<%BANNER%>

Theoretical Studies of UT-3 Thermochemical Hydrogen Production Cycle and Development of Calcium Oxide Reactant for UT-3 ...

Permanent Link: http://ufdc.ufl.edu/UFE0021178/00001

Material Information

Title: Theoretical Studies of UT-3 Thermochemical Hydrogen Production Cycle and Development of Calcium Oxide Reactant for UT-3 Cycle and Carbon Dioxide Capture
Physical Description: 1 online resource (147 p.)
Language: english
Creator: Lee, Man
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: absorbent, capture, co2, hydrogen, production, thermochemical, ut3
Mechanical and Aerospace Engineering -- Dissertations, Academic -- UF
Genre: Mechanical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Hydrogen can be a viable alternative energy carrier if it can be produced cost effectively from renewable resources. UT-3 thermochemical cycle to produce hydrogen from water is attractive because its temperature requirement is moderate (~700 degrees C) and it can be operated with solar or nuclear energy. There are still several issues that must be resolved before it becomes viable. These issues include developing high reactive surface area solid reactant structures which are able to go through large volume changes in a cycle while maintaining cyclic reactivity and strength of the solid reactants and speeding up of the hydrolysis reaction of calcium bromide, the rate limiting step in the cycle. In this study, thermodynamic feasibility investigation of the UT-3 process was conducted to determine the optimal operating conditions for high reaction rate as well as high conversion. A new calcium oxide reactant dispersed and immobilized on a yttria fabric was fabricated via an inexpensive and straightforward immobilization process. The performance of the sample was evaluated in cyclic bromination and hydrolysis reactions experimentally using the optimum conditions determined theoretically. The calcium oxide fabric showed continuous higher reactivity in four bromination reactions and the rate of hydrolysis reaction was faster than that of our calcium oxide pellets and comparable to that of calcium oxide pellets reported in the literature. The thermodynamic efficiency of the UT-3 cycle was estimated considering inert materials and incomplete conversion and heat recovery. It was found that the effects of inert materials and heat recovery on the efficiency were considerable while the influence of incomplete conversion was not significant. With heat recovery, the calculated efficiency for the calcium oxide fabric including inert materials and incomplete conversion was 52.4%. The use of calcium oxide on fabric was also studied for its application to high temperature carbon dioxide capture. The conventional calcium oxide absorbents could not maintained their performance in cyclic operations due to the reduction of active surface area. On the other hand, the new calcium oxide absorbent on fibrous alumina achieved continuous cyclic carbonation conversion over ten carbonation-calcination cycles under mild calcination condition. However, under the more severe calcination condition, its performance dropped by about eight percent after 12 cycles possibly due to the formation of Ca12Al14O33 by the reaction between calcium oxide and alumina. When calcium oxide was applied to yttria fabric, the absorbent maintained its performance for 12 cycles even under the severe calcination condition.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Man Lee.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Goswami, Dharendra Y.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021178:00001

Permanent Link: http://ufdc.ufl.edu/UFE0021178/00001

Material Information

Title: Theoretical Studies of UT-3 Thermochemical Hydrogen Production Cycle and Development of Calcium Oxide Reactant for UT-3 Cycle and Carbon Dioxide Capture
Physical Description: 1 online resource (147 p.)
Language: english
Creator: Lee, Man
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2008

Subjects

Subjects / Keywords: absorbent, capture, co2, hydrogen, production, thermochemical, ut3
Mechanical and Aerospace Engineering -- Dissertations, Academic -- UF
Genre: Mechanical Engineering thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Hydrogen can be a viable alternative energy carrier if it can be produced cost effectively from renewable resources. UT-3 thermochemical cycle to produce hydrogen from water is attractive because its temperature requirement is moderate (~700 degrees C) and it can be operated with solar or nuclear energy. There are still several issues that must be resolved before it becomes viable. These issues include developing high reactive surface area solid reactant structures which are able to go through large volume changes in a cycle while maintaining cyclic reactivity and strength of the solid reactants and speeding up of the hydrolysis reaction of calcium bromide, the rate limiting step in the cycle. In this study, thermodynamic feasibility investigation of the UT-3 process was conducted to determine the optimal operating conditions for high reaction rate as well as high conversion. A new calcium oxide reactant dispersed and immobilized on a yttria fabric was fabricated via an inexpensive and straightforward immobilization process. The performance of the sample was evaluated in cyclic bromination and hydrolysis reactions experimentally using the optimum conditions determined theoretically. The calcium oxide fabric showed continuous higher reactivity in four bromination reactions and the rate of hydrolysis reaction was faster than that of our calcium oxide pellets and comparable to that of calcium oxide pellets reported in the literature. The thermodynamic efficiency of the UT-3 cycle was estimated considering inert materials and incomplete conversion and heat recovery. It was found that the effects of inert materials and heat recovery on the efficiency were considerable while the influence of incomplete conversion was not significant. With heat recovery, the calculated efficiency for the calcium oxide fabric including inert materials and incomplete conversion was 52.4%. The use of calcium oxide on fabric was also studied for its application to high temperature carbon dioxide capture. The conventional calcium oxide absorbents could not maintained their performance in cyclic operations due to the reduction of active surface area. On the other hand, the new calcium oxide absorbent on fibrous alumina achieved continuous cyclic carbonation conversion over ten carbonation-calcination cycles under mild calcination condition. However, under the more severe calcination condition, its performance dropped by about eight percent after 12 cycles possibly due to the formation of Ca12Al14O33 by the reaction between calcium oxide and alumina. When calcium oxide was applied to yttria fabric, the absorbent maintained its performance for 12 cycles even under the severe calcination condition.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Man Lee.
Thesis: Thesis (Ph.D.)--University of Florida, 2008.
Local: Adviser: Goswami, Dharendra Y.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2009-08-31

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2008
System ID: UFE0021178:00001


This item has the following downloads:


Full Text

PAGE 1

1 THEORETICAL STUDIES OF UT-3 TH ERMOCHEMICAL HYDROGEN PRODUCTION CYCLE AND DEVELOPMENT OF CALCIUM OXIDE REACTANT FOR UT-3 CYCLE AND CARBON DIOXIDE CAPTURE By MAN SU LEE A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2008

PAGE 2

2 2008 Man Su Lee

PAGE 3

3 To my loving wife, Sam Jung; da ughter Alicia, and my parents

PAGE 4

4 ACKNOWLEDGMENTS I would like to sincerely tha nk my advisor, Dr. D. Yogi Goswami for all the advice and guidance he has always so gener ously provided. He has been a grea t mentor and teacher. I also express appreciation to my committee, Dr. Skip Ingley, Dr. William Lear, Dr. S. A. Sherif, and Dr. Samim Anghaie for their advice and s upport for the progress of my research. I would like to thank all the former memb ers and colleagues of the Solar Energy and Energy Conversion Laboratory at University of Florida for their assi stance and friendship. Specially, I thank Dr. Deepak Deshpande, Dr. Sa njay Vijayaraghavan and Dr. Nikhil Kothurkar for their valuable advice and effort. I have great appreciation for Chuck Garretson for his practical and critical support to he lp me pursue my experimental re search in the laboratory. Also I would like to thank Dr. Elias K. Stefanakos and all the staff members (specially Barbara Graham, Ginny Cosmides, Dr. Sesha Srinivasan and Dr. Nikolai Kislov) and also appreciate the friendship and cooperation of Mohammad Abutay eh, Huijaun Chen, Omatoyo Kofi Dalrymple, Gokeman Demirkaya, John Feddock, Ricardo Vas quez-Padillo, Paula Algarin-Amaris, Sam Wiejewardane, Drupatie Latchman and Jonathan Mb ah in the Clean Energy Research Center at the University of South Florida. I would like to thank Dr. Anghaie and all the staff in INSPI for the use of their facilitie s and friendly assistance. I also tha nk all the staff at PERC in UF and NNRC in USF who helped and trained me to use their equipments. I would also like to acknowledge the US Department of Energy fo r funding my researches on hydrogen production and carbon dioxide capture. I thank my parents, brother and sister for their constant encouragement and support from South Korea. Lastly, above all I thank my wi fe, Sam Jung Yang and would like to acknowledge that this doctoral degree would not be possible without her love and support.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4LIST OF TABLES................................................................................................................. ..........8LIST OF FIGURES................................................................................................................ .........9LIST OF ABBREVIATIONS........................................................................................................13ABSTRACT....................................................................................................................... ............15CHAPTER 1 INTRODUCTION..................................................................................................................171.1 Hydrogen Facts............................................................................................................. ....171.2 Motivation................................................................................................................. ........171.2.1 Issues Concerning the UT-3 Cycle.........................................................................181.2.1.1 Preparations of solid reactants......................................................................181.2.1.2 Slow reaction rate.........................................................................................181.2.1.3 Volume changes of solid reactants...............................................................191.2.1.4 Separation technology..................................................................................191.3 Objectives of Present Study..............................................................................................201.3.1 Thermodynamic Approach.....................................................................................201.3.2 Improving Performance of Solid Reactants...........................................................201.3.3 Evaluation of Thermal Efficiency..........................................................................201.3.4 Calcium Oxide Absorbent for High Te mperature Carbon Dioixde Capture..........212 HYDROGEN PRODUC TION METHODS...........................................................................222.1 Hydrogen Production from Fossil Fuels...........................................................................222.1.1 Steam Methane Reformation (SMR)......................................................................222.1.2 Partial Oxidation of Hydrocarbons (POX).............................................................232.1.3 Coal and Biomass Gasification..............................................................................242.2 Hydrogen Production from Water....................................................................................252.2.1 Water Electrolysis..................................................................................................252.2.2 Thermochemical Hydrogen Production.................................................................272.2.3 Photo-biological Hydrogen Production..................................................................282.3 Thermochemical Cycles for Hydrogen Production..........................................................282.3.1 Ispra Mark Processes..............................................................................................282.3.2 Iodine-Sulfur (IS) Cycle.........................................................................................312.3.3 ZnO/Zn Cycle.........................................................................................................322.4 Fresh Water Demands for Hydrogen Production.............................................................32

PAGE 6

6 3 THE UT-3 THERMOCHEMICAL CYCL E FOR HYDROGEN PRODUCTION...............383.1 Historical Survey of UT-3 Thermochemical Cycle..........................................................393.1.1 Studies of UT-3 processes and System..................................................................393.1.2 Development of Solid Reactant..............................................................................413.1.2.1 Development of Fe-pellets...........................................................................423.1.2.2 Development of Ca-pellets...........................................................................443.2 Strengths and Weaknesses of UT-3 Cycle........................................................................453.2.1 Strengths................................................................................................................ .453.2.2 Weaknesses.............................................................................................................453.3 Modification of UT-3 Cycle.............................................................................................464 THEORETICAL FEASIBILITY INVEST IGATIONS OF UT-3 CYCLE............................644.1 Thermodynamic Analysis of UT-3 Cycle.........................................................................644.1.1 Reaction 1: Bromination Reaction of Calcium Oxide............................................654.1.2 Reaction 2: Hydrolysis R eaction of Calcium Bromide..........................................674.1.3 Reaction 3: Bromination Reaction of Iron Oxide...................................................694.1.4 Reaction 4: Hydrolysis Reaction of Iron Bromide.................................................694.2 Optimum Conditions for UT-3 Cycle...............................................................................705 PREPARATION AND EVALUATION OF CALCIUM OXIDE REACTANT FOR UT-3 CYCLE..................................................................................................................... .....785.1 Calcium Oxide Pellets......................................................................................................785.1.1 Preparation of Calcium Oxide Pellets....................................................................795.1.2 Characterization......................................................................................................805.1.3 Kinetic Measurements............................................................................................815.2 Calcium Oxide Fabrics.....................................................................................................835.2.1 Immobilization of Calcium Oxid e on a Fibrous Yttria Fabric...............................835.2.2 Characterization......................................................................................................845.2.3 Cyclic Reaction Experiments.................................................................................845.2.4 Error Analysis.........................................................................................................865.3 Conclusion and Summary.................................................................................................866 THERMAL EFFICIENCY OF UT-3 PROCESS...................................................................996.1 Analysis Description....................................................................................................... ..996.2 Pinch Analysis............................................................................................................. .....996.3 Thermodynamic Analysis...............................................................................................1006.3.1 Ideal Case (Case 1: Complete C onversion and No Inert Materials)....................1006.3.2 Effect of Inert Materials on Efficien cy (Case 2: Complete Conversion and Including Inert Materials).............................................................................................1016.3.3 Effect of Inert Material and Incomp lete Conversion on Efficiency (Case 3: Including inert material and in complete conversion (CaO: 85%, Fe3O4: 90%))..........1026.4 Summary of Thermodynamic Analysis..........................................................................103

PAGE 7

7 7 CALCIUM OXIDE ABSORBENT FOR HIGH TEMPERATURE CARBON DIOIXDE CAPTURE..........................................................................................................1157.1 Introduction............................................................................................................... ......1157.2 Experimental Procedure..................................................................................................1177.2.1 Immobilization of Calcium Oxide on a Ceramic Fabric......................................1177.2.2 Cyclic Reaction Experiment.................................................................................1187.2.3 Characterization....................................................................................................1217.2.4 Severe Calcination Condition...............................................................................1227.3 Summary.................................................................................................................... .....1248 CONCLUSIONS AND FUTURE WORK...........................................................................1348.1 Conclusions................................................................................................................ .....1348.2 Recommendations for Future Work...............................................................................135APPENDIX CYCLIC CONVERSION PROFILES (A LUMINA, SILICA, ZIRCONIA)......137LIST OF REFERENCES.............................................................................................................139BIOGRAPHICAL SKETCH.......................................................................................................147

PAGE 8

8 LIST OF TABLES Table page 3-1 Properties of iron solid reactant pellets..............................................................................634-1 Optimum conditions for the bromination of calcium oxide...............................................774-2 Optimum conditions for the hydrolysis of calcium bromide.............................................774-3 Optimum conditions for the bromination of iron oxide.....................................................774-4 Optimum conditions for the hydrolysis of iron bromide...................................................774-5 Optimal conditions and expected conversi ons of the reactions in UT-3 cycle with 100H2O/H2........................................................................................................................775-1 Composition conditions for the pe llets with different additives........................................985-2 Characteristics and experi mental data of Ca-pellets..........................................................986-1 Thermodynamic data of the reactions in UT-3 cycle.......................................................1116-2 Energies required and rejected (Entha lpy change for heating and cooling based on latent heat and sensible heat)...........................................................................................1116-3 Thermal efficiencies with complete conversion and no inert materials (Case 1)............1116-4 Additional energies required and rejected for heating a nd cooling of inert material CaO:CaTiO3=1:1.4*, Fe3O4:Fe2TiO5=1:5** (Pellet).........................................................1126-5 Additional energies required and rejected for heating a nd cooling of inert material CaO:Y2O3=1:2*, Fe3O4: Y2O3=1:5** (Fabric)..................................................................1126-6 Thermal efficiencies with complete conversion and inert material (Case 2)...................1126-7 Additional energies required and rejected for heating a nd cooling of inert material and unreacted solid reactant, CaO:CaTiO3=1:1.4*, Fe3O4:Fe2TiO5=1:5** (Pellet)..........1136-8 Additional energies required and rejected for heating a nd cooling of inert material and unreacted solid reactant, CaO:Y2O3=1:2*, Fe3O4: Y2O3=1:5** (Fabric)...................1136-9 Thermal efficiencies including inert ma terial & incomplete conversion (Case 3)..........1147-1 Mild and severe calcination conditions............................................................................133

PAGE 9

9 LIST OF FIGURES Figure page 2-1 Natural gas steam reforming..............................................................................................342-2 Gasification-based energy conversion system...................................................................342-3 Energy demands for water electrolysis..............................................................................352-4 Simplified model of thermochemical cycle for hydrogen production...............................352-5 Photo-biological hydrogen production..............................................................................362-6 The IS thermochemical cycle.............................................................................................362-7 The ZnO/Zn cycle for hydrogen production......................................................................373-1 Flowsheet of the adiabatic UT-3 thermochemical cycle....................................................473-2 Experimental apparatus for kinetic studies........................................................................473-3 Conceptual plant design for UT-3 cycle............................................................................483-4 The MASCOT plant...........................................................................................................483-5 New flow scheme for the UT-3 process............................................................................493-6 Experimental set-up for MASCOT plant...........................................................................503-7 Pathway efficiencies....................................................................................................... ...503-8 Pelletization process for th e solid reactant pellets.............................................................513-9 Experimental apparatus for the kinetic tests of pellets......................................................523-10 Experimental apparatus for the kinetic tests of pellets......................................................533-11 Bromination and hydrolysis conversion of the Fe2O3 pellets............................................543-12 Initial conversion pr ofile of Fe-pellets...............................................................................553-13 Flowsheet for the preparation of Fe-pellets.......................................................................563-14 Conversions of the bromin ation in cyclic operation..........................................................573-15 Preparation steps of Ca-pellets...........................................................................................573-16 Conversion profiles of Ca-pellets......................................................................................58

PAGE 10

10 3-17 New preparati on method for Ca-pellets.............................................................................593-18 Comparison of the pellets by the co nventional and new modified method.......................603-19 Change of Gibbs free energy vs. temp erature for the Ca-Br cycle and water electrolysis................................................................................................................... ......603-20 Conceptual design of a fluidized bed for UT-3 cycle........................................................613-21 Conceptual molten salt-based reactor fo r hydrolysis/bromination in the CaBr cycle.....624-1 Changes of Gibbs free energy of reactions as a function of reaction temperature............714-2 Effect of excess steam on bromination r eaction of calcium oxide at 873K and 1 atm......714-3 Effect of operation temperature on brom ination reaction of calcium oxide at 1 atm with 100H2O/H2.................................................................................................................724-4 Effect of operation pressure on brominat ion reaction of calcium oxide at 873K with 100H2O/H2.........................................................................................................................724-5 Effects of excess steam on hydrolysis of calcium bromide at 1000K and 1 atm...............734-6 Effect of temperature on hydrolysis reaction of calcium bromide with 100H2O/H2 at 1 atm.......................................................................................................................... .........734-7 Effect of pressure on hydrolysis reaction of calcium bromide with 100H2O/H2 at 1000K.......................................................................................................................... .......744-8 Effect of HBr removal from equilibrium states with 100H2O/H2 at 1000K and 1 atm.....744-9 Effect of temperature on bromin ation reaction of iron oxide with 100H2O/H2 at 1 atm...754-10 Effect of operation pressure on the bromin ation reaction of iron oxide at 400K with 100H2O/H2.........................................................................................................................754-11 Effect of temperature on hydrolysis reaction of iron bromide at 1 atm with 100H2O/H2.........................................................................................................................764-12 Effect of excess steam on hydrolysis r eaction of iron bromide at 800K and 1 atm..........765-1 The SEM images of pore forming agents..........................................................................885-2 The TGA curve of a non-sintered calcium oxi de pellet at a heati ng rate of 10C /min in air......................................................................................................................... ..........885-3 Sintering steps for Ca-pellets.............................................................................................895-4 Flow diagram for the preparation of Ca-pellets.................................................................89

PAGE 11

11 5-5 The XRD pattern of the fresh calcium oxide pellet...........................................................905-6 Pore size distributions of Ca-pellets (Sample 1-4) prepared with different pore forming agents................................................................................................................. ..905-7 Laboratory Set-up.......................................................................................................... ....915-8 Schematic Diagram for Laboratory Set-up........................................................................915-9 Cyclic conversion profiles of a calcium oxide pellet (Sample: S4)...................................925-10 Changes of pore size distribution after br omination and hydrolysis in a calcium oxide pellet......................................................................................................................... ..........925-11 Comparison of pore volumes of initial, af ter bromination and after hydrolysis in a calcium oxide pellet...........................................................................................................935-12 Changes of Gibbs free energy of reac tions between calcium oxide and ceramic materials as a function of reaction temperature.................................................................935-13 Preparation steps for impregnation of calcium oxide on a yttria fabric.............................945-14 The XRD patterns of the samples......................................................................................955-15 Conversion profiles of cyc lic reactions, bromination a nd hydrolysis, of the calcium oxide fabric sample............................................................................................................955-16 The SEM images of samples..............................................................................................965-17 Comparison of the first cyclic reactions of calcium oxide fabric samples with various amounts of calcium oxide in the fabric sample.................................................................976-1 Temperature-enthalpy diagrams......................................................................................1056-2 Process flowsheet of heat a nd material in the UT-3 cycle...............................................1066-3 Temperature-enthalpy diagrams for Case 1.....................................................................1076-4 Combined composite curves (Case 2)..............................................................................1086-5 Combined composite curves (Case 3)..............................................................................1096-6 Comparison of thermal effici ency in various situations..................................................1107-1 Preparation steps for immobilization of calcium oxide on a ceramic fabric...................1257-2 Carbonation conversion of the prepared sample..............................................................1257-3 Conversion profiles of cyclic reactions............................................................................126

PAGE 12

12 7-4 Maximum conversions of carbonation reac tion of two samples loaded different calcium oxide contents, 23 wt % and 55 wt %, with the number of cycles.....................1267-5 Maximum amounts of reacted calcium oxi de in the carbonation reactions based on initial sample weight with the number of cycles.............................................................1277-6 The SEM images of a fabric sample................................................................................1277-7 The XRD patterns of the samples....................................................................................1287-8 Change of surface area in the samp le over the several cyclic reactions..........................1297-9 The SEM pictures for the sample.....................................................................................1317-10 The cyclic maximum carbonation convers ions of the samples using yttria and alumina as a substrate under the severe calcination condition at 850C and 20 wt % CO2............................................................................................................................... ....132A-1 Cyclic conversion profiles of a calcium oxide sample immobilized on an alumina fabric......................................................................................................................... .......137A-2 Cyclic conversion profiles of a calcium oxide sample immobilized on a silica fabric....138A-3 Cyclic conversion profiles of a calcium oxide sample immobilized on a zirconia fabric......................................................................................................................... .......138

PAGE 13

13 LIST OF ABBREVIATIONS a activity E equilibrium potenti al or voltage (V) F Faraday constant (96, 500 C mol-1) G change of Gibbs free energy (kJ/mol) H enthalpy (kJ/mol) H change of enthalpy change (kJ/mol) HHV Higher Heating Value (kJ/mol) HTGR High Temperature Gas-cooled Reactor IGCC Integrated Gasification Combined Cycle LHV Lower Heating Value (kJ/mol) M molecular weight n moles of electrons N number of cycle P pressure (atm) R mole fraction S change of entropy change (kJ/mol K) SMR Steam Methane Reforming T Temperature (K or C) Tmin minimum approach temperature W weight of sample Wo weight of unreacted sample X conversion

PAGE 14

14 Subscripts: a anode c cathode in input

PAGE 15

15 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy THEORETICAL STUDIES OF UT-3 TH ERMOCHEMICAL HYDROGEN PRODUCTION CYCLE AND DEVELOPMENT OF CALCIUM OXIDE REACTANT FOR UT-3 CYCLE AND CARBON DIOXIDE CAPTURE By Man Su Lee August 2008 Chair: D. Yogi Goswami Major: Mechanical Engineering Hydrogen can be a viable alternative energy ca rrier if it can be produced cost effectively from renewable resources. UT-3 thermochemi cal cycle to produce hydr ogen from water is attractive because its temperatur e requirement is moderate (~700 C) and it can be operated with solar or nuclear energy. There are still several is sues that must be resolved before it becomes viable. These issues include developing high reac tive surface area solid r eactant structures which are able to go through large volume changes in a cycle while maintaining cyclic reactivity and strength of the solid reactants and speeding up of the hydrolysis re action of calcium bromide, the rate limiting step in the cycle. In this study, thermodynamic feasibility inves tigation of the UT-3 process was conducted to determine the optimal operating conditions for high reaction rate as we ll as high conversion. A new calcium oxide reactant dispersed and immobili zed on a yttria fabric was fabricated via an inexpensive and straightforward immobilization process. The pe rformance of the sample was evaluated in cyclic brominati on and hydrolysis reactions expe rimentally using the optimum conditions determined theoretically. The calci um oxide fabric showed continuous higher reactivity in four bromination reac tions and the rate of hydrolysis re action was faster than that of

PAGE 16

16 our calcium oxide pellets and comparable to th at of calcium oxide pe llets reported in the literature. The thermodynamic efficiency of th e UT-3 cycle was estimated considering inert materials and incomplete conversion and heat re covery. It was found that the effects of inert materials and heat recovery on the efficiency were considerable while the influence of incomplete conversion was not significant. With h eat recovery, the calculated efficiency for the calcium oxide fabric including inert mate rials and incomplete conversion was 52.4%. The use of calcium oxide on fabric was also st udied for its applicati on to high temperature carbon dioxide capture. The convent ional calcium oxide absorbents could not maintained their performance in cyclic operations due to the reduction of active surface area. On the other hand, the new calcium oxide absorbent on fibrous al umina achieved continuous cyclic carbonation conversion over ten carbonation-calcination cycles under mild calcination condition. However, under the more severe calcination condition, its performance dropped by about eight percent after 12 cycles possibly due to the formation of Ca12Al14O33 by the reaction between calcium oxide and alumina. When calcium oxide was applied to yttria fabric, the absorbent maintained its performance for 12 cycles even under the severe calcination condition.

PAGE 17

17 CHAPTER 1 INTRODUCTION 1.1 Hydrogen Facts Hydrogen is the lightest gas a nd also the simplest element ever known. It is colorless, odorless, non-toxic and non-corrosi ve. Hydrogen is considered as a promising energy carrier because it is the most abundant element on earth and can be produc ed and used without or with little generation of pollution. However, hydrogen mu st be extracted from water or fossil fuels such as coal, natural gas and petroleum, sin ce hydrogen is not availa ble freely in nature. Hydrogen also represents a promising energy st orage means for renewable energy sources which are intermittent (solar radiati on and wind) or may not be conv enient for some applications (biomass). Presently about 9 million tons of hydrogen is pr oduced per year in the United States and it is mostly used for chemical industries such as fertilizer and petroche mical industries. About 150 million tons of hydrogen would be needed for fu el-cell vehicles annually by 2040, considering a transition from petroleum to hydrogen as a fuel for transportation (Argon ne National Laboratory, 2003). 1.2 Motivation For hydrogen alternative to become feasible, cost effective and pollution free methods to produce hydrogen must be developed. Currently, there are a number of methods of producing hydrogen which are available commercially or are under research. The processes currently available commercially are steam methane refo rmation (SMR), water electrolysis, partial oxidation of heavy hydrocarbons (POX) and coal gasification. Additional methods being investigated at the laborator y level include thermochemical cycles, biomass gasification, photoelectrical, photochemical and biological processes. Regardle ss of the maturity of the

PAGE 18

18 technologies, they all fall under two basic ca tegories: 1. Hydrogen production from carbon based fuels such as fossil fuels. 2. Hydrogen producti on by splitting water. The second category is more attractive in a long term prospectiv e because it is clean and renewable. A thermochemical cycle is a clean and efficient hydrogen production method because it splits water into hydrogen and oxygen using heat as the energy source. Among the several hundred thermochemical cycles proposed so fa r, UT-3 cycle was selected for this study considering its strengths de scribed in the section 3.2. 1.2.1 Issues Concerning the UT-3 Cycle In order to make the UT-3 cy cle practically feasible, there are still several barriers and issues to be investigated a nd solved, as described below. 1.2.1.1 Preparations of solid reactants UT-3 cycle is comprised of four heterogeneous reactions. In order to simplify the product separation in the cyclic system and operate th e process continuously, th e solid reactants and products should remain in the reactors while the gaseous reactants and products move from one reactor to the next. Accelerati ng the reaction rate, and increasi ng cyclic life time and durability of the solid reactants are importa nt keys for making the cycle practical. Preparation of the solid reactants with inert structural materials has been introduced and developed (Aihara et al., 1990 and 1992; Amir et al., 1993; Sakur ai et al., 1995 and 2006), but the preparation procedures seem rather complicated and expensive and the practicali ty of the pellet-type reactant is still in doubt. Lemort et al. (2006) also suggested the possibi lity of decomposition of the inert material, calcium titanate (CaTiO3), in the pellets by HBr. 1.2.1.2 Slow reaction rate In order to improve the process efficiency, the hydrolysis reaction of hydrogen bromide, which is the slowest and rate limiting step among the four reactions in the cycle, must be

PAGE 19

19 accelerated. The reaction rate can be enhanced by increasing the reaction temperature since the reaction is endothermic. However, the reaction temperature is limited because of the melting point of calcium bromine. The c onversion with excess steam and at lower pressures is expected to be higher according to the Le Chateliers princi ple. Therefore, optimal process conditions for the hydrolysis reaction must be determined to speed up the reaction rate The reaction rate may be also enhanced by providing a large surface ar ea between the gas and solid reactant particles with the elimination of diffusion resistances. 1.2.1.3 Volume changes of solid reactants The hydrolysis reaction of calcium bromide is the kinetically slowest reaction in the UT-3 cycle. Besides, it was found that about 76 perc ent increase in the mola r volume of the solid reactants (Simpson et al., 2007) from calcium oxide to calcium bromide during the bromination reaction, accounted for a significant decrease of porosity of the pe llets (Aihara et al., 1990; Lee et al., 2006). This reduction leads to low reactivity and degradation of the pellets. So, in order to overcome these disadvantages, development of a dur able porous reactant that maintains cyclic reactivity as well as has an enduring structure in cyclic transformations is required. 1.2.1.4 Separation technology The difficulty of separation of hydrogen and oxyge n from the process stream is one of the major factors affecting the efficiency of the proce ss. In the adiabatic UT-3 cycle, excess steam is used as a heat carrier (Sakurai et al., 1996a) and it was proved that the additional steam improved the hydrolysis process. Hydrogen and oxygen are produced as mixed w ith highly corrosive components such as Br2 and HBr as well as steam at very high temp eratures in the proce ss. If hydrogen and oxygen are separated from other gaseous products by a conventional condensatio n method, it will result in severe energy loss. According to Teo et al. (2 005), UT-3 cycle is economically feasible only if

PAGE 20

20 high temperature separation techno logy is realized without the unwanted energy loss. For that reason, ceramic membranes that have corrosion resistance at high temp eratures and other corrosive conditions have been investigated (Ohy a et al., 1994; Morooka et al., 1996; Ohya et al., 1997). Separation by membrane is not efficient if the concentrations of hydrogen and oxygen in the stream are low. In order to increase the se paration efficiency, either high pressure or high surface area of the membrane is required (T-Raissi, 2005). Howeve r, high operating pressure has an adverse effect on the hydrogen yi eld and increasing the surface area of the membrane is costly. 1.3 Objectives of Present Study 1.3.1 Thermodynamic Approach First of all, comprehensive theoretical feasibili ty of each reaction in the UT-3 cycle will be investigated and optimal operating conditions for high conversion and reaction rate will be determined via a thermodynamic analysis. Ther mal efficiency of the UT-3 cycle will be evaluated at the optimum conditions. 1.3.2 Improving Performance of Solid Reactants The life-time of the solid reactants is the mo st important factor for commercialization of the UT-3 cycle as mentioned earlie r. In order to make the cycle practical and cost-effective, the solid reactants must be chemically reactive and physically stable in cyclic operation and the reaction rates must be accelerated. A goal of this research is to develop a calcium oxide reactant with more favorable characteristics such as c ontinuous high cyclic reac tivity, high reaction rates and simple reactant preparation step. 1.3.3 Evaluation of Thermal Efficiency Thermal efficiency of the UT-3 cycle has been evaluated by several au thors. But the inert materials in the solid reactants, incomplete conv ersion and heat recovery were not considered as a major factor in the calculations so far. Thes e factors would have an important effect upon the

PAGE 21

21 efficiency as well as the operati onal costs since the solid reacta nts contain considerable amounts of inert materials and the conversions were found to be incomplete through the experiments. In this study, a practical thermodynami c efficiency with and without heat recovery, inclusion of inert materials, and with incomplete conversi on will be evaluated at the optimum conditions obtained by theoretical studies, to improve the pr ocess performance and reduce the uncertainties. 1.3.4 Calcium Oxide Absorbent for High Temperature Carbon Dioixde Capture Calcium oxide, an important reactant in the UT -3 cycle, is also a well-known absorbent of carbon dioxide (CO2), a major greenhouse gas. Carbon dioxide reacts with calcium oxide to form calcium carbonate in the carbonatio n reaction and the calcium oxi de is regenerated and pure carbon dioxide can be obtained th rough the calcinati on reaction. However, degradation of its performance due to the loss in active surface area, pore plugging a nd sintering of the particles must be overcome or minimized for the calcium oxi de absorbent to be pract ical. In this study, the calcium oxide immobilized on a fibrous ceramic fa bric for the UT-3 cycle was introduced to enhance the cyclic performance of high temperature carbon dioxide capture. The characteristics and cyclic performance of the proposed immobili zed calcium oxide on the fabric were examined and compared with the previ ous results in the literature.

PAGE 22

22 CHAPTER 2 HYDROGEN PRODUCTION METHODS 2.1 Hydrogen Production from Fossil Fuels Most of the hydrogen consumed presently in our world is produced from fossil fuels since it is currently the most economical and efficient way. However, fossil fuel resources are limited. Therefore these methods will lose their cost co mpetitiveness in the l ong run. Additionally, they cause environmental pollution. 2.1.1 Steam Methane Reformation (SMR) Steam methane reformation (SMR), sometimes referred to as steam reforming of natural gas, is the most commercialized way to produce hydrogen since it is the least expensive means at present. Today, about 95% of the hydrogen produ ced in the U.S. is made through SMR (U. S. DOE, 2002). Basic SMR process is composed of two chemical reactions involving methane, water, and a catalyst along with heat supply (Casper, 1978), as shown below:. Reformation of Methane 2 2 43 H CO heat O H CH ` (2-1) Water-Gas Shift reaction heat H CO O H CO 2 2 2 (2-2) Mixture of methane and steam at high temper ature (1000K~1300K) is cat alytically converted into carbon monoxide and hydrogen at the first step (steam reformi ng reaction (2-1)). The product gases are then reacted on a packed bed of catalyst to produce ca rbon dioxide and more hydrogen by the water-gas shift reaction (2-2). Approximately 71-75 mol % hydrogen with other gases such as carbon dioxide, carbon monoxide, a nd methane is obtained through these two steps. This mixture is purified by condensation and pr essure swing adsorption at high purity level.

PAGE 23

23 Overall energy and exergy efficiencies of th e SMR process that produces 97 wt % hydrogen were determined to be about 86% and 78%, respectivel y (Rosen, 1996). A simplified schematic for SMR using natural gas as a methane source is outlined in Figure 2-1. More than 80% of the hydrogen used around the world is currently produced by SMR (National Research Council and National Academ y of Engineering, 2004). The cost of hydrogen production from the SMR process strongly depends on the cost of natural gas, a major methane source (Goel et al., 2003). Moreove r, generation of carbon dioxide and release of methane during conversion are inevitable. In or der to reduce the emission of CO2, the CO2 could be sequestered with about 25-30% additional capital a nd operational costs (Goel et al., 2003). 2.1.2 Partial Oxidation of Hydrocarbons (POX) Partial oxidation (POX) is analogous to SMR while the addition of a partial oxidation reaction of hydrocarbons. Therefore, the POX pro cess includes three steps shown below (Goel et al., 2003). Partial Oxidation of Hydrocarbon heat H m nCO nO H Cm n 2 22 (2-3) Reformation of Hydrocarbon 2 22 2 H m n nCO heat O nH H Cm n (2-4) Water-Gas Shift reaction heat H CO O H CO 2 2 2 (2-5) Hydrocarbons are partially bur ned with less than the stoich iometric amount of oxygen and are converted into carbon monoxide and hydrogen in a partial oxidation reac tion (2-3) with heat rejection. The remaining hydrocarbons with st eam are catalytically converted into carbon

PAGE 24

24 monoxide and hydrogen at the steam reforming st ep using thermal energy from the partial oxidation. Air separation units would be needed to a void NOx which would be generated from the fuel combustion in air. Therefore the hydrogen production cost from POX is higher than that from the SMR process due to this additional capital cost for the separation (Goel et al., 2003). 2.1.3 Coal and Biomass Gasification Gasification of coal and biomass is consid ered as one of best methods for hydrogen production because of its environmental perfor mance, high efficiency and abundant resource available. Coal and biomass gasification for hydrogen prod uction are very similar to partial oxidation except for the operating temperature needed and the types of feedstock. Gasification requires much higher temperature (1100C -1300C) and uses various types of solid-based feedstocks such as coal, petroleum coke and biomass rather than liquid or gaseous hydrocarbon. Primary reactions of coal and biomass gasifi cation for hydrogen production are identical to the reactions, (2-3) ~ (2-5), for POX. Coal ga sification was to produce 93% hydrogen with about 59% and 49% of overall energy and exergy efficiencies, respectively (Rosen, 1996). The synthesis gas, a mixture of carbon monoxi de and hydrogen, generated by the reactions, (2-3) and (2-4), can be used to generate electricity through integrated gasification combined cycle (IGCC) or reacted with steam to increase the H2 yield by water-gas shift reaction (2-5). A schematic diagram of the process to generate both electricity and hydroge n in one gasificationbased system is illustrated in Figure 2-2. In addition to the capital cost reduction and efficiency improvement, research on the coal and biomass gasification is also focused on the development of carbon dioxide capture and

PAGE 25

25 sequestration technologies and the membra nes for hydrogen separation from other gases (National Research Council and Nationa l Academy of Engineering, 2004). 2.2 Hydrogen Production from Water In order to make the hydrogen economy feasib le or at least make hydrogen an attractive energy carrier for some applications, hydrogen mu st be produced ultimately from water via economical and environment-friendly ways since fossils fuels are limited and generate carbon dioxide. Hydrogen production from water, however, is st ill in the research and development stage except water electrolysis while most of hydroge n production technologies from fossil fuels are commercialized today. 2.2.1 Water Electrolysis Water electrolysis is a de veloped technology for hydrogen production. However, it has not been used widely because the energy source, elec tric energy, is expensive. This electrolysis technique may become more viable in the future when electricity is generated from renewable energy. Hydrogen and oxygen are produced by electrolysis at the cathode and anode, respectively. The two reactions at the cathode a nd anode are as under (Wendt, 1990): Cathode: OH H e O H2 2 22 2 (2-6) Anode: e O H O OH2 2 1 22 2 (2-7) Overall net reaction is represented as: Overall: 2 2 22 1 ) (O H y Electricit Energy O H (2-8)

PAGE 26

26 The equilibrium potentials at the cathode, Ec, and anode, Ea, at 25C are described as (Ohta, 1979): OH ca E log 059 0 828 0 (2-9) OH aa E log 059 0 401 0 (2-10) where a is the activity. Using these rela tions, the lowest electric voltage, Ea-Ec, called the reversible voltage at 25C is 1.229V. Therefore, the required mi nimum electrical energy for the water electrolysis is calculated to be 237.18kJ by the following equation: nFE G (2-11) where n is moles of electrons, F th e Faraday constant (96, 500 C mol-1) and E voltage in volts (V). The changes in total energy and electrical energy demands for the water electrolysis are approximately displayed with temp erature in Figure 2-3. The de ference between the total energy demand ( H) and the minimum electrical energy demand ( G) is the required heat energy (T S) for the electrolysis with the minimum electrical energy. As shown in Figure 2-3, the minimum electric energy demand decreases wi th temperature. Therefore, efficiency of the electrolysis can be improved by increasing the temperature because inexpensive thermal energy can replace a part of the costly electric energy. However, the improvement is limited due to the corrosion problems of the materials and the evaporation of water (Cox and Williamson, 1977). As a result, the development of high temperat ure electrolysis (solid oxide electrolyzer operated at 700 to 1000C) is underway to improve the performance by replacing th e electric energy with thermal energy (Riis et al., 2006).

PAGE 27

27 Main types of electrolyzers are alkaline, solid polymer (SPE) and solid oxide electrolyzers which are named after the electrol ytes. Current research on water electrolysis is focused on cost reduction and performance improvement. 2.2.2 Thermochemical Hydrogen Production In principle a thermochemical hydrogen producti on cycle is similar to high temperature thermal cracking in that it requires only water as the raw material and heat as the energy source. The cycle contains a combination of at least two recurring chemical reactions operated while high temperature thermal cracking has one direct decomposition reaction. By this approach, water can be separated into hydrogen and oxygen at comparatively lower temperatures than thermal cracking and in separate steps, while the sum of the reactions results in the decomposition of water. Thermochemical processes for hydrogen pr oduction were suggested in the 1960s to produce hydrogen from water as a mo re efficient or cheaper method as compared to electrolysis (Funk and Reinstrom, 1966). The concept can be simply described (Beghi 1986) as shown in Figure 2-4. Water is supplied with heat into the system which contro ls a combination of thermochemical reactions. Hydrogen and oxygen are produced separately th rough the process. All chemicals involved except water, hydrogen and oxygen ar e recycled in the process. In view of the fact that a thermochemical process does not create any pollution and may be coupled with thermal energy sources like solar ener gy, nuclear power or wast e heat, it appears to be the most attractive method to produce hydrog en. Some of the thermochemical cycles under study will be discussed later in this chapter.

PAGE 28

28 2.2.3 Photo-biological Hydrogen Production Photo-biological process for hydrogen production consists of the following two reactions, photosynthesis and hydrogen pr oduction (Riis et al., 2006): Photosynthesis 2 24 4 2 O e H O H (2-12) Hydrogen Production 22 4 4 H e H (2-13) The process uses microbes such as green alga e and cyanobacteria to promote the reaction in the presence of sunlight and water. Thr ough the natural organic processes, hydrogen and oxygen are produced. A schematic diagram of th e photo-biological hydro gen production process is displayed in Figure 2-5. Photo-biological hydrogen pr oduction requires low initial investment and low energy, however, the process is still unde r development (Goel et al., 2003). 2.3 Thermochemical Cycles for Hydrogen Production Several hundred thermochemical cycles fo r hydrogen production have been proposed and approximately eight hundred publications have been released thr ough 1999 (Funk, 2001) since the concept was introduced by Funk and Reinst rom (1964 and 1966). In the following section and chapter, several historically important and commercially potential thermochemical cycles will be discussed. 2.3.1 Ispra Mark Processes Mark 1 cycle was developed at the Joint Re search Center, Ispra based on preliminary studies on the thermochemical hydrogen production performed through thermodynamic calculations. The reactions in the Mark 1 cycle are shown below: HBr OH Ca O H CaBrK2 ) ( 22 1050 2 2 (2-14) 2 2 4502 H HgBr Hg HBrK (2-15)

PAGE 29

29 O H HgO CaBr OH Ca HgBrK 2 2 450 2 2) ( (2-16) 2 9002 1 O Hg HgOK (2-17) Both experimental kinetics studies of the reac tions in Mark 1 and corrosion tests for material selections were carried out thoroughly and the results showed that th e cycle was feasible. However research on this process was abandone d because of mercury use in the process. Subsequently, Mark 2, 3 and 6 cycles without me rcury were introduced and evaluated, but it was found that they were not promising (Beghi, 1986). During early 1970s, Mark 7, 9, 14 and 15 cy cles involving iron and chlorine were developed. Among them, Mark 15 cycle shown belo w was extensively studi ed and evaluated in detail. Mark 15 cycle 2 4 3 973 923 2 26 4 3 H HCl O Fe O H FeClK (2-18) O H FeCl FeCl HCl O FeK 2 3 2 573 473 4 34 2 8 (2-19) 2 2 593 553 32 2 Cl FeCl FeClK (2-20) 2 973 873 2 22 1 2 O HCl O H ClK (2-21) Through comprehensive experiments and analyses, th e cycle was proved to be feasible while the calculated thermal efficiency was somewhat low, about 20%, and highly co rrosive chlorine was considered to be a problem. Nevertheless, rese arch on the Mark 15 process was stopped since technical problems on the scale-up of the hydrol ysis reaction of ferr ous chloride (FeCl2) and chemically unfavorable thermal deco mposition of ferric chloride (FeCl3) were not solved (Beghi, 1986).

PAGE 30

30 Sulfur based cycles for hydrogen production such as Mark 11, 13, and 16 were also widely investigated at the Jo int Research Center. The Mark 11 cycle shown below includes an el ectrochemical reaction for the production of sulfuric acid (Brecher et al., 1977) Mark 11 cycle 2 2 2 1123 4 22 1 O SO O H SO HK (2-22) 2 4 2 393 353 2 22 H SO H O H SOK (Electrochemical step) (2-23) This two-step hybrid sulfur pr ocess is also known as the West inghouse sulfur cycle and GA-22. The cycle was successfully operate d at a bench scale and a pilot scale by Westinghouse Electric Corporation and Joint Research Center, respectively (B rown et al., 2003). The calculated overall process efficiency was about 41% (Beghi, 1986). However, in orde r to make the cycle practical, scale-up problems on the electrochemical process still need to be solv ed (Brown et al., 2003). A variation of the above cycl e is Mark 13 (Velzen et al., 1980) which is given below: 2 2 2 1100 1000 4 22 1 O SO O H SO HK (2-24) HBr SO H SO Br O HK2 24 2 370 320 2 2 2 (2-25) 2 2 3502 Br H HBrK (Electrochemical st ep) (2-26) The process was proposed by Schltz and Fieb elmann in 1974 and experimental studies including bench-scale operation were successfully conducted by th e Joint Research Center and the overall process efficiency was estimated to be about 39% (Beghi, 1986). The overall efficiency of the Mark 13-V2 process coupled with a central solar receiver system was about 21% (Bilgen and Joels, 1985). However, it was found that this hybrid cycle also had scale-up problems on the electrode system engaged in thin membranes (Brown et al., 2003).

PAGE 31

31 Mark 16 cycle uses iodine instead of bromine in the Mark 14 cycle. This cycle is also known as Sulfur-Iodine or Iodine-Sulfur cycle (IS cycle) which was investigated at the General Atomic (GA) Corp. The detailed history of the IS cycle will be discussed in the next section. 2.3.2 Iodine-Sulfur (IS) Cycle Unlike other sulfur family cycles such as Mark 11 and 13, IS cycle does not require electricity for the r eactions shown below. 2 2 2 1123 4 22 1 O SO O H SO HK (2-27) HI SO H SO I O HK2 24 2 393 2 2 2 (Bunsen Reaction) (2-28) 2 2 7232 I H HIK (2-29) The system for the cycle consists of three parts, Bunsen reaction system, H2SO4 decomposition system, and HI decomposition system (Figure 2-6) (Ogawa a nd Nishihara, 2004). The IS cycle has been studied at the Genera l Atomic (GA) Corp. and Japan Atomic Energy Research Institute (JAERI) since being pr oposed by GA in the 1970's. At JAERI, High Temperature Gas-cooled React or (HTGR) was developed for hydrogen production and the HTGR technology capable providing high temper ature heat of around 950C was attained in 2004 (Onuki et al., 2005). Also, a continuous hydr ogen production closed loop test for 20 hours was conducted in 2003 (Ogawa and Nishihara, 2004) The IS cycle and the adiabatic UT-3 cycle (Sakurai et al., 1996) were sele cted as the two promising cy cles by GA for hydrogen generation using nuclear power (Brown et al., 2003). All reactions of the IS cycle were demonstrated experimentally as well as investigated theoretically and the ove rall process efficiency was estima ted to be high, about 47% (Funk, 2001). However, there are still some challe nges to be considered and solved for commercialization. The issues are: 1) difficulty in HI decomposition due to excess steam for

PAGE 32

32 enhancing the Bunsen reaction, 2) material se lection for the corrosive chemicals, 3) high temperature of more than 900C, 4) relatively expensive material, I2, which accounts for $45M capital cost estimated for 600MW hydrogen produc tion plant and 5) uncer tain economics (Onuki et al., 2005; T-Raissi, 20 05; Vitart et al., 2008). 2.3.3 ZnO/Zn Cycle One of the most well-known two-step meta l oxide reduction-oxidation (redox) reactions for hydrogen production is solar thermal based Zn O/Zn cycle (Steinfeld, 2002). The ZnO process powered by concentrated solar thermal energy is illu strated in Figure 2-7. As shown in the figure, the process consists of two reactions, a high en dothermic thermal decomposition reaction of ZnO and an exothermic hydrolysis reaction of zinc Since the decomposition of ZnO requires very high temperature, a concentrating solar chemical reactor was specially de signed by Haueter et al. (1999) and the experimental tests and de velopments on the reactor are ongoing. 2.4 Fresh Water Demands for Hydrogen Production Shortage of fresh water is one of the greatest problem s facing the world today. If hydrogen economy is realized in the future, it will be on the basis of hydrogen production from water. Only fresh water can be used as the feedstock to produce hydrogen although saline water can be used as the cooling water (Webber, 2007) Thus, desalination of water may be necessary, which is very energy intensive. Water consum ption was estimated for hydrogen production by electrolysis and SMR an d compared to that for refining pe troleum by Webber (Webber, 2007). It was estimated that water electrolysis consumes about 2.38 gallons of fresh water as a feedstock to produce 1kg of hydrogen while SM R requires 1.19 gallons of fres h water as a feedstock and an additional 3.5 gallons of fresh water for excess steam (Spath and Mann, 2001). Webber (2007) estimated that direct water consump tion for hydrogen production via electrolysis is comparable to that for refining an equivalent amount of gasoline. Seawater electrolysis was

PAGE 33

33 investigated for the reduction of the fresh wa ter consumption (Bockris, 1975; Williams, 1975; Bennett, 1980), however, corrosion and contamina tion on the electrodes due to the undesirable gas product, chlorine, and impurities and the release of the chlorine gas must be overcome. Like water electrolysis, the thermochemical hydrogen production process also uses water as the raw material to produ ce hydrogen. However, high temperat ure steam is needed for the thermochemical process instead of liquid water. Thus, fresh water may not be necessary for the thermochemical process since the saline water can be heated to supply pure steam to the process from an additional distiller.

PAGE 34

34 Figure 2-1. Natural gas steam re forming. Adapted from National Research Council and National Academy of Engineering, 2004. The Hydrogen Economy: O pportunities, Costs, Barriers, and R&D Needs The National Academies Pre ss. (Figure G-1, page 199). Figure 2-2. Gasification-based energy conversion system. Adapted from Stiegel, G. J., Ramezan, M., 2006. Hydrogen from coal gasification: an economical pathwa y to a sustainable energy future. Internat ional Journal of Coal Geology 65, 173-190 (Figure 2, page 177). Feed Sulfur removal Steam reforming Gas shift reaction Pressure swing adsorption Steam Fuel Fuel H2 Product

PAGE 35

35 Figure 2-3. Energy demands for water electro lysis. Adapted from Ohta, T., ed., 1979. SolarHydrogen Energy Systems Oxford and New York, Pergamon Press. Figure 2-4. Simplified model of thermo chemical cycle for hydrogen production Temperature Ener gy G (Electrical Energy Demand) H (Total Energy Demand) T S (Heat Demand) Chemical Reactions H2O H2 1/2 O2 Heat

PAGE 36

36 Figure 2-5. Photo-biol ogical hydrogen production. Adapted from Riis, T., Hagen, E. F., Vie, P. J. S., Ulleberg, O., 2006. Hydrogen production R& D: priorities and gaps. International Energy Agency (IEA). (Figure 6, page 13). Figure 2-6. The IS thermochemical cycle. Adapte d from Ogawa, M., Nish ihara, T., 2004. Present status of energy in Japan and HTTR project. Nuclear Engineering and Design 233, 510 (Figure 6, page 9).

PAGE 37

37 Figure 2-7. The ZnO/Zn cycle for hydrogen production. Adapted from Weidenkaff, A., Reller, A. W., Wokaun, A., Steinfeld, A., 2000. Th ermogravimetric analysis of the ZnO/Zn water splitting cycle. Thermochimica Acta 359, 6975 (Figure 1, page 70).

PAGE 38

38 CHAPTER 3 THE UT-3 THERMOCHEMICAL CY CLE FOR HYDROGEN PRODUCTION UT-3 cycle is one of the very few thermoch emical cycles that have been studied, both theoretically and experimenta lly, since being proposed in 19 78 (Kameyama and Yoshida, 1978). This cycle requires a relatively lower maximum te mperature, 730C, than the other cycles. Thus, concentrated solar thermal energy as well as nuclear power can drive the process. The cycle is comprised of four gas-solid reactions, two of which involve calcium compounds and the remaining two, iron compounds as shown below. ) ( 2 1 ) ( ) ( ) (2 2 600 2g O s CaBr g Br s CaOC (3-1) ) ( 2 ) ( ) ( ) (730 2 2g HBr s CaO g O H s CaBrC (3-2) ) ( ) ( 4 ) ( 3 ) ( 8 ) (2 2 2 210 4 3g Br g O H s FeBr g HBr s O FeC (3-3) ) ( ) ( 6 ) ( ) ( 4 ) ( 32 4 3 560 2 2g H g HBr s O Fe g O H s FeBrC (3-4) Each of four reactions occurs in a separate reactor that contains a solid reactant. Solid products produced remain in the reactor for the next reac tion which simplifies the separation steps. Only the gaseous reactants and produc ts move through the reactors. Reactions, (3-1) and (3-2), involving calcium compounds, have been extensively investigat ed since the hydrolysis of calcium bromide, reaction (3-2), is the kinetica lly slowest reaction and the rate-limiting step (TRaissi, 2005) as well as the highest-temperature reaction. Moreover, repetitive substantial volume changes of the solid reactants as they cycle between oxide and bromide forms adversely affect the physical stability and chemi cal reactivity of the solid reactants. A bench scale plant, named MASCOT (Mode l Apparatus for the Study of Cyclic Operation in Tokyo), was constructed and ope rated with continuous hydrogen production during several test runs (Nakayama et al., 1984). Adiabati c UT-3 cycle shown in Figure 3-1 (Sakurai et

PAGE 39

39 al., 1996) was selected together with the IS cycle as one of the two final cycles by General Atomics (GA) Corp. for hydrogen generation usi ng nuclear power (Brown et al., 2003). There are still several technical issues to be resolved even though extensive research on the process has already been done. These issues were outlined in chapter 1. 3.1 Historical Survey of UT-3 Thermochemical Cycle 3.1.1 Studies of UT-3 processes and System Kameyama and Yoshida (1978) introduced th e UT-3 cycle in 1978 based on the Gibbs free energy change as a function of the reaction temper ature. Experimental ki netic studies using the apparatus in Figure 3-2 showed very low convers ions in the hydrolysis reactions of calcium bromide and iron bromide, (3-2) and (3-4) while bromination reactions of calcium oxide and iron oxide, (3-1) and (3-3) were co mpleted. Separation techniqu es of oxygen and hydrogen using condensation were mentioned and thermal efficien cy of the cycle was estimated to be 36.8% on the basis of the HHV of hydrogen. A first conceptual plant design of the UT-3 cycle coupled wi th a high temperature nuclear power plant was suggested by Kameyama and Yoshida (1981). Figure 3-3 describes the conceptual plant design which cons ists of three parts: a calcium tower, an iron tower and a heat exchanger tower. A honeycomb shaped solid reac tant was made with an inert material and packed in a tubular reactor. This approach was chosen to make it easy to select the reactor material since the outside part of the honeycomb would remain in as oxide while the inside part would cycle between the brom ide and the oxide forms. The solid reactant was mixed with inert bindi ng materials to increase its strength. Fe3O4 pellets were fabricated by mixing magnetite pow der with silicasol, gell ing by adding ammonium carbonate, molding, drying and sinter ing. A kinetic study of bromina tion of the iron oxide pellets was conducted, and a simulation was performed to evaluate the reactor performance using the

PAGE 40

40 experimental data. It was found through the simulation studies th at the honeycomb-shaped solid reactant was feasible for the process. In 1984, a bench-scale UT-3 plant name d MASCOT (Model Apparatus for Studying Cyclic Operation in Tokyo) was built to produ ce 3 L/hr of hydrogen continuously as shown in Figure 3-4. In several test r uns, 2L/hr of hydrogen was continu ously produced (Nakayama et al., 1984). Aochi et al. (1989) designed a commercial size plant (20,000 Nm3/h of H2) of the UT-3 cycle conceptually and estimated the thermal efficiency of the plant to be about 40%, provided the efficiency of power generation from recovered heat was more than 25%. A new flow scheme shown in Figure 3-5 was suggested by Kameyama et al. (1989). The system consists of four reactors in series with tw o separators. At first, a ll four solid reactants are placed in each reactor. Steam is introduced into the reactor (1) which contains calcium bromide. Hydrolysis of the calcium bromide with water produces calcium oxide in solid form, which remains in the reactor while the gaseous product, HBr, and the resi dual gases move to the reactor (4). From the reactor (4), hydrogen is separa ted from the other gases. Bromine and water produced in the reactor (3) flow into the reactor (2). Sakurai et al. (1992) reported contin uous operation for 11 cycles, generating H2 and O2 in the ratio of 2:1. As illustrated in Figure 3-6, the experimental set-up was composed of Ca and Fe packed reactors. New Fe pellets showed improve d performance by changing the inert materials and the mixing ratio of the r eactant to the inert compounds. An adiabatic UT-3 Cycle was suggested by Sakur ai et al. (1996a) and a flowsheet of the cycle using a nuclear reactor is shown in Figure 3-1. The syst em would use excess steam and nitrogen as the heat carriers between adiabatic reactors and a gas stream including the excess

PAGE 41

41 steam, nitrogen and reactant gases was cooled or heated through a heat exchange before entering each adiabatic reactor. Hydrogen and oxygen would be separated by a zirconium-silica membrane at high temperature. A computer simulation with ASPENPLUS was conducted and the first and second law effici encies were found to be 48.9% and 53.2%, respectively. The adiabatic cycle has the following advantages. 1) E nhanced efficiency, 2) Reduced total heat duty of the heat exchanger, 3) Reduced power consum ption and 4) mitigated criteria of material selection for the reactor. Furthermore, the adiabatic UT-3 cycle was coupled with a solar heat source (Sakurai et al., 1996b). A continuous operation concept was cons idered using thermal storage for night operation. The overall thermal efficiency was estimated to be 49.5% and the exergy efficiency was estimated to be 52.9% by ASPEN-PLUS. Teo et al. (2005) conducted a critical eval uation on the efficiency of UT-3 process considering solar energy and ot her high temperature heat as en ergy sources (Figure 3-7) and compared the efficiencies with that of water electrolysis. As shown in Figure 3-7, instead of the HHV ba sed efficiency (49.5%) previously reported, a LHV based efficiency of 42% was used fo r the calculations. They conducted detailed calculations considering equipment efficiencies (compressors, heat exchangers), separation membranes and associated pressure losses, inco mplete conversions for the reactions and the impracticality of isothermal ope ration in the reactors (Teo et al., 2005). They reported the upper efficiency of UT-3 cycle based on a high temper ature heat to be around 13 %, which is much lower than the previous estimates (Sakurai et al., 1996a and 1996b). 3.1.2 Development of Solid Reactant The development of superior solid reactant has been regarded as one of the most important keys for the practicality of the UT-3 cycle. Ev en in the first proposal of the UT-3 process

PAGE 42

42 (Kameyama and Yoshida, 1978) the importance of developing durable solid reactants was emphasized. The following attempts on the iron a nd calcium reactants have been tried to improve the performance of the solid reactants. 3.1.2.1 Development of Fe-pellets In the first study of the UT-3 cycle (Kameyama and Yoshid a, 1978), three types of iron pellets using glass beads, bentoni te and kaolin as supporting mate rials were prepared by mixing, molding, drying and sintering. E ach pellet was tested in order to find out a durable and reliable supporting method for the iron solid reactant. Among them, the iron pellets supported by bentonite showed the best cyclic reactivity. In 1981, Fe3O4 pellets with a different inert material were also fabricated (Kameyama and Yoshida, 1981). A mixture of Fe3O4 powder and the inert binding ma terial, silicasol (Cataloid-S3OH), was gelled by adding a saturated ammoni a carbonate solution. The mixture was then formed into a sphere of 1.0 cm diameter and dr ied at 80C. Next, the sphere was sintered at 900 C. For the bromination reaction of the prepar ed iron oxide pellets, gaseous HBr and steam were supplied into the reactor where the pellets we re placed in. Using the experimental results, a simulation study of the performance of the newl y introduced honeycomb-shaped solid reactant was carried out. Figure 3-8 shows the fabricati on process of the CaO and Fe2O3 pellets used by Yoshida et al. (1990). Using the pellets thus fabricated, they experimentally evaluate d the reaction rates of all the four reactions. Their experime ntal set-up is shown in Figure 3-9. In order to make the pellets porous, Amir et al. (1992) mixed the magnetite powder with graphite powder and cellulose al ong with the inert materials such as zirconia and silica. They measured the pore volumes of the pellets with various graphite co ntents using mercury porosimeter. Figure 3-10 shows th e relationship of pore volume and the graphite contents. Based

PAGE 43

43 on the plots, the total cumulativ e pore volume was found to increas e linearly with the graphite content. Figure 3-11 shows the effects of the graphite addition on th e bromination and hydrolysis conversion of the Fe2O3 pellets. It was observed that the conv ersion of pellets with 20% graphite was more than 45% in 2 hours. However, the conv ersion was decreased in th e cyclic operations. In order to develop iron reactant pellets strong enough to overcome the cyclic volume change between oxide and bromide form, Amir et al. (1993) prepared and evaluated the iron reactants. The results such as chemical compos ition, hardness, porosity, and reactivity of the pellets are shown in Table 3-1. Through the studies on the cyclic reactivity, it was found that degrad ation of reactivity occurred over a number of cycl es due to sublimation of FeBr2, aggregation of Fe3O4 and formation of inert compounds by reaction between th e reactant and the su pport materials. Nakajima et al. (2000) fabricated Fe-pelle ts by an alkoxide method. The Fe-pellets contained reactant particles, Fe2O3, dispersed between the inert binder materials, Fe2TiO3. The actual reactant, Fe3O4, was generated first by the following reactions. 2 2 2 573 473 3 23 9 6 18 3 Br O H FeBr HBr O FeK (3-5) 2 4 3 873 823 2 22 12 2 8 6 H HBr O Fe O H FeBrK (3-6) Figure 3-12 shows the improvement on the initi al conversion by adapting the alkoxide method for the preparation of Fe-pellets instead of the old powder mixi ng method. It was observed that the rates of bromination and hydrolysis reactions of the iron reactant we re accelerated at least twice as fast. Recently, Sakurai et al. (2006) pub lished more detailed experiment al studies on the Fe-pellets. Their pelletization procedure is shown in Figure 3-13. The figure shows the conversion trend of

PAGE 44

44 the bromination of iron pellets prepared by the alkoxide method and the conventional powder mixing method. It can be seen that the convers ion of the pellet by the alkoxide method was well maintained in 10 cycles while the pellets usi ng powder mixing method degraded significantly. 3.1.2.2 Development of Ca-pellets Compared to the preparation step of Fe-pelle ts, the alkoxide chemistry for Ca-pellets was introduced comparatively early. In 1990, Aihara et al. (1990) suggested the pe lletization process using the alkoxide method (Fi gure 3-15). Through the new met hod, they achieved significant improvement in both bromination and hydrol ysis reaction rates of Ca-pellets. Figure 3-16 shows a comparison between the new pellets by the alkoxide chemistry and the old type pellets by a conven tional powder mixing method. It was observed from this figure that the pellets by the new method had a much higher reaction rate than the ones by the conventional preparation step. The time to maxi mum conversion was reduced significantly, and degradation was not obser ved in 6 cycles. In order to further improve the hydrolysis ra te, Sakurai et al. (1995) synthesized the CaO and CaTiO3 separately and then mixed the two com pounds to form the final product (Figure 317). Through the new process, the CaO agglomerat es were smaller than those produced from the previous method. Lauric acid was used as a pore foaming agent in this preparation to make CaO agglomerates finer and increase the surface area. From Figure 3-18(A), the volum e of micropores (pores with diameter less than 0.5 micrometers) was almost the same but the volume of macropores (pores with diameter greater than 0.5 micrometers) was increased around 4 times by applying the new method. So, one can conclude from the figure that the macropores were significantly increased by the lauric acid addition and the separated process chain and as a result, the rate of hydrolysis was approximately three times faster (Figure 3-18(B)).

PAGE 45

45 3.2 Strengths and Weaknesses of UT-3 Cycle 3.2.1 Strengths It is attractive because the UT-3 cycle is composed of only gas-solid reactions, which would simplify the separation of ga seous products from solid products. The process requires relatively low temperatur es for the process which can be achieved through solar concentrator s or nuclear reactors. The chemistry and kinetics of each reaction in the cycle have been studied extensively and well documented. The UT-3 cycle is one of a few processes wh ich were operated at a bench-scale plant. Involved elements such as calcium, iron, a nd bromine are inexpensive and abundant and no precious metals are needed in this cycle. (Average price of bromine was $2/kg, while average price of iodine for IS cycle was a bout $20/kg in 2007 (Bro mine Statistics and Information, 2008; Iodine Statis tics and Information, 2008) 3.2.2 Weaknesses There are weaknesses of the UT-3 cycle as well. For the process to be feasible, the solid reacta nts must be physically stable and chemically active while cycling between oxide and bromide forms. The excess steam needed for thermal energy and the hydrolysis reaction would decrease the performance of membrane separation. Hydrolysis reaction of calcium bromide is thermodynamically unfavorable reaction which means the conversion would be low at equilibri um state. However, the ultimate conversion efficiency would be higher si nce the gaseous product, HBr, is removed immediately from the reactor by the gaseous stream as soon as it is produced. This assumption would be verified in the chapter 4 and 5. Bromine and hydrogen bromide are circulated in the system at high temp erature, so special materials of construction are required. Teo et al. (2005) predicted a much lower effi ciency of less than 13% for the UT-3 cycle although this prediction was based on some assumptions which may not be correct.

PAGE 46

46 3.3 Modification of UT-3 Cycle Recently, in order to simplify or overcome some barriers in the UT-3 cycle, a few modifications of the UT-3 cycle were suggested and investigated. Doctor et al. (2002) proposed a modified 3-stage UT -3 cycle named Ca-Br cycle to simplify the process. The Ca-Br process uses a single-stage HBr dissociation by a commercial HBr electrolyzer or plasma dissociation instead of the two-steps HBr dissociation involving iron oxide and iron bromide. From the data in the Fi gure 3-19, HBr electrol ysis was expected to require about 48% lower electric ity than water electrolysis. In 2006, instead of the packed bed reactor, a fluidized bed reactor for the gas-solid reaction was proposed by Lemort et al. (2006) in order to improve the reaction kinetics and avoid the expensive preparation steps. The process effici ency was estimated to be 22.5%, provided the membrane technology for high temp erature separation was develope d. The conceptual design is shown in Figure 3-20. Simpson et al. (2007) suggest ed an innovative concept (Figure 3-21) based on molten calcium bromide with melted calcium oxide and conducted preliminary experiments to support the idea

PAGE 47

47 Figure 3-1. Flowsheet of the adiabatic UT-3 th ermochemical cycle. Adapted from Sakurai, M., Bilgen, E., Tsutsumi, A., Yoshida, K ., 1996a. Adiabatic UT-3 thermochemical process for hydrogen production. Internati onal Journal of Hydrogen Energy 21, 865870 (Figure 1, page 866). Figure 3-2. Experimental apparatus for kinetic studies. Adapted from Kameyama, H., Yoshida, K., 1978. Br-Ca-Fe water-decomposition cycl es for hydrogen production. Proc. of the 2nd World Hydrogen Energy Conf., Zurich, Switzerland, 829-850 (Figure 2, page 840).

PAGE 48

48 Figure 3-3. Conceptual plant de sign for UT-3 cycle. Adapted fr om Kameyama, H., Yoshida, K., 1981. Reactor design for the UT-3 ther mochemical hydrogen production process. International Journal of Hydrogen Ener gy 6, 567-575 (Figure 1, page 569). Figure 3-4. The MASCOT plant. Adapted from Nakayama, T., Yoshioka, H., Furutani, H., Kameyama, H., Yoshida, K., 1984. MASCOT a bench-scale plant for producing hydrogen by the UT-3 thermochemical decom position cycle. International Journal of Hydrogen Energy 9, 187-190 (Figure 1, page 188).

PAGE 49

49 Figure 3-5. New flow scheme for the UT-3 pr ocess. Adapted from Kameyama, H., Tomino,Y., Sato, T., Amir, R., Orihara, A., Aihara, M ., Yoshida, K., 1989. Process simulation of MASCOT plant using the UT-3 thermo chemical cycle for hydrogen production. International Journal of Hydrogen Ener gy 14, 323-330 (Figure 1, page 324).

PAGE 50

50 Figure 3-6. Experimental set-up for MASCOT plant. Adapted from Sakurai, M., Aihara, M., Miyake, N., Tsutsumi, A., Yoshida, K., 1992. Test of one-loop flow scheme for the UT-3 thermochemical hydrogen production process. International Journal of Hydrogen Energy 17, 587-592 (Figure 13, page 591). A B Figure 3-7. Pathway effi ciencies. A) Solar energy to hydroge n. B) High temperature heat to hydrogen. Adapted from Teo, E. D., Brandon, N. P., Vos, E., Kramer, G. J., 2005. A critical pathway energy effi ciency analysis of the thermochemical UT-3 cycle. International Journal of Hydrogen Ener gy 30, 559-564 (Figure 2, page 561).

PAGE 51

51 Figure 3-8. Pelletization process for the solid reactant pellets. Aadapted from Yoshida, K., Kameyama, H., Aochi, T., Nobue, M., Aiha ra, M., Amir, R., Kondo, H., Sato, T., Tadokoro, Y., Yamaguchi, T., Sakai, N ., 1990. A simulation study of the UT-3 thermochemical hydrogen production proce ss. International Journal of Hydrogen Energy 15, 171-178 (Figure 3, page 172).

PAGE 52

52 Figure 3-9. Experimental apparatu s for the kinetic tests of pellets. Adapted from Yoshida, K., Kameyama, H., Aochi, T., Nobue, M., Aiha ra, M., Amir, R., Kondo, H., Sato, T., Tadokoro, Y., Yamaguchi, T., Sakai, N ., 1990. A simulation study of the UT-3 thermochemical hydrogen production proce ss. International Journal of Hydrogen Energy 15, 171-178 (Figure 2, page 172).

PAGE 53

53 Figure 3-10. Experimental appara tus for the kinetic tests of pellets. Aadapted from Amir, R., Sato, T., Yoko Yamamoto, K., Kabe, T., Kameyama, H., 1992. Design of solid reactants and reaction ki netics concerning the iron compounds in the UT-3 thermochemical cycle. International J ournal of Hydrogen Energy 17, 783-788 (Figure 2, page 785).

PAGE 54

54 A B Figure 3-11. Bromination and hydr olysis conversion of the Fe2O3 pellets. A) Bromination. B) Hydrolysis of the Fe2O3 pellets by adding graphite. Adap ted from Amir, R., Sato, T., Yoko Yamamoto, K., Kabe, T., Kameyama, H., 1992. Design of solid reactants and reaction kinetics concerning th e iron compounds in the UT-3 thermochemical cycle. International Journal of Hydrogen Energy 17, 783-788 (Figure 3 and 4, page 785). With Graphite 20w% Without Graphite Without Graphite With Graphite 20w%

PAGE 55

55 Figure 3-12. Initial conversion pr ofile of Fe-pellets. Adapted from Nakajima, R., Kikuchi, R., Tsutsumi, A., 2000. Improvement of th e Fe-pellet reactiv ity in the UT-3 thermochemical hydrogen production cy cle. Hydrogen energy progress XIII: Proceedings of the 13th World Hydrogen Energy Conf erence, Beijing, China.303-307 (Figure 8, page 307).

PAGE 56

56 Figure 3-13. Flowsheet for the preparation of Fe-p ellets. Adapted from Sakurai, M., Ogiwara, J., Kameyama, H., 2006. Reactivity improveme nt of Fe-compounds for the UT-3 thermochemical hydrogen production process. Journal of Chemical Engineering of Japan 39, 553-558 (Figure 1, page 554).

PAGE 57

57 Figure 3-14. Conversions of the bromination in cyclic operation. Adapted from Sakurai, M., Ogiwara, J., Kameyama, H., 2006. Reactivity improvement of Fe-compounds for the UT-3 thermochemical hydrogen production proc ess. Journal of Ch emical Engineering of Japan 39, 553-558 (Figure 6, page 556). Figure 3-15. Preparation steps of Ca-pellets. Adapted from Aihara M., Umida, H., Tsutsumi, A., Yoshida, K., 1990. Kinetic study of UT -3 thermochemical hydrogen production process. International Jour nal of Hydrogen Energy 15, 7-11 (Figure 8, page 9).

PAGE 58

58 A B Figure 3-16. Conversion profiles of Ca-pellets. A) Bromination. B) Hydrolysis of Ca-pellets. Adapted from Aihara, M., Umida, H., Tsut sumi, A., Yoshida, K., 1990. Kinetic study of UT-3 thermochemical hydrogen producti on process. Interna tional Journal of Hydrogen Energy 15, 7-11 (Figure 10 and 11, page 10). Alkoxide method Conventional method Alkoxide method Conventional method

PAGE 59

59 Figure 3-17. New preparation me thod for Ca-pellets. Adapted fr om Sakurai, M., Tsutsumi, A., Yoshida, K., 1995. Improvement of Ca-pel let reactivity in UT-3 thermo-chemical hydrogen production cycle. In ternational Journal of Hydrogen Energy 20, 297-301 (Figure 6, page 299).

PAGE 60

60 A B Figure 3-18. Comparison of the pellets by the co nventional and new modified method. A) Pore volume. B) Hydrolysis rate Adapted from Sakurai, M., Tsutsumi, A., Yoshida, K., 1995. Improvement of Ca-p ellet reactivity in UT-3 thermo-chemical hydrogen production cycle. International Journal of Hydrogen Energy 20, 297-301 (Figure 7 and 11, page 299 and 301). Figure 3-19. Change of Gibbs free energy vs. temperature for the Ca-Br cycle and water electrolysis. Adapted from Doctor, R. D ., Marshall, C. L., Wade, D. C., 2002. Hydrogen cycle employing calcium-bromine a nd electrolysis. ACS Division of Fuel Chemistry, Preprints 47, 755-756.

PAGE 61

61 Figure 3-20. Conceptual design of a fluidized bed for UT-3 cycle. Adapted from Lemort, F., Lafon, C., Dedryvre, R., Gonbeau, D., 2006. Physicochemical and thermodynamic investigation of the UT-3 hydrogen pr oduction cycle: A ne w technological assessment. International Journal of H ydrogen Energy 31, 906-918 (Figure 8, page 914).

PAGE 62

62 Figure 3-21. Conceptual molten salt-based reacto r for hydrolysis/brominati on in the CaBr cycle. Adapted from Simpson, M. F., Utgikar, V., Sachdev, P., McGrady, C., 2007. A novel method for producing hydrogen based on the Ca Br cycle. Interna tional Journal of Hydrogen Energy 32, 505-509 (Figure 1, page 506).

PAGE 63

63 Table 3-1. Properties of iron solid reactant pellets. Adapted from Amir, R., Shiizaki, Yamamoto K., Kabe, T., Kameyama, H., 1993. Design development of iron solid reactants in the UT -3 water decomposition cycle based on ceramic support materials. Internationa l Journal of Hydrogen Energy 18, 283-286 (Table 2, page 285).

PAGE 64

64 CHAPTER 4 THEORETICAL FEASIBILITY INVESTIGATIONS OF UT-3 CYCLE Experimental feasibility studies of UT-3 cycle have been conducted and documented well, but rarely using a thermodynamic analysis appr oach except for one paper recently published by Lemort et al. (2006). In this chapter, the thermo dynamic conditions for each reaction in the cycle were comprehensively examined and discussed. This thermodynamic feasibility analysis will contribute to the determination of the optimal operating conditions fo r high conversion and reaction rate to improve the process efficiency. 4.1 Thermodynamic Analysis of UT-3 Cycle As introduced in the previous chapter, the UT-3 cycle consists of the following four heterogeneous reactions. (R1) ) ( 2 1 ) ( ) ( ) (2 2 600 2g O s CaBr g Br s CaOC (4-1) (R2) ) ( 2 ) ( ) ( ) (730 2 2g HBr s CaO g O H s CaBrC (4-2) (R3) ) ( ) ( 4 ) ( 3 ) ( 8 ) (2 2 2 210 4 3g Br g O H s FeBr g HBr s O FeC (4-3) (R4) ) ( ) ( 6 ) ( ) ( 4 ) ( 32 4 3 560 2 2g H g HBr s O Fe g O H s FeBrC (4-4) A reaction is spontaneous for negative values of G, while it is non-spontaneous with positive values. It is useful to find favorable co nditions, though the spontaneity is not related to the reaction rate (Meites, 1981). In order to check whether th e reactions in th e cycle will occur spontaneously, the changes of Gibbs free energy as a function of reaction temperature were calculated and are plotte d in Figure 4-1. The thermodynamic properties in the JANAF Tables (Chase et al., 1995) were us ed for the computation. Based on the Figure 4-1, the bromination r eactions of calcium oxi de and iron oxide, R1 and R3, are spontaneous, while the hydrolysis re actions of calcium bromide and iron bromide,

PAGE 65

65 R2 and R4, are non-spontaneous. The hydrolysis r eactions are advantageous at high temperature but it is limited by the melting points of the br omide forms. Thus, various parameters were considered to find the optimal operating conditi ons for the reactions. Detailed calculations conducted via an equilibrium module which uses the Gibbs energy minimization algorithm in FactSage thermochemical software and databa ses (FactSage, Web versio n) and analysis are described separately for each reaction below. In these results, minor products were neglected. 4.1.1 Reaction 1: Bromination Reaction of Calcium Oxide (R1) ) ( 2 1 ) ( ) ( ) (2 2 600 2g O s CaBr g Br s CaOC (4-1) Based on the changes of free energy as a functio n of temperature (Fig ure 4-1), bromination reaction of calcium oxide is very favorable. Th is was confirmed by the experimental results of Aihara et al. (1992) and Sakurai et al. (1995). However, the ea rlier researches al so found that the bromination process was terminated with roughl y 70% conversion, mainly due to pore plugging in less than 5 minutes. Therefore, a high porous reactant with high su rface area free from pore plugging must be developed in order to increase the bromination conversion. Aihara et al. (1992) investig ated the effects of steam c oncentration on the bromination reaction of calcium oxide pellets experimentally They reported that the bromination conversion was well maintained in the pres ence of approximately thirty times excess steam per hydrogen. However, according to Sakurai et al. (1996b), at least one hundr ed times of steam per hydrogen yield should be used to operate the adiabatic process without any ot her external heat input to the reactors. Therefore, in this study, the effect of steam concentration on the br omination reaction of calcium oxide was evaluated up to five hundred times the hydrogen using FactSage thermochemical software and databases (FactS age, Web version). The simulation results are

PAGE 66

66 illustrated in Figure 4-2. As shown in Figure 4-2, only small adverse effect of excess steam up to one hundred times excess supply on the bromination reaction was observed. With one hundred times excess steam, the e ffects of temperature and pressure on the reaction progress were also ca lculated and are displayed in Figure 4-3 and Figure 4-4, respectively. Based on Figure 4-3, high conversion was achieve d at temperatures in the range 600-900K. The influence of temperature on the reaction was slight in the range. Ho wever, Aihara et al. (1992) observed the degradation at 823K in their experimental study due to the formation of hydrates of calcium bromide while the reactivity was ma intained at 873K w ithout the formation of hydrates. The effect of operation pressure on the reaction progress is illustrated in Figure 4-4. High pressure did not enhance conversi on much, as expected according to the Le Chateliers principle. Moreover, formation of calcium hydroxide (Ca(OH)2) was observed at high pressures over 5 atm due to the reaction of calcium oxide and steam Under low pressure, the conversion decreased dramatically. The theoretical analysis of th e bromination reaction of calcium oxide is summarized here. In theory, the bromina tion reaction of calcium oxide is spontaneous and can be completed thermodynamically if there is no barrier. Very subtle adverse effect of excess steam up to one hundred times on the bromination was observed in the simulation study. Influence of temperature on the conversion of the reaction was insi gnificant in the range 600-900K. The conversion was decreased noticeably under vacuum pressure while it was not much enhanced with pressure over 2 atm. Based on our theoretical analys is and experimental data from the literature optimum conditions for the bromination of calcium oxide we re determined and are presented in Table 4-1.

PAGE 67

67 4.1.2 Reaction 2: Hydrolysis Reaction of Calcium Bromide (R2) ) ( 2 ) ( ) ( ) (730 2 2g HBr s CaO g O H s CaBrC (4-2) Figure 4-1 tells us that the hydrolysis reaction of calcium bromide (R2) is thermodynamically unfavorable. Actually, this reac tion is the slowest and the rate limiting step. Therefore, speeding up this reaction is the key to improve the performance of the cycle. Several experimental (Aihara et al ., 1990; Sakurai et al., 1995; Lee et al., 2006 and 2007) and theoretically studies (Lemort et al., 2006) to acce lerate the reaction rate ha ve been reported in the literature. Conversion seems to increase with elevating temperature, sin ce the reaction is endothermic. Besides, the conversion with ex cess steam and at lower pressure s is expected to be higher according to the Le Chateliers principle. The effect of excess steam on hydrolysis react ion of calcium bromide at 1000K and 1 atm is shown in Figure 4-5. The hydr olysis reaction was effected significantly by the excess steam and the conversion was increased with the addition of excess steam as expected according to the Le Chateliers principl e. From figure 4-5, onl y 9.5 percent conversion was expected with 100 times excess steam at atmospheric pressure. This conversion is very close to the value (9.45%) reported by Lemort et al. (2006). The effect of temperature on the hydrolysis re action of calcium bromide with one hundred times excess steam was estimated and is di splayed in Figure 4-6. The conversion was significantly increased with the elevation of te mperature; however, the reaction temperature is limited by the melting point of calcium bromid e at about 1000K as shown in the plot. The influence of pressure was also calculated and the result is di splayed in Figure 4-7.

PAGE 68

68 Based on the computation, the conversion was highly dependant on the operation pressure as suggested before (Lemort et al., 2006). If the pressure reduces to 0.01 atm, about 96 percent conversion can be achieved even though the low pressure condition may be impractical as well as inefficient for mass production of hydrogen. Considering the operation principle of packed-bed reactor involving gas-solid heterogeneous reaction, however, the ultimate conversion efficiency may be higher even at atmospheric pressure since the gaseous products are removed immediately from the reactor by the gaseous stream as soon as they are produced This assumption was verified by calculating the effect of consecutive hydrogen bromide (HBr) removal from the equilibrium conditions. As shown in Figure 4-8, it was observed that the conversion could be completed theoretically by removing the product gas, HBr, from each equilibrium state. The hypothesis will be verified by means of the experimental tests in chapter 5. Therefore, considering the advantage of the gas-solid reaction in the packed bed system atmospheric pressure rather than 0.01 atm is more practical for this reaction. The theoretical analysis of the hydrolysis r eaction of calcium bromid e is summarized here. Thermodynamically, the hydrolysis reaction of calcium bromide is unfavorable. The hydrolysis of calcium bromide is enha nced with the addition of excess steam significantly. High temperature and low pressure favors higher conversion though the temperature is constrained due to the melti ng point of calcium bromide. 96 percent conversion is expect ed with excess steam, 100H2O/H2, at 1000K and 0.01 atm even though the low pressure condition is not feasible for the pr actical application. The conversion is expected to be complete d by removing the product gas, HBr, from each equilibrium state even at atmospheric pressure. Based on the gas-solid reaction advantage in packed bed system and economic performance, 1 atm rather than 0.01 atm is more practical pressure condition for the reaction.

PAGE 69

69 Through theoretical analysis, optimum conditio ns for the hydrolysis of calcium bromide are presented in Table 4-2. 4.1.3 Reaction 3: Bromination Reaction of Iron Oxide (R3) ) ( ) ( 4 ) ( 3 ) ( 8 ) (2 2 2 210 4 3g Br g O H s FeBr g HBr s O FeC (4-3) The bromination reaction of ir on oxide seems to be favorab le only at low temperature range (Figure 4-1). The effect of temperature on the bromina tion of iron oxide was computed. One hundred times excess steam per hydrogen was included in the reactant gases in the calculation. The results are plotted in Figure 4-9. The figur e shows that the conver sion increases as the temperature decreases and an undesirable substance, Fe2O3, is formed between 400K and 800K. So, 400K was selected as the optimal temperat ure as suggested by Lemort et al. (2006). The effect of pressure on the bromination of iron oxide is illust rated in Figure 4-10. Judging from this plot, convers ion can be reached almost completely by promoting the bromination of Fe2O3 by HBr at pressures slightly above 3 atm. The theoretical analysis of the brominati on reaction of iron oxide is summarized here. In theory, the brominatio n of iron oxide is favorab le at low temperature Almost complete conversion is r eached at a pressure of 3 atm. All things considered, optimum conditions for the bromination of iron oxide are presented in Table 4-3. 4.1.4 Reaction 4: Hydrolysis Reaction of Iron Bromide (R4) ) ( ) ( 6 ) ( ) ( 4 ) ( 32 4 3 560 2 2g H g HBr s O Fe g O H s FeBrC (4-4) R4, hydrolysis of iron bromide, is thermodynami cally unfavorable similar to the hydrolysis of calcium bromide. High temperature, excess stea m and low pressure seem to be effective in

PAGE 70

70 improving the conversion according to the change of Gibbs free energy versus temperature and the Le Chateliers principle. The influence of temperature on the hydrolys is of iron bromide was calculated and is plotted in Figure 4-11. As expected, the conversio n is strongly dependant on the temperature in the temperature range below 800K. Above 900K however, the conversion was decreased gradually as the temperature increased due to th e sublimation of iron brom ide as reported in the literature (Amir et al., 1993). Figure 4-12 shows the effect of excess steam on the reaction and it is observed that supplying excess steam is very effec tive in increasing the conversion. The theoretical analysis of the hydrolysis r eaction of iron bromide is summarized here. Thermodynamically, the hydrolysis reaction of iron bromide is unfavorable. The highest conversion was ach ieved at 800K without the su blimation of iron bromide The excess steam enhanced the reaction significantly. Through theoretical analysis, the optimum cond itions for the hydrolysis of iron bromide are presented in Table 4-4. 4.2 Optimum Conditions for UT-3 Cycle Considering excess steam and other thermodyna mic parameters, the determined optimum conditions and expected conversio ns of each reaction in UT-3 cy cle are summarized in Table 4-5 subject to no physical barrier of the reactions such as reduction of active surface area and diffusion resistance due to sintering of reactant granules and encapsulation by solid product.

PAGE 71

71 -200 -150 -100 -50 0 50 100 150 200 250 300 400500600700800900100011001200 Temperature [K]Gibbs Free Energy [kJ/mol] R1 R2 R3 R4 Figure 4-1. Changes of Gibbs fr ee energy of reactions as a f unction of reaction temperature 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0151050100300500 Excess steam/Hydrogen [times]Equilibrium Amount [mol] 0 10 20 30 40 50 60 70 80 90 100Conversion [%] CaO Br2 CaBr2 O2 Conversion Figure 4-2. Effect of excess steam on bromina tion reaction of calcium oxide at 873K and 1 atm

PAGE 72

72 0.00 0.20 0.40 0.60 0.80 1.00 1.20 6007008008739009451000 Temperature [K]Equilibrium Amount [mol] 0 10 20 30 40 50 60 70 80 90 100Conversion [%] CaO Br2 CaBr2 O2 Ca(OH)2 Conversion Figure 4-3. Effect of operati on temperature on bromination reac tion of calcium oxide at 1 atm with 100H2O/H2 0.00 0.20 0.40 0.60 0.80 1.00 1.20 0.0050.010.050.10.512510 Operation Pressure [atm]Equilibrium Amount [mol] 0 10 20 30 40 50 60 70 80 90 100Conversion [%] CaO Br2 CaBr2 O2 Ca(OH)2 Conversion Ca(OH)2 Figure 4-4. Effect of op eration pressure on bromination reacti on of calcium oxide at 873K with 100H2O/H2

PAGE 73

73 0.0 0.2 0.4 0.6 0.8 1.0 1.2 0151050100500 Excess steam [H2O/H2]Equilibrium Amount [mol] 0 10 20 30 40 50Conversion [%] HBr CaO CaBr2 conversion Figure 4-5. Effects of excess steam on hydrol ysis of calcium bromide at 1000K and 1 atm 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 60070080090010001050110011501200 Temperature [K]Equilibrium Amount [mol] 0 10 20 30 40 50 60 70 80 90 100Conversion [%] CaO HBr CaBr2(S) CaBr2(L) Conversion Figure 4-6. Effect of temper ature on hydrolysis reaction of calcium bromide with 100H2O/H2 at 1 atm

PAGE 74

74 0.0 0.5 1.0 1.5 2.0 2.5 0.0010.0050.010.050.10.51 Pressure [atm]Equilibrium Amount [mol] 0 10 20 30 40 50 60 70 80 90 100Conversion [%] CaO HBr CaBr2(S) Conversion Figure 4-7. Effect of pressure on hydroly sis reaction of calcium bromide with 100H2O/H2 at 1000K 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1234567891011 Stage No.Equilibrium Amount [mol] 0 10 20 30 40 50 60 70 80 90 100Conversion [%] CaO HBr CaBr2(S) Conversion Figure 4-8. Effect of HBr remova l from equilibrium states with 100H2O/H2 at 1000K and 1 atm

PAGE 75

75 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0 9.0 400500600700800900100011001200 Temperature [K]Equilibrium Amount [mol] 0 10 20 30 40 50 60 70 80 90 100Conversion [%] Fe3O4 HBr FeBr2 Fe2O3 Br2 FeBr2(g) Conversion Figure 4-9. Effect of temp erature on bromination reac tion of iron oxide with 100H2O/H2 at 1 atm 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7. 0 0.0010.0050.0100.0500.1000.5001.0002.0003.0005.000 Pressure [atm]Equilibrium Amount [mol] 0 10 20 30 40 50 60 70 80 90 10 0 Conversion [%] Fe3O4 HBr FeBr2 Fe2O3 Br2 Conversion Figure 4-10. Effect of operation pressure on the bromination reacti on of iron oxide at 400K with 100H2O/H2

PAGE 76

76 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 60070075080090010001050110011501200 Temperature [K]Equilibrium Amount [mol] 0 10 20 30 40 50 60 70 80 90 100Conversion [%] Fe3O4 HBr FeBr2(s) FeBr2(g) H2 Conversion Figure 4-11. Effect of temper ature on hydrolysis reaction of iron bromide at 1 atm with 100H2O/H2 0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 0151050100500 Excess steam [H2O/H2]Equilibrium Amount [mol] 0 10 20 30 40 50 60 70 80 90 100Conversion [%] Fe3O4 HBr FeBr2(s) FeBr2(g) H2 Conversion Figure 4-12. Effect of excess steam on hydrolys is reaction of iron bromide at 800K and 1 atm

PAGE 77

77 Table 4-1. Optimum conditions for the bromination of calcium oxide Reaction Conditions Optimum value Excess Steam Quantity 100H2O/H2 Temperature 873K Pressure 1 atm Table 4-2. Optimum conditions for the hydrolysis of calcium bromide Reaction Conditions Optimum value Excess Steam Quantity 100H2O/H2 Temperature 1000K Pressure 1 atm Table 4-3. Optimum conditions for the bromination of iron oxide Reaction Conditions Optimum value Excess Steam Quantity 100H2O/H2 Temperature 400K Pressure 3 atm Table 4-4. Optimum conditions for the hydrolysis of iron bromide Reaction Conditions Optimum value Excess Steam Quantity 100H2O/H2 Temperature 800K Pressure 1 atm Table 4-5. Optimal conditions and expected conv ersions of the reactions in UT-3 cycle with 100H2O/H2 No. Reaction Temperature [K] Pressure [atm] Expected Conversion R1 CaO(s) + Br2(g) CaBr2(s) + 0.5O2(g) 873 1 94.8 % R2 CaBr2(s) + H2O(g) CaO(s) + 2HBr(g) 1000 1 Completed* R3 Fe3O4(s) + 8HBr(g) 3FeBr2(s) + Br2(g) + 4H2O(g) 400 3 99.0 % R4 3FeBr2(s) + 4H2O(g) Fe3O4(s) + 6HBr(g) + H2(g) 800 1 95.7 % The conversion was expected to be completed by rem oving the product gas according to the Le Chateliers principle.

PAGE 78

78 CHAPTER 5 PREPARATION AND EVALUATION OF CALCIUM OXIDE REACTANT FOR UT-3 CYCLE Calcium oxide is involved in two of the four gas-solid reactions in UT-3 cycle. It is converted to calcium bromide by bromination re action with bromine, and the produced calcium bromide is reconverted into the original calc ium oxide during hydrolysis reaction with water. The cyclic reactions must be continuously maintained at high conversion and rates for the UT-3 cycle to be practically viable. Aihara et al. (1990) pr oposed and made calcium oxide pellets that contain calcium oxide particles dispersed and imm obilized in an inert matrix of calcium titanate (CaTiO3) through a procedure based on alkoxide chemistry and solgel process. Then they evaluated the pellets in cyclic bromination-hydrolysis reactions and found that the performance was enhanced considerably. However the preparation procedures seem rather complicated and expensive and the practicality of the pellet-type reactan t is still in doubt in the long operation. Moreover, Lemort et al. (2006) suggested the possi bility of decomposition of the inert material, calcium titanate (CaTiO3) in the pellets by HBr. In this study, conventional calcium oxide pell ets and new developed calcium oxide fabric in which calcium oxide dispersed and immobiliz ed on a yttria fabric were fabricated and evaluated experimentally. 5.1 Calcium Oxide Pellets The procedure for conventional pellets was de veloped based on the alkoxide chemistry and sol-gel process of Aihara et al. (1990 and 1992) and Sakurai et al. ( 1995). The calcium oxide pellets contain calcium oxide reactant dispersed and immobilized in an inert material, calcium titanate (CaTiO3). In this study, pellets were fabricated and their characteri stics and performance were evaluated.

PAGE 79

79 5.1.1 Preparation of Calcium Oxide Pellets A precursor of CaO and CaTiO3 with a mole ratio of 1:2 was synthesized using alkoxide process followed by Sol-Gel technique describe d by Aihara et al. (1990) Then, predetermined amounts of corn starch, stearic ac id, and graphite powders were added into the solution as pore forming agents to increase the porosity in this study while graphite powder and lauric acid were used previously (Sakurai et al., 1995). The am ounts of various pore forming agents added are listed in Table 5-1. The slurry was dried via na tural evaporation at room temperature for 12h. The resultant powder was crushed, screened and compressed into a specially fabricated cylindrical shaped mold via a uni-axial hydraulic pr ess. The fabricated pelle ts were then sintered in a furnace in order to strengthen them and bur n out the pore forming agents. Heating schedule for the sintering was controlled to ensure mild de-binding of the pore forming agents since rapid burning of additives is a cause of breakage of powder compacts. TG analysis was performed in air by Perkin-Elmer TGA-7 TG Analyzer in or der to determine cont rolled sintering step. The particle sizes and shape of pore forming ag ents can be observed with SEM images in Figure 5-1. Based on the micrographs, corn starch and stearic acid are sphere-shaped particles which have approximately 10 m and 200 m in diameter, respectively, while graphite has arbitrary shape. Figure 5-2 shows a TGA curve of a non-sintered calcium oxide pe llet at a heating rate of 10C /min in air. In the TGA curve, it was observed that residual ethanol and moisture were removed in the first step (60-250 C) and additives were burnt out in the second (250-400C) and third steps (650-700C). Considering auto-ignition or thermal decomposition temperature of the additives, it is possible to tell that corn starch, stearic aci d and graphite were decomposed during the forepart of the second step, the late stage of second step a nd the third step, respectively.

PAGE 80

80 On the basis of the TGA result shown in Figur e 5-3, the heating prof ile for sintering was determined. The furnace temperature was slowly incr eased in the range of burning of additives to decrease the adverse effect of the thermal decomposition of the a dditives on the final strength of the pellets. The cylindrical shaped calcium oxide pellet s approximately 4mm in diameter and 10mm long were obtained after sinteri ng. A brief procedure of the Ca-p ellets preparation is shown in Figure 5-4. 5.1.2 Characterization The fabricated pellets were subsequently characterized using XRD experiments and Mercury porosimeter. The composition of the pellets was investig ated by XRD experiments to verify the presence of both calcium oxide, CaO, and calcium titanate, CaTiO3. XRD peaks shown in Figure 5-5 indicate the presence of calcium oxide and calcium titanate in the pellet. The pore size distributions of the pellets were measured by Quantachrome Autoscan 60 Mercury porosimeter from Quantachrome instru ments. The pore size distributions of each sample are shown in Figure 5-6. A definite incr ease of total pore volume was observed by adding pore forming agents. Addition of stearic acid increased the pores greater than 5 m while the addition of graphite and corn starch contributed to the fo rmation of pores less than 5 m. On the surface of the pellets with added stearic acid (samples 3 and 4), macroscopic pores were detected. On the other hand, it was found that the additi on of the pore forming agents decreased the strength of the pellets. From the pore size dist ributions shown in Figure 5-6, it was ascertained that the pore size distribution of pellets was strong ly influenced by the type of additives used.

PAGE 81

81 5.1.3 Kinetic Measurements The bromination and hydrolysis reactions of the selected Ca -pellets were conducted using laboratory experimental set-up s hown in Figure 5-7. The schematic diagram of the facility is shown in Figure 5-8. A CaO reactant-pellet was placed in a plat inum basket that was hung from a selfconstructed precision spring bala nce which was purged with nitr ogen. The reactor temperature was controlled by a K-type thermocouple installe d right under the basket. Bromine and steam were supplied as gases and nitrogen was used as a carrier gas. The weight change of the pellet during the reaction is proportiona l to the spring deflection, whic h was detected by a microscope. The conversion, a measure of the progress of a reaction, is define d as the moles of a species that have reacted over the moles of the sa me species initially present. Therefore, it is represented as follows (Aihara et al., 1990); ) ( ) (2CaO CaBr CaO o o CaOM M R W W W M X (5-1) where M is molecular weight, W is wei ght of the sample during reaction, W0 is weight of a fresh sintered sample, and RCaO is weight fraction of calcium oxide in the fresh sample. The weight change of the sample, W-W0, was determined using the spring constant and deflection of the spring measured by a microscope. Table 5-2 gives the characteristics measured by the porosimeter and the experimental data by the kinetic measurements. Sample 1 had the largest surface area since small pores account for the surface area. The hydrolysis rate was accelerated by increasing the pore volume greater than 5 m since these pores contribute to the diffu sion characteristics inside the pellet. However, our measured rate was still slower than the best result reported by Sakurai et al. (1995).

PAGE 82

82 Cyclic experimental tests were conducted using one of the samples with excellent characteristics in order to see the practical feasibility of the pellets. The temperatures for bromination and hydrolysis reaction were select ed as 600C and 700C, respectively, based on the thermodynamic analysis and previous studie s. For the bromination reaction, bromine and water were supplied while only steam was s upplied for hydrolysis reaction. The bromine concentration for bromination a nd water concentration for hydrolysis were 2.6 mol % and 90 mol %, respectively. The feed for the bromin ation reaction contained approximately 27 times excess steam per mole of hydroge n production. Nitrogen was used as a carrier gas and a purge gas of the spring balance. Figure 5-9 shows two cyclic conversion profiles of the br omination and hydrolysis reactions. It took less than ten minutes to re ach the maximum conversion in the bromination reaction. The hydrolysis reacti on was much slower as compared to bromination reaction. Degradation was observed during th e cyclic operations. 70% conversi on was attained in the first cycle, while about 50% conversi on was achieved in the second. It was also observed that the hydrolysis rate was retarded in the second cycle. In order to investigate the mechanism of degradation, the changes in the pore size distributions after brominati on and hydrolysis processes we re measured using a Mercury porosimeter. As shown in Figure 5-10 and Figure 5-11, the pore volume was reduced by almost half in the process of bromination possibly due to expansion of the solid reactant. The pores greater than 5 m were decreased by about 35% after br omination and were regenerated about 20% after hydrolysis. The pores less than 5 m were much more seriously affected by the reactions. The pores less than 5 m were decreased by about 55% after bromination and only 9 % were recovered after hydrolysis. From the ch anges in the pore volum e distribution, it was

PAGE 83

83 ascertained that a reduction in the pore volume lead s to the degradation of the chemical reactivity in the cyclic operation. 5.2 Calcium Oxide Fabrics The concerns about attrition and degradation of the pellets are still present and a simpler and inexpensive preparation step of the reactant is preferable, if possible. Further, the pellet-type reactant is inherently unfavorab le for the gas-solid reaction incl uding significant volume change because it increases mass transfer resistances. He nce, a new type of calcium oxide reactant was fabricated on a fibrous yttria fabric via a comparatively straightfo rward and inexpensive immobilization process. In the beginning, several ceramic materials such as alumina (Al2O3), silica (SiO2), zirconia (ZrO2) and yttria (Y2O3) were considered as a substrate. Bu t the preliminary e xperiments showed that the performance of all of th ese materials except yttr ia severely degraded at the second cycle possibly due to the formation of inert materials such as calcium aluminate, calcium silicate and calcium zirconate due to reactions between cal cium oxide and the subs trate materials. The experimental results using alumina, silica and zirconia are given in Appendix A. According to the changes of Gibbs free energy of reactions between calcium oxi de and the ceramic materials in Figure 5-12, the formation of the inert materials is very feasible. Hence, yttria was selected as the substrate material for the fabric. 5.2.1 Immobilization of Calcium Oxide on a Fibrous Yttria Fabric A new procedure for the immobilization of cal cium oxide on a fibrous yttria fabric was developed by modifying the proce ss presented earlier (Lee et al ., 2007) and is shown in Figure 513. The nano-sized precipitated calcium carbonate (PCC; average particle size = 70 nm) surface treated with stearic acid to enhance dispersi bility was selected as the starting material. The PCC was blended with ethyl alcohol by mechanical stirring for 10 minutes and ultrasonicated for

PAGE 84

84 another 10 minutes to reduce a gglomeration. The suspension was dropped onto a dried yttria fabric (YF-50 from Zircar Zirconia, Inc., USA) using a transfer pipette. Then the fabric was dried at 100C for 30 minutes a nd subsequently sintered at 90 0C for 20 hours in air. During sintering, the PCC on the yttria fabric was converted into ca lcium oxide. The size of the sintered fabric sample was roughly 30mm (L) x 10mm (W) x 1.27 mm (T). 5.2.2 Characterization The calcium oxide particles att ached well on the yttria fiber in the fresh sintered sample since no particles fell off from shaking or slight impact. The chemical compositions of the calcium oxide fabric samples were examined by XRD using Philips MRD X'Pert System The X-ray diffraction patterns of the fresh and brominated samples are shown in Figure 5-14. According to the XRD data, it was confirmed that all of the calcium carbonate was converted to calcium oxide during the sintering and most of the calcium oxide reacted with bromine to form calci um bromide through the bromination reaction. 5.2.3 Cyclic Reaction Experiments The cyclic reactions, bromination and hydrol ysis reactions, were conducted using the experimental set-up illustrated in Figure 5-7. The prepared sample was suspended from a selfconstructed precision spring bala nce in the reactor whose temp erature was controlled by a thermocouple installed right under the sample. The bromination and hydrolysis reaction were conducted at 600C and 700C, respectively similar to the earlier experime nts with pellets. The feed compositions were also same. Bromine and water were supplied for the bromination reaction, while only steam was supplied for the hydrolysis reaction. Th e bromine concentration was 2.6 mol %, which included a corresponding wate r of 27 times excess steam, for bromination. Water concentration for hydrolysis was 90 mol %. Nitrogen was also supplied as a carrier gas of the reactant gases and a purge gas of the spring balance.

PAGE 85

85 The cyclic conversion profile of bromination and hydrolysis reactions is shown in Figure 5-15. About 85 percent conversion of the brominati on reaction was obtained in 10 minutes, and it was maintained at the same level through the f ourth cycle. This conve rsion was approximately 5% higher than that obtained from the pellets in the earlier studies (Lee et al., 2006; Sakurai et al., 1995). The hydrolysis rate was about 20 percent highe r as compared with that of pellets in our studies (Lee et al., 2006) and comparab le to that of the pellets in the literature (Sakurai et al., 1995). The structure of calcium oxide fabric sample s was examined by SEM using Hitachi S-800. SEM images of the bare yttria fabric, fresh calci um oxide fabric sample, and brominated sample are displayed in Figure 5-16. The di ameter of the yttria fibers on the original fabric is in the range of approximately 8 to 10 m (Figure 5-16 (A)). As illustrated in the Figure 5-16(B), the surface of the yttria fiber in a fresh sample wa s tightly covered by interconnected calcium oxide particles with a few hundred nanometer diameter. The changes in the structure of the sample by the bromination reaction were very noticeable, as shown in Figure 5-16(C). After bromination, it was observed that the volume of the particles in creased from calcium oxide to calcium bromide and the brominated particles were merged or overlapped. The amount of calcium oxide in the fresh fa bric sample was about 11 wt %, which is almost half of the quantity in the calcium oxide pe llet. Lesser inert material is preferable as long as there is no adverse effect on the performance of the reactant in order to reduce the preparation cost and operational energy. For this reason, an attempt was made to increase the calcium oxide content through repetitive impregnation steps. Th e impregnation steps resulted in a much higher calcium oxide content in the samp le (22 wt %). Cyclic performan ce of the samples thus prepared was examined for the first cycle. The results for the samples with various calcium oxide contents

PAGE 86

86 (8.7, 10.7, 11.9, 22 wt %) are compared in Figur e 5-17. As shown in the figure, the hydrolysis rate was retarded by nearly a factor of two as the content increased by two times, while the maximum bromination conversion was not affected much. 5.2.4 Error Analysis The resolution of the balance to determine th e weight fraction of calcium oxide in the sample is 0.1 mg. The spring rate of the spring hung in the balance was measured to be 13.30 0.03 mm/g through repeated microscopic measur ements using standard masses. Hence, the uncertainty in measurement of the spring defl ection by the microscope was 0.03 mm which is approximately equivalent to 2.2 mg. The combin ed uncertainties of each measurement in cyclic reactions were graphically represen ted as the error bar in figures. 5.3 Conclusion and Summary Porous Ca-pellets for UT-3 thermochemical cycle were prepared and characterized. A set of Ca-pellet samples with differe nt amounts of pore forming agents was fabricated and compared with one another with respect to pore size distribution, convers ion and reaction rate. From the characterization and kinetic studies it was ascertained that the am ount and type of additives have an important effect on the pore volume, and in creasing the volume of pores greater than 5 m speeds up the hydrolysis rate of the Ca-pellets. Degradation of the bromination reaction was observed during the cyclic experiment. 70% conver sion in the first cycle dropped to 50% in the second. It was also observed th at the hydrolysis rate was reta rded in the second cycle. The degradation in the cycl ic operation was found to be caused by a reduction in the pore volume based on the changes in th e pore volume distribution. An inexpensive and straightforward calcium oxide immobilization process on a yttria fabric was developed. Based on the experimental results in th e cyclic reactions, the calcium oxide reactant on the yttria fabric had continuou s higher reactivity (~ 85 %) in the bromination

PAGE 87

87 reaction during four cycles, and th e rate of hydrolysis reaction was comparable to that of calcium oxide pellets.

PAGE 88

88 A B C Figure 5-1. The SEM images of pore forming agen ts. A) Corn starch. B) Graphite. C) Stearic acid. 0 10 20 30 40 50 60 70 80 90 1002 0 100 2 00 3 00 400 500 600 700 800 9 00Temperature[ ]Weight loss [%] Figure 5-2. The TGA curve of a nonsintered calcium oxide pellet at a heating rate of 10C /min in air

PAGE 89

89 0 200 400 600 800 1000 1200 05101520 Time[h]Temperature [ ] Figure 5-3. Sintering steps for Ca-pellets Figure 5-4. Flow diagram for th e preparation of Ca-pellets Metal pure Calcium Ca Ethanol, C 2 H 5 OH Calcium ethoxide ethanol, Ca(OC2H5)2 Add Titanium Isopropoxide, Ti(C3H7O)4 Mixture of CaOand CaTiO 3 p recursors Add20%Water/EthanolSolution Add corn starch, stearic acid, and graphite powder Dr y slurr y via eva p oration Die p ressin g Sinterin g Porous Ca-Pellets

PAGE 90

90 1 10 100 1000 10000 1015202530354045505560657075802 Intensity CaO CaO CaTiO3 CaTiO3 CaTiO3 CaTiO3 CaTiO3 CaTiO3 CaTiO3 CaTiO3 CaTiO3 CaTiO3 CaTiO3 CaTiO3 CaTiO3 CaTiO3 CaO CaOCaO CaO Figure 5-5. The XRD pattern of the fresh calcium oxide pellet 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 110100100010000100000 Pore Diameter [nm]Cumulative Specific Pore Volume [cc/g] Cumulative Specific Pore Volume [cc/g] Figure 5-6. Pore size di stributions of Ca-pellets (Sample 1-4) prepared with different pore forming agents S1 S2 S3 S4

PAGE 91

91 Figure 5-7. La boratory Set-up Figure 5-8. Schematic Diag ram for Laboratory Set-up

PAGE 92

92 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 0102030405060708090100110120130140Time[min]Conversion, X 1st Cycle 2nd Cycle Figure 5-9. Cyclic conversi on profiles of a calcium oxide pellet (Sample: S4) 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 110100100010000100000 Pore Diameter [nm]Cumulative Specific Pore Volume [cc/g]Initial pellet After bromination After hydrolysis Figure 5-10. Changes of pore size distribution after brominati on and hydrolysis in a calcium oxide pellet Hydrolysis Bromination

PAGE 93

93 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Initial pelletAfter brominationAfter hydrolysisPore Volume [cc/g] 0 ~ 5 m 5 m ~ Figure 5-11. Comparison of pore volumes of initial, after bromination and after hydrolysis in a calcium oxide pellet -1000 -900 -800 -700 -600 -500 -400 -300 -200 -100 0 29830040050060070080090010001100Temperature (K)Delta G (kJ/mol) CaO + SiO2 = CaSiO3 CaO + Al2O3 = CaAl2O4 CaO + ZrO2 = CaZrO3 Figure 5-12. Changes of Gibbs free energy of reactions between calcium oxide and ceramic materials as a function of reaction temperature

PAGE 94

94 Figure 5-13. Preparation step s for impregnation of calcium oxide on a yttria fabric Precipitated Calcium Carbonate (PCC) Ethyl alcohol, C2H5OH Mechanical Stirring (10 min) Ultrasonication (10 min) Impregnation by dropping on a yttria fabric Drying (100C, 30min) Sintering (900C, 20 hours) Calcium oxide dispersed and immobilized on a yttria fabric

PAGE 95

95 A B Figure 5-14. The XRD patterns of th e samples. A) Fresh sintered. B) After bromination reaction. yy -0.2 0 0.2 0.4 0.6 0.8 1050100150200250300350400450500550600Time[min]Conversion, X Figure 5-15. Conversion profiles of cyclic reactions, bromination and hydrolysis, of the calcium oxide fabric sample

PAGE 96

96 A B C Figure 5-16. The SEM images of samples. A) Bare yttria fabric. B) Fres h calcium oxide fabric sample. C) Brominated sample.

PAGE 97

97 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 020406080100120140160180200220240260Time[min]Conversion, X 8.7 wt % 10.7 wt % 11.9 wt % 22 wt % Figure 5-17. Comparison of the fi rst cyclic reactions of calcium oxide fabric samples with various amounts of calcium oxi de in the fabric sample

PAGE 98

98 Table 5-1. Composition conditions for th e pellets with different additives Additives (wt %) Sample Corn a Starch Stearic b acid Graphite b S1 0 0 0 S2 0 0 100 S3 0 100 100 S4 100 50 50 a The content of corn starch is expressed as weight percent of calcium. b The contents of stearic acid and graphite are expressed as wei ght percent of the sintered pellet. Table 5-2. Characteristics and e xperimental data of Ca-pellets Sample Surface area (m2/g) Total pore volume (cc/g) Max. Conversion Hydrolysis time (min) S1 7.61 0.6587 0.65 > 400 S2 6.69 1.213 S3 4.01 1.501 0.81 130 S4 5.58 1.508 0.71 120

PAGE 99

99 CHAPTER 6 THERMAL EFFICIENCY OF UT-3 PROCESS 6.1 Analysis Description The thermal efficiency of UT-3 cycle was evaluated thermodynamically. This parametric study focused on the effects of inert material amounts in the solid reactants, incomplete conversion and heat recovery on th e efficiency. These parameters we re not considered as a major factor for UT-3 process in great depth by ot her researchers, but the factors would have significant influence on the efficien cy as well as the operation cost in the actual scale plants since the solid reactants contain considerable amounts of inert materials and all reactions were found to be fractionally completed according to the experiments in this and previous studies. The thermal efficiencies ( ) were calculated based on the higher heating value (HHV) of the produced hydrogen according to the following equation: inH H of HHV2 (6-1) where, HHV of H2 = 286kJ/mol and Hin = Total heat input per mole of hydrogen to the system. In the analysis, one mole of hydrogen produc tion was considered a nd operating pressure was 1 atm. Assumptions for th is study are listed below: Negligible pressure drop, pumping work, and heat loss No excess water No separation work In pinch analysis, the mini mum approach temperature ( Tmin) is assumed to be 20K. 6.2 Pinch Analysis In order to determine feasible heat recovery in the process, pinch anal ysis was employed in this thermodynamic analysis. The pinch analysis also known as pinch point analysis has been used as a simple and efficient means to improve energy saving and reduce initial capital cost via optimization of the heat exchanger network in th e industrial applications (Ebrahim et al., 2000)

PAGE 100

100 since introduced by Linnhoff and Flower (1978). In the pinch anal ysis, hot and cold composite curves which represent heat availability and dema nd in the process are plotted with a gap by the minimum approach temperature ( Tmin) on a temperature-enthalpy diagram shown in Figure 61(A). The maximum amount of the possible heat recovery, the overlap region between the two composite curves, is indicated in Figure 6-1(B) (National Re sources Canada, 2003). 6.3 Thermodynamic Analysis The thermodynamic data for the four reactions in the UT-3 cycle were calculated using FactSage thermochemical software and databa ses (FactSage, Web version) and are summarized in Table 6-1. The optimized reaction temperatures determined in Chapter. 4 were employed. As shown in the Table 6-1, two bromination reactio ns, R1 and R3, are exothermic which release heat. On the other hand, the two hydrolys is reactions, R2 and R4, ar e endothermic which require thermal energy to drive the reactions. The total he at input for exothermic chemical reactions was 593.5kJ/mol and the total rejected heat from exothermic reactions was 345.95kJ/mol. The rejected heat from the exothermic reactions can be utilized for the endothermic reactions or heating of the reactant steam. The material and heat flow sheet for the whol e process is illustrated in Figure 6-2. The cooling and heating of all reacta nts and products for the following reactions are in dicated and the inert materials and incompletion of the conversion ar e not taken into account in the figure. It was assumed that one mole of water is supplied and one mole of hydrogen and a half of oxygen were produced at ambient temperature. 6.3.1 Ideal Case (Case 1: Complete Co nversion and No Inert Materials) Ideally, UT-3 process may be operated as s hown in Figure 6-2. One mole of hydrogen can be produced with the chemicals in the flowsheet provided all the reacti ons are completed and no inert materials are required in the solid reactants.

PAGE 101

101 The enthalpy changes for the heating and coo ling of the materials were calculated in consideration of the sensible a nd latent heats in the process. The sensible enthalpies were determined using a mean heat capacity over th e temperature range. The required energy for heating was 261.3kJ and the reje cted heat from cooling pro cesses was 228kJ to produce one mole of hydrogen, respectively. The amount of recovered heat from the coo ling processes and exothermic reactions was calculated via pinch analysis. The hot composite curve represents heat available in the process and cold composite curve represents heat demand for the process are plo tted in Figure 6-3(A) and (B), respectively. The combined composite curves with the minimum approach temperature (20K) are plotted in Figure 6-3(C). According to the pinch analysis, the maximum amount of recovered heat was 324kJ. The thermal efficiencies were calculated with and without heat r ecovery. The calculated thermal efficiencies with and without heat recovery were 53.9 and 33.5% as shown in Table 6-3. 6.3.2 Effect of Inert Materials on Efficiency (Case 2: Comple te Conversion and Including Inert Materials) The solid reactants are integrated forms of actual reactants and inert materials in order to make the reactants reactive and durable in th e cyclic transformation. Calcium titanate (CaTiO3, Aihara et al., 1999) and iron titanate (Fe2TiO5, Sakurai et al., 2006) were used for calcium oxide pellets and iron oxide pellets as the inert material, respectively. In this study, yttria fibers play the role of supporting the calcium oxide reactant. The amount of the inert material was greater than the reactant in the final forms of pellets a nd fabrics. Hence, the inert materials incorporated in the solid reactants should also be included in the calculation of enthalpy changes for heating and cooling. The efficiency comparison of the pell ets of Aihara et al. ( 1999) and Sakurai et al. (2006) and fabrics in this study is somewhat difficult since the t ypes of inert material and the

PAGE 102

102 molar ratios of the actual reactants to the inert ma terials are different. The molar ratios of actual reactants to inert materials and the additional ener gies for the other researchers (pellet) and this study (fabric) were calculated and are listed in the Tables 6-4 and 6-5, respectively. As shown in the tables, the absolute values of energies for heating and cooling of the inert material in each table were same because the inert materials were not consumed or generated during the processes. It was seen that there is a discrepancy of 121.9 kJ (=410.4kJ-288.5kJ) between the values of other researchers and this study because of the di fference in the heat capacities of the inert materials and reactant/inert molar ratios. For pinch analysis, the combined composite curv es for pellet and fabric samples with inert materials are plotted in Figure 64(A) and (B), respectively. The thermal efficiencies including the effect of the inert materials are shown in Table 6-6. The 33.5% efficiency without h eat recovery dropped to 22.6% and 25% for the pellet type reactants and the fabric type r eactants, respectively. With heat recovery, the 53.9% efficiency was slightly reduced to 52.0% and 52.6%, respecti vely because most of the rejected heat from the cooling of the inert material was recovered in the pinch anal ysis. However, the minimization of the inert materials is prefer able even with heat recovery since the additional capital and operational costs for heat exchangers to recover th e rejected heat is required. The reason of the somewhat greater drop in the efficiency for the pe llets is mainly due to the higher heat capacities of the inert materials, CaTiO3 and Fe2TiO5 compared to that of yttria. 6.3.3 Effect of Inert Material and Incomplete Conversion on Efficiency (Case 3: Including inert material and incomplete conversion (CaO: 85%, Fe3O4: 90%)) The effect of incomplete conversion as well as inert material on the thermal efficiency was evaluated. The experiments revealed incompletio n of each bromination reaction possibly due to the diffusion limitation of gaseous reactants into the core of the solid reactant. The maximum

PAGE 103

103 conversions of calcium oxide and iron oxide were observed to be 85% (this study) and 90% (Sakurai et al., 2006) in the br omination reaction, respectively. Hence, these values were employed in this calculation. Additional reactan ts were added to produce one mole of hydrogen due to the incomplete conversion. The energies required and rejected for heating and cooling of inert material, respectively, and the additional unr eacted solid reactant are listed in Tables 6-7 and 6-8 for the pellet type reactants from other re searchers and for fabric type reactants of this study, respectively. For pinch analysis, the combined composite curv es for pellet and fabric samples with inert materials and incompletion conversion are plotte d in Figure 6-5(A) and (B), respectively. The thermal efficiencies includ ing the effects of inert mate rial & incomplete conversion were calculated and are listed in Table 6-9. The effects of incomplete conversion on the efficiencies of pellets and fabr ics were not great due to the high conversions employed in the calculation. The efficiencies dropp ed by about 1% without heat r ecovery and by 0.2% with heat recovery due to incomplete conversion. 6.4 Summary of Thermodynamic Analysis The efficiencies of UT-3 cycle have been evaluated in the earlier studies; however, inert materials, incomplete conversion and heat recovery were not considered in those analyses. In this chapter, a thermodynamic analysis has been carried out in order to determine the effect of the inert materials, incomplete conversion and heat recovery on the efficiency. The analyses used the experimental data on the amounts of inert materials in the pellets and fabrics and the maximum bromination conversions of calcium oxid e and iron oxide reactants. The results of the thermodynamic analysis are illustra ted in Figure 6-6. Without heat recovery, the efficiency with complete conversion and no inert materials was 33.5% and the efficiencies for pellet and fabric reactants including inert material s were 22.6% and 25.0%, respectively. The efficiencies for

PAGE 104

104 pellet and fabric reactants incl uding inert materials and incomp lete conversion were 21.6% and 24.1%, respectively, without heat recovery. With heat recovery, the efficiency with complete conversion and no inert materials wa s 53.9% and the efficiencies fo r pellet and fabric reactants including inert materials were 52.0% and 52.6%, respectively. The efficiencies for pellet and fabric reactants including inert materials and incomplete conversion were 51.8% and 52.4%, respectively, with heat recovery. The efficiency differences with and without heat recovery were in the range of 20.4-30.2%. The inert materials accounted for 10. 9 and 8.5% reductions of the efficiency for the pellet and fabric reactants, re spectively, without heat recovery. On the other hand, the effect of the inert materials on the effici encies was not considerable with heat recovery since most of the rejected heat from the cooli ng of the inert material was found to be recovered in the pinch analysis. The effici encies for the pellet and fabric reactants only dropped by 1.9 and 1.3%, respectively. The thermal efficiencies of the pellets were slightly lower than those of fabric type reactants in this study whether or not heat recovery was used, while the amounts of the inert materials are somewhat smaller. The re ason of the lower efficiencies is higher heat capacities of inert materials, CaTiO3 and Fe2TiO5, in the pellets compared to the yttria fabric. The influence of incomplete conversion was not significant because of the high conversions, 85% and 95%, used for this estimation. To sum up it was found that the eff ect of heat recovery and inert materials cannot be underest imated in the calculati on of thermal efficiency of the cycle. Thus, in order to increase the efficiency, heat re covery must be employed and the use of the inert materials should be minimized.

PAGE 105

105 A B Figure 6-1. Temperature-enth alpy diagrams. A) Hot and cold composite curves and B) maximum heat recovery on temperature-en thalpy diagram. Adapted from National Resources Canada, 2003. Pinch analysis: for the effici ent use of energy, water & hydrogen CANMET Energy Technology Centre-Var ennes. (Figure 3-5, page 25). Available at http://cetc-varennes.nrcan.gc.ca/ fichier.php/codectec/En/2003-140/2003140e.pdf 60 100 150 T ( ) 2000 4000 6000 H (kW) Qrequired Maximum Heat Recovery Pinch Tmin Qrejected 60 100 150 T ( ) 2000 4000 6000 H (kW) Cold Composite Curve Hot Composite Curve Pinch Tmin

PAGE 106

106 Figure 6-2. Process flowsheet of heat and material in the UT-3 cycle H2 HEAT H2O 0.5O2 CaO CaBr2 (R2) CaBr2(s) + H2O(g) CaO(s) + 2HBr(g)(1000K, Endothermic) (R1) CaO(s) + Br2(g) CaBr2(s) + 0.5 O2(g) (873K, Exothermic) (R4) 3FeBr2(s) + 4H2O(g) Fe3O4(s) + 6HBr(g) + H2(g) (800K, Endothermic) Br2 2HBr (R3) Fe3O4(s) + 8HBr(g) Br2(g) + 3FeBr2(s)+ 4H2O(g) (400K, Exothermic) Fe3O4 3FeBr2 4H2O 6HBr H1 H2 H3 H4 H5 C1 C2 C3 C4 C5 C6

PAGE 107

107 A200400600800 900 700 500 300 Enthalpy [kJ]Temperature [K]0 Hot composite curve 900 700 500 300Temperature [K]800 600 400 200 Enthalpy [kJ] 0 Cold composite curve B C 324.0 kJ 249.5 kJ530.6 kJ Waste Heat Heat RecoveredHeat RequiredTemperature [K]Enthalpy [kJ] 1000 200400600800 300 500 700 900 0 Cold composite curve Hot composite curve Figure 6-3. Temperature-enthalpy diagrams for Case 1. A) Hot composite curve. B) Cold composite curve. C) Combined composite curves for pinch analysis.

PAGE 108

108 Temperature [K]300 500 700 900 Enthalpy [kJ] 1000 200400600800 0 1400 12001600715.2 kJ 268.7 kJ 549.9 kJWaste Heat Heat Recovered Heat Required Cold composite curve Hot composite curve A 599.7 kJ 543.4 kJHeat RecoveredHeat Required Waste Heat Temperature [K]300 500 700 900 Enthalpy [kJ] 1000 200400600800 0 1400 1200 Cold composite curve Hot composite curve 262.2 kJ B Figure 6-4. Combined composite curves (C ase 2). A) Pellet and B) Fabric samples

PAGE 109

109 1600 12001400 0 800 600 400 2001000 Enthalpy [kJ] 900 700 500 300Temperature [K] Waste Heat271.3 kJHeat Required Heat Recovered552.6 kJ 770.3 kJ Cold compos ite curve Hot composite curve A 642.8 kJ 545.3 kJHeat RecoveredHeat Required 264.1 kJWaste Heat Temperature [K]300 500 700 900 Enthalpy [kJ] 1000 200400600800 0 1400 1200 Cold composite curve Hot composite curve B Figure 6-5. Combined composite curves (C ase 3). A) Pellet and B) Fabric samples

PAGE 110

110 33.5% 53.9% 22.6% 21.6% 51.8% 25.0% 52.6% 24.1% 52.4% 52.0%0% 10% 20% 30% 40% 50% 60% No heat recoveryHeat recoveryThermal Efficiency CASE 1: No inert material & complete conversion CASE 2: Include inert material (pellet) CASE 2: Include inert material & incomplete conversion (pellet) CASE 3: Include inert material (fabric) CASE 3: Include inert material & incomplete conversion (fabric) Figure 6-6. Comparison of thermal e fficiency in various situations

PAGE 111

111 Table 6-1. Thermodynamic data of the reactions in UT-3 cycle No. Reaction Rx Temp. H [kJ/mol] G [kJ/mol] Remarks R1 CaO(s) + Br2(g) CaBr2(s) + 0.5O2(g) 873K -73.55 -38.22 Exothermic R2 CaBr2(s) + H2O (g) CaO(s)+ 2HBr(g) 1000K 211.7 104.3 Endothermic R3 Fe3O4(s) + 8HBr(g) Br2(g) + 3FeBr2(s)+ 4H2O(g)400K -272.4 -149.6 Exothermic R4 3FeBr2(s) + 4H2O(g) Fe3O4(s) + 6HBr(g) + H2(g)800K 381.8 146.3 Endothermic Table 6-2. Energies required and rejected (Enthalpy change for heating and cooling based on latent heat and sensible heat) Temperature [K] Cp [J/(K-mol)] Enthalpy change [J] Remark Heating or cooling Initial FinalInitial Final Latent Sensible H1 CaBr2(s) CaBr2(s) 873 100085.15 88.61 11033.8 H2 Br2(g) Br2(g) 400 873 36.67 37.66 17579.8 H3 3FeBr2(s) 3FeBr2(s) 400 800 82.50 91.41 104343 3 mol H4 4H2O(g) 4H2O(g) 400 800 34.27 38.73 58395.2 4 mol 298 373 75.38 76.00 40626.0 5676.64 H5 H2O(l) H2O(g) 373 100034.05 41.26 23609.7 Energy required for heating 261.3 kJ C1 CaO(s) CaO(s) 1000 873 53.51 52.89 -6756.59 C2 2HBr(g) 2HBr(g) 1000 400 32.31 29.22 -36912.6 2 mol C3 Fe3O4(s) Fe3O4(s) 800 400 266.7 174.0 -88144.6 C4 6HBr(g) 6HBr(g) 800 400 31.07 29.22 -72340.8 6 mol C5 0.5O2(g) 0.5O2(g) 873 298 34.20 29.38 -9139.34 0.5 mol C6 H2(g) H2(g) 800 298 29.63 28.84 -14675.0 Rejected heat from cooling -228 kJ Table 6-3. Thermal efficiencies with comple te conversion and no inert materials (Case 1) No recovery Heat recovery Thermal energy for reactions 593.5 kJ/mol 593.5 kJ/mol Total energy for heating 261.3 kJ./mol 261.3 kJ./mol Recovered thermal energy 324 kJ/mol Thermal efficiency 33.5% 53.9%

PAGE 112

112 Table 6-4. Additional energies required and rejected for heating and cooling of inert material CaO:CaTiO3=1:1.4*, Fe3O4:Fe2TiO5=1:5** (Pellet) Temperature Cp [K] [J K-1mol-1] Enthalpy change [J/mol] Heating or cooling Initial FinalInitial Final Sensible Total enthalpy change [J] H6 CaTiO3(s)CaTiO3(s) 873 1000128.8 130.4 16458.8 23042.3 H7 Fe2TiO5(s)Fe2TiO5(s) 400 800 182.0 205.4 77473.2 387366 Energy required for heating 410.4 kJ C7 CaTiO3(s)CaTiO3(s) 1000 873 130.4 128.8 -16458.8 -23042.3 C8 Fe2TiO5(s)Fe2TiO5(s) 800 400 205.4 182.0 -77473.2 -387366 Rejected heat from cooling -410.4 kJ Molar ratio of CaO to CaTiO3 in calcium oxide pellets. (Aihara et al., 1999). ** Molar ratio of Fe3O4 to Fe2TiO5 in iron oxide pellets. (Sakurai et al., 2006). Table 6-5. Additional energies required and rejected for heating and cooling of inert material CaO:Y2O3=1:2*, Fe3O4: Y2O3=1:5** (Fabric) Temperature Cp [K] [J K-1mol-1] Enthalpy change [J/mol] Heating or cooling Initial Final InitialFinal Sensible Total enthalpy change [J] H6 Y2O3(s)Y2O3(s) 873 1000 124.6 126.9 15970.5 31941 H7 Y2O3(s)Y2O3(s) 400 800 131.8 124.7 51306.8 256534 Energy required for heating 288.5 kJ C7 Y2O3(s)Y2O3(s) 1000 873 126.9 124.6 -15970.5 -31941 C8 Y2O3(s)Y2O3(s) 800 400 124.7 131.8 -51306.8 -256534 Rejected heat from cooling -288.5 kJ Molar ratio of CaO to CaTiO3 in calcium oxide fabrics. (this study). ** The molar ratio of Fe3O4 to Y2O3 was assumed same as the value of iron pellet. (Sakurai et al., 2006). Table 6-6. Thermal efficiencies with comple te conversion and iner t material (Case 2) Pellets (other researches*) Fabrics (this study) No recovery Heat recoveryNo recovery Heat recovery Thermal energy for reactions 593.5 kJ/mol 593.5 kJ/mol 593.5 kJ/mol 593.5 kJ/mol Total energy for heating 671.7 kJ./mol 671.7 kJ./mol 549.8 kJ./mol 549.8 kJ./mol Recovered thermal energy 715.2 kJ/mol 599.7 kJ/mol Thermal efficiency 22.6% 52.0% 25.0% 52.6% From Aihara et al., 1999 and Sakurai et al., 2006.

PAGE 113

113 Table 6-7. Additional energies required and reject ed for heating and cooling of inert material and unreacted solid reactant, CaO:CaTiO3=1:1.4*, Fe3O4:Fe2TiO5=1:5** (Pellet) Temperature Cp [K] [J K-1mol-1] Enthalpy change [J/mol] Heating or cooling Initial Final Initial Final Sensible Total enthalpy change [J] H6 CaTiO3(s) CaTiO3(s) 873 1000 128.8 130.4 16458.8 27189.9 H7 Fe2TiO5(s) Fe2TiO5(s) 400 800 182.0 205.4 77473.2 429976 H8 0.18CaO(s) 0.18CaO(s) 873 1000 52.89 53.51 6756.59 1216.19 H9 0.11Fe3O4(s) 0.11Fe3O4(s) 400 800 174.0 266.7 88144.6 9695.91 Energy required for heating 468.1 kJ C7 CaTiO3(s) CaTiO3(s) 1000 873 130.4 128.8 -16458.8 -27189.9 C8 Fe2TiO5(s) Fe2TiO5(s) 800 400 205.4 182.0 -77473.2 -429976 C9 0.18CaO(s) 0.18CaO(s) 1000 873 53.51 52.89 -6756.59 -1216.19 C10 0.11Fe3O4(s) 0.11Fe3O4(s) 800 400 266.7 174.0 -88144.6 -9695.91 Rejected heat from cooling -468.1 kJ Molar ratio of CaO to CaTiO3 in calcium oxide pellets. (Aihara et al., 1999). ** Molar ratio of Fe3O4 to Fe2TiO5 in iron oxide pellets. (Sakurai et al., 2006). Table 6-8. Additional energies required and reject ed for heating and cooling of inert material and unreacted solid reactant, CaO:Y2O3=1:2*, Fe3O4: Y2O3=1:5** (Fabric) Temperature Cp [K] [J K-1mol-1] Enthalpy change [J/mol] Heating or cooling Initial Final Initial Final Sensible Total enthalpy change [J] H6 Y2O3(s) Y2O3(s) 873 1000 124.6 126.9 15970.5 37690.4 H7 Y2O3(s) Y2O3(s) 400 800 131.8 124.7 51306.8 284753 H8 0.18CaO(s) 0.18CaO(s) 873 1000 52.89 53.51 6756.59 1216.19 H9 0.11Fe3O4(s) 0.11Fe3O4 400 800 174.0 266.7 88144.6 9695.91 Energy required for heating 333.4kJ C7 Y2O3(s) Y2O3(s) 1000 873 126.9 124.6 -15970.5 -37690.4 C8 Y2O3(s) Y2O3(s) 800 400 124.7 131.8 -51306.8 -284753 C9 0.18CaO(s) 0.18CaO(s) 1000 873 53.51 52.89 6756.59 -1216.19 C10 0.11Fe3O4(s) 0.11Fe3O4(s) 800 400 266.7 174.0 88144.6 -9695.91 Rejected heat from cooling -333.4kJ Molar ratio of CaO to Y2O3 in calcium oxide fabrics. (this study). ** The molar ratio of Fe3O4 to Y2O3 was assumed same as the value of iron pellet. (Sakurai et al., 2006).

PAGE 114

114 Table 6-9. Thermal efficiencies including iner t material & incomplete conversion (Case 3) Pellets (other researches*) Fabrics (this study) No recovery Heat recovery No recovery Heat recovery Thermal energy for reactions 593.5 kJ/mol 593.5 kJ/mol 593.5 kJ/mol 593.5 kJ/mol Total energy for heating 729.4 kJ./mol 729.4 kJ./mol 594.7 kJ./mol 594.7 kJ./mol Recovered thermal energy 770.3 kJ/mol 642.8 kJ/mol Thermal efficiency 21.6% 51.8% 24.1% 52.4% From Aihara et al., 1999 and Sakurai et al., 2006.

PAGE 115

115 CHAPTER 7 CALCIUM OXIDE ABSORBENT FOR HIGH TEMPERATURE CARBON DIOIXDE CAPTURE 7.1 Introduction The emissions from fossil fuel power plants ar e one of the largest sources of anthropogenic carbon dioxide emissions in the atmosphere. The carbon dioxide from the power plants can be separated from the sources via the post-com bustion, pre-combustion and oxyfuel combustion carbon dioxide capture processes (Metz et al., 2005). Among them, the pre-combustion process is considered as a feasible way to capture car bon dioxide in the coal gasification and the steam methane reforming (SMR) processes. The conventional coal gasification and SMR consist of the following reformation and water-gas shift reactions: Reformation 2 2yH CO O H CHx (7-1) (coal gasification: x=0, y=1, SMR: x=4, y=3) Water-Gas Shift reaction 2 2 2H CO CO O H (7-2) Hydrogen yield from these processes can be increa sed using absorbents to react with or absorb carbon dioxide since the forward equilibrium sh ift would occur by removing carbon dioxide in the water-gas shift reaction (Bal asubramanian et al., 1999; Lin et al., 2002). Various absorbents have been introduced and studied, but calcium ox ide based absorbents seem very promising in consideration of the operating temperature, pre ssure and capture capacity (Gupta and Fan, 2002). The zero-emission coal (ZEC) process using the carbonation/ calcination reactions for carbon dioxide sequestration and fo r higher hydrogen yield was intr oduced and discussed by S owinski (2006). The author stated that th e process is attractive because el ectricity can be generated with high efficiency and without emission of carbon dioxide, though improveme nt of stability and

PAGE 116

116 cyclic kinetics of CaO/CaCO3 bed are required. In the thermodynamic study of Mahishi et al. (2005), the heat duty for gasifica tion was reduced by almost 42% vi a in-situ heat transfer since CO2 absorption is exothermic, and the hydrogen yield increased by about 19% while production of carbon dioxide was reduced by 50.2%. Similarly, Satrio et al. (2005) reported that hydrogen yield was increased 20% with a combined cata lyst and absorbent for methane reforming and Hanaoka et al. (2005) also f ound that hydrogen yield was signif icantly increas ed with the reduction of carbon dioxide and methane by the c ontact between the biomass and calcium oxide on gasification. For their study, th e authors fabricated and experi mentally tested core-in-shell spherical pellets which consist of calcium oxide co re and alumina shell contains nickel catalyst. The carbonation and calcination reacti ons are described as follows: Carbonation: ) ( ) ( ) (3 2s CaCO g CO s CaO mol kJ HK/ 7 169973 (exothermic) (7-3) Calcination: ) ( ) ( ) (2 3g CO s CaO s CaCO mol kJ HK/ 3 1661173 (endothermic) (7-4) Carbon dioxide reacts with calcium oxide to form calcium carbona te in the carbonation reaction and the calcium oxide is regenerated and pure carbon dioxide can be obtained through the calcination reaction. However, substant ial volume changes be tween carbonate (36.9 cm3/mol) and oxide forms (16.9 cm3/mol) are induced by these ga s-solid reactions (Stanmore and Gilot, 2005). These stru ctural and thermal stresses caused by the cyclic carbonationcarbonation reaction lead to the lo ss in active surface area, pore plugging and sintering of the particles in the absorbent. As a result, the pe rformance of the absorbent was found to degrade seriously in the cyclic operation (Barker, 1973; Borgwardt, 1989). This degradation should be overcome or minimized for the calcium oxide absorbent to be practical. For that reason, various struct ural forms of calcium oxide to capture carbon dioxide such as dolomite (CaCO3MgCO3) (Curran et al., 1967; Dobner et al., 1977; Silaban et al., 1996),

PAGE 117

117 calcium oxide dispersed in porous inert calcium titanate (CaTiO3) matrix (Aihara et al., 2001), impregnated in porous alumina granules (F eng et al., 2006), or mixed with mayenite (Ca12Al14O33) (Li et al., 2005 and 2006) or nano-sized alumina (Al2O3) particles (Wu et al., 2008) and core-in-shell catalysis/sorbent (S atrio et al., 2004, 2005 and 2007) have been introduced and investigated to improve the cyclic performanc e of the absorbents for carbon dioxide capture. Most of these attempts show ed better cyclic perfor mance than pure calcium oxide thanks to the inert materials, but those br ing other drawbacks, comp lexity in preparation, high cost for the synthesis and low content of calcium oxide in the inert materials. In this study, a relatively simple and cost e ffective immobilization pr ocedure of nano-sized calcium oxide particles on a fibrous ceramic fabric which acts as a support of the calcium oxide particles was introduced to enhance its cyclic performance. The characteristics and cyclic performance of the proposed im mobilized calcium oxide on the fabric were examined and compared with other results in the literature. 7.2 Experimental Procedure 7.2.1 Immobilization of Calcium Oxide on a Ceramic Fabric A new procedure for the immobilization of calcium oxide in a ceramic fiber was developed. Calcium carbonate (CaCO3) was selected as a starting ma terial. The nano-sized calcium carbonate powder is readily available and inexpens ive. In order to improve the homogeneity and dispersibility of the particles, precipitated nano-particulate calcium carbonate (PCC; average particle size = 70 nm) surface tr eated with stearic acid was procured from the manufacturer (Specialty Minerals, USA). The PCC was blended with ethyl alcohol under mechanical stirring and the suspension was ultrasonicated subseque ntly to improve dispersibility and reduce agglomeration of the powders. The resultant sl urry was dropped onto a dried ceramic fabric

PAGE 118

118 (alumina or yttria) using a transfer pipette. The impregnated fabric wa s dried at 100C for 30 min and sintered at 800C for 12 hours in air. The preparation steps are show n in brief in Figure 7-1. 7.2.2 Cyclic Reaction Experiment The cyclic carbonation-calcina tion reactions were conducted in a Thermogravimetric (TG) Analyzer (Model: SDT-Q600). The cyclic experiments were performed in various conditions. The prepared samples were evaluated under the is othermal condition at 7 50C and at different temperatures of 700C and 850C for car bonation and calcination, respectively. 20 vol. % of carbon dioxide in nitrogen was supplied for the severe calcination wh ile pure nitrogen was supplied for the mild calcination. The reaction conditions for the m ild and severe calcinations are summarized in Table 7-1. The degree of carbonation conversi on (X) of the calcium oxide in the sample was defined as moles of calcium oxide reacted with carbon dioxide over moles of calcium oxide initially present: ) ( ) (2oxide calcium initial n CO with reacted oxide calcium n X (7-5) The moles of calcium oxide react ed with carbon dioxide were calcu lated from the weight change of the sample measured by the TG analyzer. The trend of conversion of the carbonation r eaction and the temperature for the calcium oxide fabric sample with respect to the reaction time are plotted in Figur e 7-2. Alumina fabric was used as a substrate. There was a 2.5% wei ght loss as the furnace was heated at a rate of 20C/min to 750C under a pure nitrogen atmosphere It was believed that the weight loss is possibly caused by the thermal d ecomposition of calcium hydroxide that might have formed by the reaction between calcium oxi de and moisture. Once the temperature reached 750C, 20 vol % CO2 in nitrogen was delivered for the carbonation reac tion. Like other trends in the literature, the

PAGE 119

119 carbonation reaction was very rapid in the initial stage while it decreased as time passed since the calcium oxide particles are covered with a layer of calcium carbonate, which imposes a limitation of the reactant gas diffusion (Bhatia and Perlmutter, 1983; Mess et al., 1999). 60% conversion was attained within approximately 20 minutes while 80% conversion to ok two hours. Therefore 20 minute duration for the carbonation r eaction was chosen for th e cyclic experiment in consideration of the effectiveness. The conversion and temperature profiles for cy clic carbonation/calcina tion reactions of the calcium oxide sample (alumina fabric) are illu strated in Figure 7-3. For calcination, pure nitrogen at 750C was applied. The conversion began at comparatively low level (about 59%), but the value gradually in creased in the first couple of cycles and seemed to stabilize at about 75%. The increase in the conversi on in the early stage was possibl y due to the increase in the surface area by the initial structur al transformation that was caused by the volume change. Each profile consists of the first fa st stage and the second sluggish stage of the carbonation reaction. It can be observed that the temp erature fluctuated very sli ghtly since the carbonation and calcination reactions are exotherm ic and endothermic, respectivel y. The calcination reactions were completed within at most 5 minutes. The degree of conversion of the sample was maintained at the same level after several cycles and showed no signs of decrease even after 13 cycles. The absolute capacity of carbon dioxide capture based on the to tal sample weight is not high due to the low calcium oxide content in the samples, about 23 wt %. Therefore the impregnation step in Figure 7-1 was repeated fo r higher calcium oxide cont ent. It was confirmed that the content of calcium oxide in the fabr ic increased in proporti on to the number of the impregnation steps. The maximum carbonation c onversions of two samples with different

PAGE 120

120 calcium oxide contents, 23 wt % and 55 wt %, in the alumina fabric under the mild condition are plotted in Figure 7-4. The initial maximum carbona tion conversions were comparatively low, but the value increased and maintained at the same level. The sample with 23 wt % calcium oxide content attained about 75% car bonation conversion after 13 car bonation-calcination cycles and the carbonation conversion of the sample with 55 wt% calcium oxide reached about 62% after 10 cycles. It is possible that the lower conversion of the sample w ith higher calcium oxide contents is caused by higher diffusion resistance in the absorbent. The weight resolution of TGA instrument is 0.1 g. Hence, the error due to the TGA instrument in measurement error was negligible compared to the amount of the weight change of the samples. The maximum amounts of reacted calcium oxide in the two samples ar e measured against those in the pure calcium oxide forms from the previous re searchers in Figure 7-5. Two empirical curve fits of the e xperimental test data for vari ous types of calcium oxide under different conditions from previous studies (Barker, 1973; Curran et al., 1967; Aihara et al., 2001; Shimizu et al., 1999; Silaban et al., 1995) we re developed (Abanade s and Alvarez, 2003; Abanades, 2002). One of the curve fit equations is given below (Abanades and Alvarez, 2003), which is also reproduced in the figure without actual experimental da ta points in order to compare our results with previous studies in the literature. 17 0 77 0 83 0 NX (7-6) The fitted curve from the literature dropped steeply w ith the number of cycles while the values of the samples in this study gradually increased fo r the first few cycles and remained stable. The absolute capacities of carbon dioxide capture of the sample with 55 wt % calcium oxide content exceeded that of the conventiona l pure calcium oxide after the sixth cycle. The amount of carbon

PAGE 121

121 dioxide capture of the sample with 23 wt % cal cium oxide is lower due to the lower calcium oxide content in samples but the value of the samp le is also expected to exceed the fitted curve approximately after the 25th cycle. 7.2.3 Characterization The structure of the impregnated sample was examined under SEM. The observation focused on the structural changes over the cycl ic reactions. The composition of the sample was investigated by X-ray diffraction (XRD) experiments to confirm th e conversion of calcium oxide into calcium carbonate during carbonation a nd the regeneration of calcium oxide after calcination. The structure of the impregnated sample was examined under SEM. Figure 7-6 shows the SEM images of an unaltered alumina fabric and a fresh sintered calcium oxide mat. The original alumina fabric consists of micr on sized alumina fibers which have a diameter of about 2 to 5 m (Figure 7-6( A)). It was observed that calciu m oxide particles are supported by the alumina fibers and interconnected microstructu res of the calcium oxide were formed between the alumina fibers in the mat as shown in Figur e 7-6(B). Some threads are exposed, while most others are buried by the calcium oxide particulate aggregates. It was observed that the P CC was converted into calcium oxide during the sintering process by XRD analysis. Figure 7-7 shows X-ray diffraction patterns of the sample (23 wt % CaO) (A) fresh sinter ed (B) after the 10th carbonation and (C) after the 10th calcination. The XRD data indicate that the samples after sintering and the 10th calcination contain only calcium oxide while calcium carbonate accounts for the majo r compound along with small quantities of unreacted calcium oxide after the 10th carbonation. The changes in the surface area during cyclic operation of the sample were measured by nitrogen adsorption in an AUTOSORB-1 instru ment from Quantachrome Instruments. The

PAGE 122

122 surface areas of the sample with 23 wt % calcium oxide in the alum ina fabric over several cyclic reactions are shown in Figure 7-8. Considering the molar volumes of calcium oxide and calcium carbonate, it was expected that the surface area would diminish during the carbonation reaction and would be regenerated after the calcination reaction. The fresh sample began with a comparatively low surface area (9.2 m2/g). It decreased to 5.3 (m2/g) after carbonation due to the volume expansion of the particles as expected, but it increased dr astically to about 18.3 m2/g after one cycle. Afte r that, values of the surface area of the sample after calcination remained in the range of 16-19 m2/g. Based on the figure, one can conclude from the figure that the big jump of the surface ar ea in the initial stage was caused by the permanent partial structural breakage and void generation owing to the volume contraction from carbonate to oxide form. Figure 7-9 shows the magnified images of the freshly sintered sample (23 wt % CaO) and the sample after the 10th carbonation and calcination reactions. From images (a) and (b), it was observed that the calcium oxide particles which have a diameter of about 150 nm, are dispersed in the alumina fabric. After the 10th carbonation reaction (c and d), the structure was comparatively closed-packed with larger particle s possibly due to the volume expansion of the particles. The structure after the 10th calcination seems to consist of more interconnected agglomerates of the calcium oxide particles with a stabilized high surface area in the images (e) and (f). 7.2.4 Severe Calcination Condition Practically, the severe condition for calcinati on reaction is preferab le for carbon dioxide sequestration or utilization since higher c oncentration of carbon dioxide from carbonated absorbents can be obtained. However, it was observed that the severe calcination under higher temperature and presence of car bon dioxide pulled down the perf ormance of the calcium oxide

PAGE 123

123 absorbents significantly (Li et al., 2005 and 2006; Wu et al., 2008; Silaban et al., 1995; Grasa et al., 2007). Under high temperature, it was proved that the mixture of calcium oxide and alumina react to form a new compound, Ca12Al14O33 (Wu et al., 2008), in the co mposite of calcium oxide and alumina. This new inert materials could be a reason for the degrad ation under the severe calcination condition. For this reason, a yttria fabric was also in troduced as a substrate in place of the alumina fabric. The cyclic carbonation-calcination experime nts of the two samples using different fabric materials, alumina and yttria, we re conducted under the severe cal cination condition in Table 7-1 and the maximum conversion trends of the sample s are illustrated in Fi gure 7-10. As shown in the figure, the maximum carbonation conversions (about 55%) of the sample using yttria fabric as a substrate showed no sign of degradation over the 12 cycles while those of the sample using alumina dropped by about eight percent after 12 cy cles from the maximum of 59%,. Li et al. (2005) observed the decline of the carbonation co nversion due to the presence of 14 wt % CO2 in calcination process at 850C becau se the sintering rate of calcium oxide particles at a given temperature was accelerated by the presence of CO2 (Borgwardt, 1989). Judging from these experimental results and the previous literature, it was concluded that the yttria fabric is superior to the alumina fabric as a substrate for calcium oxide absorbent at the severe calcinations condition due to the possibil ity of the formation of Ca12Al14O33 by the reaction between calcium oxide and alumina under the high temperature over 800C. Also this new proposed calcium oxide reactant using yttria was found to be better than the Ca-based CO2 absorbent, (CaO/ Ca12Al14O33), proposed by Li et al. (2005) since the cyclic performance was maintained even with 20% CO2 at 850C.

PAGE 124

124 7.3 Summary In order to minimize the degrad ation of the absorption capacity in cyclic operation due to the pore plugging and sintering of particles, the calcium oxide on fi brous ceramic fabricated via a simple and effective immobilization process deve loped for the UT-3 cycle was applied to carbon dioxide capture at high temperat ure. The prepared samples were characterized and evaluated by various analytical and experimental tools co mprehensively. Two samples with 23 wt% and 55 wt% calcium oxide contents in the alumina fabric achieved continuous cyclic carbonation conversions, about 75% and 62% over 13 and 10 carbonation-calcination cy cles under the mild calcinations conditi on at 750C in N2. Under the more severe calci nation condition at 850C and 20 wt% CO2 in N2, it was confirmed that the yttria fabr ic was superior to the alumina as a substrate for carbon dioxide capture. The reactivity of the calcium oxide absorbent using the yttria fabric was maintained at the same leve l in 12 cycles. The stabilized conversion was about 55%. On the other hand, the sample using the al umina fabric showed degradation by about eight percent from the maximum, 59%, after 12 cy cles possibly due to the formation of Ca12Al14O33.

PAGE 125

125 Figure 7-1. Preparation steps for immobiliz ation of calcium oxide on a ceramic fabric 0 0.2 0.4 0.6 0.8 1 020406080100120140 Time (min)Conversion (X) 0 100 200 300 400 500 600 700 800Temperature (C) Figure 7-2. Carbonation conversi on of the prepared sample Precipitated Calcium Carbonate ( PCC ) Ethyl alcohol, C2H5OH Mechanical Stirring (10 min) Ultrasonication ( 10 min ) Dropping the Slurry Drop on a Ceramic Fabric Dr y in g ( 100C 30min ) Sinterin g ( 800C 12hours ) Calcium Oxide Absorbent

PAGE 126

126 0 0.2 0.4 0.6 0.8 1 050100150200250300350400450 Time (min)Conversion (X) 0 100 200 300 400 500 600 700 800Temperature (C) Figure 7-3. Conversion prof iles of cyclic reactions 0 0.2 0.4 0.6 0.8 012345678910111213 Number of cycles (N)Carbonation conversion (X ) 23 wt% of CaO 55wt% of CaO Figure 7-4. Maximum conversions of carbonation reaction of tw o samples loaded different calcium oxide contents, 23 wt % and 55 wt %, with the number of cycles

PAGE 127

127 0 0.2 0.4 0.6 0.8 1 012345678910111213 Number of cycles (N)Reacted CaO [g] / Initial calcined sample weight [g] Reported by previous studies 23 wt% (this study) 55 wt% (this study) Figure 7-5. Maximum amounts of reacted calciu m oxide in the carbonation reactions based on initial sample weight with the number of cycles A B Figure 7-6. The SEM images of a fabric sample. A) An original alumina fabric. B) Fresh sintered sample.

PAGE 128

128 A B C Figure 7-7. The XRD patterns of the samples. A) Fresh. B) After 10th carbonation. C) After 10th calcination (10 cycles).

PAGE 129

129 Figure 7-8. Change of surface area in the sample over the several cyclic reactions 0 5 10 15 20 25 0246810 Number of cyclesSurface area (m2/g) after 1st carbonation

PAGE 130

130 A B C Figure 7-9. The SEM pictures for the sample (A and B) Fresh. (C and D) After 10th carbonation. (E and F) After 10th calcination (10 cycles).

PAGE 131

131 D E F Figure 7-9. Continued

PAGE 132

132 0 0.2 0.4 0.6 012345678910111213 Number of cycles (N)Carbonation conversion (X) Alumina (54 wt% CaO) Yttria (51 wt% CaO) Figure 7-10. The cyclic maximu m carbonation conversions of the samples using yttria and alumina as a substrate under the severe calcination condition at 850C and 20 wt % CO2

PAGE 133

133 Table 7-1. Mild and severe calcination conditions Carbonation Calcination Temperature Gas feed DurationTemperature Gas feed Duration Mild condition 750C 20 vol. % of CO2 in N2 20 min 750C Pure N2 10 min Severe condition 700C 20 vol. % of CO2 in N2 30 min Ramp 15C/min from 700C to 850C 20 vol. % of CO2 in N2 10 min

PAGE 134

134 CHAPTER 8 CONCLUSIONS AND FUTURE WORK 8.1 Conclusions This study began with the goal of thermodynami c feasibility investig ation of UT-3 cycle and the development of calcium oxide reactant wi th better characteristics and performance. The feasibility of each reaction in the UT-3 cycl e was examined via theoretical thermodynamic approaches in order to determine the optimal ope rating conditions for high reaction rate as well as high conversion. The effects of excess steam, temperature and pressure on the conversion and chemical compositions at the equilibrium stat e were investigated. The major findings were: Thermodynamically, the two hydrolysis react ions are unfavorable while the two bromination reactions are favorable in the UT-3 cycle. The excess steam enhanced the hydrolysis reac tions of calcium bromide and iron bromide significantly while no serious a dverse effect of the excess steam on the bromination was observed. The temperature for hydrolysis reaction of calcium bromide is limited by the melting temperature of 1000K for calcium bromide. The conversion of the hydrolysis reaction of cal cium bromide is expected to be complete by continuously removing the product gas, HBr, from the equilibrium state even at atmospheric pressure. Porous calcium oxide pellets for the UT-3 cycle were prepared and characterized experimentally. The effects of pore forming agen ts on the characteristics and performance of pellets were investigated. From the characterization and kinetic st udies, it was ascertained that the amount and type of additives had an importa nt effect on the pore volume and increasing the volume of pores greater than 5 m speeded up the hydrolysis rate of calcium bromide in the pellets. Subsequently, an inexpe nsive and straightforward calcium oxide immobilization process on yttria fibrous fabric was developed. The perf ormance was evaluated in cyclic bromination and hydrolysis reactions experimentally. Based on the e xperimental results in th e cyclic reactions, the

PAGE 135

135 calcium oxide dispersed and immobilized on the fibrous yttria fabric had continuous higher reactivity (~ 85%) in four bromination reacti ons and the rate of hydrolysis reaction was comparable to that of calcium oxide pellets. The thermodynamic efficiency of UT-3 cycle wa s investigated considering inert materials, heat recovery and incomplete conversion. The e ffect of heat recovery on the efficiency was considerable, which accounted for 20.4-30.2% e fficiency discrepancy. The inert materials accounted for 10.9 and 8.5% reductions of the effi ciency for the pellet and fabric reactants, respectively, without heat recovery. However, the influence of incomplete conversion was not significant. Also it was found that the fabric type reactant was so mewhat better than the pellet type reactant for thermal efficiency. The developed calcium oxide fabrics were us ed for carbon dioxide capture in coal or biomass gasification, SMR process and conventi onal coal power plants. A new type calcium oxide absorbent was fabricated on fibrous alum ina and the cyclic carbonation conversion was maintained over ten carbonation-calcination cycles under mild calcinations condition. Under the severe calcination condition, th e carbonation conversion of the calcium oxide sample using yttria fabric was maintained at 56% through 12 cycles while those of the sample using alumina dropped by about eight percent af ter 12 cycles from the maximum of 59%, possibly due to the formation of Ca12Al14O33 by the reaction between calcium oxide and alumina. 8.2 Recommendations for Future Work The newly developed calcium oxide reactant for the UT-3 cycle was tested in four cyclic bromination and hydrolysis reactions. In order to make the process commercially feasible, however, several thousands cycles are needed without severe degrad ation of the solid reactant. A new preparation method for iron oxide is also ne eded to be developed because the existing iron pellet is also made through an expensive and complicated process.

PAGE 136

136 The process efficiency should be es timated by including the hydrogen and oxygen separation work, pumping work, actual process pre ssure, excess water and realistic heat match in the calculations. Heat recovery a nd reduction of the inert material should be investigated in order to increase the efficiency. The excess steam was found to enhance hydrolys is reactions of calcium bromide and iron bromide significantly in this study. However, ther e may be economic penalty associated with the use of excess steam. It is recommended that eco nomic and technical considerations of the excess steam must be studied with regard to the co st and size of the building and the separation efficiency of the high temperature membrane. The prepared calcium oxide absorbent for hi gh temperature carbon diox ide capture showed continuous high reactivity in the cyclic ca rbonation-calcination cycl e even under severe calcinations condition. However, in order to obtain pure carbon dioxide for sequestration from the calcination reaction, cyclic experiments under the more severe calcina tion condition at higher temperature of at least 950C under 100% concen tration of carbon dioxide would be needed. Future work must also consider the economics of the calcium oxide absorbent as compared with other type of carbon di oxide absorbents.

PAGE 137

137 APPENDIX CYCLIC CONVERSION PROFILES (ALUMINA, SILICA, ZIRCONIA) Several ceramic materials such as alumina (Al2O3), silica (SiO2), zirconia (ZrO2) and yttria (Y2O3) were considered as a substrate of calcium oxide reactant for UT-3 cycle. Among them, the yttria fabric was selected as the substrate for the calcium oxide react ant through preliminarily experiments. The cyclic bromination and hydr olysis profiles of calcium oxide samples immobilized on alumina, silica and zirconia fabr ics are shown in Figure A-1, A-2 and A-3, respectively. It was observed that the performance of all samples severely degraded at the second cycle, which is possibly due to the formation of inert materials such as calcium aluminate, calcium silicate and calcium zirconate due to reactions between calcium oxide and the substrate materials. 0 0.1 0.2 0.3 0.4 0.5 0.6 0102030405060708090100110120130140150160 Time[min]Conversion, X 2nd cycle 1st cycle Figure A-1. Cyclic conversion pr ofiles of a calcium oxide samp le immobilized on an alumina fabric

PAGE 138

138 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 01020304050607080 Time[min]Conversion, X 2nd cycle 1st cycle Figure A-2. Cyclic conversion prof iles of a calcium oxide sample immobilized on a silica fabric 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0102030405060708090100110 Time[min]Conversion, X 2nd cycle 1st cycle Figure A-3. Cyclic conversion pr ofiles of a calcium oxide samp le immobilized on a zirconia fabric

PAGE 139

139 LIST OF REFERENCES Abanades, J. C., 2002. The maximum capture efficiency of CO2 using a carbonation/calcination cycle of CaO/CaCO3. Chemical Engineering Journal 90, 303-306. Abanades, J. C., Alvarez, D., 2003. C onversion limits in the reaction of CO2 with lime. Energy & Fuels 17, 308-315. Aihara, M., Umida, H., Tsutsumi, A., Yo shida, K., 1990. Kinetic study of UT-3 thermochemical hydrogen production proce ss. International Journal of Hydrogen Energy 15, 7-11. Aihara, M., Sakurai, M., Tsutsumi, A., Yoshid a, K., 1992. Reactivity improvement in the UT-3 thermochemical hydrogen production process. International Journal of Hydrogen Energy 17, 719-723. Aihara, M., Nagai, T., Matsushita, J., Negi shi, Y., Ohya, H., 2001. Development of porous solid reactant for thermal-ener gy storage and temperature upgrade using carbonation/decarbonation reaction. Applied Energy 69, 225-238. Amir, R., Sato, T., Yoko Yamamoto, K., Kabe, T., Kameyama, H., 1992. Design of solid reactants and reaction ki netics concerning the iron compounds in the UT-3 thermochemical cycle. International Journal of Hydrogen Energy 17, 783-788. Amir, R., Shiizaki, Yamamoto, K., Kabe, T., Kameyama, H., 1993. Design development of iron solid reactants in the UT-3 wa ter decomposition cycle based on ceramic support materials. International J ournal of Hydrogen Energy 18, 283-286. Aochi, A., Tadokoro, T., Yoshida, K., Kameyama, H., Nobue, M., Yamaguchi, T., 1989. Economical and technical evaluation of UT-3 thermochemical hydrogen production process for an industrial scale pl ant. International Journal of Hydrogen Energy 14, 421-429. Argonne National Laboratory, 2003. Basic re search needs for the hydrogen economy. Report on the Basic Energy Sciences Wo rkshop on Hydrogen Production, Storage, and Use, available at http://www.sc.doe.gov/bes/hydrogen.pdf Balasubramanian, B., Ortiz, A. L., Kaytakogl u, S., Harrison, D. P., 1999. Hydrogen from methane in a single-step process. Chem Eng Sci 54, 3543-3552. Barker, R., 1973. The reversib ility of the reaction CaCO3 CaO CO2. J. Appl. Chem. Biotechnol 23, 733-742. Beghi, G. E., 1986. A decade of research on thermochemical hydrogen at the joint research center, Ispra. Internationa l Journal of Hydrogen Energy 11, 761-771.

PAGE 140

140 Bennett, J. E., 1980. Electrodes for generation of hydr ogen and oxygen from seawater. International Journal of Hydrogen Energy 5, 401-408. Bhatia, S. K., Perlmutter, D. D., 1983. Effect of the product layer on the kinetics of the CO2-lime reaction. AIChE Journal 29, 79-86. Bilgen, E., Joels, R. K., 1985. An assessment of solar hydrogen production using the mark 13 hybrid process. International Journal of Hydrogen Energy 10, 143-155. Borgwardt, R. H., 1989. Calcium oxide sint ering in atmospheres containing water and carbon dioxide. Ind. Eng. Chem. Res. 28, 493-500. Bockris, J. O. M., 1975. Energy: the Solar Hydrogen Alternative John Wiley & Sons, New York. Brecher, L. E., Spewock, S., Warde, C. J ., 1977. The westinghouse sulfur cycle for the thermochemical decomposition of water. International Jo urnal of Hydrogen Energy 2, 7-15. Bromine Statistics and Info rmation, 2008. Available at http://minerals.usgs.gov/m inerals/pubs/commodity/bromine/mcs-2008-bromi.pdf Brown, L. C., Besenbruch, G. E., Lentsch, R. D., Schultz, K. R., Funk, J. F., Pickard, P. S., et al., 2003. High efficiency genera tion of hydrogen fuels using nuclear power. GA-A24285. Prepared under the Nuclear En ergy Research Initi ative Program for the US Department of Energy. Casper, M. S., ed., 1978. Hydrogen Manufacture by Electr olysis, Thermal Decomposition and Unusual Techniques Noyes Data Corporation. Chase, M. W. Jr., Davies, C. A., Downey, Jr., Fr urip, D. J., McDonald, R. A., Syverud, A. N., 1985. JANAF Thermochemical Tables 3rd ed., Journal of Physical and Chemical Refercence Data 14, Supplement No. 1. Cox, K. E., Williamson, K. D., 1977. Hydrogen: Its technology and implication, Vol. 1 Hydrogen Production Technology, CRC PRESS, Inc. Curran, G. P., Fink, C. E., Gorin, E., 1967. Carbon dioxide-acceptor gasification process: studies of acceptor properties. Adv. Chem. Ser. 69 141-165. Dobner, S., Sterns, L., Graff, R. A., S quires, A. M., 1977. Cyc lic calcinations and recarbonation of calcined dolomite. Ind. Eng. Chem., Process Des. Dev. 16, 479486. Doctor, R. D., Marshall, C. L., Wade, D. C., 2002. Hydrogen cycle employing calciumbromine and electrolysis. ACS Division of Fuel Chemistry, Preprints 47, 755-756.

PAGE 141

141 Ebrahim, M., Kawari, A., 2000. Pinch technology: an efficient tool for chemical-plant energy and capital-cost savi ng. Applied Energy 65, 45-49. FactSage, Web version, Thermo-chemical database software, Thermfact, GTT Technologies, Germany, Available at http//www.factsage.com/ Feng, B., Liu, W., Li, X., An H., 2006. Overcoming the problem of loss-in-capacity of calcium oxide in CO2 capture. Energy & Fuels 20, 2417-2420. Funk, J. E., Reinstrom, R. M., 1964. Final report energy depot electrolysis study. TID 20441(EDR3714), Vol. 2, Supplement A. Funk, J. E., Reinstrom, R. M., 1966. Energy requirements in the production of hydrogen from water. Industrial and Engin eering Chemistry Process Design and Development 5, 336-342. Funk, J. E., 2001. Thermochemical hydrogen pr oduction: past and present. International Journal of Hydrogen Energy 26, 185-190. Goel, N., Mirabel, S., Ingley, S., Goswam i, Y., 2003. Solar hydrogen production. Advanced in Solar Energy. Grasa, G., Gonzalez, B., Alonso, M., Abanades, J. C., 2007. Comparison of CaO-based synthetic CO2 sorbents under realistic calcina tion conditions. Energy & Fuels 21, 3560-3562. Gupta, H., Fan, L., 2002. Carbonation-calcina tion cycle using high reactivity calcium oxide for carbon dioxide sepa ration from flue gas. Ind. Eng. Chem. Res. 41, 40354042. Hanaoka, T., Yoshida, T., Fujimoto, S., Kamei, K., Harada, M., Suzuki, Y., Hatano, H., Yokoyama, S.-Y., Minowa, T., 2005. H ydrogen production from woody biomass by steam gasification using a CO2 sorbent. Biomass and Bioenergy 28, 63-68. Haueter, P, Moeller, S, Palumbo, R, St einfeld, A., 1999. The production of zinc by thermal dissociation of zinc oxide-solar chemical reactor design. Solar Energy 67, 161-167. Iodine Statistics and Information, 2008, Available at http://minerals.usgs.gov/minerals/pu bs/commodity/iodine/mcs-2008-iodin.pdf Kameyama, H., Yoshida, K., 1978. Br-Ca-Fe water-decomposition cycles for hydrogen production. Proc. of the 2nd World Hydr ogen Energy Conf., Zurich, Switzerland, 829-850. Kameyama, H., Yoshida, K., 1981. Reactor design for the UT-3 thermochemical hydrogen production process. International Journal of Hydrogen Energy 6, 567575.

PAGE 142

142 Kameyama, H., Tomino,Y., Sato, T., Amir, R ., Orihara, A., Aihara, M., Yoshida, K., 1989. Process simulation of MASCOT pl ant using the UT-3 thermochemical cycle for hydrogen production. Internati onal Journal of Hydrogen Energy 14, 323330. Lee, M. S., Goswami, Y., Hettinger, B., Vijayaraghavan, S., 2006. Preparation and characteristics of calcium oxide pellet s for UT-3 thermochemical cycle. Proceedings in 2006 ASME International Mechanical Engineering Congress and Exposition. Lee, M. S., Goswami, Y., Kothurkar, N., St efanakos, E. K., 2007. Fabrication of porous calcium oxide film for UT-3 thermo chemical hydrogen production cycle. Proceedings in Energy Sustainability 2007, ASME. Lemort, F., Lafon, C., Dedryvre, R., Gonbeau, D., 2006. Physicochemical and thermodynamic investigation of the UT-3 hydrogen production cycle: A new technological assessment. International Journal of Hydrogen Energy 31, 906-918. Li, Z.-S., Cai, N.-S., Huang, Y.-Y., Han, H.-J ., 2005. Synthesis, experimental studies, and analysis of a new calcium-based carbon dioxide absorbent. Energy & Fuels 19, 1447-1452. Li, Z.-S., Cai, N.-S., Huang, Y.-Y., 2006. Eff ect of preparation temperature on cyclic CO2 capture and multiple carbonation-calci nation cycles for a new ca-based CO2 sorbent. Ind. Eng. Chem. Res. 45, 1911-1917. Lin, S., Harada, M., Suzuki, Y., Hatano, H., 2002. Hydrogen production from coal by separating carbon dioxide during gasification. Fuel 81, 2079-2085. Lin, S.-Y., Suzuki, Y., Hatano, H., Harada, M., 2002. Developing an innovative method, HyPr-RING, to produce hydrogen from hydrocarbons. Energy Conservation and Management 43, 1283-1290. Linnhoff, B., Flower, J. R., 1978. Synthesis of heat exchange r networks: Part I and II. AIChE Journal 24, 633-642. Mahishi, M. R., Sadrameli, M. S., Vijayaraghavan, S., Gosw ami, D. Y., 2005. Hydrogen production from ethanol: A thermodynamic analysis of a novel sorbent enhanced gasification process. American Society of Mechanical Engineers, Advanced Energy Systems AES 45, 455-463. Meites, L., 1981. An Introduction to Chemical Equilibrium and Kinetics Pergamon Series in Analytical Chemistr y, Vol. 1, Pergamon Press. Mess, D., Sarofim, A. F., Longwell, J. P., 1999. Product layer diffusion during the reaction of calcium oxide with carbon dioxide. Energy & Fuels 13, 999-1005.

PAGE 143

143 Metz, B., Davidson, O., Coninck, H., Loos, M., Meyer, L. (Eds.), 2005. IPCC special report on carbon dioxide capture and storage Cambridge University Press. Morooka, S., Kim, S. S., Yan, S., Kusa kabe, K., Watanabe, M., 1996. Separation of hydrogen from an H2-H2O-HBr system with a SiO2 membrane formed in macropores of an -Alumina support tube. Intern ational Journal of Hydrogen Energy 21, 183-188. Nakajima, R., Kikuchi, R., Tsutsumi, A., 2000. Im provement of the Fe-pellet reactivity in the UT-3 thermochemical hydrogen produc tion cycle. Hydrogen energy progress XIII: Proceedings of the 13th World Hydrogen Energy Conference, Beijing, China. Nakayama, T., Yoshioka, H., Furutani, H., Kameyama, H., Yoshida, K., 1984. MASCOTa bench-scale plant fo r producing hydrogen by the UT-3 thermochemical decomposition cycle. Inte rnational Journal of Hydrogen Energy 9, 187-190. National Research Council and Nati onal Academy of Engineering, 2004. The Hydrogen Economy: Opportunities, Costs, Barriers, and R&D Needs The National Academies Press. National Resources Canada, 2003. Pinch analysis : for the efficient use of energy, water & hydrogen, CANMET Energy Technology Ce ntre-Varennes, available at http://cetc-varennes.nrcan.gc.ca/fi chier.php/codectec/E n/2003-140/2003-140e.pdf Ogawa, M., Nishihara, T., 2004. Present stat us of energy in Japan and HTTR project. Nuclear Engineering and Design 233, 5-10. Ohya, H., Nakajima, H., Togami, N., Aiha ra, M., Negishi, Y., 1994. Separation of hydrogen from thermochemical processes using zirconia-silica composite membrane. International Journal of Hydrogen Energy 97, 91-98. Ohya, H., Nakajima, H., Togami, N., Ohashi, H., Aihara, M., Tanisho, S., Negishi, Y., 1997. Hydrogen purification using zirc onia-silica composite membranes for thermochemical process. International Journal of Hydrogen Energy 22, 509-515. Ohta, T., ed., 1979. Solar-Hydrogen Energy Systems Oxford and New York, Pergamon Press. Onuki, K., Inagaki, Y., Hino, R., Tachiban a, Y., 2005. Research and development on nuclear hydrogen production using HTGR at JAERI. Progress in Nuclear Energy 47, 496-503. Riis, T., Hagen, E. F., Vie, P. J. S ., Ulleberg, O., 2006. Hydrogen production R&D: priorities and gaps. International Energy Agency (IEA). Rosen, M. A., 1996. Thermodynamic comparison of hydrogen production process. International Journal of Hydrogen Energy 21, 349-365.

PAGE 144

144 Sakurai, M., Aihara, M., Miyake N., Tsutsumi, A., Yoshida, K., 1992. Test of one-loop flow scheme for the UT-3 thermochemical hydrogen production process. International Journal of Hydrogen Energy 17, 587-592. Sakurai, M., Tsutsumi, A., Yoshida, K., 1995. Improvement of Ca-pellet reactivity in UT-3 thermo-chemical hydrogen production cycle. Internati onal Journal of Hydrogen Energy 20, 297-301. Sakurai, M., Bilgen, E., Tsutsumi, A ., Yoshida, K., 1996a. Adiabatic UT-3 thermochemical process for hydrogen pr oduction. International Journal of Hydrogen Energy 21, 865-870. Sakurai, M., Bilgen, E., Tsutsumi, A., Yosh ida, K., 1996b. Solar UT-3 thermochemical cycle for hydrogen production. Solar Energy 57, 51-58. Sakurai, M., Ogiwara, J., Kameyama, H ., 2006. Reactivity improvement of Fecompounds for the UT-3 thermochemical hydrogen production process. Journal of Chemical Engineering of Japan 39, 553-558. Satrio, J. A., Shanks, B. H., Wheelock, T. D., 2004. Application of combined catalyst/sorbent on hydrogen generation from biomass gasification. AIChE Annual Meeting Conference Proceedings, November 7-12, Austin, TX. Satrio, J. A., Shanks, B. H., Wheelock, T. D., 2005. Development of a novel combined catalyst and sorbent for hydrocarbon re forming. Industrial & Engineering Chemistry Research 44 3901-3911. Satrio, J. A.,Shanks, B. H., Wheelock, T. D ., 2007. A combined catal ysis and sorbent for enhancing hydrogen production from coal or biomass. Energy & Fuels 21, 322326. Shimizu, T., Hirama, T., Hosoda, H., Kitano, K., Inagaki, M., Tejima, K., 1999. A twin fluid-bed reactor for removal of CO2 from combustion proce sses. Trans. IChemE 77 (Part A), 62-68. Simpson, M. F., Utgikar, V., Sachdev, P., McGrady, C., 2007. A novel method for producing hydrogen based on the CaBr cycl e. International Journal of Hydrogen Energy 32, 505-509. Silaban, A., Harrison, D. P., 1995. Hi gh temperature capture of carbon dioxide: Characteristics of the reversible reaction between CaO(s) and CO2 (g). Chem. Eng. Comm. 137, 177-190. Silaban, A., Narcida, M., Harrison, D. P., 1996. Characteristics of th e reversible reaction between CO2(g) and calcined dolomite. Ch em. Eng. Comm. 146, 149-162. S owinski, G., 2006. Some technical issu es of zero-emission coal technology. International Journal of Hydrogen Energy 31, 1091-1102.

PAGE 145

145 Spath, P. L., Mann, M. K., 2001. Life Cy cle Assessment of Hydrogen Production via National Gas Steam Reforming. Nati onal Renewable Energy Laboratory. Stanmore, B.R., Gilot, P., 2005. Review-cal cination and carbonation of limestone during thermal cycling for CO2 sequestration. Fuel Processing Technology 86, 17071743. Steinfeld, A., 2002. Solar hydrogen produc tion via a two-step water-splitting thermochemical cycle based on Zn/ZnO re dox reactions. International Journal of Hydrogen Energy 27, 611-619. Stiegel, G. J., Ramezan, M., 2006. Hydrogen from coal gasification: an economical pathway to a sustainable energy futu re. International Journal of Coal Geology 65, 173-190. Teo, E. D., Brandon, N. P., Vos, E., Kram er, G. J., 2005. A critical pathway energy efficiency analysis of the thermochemical UT-3 cycle. Intern ational Journal of Hydrogen Energy 30, 559-564. T-Raissi, A., 2005. Technoeconomic analysis of area II hydrogen production. DOE final report to U.S. DOE under DEFC3600GO10603. U. S. DOE, 2002. A national Vision of am ericas transition to a hydrogen economy-to 2030 and beyond. United States Depart ment of Energy, available at http://www1.eere.energy.gov/hydrogena ndfuelcells/pdfs/vision_doc.pdf Velzen, D. V., Langenkamp, H., Schtz, G., La londe, Flamm, D. J., Fiebelmann, P., 1980. Development and design of a continuous laboratory-scale plant for hydrogen production by the Mark-13 cycle. International Journa l of Hydrogen Energy 5, 131-139. Vitart, X., Carles, P., Anzieu, P., 2008. A gene ral survey of the potential and main issues associated with the sulfur-iodine ther mochemical cycle for hydrogen production using unclear heat. Progress in Nuclear Energy 50, 402-410. Webber, M. E., 2007. The water intensity of the transitiona l hydrogen economy. Environmental Research Letters 2, 034007 (7pp). Weidenkaff, A., Reller, A. W., Wokaun, A ., Steinfeld, A., 2000. Thermogravimetric analysis of the ZnO/Zn water splitting cycle. Thermochimica Acta 359, 69-75. Wendt, H., ed. 1990. Electrochemical Hydrogen T echnologies: Electrochemical Production and Combustion of Hydrogen Elsevier, Amsterdam. Williams, L.O., 1975. Electrolysis of sea-water. Hydrogen Energy (T.N. Veziroglu, Ed.), Plenum, New York, 417.

PAGE 146

146 Wu, S. F., Li, Q. H., Kim, J. N., Yi K. B., 2008. Properties of a nano CaO/Al2O3 CO2 sorbent. Ind. Eng. Chem. Res 47, 180-184. Yoshida, K., Kameyama, H., Aochi, T., Nobue, M., Aihara, M., Amir, R., Kondo, H., Sato, T., Tadokoro, Y., Yamaguchi, T., Sakai, N., 1990. A simulation study of the UT-3 thermochemical hydrogen production process. International Journal of Hydrogen Energy 15, 171-178.

PAGE 147

147 BIOGRAPHICAL SKETCH Man Su Lee was born in 1974 in Gokseong, a small town in South Korea. He was raised primarily in Seoul, South Korea. He received his Bachelor of Engin eering and Master of Engineering in mechanical engineering from Ch ung-Ang University, Seoul, South Korea, in 1997 and 1999, respectively. He worked for R&D Institute of Unisem, Osan, South Korea, as a development researcher for 3 years. Then he mo ved to Daewoo Electronics Co. and worked there for 2 years as a researcher. In August 2004 he en rolled in the doctoral program in mechanical engineering at University of Florida. He joined Dr. Go swamis Solar Energy and Energy Conservation Laboratory and has been worki ng on the thermochemical hydrogen production and high temperature carbon dioxide ca pture. Upon graduation, he w ould like to pursue a career in the field of clean energy research.