
) Pco )( aCr (
k
4 3
4
1
4
3
=
Decarburization occurs on the surface of rising
bubbles that form either from the injection of inert
gas, or on the surface of chromium oxide particles
being reduced and generating CO. A minimum
bath depth or bubble residence time is required for
the bubbles to become saturated with CO. Three
techniques are used to minimize chromium
oxidation. In addition to thermodynamic
considerations, kinetic considerations (degree of
slag and metal mixing) are critical in all three
approaches described below:
1. In the early arc furnace stainless practice, high
temperatures (about 1900°C) were used to
make decarburization more favorable. Figure 4
shows this relationship. However, such high
temperatures led to operational difficulties and
high refractory costs.
5
EQ @ 1705
0
C
& 1.0 atmCO
EQ @ 1815
0
C
& 1.0 atm CO
EQ @ 1705
0
C
& 0.1 atm CO
EQ @ 1815
0
C
& 0.1 atm CO
0 4 8 12 16 20
Chromium (%)
0.002
0.005
0.01
0.02
0.05
0.1
0.2
0.5
Carbon
(%)
Figure 4: Carbon-Chromium Equilibrium Curves
2. The AOD and similar duplex converter
processes use the dilution principle to minimize
chromium oxidation. The injection of inert gas
(argon or nitrogen) lowers the partial pressure
of CO in the bath, thus allowing higher
chromium contents to be in equilibrium with low
carbon contents. Figure 5 illustrates the results
of tests using several dilution ratios. Note the
distinct break points in the decarburization
reaction indicating a rapid decrease in carbon
removal efficiency. These points are related to
equilibrium carbon monoxide pressures and
indicate that a mixture containing higher
amounts of inert gas should be used in order to
reach lower carbon contents without excessive
chrome oxidation.
Vacuum processes also result in lower partial
pressure of CO, allowing higher chromium
contents to be in equilibrium with low carbon
contents. This approach is particularly effective
when carbon contents are below 0.04 percent.
While dilution processes can also obtain very
low carbon levels, relatively large amounts of
inert gas are required.
100
48
17
4.6
pct. O
2
in
Feed Gas
0 4 8 12 16 20
Chromium (%)
0
0.004
0.01
0.04
0.1
0.4
1.0
Carbon
(%)
Figure 5: Effect of Dilution Ration on Carbon Content
3. Stainless steel slags participate in
decarburization and other reactions. Any
chromium oxide that is not reduced by carbon
ends up in the slag forming a complex spinel.
The reduction of oxides (chromium, iron,
manganese, etc.) with silicon or aluminum to
optimize recovery is dependent on many factors
including: slag composition and basicity,
temperature, vessel mixing conditions, and
dissolution kinetics.
Raw Materials Savings, Productivity
Increases, and Quality Improvements
To varying degrees, all secondary stainless steel
refining processes result in:
1. The use of least cost charge materials. The
cost of materials amounts to as much as 85
percent of total stainless melting and refining
operating costs. Significant increases in
alloy recoveries and product yields result in
decreased total operating costs that include
the melting, refining, casting, rolling, and
finishing steps.
2. The utilization of the arc furnace primarily as
an efficient melting unit, which reduces arc
furnace charge-to-tap time by as much as
50 percent (increasing productivity up to
6
100 percent). Power, electrode, and
refractory consumptions are also reduced.
3. Significant quality improvements (surface
and internal cleanliness and mechanical
properties) associated with low levels of
sulfur, dissolved gases (oxygen, nitrogen,
and hydrogen), carbon, and volatile tramp
element (Zn, Pb, and Bi) contents.
All three of the above benefits are enhanced by
the ability of secondary steelmaking processes to
achieve very tight control of chemical composition
and temperature.
Equipment
Vessel Design
Figure 2 is a schematic illustration of the various
types of secondary steelmaking equipment.
Converter vessels typically are refractory-lined
cylindrical shells with concentric or sliced conical
sections at the top of the barrel section. They are
sometimes described as pear-shaped, and have
specific volumes (internal volume in cubic
meters/metric tons bath weight) ranging from 0.40
to 0.80. They are typically top and bottom blown
or top and side blown. In larger vessels, top
blowing is carried out with water-cooled lances
positioned from 1.5 to 4 meters off the bath surface
and with single or multiple nozzles capable of flows
ranging from subsonic to supersonic (the latter when
using a converging diverging nozzle). Top oxygen
reacting with the bath can range from 30 to 100
percent of that blown, depending on lance height,
nozzle design, and the gas pressure used in order to
obtain the amount of post combustion (burning of
CO to CO
2
in the headspace of the converter)
and/or decarburization rate desired.
Side blown converters use annular tuyeres (two
to seven), and bottom blowing ones use either
annular tuyeres (two to four) or porous plugs. (The
latter are used when only inert gases are blown
from the bottom.) Cooling gases, required to form
a metal or oxide accretion (called a knurdle or a
mushroom) at the tuyere tip to protect the tuyere
and surrounding refractory, are blown through the
outer annulus (shroud), and process gas
(oxygen/inert mixtures) through the inner annulus.
Figure 6 is a schematic of a patented converter
tuyere. The vessels normally rotate 360 degrees
and the drive systems are similar to those used in
LD (BOF) vessels. Side blown converters tend to
use removable shells as opposed to stationary
vessels.
Figure 6: Schematic of AOD Tuyere
Vacuum stainless decarburization processes are
typically carried out in a ladle, which in the case of
the VOD, has one to three porous plugs in the
bottom, a splash cover, and is enclosed in a tank.
The RH-OB or KTB processes involve “lifting”
metal from a conventional ladle up through one leg
of a vacuum vessel having a small tuyere for argon
injection, exposing the metal to the vacuum, and
circulating the metal down the other leg back into
the ladle. Oxygen is top blown through a lance
embedded in the refractory-lined vacuum vessel.
The AOD-VCR process uses a standard AOD
converter (including top lance and side blown
tuyeres) and a water-cooled vacuum hood that can
be sealed to a flange on the top of the converter. A
steam plant is required for vacuum units (steam
ejector pumps are used to pull vacuums as low as
0.3 torr).
The major advantages of converter equipment
relative to vacuum are ease of maintenance and slag
handling ability. The major advantage of vacuum
equipment is the ability to reach lower levels of
carbon and nitrogen without excessive argon
consumption.
7
Gas Control Systems
Most secondary steelmaking processes use
microprocessor-based digital electronic systems to
accurately control and measure gas flow rates.
These systems are capable of data logging and
compatible with process control computer
programs. Formerly common push button control
systems are being replaced by touch screen
controls. Process control computer models (static,
dynamic, or hybrid) may use artificial intelligence to
control endpoint carbon and temperature levels,
and may eventually obviate the need for sublances
and/or off-gas analyses. At a minimum, these
programs advise the operator of blowing
procedures, alloy and flux addition requirements,
and nitrogen switch-points, as well as provide data
logging. Many operators integrate the process
control computer model with the gas control system
(valve rack) and some with other vessel functions
(vessel rotation, additions and fume system
interfaces, sampling, etc.). These programs can be
integrated with automated sampling equipment
and/or equipment for off gas analyses. The degree
of control and process algorithm sophistication
varies. However, savings similar to those shown in
Table IV for an AOD converter are typical.
Table IV
Improvement in Operation Efficiencies Due to
Process Computer Control (Index)
Before
Process
Computer
After
Process
Computer
Operation time 100 89
Si consumption 100 87
Refractory consumption 100 92
Ar consumption 100 90
Accuracy of alloy
composition
100 82
Gases used in converter processing may
include: oxygen, nitrogen, argon (pure and crude),
air, carbon dioxide, and hydrocarbons (sometimes
liquid). Oxygen or nitrogen may be supplied from
on-site plants (cryogenic or non-cryogenic) or liquid
gases transported to the plant site. Total gas flow
rates (top and bottom blown) range from 0.5 to 3.5
cubic meters/min/ton. Vacuum processes typically
use much lower total input rates, e.g., 0.02 to 0.6
cubic meters/min/ton (the typical top oxygen blow
rate is 0.3 to 0.4 cubic meter/min/ton), and seldom
use gases other than oxygen, nitrogen, and argon.
Auxiliary Systems
As in most steelmaking facilities, materials
handling (shop layout, flow of materials, etc.) is
extremely critical in secondary steelmaking systems
in order to minimize operating costs and maximize
productivity. Auxiliary systems closely associated
with converters include alloy and flux additions
systems; slag-off equipment before and/or after
converter charge; automated in-situ sampling and
temperature (sub lance); and off gas measurement
devices and a chemical laboratory.
Total alloys added during converter
decarburization for chemistry control, weight build-
up, and/or cooling typically range from 5 to 30
percent of the tap weight and include carbon and
stainless scrap, nickel, chromium, manganese,
molybdenum, etc. Alloy additions (primarily
ferrosilicon) are also used in the reduction period.
Fluxes added in the process range from 6 to 12
percent of metal tap weight. State-of-the-art
converters have overhead additions systems
capable of making additions without oxygen blow
interruption, either continuously or in batches.
Continuous alloy addition rates typically range from
20 to 35 kg/min/ton. Alloy and flux additions
systems are also used in VOD units to a lesser
degree; the additions range from 4 to 8 percent of
tap weight.
In order to minimize converter operating time
and costs, most steelmakers slag off the transfer
ladle tapped from the arc furnace prior to charging
the converter. There is a tendency for steelmakers
to use mechanical rakes for this procedure to obtain
a consistent degree of slag off without metallic yield
loss and minimal exposure to operators. Some
operators use the same or a second rake to slag off
the converter after tap, particularly if the steel will
be transferred to a VOD (such rakes are typically
use by VOD operators).
8
Some modern converter operators use a sub-
lance to obtain temperatures and metal samples
without blow interruption. Normally, these
measurements are limited to the decarburization
period; however, at least one operator uses a “no
tilt” practice in which all sampling throughout the
heat is made by the sublance. Off gas analyses, in
which infra-red, mass spectrometer, and/or
parametric oxygen measurement techniques are
used to determine the rate of carbon removal and
carbon content at any point in the blow, are used by
several converter operators. Sublances for
sampling and temperature are also used in VOD
equipment.
Analytical Facilities
Modern secondary steelmaking facilities must
be capable of preparing, analyzing, and reporting
metal and, in some shops, slag chemistries.
Typically, samples are obtained, transferred to the
lab via pneumatic tube, analyzed on an X-ray
machine or spectrograph (Leco or similar type
equipment is used for carbon, sulfur, and nitrogen),
and reported back to the control room on a
computer screen. Ideally, a carbon analysis should
require about 4 minutes and a complete metal
analysis should require a maximum of 8 minutes.
Several converter operators analyze slag samples
online.
Refractories
The most common types of refractories used in
secondary steelmaking units are dolomite,
magnesite chrome, magnesite, and/or synthetic
versions of these bricks. Magnesite carbon bricks
are used primarily in some converters that are
tapped at high carbon levels (e.g. converter/VOD
processes). Linings are zoned by quality and
thickness. The brick used in converters is normally
thicker than that used in the ladles of vacuum
processes. Bottom blown vessels are thickest at
the bottom, and side blown converters are thickest
in the tuyere pad area. The total refractory weight
is greater in bottom blown converters due to the
greater specific volume requirements (splash).
Tuyere wear rates, using optimum practices, are
similar in both side and bottom blown converters.
The high temperatures generated locally at the tip of
the tuyeres, the high degree of bath agitation, and
the alternating formation of both highly oxidizing and
reducing slags place significant demands on
refractories used in the converter processes.
Converter refractory costs typically account for
roughly 25 percent of operating costs. The choice
of refractory brick is dependent on the specific
blowing and slag practices, final product chemistry
specifications (grades), production rates (number of
heats/day), required casting temperatures, and brick
availability and cost.
Fume Systems
Emissions collection and treatment equipment
for converters normally consists of a hood,
associated ducting, and a bag house; however,
there are several types of hood designs, and both
wet and dry precipitators are sometimes used.
Large, modern converters usually use a “semi-
closed” capture water-cooled hood (located about
0.2 to 0.4 meters above the vessel mouth). These
types of hoods are integrated with additions
systems chutes and both top lances and sublances.
Non-water-cooled hoods may be located about 1
meter above the mouth. Another variation involves
the use of an accelerator stack above the vessel that
acts as a chimney to direct the fume to a roof
canopy.
Unless the top lance is used for post
combustion, the gas evolved from the converter is
primarily CO along with argon or nitrogen. A
major portion of the combustion and cooling takes
place at the mouth exit and/or at the hood intake.
Further cooling takes place in the duct. By the time
the gas reaches the baghouse, the temperature is
typically 120°C and it is composed entirely of CO
2
and inert gas. Solids emissions from the converter
are primarily particulates of metallic (e.g., iron,
chromium, and manganese) and non-ferrous oxides
(e.g., lime and silica). Total solids emissions
average between 6 to 10 kg/ton of metal, and
loading is about 50 grams/cubic meter. Baghouse
dust is normally handled by commercial processors.
The total cost of the fume system is
approximately 40 percent of a new converter
9
installation cost. In vacuum processes, the fume
collection system is an integral part of the equipment
and typically is used in conjunction with a water
treatment plant.
Vessel Operation
There are several common steps during
operation of all secondary steelmaking processes
including decarburization, reduction, and refining.
The overall carbon removal efficiency (CRE) is
typically about 50 percent, and the metallic oxides
formed are reduced in the reduction period.
Refining takes place in both the decarburization and
reduction periods.
Decarburization
In converter processes, the vessel is usually
charged with deslagged metal and blown at
oxygen/inert ratios of 3/1 to 5/1 through the tuyeres
and 100 percent oxygen through the top lance. The
adaptation of the top lance to bottom or side blown
converters has significantly reduced converter
processing costs and increased productivity. An
example of the effect of top blowing with a “hard
blow” (essentially 100 percent oxygen reacting with
the bath) on decarburization time relative to
operation without a top lance in a side blown
converter is shown in Figure 7. Use of a “soft
blow” (e.g., with 60 percent oxygen reacting with
the bath and 40 percent burning CO to CO
2
in the
headspace) would result in a shorter time decrease
(about 70 percent of the “hard blow” case).
However, the additional heat generated and
transferred to the bath could be used to decrease
silicon consumption, increase the amount of scrap
melted, and/or lower the arc furnace tap
temperature. Some operators use a lance design
that permits simultaneous “hard and soft” blowing.
AOD Process
4 tuyeres, gas rate: 1 Nm
3
/t-min
Refining time: 42 min
KCB-S Process
Gas rate: 1.35 Nm
3
/t-min
Refining time: 23.5 min
Refining Time (min)
0 10 20 30 40 50
0
0.5
1.0
1.5
Carbon
Content
(%)
Figure 7: Decarburization Comparison of an AOD
Without the Top Lance Versus an AOD With the Hard
Blow Top Lance
Lime is added just before and immediately after
metal is charged. During the initial step(s), alloys
are added for weight and chemistry adjustment, and
to control the temperature to a maximum of
1700°C. Start carbon in converters can vary from
0.7 to 4.5 percent (typically 1.0 to 2.5 percent).
Start silicon ranges from 0.2 to 0.4 percent (in
some cases as high as 3 percent). Silicon is oxidized
in the initial stage of the blow. The oxygen/inert ratio
in the tuyeres is progressively decreased (stepwise
or in some installations, continuously) as the carbon
content decreases (e.g., 1/1 at 0.3 percent carbon,
1/3 at 0.15 percent carbon, 1/5 at 0.04 percent
carbon, and pure argon [argon decarb] at 0.015
percent carbon in austenitic grades) in order to
minimize chromium oxidation and control
temperature. Top blowing is usually limited to the
first step (3/1 ratio); however, if using mixed gas
blowing (oxygen/inert ratios of 2/1 or 1/1), top
lancing can be continued down to a carbon level of
roughly 0.15 percent (in austenitic grades). Figure
8 shows an example of mixed gas blowing in a side
blown converter. Several vessels use the top
heating process that uses an oxy-fuel burner lance
to add heat and/or decarburize the bath in the initial
blowing steps. Sampling and temperature
measurement normally takes place when the carbon
level is about 0.1 percent (0.04 percent when
making ELC grades). Depending on the aim
10
nitrogen, nitrogen can be used as the inert gas until a
calculated carbon level and oxygen/inert ratio is
reached (carbon ratio switchpoint). In very low
nitrogen grades (0.010 percent nitrogen and 12
percent chromium), no nitrogen is used, and in high
nitrogen grades (0.24 percent nitrogen and 18
percent chromium), no argon is required. The
length of the decarb period is determined by the
start carbon and silicon level, blowing rates, the
amount and efficiency of additions, and the aim
carbon content. The decarb time ranges from 20 to
35 minutes in modern converters (start carbon 1.5
to 2.5 percent, and aim carbon 0.04 percent).
3. Step 2. Step 1. Step
4. Step 5. Step
Reduction and
Desulfurization
O
2
lance
Ar
lance
Ar
tuyere
0 50 100
Process Time(%)
Blowing
Speed
(m
3
min)
Decarburization
O
2
lance
O
2
lance
O
2
lance
O
2
tuyere
O
2
tuyere
O
2
tuyere
O
2
tuyere
O
2
tuyere
Ar
tuyere
Ar/N
2
lance
Ar/N
2
lance
Ar/N
2
tuyere
Ar/N
2
tuyere
Ar
tuyere
0
50
100
Figure 8: Mixed Gas Blowing in a Side Converter
The deslagged metal charged to the VOD is
typically around 0.3 percent carbon and less than
0.1 percent silicon. All of the oxygen is blown from
the top and argon is injected via porous plug(s).
Since the blowing rate in a ladle is limited, the
decarb time is typically 45 to 60 minutes.
Reduction
After converter decarburization, a reduction mix
consisting of silicon and/or aluminum and fluxing
agents (lime and spar) is added and stirred with
inert gas for 5 to 8 minutes. Using 2 to 3 kg
aluminum/ton, 2 to 3 kg spar/ton, a slag basicity of
1.5 to 1.7, and a temperature of 1700°C, sulfur
contents of 0.003 to 0.005 percent can be obtained
with start sulfurs as high as 0.03 to 0.04 percent. If
lower sulfur levels are required, the reduction slag is
decanted, and a second slag is added and stirred
for 3 to 4 minutes to obtain sulfur levels less than
0.001 percent. Ideally, the chemistry after
reduction should be within aim specifications;
however, if trim additions are required, they can be
made with a 2-minute stir after reduction, during a
second slag for desulfurization, or in the tap ladle.
Obviously, slag composition and fluidity control
is very critical in secondary steelmaking processes
with respect to refractory wear, alloy recovery, and
desulfurization. However, in recent years, the effect
of slag chemistry on decarburization and nitrogen
removal, as well as the need to consider the slag in
process control models, has been recognized, and
appropriate practices have been developed by
several operators.
Design maximum tap-to-tap times in modern
converter installations for all grades is 60 minutes or
less in order to match the casting times necessary
for sequence casting.
Post Converter Treatment
In some grades, final deoxidation additions and
the addition of reactive alloys such as titanium are
made in the converter tap ladle during tap. In
recent years, an increasing number of producers use
ladle treatment (ladle metallurgy) stations, ladle
furnaces, or vacuum systems. The primary function
of such facilities is to reduce the converter tap-to-
tap time, improve quality, and/or provide a buffer
between the converter and continuous caster. Ladle
treatment stations normally include automated
additions systems, an injection lance or a porous
plug for stirring with inert gas, and wire-feeding
equipment. Some producers use lances to inject
lime, spar, and/or calcium alloys for quality
improvements that include: reduced inclusion
content, improved weldability, and decreased
oxygen levels. Wire feeding is used to add alloys of
titanium, aluminum, calcium, and/or sulfur to
improve alloy recoveries and quality (decreased
oxygen content, inclusion shape control, etc.) on
certain grades. The use of ladle treatment facilities
in combination with modern process control
11
computer models permits tapping immediately after
reduction (often without waiting for sample
analyses).
Some operators use ladle furnaces to provide
additional heat if required to perform the above
operations. One drawback of ladle furnaces is the
problem of carbon pick-up from the electrode arcs
(minimum pick-up is typically 20 ppm). Despite
developments of practices to control slag thickness
and stirring rates to minimize carbon pick-up,
production of ultra-low carbon grades with a ladle
furnace is not practical. With the advent of
improved process control computer models, most
large vessel operators do not have ladle furnaces.
If they have such equipment available, they tend to
use them for reheating in only 10 to 20 percent of
the heats cast.
A number of operators use vacuum
decarburization units following converter tapping for
special grades such as ultra-low carbon and
nitrogen ferritic grades. Production of stainless steel
using the arc furnace (or other hot metal source)
with a converter for decarburization to about 0.1 to
0.3 percent carbon, followed by transfer to a VOD
or final decarb, and finishing is shown as the
converter/VOD process in Figure 3 and is
sometimes referred to as the “triplex process.”
(Refer to the “Other Converter Processes” section
for more information.)
Specific Refining Processes
AOD
Presently, there are over 60 AOD steel mill
installations worldwide producing about 69 percent
of the total Western World stainless (see Figure 3).
There are a similar number of AOD foundry
installations. There are more than 400 publications
in the literature relative to the AOD process.
Additional details of the historical development,
fundamental thermodynamics, equipment, and
process are summarized in “Pneumatic Steelmaking,
Volume Two, The AOD Process,” edited by S.K.
Mehlman, Iron and Steel Society, 1991.
The rapid acceptance of the AOD process by a
majority of worldwide stainless producers is
indicative of its capability to economically produce
high quality product from least cost charge
materials. Continued new developments in AOD
practices and equipment have kept the process
competitive with respect to operating costs and
productivity relative to other processes. These
include: the development of optimum techniques for
top blowing; reliable process control models to
predict carbon and temperature endpoints; brick
thickness zoning increases in the tuyere pad area;
and practices to optimize economic production of
ultra-low carbon and nitrogen low-chrome ferritic
grades. Note that in this presentation, the Krupp
Combined Blowing Process (KCB) is considered a
modification of the AOD Process.
With respect to the latter development, all of
the producers of ultra-low carbon and nitrogen
ferritic grades (409L and 439L) listed in Table III
(outside of those in Japan) are AOD producers.
Table V illustrates the carbon and nitrogen levels
obtained in the tap ladle on such grades. Carbon
and nitrogen levels similar to those obtained on 16
to 18 percent chromium ferritic grades are also
obtained on austenitic grades of similar chromium
content.
Table V
Carbon and Nitrogen Levels Obtained in the AOD
on Ferritic Grades (Ladle)
11-13% Cr
Grades
16-18% Cr
Grades*
Typical Best Typical Best
Carbon
(ppm)
70 30 70-100 60
Nitrogen
(ppm)
80-110 60 100-130 80
* Similar results obtained on austenitic grades.
Several other promising practices that still under
development and/or not widely adopted by AOD
operators to date include: stainless steel
dephosphorization, powder injection through
tuyeres, use of irregular-shaped tuyeres, and slag
splashing.
Table VI shows the main characteristics and
estimated costs of practices developed for stainless
12
dephosphorization. In addition to phosphorus
reduction, the calcium carbide/spar techniques
(reducing practice) also results in significant
reduction in levels of arsenic, antimony, and tin.
While this technique has proven successful in trials
on converters ranging in size from 3 to 120 tons, a
major drawback to widespread routine use of this
practice is the practical and economic aspects of
slag handling. If the phosphide in the slag is not
oxidized before the slag cools to less than 200°C, it
will react with moisture to form phosphine (PH
3
), a
toxic gas.
Table VI
Comparison of AOD Practices to Remove
Phosphorus
Practice Flux
(kg/ton)
Dephos-
phorization
(%)
Cost
Increase
1
(%)
CaC
2
/spar
(reducing)
30/2.5 75 (3-7)
2
BaCO
3
/BaCl
2
3
(oxidizing)
55/55 60 16-26
LiCO
3
/spar
4
/
mill scale
(oxidizing)
6/30/20 63 (60)
1. Operating plus materials plus materials (does not
include savings in least cost charged selection).
2. Does not include the cost of disposing of hazardous
slag.
3. Optimum start carbon content is between 1.5 to 2.5
percent.
4. Effective only at a start carbon of about 4 percent.
KTN has modified its 80-ton AOD converters
to permit lime powder injection through the tuyeres.
However, they have not been able to inject the bulk
of the required lime by this technique. Lime
injected during the late decarb steps lowers the
temperature at the tip of the tuyeres and decreases
carbon removal efficiency. Apparently, the
operating costs and/or maintenance associated with
the equipment outweigh any benefits from partial
lime addition through the tuyeres, and KTN has
discontinued the practice. However, over the past
several years, Praxair and four major sheet
producers have conducted laboratory and plant
scale trials injecting relatively small amounts of iron
ore and flux through AOD tuyeres with promising
results relative to improving CRE in the low carbon
range and desulfurization to very low levels using a
single slag (without significant amounts of
aluminum).
The use of irregular-shaped tuyeres evolved
from laboratory and 90-ton plant scale trials at
Sumitomo in the early l980s. These trials indicated
significant improvement in refractory life and
reduced inert gas stirring requirements using tuyere
shapes which minimized jet collapse (back attack).
Praxair’s model studies and small plant scale trials
confirmed these findings and suggested that the
optimum design would be a severely flattened shape
as shown in Figure 9a. Engineering problems
associated with the manufacture of such a tuyere
and tuyere brick, as well as the question of how to
assure equal distribution of shroud gas over the
perimeter of the tuyere, have hindered widespread
adoption of this ideal. However, several AOD
operators successfully use a standard tuyere with a
rod inserted in the center tube as shown in Figure
9b (a variation of the same concept).
9a
9b
Figures 9a and 9b: Schematic of Irregular-Shaped
Tuyeres
Current development work includes efforts by
Praxair to adapt “slag splashing” techniques used in
the LD (BOF) to the AOD. This would probably
require the use of a tap hole in the AOD.
Table VII summarizes typical and best results
currently obtained in operating AOD vessels. It
should be recognized that there are inherent
difficulties in making generalizations applicable to
AOD operators worldwide since there are wide
variations in raw materials, product mixes,
13
equipment, practices, and economic conditions.
Further, some of the improvements used by the
majority of AOD operators have not been
universally adopted.
Table VII
AOD Consumption when Producing AISI 304 in
an 80-Ton Vessel*
Typical Best
Argon (Nm
3
/ton) 12 9
Nitrogen (Nm
3
/ton) 9-11 9
Oxygen (Nm
3
/ton) 25-32 ---
Lime (kg/ton) 50-60 42
Spar (kg/ton) 3 2
Aluminum (kg/ton) 2 1
Silicon (reduction) (kg/ton) 8-9 6
Brick (kg/ton) 5-9 2
Decarbmetallics (kg/ton) 135 ---
Charge to tap time (min) 50-80 40
Total Cr yield
(% EAF/AOD)
96-97 99.5
Total Mn yield
(% EAF/AOD)
88 95
Total metallic yield
(% EAF/AOD)
95 97
* 1.8 start carbon, 0.05 percent N
2
, 0.005 percent S
Other Converter Processes
Figure 3 indicates that converter processes
(defined as those in which converters are directly
fed from melting units and having no vacuum
facilities) other than AOD accounted for 5 percent
of the total Western World stainless steel
production in l996. Other converter processes
currently in use include the CLU, K-OBM-S,
MRP, ASM, and several top blown oxygen/bottom
blown argon converter processes (including SRF).
The bulk of the stainless production in l996 in
this category was made in the CLU at Columbus in
South Africa (280,000 tons), at the Degefors plant
of Avesta Sheffield in Sweden (180,000 tons), in
the ASM at Jindal in India (90,000 tons), and in
NKK's SFR in Japan (90,000 tons).
The CLU converter process was developed by
Uddeholm in Sweden and Creusot Loire in France.
The first commercial installation was made at
Degefors in Sweden (now part of Avesta Sheffield)
in l973. The process is bottom blown and
originally, steam was used as a diluent throughout
the decarb steps, and argon consumption was
reduced by about 80 percent relative to the AOD
process. However, the silicon consumption
increased by 250 percent due to increased metallic
oxidation below 0.18 percent carbon. The process
has evolved to use various amounts of nitrogen,
argon, and steam depending on nitrogen and
hydrogen aims. The process is used by Samancor
in the production of medium carbon ferrochrome in
South Africa and at Maharastra in India. The most
recent installation (l995) was the 100-ton converter
at Columbus in South Africa (note this unit has a
top lance). Table VIII shows results obtained
during the commissioning period. The justification
for the Columbus installation was based on their
lack of cooling material. (The lack of argon
availability and/or high argon price are other
reasons to justify the CLU process.) While argon
consumption is lower, silicon and refractory
consumptions are higher, and maintenance
requirements (associated with steam generation) are
greater than in other converter processes. Fume
requirements for a given oxygen input rate are
greater for the CLU (because of the hydrogen
involved).
Table VIII
CLU Consumption
(100-ton Vessel During Commissioning)
304.00 409.00
Aim carbon % 0.03 0.01
Aim nitrogen ppm 350.00 100.00
Melt-in carbon % 1.65 0.96
Melt-in silicon % 0.20 0.13
Oxygen Nm
3
/ton 27.70 22.40
Nitrogen Nm
3
/ton 13.50 1.70
Table VII
CLU Consumption
(100-ton Vessel During Commissioning)
304.00 409.00
Steam Nm
3
/ton 10.40 6.00
Hydrogen ppm 5.90 3.80
Argon Nm
3
/ton 4.00-7.00 17.10
Silicon (reduction) kg/ton 15.50 15.90
14
The K-OBM-S process evolved from
Kawasaki's K-BOP process (described in the
“Converter/Vacuum Processes” section) which, in
turn, evolved from the OBM practice as formerly
used on stainless steel. K-OBM-S is similar to the
modern AOD (top blown with a lance and either
bottom or side blown with tuyeres). Its features
include: high blow rates; the addition of energy
sources such as carbon; and the use of
hydrocarbon shroud gases, post combustion, and
lime powder injection (all of which may be included
in the modern AOD depending on the steelmakers’
needs and the degree of flexibility desired). The K-
OBM-S is promoted as a bottom blown process;
however, two of the first four installations (POSCO
in Korea and Bolzano in Italy) are side blown. In
these installations, the K-OBM-S is used to finish
the bulk of the production, which is then transferred
directly to the ladle treatment station and/or
continuous caster. In the other two installations,
ISCOR and Microsteel in South Africa, the
converter is bottom blown to an intermediate
carbon content and then transferred to a VOD
(triplex process). All four of these installations were
started up in l996.
Proponents of side blowing claim that higher
carbon removal efficiencies are obtained in the
range of 0.1 to 0.005 percent carbon (due to longer
inert gas bubble residence time), and improved
desulfurization is obtained in the range of 0.005 to
less than 0.001 percent sulfur (due to improved
mixing). There is a tendency for bottom blown
converter operators to use an EF-Converter-VOD
process route (triplex) in order to decrease argon
and silicon consumption and process time when
making low carbon grades. The MRP and ASM
processes, developed by the major German steel
mill equipment suppliers (MDH and MAN GHH),
are similar to the AOD process except they are
bottom blown (with or without a top lance). In
some installations, the MRP process is operated
using alternate periods of oxygen and inert blowing
rather than mixtures (the concept is similar to argon
decarb in the AOD). The use of MRP or ASM
without VOD is essentially limited to companies in
India and several small installations in Germany. To
the extent that these vessels are operated as an
AOD, operating costs are similar to that of AODs
on regular carbon grades.
NKK's SRF converter (120 tons) is essentially
a top and bottom blown LD. (Tuyeres are used to
inject inert gas in the bottom) What makes this
process unique is the process routing, which is
nickel ore converter-chrome ore converter-SRF.
Despite high operating costs and relatively low
productivity, materials (and overall) costs are less
than those of other converter processes.
Vacuum Decarburization Processes
VOD and other vacuum processes (fed directly
from the EF or LD) accounted for about 6.8
percent of the Western World production in l996
(see Figure 3). The bulk of this production was
made in Japan (about 400,000 tons) with significant
amounts being made in China and Canada, and to a
lesser extent in the United States, the United
Kingdom, Germany, France, and India. Most
vacuum production (probably 98 percent) was
made using the VOD process. As previously
noted, the process was developed at Witten
(Thyssen) in Germany between 1962 and l967. Its
major attributes include a minimal consumption of
argon (e.g., about 1 cubic meter/ton of steel), and
the elimination of nitrogen pick-up during tap
associated with converter processes (since the
VOD ladle is the casting ladle). Silicon
consumption in the VOD is only 3 kg/ton; however,
because of the 0.3 percent carbon and less than 0.1
percent silicon VOD charge requirement, an
additional 3 kg/ton of silicon is used to minimize
chromium losses in the arc furnace. The major
drawbacks relative to converter processes are
higher refractory consumption, a lower productivity
rate in both the EF and VOD (EF time was
reduced by 25 percent versus 50 percent, and
VOD had a charge-to-tap time of 50-70 minutes
versus 40-60 in the AOD), less flexibility relative to
the use of least cost charge mixes (EF carbon
ideally is about 0.3 percent versus 1.8 percent,
silicon content is less than 0.1 percent versus 0.3
percent, and decreased desulfurization capability),
15
and lower scrap melting capability. Refractory
consumption is higher in the VOD than in modern
converters. Maintenance and operating costs
associated with steam production are also
drawbacks.
In the early l970s, Kawasaki Steel developed
the SS-VOD, which included modifications to the
standard VOD such as multiple porous plugs and
argon O-ring sealing. This process is now used
primarily for making extra-low carbon and nitrogen
ferritic grades having chrome contents between 22
and 30 percent to reach maximum carbon and
nitrogen levels of 30 and 60 ppm, respectively.
Maximum carbon and nitrogen levels reached on 18
percent chromium grades were 15 and 30 ppm,
respectively. However, the typical treatment time is
120 to 180 minutes. By way of comparison on 18
percent chrome grades, standard VODs obtain
maximum carbon and nitrogen levels of 60 and 90
ppm, respectively (just slightly below those
obtained in the AOD as indicated in Table V).
Recently, several Japanese VOD operators have
adopted the PB Process (powder blowing through
a top lance onto the bath under vacuum) using iron
ore to improve CRE and/or flux to improve
desulfurization. In both applications, the time of the
VOD process is decreased.
Up until mid-l995, a significant amount of
stainless steel was produced in the KTB or
modified RH-OB process at Kawasaki Chiba (see
Figure 1). Following K-BOP converter
decarburization to about 0.12 percent carbon, steel
was transferred to the KTB for final decarb and
finishing. In the new shop, the VOD replaced the
KTB in a triplex route (see the “Converter/Vacuum
Processes” section). An earlier version of this
process at Nippon Steel Muroran was discontinued
in l993.
Converter/Vacuum Processes
In l996, converter/vacuum processes accounted
for about 19.5 percent of the Western World
stainless production (see Figure 3). Process routes
in this category include converter/vacuum units that
are fed from any hot metal source (e.g., EF, BF,
Corex, liquid FeCr, etc.). Essentially, any type of
converter can be used to supply the VOD. The
AOD-VCR (combined AOD and vacuum reactor)
is also included in this category.
The bulk of the l996 production in this category
was made in the MRP/VOD at ALZ (580,000
tons), K-BOP/VOD at Kawasaki (600,000 tons),
LD/VOD at Nisshin (540,000 tons), and
MRP/VOD at Yieh United (300,000 tons).
Significant amounts were also made in the
AOD/VCR at Daido (240,000 tons), LD/VOD at
Nippon Steel Yawata (200,000 tons), AOD/VOD
at Sumitomo (140,000 tons), MRP/VOD at
Acesita (100,000 tons), and the AOD/VOD at
Mukand (80,000 tons).
The K-BOP process as used at Kawasaki
Chiba is unique in that the process route, BF-
Converter I-Converter II-VOD, uses one converter
(SRF) for chrome ore reduction. The vessel is
charged with blast furnace iron, chromium ore (up
to 11 percent of the chromium), coke, lime, and
stainless scrap or HCFeCr. This top and bottom
(tuyere) blown converter utilizes a tap hole, post
combustion lance, and lime injection through the
tuyeres. The second top and bottom blown K-
BOP is used for decarburization (uses a sublance,
OG or close-fitting fume hood, tap hole, and off gas
analyses). Note that this process as well as all
others that use a blast furnace or Corex iron must
use a dephosphorization step (torpedo car or ladle)
between the BF and the converter.
In the MRP/VOD process (and in most
LD/VOD processes), heats are always tapped from
the converter at relatively high carbon contents (0.3
percent) and transferred to the VOD for finishing,
as opposed to AOD/VOD or K-OBM-S/VOD
facilities that can finish most grades without vacuum
treatment. All oxygen is top blown. (Only inert gas
is injected in the bottom via a porous plug or
elements.) The major advantages of the
MRP/VOD and other triplex processes is lower
consumption of argon (3 to 4 cubic meters/ton) and
silicon (5 to 8 kg/ton), and the elimination of
nitrogen pick-up during tapping associated with
duplex converter processes. These differences are
particularly significant when making ultra-low
16
carbon and nitrogen grades. The major
disadvantages include: inability of the VOD to attain
the cycle times desirable for sequence casting
(particularly in ultra-low carbon and ultra-low
nitrogen grades); decreased scrap melting
capability; lower total chromium and metallic yields;
higher costs associated with maintenance and costs
of steam production; and higher total
converter/VOD refractory costs.
AOD-VCR
This process, developed at Daido Shibukawa in
l989, combines AOD and VOD in the same
converter. It is used at Daido Chita and was
adapted to the AOD at Nippon Steel Hikari last
June. Predecessors to this process were the Ugine
vacuum converter process developed in the mid-
l970s and the VODC or VODK process
developed at Witten between l973 and 1976. The
VODC process is used in a number of small
foundries and/or mills in Finland, China, and India.
The VODC is typically blown with a top oxygen
lance and with a single tuyere in the vessel bottom.
The AOD-VCR operates as a conventional
AOD down to a carbon level of 0.08-0.10 percent.
The process is stopped for sampling and a vacuum
lid is put into place (sealed to a flange located about
halfway up the conical section of the converter). A
vacuum is pulled and used for the remainder of
decarb and reduction (desulfurization is carried out
in the transfer ladle prior to AOD charge). The
major advantages of this process relative to
converter processes are decreased argon and
silicon consumption. The disadvantages include:
higher refractory consumption, a decreased ability
to melt scrap, and maintenance and costs
associated with steam production. The actual time
of the operation (including turndown for sampling,
exchanging hoods, lip seal maintenance, pulling and
breaking vacuum, etc.) is longer than that for a
modern AOD on regular carbon grades. Relative
to having separate converter and VOD units, the
AOD/VCR has higher operating costs (silicon,
refractory, and argon), lower productivity, and
higher nitrogen contents. However, capital costs
may be somewhat lower.
Process Route Selection
Choosing the most practical and economic
process route for stainless steelmaking depends on
many factors including consideration of: raw
material availability and costs; productivity
requirements; product mix; existing shop equipment
and logistics; and capital and operating costs.
Other important factors may include: general
business conditions; company financial status;
company mergers and/or consolidation of
steelmaking facilities; plant closures; competency of
available manpower; and social and political
concerns. The fact that many of the above factors
may change over time suggests choosing the most
flexible process route.
As noted previously, the melting and refining
equipment and the process chosen should have the
ability to melt the least cost charge (since as much
as 85 percent of melting and refining costs are
materials costs). The availability and cost of certain
raw materials may be critical to process selection.
For example, the availability of a blast furnace or
Corex hot metal in an integrated steel plant, and/or
expensive power or an erratic power supply may
favor hot metal supply to converters or
converter/vacuum systems, as would the lack of
stainless scrap. Hot metal supply means high start
carbon and silicon levels, which also favor a high
blowing rate converter or converter/vacuum
systems. Lack of stainless scrap or other alloys for
cooling the bath may favor a less heat efficient
converter process. High argon costs would tend to
favor converter/vacuum or vacuum rather than
converter-only processes.
Maximum productivity usually can be obtained
using converter processes or converter/vacuum
processes. Product mix is a factor. (A high
production of high chromium, ultra-low carbon, and
nitrogen grades would favor converter/vacuum
processes. Grades that require ultra-low sulfur
levels would tend to favor converter processes.)
Existing shop equipment and logistics, both
upstream and downstream of the converter, are
critical when selecting secondary steelmaking
17
equipment for existing shops. Considerations
include: EF capacity and power rating; crane
capacity and availability; baghouse capacity; slag
handling facilities; the availability of a ladle furnace
and/or ladle treatment facilities; continuous casting
requirements; and the availability of steam and/or
water treatment plants.
For example, if the EF capacity is insufficient
for the required production levels, the refining
process should be able to melt large amounts of
cold materials. Precise control of chemistry,
temperature, and weight is more critical in the
EF/VOD process. Limitations in crane and/or
baghouse capacity could restrict the choices of raw
materials, vessel size, blowing rates, etc. The
availability of a ladle treatment station or ladle
furnace results in shorter charge-to-tap time in the
converter. Heats can be tapped immediately after
reduction and precise adjustments of chemistry,
weight, and temperature necessary for sequence
casting can be made in the ladle. Of course, shop
layout and logistics are also critical aspects of new
shop design.
In some recent installations, capital spending
considerations have eclipsed those of all other
factors including flexibility and operating costs in the
selection of process route. For major turnkey
installations (which may include steel melting,
refining and casting, hot and cold rolling, anneal and
pickle lines, and the associated technology), the
choice of equipment suppliers is limited to three or
four large firms. Each of these suppliers
aggressively promotes its own secondary
steelmaking process as superior relative to
operating costs, quality, productivity, etc..
However, several of these suppliers will
manufacture and install a version of other types of
secondary steelmaking if the steelmaker insists.
Depending on geographical location, order
backlogs, previous equipment sales, manpower
availability, etc., any one of these suppliers may be
in a better position to manufacture and install
equipment at a substantially lower price than the
others.
At present, a large majority of stainless steel
producers use the arc furnace to supply metal to
various secondary refining processes. However, as
can be seen in Figure 10, numerous choices may be
feasible. All shown are presently in use by some
operators. Eventually, all of these routes lead to a
converter or an arc furnace. Figure 11 illustrates
various process routes leading from either the
converter or arc furnace to the continuous caster.
The most common type of continuous casting
machine in use today is the slab caster. However,
thin slab casting has been adopted by several
stainless producers (including Avesta, AST, Armco,
and Nucor), and the first production strip casting
machine is scheduled to start up this September at
Nippon Steel Hikari. The incentive for these
techniques is the partial or complete elimination of
reheating and hot and cold rolling equipment and
processing used in standard finishing operations.
However, some quality and/or practical problems
have yet to be solved for thin slab casting in the 50
to 80 mm thickness range, and the feasibility of strip
casting, particularly as related to quality, remains to
be proven on a production scale. Process selection
for existing or new stainless producers that are
evaluating upgraded or new facilities relative to hot
metal supply or type of continuous caster is very
complex. However, the choice of the most flexible
and economic refining process route for a high
productivity sheet producer making the full range of
stainless grades seems clear. There should be a
converter that has the ability to fully decarb and fully
reduce essentially all grades with a maximum tap-
to-tap time of 60 minutes. Grades tapped from the
converter when operated as above would be
transferred to a ladle treatment station (a ladle
furnace is probably not necessary for large vessels)
and later to the caster. For some grades requiring
ultra-low carbon and nitrogen levels, for scheduling
optimization, and/or in periods when certain raw
materials are not available or are high in cost, the
converter would be tapped at roughly 0.3 percent
carbon, partially reduced, and transferred to the
VOD for final decarb and finishing prior to casting.
18
Converter
Cr Ore Reduce
(coke)
Converter
EF
BF
Mini BF
Corex
Ladle
Converter
Converter
EF
Converter
Converter
EF
Converter
Converter
Liquid Fe Cr
SAF
Converter
Mild Steel Scrap
Liquid Fe Ni
SAF
Ni Ore
Reduce
Cr Ore Reduce
(coke)
Ladle
Figure 10: Metal Source for Secondary Refining
It is interesting to note that of 17 stainless
installations (over 40 tons capacity) in which new or
replacement refining equipment has been installed
since the beginning of l995, there were nine side
blown converters (seven AOD and two K-OBM-
S), two EF/MRP/VOD units, one bottom blown
K-OBM-S/VOD unit and one each of the
EF/VOD, EF/CLU, BF/LD/VOD, EF/AOD/VOD,
and AOD-VCR units.
Ladle Furnace
or
Ladle Treatment
Station
Thin Slab Caster
50-125 mm
Slab Caster
140-250 mm
Strip Caster
2-5 mm
VOD
Converter
EF
Figure 11: Stainless Refining and Casting Options
Future Trends
As noted previously, the average worldwide
growth rate of stainless steel production is
projected to be about 5 percent through the year
2000; however, this will vary from 3 percent in
Europe and the United States to as much as 15
percent in developing countries. Despite estimates
that present production is less than 80 percent of
existing capacity, existing and new entrants in the
stainless business will install new facilities. It is
projected that 2 million tons of new annual capacity
will be installed over the next 5 years in China,
India, Korea, and other Asian countries. Projects
now under consideration in the United States could
add another 1.2 million tons annually if successfully
financed.
Existing producers will restructure their
operations, upgrade equipment, and optimize
operating costs. They will increase efforts to
educate and promote the mechanical properties, life
cycle cost, and environmental benefits of stainless to
end users and the public in order to increase
demand growth. Further mergers and plant
shutdowns among large companies will occur.
Along with a predicted geographical shift in
stainless production, the percentages of stainless
made by converter/VOD process routes will
increase at the expense of converter-only and
vacuum-only processes. Hot metal supply from
sources other than the arc furnace will become
more common, particularly in developing countries
such as China and India.
19
Assuming that political, social, and economic
conditions in Eastern Europe and the CIS will
improve, existing steel plants will upgrade
equipment and increase production at least to their
former levels of 2 million tons annually.
Process optimization will continue in existing
and new secondary steelmaking installations as
follows:
1. New generations of computer programs will be
applied to more accurately predict end point
carbon and temperature control; this will
continue the trend toward shorter vessel tap-to-
tap times, and increased utilization of post
converter treatment and optimizing sequence
casting. The use of on-line slag analyses will be
expanded to facilitate these improvements.
2. Continued evolution and development of top
lance designs will result in optimization of post
combustion in both hard and soft blow
practices. There will be further application of
fuels and mixed gases in top lances.
3. The use of noncryogenic on-site gas production
facilities (pressure swing absorption and
membrane) to supply oxygen and nitrogen may
become more common. Eventually, argon
recycling in combination with a closed hood,
off-gas, and a heat recovery system may be
applied in a new installation.
4. The application of irregular-shaped tuyeres,
stainless dephosphorization, powder injection,
and slag splashing in converters will be initiated
and/or expanded.
Acknowledgements
The author wishes to thank Doctors Allen Chan
and Ian Masterson, and Messrs. Stewart Mehlman
and Jim Wiencek of Praxair, and Dr. Bal Patil of
Allegheny Ludlum for their contributions to this
presentation.

Sponsor Documents

Or use your account on DocShare.tips

Hide

Forgot your password?

Or register your new account on DocShare.tips

Hide

Lost your password? Please enter your email address. You will receive a link to create a new password.

Back to log-in

Close