Low Cost CSP
Content
FLORIDA STATE UNIVERSITY FAMU–FSU COLLEGE OF ENGINEERING
LOW-COST CONCENTRATING SOLAR COLLECTOR FOR STEAM GENERATION
By JOHN DASCOMB
A Thesis submitted to the Department of Mechanical Engineering in partial fulfillment of theof requirements for the degree Master of Science
Degree Awarded: Spring Semester, 2009
The members of the Committee approve the Thesis of John Dascomb defended on March 26, 2009.
Anjaneyulu Krothapalli Professor Directing Thesis
Brenton Gresk Greskaa Committee Member
Juan Carlos Ord´oo˜˜n nez ez Committee Member
William S. Oates Committee Member
The Graduate School has verified and approved the above named committee members. ii
TABLE OF CONTENTS
List of Tables Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
v
List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
vi
Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
ix
1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
1.1 Moti Motiv vation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.2 Lite Literatur raturee Revi Review ew . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Previous ious Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1.3 Prev
1 1 3 9
1.4 Proje Project ct Goa Goall . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
11
2. BACKGROUND . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
13 13 14 23 28 34
2.1 2.2 2. 2.33 2.4 2.5
Introduction Introducti on to the the Solar Spectru Spectrum m . Solar Sol ar Geom Geometr etry y . . . . . . . . . . . . Opti Op tics cs . . . . . . . . . . . . . . . . . Receive Rece iverr Desig Design n . . . . . . . . . . . . Steam Ste am Boiler Boiler Desi Design gn . . . . . . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
. . . . .
3. EXPERIMENTAL SETUP . . . . . . . . . . . . . . . . . . . . . . . . . . . .
3.1 3.2 3.3 3.4 3.5 3.6
Parabolic Parabo lic Dis Dish h . . . Reflectiv Refle ctivee Surfac Surfacee . . Fram ramee . . . . . . . . Recei Re ceive verr . . . . . . . Trac racking king . . . . . . . Data Acquisit Acquisition ion and
. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Instrumentation Instrumen tation
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
. . . . . .
4. RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . . . . . . . . . . .
4.1 4.2 4.3 4.4 4.5 4.66 4. 4.7
Thermal Efficien Thermal Efficiency cy Result Resultss . System Syste m Charact Characteriza erization tion . . Optical Opti cal Efficie Efficiency ncy . . . . . . Boiler Boi ler Effic Efficien iency cy . . . . . . Insulation Insula tion Efficie Efficiency ncy . . . . Erro Er rorr . . . . . . . . . . . . . Costt Anal Cos Analysi ysiss . . . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
iii
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
. . . . . . .
39 39 41 42 44 46 47 50 50 58 63 63 63 64 64
5. CONCLUSION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
5.1 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
67 68
APPENDICES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
69
A. RECEIVER DRAWINGS . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
69
B. CONNECTION ARM DRAWINGS . . . . . . . . . . . . . . . . . . . . . . . C. CONCENTRATOR CONNECTION RING DRAWINGS DRAWINGS . . . . . . . . . . . .
74 77
D. FOUNDATION DRAWINGS . . . . . . . . . . . . . . . . . . . . . . . . . . .
80
REFERENCES REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
83
BIOGRAPHICAL SKETCH . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
85
iv
LIST OF TABLES
4.1 Ave Average rage Values Values for Testin Testingg Performed Performed Feb Feb 23, 2009 . . . . . . . . . . . . . .
59
4.2 Opt Optica icall Effic Efficien iency cy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
63
Absorbe orberr Effic Efficien iency cy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Abs
64
Materi erial al Cos Costs ts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4.4 Mat
66
v
LIST OF FIGURES
1.1 World HDI vs. Electricit Electricity y Consump Consumption tion [1] . . . . . . . . . . . . . . . . . . .
1
1.2 Pifre Pifre’s ’s 1878 Sun-Po Sun-Power wer Plan Plantt Driving Driving a Printing Printing Press Press [2 [2] . . . . . . . . . . .
3
1.3 Boi Boing/ ng/SES SES DECC DECC Dish Dish Stirl Stirling ing Syste System m [3 [3] . . . . . . . . . . . . . . . . . . .
4
Process Flow and and Corresponding Corresponding Thermody Thermodynamic namic State State for Shenand Shenandoah oah Total Total 1.4 Process Energy Project [4 [4] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
6
1.5 Proce ocess ss Fl Floow an and d Th Ther ermod modyn ynam amic ic St Stat atee fo forr Jo John hnso son n an and d Jo John hnso son n So Sola larr 1.5 Pr Facility [4 [4] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
7
Overa erall ll View View of Sola Solarr 1, at Barst Barstow ow,, CA CA [4] . . . . . . . . . . . . . . . . . . 1.6 Ov
8
1.7 Solar Configu Configuratio ration n of INDITEP INDITEP Project Sevil Seville, le, Spain Spain [6 [6] . . . . . . . . . .
8
1.8 Sc Schem hemati aticc of Sol Solar ar 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
9
1.9 Conce Concentr ntrator ator Assem Assembly bly for Solar 1 [8] . . . . . . . . . . . . . . . . . . . . . .
10
1.10 Rece Receive iverr Assembly Assembly for Solar 1 [8] . . . . . . . . . . . . . . . . . . . . . . . .
11
2.1 Sol Solar ar Spectr Spectrum um at the the Surfac Surfacee of the the Earth Earth . . . . . . . . . . . . . . . . . . .
14
2.2 Ave Average rage Insola Insolation tion for for Tallaha Tallahassee ssee [7] . . . . . . . . . . . . . . . . . . . . . .
15
Declinatio ination n Angle Angle at Summe Summerr Solstic Solstice[ e[88] . . . . . . . . . . . . . . . . . . . . 2.3 Decl
16
2.4 Decl Declinatio ination n Angle Angle Versus Versus Day of Year Year [ [99] . . . . . . . . . . . . . . . . . . . .
17
2.5 Sid Sidere ereal al Tim Timee Shi Shift ft [5] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
17
2.6 Shif iftt in in Sol Solar ar Ti Time me . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.6 Sh
18
2.7 Anale Analemma mma (Y (Yearly early Solar Shift Shift)) . . . . . . . . . . . . . . . . . . . . . . . . .
18
Horizo izont ntal al Coordi Coordinat natee Syste System m [18 [18]] . . . . . . . . . . . . . . . . . . . . . . . . 2.8 Hor
20
Altitude ude Angle Angle vs. Time of Day Day for Tallahas Tallahassee, see, Florida Florida . . . . . . . . . . . . 2.9 Altit
20
vi
2.10 Azim Azimuth uth Angle vs. Time of Day for Tallahasse Tallahassee, e, Florida . . . . . . . . . . .
21
2.11 Alt-A Alt-Azim zimuth uth Mount Mount . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
22
Concentrator ator with an Equat Equatorial orial Mount Mount Syste System m [ [22] . . . . . . . . . . . 2.12 Solar Concentr
23
2.13 Rece Receive iverr Temp. Temp. vs. Conc Concent entratio ration n Ratio . . . . . . . . . . . . . . . . . . . .
24
2.14 Specula Specularr Reflectance Reflectance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
25
2.15 Ray tracing tracing on a parabola . . . . . . . . . . . . . . . . . . . . . . . . . . . .
26
2.16 Int Interce ercept pt factor vs. Apertu Aperture re Diameter Diameter for Ty Typical pical Dish Dish . . . . . . . . . . . .
28
2.17 Cav Cavity ity Receive Receiverr [8 [ 8] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
30
2.18 Radian Radiantt Flux of a Blackbody Blackbody [2] . . . . . . . . . . . . . . . . . . . . . . . . .
32
2.19 Con Conve vection ction Loss vs. Tilt Angle for Open and Cov Covered ered Cavitie Cavitiess . . . . . . . .
35
Tube Boiler [ Boiler [11 11]] . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2.20 Water Tube
36
2.21 Typical Typical Boiling Boiling Curve [10 [10]] . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3.1 Bare Conc Concent entrator rator Wedge . . . . . . . . . . . . . . . . . . . . . . . . . . . .
37 40
3.2 Fully Assem Assembled bled Conc Concent entrator rator . . . . . . . . . . . . . . . . . . . . . . . . . .
41
3.3 Mou Mount nting ing Fram ramee . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
42
3.4 Rece Receive iverr Assemb Assembly ly and and Connect Connection ion Ring Ring . . . . . . . . . . . . . . . . . . . .
43
Energy rgy Abso Absorpt rption ion vs. vs. Aper Apertur turee Diamet Diameter er . . . . . . . . . . . . . . . . . . . 3.5 Ene
45
3.6 Exp Explode loded d View View of of Rece Receiv iver er . . . . . . . . . . . . . . . . . . . . . . . . . . . .
45
3.7 Ass Assem emble bled d Rec Receiv eiver er . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
46
3.8 Ass Assem emble bled d Linea Linearr Actua Actuator torss . . . . . . . . . . . . . . . . . . . . . . . . . . .
47
3.9 Dua Duall Axis Axis Trac Trackin kingg Senso Sensorr . . . . . . . . . . . . . . . . . . . . . . . . . . . .
48
3.10 Sch Schemati ematicc of Test Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . .
49
3.11 Epple Eppley y Nomina Nominall Incide Incidence nce Pyrheliomete Pyrheliometerr . . . . . . . . . . . . . . . . . . . .
49
4.1 Testin estingg Perfo Performed rmed Feb 23, 23, 2009 . . . . . . . . . . . . . . . . . . . . . . . . .
51
4.2 Testin estingg Perfo Performed rmed Marc March h 02, 2009 2009 . . . . . . . . . . . . . . . . . . . . . . . .
51
estingg Perfo Performed rmed Marc March h 04, 2009 2009 . . . . . . . . . . . . . . . . . . . . . . . . 4.3 Testin
52
vii
4.4 Superhe Superheated ated Steam Product Production ion Marc March h 09, 09, 2009 . . . . . . . . . . . . . . . .
53
4.5 Testin estingg Performed Performed March March 02, 2009 2009 Display Display of Salt Bath Phase Phase Change . . .
54
4.6 Testin estingg Perfo Performed rmed Sept 04, 2008 2008 0.7 LPM . . . . . . . . . . . . . . . . . . .
55
4.7 Testin estingg Perform Performed ed Februa February ry 24, 2009 2009 1.0 LPM LPM . . . . . . . . . . . . . . . . .
56
4.8 Testin estingg Perform Performed ed Feb Feb 22, 2009 1.5 1.5 LPM LPM . . . . . . . . . . . . . . . . . . . .
57
4.9 Testin estingg Perform Performed ed Feb Feb 19, 2009 1.3 1.3 LPM LPM . . . . . . . . . . . . . . . . . . . .
57
estingg Pe Perform rformed ed Februar February y 26, 2009 1.1 LPM . . . . . . . . . . . . . . . . . 4.10 Testin
58
4.11 Int Interce ercept pt Factor Factor for Solar 2 . . . . . . . . . . . . . . . . . . . . . . . . . . .
60
4.12 Distr Distributi ibution on of Thermal Losses . . . . . . . . . . . . . . . . . . . . . . . . .
62
viii
ABSTRACT
Concen Con centra tratin tingg sol solar ar pow power er (CS (CSP) P) is a uni unique que ren renew ewabl ablee ene energy rgy tec techno hnology logy.. CSP systems have the ability to provide electricity, refrigeration and water purification in one unit. uni t. Thi Thiss tec techno hnolog logy y wil willl be ext extrem remely ely helpf helpful ul in imp impro rovin vingg the quali quality ty of life life for many people around the world who lack the energy needed to live a healthy life. An economic parabolic dish concentrating system was built at the Sustainable Energy Science and Engineering Center (SESEC) at Florida State University in Tallahassee, Florida. The goal of the projec projectt wa wass to prov provide ide 6.67 kW of therm thermal al ene energy rgy.. Thi Thiss is the amou amount nt of energy required to produce 1 kW of electricity with a conventional micro steam turbine. The system had a price goal of $1000 per kW and must be simple enough to be maintained by non-technical personnel. A 14 m2 fiberglass parabolic concentrator was made at SESEC to ensuree simplicit ensur simplicity y of production and operati operation. on. The concen concentrato tratorr was coate coated d with a highly reflectiv refle ctivee p polyme olymerr film. The cavi cavity ty type receiv receiver er was filled with sodium nitrate to act as a heat storage and trans transfer fer medium. The collect collection ion efficienc efficiency y of the cav cavity ity was estimat estimated ed at 70%. The gross thermal conv conversio ersion n efficie efficiency ncy of the system was 39%, whic which h represen represented ted a 333% improvement over the first concentrator assembled at SESEC. At peak insolation 5.46 kW of thermal energy was produced. The material cost for the system was $3,052.
ix
CHAPTER 1 INTRODUCTION 1.1 1. 1
Moti Motiv vatio ation n
The world is dependent upon energy. People’s energy use directly correlates to their grade of health care, life expectan expectancy cy and educa education. tion. These are important facto factors rs that dete determine rmine a person’ss qualit person’ quality y of life. One quanti quantitativ tativee measur measuree of life qualit quality y is the Human Dev Developm elopment ent Index (HDI). The HDI combines life expectancy, literacy, education and GDP per capita for different countries. Figure Figure 1.1 1.1 displays displays the HDI versus electricity use for different countries and one can see the clear correlation between electricity consumption and HDI. Electricity allows people access to refrigeration for food and medicine, energy for cooking
Figure 1.1: World HDI vs. Electricity Consumption [ [11]
1
and cleaning water, and allows people to read and study at night when there is little work that can be done outside. A small amount of electricity can dramatically change the life of a pers person on who has had none. It is est estima imated ted that the pow power er requi requirem remen entt for basic heal health thy y This is is les lesss than 1% functioning in rural communities is about 0.08 kWh/day/person [12 [ 12]]. Th of an average person’s usage in the United States, yet many people can not afford or do not have access to even this small amount. Nearly two billion people live in rural areas without access to electrical grids. Developing an infrastructure in these remote areas is usually not feasible due to the extreme distance from existing existing electri electricc grids. Buildi Building ng new p pow ower er plants in these areas is not cost effectiv effectivee due to the relatively low electricity consumption. Economical, small-scale, distributed energy systems can fulfill the need and renewable energy is ideally suited for this purpose. Many under-developed areas around the world receive large amounts of sunlight. Northern Africa and Central Asia receive as much as 7.5 kWh/m2 /day. There is great opportunity to use solar power to provide basic energy needs in these regions. The two most prominent solar energy technologies are photovoltaic (PV) and concentrated solar power (CSP). PV sy syst stem emss ar aree ben benefi efici cial al bec becau ause se they they ca can n be sc scal aled ed to an any y size size,, bu butt th they ey are are co cost stly ly an and d solely produce electricit electricity y. CSP systems can pro provide vide electr electricit icity y as we well ll as therm thermal al pow power. er. This thermal power can be efficiently used for cooking, water distillation and absorption refrigeration cycles. The drawback to these systems is that the most efficient solar thermal systems currently have an installation cost of $10,000/kW [13 [ 13]]. An econ economi omicc CSP syste system m could provide rural areas with electricity and energy needs to dramatically improve their quality of life. This project is a continuation of work performed at the Sustainable Energy Science and Engineering Cen Engineering Center ter (SESEC (SESEC)) to build an economic CSP system system.. The first system built at SESEC constant constantly ly provide provided d roughly 2.5 kW of therm thermal al energy energy.. 6.67 kW is the minim minimum um input power power require required d for a micro-ste micro-steam am turbine to produce 1 kW of elec electric trical al power. The goal of this project is to provide at least 6.67 kW of thermal power at an installation cost of $1,000. The system must be easy enough to be constructed and maintained by non-technical personnel.
2
1.2 1. 2 1. 1.2. 2.1 1
Lite Litera ratu ture re Revi Review ew
Br Brie ieff Hist Histor ory y of Sola Solar r Ther Therma mall Po Pow wer
Conc Co ncen entr trat atin ingg so sola larr po pow wer is a me meth thod od of in incr crea easi sing ng sola solarr po pow wer de dens nsit ity y.
Us Usin ingg a
magnifying glass to set a piece of paper on fire demonstrates the basic principle of CSP. Sunlig Sun light ht shin shining ing on the curv curved ed glass is con conce cent ntrat rated ed to a sma small ll p poin oint. t. Whe When n all the hea heatt energy that was spread across the surface of the magnifying glass is focused to a single point, the result is a dramat dramatic ic rise in temper temperature ature.. The paper will reach temper temperature aturess above 451 ◦
F and com combust. bust. CSP has b been een theori theorized zed and contemp contemplated lated by inv inven entors tors for thousands of
years. It is possible that as far back as ancient Mesopotamia, priestesses used polished golden ve vesse ssels ls to ign ignite ite altar fires. The first docu documen mented ted use of con concen centra trate ted d pow power er com comes es fro from m the great Greek scientist scientist Arch Archimede imedess (287-2 (287-212 12 B.C.). Stori Stories es of Arc Archimed himedes es repelling the invading Roman fleet of Marcellus in 212 B.C. by burning their ships with concentrated solar rays were told by Galen (A.D. 130-220) [2 [2]. In the seventeenth century, Athanasius Kircher (1601-1680) set fire to a woodpile at a distance in order to prove the story of Archimedes [2]. Thi Thiss is con consid sidere ered d the begin beginnin ningg of mode modern rn sol solar ar con concen centra tratio tion. n. Sol Solar ar con concen centra trator torss 14]. ]. Th They ey then began being used as furnaces in chemical and metallurgical experiments [14 were preferred because of the high temperatures they could reach without the need for any fuel. Furthe urtherr applic applications ations opened for conc concent entrate rated d pow power er when August Mouc Mouchot hot pionee pioneered red
Figure 1.2: Pifre’s 1878 Sun-Power Plant Driving a Printing Press [2 [ 2]
generating low-pressure steam to operate steam engines between 1864 and 1878. Abel Pifre made one of his solar engines operate a printing press in 1878 at the Paris Exhibition [ 2], 3
Figure 1.3: Boing/SES DECC Dish Stirling System System [3]
but after extensi extensive ve test testing ing he dec declar lared ed the syste system m too expe expensi nsive ve to be fea feasib sible. le. His press 1.2.. Pif Pifre’ re’ss and Mouc Mouchot hot’s ’s res resear earch ch began a bur burst st of grow growth th for sol solar ar is shown in Figure 1.2 concentrators. The early twentieth century brought many new concentrating projects varying from solar pumps pum ps to ste steam am pow p ower er gener generato ators rs to wa water ter dist distill illati ation. on. A 50 kW sol solar ar pump mad madee by Shum Sh uman an and and Bo Boys ys in 19 1912 12 was used used to pu pump mp irri irriga gati tion on water ater from from th thee Nile Nile.. Mi Mirr rror ored ed troughs were used in a 1200 m 1200 m 2 collector field to provide the needed steam [15 [15]]. In 1920 J.A. Harrington used a solar-powered steam engine to pump water up 5 m into a raised tank. This was the first documented use of solar storage. The water was stored for continual use as power for a turbine inside a small mine [15 [15]. ]. Concentrating technology had made a huge leap from the nineteenth century but was halted by World War II and the resulting explosion of cheap fossil fuels. The advantages of solar power lost their luster and the technology would merely inch forward for nearly five decades. Starting in the late seventies and early eighties, solar power came back to the forefront of researchers’ researchers’ agend agendas as with oil and gas short shortages. ages. In 1977 in Shenandoah Shenandoah,, GA, 114 7-meter parabolicc dishes were used to heat a silic paraboli silicon-bas on-based ed fluid for a steam Rankine cycle cycle.. The plan plantt also supplied waste waste heat to a lithium bromide absorpt absorption ion chille chiller. r. The plant plantss total therm thermal al efficiency was 44%, making it one of the most efficient systems ever implemented [ 4]. More modern systems like the Department of Energy’s Dish Engine Critical Components (DECC) project, which was built at the National Solar Thermal Test Facility, consisted of a 89 m2 dish with a peak system electrical efficiency of 29.4% [[33]. Th This is sys syste tem m ut util iliz izes es the hig high h 4
efficiency of the sterling engine to convert the heat generated into electricity. This efficiency is unmatched by any concentrator that utilizes a steam cycle, with one or two working fluids.
1. 1.2. 2.2 2
So Sola lar r S Stea team m Gen Genera erati tion on
Solar thermal systems can utilize the Rankine cycle to produce electricity. This is done by creating creating steam using solar energy and passing it through a turbi turbine. ne. The most common modern technique to produce steam for electricity production or industrial needs has been to utilize a heat transfer fluid. Usually a salt or oil is pumped through a solar field to heat up the flui fluid. d. The hot oil is then passe passed d thr through ough a hea heatt exc exchan hanger ger with with wa water ter to gen genera erate te steam. ste am. The benefi benefitt of a hea heatt tra transf nsfer er fluid is tha thatt it remai remains ns liquid at hig high h tem tempera peratur tures, es, which whic h increases heat transf transfer er and ease of pumping. Oils hav havee a liquid wo working rking range up to 400 C, salts up to 600 C and liquid metals can reach much higher temperatures. ◦
◦
The Solar Total Energy Project in Shenandoah, Georgia used a heat transfer fluid to create crea te electric electricity ity and process steam for a text textile ile factory factory.. The heat transfer fluid used was Syltherm Sylth erm 800, a silic silicon-bas on-based ed fluid produce produced d by Dow-C Dow-Cornin orningg Corpora Corporation. tion. The Sylther Syltherm m carried 2.6 MW of heat from 114 7 m diameter concentrating parabolic dishes. The fluid was pumped to cav cavity ity rece receive ivers rs that we were re placed at the focal p point oint of each dish. The collec collector tor efficien effic iency cy of the conce concent ntrat rators ors was 76.9 76.9%. %. Thr Three ee hea heatt exc exchan hanger gerss we were re the then n uti utiliz lized ed to preheat, boil and superheat water before it was passed through a steam turbine. A portion of the steam generated was used directly by the factory. The remaining hot water was used as a heat sink for a lithium bromide absorption chiller. The electricity generation efficiency was 11.8%. The total heat work efficiency for all of the processes was 44.6% [4 [ 4]. Figure Figure 1.4 shows the process flow and corresponding state for the system. Even though using an oil or salt as a heat transfer fluid is the most popular choice, water wa ter is sti still ll used in som somee cas cases. es. The John Johnson son and Joh Johnso nson n Sol Solar ar Proces Processs Hea Heatt Sys System tem used pressurized pressurized wat water er as the heat transfer fluid in a parabolic trough plant. The facili facility ty had 1070 m2 of parabolic troughs which heated water pressurized at 310 psi to 200 C. ◦
The water was stored in a tank where flash steam was created 4 times a day for industrial processes. process es. To create steam, the press pressurize urized d water was throttl throttled ed down to 125 psi wher wheree it vapor aporize ized d add additi itiona onall fe feed ed wa wate ter. r. The satur saturate ated d ste steam am tha thatt wa wass prod produce uced d wa wass sen sentt to do Figure 1.5 1.5 shows shows the process work. Thermal collection efficiency for the system was 30% [4 [4]. Figure flow and corresponding thermodynamic state for this facility. 5
Figure 1.4: Proces Figure Processs Flo Flow w and Correspondi Corresponding ng Therm Thermodynamic odynamic State for Shena Shenandoah ndoah T Total otal Energy Energ y Project Project [4]
6
Figure 1.5: Process Flow and Thermodynamic State for Johnson and Johnson Solar Facility [4]
Direct steam generation (DSG) can be a more efficient and economic way of producing steam from solar collectors. Eliminating storage and heat exchangers decreases losses, capital investment and maintenance. The most famous case of a solar facility producing direct steam was the Solar 1 plant built in Barstow, CA. Solar 1 was a solar power tower that produced 10 MW of electricity in 1982. The receiver sat nearly 100 m above ground and was powered by 1818 39 m2 co coll llec ecto tors rs.. Th Thee 13 13.7 .7 m hi high gh an and d 7 m diam diamet eter er rec recei eiv ver was ma made de of 69 mm alloy alloy tubes. The tubes were plac placed ed ve verti rtical cally ly,, we welde lded d tog togeth ether er and coa coated ted with an 1.6 shows an overall view of the Solar 1 facility. The surface of the absorptive paint. Figure Figure 1.6 receiver reached temperatures up to 620 C. Water was pumped through the tubes where it ◦
was vaporized and superheated to 516 C. The steam was then passed through a turbine to produce electricity. The maximum net monthly electrical efficiency was 15%. A more modern example of DSG comes from the INDITEP project built near Seville, Spain Spa in in 200 2003. 3. The 5 MW parabo parabolic lic trou trough gh p pow ower er plan plantt use usess DSG to run stea steam m tur turbin bines. es. The pilot plant is testing the efficiency of both saturated and superheated steam production. For th thee sa satu tura rate ted d ca case se,, water ater is pu pumpe mped d at 77 ba barr th thou ough gh a sola solarr co coll llec ecto torr fie field ld.. Th Thee steam/water mixture exits the collectors with a quality of 0.85 at 285 C. A steam separator ◦
collects the steam and sends it to the turbine. The liquid water is recycled and passed back through the collector field. For the superheated case, the collected steam is passed through an additional set of collectors where it is superheated to 400 C. Superheating the steam ◦
increases the turbine efficiency. Figure 1.7 Figure 1.7 shows shows the two configurations. Collection efficiency 7
Figure 1.6: Overall View of Solar 1, at Barstow, CA [4]
(a) Saturated Steam Cycle
(b) Superheated Steam Cycle
Figure 1.7: Solar Configuration of INDITEP Project Seville, Spain [ Spain [66]
for the saturated saturated steam configu configuratio ration n was 66.9% and 65.1% for the superhea superheated ted case. [6]. The reduced efficiency of the superheated case can be attributed to the high temperature of the receiver line, which increases thermal losses. The net electrical efficiency of the plant was 16.4% for the superhe superheated ated case and 16.2% for the satur saturated ated case. The increa increased sed turbine efficiency of the superheated case is offset by the increased thermal losses during collection. The net increase in efficiency is only 0.2% more for the superheated case than the saturated case.
8
Figure 1.8: Schematic of Solar 1
1.3 1. 3
Prev Previo ious us Work ork
An economic solar dish system was built in 2006 by graduate student C. Christopher Newton. This first dish system built at SESEC , nicknamed Solar 1, attempted to complete the same goal as the one addressed in this paper, which is to provide 1 kW of electrical energy ene rgy for an ins instal tallat lation ion cost of onl only y $10 $1000 00 dol dollar lars. s. Fig Figure ure 1.8 1.8 shows a schematic of the the system. system. A par parabol abolic ic con conce cent ntrat rator or refl reflect ected ed sol solar ar rad radiat iation ion to a cen centra trall rec receiv eiver. er. The receive rece iverr produce produced d int intermit ermitten tentt steam that was injected into a steam turbine turbine.. The turbine was connected to an electric generator that produced electricity. The concentrator used was a fiberglass Channel Master satellite dish with an aperture 1.9 shows shows the operational conc concent entrator rator.. Solar 1 pivote pivoted d on a diameter diamet er of 3.66 m. Fig Figure ure 1.9 steell alt-az stee alt-azimut imuth h type frame. Tw Twoo indepen independen dentt linea linearr actua actuators tors move the dish throu throughout ghout the day. day. The actuato actuators rs were cont controlle rolled d autom automatical atically ly by a set of photophoto-sensin sensingg mo modules dules.. The modules consisted of light sensing LEDs that sent signals to the actuators when the module was not oriented normal to incoming solar radiation. The power for the sensors and actuators was provided by 2 small thin film photovoltaic panels, which charged two 24 V deep cell batteries. The reflective material used to coat the fiberglass surface was aluminized mylar, which has an optical reflectivity of 76%. The system utilized an external type receiver that had an absorber diameter of 15 cm.
9
Figure 1.9: Concentrator Assembly for Solar 1 [8 [ 8]
The absorber was coated with a high temperature black paint, which has an absorptivity and emissiv emissivit ity y both equa equall to 90%. The rec receiv eiver er was filled filled wit with h dra draw w sal saltt to act as a hea heatt storagee and heat transfer medi storag medium. um. Dra Draw w salt is a 1:1 molar ratio of potassium nitr nitrate ate and sodium nitrate. The melting temperature of this eutectic mixture is 223 C. ◦
The heat exchanger in the receiver consisted of an abbreviated water tube boiler, which consisted of copper tubes coiled around the outer rim of the salt bath. They connected to a water wat er drum in con contact tact with the flat absorber at the base of the rece receive iver. r. The steam in the water wat er drum exit exited ed through copper tubes in the cente centerr of the rece receive iver. r. Figur Figuree 1.10 shows 1.10 shows the Solar 1 receiver assembly. The maximum steady state thermal output for the system was 1 kW. The resulting thermal ther mal con conve version rsion efficienc efficiency y was estimated estimated at 9.03%. To produce electr electricit icity y, wat water er was held in the receiver where it pressurized and released in intervals at a 6.67% duty cycle. The maximum electrical conversion efficiency was estimated at 1.94%, which equated to a turbine turbi ne efficienc efficiency y near 15%. The maxim maximum um gross electric electricity ity product production ion by Solar 1 was 220 W. The majority of the thermal losses for Solar 1 are believed to be from three major factors. First,, the concen First concentrato tratorr efficiency was extre extremely mely poor. The reflect reflected ed focal area was muc much h largerr than the receiv large receiver. er. Muc Much h of the radiatio radiation n was refle reflected cted onto the side of the receiv receiver, er, which was heavily insulated and thus lost. Second, the reflective material had incredibly poor weathering abilities. In the few months before testing the surface was exposed to the elements and pollution pollution in the atmos atmosphe phere. re. Dur During ing final testi testing ng the myl mylar ar wa wass vis visibl ibly y clo cloudy udy and 10
Figure 1.10: Receiver Assembly for Solar 1 [8 [8]
turning yellow, clearly indicating a greatly reduced reflectivity. The last major factor hurting system efficiency was the absorber. The flat plat absorber, although compact, was extremely exposed and lost tons of energy to the environment through convection and radiation. This type of receiver is extremely sensitive to wind, and the results show dramatic drops in receiver temperature with any wind at all. These three factors are primarily responsible for the low thermal conversion energy and must be improved in the second concentrator if it is to be successful.
1.4
Projec oject Goal
The first concentrator system built at SESEC, Solar 1, attempted to satisfy all of the resear res earch ch objecti objective vess hel held d by SESEC for thi thiss proje project. ct. Mir Mirror ror and boil boiler er ine ineffici fficienc encies ies held the system from producing enough steam to continuously continuously run a micr micro-ste o-steam am turbine. The minimum thermal input to run a micro-steam turbine is near 5 kW, 5 times what was produce prod uced d by Solar 1. Furt urther hermor more, e, a mic microro-ste steam am tur turbin binee wit with h 15% the therma rmall effic efficien iency cy requires 6.67 kW of thermal energy to produce 1 kW of electricity. For the project discussed in this thesis, it was decided to focus only on the thermal energy generation and neglect electrical generation. Once enough thermal energy is being produced, a micro-steam turbine 11
will then by implemented. The goal of this project is to rebuild the concentrator and receiver assembly to produce 6.67 kW of thermal energy for an installation cost of $1000 dollars.
12
CHAPTER 2 BACKGROUND 2.1 2. 1
Intr Introdu oduct ctio ion n to the the Sol Solar ar S Spec pectr trum um
The Sun is an enormous atomic reactor, constantly fusing hydrogen nuclei together to form helium. In the process it gives off huge amounts of energy in the form of electromagnetic radiation. This radiated energy can travel infinite distances to nearby planets, such as Earth, or planets millions of light years away. The Sun emits 4 1026 W of energy and only 1. 1 .7 1017
×
×
W reaches the Earth [14 14]. ]. The Eart Earth h rec receiv eives es less than one bil billio liont nth h of the Sun Sun’s ’s powe powerr output. Electromagnetic (EM) radiation is a self propagating wave that carries energy through space. spa ce. EM radia radiatio tion n is class classifie ified d into into types acc accord ording ing to the fre freque quency ncy of the wa wave ve.. At less than 1 pm, gamma rays encompass the smallest wa wavel velength engthss of EM radiation radiation.. Radio wave wa vess hav havee the longe longest st wav waveleng elengths ths and start at 1 mm. In b betw etween een these tw twoo limit limitss there are classifications such as microwaves, terahertz, infrared, visible, ultraviolet and X-rays. The wavelength of the radiation emitted by the sun varies and can be approximated as a black body at 5800 K[ K[2]. Fi Figu gure re 2.1 shows the relative energy intensity as a function of wave wa velengt length h for Solar radiation. Only 43% of the radiat radiation ion emitted is visib visible le light, betwe between en the wave wavelen length gthss of 380 380-75 -7500 nm. Of the Sol Solar ar spec spectru trum m 8% of the radia radiatio tion n is ult ultra ravio violet let,, less than 380 nm, and 49% is infrared, 750 nm to 1 mm. Solar Sol ar irr irradi adiati ation, on, Ir , is defi defined ned as ext extrat raterr errest estria riall rad radiat iation ion tha thatt str strik ikes es the Ear Earth’ th’ss atmosp atm ospher here. e. Thi Thiss valu aluee has been mea measur sured ed as 136 13677 W/m2 [4]. Thi Thiss amoun amountt of rad radiat iation ion varie ariess th thro roug ugho hout ut the the yea ear, r, du duee to Ea Eart rth’ h’ss sl sligh ightl tly y elli ellipt ptic ical al orbi orbitt an and d ax axia iall tilt tilt..
As
extraterrestrial radiation reaches Earth, the atmosphere will reflect some of the radiation back into space, as well as absorb some of the energy. The atmosphere reduces the amount of incide incident nt solar rad radiat iation ion,, whi which ch is ref referr erred ed to as ins insola olatio tion. n. Ins Insola olatio tion n can be cla classi ssified fied 13
Figure 2.1: Solar Spectrum at the Surface of the Earth
into into tw twoo ty types: pes: dir direct ect norma normall ins insola olatio tion n and diffus diffusee ins insola olatio tion. n. Dir Direct ect norma normall ins insola olatio tion n is the radiation that tra travel velss dire directly ctly through the atmosphere atmosphere without inter interfere ference. nce. Diffus Diffusee insolation is scattered by particles in the atmosphere, eventually hitting the Earth at random angles, an effect that can be observed on a cloudy day. For reasons which will be explained below,, this project deals exclusi below exclusive vely ly with direct normal insolation. insolation. The av average erage amoun amountt of direct normal insolation throughout the day for Tallahassee can be seen in Figure 2.2 Between 09:00 and 17:00 the average is about 450 W/m 2 .
2.2 2. 2
Sola Solar r Geom Geomet etry ry
The Earth is rotating about the Sun in an elliptical orbit at a rate of 1 rotation every 365 + 1/ 1/4 da days ys.. Th Thee Ea Eart rth h is on av aver erag agee 1.5
× 1011meters away from the Sun [18 [18]].
The
Earth’s axis of rotation is tilted at an angle of 23.45 with respect to its orbital plane. This ◦
tilt causes the angle at which sunlight hits the Earth to change throughout the year. For the months of June through September the northern hemisphere is tilted toward the Sun and the Sun’s rays shine normal to the surface. In the winter months, the northern hemisphere is tilted away from the Sun and sunlight hits the surface at lesser angles. As the incident sun angle moves away from zenith angle (θ ( θ) equal to zero(directly overhead), sunlight is spread 14
Figuree 2.2: Av Figur Average erage Insolation Insolation for Tal Tallahass lahassee ee [7]
over over a lar larger ger area and is les lesss inten intense. se. Thi Thiss rel relati ations onship hip is des descri cribed bed in uni units ts of Air Mass (AM). An Air Mass equal to 1 is when the Sun is directly overhead shining normal to the surface, surfa ce, called zenith. Equat Equation ion 2.1 2.1 gives gives the equation for Air Mass in terms of zenith angle ]. The equation is applicable for angles between 0 and 70 . At a zenith angle of 60 (θ) [18 [ 18]. ◦
◦
◦
the Air Mass is 2, which means incident radiation is half of what it is when the Sun is at zenith. AM =
1 cos((θ) cos
(2.1)
The position of the Sun cha changes nges conside considerably rably throug throughout hout the year. The differe different nt types of solar motion are described in the following sections.
2. 2.2. 2.1 1
Di Diur urna nall Rota Rotatio tion n
The first and most basic motion of the Sun is diurnal, or daily, rotation. The sun makes an apparent pass around Earth every 24 hours, rotating around an axis called the celestial axis. This axis differs at times from our magnetic axis by as much as 6 . If an observer were ◦
standing on the celestial equator during the equinox, the sun would pass directly overhead going goi ng from East to West est.. If the obser observe verr were to line up a tel telesc escope ope on an axis point pointed ed directly at celestial north, the rotation of the telescope following the Sun would be 1 every ◦
4 minutes, or 360 every day. ◦
15
Figure 2.3: Declination Angle at Summer Solstice[8 Solstice[8]
2. 2.2. 2.2 2
De Decl clin inat atio ion n
The apparent height of the Sun in the sky is described by declination. Figure 2.3 2.3 shows shows declination during the summer solstice when it is at its maximum of 23.45 . As th thee Ear Earth th ◦
follows its orbit, the declination will drop to a minimum of -23.45 . Dec Declin linati ation on ma make kess ◦
2.4.. Declination can be referred to as the apparent this cycle every year as seen in Figure Figure 2.4 longit lon gitudi udinal nal mov moveme ement nt of the sun in the sky thr throug oughou houtt the yea year. r. The next sect section ion will describe the annual latitudinal shift.
2.2 2. 2.3
So Sola lar r Ti Time me
When the Earth is moving away from the Sun in its elliptical path, it takes a little longer for for th thee ea eart rth h to rota rotate te and view the Sun at the the same same posit positio ion n in the sky sky. Th This is effe effect ct is displayed in Figure Figure 2.5 2.5.. The difference difference in time it takes betw between een position 2 and 3 giv gives es the difference in solar time to mean time. Mean time is the time we keep with clocks on Earth. Solar time will drift behind and then ahead of mean solar time. This shift in time is shown 2.6.. Thi Thiss eff effect ect shif shifts ts the p posi ositio tion n of the sun by as mu much ch as in Figure Figure 2.6 year and it must be accounted for when tracking the sun.
± 4
◦
throughout the
Other smaller factors affecting the position of the Sun are refraction and dispersion, which are caused by light traveling through the atmosphere. Their effect causes the actual position to differ from the apparent apparent position. The differ difference ence can be as much as 34 arc minut minutes es during 16
Figure 2.4: Declination Angle Versus Day of Year [9 [ 9]
Figure 2.5: Sidereal Time Shift [5 [5]
sunrise and sunset, which is slightly larger than the viewed size of the sun. The net effect of the two major sources of solar shift can be seen in Figure 2.7 2.7.. This his unsymm uns ymmetr etrica icall figu figure re 8 is refer referred red to as an ana analem lemma. ma. It show showss the posit position ion of the Sun at the same mean time every day of the year. Ther Theree are 3 paramete parameters rs that create this image. The two greatest factors, declination and equation of time, have already been discussed. The third parameter distorts the figure-8 from being symmetric horizontally and is caused by the difference in angle between the apse line and the line of solstices. 17
Figure 2.6: Shift in Solar Time
Figure 2.7: Analemma (Yearly Solar Shift)
18
2.2.4 2.2 .4
Cel Celest estial ial Coor Coordina dinate te systems systems
In ast astro rono nom my it is im impor porta tant nt to pinpo pinpoin intt th thee ex exac actt pos posit ition ion of bodi bodies es in th thee sky sky. Astron Ast ronome omers rs do thi thiss by first imagi imaginin ningg a gia giant nt orb whi which ch surr surroun ounds ds the eart earth. h. Thi Thiss orb is called the celestial celestial sphere sphere.. All stars and planets are imagine imagined d to be a point on this sphere. To find a particular star two angles are needed to denote the position of a star on the celestial sphere, spher e, which can be imagin imagined ed as tw twoo plane planes. s. Where these plane planess int interse ersect, ct, a vec vector tor is born which whic h ”points” to a particula particularr location on the celes celestial tial sphere sphere.. The length of the vector is not important important for our applicat application, ion, only the dire direction ction of the b body ody in the sky sky.. The celes celestial tial coordinate coordin ate syste system m giv gives es coordin coordinates ates based on differ different ent refere reference nce frames. Eac Each h syste system m tak takes es a different major axis and plane, and describes the location of objects in space using two coordinates. There are two different systems that will be discussed in this review: horizontal and equatorial. Horizontal Coordinates
The first coordinate system discussed here is Horizontal, or alt-azimuth. In this system a person standing on the surface of the Earth takes their perceived horizon to be a flat plane. The line normal to this surface, directly above the person, is called the zenith. The altitude angle describes the angle a body is above the horizontal plane. Everything above the plane is positive, and everything below the plane is negative. The other parameter, azimuth angle, describes the angle a body is from a plane parallel to zenith connecting celestial north-south. This plane is chosen to be 0 azimuth. Positive angles run in the clockwise direction, making ◦
2.8,, we can see see γ s represents positive azimuth, and and γ this a left-handed left-handed syste system. m. In Figure Figure 2.8 represents negative azimuth. Also in Figure 2.8 Figure 2.8 we we see that α that αs represents the altitude angle. With these two angles, the direction normal to the Sun can be found. Figures 2.9 and and 2.10 2.10 show the altitude and azimuth angle for the Sun for the equinox and solstices. Equatorial Coordinates
The second coordinate system reviewed reviewed is calle called d equat equatorial. orial. In the equatori equatorial al coordin coordinate ate system, syste m, the fundam fundament ental al plane lies along the celestial celestial equator equator.. The normal vec vector tor in this system syste m is the rotation rotational al axis of the eart earth. h. The celest celestial ial sphere is parti partitione tioned d into sections analogo ana logous us to our geogr geograph aphica icall coor coordin dinate ates. s. Dec Declin linati ation on (δ ) describes the angle above or below the celestial equator like latitude. The other coordinate, which is likened to longitude, 19
Figure 2.8: Horizontal Coordinate System [18 [18]]
Figure 2.9: Altitude Angle vs. Time of Day for Tallahassee, Florida
20
Figure 2.10: Azimuth Angle vs. Time of Day for Tallahassee, Florida
is called hour angle (H). The hour angle uses units of hours, minutes, and seconds to partition the 360 spher spheree into 24 hours. One hour angle equa equals ls 15 of arc. Following these coordinates ◦
◦
simplifies the apparent motion of the sun. The change in hour angle is constant throughout the day at 1 every 4 minutes. The declination changes by no more than 4 tenths of a degree ◦
per day. 2.2 and and 2.3 2.3.. These To convert between the two coordinate systems we may use Equations 2.2 equations give altitude and azimuth angle for the sun given the observers latitude (Φ), declination (δ (δ ) and hour angle (H) [16 [ 16]. ]. sinα = sinα = sinδ sinδ cosγ
× sin sinΦ Φ + cosδ + cosδ × cos cosΦ Φ × cosH
× cosα cosα = = cos cosΦ Φ × sinδ − sin sinΦ Φ × cosδ × cosH
(2.2)
(2.3)
Using Equations 2.2 Equations 2.2 and 2.3 we can easily plot the motion of the sun throughout the entire year.
21
Figure 2.11: Alt-Azimuth Mount
2.2.5 2.2 .5
Cel Celest estial ial Trac rackin king g M Moun ounts ts
The alt-azimuth system requires two independent axes of tracking, one which rotates on the horiz horizonta ontall plane and a second that rotates vert vertically ically.. These motio motions ns are shown in 2.11.. The mec mechan hanics ics of suc such h a mou mount nt are very simp simple, le, and can easi easily ly be des design igned ed Figure 2.11 Figure to handle large systems. Most modern concentrators use horizontal mounts for this reason. The drawback is that horizontal mounts require computers to control the positioning of the axis to align with the Sun. Equatorial mounts also have two axes of tracking, which are declination and hour angle, but the they y als alsoo have have the add added ed req requir uireme ement nt that the hour ang angle le mu must st rot rotate ate parall parallel el to the Ear Earth’ th’ss rot rotati ational onal axi axis. s. Thi Thiss add added ed req requir uireme ement nt cre create atess eng engine ineer ering ing hu hurdl rdles es for thi thiss type type of mou mount nt.. The decl declina inatio tion n axi axiss is not inde independ penden entt of the hou hourr ang angle le or the cele celesti stial al rotationa rotati onall axi axis. s. As such such,, the siz sizee of the moun mountt must be lar larger ger than the conc concen entra trator tor.. A solar concentrator utilizing an equatorial mount is displayed in Figure 2.12 Figure 2.12.. The equatorial mount mou nt is ve very ry cum cumbers bersome ome and beco becomes mes expe expensi nsive ve for lar large ge con conce cent ntrat rators ors.. His Histor torica ically lly though, telescopes have used the equatorial mount because of the simple tracking motion. Despite the difficulties with equatorial mounts, they do have appealing aspects to them. Because of the extremely small declination movement in equatorial coordinates, the motion requir req uired ed to tra track ck a cel celest estial ial body eac each h day day is reduc reduced ed to one one.. Wit With h a sin single gle simp simple le motor or mechanism that can rotate the hour angle at a rate of 15 per hour, an object in the sky ◦
can by effectively tracked. No computers with complex equations to describe the movement
22
Figure 2.12: Solar Concentrator with an Equatorial Mount System [2 [ 2]
are necessary necessary..
2.3 2. 2.3. 3.1 1
Optics
Co Conc ncen entra trati tion on
2.4 [2]. The concentration ratio of a parabolic concentrator (PC) is given by Equation 2.4 Ac is the projected area of the concentrator, and A r is the receiver area. C R =
Ad Ar
(2.4)
The maximum possible concentration ratio for a 3-D collector with a source half angle of θ of θ is given in Equation 2.5 Equation 2.5 [ [17 17]].
1
C Rmax =
(2.5)
sin θ2 19]]. For a concentrator using the Sun as its source, the half angle is approximately 1/ 1/4 [19 ◦
This means the maximum concentration with a 3-D concentrator is approximately 5000. The higher the concentration ratio, the higher the maximum temperature will be. Achieving high temperatures temperatures is a unique charac characteri teristic stic to PCs. Ideal Ideally ly,, the maxim maximum um temperatur temperaturee achiev ach ievable able by a conc concent entrator rator is the sourc sourcee temperature temperature.. Equat Equation ion 2.6 gives the absorber temperature T temperature T aabs bs for a PC neglecting conduction and convection. T abs = T = T s [(1
− η) ×
ηopt abs 23
2 1/4
× C R × sinθ ]
(2.6)
Figure 2.13: Receiver Temp. vs. Concentration Ratio
given the following values: T SS = Temperature of the source (K) η = Efficiency of transferring heat to working fluid ηopt = Optical efficiency of concentrator system abs = Emissivity of absorber The maximum temperature of a receiver versus concentration ratio is graphed in Figure 2.13 Figure 2.13 under usual concentrator conditions.
2. 2.3. 3.2 2
Refle Reflect ctio ion n
The manipulation of light and other EM waves can be achieved in a number of ways; it can reflect off of a surface, transmit through a material without effect, or be absorbed by the surface. surface. With absorpt absorption, ion, the EM wav waves es increase the energy of the imping impinging ing surface surface.. 2.7 describes describes a particular surface depending upon it’s coefficient of reflectivity, reflectivity, ρ, Equation 2.7 Equation transmittance, τ transmittance, τ ,, and aabsorptivity bsorptivity,, γ . Each coefficient describes the relative effect the surface has on the radiation that impinges upon it. For example a surface with τ = τ = 1 would transmit all radiation radiation through the material and none wou would ld be refle reflected cted or absorbed. Coeffici Coefficient entss of a material change depending on radiation wavelength, surface temperature and incident angle. 24
Figure 2.14: Specular Reflectance
To simplify analysis, coefficients are averaged for regions of interest and given a single value. ρ + τ + τ + γ = 1
(2.7)
There are two types of reflection; diffuse and specular. A diffuse surface reflects the light rayss in a num ray number ber of diffe differen rentt direction directions. s. Specul Specular ar reflector reflectorss are “mirror” like and the angle of incidence (θ (θi ) equals the angle of reflection (θ (θr ), as giv given en by Sne Snell’ ll’ss law. law. For a spec specula ularr reflecting surface, a method called ray-tracing can accurately determine the path of reflected EM waves. waves. Ra Ray y tra tracin cingg red reduce ucess a lar large ge field field of partic particles les tra trave velin lingg thr throug ough h a sys system tem to a discre dis crete te nu numbe mberr of nar narro row w beam beamss cal called led rays. rays. Sol Solvin vingg the path of the these se ra rays ys acc accura urate tely ly 2.14 shows shows the path of light predicts predi cts the move movemen mentt of an entire field of particl particles. es. Figur Figuree 2.14 striking a mirror for a specular reflecting surface. A parabola is a curve generated by the intersection of a right circular cone and a plane parall par allel el to an ele elemen mentt of the cur curve ve.. Equ Equati ation on 2.8 2.8 describes a parabola that is symmetric about the y-axis. y = ax = ax2
(2.8)
EM rays incoming parallel to the symmetric axis intersect at a single point called the focus or focal point. point. This This poin pointt is fou found nd by using Equa Equatio tion n 2.9 2.9.. Ra Ray y tra traci cing ng on a pa para rabol bolaa is displayed in Figure Figure 2.15 2.15.. In this figure the rim angle, ((φ φ), is defined as the incidence angle of light striking the outer rim of the concentrator. f =
1 4a
25
(2.9)
Figure 2.15: Ray tracing on a parabola
2. 2.3. 3.3 3
Op Opti tica call Effici Efficien ency cy
An ideal concentrato concentratorr focuses all ligh lightt that hits its surface to a single point point.. In realit reality y, errors err ors in the opti optics cs of the dish ske skew w the ligh lightt inc increa reasin singg the size of the focal beam. Tota otall angular error due to optics can be calculated using Equation 2.10 2.10 [ [77]. σtot =
(2
× σconc)2 + (σ (σtracking )2 + (σ (σrefl )2 + (σ (σabs )2 + (σ (σsun )2
(2.10)
where angular errors are due to: σsun - suns rays not being perfectly parallel σtracking - concentrator alignment with sun σconc - concentrator surface irregularities σrefl - non specular reflector σabs - receiver alignment with focal point By determining the total angular error in a concentrator we can accurately determine the flux capture fraction, fraction, Γ. The flux capt capture ure fractio fraction n relat relates es the percen percentage tage of ligh lightt refle reflected cted 2.11 gives gives the value to a desir desired ed circula circularr area at the focal point with diameter (d). Equat Equation ion 2.11 of the flux capture fraction. Γ =1
− 2 × Qx
where: Qx = f = f x
× (b1 × t + b + b2 × t2 + b3 × t3 + b4 × t4 + b5 × t5 ) 2
f x =
x √ 21 π × exp exp((− ) 2 × 26
(2.11)
t =
1 (1 + r + r
× x)
x = n/ = n/22 n = arctan = arctan((d
cos(φ) × (2/σ × cos( (2/σtot) p )
and: r = . = .2316419 2316419 b1 = 0.379381530 b2 =
−0.356563782
b3 = 1.781477937 b4 =
−1.821255978
b5 = 1.330274429 d = Receiver diameter φ = Rim angle p = p = Distance from focal point to rim of the paraboloid The standard angular error found in state of the art paraboloid concentrators, (σ ( σ ), is 6.7 mrad [7]. This translates to an intercept factor of 1 at an aperture diameter of 4.8 in. This mrad means that 100% of the reflected radiation is focused to a circular area with a diameter of 4.8 in, cen centere tered d at the focal point. As the total angula angularr error increase increasess it tak takes es an incr increasing easing aperturee diame apertur diameter ter to captu capture re the same amount of radiation. Figur Figuree 2.16 shows 2.16 shows the intercept factor for differing aperture diameters. The optical efficiency of a concentrator system (η (ηopt ) describes a CPCs ability to reflect and absorb solar radiation. The optical efficie efficiency ncy is the product of the reflect reflection ion of radia radiation tion off the reflective surface, η surface, ηrefl , and absorption of radiation to the receiver, η receiver, ηabs . This eefficie fficiency ncy is shown in Equation 2.12 Equation 2.12.. = η refl ηopt = η
× ηabs
(2.12)
A perfect system would have a reflectivity coefficient for the concentrator equal to 1 and an abs absorp orptiv tivit ity y coeffi coefficie cient nt for the absor absorber ber als alsoo equ equal al to 1. Re Reflec flector torss wil willl alw alway ayss abs absorb orb small amounts radiation radiation and absorbers will refle reflect ct some radia radiation tion reducing the total system efficiency. 27
Figure 2.16: Intercept factor vs. Aperture Diameter for Typical Dish
Combining the optical efficiency with the flux capture fraction of a dish provides the collection efficiency efficiency,, η collection . The collection efficiency describes the CPCs ability to collect available insolation and absorb that radiation into the receiver. Collection efficiency is found using Equation Equation 2.13 2.13.. = η opt ηcollection = η
2.4 2. 4
×Γ
(2.13)
Rece Receiv iver er Desi Design gn
The receiver is the part of the system that converts solar radiation to heat energy in a working fluid. The receiver consists of an absorber, heat exchanger and possibly heat storage. The absorber absorber is the imping impinging ing sur surfac facee for reflec reflecte ted d sol solar ar rad radiat iation ion to str strik ike. e. Rad Radiat iation ion is absorbe abs orbed d in into to the absor absorber ber mat materi erial al as hea heat. t. The heat exch exchang anger er tra transf nsfers ers the ene energy rgy to a working fluid that carries carries the energ energy y out of the receiv receiver. er. Equat Equation ion 2.14 2.14 shows shows an energy balance for a receiver. = Q abs Qout = Q
− Qloss
where: Qout = Useful energy transfered to working fluid 28
(2.14)
Qabs = Energy collected by the absorber Qloss = Receiver energy losses and total receiver efficiency η efficiency η rec , is given by Equation 2.15 Equation 2.15.. Qout ηrec = Qabs
2. 2.4. 4.1 1
(2.15)
Abso Absorbe rber r
Theree are 2 ty Ther types pes of absorbers; external and cavit cavity y. An external absorber is esse essenti ntially ally a flat plate. The refle reflected cted radi radiation ation impin impinges ges on the plate and heats the surface surface.. Exte External rnal absorbers are extremely simple and cheap but have many inefficiencies. External absorbers are directly directly expose exposed d to am ambie bient nt air and at hig high h tem tempera peratur tures es con conve vect ction ion los losses ses can b bee extre ext reme. me. Tempe emperat rature ure stra stratifi tificat cation ion will also occur insid insidee the rec receiv eiver er.. Wit With h all of the heating being done solely on one end of the receiver, the internal temperature will vary depending on the distance away from the absorber surface. Consequently, the heat exchanger efficien effic iency cy can be red reduce uced. d. The last dra drawba wback ck disc discuss ussed ed is effe effecti ctive ve abso absorpt rption ion.. For an external absorber, if any radiation is reflected off the surface, it is reflected away from the absorber and lost. These problems can be reduced by using a cavity absorber. 2.17 shows shows a rece receive iverr with a cav cavity ity type absorber absorber.. The absorber is rece recessed ssed inside Figure 2.17 Figure the receiver. Solar radiation is reflected by the concentrator through an opening, referred to as the receiver aperture, and is collected on the absorber surface. Most modern receivers are of this type type.. Cav Cavity ity absorbers are more expensiv expensivee and complicat complicated ed than exte external rnal absorbers, but are muc much h more effici efficient ent.. If any radia radiation tion reflect reflectss off the absorber surfac surfacee it is refle reflected cted onto onto anothe anotherr par partt of the abs absorbe orber. r. In this wa way y, the absor absorber ber has an incre increase ased d ch chanc ancee of absorbing absorb ing radiation each time light strike strikess its surface. This effect is quan quantifie tified d in Equation 2.16. cavity ity rece receive iverr with an inner surfa surface ce 5 times the aperture area area,, and 2.16. If we assumed a cav surface surfa ce absorptiv absorptivity ity of 0.7, the effec effectiv tivee absorp absorptivit tivity y wou would ld increase to 0.92. The effectiv effectivee absorptivity of a cavity is always higher than that of a flat plate. αs (2.16) αs + (1 αs ) Aa /Ai A receiv receiver er is compl completel etely y insulated excep exceptt for the absorber. As absorber size increases increases,, αef f =
−
·
energy intercepted intercepted from the concentrator increases, as given by the intercept factor. Although 29
Figure 2.17: Cavity Receiver [[88]
as thi thiss hap happens pens,, the therma rmall los losses ses to the envi environ ronmen mentt als alsoo inc increa rease. se. To opt optimi imize ze abs absorbe orberr 2.17 is used. used. In this equa equatio tion n abs absorbe orbed d rad radiat iation ion ((Q size Equation Equation 2.17 Qin ) is a function of the characteristic absorber diameter “L” and the angular error of the concentrator “Γ”. Energy losses (Q (Qloss ) are primarily a function of absorber diameter for an ideal cavity. Qopt = Q = Qin (L, Γ)
2. 2.4. 4.2 2
− Qloss(L)
(2.17)
Heat Heat St Stor orag age e
Many app Many applic licati ations ons for wo worki rking ng flui fluids ds req requir uiree a ste steady ady ther thermal mal output output.. Hea Heatt ma may y be stored store d in the receiv receiver er to act as a damper for heat transfer transfer.. When momen momentary tary cloud cov cover er blocks energy input to the system the working fluid can obtain energy from the stored heat. If a receiver is producing steam to run a turbine and the thermal input decreased, the turbine bladess could b blade bee damaged if liquid water is injecte injected. d. A system equipped with heat storage could generate the steam after decreased solar input, decreasing the risk of outputting liquid water. Energy Ene rgy can be sto stored red in an any y mat materi erial al by incr increas easing ing its tem tempera peratur ture. e. The amoun amountt of energy ene rgy store stored d is pro proport portion ional al to the mass, hea heatt cap capaci acity ty and tem tempera peratur turee inc increa rease. se. The storage capacity of a mass can be made more efficient if it undergoes a phase change. This 30
effectt incr effec increases eases thermal storage without increasi increasing ng rece receive iverr temper temperature ature.. The total ener energy gy stored in a material that undergoes a phase change is given in Equation 2.18. 2.18. ∗
Qs = m = m[[C solid solid (T
∗
− T 1) + λ + λ + + C C liquid liquid(T 2 − T )]
(2.18)
where: Qs = Energy storage m= Mass of material C solid solid = Heat capacity of solid phase C liquid liquid= Heat capacity of liquid phase T 1 = Initial Initial mater material ial temper temperature ature T = Phase change temperature temperature ∗
T =Final material temperature 2
2. 2.4. 4.3 3
He Heat at Tran ransf sfer er
Most heat transfer transfer in a conc concent entratin ratingg colle collector ctor system occurs at the receiv receiver. er. Ener Energy gy from insolation is reflected onto the absorber and leaves the system via the working fluid and thermal ther mal losses. Heat is lost through all three modes of heat trans transfer; fer; radiati radiation, on, conducti conduction on and convection. The thermal efficiency of a concentrator system (η ( ηtherm) is given by Equation 2.19.. 2.19 ηtherm = Q = Qout /Qin = 1
− QQL
in
in this equation:
= Adish I Qin = A
·
QL = Q = Qrad + Q + Qconv + Q + Qcond and: Qin = Energy incident on dish I= Direct normal insolation Qout = Energy absorbed by working fluid 31
(2.19)
Figure 2.18: Radiant Flux of a Blackbody Blackbody [2]
2. 2.4. 4.4 4
Radi Radiat atio ion n
All bodies emit radiation. The amount of energy a body emits depends upon the surface temperature and emissivity, emissivity, . Emissivit Emissivity y is a material proper property ty that descr describes ibes the rate of emission for a surface relative to a blackbody. A blackbody is a material that emits energy at a rate prescribed by Planck’s law of blackbody radiation, as shown in Figure 2.18. 2.18. 2.20 gives gives the net energy loss due to radiation for a cavity receiver[ 7]. Inside a Equation 2.20 Equation cavity cav ity some radiation will reflect back to the cavit cavity y wal walls. ls. To accou account nt for this effect, a term called effective emissivity ( (ef f ) is used. Equation 2.21 Equation 2.21 gives gives effective emissivity for a cavity receiver [7 [7].
· · · − T 4 )
Qrad = = ef f σ Aa (T s4 ef f =
cav + (1
−
cav cav ) (Aa /Acav )
In this equation: ef f = Effective emissivity of cavity cav = Emissivity of cavity surface σ = Stefan-Boltzmann constant 32
·
(2.20)
∞
(2.21)
Aa = Surface area of aperture Acav = Surface area inside cavity T s = Surface temperature of receiver T = Temperature of surroundings ∞
2. 2.4. 4.5 5
Co Con nvec ectio tion n
Convection transfers energy from the absorber surface directly to the air in contact with it it.. Wh When en the air air is stat statio iona nary ry it is refe referr rred ed to as na natu tura rall co conv nvec ecti tion on an and d wh when en it is in motion it is referred to as forced convection. Equation 2.22 Equation 2.22 gives gives the heat loss due to natural 20]. ]. convection for a cavity [ cavity [20 Qcond = ¯h Acav (T s
·
·
− T amb amb )
N¯uf kamb ¯h = L
·
N u = 0.088 Gr
·
1 3
(2.22)
.18 · cos θ2.47 · (L/Dcav )s · (T s/T amb amb )
in this equation: s = 1.12 0.98 (L/Dcav ) ¯ff ) L3 g β (T s T Gr = Gr = υ2
−
· −
· ·
·
and: h ¯ = Average heat transfer coefficient Acav = Cavity area T s = Temperature of receiver surface T amb amb = Ambient fluid temperature ¯f f = Average fluid temperature T Nu= Nusselt number Gr= Grashoff number kf = Thermal conductivity of ambient fluid L= Chara Character cteristic istic length of apertur aperturee opening (diamete (diameter) r) Dcav = Diameter of cavity 33
(2.23) (2.24)
υ = Kinematic viscosity θ = Angle receiver makes with zenith (at (at θ = 0 the receiver is horizontal) ◦
g= Gravity β = Volumetric thermal expansion coefficient At a tilt angle of θ = 90 (v (vertic ertical), al), the air inside a cavit cavity yb become ecomess trappe trapped. d. The hot, ◦
buoyant air inside the cavity cannot escape and convection losses are nearly eliminated. As the tilt angle decreases (θ (θ < 90 ) a dramatic increase in heat loss occurs as the lighter air ◦
begins escaping from the cavity. In this case buoyancy works to aid in heat loss. To reduce convective losses a translucent cover can be placed over the aperture. A cover allows the insolation to transmit through the material while trapping gas inside. Convection would still occur at the surface of the cover, but it would be reduced due to the reduction in exposed surface area. Convective loss can be calculated in this case by modeling the cover as an inclined flat plate. Using equation 2.25 equation 2.25 gives gives the Nusselt number for an inclined flat plate, which can be used with equation 2.22 2.22 to to calcu calculate late conv convecti ective ve losses. Figur Figuree 2.19 compares 2.19 compares the energy loss for an open and cov covered ered cavi cavity ty.. This example assume assumess an ap apertu erture re diameter of 30 cm, and cavity area ratio of 5. N¯u = 0.56( 56(Gr Gr P r cos θ)
·
·
1 4
(2.25)
where: θ < 88 ; 105 < Gr P r cos θ < 10 11 ◦
·
·
for β which which is evaluated All properties are evaluated at a reference temperature temperature T e , except for at at T ββ : T e = T = T w T β β = T
∞
2.5 2. 5
− 0.25( 25(T T s − T ) + 0. 0.50( 50(T T s − T ) ∞
∞
Stea Steam m Boil Boiler er Desi Design gn
A boi boile lerr is a cl clos osed ed ves esse sell wh wher eree water ater or an anot othe herr wor orki king ng flu fluid id is he heat ated ed..
Mo Most st
conventional boilers burn some fuel and pass the superheated exhaust gases into a boiler to boil a working fluid. Designs of boilers are classified by their heat transfer process. 34
Figure 2.19: Convection Loss vs. Tilt Angle for Open and Covered Cavities
Fire tube boilers look similar to conventional hot water heaters. Water is kept in a sealed container with empty space above it to hold the steam. Hot exhaust gases are passed through tubes inside the container, heating the water on the outside surface of the tubes. This causes water to boil and steam to accumulate at the top of the vessel. Water tube boilers have exactly the opposite design. Hot exhaust gases are sent into an open vessel vessel and the wat water er is pas passed sed thro through ugh tubes in the ves vessel sel.. Wate aterr is hea heated ted on the inside surface of the tubes, where it is converted to steam. This water/steam mixture then flows together through the tubes. A conventional design of a water tube boiler can be seen partss of the wat water er tube boiler are are:: the ste steam am drum, mu mud d dru drum, m, in Figure 2.20 Figure 2.20.. The major part downcomer tube and riser tube. Water flowing through the risers is heated by the flue gases. Hot water and vapor rises into the steam drum where they are separ separated ated.. The steam exits the top of the drum and the liquid water falls due to buoyancy down the downcomer tube. Incoming feedwater can be pumped into the steam drum or mud drum.
35
Figure 2.20: Water Tube Boiler [ Boiler [11 11]]
2. 2.5. 5.1 1
Boil Boilin ing g
Phase change of a working fluid only occurs at a contact interface. Boiling can occur at the solid-li solid-liqui quid d con contac tact, t, as in wa water ter in a hot pan, as we well ll as a liq liquid uid-v -vapor apor inte interfa rface. ce. For example, if hot gas was passed over a pool of water, boiling would occur at the surface of the water. We will only be concerned about the first case for this application. At the solid-liquid interface two types of boiling have been identified: nucleate and film. Nucleate boiling refers to the formation and release of steam bubbles on the solid surface with liquid water still wetti we tting ng the con contac tactt sur surfac face. e. Fil Film m boil boiling ing occur occurss whe when n the wat water er flow rate in the tube in contact con tact with the wall is not high enough to remov removee steam bubbles being produced. Steam 21]. ]. This film acts accumulates until the tube wall is covered with a continuous film of vapor [ vapor [21 as an insula insulator tor separat separating ing the heating surface from the work working ing fluid. Bey Beyond ond this p poin oint, t, increasing the temperature of the tube surface results in a net decrease in heat transfer. This point is marked B in Figure 2.21 Figure 2.21.. This figure shows the boiling curve for heat transfer coefficie coeffi cient nt (h) and net hea heatt tra transf nsfer er (q). It can be see seen n tha thatt the heat trans transfer fer coeffic coefficien ientt decreases after point point B . This drop precedes the reduction in net heat transfer because the
film layer layer is still still ve very ry thin. As the film lay layer er incre increase ases, s, the effe effect ct on net heat tran transfe sferr is observed obser ved.. An important aspect of boiler design is making sure that enough fluid is flow flowing ing through the boiler tubes to prevent film boiling. A well designed boiler has a steady stream of water flowing up the riser tubes constantly 36
Figure 2.21: Typical Boiling Curve [10 [ 10]]
feedin fee dingg the stea steam m dru drum. m. The There re are tw twoo wa ways ys flow can occu occurr ins inside ide the boiler boiler:: for forced ced and natural natur al circ circulati ulation. on. Force orced d circ circulati ulation on emplo employs ys pumps inside the boiler to force flow in the desired desir ed direction direction.. Natur Natural al circu circulatio lation, n, whic which h is common commonly ly used in indust industrial rial b boiler oilers, s, utili utilizes zes the natural density difference of the cold water in the downcomers and the less dense hot water wat er vapor mixture in the riser risers. s. The pressur pressuree differenc differencee in the two secti sections ons creates the force needed to keep flow moving through the system. Steam Ste am sep separa aratio tion n is also an impo importa rtant nt aspec aspectt to design designing ing a ste steam am dru drum. m. If the ste steam am does not separate from the water the steam outlet tube will contain unwanted liquid that can damage turbine blades or adversely affect other applications that require pure steam. An open boiler has an uno unobst bstruc ructed ted int interf erface ace betw betwee een n the liqu liquid id wa water ter and ste steam. am. If the steam velocity leaving the water’s surface is low (less than .9 m/sec) the steam bubbles will separate separ ate from the liquid dropl droplets ets [9]. If the rate is too high, water dropl droplets ets will be carried intoo the steam outlet line. For a unit with a higher ve int velocit locity y of steam product production, ion, gravit gravity y alone alo ne is not enou enough gh to sep separa arate te the steam from the liqui liquid. d. In this case steam separat separators ors must be utilized to facilitate gravitational separation. Simple types of steam separators form a barrier between the steam escaping the liquid and the exit tube. This forces the steam to take a longer path giving more time for gravity to separate the steam from any liquid that 37
has been carried with the steam.
38
CHAPTER 3 EXPERIMENTAL SETUP
A 14 m2 par paraboli abolicc dis dish h sol solar ar con concen centra trator tor wa wass fab fabric ricate ated d and tes tested ted at the SES SESEC EC facili fac ilitie tiess at Flo Florid ridaa Sta State te Uni Unive versi rsity ty.. The conc concen entra trator tor is ref referr erred ed to as Sol Solar ar 2 and it consists of a new concentrator and receiver assembly. The original steel frame and tracking system were kept from the first concentrator built at SESEC, Solar 1. This chapter includes details on design and fabrication for the concentrator system.
3.1 3. 1
Parabo araboli lic c Dish Dish
The parabolic dish was constructed out of fiberglass, which was chosen because of its abilit abi lity y to for form m to any any mol mold. d. The ent entire ire conc concen entra trator tor syst system em was limite limited d in siz sizee by the parabol par abolic ic dish. The larg largest est mold that was pract practica icall to make in hou house se was limited limited to a 1.2 m (4 ft) x 2.4 m (8 ft) plyw plywood ood she sheet. et. It was import importan antt to fabri fabricat catee the sys system tem by hand to ensur ensuree cost efficien efficiency cy and simpl simplicit icity y of man manufact ufacturing uring and assem assembly bly.. To maximi maximize ze the concentrator size, small sections were made and fitted together to form a continuous dish. The dish was comprised of eleven identical pieces and each piece was a 32 degree wedge of the full dish. This left 8 degrees open for the receiver receiver arm, which will be discussed below. below. The effective radius at the outer rim of the dish was 2.13 m and the greatest width of each section was 1.1 m. Each section was handmade at SESEC using vacuum molding techniques. To create the mold that mimick mimicked ed the parabolic shape, a plug was first needed needed.. A plug is a piece that replicates the outer contour of the final product. The plug consisted of 0.16 cm thick plywood sheet pressed onto verticle plywood ribs. The ribs were shaped to the profile of the dish given in Equation 3.1 3.1.. 1
2
y = 8.53 53m m
× x
39
(3.1)
Figure 3.1: Bare Concentrator Wedge
A thin plywood sheet was placed onto the ribs, taking the exact shape of the panels desired. A plywood box framed the plug, sealing it tight. This held the liquid plaster that was poured into the cavity. The plaster took the negative shape of the plug and created a mold. To construct the fiberglass concentrator sections, a foam sheet was laid between sheets of fiberglass cloth. The foam core stren strengthen gthenss the panels without greatl greatly y incre increasing asing weigh weightt because the foam does not absorb any of the hardening resin. The fiberglass and foam were then the n laid on the mold mold,, tak taking ing the panel panelss fina finall sha shape. pe. A pla plasti sticc bag was wrap wrapped ped aroun around d both the fiberglass fiberglass and mold and a vacu acuum um was pulle pulled d on the bag. Suc Suctio tion n fro from m the bag held the fiberglass tightly tightly to the con contour tour of the mold. Resin wa wass then drawn into the bag, soaking into the sheets of fiberglass. The result after hardening was a strong, lightweight and durable panel. Each of the eleven sections weighed approximately 17 lbs. Figure Figure 3.1 3.1 shows shows a completed completed wedge after it had b been een sanded. All of the wed wedge ge pieces wer weree bolted together together,, then bolted to the steel frame discussed below. A focal length of 2.1 m was chosen to give the dish a rim angle of 53 . A ggre reat ateer ◦
rim angle (shorter focal length) would decrease the acceptance angle of radiation into the absorber cavity. A lesser rim angle (longer focal length) would increase strain on the receiver arm and cause the optics to be more sensit sensitive ive to error in the dish. The finished conce concentr ntrator ator is shown in Figure 3.2 Figure 3.2
40
Figure 3.2: Fully Assembled Concentrator
3.2 3. 2
Refle Reflect ctiv ive e Su Surf rfac ace e
Thee re Th refle flect ctiv ivee ma mate teri rial al used used fo forr this this projec projectt is a silv silver er pol polym ymer er film film pr produ oduce ced d by ReflecT Refl ecTec ech. h. Thi Thiss film refle reflects cts 94% of the solar spect spectrum rum and las lasts ts 10 ye years ars in an out outdoor door environm env ironment ent.. This particu particular lar material was chose chosen n for its extr extremel emely y high reflect reflectivit ivity y and flexibi flex ibilit lity y of app applic licati ation. on. The film has an adh adhesi esive ve backing backing that can be app applie lied d eas easily ily to any any flat surf surface ace.. One dra drawba wback ck of the produ product ct is tha thatt the film is ve very ry thin and ext extrem remely ely susceptibl susce ptiblee to ‘prin ‘print-thr t-through’. ough’. This occurs when the film contou contours rs to imperf imperfecti ections ons on the applied surface and results in decreased optical clarity. This is a major problem when using fibergla fiber glass ss as the conce concent ntrat rator or str struct ucture ure.. Fibe Fibergl rglass ass res resin in har harden denss to the cloth cloth,, lea leavin vingg a crisscross pattern of hardened fiberglass strands, which is a less than ideal surface for the ReflecTech. Therefore, great care was given to surface preparation of the fiberglass. Initially, all of the pane panels ls were cov covere ered d and sande sanded d wit with h Bon Bondo do surfa surface ce prep. Thi Thiss smoo smoothe thed d out the print pattern pattern of the fibergla fiberglass ss panel panels. s. How Howeve ever, r, the durabili durability ty of the surface prep was questioned during the painting process, and the panels were ultimately sanded down with an industrial sander. A base paint was then sprayed directly to the sanded fiberglass. This 41
Figure 3.3: Mounting Frame
still left a small amoun amountt of ‘prin ‘print-th t-through’ rough’ on the reflec reflectiv tivee surface. The optical error due to this effect is believed to be the same order of magnitude as the error in the curvature of the dish surface. surface. Becau Because se the print through erro errorr is rando random m it did not greatly amplify the surface error. This situation is acceptable given the advantages that the material brings. A better surface preparation solution needs to be found if the curvature of the surface slope was corrected.
3.3
Frame
The horizontal mount system has two axes of movement that independently track the altitude and azimuth motion of the Sun. Figure 3.3 Figure 3.3 shows the mount system used for Solar 2. The azimu azimuth th pivot marked “D” follows the east-we east-west st movem movement ent of the sun throughou throughoutt the day. day. This piv pivot ot is made from 0.6 cm thick steel L-bra L-brack ckets ets that are 5.1 cm wide wide.. The azimuth pivot is controlled by a linear actuator which is mounted to the altitude pivot under the point marked marked “C”. The altitu altitude de pivot follo follows ws the north-sout north-south h movement movement of the Sun. This pivot is made from 0.6 cm square tubing that is 10 cm wide, with 1.3 cm steel plates extending up from the ends of the tubing. The altitude pivot is mounted to the zenith pivot marked “B” in Fig 3.3 3.3.. Thi Thiss piv pivot ot is normally stationary and bolted tight. This section has the ability to rotate around the zenith axiss to align with cele axi celesti stial al North. North. The zen zenith ith piv pivot ot is mad madee fro from m 0.6 cm ste steel el plate platess and 42
Figuree 3.4: Rece Figur Receive iverr Assem Assembly bly and Connectio Connection n Ring
welded to a sleeve cap which fits over a 12.7 cm nominal steel pipe. This steel pipe is bolted by three 1.6 cm bolts to the 15.2 cm (6 in) XS nominal steel foundation pipe marked “A”. The foundation foundation pipe rises 1.2 m above a 10.2 cm deep conc concrete rete slab slab.. It exten extends ds 4.3 m int intoo a 0.5 m diameter conc concrete rete backfille backfilled d hole. The founda foundation tion is designed to handle the forces and moments moments of a 4.3 m diameter dish in winds up to 49 mps (110 mph). Full engine engineering ering drawings for the foundation can be found in Appendix E. Atop At op the moun mountt sys system tem sits the con connec nectio tion n rin ring. g. Thi Thiss 0.9 0.955 cm thic thick k ste steel el ring is 10. 10.22 3.4 shows shows the cm wide and 1.47 m in diame diameter ter.. It is we welde lded d to the azi azimu muth th fram frame. e. Fig Figure ure 3.4 connection conne ction ring, marke marked d “F”. This ring has 0.95 cm slots mach machined ined out in which L-brac L-bracke kets ts are bolted bolted to. Tw Twelv elvee 0.95 cm L-brac L-bracke kets, ts, 10.2 cm in lengt length, h, bolt to the ste steel el ring to the flange of each panel. The receiv receiver er is hel held d in place by a sin single gle suppo support rt arm. The arm is con connec nected ted to a 7.6 cm square square ste steel el tube we welde lded d to the azi azimu muth th frame frame.. Thi Thiss tub tubing ing is mar marke ked d ”E” in Fig Figure ure 3.4. Ext 3.4. Extend ending ing ver vertic ticall ally y at the end of thi thiss pie piece ce are tw twoo ste steel el plate plates. s. Thr Three ee 1.6 cm holes are drilled through the plates. Between the plates, 7.6 cm aluminum tubing is inserted and
43
bolted bolt ed in. Tw Twoo of the bolt boltss can be rem remov oved ed to allow allow the rece receiv iver er arm to pivo pivott aro around und the remaining bolt, lowering the receiver to ground level. This allows easy access for repairs and changes cha nges to configurati configuration on during testing. The aluminum aluminum arm is marked ”G” in Figur Figuree 3.4 3.4.. It is made from two 7.6 cm pieces of aluminum square tubing. The vertical arm section is 1.92 m long and the angl angled ed sect section ion is 1.33 m lon long. g. An alumi aluminu num m plate is welded welded at the end of the arm. This plate has an arra array y of bolt holes aligned verti vertically cally that allo allow w the rece receive iverr to be attached at differing heights. For different arrangements of aperture covers, the receiver receiver can be adjusted closer and further from the concentrator focal point.
3.4
Recei eceiv ver
The receiver for this system acts as an absorber, boiler, and heat storage unit. A cavity type absorber absorber is used due to its high absorptio absorption n efficiency and low heat loss. Surro Surrounding unding the absorber inside the receiver is 10 kg of sodium nitrate. This salt acts as a heat transfer and storage media and 0.6 cm diamet diameter er copper tubing was coile coiled d through it. The wor working king fluid is pumped through the tubing where heat is transfered to the fluid. To find the optimum cavity size, heat losses and incident radiation must be estimated based on aperture (cavity opening) diameter. Figure 3.5 Figure 3.5 shows shows the estimated energy absorption versus aperture diameter. diameter. Heat loss is depende dependent nt upon cav cavity ity geometr geometry y, orien orientatio tation, n, and surface surface emi emissi ssivit vity y. Cal Calcul culati ations ons were made ass assumi uming ng neg neglig ligibl iblee win wind. d. In rea realit lity y the convec con vectiv tivee term will b bee great greater, er, making the losses almost equal. The inciden incidentt radia radiation tion was estima est imated ted for a con concen centra trator tor with an err error or 4 tim times es tha thatt of the indus industri trial al sta standa ndard. rd. Thi Thiss is one of the major dra drawba wback ckss of a lo low w bud budget get.. For perf perfec ectt con concen centra trator tor effici efficienc ency y, the 3.5 takes takes the shape of a step function, instantly reaching incident radiation term in Figure Figure 3.5 maximum radiation of about 6150 W. The large slope in the curve shows the effect of decrea dec reased sed efficie efficiency ncy.. At the opt optim imum um dia diamet meter er of 15 cm, onl only y 80% of refl reflect ected ed rad radiat iation ion passes through the aperture and the estimated Sun to absorbed energy efficiency is 42%. The cavity depth was chosen such that 85% of reflected radiation strikes the cylindrical cavit ca vity y abs absorbe orberr wa wall, ll, and 15% of radiat radiation ion was absor absorbed bed by the top of the cav cavit ity y. The resulting resul ting cavit cavity y depth is 15 cm. For deeper cavitie cavitiess less radiati radiation on will strik strikee the top of the cavity, resulting in an uneven temperature distribution inside the cavity. For shorter cavities the concentration ratio will be higher, increasing cavity temperature as well as thermal losses. Figure 3.6 Figure 3.6 shows an exploded view of the receiver assembly. 44
Figure 3.5: Energy Absorption vs. Aperture Diameter
Figure 3.6: Exploded View of Receiver
45
Figure 3.7: Assembled Receiver
The cylindrical walls of the receiver are made of 0.16 cm thick stainless steel tubing. The exterior tubing is 29.2 cm long. The tubing is topped by a flange that is 0.64 cm thick with an outer diameter diameter of 31. 31.88 cm. Due to err errors ors in the conc concen entra trator tor,, not all of the refl reflect ected ed radiation radiat ion passes through the aperture opening. SA portion of radiat radiation ion was absorbed by the bottom ring of the rece receive iver, r, and a p portion ortion comp completel letely y missed the absorber. The bottom ring of the rec receiv eiver er has an out outsid sidee dia diamet meter er of 25. 25.44 cm and is not insul insulate ated. d. Bot Both h the exte exterio riorr surface surfa ce of the cavi cavity ty and the bottom ring are chemic chemically ally bonded with black chro chrome. me. This material has an absorptivity of 90% for solar radiation and an emissivity of 15% for relatively low temperatures compared to that of the Sun. Once the solar energy has been absorbed it is conducted through the metal walls and conducted through the sodium nitrate that fills the receive rece iver. r. Immer Immersed sed in the salt is 15 m of coiled 0.6 cm diameter diameter copper tubing. Water enters enters the receivers top flange and flows through the loosely coiled tubing around the exterior of the receiver. Once the coil reaches the bottom, it is tightly coiled around the center cylinder and this inn inner er coil ex exits its the rec receiv eiver er through through the top flan flange. ge. The exte exterio riorr of the boile boilerr is wrapped wra pped in 3 cm of The Therma rmall Cer Cerami amics cs Kao Kaowoo wooll Bla Blank nket et ins insula ulatio tion. n. The fully assem assemble bled d .7 receiver is shown in Figure 3 Figure 3.7
3.5 3. 5
Trac racking king
Movement of the parabolic dish was accomplished by two satellite dish linear actuators. A SuperJack Pro Band HARL3018 was used for altitude tracking, and a VBRL3024 was used for azimuth tracking. The HARL3018 has a stroke length of 45 cm and dynamic load of 252 46
Figuree 3.8: Assem Figur Assembled bled Linear Actuators
kg. The VBRL3024 has a stroke length of 61 cm and a dynamic load of 680 kg. A stronger actuator was needed on the azimuth axis due to the extreme angles needed in tracking, specifically when approaching horizontal. The large moment caused by the receiver at these angles requires requires an extr extremel emely y pow powerfu erfull actua actuator. tor. Po Power wer for the actuator actuatorss was provi provided ded by the local elec electric trical al grid. How Howeve ever, r, this syste system m can be pow powered ered remot remotely ely by a small PV panel and two 24V batteries, as was done on Solar 1. Figure 3.8 Figure 3.8 shows the assembled actuators. The tracking system was controlled by a dual axis tracking sensor shown in Figure 3.9 3.9.. The LED3X sensor was was built by Red Rock Ener Energy gy.. It has four LEDs attac attached hed to a circu circuit it board. boar d. The rela relativ tivee inte intensi nsity ty of rad radiat iation ion for eac each h LED is mea measur sured ed and whe when n the there re is a disparity between two LEDs the circuit allows movement of the actuators to re-align the system.
3.6
Data Data Acqu Acquisi isitio tion n and and Instr Instrume umen ntation tation
System Syste m test testing ing consisted of running water through the heated boiler. The temperat temperature ure rise and mass flow rate of the water was measured to calculate thermal power output. Three thermocouples were used for data acquisition. The thermocouples were placed at the water inlet,, inside the therm inlet thermal al bath, and at the exit of the recei receiver ver.. A schemat schematic ic of the recei receive verr 47
Figure 3.9: Dual Axis Tracking Sensor
during testing is shown in Figure 3.10 Figure 3.10.. The thermocouples used were Omega K-type (CASS18U-6-NHX) and were connected to data acquisition hardware using K-type thermocouple wires. wire s. A National Instrum Instrument entss Signal Condi Conditionin tioningg Board (SCB(SCB-68) 68) routed the signal to a National Nation al Instrumen Instruments ts data acqui acquisitio sition n card, model DAQC DAQCard ard -6024e. The card had a 12 bit resolution with sampling rate of 1 kS/s. The FSO during testing was 10 V. The DAQ card had a corresponding absolute accuracy during testing of 2.44 mV. Both the water and pumping power for testing was provided from the local water tap. The average pressure at water inlet was 60 psi, and the flow rate was controlled and measured using usi ng a bal balll floa floatt flow flow met meter er with con contro troll valv alve. e. The va valv lvee had a flow flow ran range ge of 2-2 2-255 GPH with an accuracy of 4%. Measurements were taken by hand and logged into the LabVIEW
±
software.
A pyrhe pyrhelio liomet meter er was use used d to mea measur suree dir direct ect norma normall ins insolat olation ion durin duringg tes testin ting. g. The instrument used was an Eppley Model NIP. This pyrheliometer was mounted on a powerdriven equatorial mount for continuous readings. The unit has a sensitivity of approximately 8 µV/Wm µ V/Wm 2, and holds linearity of 0.5% from 0 to 1400 Wm 2. The unit can be seen in −
Figure 3.11 Figure 3.11..
−
±±
48
Figure 3.10: Schematic of Test Setup
Figure 3.11: Eppley Nominal Incidence Pyrheliometer
49
CHAPTER 4 RESULTS AND DISCUSSION
This chapter contains results and analysis for testing performed at SESEC facilities in the Spring of 2009. Tests were perform performed ed to dete determine rmine the thermal conv conversio ersion n efficie efficiency ncy of the conce concentr ntrator ator system. Water was pumped throu through gh the centr central al rece receive iverr where it was heated and the flow rate and temperature rise of the water were recorded to determine the collected energy. To calculate solar energy input, a pyrheliometer was used to measure direct beam insolation. With this information the thermal collection efficiency for the system was found.. The system was cha found charact racteriz erized ed by estimatin estimatingg therm thermal al losses for eac each h sub-c sub-componen omponent. t. The chapter concludes with an economic analysis.
4.1 4. 1 4.1.1 4.1 .1
Ther Therma mall Effici Efficien ency cy Re Resu sult ltss
Ste Steady ady Stat State e T Therm hermal al Test esting ing
Tests were performed to dete determine rmine total therm thermal al con conver version sion efficiency efficiency.. In these tests, water inlet water at 20 C was heated by the receiver to temperatures under 100 C. By not ◦
◦
flashing the working fluid, a steady inlet flow was achieved that allowed accurate collected energy ene rgy mea measur sureme ement nts. s. Re Resul sults ts fro from m ste steady ady sta state te the therma rmall con conve versi rsion on tes tests ts are sho shown wn in 4.1 through through 4 4.3 .3.. The figures show outlet water temperature, salt bath temperature Figures 4.1 Figures and insolation insolation during each test. The ave average rage conv conversio ersion n efficiency at a cav cavity ity angle of 53 degrees was 39%.
4. 4.1. 1.2 2
Su Superh perhea eated ted Steam Steam P Prod roduc ucti tion on
A series of tests were conducted to produce superheated steam. In these tests the receiver was heated without water flowing through it, until it reached a temperature of 400 C. Water ◦
50
Figure 4.1: Testing Performed Feb 23, 2009
Figure 4.2: Testing Performed March 02, 2009
51
Figure 4.3: Testing Performed March 04, 2009
was then allowed allowed into the b boiler oiler wher wheree it flashe flashed d to steam steam.. The steam was then superhea superheated ted in the boiler b oiler coil coilss b befor eforee exiting the recei receiver ver.. The tempera temperature ture of the exiting steam can be seen in Figure 4.4 Figure 4.4.. Intermitt Intermittent ent cloud cove coverr prese present nt during test testing ing caused the thermal bath temperature to fluctuate, but it was stable enough to continue producing superheated steam. Superheated steam was produced as long as the salt bath was a liquid. Figure 4.5 Figure 4.5 shows the salt and ste steam am tem tempera peratur tures es durin duringg tra transi nsitio tion n fro from m liq liquid uid to sol solid. id. In thi thiss tes testt the receiver was heated to 375 C befo before re being poin pointed ted aw away ay from the Sun Sun.. The tempe temperat rature ure ◦
dropped as the water extr extracte acted d heat from the rece receive iver. r. The salt bath ev event entually ually liquefied liquefied when it cooled to 290 C. Heat transfer plummeted during this process and the boiler output ◦
was saturated saturated steam instead of superheate superheated d steam steam.. This is because the solidifie solidified d salt cannot transfer enough heat to superheat the steam. For steam turbine operation it will be necessary to have a liquid salt bath. As water entered the boiler and flashed to steam, it expanded, increasing the pressure inside ins ide the boiler tubes tubes.. This This pus pushed hed bac back k on the inl inlet et line and sto stopped pped the flow of wa wate ter. r. When steam then exited the boiler, the pressured dropped below the inlet pressure and water
52
Figure 4.4: Superheated Steam Production March 09, 2009
then flowed back into the boiler. This resulted in an oscillating flow rate. The manual flow meter used was incapable of measuring unsteady flow of this type and as result no flow rate data is prese presente nted d here. An automated data logging flow meter that can handle fluctuatin fluctuatingg flow flow rat rates es must be use used d to get an acc accura urate te readi reading. ng. Thi Thiss wil willl be req requir uired ed to dis disco cove verr the conversion efficiency of the system while producing superheated steam.
4. 4.1. 1.3 3
Th Therm ermal al St Stor orag age e
Thermal Ther mal storage in the receiv receiver er is int intende ended d to dampen the effects of cha changing nging insolatio insolation. n. The objective is to keep the output temperature relatively steady during intermittent cloud cover, as was demonstrated with the production of superheated steam as shown in Figure 4.4.. The out output put tem tempera peratur turee droppe dropped d only 54 C when the Sun was blocked by clouds 4.4 ◦
for nearly nearly 15 min minute utess fro from m 13: 13:34 34 to 13:50. The leng length th of tim timee the rece receiv iver er stay stayss hea heated ted will depend on the duration of the blocked sunlight and the amount of energy removed by the workin workingg flui fluid. d. The outpu outputt ste steam am rem remain ained ed superh superheat eated ed for the dura duratio tion n of the clo cloud ud cover. cov er. If the outpu outputt steam had been conne connected cted to a turbi turbine, ne, the therm thermal al storage would have 53
Figure 4.5: Testing Performed March 02, 2009 Display of Salt Bath Phase Change
4.6,, which shows testing ensure ens ured d con contin tinuou uouss oper operati ation. on. A sim simila ilarr effect effect is see seen n in Fig Figure ure 4.6 with intermitt intermitten entt cloud cove coverr when the salt bath was a solid. The tempera temperature ture drop was slower than in the superheated case due to the lower water flow rate. Heatt is also sto Hea stored red in the phas phasee chang changee of the bat bath h sal salt. t. Whe When n the boile boilerr rea reach ches es the phase change temperature of the salt, the bath temperature is momentarily constant. This increases the time that the boiler can produce superheated steam when cooling down. Figure 4.5 shows 4.5 shows the output and salt bath tempera temperature ture durin duringg phase change change.. It can b bee seen that the amount of heat released during phase change is small. This is due to the fact that most of the thermal mass in the receiver is in the stainless steel, while the salt holds only a small fraction of the heat in the receiver. This has unintended consequences during warm up and cool down and is discussed below. A clear example that the salt bath holds only a portion of the thermal mass is shown in Figure 4.7 Figure 4.7.. In this figure two different salt bath heating rates can be seen; A and B. Rate A corresponds to heating of the entire receiver, salt and steel, which occurs at 65 C/hr. Rate ◦
B corresponds to when the heat has been drawn from the salt bath by the working fluid, but 54
Figure 4.6: Testing Performed Sept 04, 2008 0.7 LPM
the steel is still hot. In this case the salt heats up at 410 C/hr, over 6 times that of rate A. ◦
This shows that when the salt bath and steel housing are at the same temperature, most of the heat goes to heating the steel. This is detrimental because the steel does not give up its heat quickly to the working fluid and therefore, it is wasted energy.
4.1.4 4.1 .4
Warmarm-up up and Cool Cooling ing-do -down wn
The large thermal mass of the receiver generally hurts performance. The worst effect on the system can be seen during warm-up of the receiver. Figure 4.8 4.8 shows shows the initial warm up temperature temper ature profi profile. le. The heati heating ng rate for the salt bath was 65 C/hr, which was the same ◦
as rate ‘A’ marked in Figure 4.7 Figure 4.7.. For steady state testing the salt bath reached equilibrium at 250 C. It took 3.8 hours for the receiver to heat up to equilibrium temperature when ◦
water was flowing through the receiver from start-up. The quickest way to heat the receiver was to do so without water flowing through it. After the receiver had reached 250 C, it was ◦
removed from the focal point and water was allowed to pass through it. Initially the water flashed flashe d to steam and the flow rate was reduc reduced ed considera considerably bly.. Afte Afterr wai waiting ting for the salt bath 55
Figure 4.7: Testing Performed February 24, 2009 1.0 LPM
to cool so that no steam exited the receiver, it was again placed in the focal point where the salt was heated at rate B until it reached equilibrium. This complicated warm-up procedure could have been avoided if less steel was used in the construction of the receiver.
4. 4.1. 1.5 5
Trac racki king ng Err Error or
The trackin trackingg sen sensor sor used for test testing ing was not alwa always ys acc accura urate. te. The LED tra track cker er wa wass mounted to the receiver arm. During testing the large amounts of torque placed on the arm by the receiver caused the arm to bend. In particular, when the azimuth angle was greater than 45 the receiver receiver arm bending caused the track tracker er mounte mounted d to it to err. Figur Figuree 4.9 4.9 shows shows ◦
the tracking tracking error during late hours. The oscilla oscillating ting recei receiver ver temper temperature aturess are seen with relati rel ation on to the stead steady y ins insola olatio tion. n. The trac trackin kingg sys system tem has a diffi difficul cultt tim timee cor correc rectin tingg for changes in both altitude and azimuth angle. Trac racking king error also occurs with int intermit ermitten tentt cloud cove cover. r. The trac tracking king mechani mechanism sm was design des igned ed to p poin ointt at the bri brigh ghtes testt spot in the sky sky.. Whe When n clo clouds uds cov cover er the Sun Sun,, thi thinne nnerr parts of the cloud shine the brightest. brightest. The trac tracking king syste system m then points to this spot. Afte Afterr the cloud cloud has pas passed sed,, the syst system em is not aime aimed d at the Sun and must cor corre rect ct itse itself. lf. Thi Thiss 56
Figure 4.8: Testing Performed Feb 22, 2009 1.5 LPM
Figure 4.9: Testing Performed Feb 19, 2009 1.3 LPM
57
Figure 4.10: Testing Performed February 26, 2009 1.1 LPM
After short cloud cov cover, er, the temperat temperature ure of the salt phenomenon can be seen in Figure 4.10 Figure 4.10.. After bath and water outlet continued to decrease until the system corrected itself.
4.2
System System Chara Characte cteri rizat zation ion
Ener En ergy gy los losse sess fo forr ea eacch subs subsys yste tem m ar aree es esti tima mate ted d bel below ow..
Ca Calc lcul ulat atio ions ns are are ma made de to
determine dete rmine conce concentr ntration ation,, radiat radiation ion and con conve vectiv ctivee losse losses. s. Analys Analysis is was done for the case of testing performed on Febru ebruary ary 23, 2009, shown in Figur Figuree 4.1. 4.1 . All variables were averaged and are shown in Table 4.1 Table 4.1..
4. 4.2. 2.1 1
Mirr Mirror or Lo Loss ss
The Reflectech film has a solar reflectance of 94%. At an average insolation of 946 W/m 946 W/m 2 , incident radiation was 13.2 kW. The loss due to the reflector was 6%, which corresponds to 792 W. As such, 12.4 kW was reflected by the mirror.
58
Table 4.1: Average Values for Testing Performed Feb 23, 2009 Cavity Angle (deg) Concentrator Area m2 Absorber Surface Temp ( C) Ambient Temp ( C) Water Flow Rate (LPM) Inlet Temp ( C) Outlet Temp ( C) Salt Bath Temp ( C) Insolation (W/m2 ) Qin (kW) Qout (kW) Thermal Efficiency (%) ◦
◦
◦
◦
◦
4.2.2 4.2 .2
53 13.98 800 20 1.5 19.5 69.2 243 946 13.2 5.2 39
Con Concen centrat trator or Geo Geomet metry ry Loss Loss
Angular error was estimated for the fiberglass dish to determine the intercept factor. The error was estimated to the nearest integer multiple of the industrial standard error, which is 6.7 mrad. A series of boards with a center hole cut out were were constructe constructed. d. The cen center ter hole was cut with the diameter equal to the 99% cutoff diameter for a particular standard error. For the case of 5 standard errors (5*6.7 mrad angular error) the 99% cutoff diameter for a concentrator with a focal length of 2.13 meters is 38.5 cm. This means 99% of the reflected lightt should shine inside this circl ligh circle. e. The fiberglass concen concentrato tratorr built corres corresponded ponded closest to an angular error 6 times the industrial standard. The concentrator angular error was 40.2 4.11 with with respect to +/- 6.7 mrad. The inte intercep rceptt factor for this error is plotted in Figure Figure 4.11 aperture diameters. It can be seen that 60%, or 7.44 kW, of the radiation strikes inside the cavity aperture and 25%, or 3.1 kW, of radiation strikes the absorber plate. The remaining 15% of reflected radiation was lost completely, which amounted to 1.86 kW.
4. 4.2. 2.3 3
Ab Abso sorp rptio tion n Lo Loss ss
When impinging radiation reflects off the absorber it is considered an absorption loss. The absorptivity coefficient of the cavity and plate surface was 90%. As such, 10%, or 310 W, of
59
Figure 4.11: Intercept Factor for Solar 2
radiation radiat ion that hit the absorber plate was lost. The cavit cavity y surface had the same absorptiv absorptivity ity as the plate, but it recaptured some of the reflected radiation. When light was reflected off the surface surface it hit anoth another er surfa surface, ce, givin givingg it a mu much ch high higher er chan chance ce of bein beingg abs absorbe orbed. d. The effective absorptivity of the cavity was 98.7%. Only 164 W was lost by reflection out of the cavity. The total absorption losses were calculated to be 474 W.
4. 4.2. 2.4 4
Ra Radi diat atio ion n Lo Loss ss
Radiation Radia tion losses were estimate estimated d for the cav cavity ity and plate absorber separate separately ly.. Equat Equation ion 2.20 2.20 gives gives energy lost due to radiation. For a cavity, the radiated energy is calculated using 2.21.. The emissiv emissivity ity of the black chro chrome me coating the effective emissivity given in Equation Equation 2.21 on the absor absorber ber surfa surface ce is 0.15 0.15.. For the abso absorber rber plat plate, e, the amou amount nt of rad radiat iated ed energ energy y at 800 C is 0.6 kW. The cavity helped to reduce radiation losses by recapturing radiated energy. ◦
Even though the cavity surface area is nearly 4 times that of the absorber plate, the losses were only slightly higher at 0.85 kW. The total loss from the absorber due to radiation was calculated at 1.5 kW. 60
4. 4.2. 2.5 5
Co Con nvec ectio tion n Lo Loss ss
Natura Nat urall con conve vecti ction on wa wass the large largest st con contri tribut butor or to hea heatt los loss. s. At a ca cavit vity y ang angle le of 53
◦
buoyancy forces kept air continuously moving over the absorber, taking great amounts of energy loss due to natural convection from a cavity can be found. heat. Using Equation 2.22 Equation 2.22 energy The convective heat loss from the cavity was calculated at 1.3 kW. The absorber plate was modeled as an inclined flat plate. plate. Equat Equation ion 2.24 2.24 gives gives the average Nusselt number that must be used with Equation Equation 2.22 2.22 to find con conve vection ction from an incli inclined ned plate. Natur Natural al conve convection ction from the absorber plate was calculated at 1.1 kW. The total heat loss due to convection was 2.4 kW.
4. 4.2. 2.6 6
Re Rema maini ining ng Th Ther erma mall L Los osss
A sim simple ple Firs Firstt La Law w ene energy rgy balan balance ce rev reveal ealss tha thatt an add additi itiona onall 1.1 kW are lost lost.. Thi Thiss is due to a com combinat bination ion of man many y small losse losses. s. The largest being reflect reflector or wear, conduct conduction ion and imperfect imperfe ct insulati insulation. on. The mirror mirrored ed surface had a maxim maximum um reflecti reflective ve efficie efficiency ncy of 94%. In realit rea lity y the effic efficien iency cy of the mirr mirrore ored d sur surfac facee on Solar 2 wa wass not this high. Dus Dustt and dirt were blown onto the reflector and blocked sunlight. Discolorations formed on sections of the mirrorr after water pooled on them for months mirro months.. If the we wear ar was enough to redu reduce ce reflecti reflectivit vity y to 90%, an additional 528 W were lost. Conduc Con ductio tion n occu occurre rred d at the joini joining ng of the rece receiv iver er and the con connec nectio tion n arm arm.. Thi Thiss wa wass reduced as much as possible by placing insulation between the plates that are bolted together. The insulation will conduct a small amount, but the stainless steel bolts that directly connect the receiver to the arm are good conductors of heat. The final major ma jor contr contributor ibutor is the insula insulation tion surroun surrounding ding the receiv receiver. er. The insulati insulation on was not one con contin tinuous uous piece cov coverin eringg the recei receiver ver.. Hot air could wor work k its wa way y out of the insulation carrying heat with it. The insulation itself heated a small amount and convected heat to the ambient air.
4. 4.2. 2.7 7
Re Redu duci cing ng Lo Loss sses es
Figure 4.12 shows Figure 4.12 shows a breakdown of the relative losses for Solar 2. Convection, radiation and concentr conce ntrator ator geomet geometry ry error make up 71% of all of the losses. Eac Each h of these thre threee systems is primarily dependent upon the angular error of the concentrator, which determines the 61
Figure 4.12: Distribution of Thermal Losses
absorber aperture aperture diameter. If the apertur aperturee radius was reduc reduced, ed, these losses would decrease exponentially exponen tially.. By incre increasing asing the optic optical al effici efficiency ency of the concen concentrato trator, r, cav cavity ity losses will be greatly great ly reduced reduced.. Incre Increasing asing the efficienc efficiency y of the concen concentrato tratorr should b bee the prima primary ry focus to increase total system efficiency. Abso Ab sorp rpti tiv ve loss losses es,, wh whic ich h co cont ntri ribu bute ted d 6% of all all loss losses es ca can n also also be redu reduce ced. d. 64 64% % of absorp abs orptio tion n los losses ses we were re fro from m the flat pla plate te abs absorbe orber, r, ev even en tho though ugh onl only y 25% of refl reflect ected ed radiation radiat ion struck the flat plate. By optimizin optimizingg the cav cavity ity apertur aperturee to the concen concentrato tratorr the flat plate absorber will be unnecessary. This will automatically eliminate a large portion of the absor absorption ption losses witho without ut incre increasing asing system cost. The remaining 23% of losses came from sources that will be more complicated to reduce. Mirror losses can not be decreased easily as it will be very difficult to find a material with such a high reflectivity and weathering capabilities for a reasonable price. Lastly, insulation losses can be minimized by building an airtight housing around the insulation. The insulation used was open to the elements. Wind and rain wear down the insulation and destroyed parts of it. A housing can be built to keep air from ent entering ering the insula insulation tion and weath weather er out, but it will increase system cost and complexity.
62
Table 4.2: Optical Efficiency Cavitty Ab Cavi Abso sorbe rberr Plat Platee Total otal Mirror Efficiency (%) 94 94 94 Intercept Efficiency (%) 60 25 85 Absorption Efficiency (%) 98.7 90 95 Optical Efficiency (%) 55 21 76
4.3 4. 3
Opti Optica call Effici Efficien ency cy
Optical efficiency describes the system’s ability to absorb radiation that strikes normal to the concen concentrat trator or apertur aperture. e. Colle Collection ction efficien efficiency cy is affect affected ed by the mirro mirror, r, conce concentr ntrator, ator, absorbe abs orberr abs absorp orptiv tivit ity y and the inter interce cept pt fac factor tor.. The inte interce rcept pt fac factor tor is a fun funct ction ion of the 4.2 shows shows the angular error of the concen angular concentrato tratorr and apertur aperturee area of the absorber absorber.. Table able 4.2 optical optic al efficienc efficiency y for Solar 2. 76% of all of the radiati radiation on that struck the mirror mirrored ed surface was absorbed by the receiver, with 55% of incident radiation being absorbed by the cavity and 21% by the flat plate.
4.4 4. 4
Boil Boiler er Effici Efficien ency cy
Boiler efficiency describes the receiver’s ability to take impinging radiation and convert it into useful energy. energy. Table able 4.3 4.3 shows shows the efficiency of the cavity, absorber plate and total boiler. It can be seen that the efficiency of the cavity is much greater than that of the absorber plate. The cavity had a conversion efficiency of 70%, but the total boiler conversion efficiency was only 62%. If the angular error of the concent concentrator rator wa wass decrease decreased d so that all the radiation enters the cavity, the boiler efficiency would increase to 70%.
4.5 4. 5
Insu Insula lati tion on Effici Efficien ency cy
The final efficiency factor needed to completely describe the system is the insulation efficiency. This factor takes into account all of the small losses described in section 4.2.7. 4.2.7. The relative impact of the 1.1 kW lost due to insulation efficiency is found by solving Equation 4.1 4.1.. Solvi Solving ng for insulati insulation on efficien efficiency cy in this equati equation on reveals reveals it to b bee 83%. The syst system em can 63
Table 4.3: Absor Absorber ber Efficie Efficiency ncy Cavitty Ab Cavi Abso sorbe rberr Plat Platee Total otal Absorber Area (cm (cm ) 910 320 1230 Absorbed Radiation (W) 7,160 2,910 10,070 Convection Loss (W) 1,300 1,100 2,400 Radiation Loss (W) 850 600 1,450 Absorber Efficiency (%) 70 42 62 2
now completely be described by these three sub-system efficiencies. ηinsulation
× ηoptical × ηabsorber = ηthermal 4.6
(4.1)
Error
Error in calculation was caused by the flow meter and the thermocouples. The flow meter had a relat relative ive err error or of 4%. The ther thermocouple mocoupless had an absolute err error or of 1.5 degre degrees es for the temperature temper aturess used in testing, whic which h equates to a relativ relativee error of 3% each. Absolu Absolute te error in efficiency calculations is 4.2%.
4.7 4. 7
Cost ost Analys alysiis
4.4 shows shows the final cost breakdown The final material cost for Solar 2 was $3062 and Table 4.4 for each subsystem. Costs can be brought down in many areas. One way to reduce material cost would be to redesign the frame. The frame used was recycled from a commercial satellite dish. The cost of the steel wa wass estimate estimated d at pric prices es in August 2008. A frame design par particul ticular ar for this project could prove prove cheaper cheaper by usi using ng les lesss ste steel. el. The rece receiv iver er arm used was onl only y necessary for testing purposes. The final commercial design should have a configuration that holds the receiver in more than one location, greatly lowering the strength requirement of each eac h support and total cost. Mate Materials rials used for the receiv receiver er could also be ch changed. anged. Stainl Stainless ess steel was used for the boiler to guarantee the prototype did not corrode, which decreases absorp abs orptio tion n and heat tra transf nsfer er.. Che Cheaper aper stee steels ls cou could ld be use used d if they we were re found to have have no adverse adv erse effec effectt to syste system m efficiency efficiency.. Clea Clearly rly the cost of the system can be reduc reduced, ed, but it is 64
unsure whether it will be possible to reduce the system cost to the original goal of $1000. The dramatic rise in the cost of materials, even in the period of fabricating this system, will make it difficult to reach this goal.
65
Table 4.4: Material Costs Frame and Foundation Base Tubing 6” SCH 80 Steel Tubing 17’ Concrete 1 sq yd bulk Connection Tubing 5” SCH 840 Steel Tubing 28’ Base Top .5” Steel Plate 6” x 4.5’ Altitude Axis 1/2” x 4” Steel Plate 13’ Azimuth Axis 2 x 2 x 1/4” Square Tube 13.5’ Connection Miscellaneous Hardware SubSub-T Total otal 18” Actuator 24” Actuator Tracking Sensor
Tracking System Pro Band HARL3018 Pro Band HARL3024 Dual Axis LED3X SubSub-T Total otal
Concentrator Foam Core $1.67 per sq ft Fiberglass Cloth $0.28 per sq ft x 4 layers Resin $2.92 per lb 1:1 glass to resin weight ratio Surface Prep Feather Fill Reclectech bulk price Cost Co st Per Sq Ft Sub Total otal Receiver Support Structure Aluminium Tubing 33”” Square x 147” Steel Tubing 3” Square x 44” Steel Support Plates 2 x .5” 3” x 8” SubSub-T Total otal Receiver Outer Tubing 10” OD SS Cavity Tubing 6” OD SS Flanges 1/4” SS Plate 12” x 24” Copper Tubing 1/4” x 50’ Salt 10 kg Sodium Nitrate SubSub-T Total otal Total
66
Cost 350.00 100.00 44.11 84.62 146.64 88.95 100.00 $9 $914 14.3 .322 Cost 69.95 169.95 114.00 $3 $353 53.9 .900 Cost 267.20 177.79 169.36 249.80 216.00 $6. $6.75 $1080. $1080.15 15 Cost 149.94 38.72 24.35 $2 $213 13.0 .011 Cost 147.00 54.50 88.12 45.98 165.24 $5 $500 00.8 .844 $3062.22
CHAPTER 5 CONCLUSION
A 14 m2 parabolic dish concentrator, nicknamed Solar 2, was built at SESEC. The system used a cavity type receiver with 10 kg of sodium nitrate to act as heat storage. The goal of the system was to provide 6.67 kW of thermal energy, enough to provide 1 kW of electricity with a micr micro-ste o-steam am turbine. Man Many y impro improve vemen ments ts were made from the first concent concentrator rator system syste m built at SESEC, which prov provided ided only 1 kW of therm thermal al energy energy.. The concen concentrat trator, or, mirror and receiver were all redesigned to increase thermal conversion efficiency. The gross thermal efficiency of the system at a cavity angle of 53 degrees was 39%. This was a 333% increa increase se from the first syste system m assemble assembled d at SESEC SESEC,, Solar 1. Efficie Efficiency ncy was improved in multiple areas from Solar 1. First, the mirror efficiency was increased to possibly double dou ble that of Sol Solar ar 1. The sec second ond major imp impro rove vemen mentt wa wass mad madee to the con concen centra trator tor.. A largerr p large perce ercenta ntage ge of radiat radiation ion struck the absorber of Solar 2. The last ma major jor impro improvem vemen entt was made to the absorber. A cavity type absorber was used instead of a flat plate. This not only increased absorption but greatly decreased radiation and convection losses. Thermal losses were dete Thermal determine rmined d for eac each h compone component nt of Solar 2. The three larges largestt loss types were natural convection off the absorber, concentrator error and radiation from the absorber. These losses can be reduced without increasing system complexity or cost. Other losses were due to mirro mirrorr effici efficiency ency,, mirro mirrorr wea wear, r, absorp absorptivit tivity y, imperf imperfect ect insulation and receiv rec eiver er condu conducti ction on to the suppo support rt arm. The These se losse lossess mad madee up 29% of all los losses ses and wil willl be more difficult to reduce without increasing system cost. The correlation between the concentrator and absorber aperture area requires that any improvem impro vemen ents ts to either of the systems be made simult simultaneou aneously sly.. The concen concentrat trator or mus mustt be improved first, increasing the amount of reflected radiation that strikes inside the receiver cavity cav ity.. When this is accom accomplishe plished, d, the absorbe absorberr diameter can be optim optimized ized to maximize 67
radiation radiat ion colle collection ction while minimizin minimizingg radiat radiation ion and con conve vectiv ctivee losses losses.. Costs will actua actually lly b bee decreased here by using less materials in the receiver. The goal of prod produci ucing ng 6.67 kW of the therma rmall pow power er was not met met.. To ac achie hieve ve this this,, the thermal ther mal efficienc efficiency y of the system must be incre increased ased to 48%. The receiv receiver er used was optimized for a conc concent entrator rator that had an angula angularr error of thre threee times the indust industrial rial standar standard. d. Becau Because se the actual concentrator had a higher error, the receiver did not maximize the use of incoming radiation. Although the cavity absorber efficiency was 70%, the flat plate was inefficient in absorbing absorb ing radia radiation. tion. By impro improving ving the conc concent entrator rator,, more radiation would strik strikee inside the cavity and it is believed that the system efficiency would increase to the intended goal. The cost of the system exceeded exceeded the $1000 ob object jective ive set for this project. The materi material al costs for Solar 2 was $3,062. It is believed that the cost of materials can be brought down, but it is unclear whether reaching the goal is possible. The rising cost of materials, even in the span of constructing the concentrator, and the addition of a steam turbine and generator will make it difficult to reduce material costs to $1000.
5.1 5. 1
• Impr Improv ovee
Futur uture e Work ork
con concen centra trator tor geo geomet metry ry..
Thi Thiss wil willl mak makee the large largest st diff differe erence nce in sys system tem
efficiency.
• Decrease Decrease cavit cavity y apertur aperturee size size..
With an impro improved ved conce concentr ntrator, ator, the absorber radius
should be decr decreased eased to tak takee advan advantage tage of the improv improved ed optical efficienc efficiency y. This will decrease radiation and convection.
• Eliminate the flat plate absorber. Use only the cavity absorber and insulate all other surfaces on the receiver.
• Add a pump. A water pump will allow for steady steam production. • Correct Correct trac tracking king system. Progr Program am safety procedu procedures res to kee keep p trac tracking king system from moving when clouds block sunlight.
• Incre Increase ase boiler support. By stiffeni stiffening ng the receiv receiver er arm, the trackin trackingg will err less in the morning and evening.
• Add steam turbine. The system is now ready to test with a micro steam turbine. 68
APPENDIX A RECEIVER DRAWINGS
69
70
71
72
73
APPENDIX B CONNECTION ARM DRAWINGS
74
75
76
APPENDIX C CONCENTRATOR CONNECTION RING DRAWINGS
77
78
79
APPENDIX D FOUNDATION DRAWINGS
80
81
82
REFERENCES
Energy gy Conv Conver ersio sion n [1] [1] Dr Dr.. An Anja jane neyu yulu lu Kr Krot otha hapa pall lli. i. Ener //sesec.fsu.edu/lectures.html,, 2005. (document), //sesec.fsu.edu/lectures.html (document), 1.1
Sy Syst stem ems s .
http :
[2] Aden B Meinel and M Marjorie arjorie P P.. Meilen. Applied Solar Energy, An Introduction . AddisonWesley Publishing Company, Inc. Philippines, third edition edition, 1976. (document) (document),, 1.2.1,, 1 1.2 .2,, 1.2.1 1.2.1,, 2 2.1 .1,, 2.12, 2.12, 2.3.1, 2.3.1, 2.18 1.2.1 [3] U.S. Department of Energy’s Sun Lab. http : //www.energyl //www.ener gylan.sandia.gov/sunlab/P an.sandia.gov/sunlab/PDF DF s/boeing.pd s/boeing.pdf f ,, Aug August ust 2001. (document), (document), 1.3 1.3,, 1.2.1
∗
[4] [4] W. W.B. B. St Stin inee and R. R.W. W. Ha Harr rrig igan an.. Solar Solar Ener Energy gy System System and Design Design , 19 1985 85.. http : //www.powerfromthesun.net.htm.. (document) //www.powerfromthesun.net.htm (document),, 1.2.1, 1.2.1, 1.2.2, 1.2.2, 1.2.2, 1.2.2, 1.4 1.4,, 1.5 1.5,, 1.6, 1.6 , 2.1 [5] The Nebraska Astronomy Applet Pro je ject. (document),, 2.5 //astro.unl.edu/naap/motion33/sidereals ynodic.html //astro.unl.edu/naap/motion ynodic.html.. (document)
http
:
[6] M. Eck and E. Zarza Zarza.. Satur Saturated ated steam proces processs with direct steam genera generating ting parabolic (document),, 1.7, 1.7 , 1.2.2 troughs. Solar Energy , 80:1424–1433, 2006. (document) [7] Matt Ander Anderson. son. Compu Computer ter simu simulation lation of a solar dish powe powerr system for sustainabl sustainablee rural (document),, development. developmen t. Master’s thesis, Royal Institute of Technology echnology,, 20 2007. 07. Page 42. 42. (document) 2.2 2.2,, 2.3.3 2.3.3,, 2.3.3 2.3.3,, 2.4.4 [8] C. Christop Christopher her Newton Newton.. A concent concentrated rated solar therm thermal al energy system system.. Maste Master’s r’s thesis, Florida State University, 2007. (document) (document),, 1.9, 1.9 , 1.10, 1.10, 2.3, 2.3 , 2.17 [9] J.W. Spencer Spencer.. Fourie ourierr series repres represent entation ation of the p positio osition n of the sun. Journal of Solar Energy , 1971. (document) (document),, 2.4 Wolver erin inee Engin ngineeerin ring Da Data ta Book Book II . [1 [10] 0] Inc Wolve olveri rine ne Tube ube.. Wolv //www.wlv.com/products/databook/ch551 .pdf //www.wlv.com/products/databook/ch .pdf ,, 2001. (document) (document),, 2.21
http :
[11] Spirax Sar Sarco co USA. USA. http : http : //www.spiraxsarco.com/resources/steam //www.spiraxsarco.com/resources/steam engineering tutorials/the boiler house/water tube boilers.asp boilers.asp.. (document) (document),, 2.20
−
−
−
−
−
−
[12] Brenton Gresk Greskaa and Anjaneyulu Krothapalli. Tri-generation concen concentrating trating solar power system syste m for rural applic application ations. s. Tec echnica hnicall Report ES2007ES2007-36169, 36169, Sustainabl Sustainablee Energy Science and Engineering Center, 2007. 1.1
83
[13] T. Mancini Mancini,, P. Helle Heller, r, B. Butle Butler, r, B. Osborn, W. Schie Schiel, l, V. Goldber Goldberg, g, R. Buc Buck, k, R. Div Diver, er, C. Andraka, Andraka, and J. Moreno. DishDish-stirl stirling ing systems: An over overview view of develo developmen pmentt and status. Journal of Solar Energy , 125:135–151, 2003. 1.1 [1 [14] 4] Ri Riccha hard rd B. Ba Bann nner erot ot,, John John R. Ho How wel ell, l, an and d Ga Gary ry C. Vlie Vliet. t. Solar-Thermal Energy Energy 1.2.1,, 2.1 Systems, Analysis and Design . McGraw-Hill, Inc., 1982. 1.2.1 [15] L.C. Spence Spencer. r. A comprehe comprehensiv nsivee review of small solarsolar-pow powered ered heat eng engines: ines: Par Partt I. A history of solar-powered devices up to 1950. Journal of Solar Energy , 43(4):191–196, 1989. 1.2.1 [16] Robin M. G Green reen.. Spherical Astronomy . Cambridge Cambridge Unive Universit rsity y Press Press,, 1985. 1985. 2.2.4 [1 [17] 7] Fra rank nk Kr Krei eith th an and d Ja Jan n F. Kr Krei eide der. r. Princi Principle pless of Solar Solar Engine Engineerin ering g . Publishing Corporation, New York, 1978. 2.3.1
He Hemi misp sphe here re
[18] John A Duffie and Willi William am A Beckm Beckman. an. Solar Engineering of Thermal Processes . John Wiley & Sons, Inc., second edition, 1991. (document), (document), 2.2 2.2,, 2.8 [19] Donald Rapp. Solar Energy . Prentice-Hall, Inc. 1981. 2.3.1 [20] J. P P.. Holman. Heat Transfer . McGraw-Hill, ninth edition edition, 2002. 2.4.5 [21] Babcock & Wilco Wilcox x Compan Company y. Steam, Its Generation and Use . Geo Mc Mckib kibbin bin & Son, New York, 1955. 2.5.1
84
BIOGRAPHICAL SKETCH
John Dascomb
John Dascomb completed his Bachelors degree in Mechanic Mechanical al Engineering at Florida State University in the Spring of 2006. Under the advisement of Prof. Anjaneyulu Krothapalli he completed his Masters degree in the Spring of 2009. John’ss resea John’ researc rch h int interes erests ts inclu include de rene renewab wable le ener energy gy,, com combusti bustion, on, and heat trans transfer. fer. He will continue his studies under Dr. Krothapalli where he is currently enrolled in the doctoral program.
