-650/2-73-048a
\
ember 1973
Environmental Protection Technology Series
-------
EPA-650/2-73-048a
EVALUATION
OF THE HUIDIZED-BED
COMBUSTION PROCESS
VOLUME I -
PRESSURIZED FLUIDIZED-BED COMBUSTION
PROCESS DEVELOPMENT AND EVALUATION
by
D.L. Keairns, D.H. Archer, J.R. Hamm,
R.A. Newby, E.P. O'Neill, J.R. Smith, and
W.C. Yang
Westinghouse Research Laboratories
Pittsburgh, Pennsylvania 15235
Contract No. 68-02-0217
ROAP No. 21ADB-09
Program Element No. 1AB013
EPA Project Officer: P.P. Turner
Control Systems Laboratory
National Environmental Research Center
Research Triangle Park, North Carolina 27711
Prepared for
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
WASHINGTON, D.C. 20460
December 1973
Environmental Proteetioa Agency
Region Vs Library
230 Soutft Dearboz-a Street
Chicago 8 mi-sols 60604
-------
This report has been reviewed by the Environmental Protection Agency and
approved for publication. Approval does not signify that the contents
necessarily reflect the views and policies of the Agency, nor does
mention of trade names or commercial products constitute endorsement
or recommendation for use.
mviHOHMEHTAL PROTECTION AGEJTC3t
-------
ACKNOWLEDGMENTS
The results, conclusions, and recommendations presented in
this volume represent the combined work and thought of many persons at
Westinghouse and the Office of Research and Development (ORD). Other
ORD contractors have freely shared with us their ideas and the results of
their research and development effort.
In particular, we want here to express our high regard for
and acknowledge the contribution of personnel at Westinghouse Research
Laboratories and the personnel at ORD who conceived the overall fluidized
bed combustion boiler effort and who have defined, monitored, and supported
the efforts of Westinghouse and others on the program. Mr. P. P. Turner,
Chief of the Advanced Process Section, has served as EPA project officer
on our work. Numerous enlightening and helpful discussions have been
held with Mr. Turner; with section members D. Bruce Henschel and Sam
Rakes; and with R. P. Hangebrauck, Chief of the Demonstration Projects
Branch. Personnel from the Westinghouse Research Laboratories have made
significant contributions. Mr. E. J. Vidt participated in the evaluation.
Mr. R. E. Brinza and Mr. C. Spangler assisted in the collection of data.
Drs. L. E. Brecher and C. R. Wolfe participated in the design and
development of the pressurized thermogravimetric analysis system.
Mr. W. F. Kittle contributed to each stage of the design, development
and use of the pressurized thermogravimetric analysis system. Westing-
house Power Generation Systems, Gas Turbine Systems Division, Heat
Transfer Division, Power Generation Services, and Instrumentation
Systems Division personnel were consulted and participated in the
evaluation.
We gratefully acknowledge the work of our secretary, Ms. Sylvia
Nalepa. We are also grateful to Ms. Nancy Berkowitz who coordinated
the final editing and production work and to Ms. Deborah Conrad for
typing the editorial material.
-------
PREFACE
The Office of Research and Development (ORD) of the United
States Environmental Protection Agency has organized and is sponsoring
a fluidized bed fuel processing program. Its purpose is to develop and
demonstrate new methods for utilizing fossil fuels — particularly
coal and oil — in utility power plants. These methods should:
• Meet environmental goals for S09, NO , ash, smoke emissions,
£. X
and wastes
• Compete economically with alternative means for meeting
these abatement goals.
Westinghouse Research, under contract to the Office of Research
and Development (ORD) of the Environmental Protection Agency, is carrying
out a study to evaluate and develop fluidized bed combustion and oil
gasification in air pollution abatement. The goals of this work are
• To identify which fluidized processes might be economically
employed in utility power plants or industrial boilers to
reduce S0_, particulate, and NO emissions
^> X
• To assist in planning and implementing a program to
develop the fluidized bed systems deemed effective in
air pollution control and economical in steam/power production.
Tasks in the evaluation of fluidized bed combustion set forth
by EPA which have been completed at Westinghouse under a previous
contract are
• To search the technical and patent literature in fluidized
bed combustion, to canvass commercial organizations with
expertise pertaining to this field, and to survey the
market for industrial boilers and utility power systems
-------
• To design a fluidized bed industrial boiler and two
fluidized bed utility power systems — one 300 MW
capacity, the other 600 MW; and to provide performance
and cost projections for the equipment designed
• To provide technical consultation and assistance on the
ORD fluidized bed fuel processing program, including
both combustion and gasification
• To conceptualize a development plant which will prove
the configurational or operating features of the fluidized
bed boiler designs not confirmed by prior art or experience
• To assess the effectiveness and economics of an
atmospheric-pressure, fluidized bed oil gasification-
combustion system and to aid in planning for a demonstration
installation of such a system.
The results of these surveys, designs, evaluations, and
comparisons were published in a three-volume report, "Evaluation of
the Fluidized Bed Combustion Process," in November 1971 under contract
No. CPA 70-9.
These results provided the basis for the work performed from
July 1971 to May 1973 which is described in this four-volume report.
Tasks in the evaluation of the fluidized bed combustion process set
forth by EPA under contract 68-02-0217 have focused on the development
of pressurized fluidized bed combustion for power generation and flui-
dized bed oil gasification for power generation. These tasks, which are
presented here, have included:
• Extensive process evaluation studies of the pressurized
fluidized bed combustion system for power generation.
These investigations predict the sensitivity of operating
and design parameters selected for the base power plant
design on plant economics; provide additional experimental
data on sulfur removal and sorbent regeneration using
vi
-------
limestones and dolomites; project economics and performance
for various sulfur removal systems, both regeneration/sulfur
recovery and once-through sorbent; establish a plant operation
and control philosophy and evaluate alternative pressurized
fluid bed combustion concepts, both economics and performance.
Preparation of preliminary plans and a cost estimate for a
30 MW (equivalent) pressurized fluidized bed combustion
boiler development plant. The design provides sufficient
detail to locate a suitable plant site and to obtain a fixed
cost bid for the preparation of detailed plans. The plant
will provide the capability for studying the remaining
technical problems effectively in order to achieve the
greatest potential for reducing emissions and for generating
economical electrical energy.
Identification of a project team to demonstrate fluidized
bed oil gasification/desulfurization for power generation.
A cooperating utility — New England Electric System —
has been identified to carry out a 50 MW demonstration
plant program. Further process evaluation has been
carried out and experimental data obtained on sulfur
removal and spent stone disposal.
Evaluation of pressurized oil gasification for combined
cycle power generation. Oil gasification process concepts
and options have been reviewed, material and energy balances
projected, performance projected,and capital and energy
costs estimated.
Provision of technical consultation and assistance on the
ORD fluidized bed fuel processing program, including both
combustion and gasification processes. Technical and
economic comparisons have been carried out on various
fluidized bed fuel processing systems and various conventional
means of steam/power generation.
VII
-------
This volume (Volume I) contains new data on pressurized fluid
bed combustion and an evaluation of the current technology state.
Volume II contains the appendices to this volume. Volume III presents
the preliminary design of a 30 MW (equivalent) pressurized fluid bed
boiler development plant. Volume IV includes the work on fluidized bed
oil gasification/desulfurization.
viii
-------
VOLUME I
PRESSURIZED FLUID BED COMBUSTION PROCESS DEVELOPMENT EVALUATION
TABLE OF CONTENTS
Page
1. SUMMARY 1
1.1 Economic Comparison 8
1.2 Environmental Impact 11
1.3 Energy Resources 11
1.4 Technical Uncertainties 11
1.5 Development Status 13
U..6 Comparison with Alternative Technology 16
2. ECONOMIC SENSITIVITY 17
2.1 Basis for Sensitivity Analysis 18
2.2 Operating Conditions 28
2.2.1 Bed Temperature 28
2.2.2 Fluidizing Velocity 34
2.2.3 Excess Air 44
2.2.4 Operating Pressure 51
Y/2.3 Boiler Design 53
2.3.1 Heat Transfer Surface Configuration 53
2.3.2 Heat Transfer Coefficient 68
2.3.3 Tube Materials 75
2.3.4 Module Capacity 75
i/2.4 Particulate Removal System Economics 78
2.4.1 Range of Dust Loading and Particle Size Distribution 78
2.4.2 Gas-Turbine Specifications 78
2.4.3 Effect of Dust Loading and Particle Size on Cost 82
2.4.4 Effect of Gas Flow Rate on Cost 85
V2.5 Power Plant 93
2.5.1 Gas-Turbine Inlet Temperature 93
2.5.2 Steam Temperature 97
2.6 Assessment 97
2.7 Conclusions 101
ix
-------
Page
«
3. LABORATORY SUPPORT STUDIES 103
3.1 Introduction 103
3.2 Apparatus 106
3.2.1 Modification of the TG Apparatus 108
3.2.2 Operation 110
3.2.3 System Performance 113
3.3 Process Variables 114
3.4 Previous Laboratory Studies 116
3.5 Summary of the Results 116
3.5.1 Pressure 116
3.5.2 Sulfur Dioxide Concentration 117
3.5.3 Temperature 117
3.5.4 Agreement with Previously Reported Data 117
3.5.5 Stone Type 117
3.5.6 Particle Size 119
3.5.7 Oxygen Concentration 119
3.5.8 Stone Pretreatment 119
3.6 Experimental Program 119
3.6.1 Sulfation of Dolomites 119
3.6.2 Atmospheric Pressure Tests 122
3.6.3 Dolomite 1337 at 10 Atmospheres 128
3.6.4 Sulfation of Tymochtee Dolomite at 10 Atmospheres 130
3.6.5 Effect of Low Sulfur Dioxide Levels 130
3.7 Theory 136
3.8 Kinetics of Dolomite Sulfation 137
3.9 Qualitative Comparison with Fluidized Bed Results 139
3.10 Quantitative Comparison with Fluidized Bed Results 142
3.11 Comparison of NCB Data and Westinghouse Data 144
3.12 Regeneration 147
3.13 Previous Regeneration Studies 148
3.14 Conclusions from the Regeneration Experiments 150
3.15 Two-step Regeneration Experiments 151
3.16 Exploratory Runs 157
-------
Page
3.17 Tests with Powdered Calcium Sulfide 157
3.18 Tests with Sulfided Dolomite 158
3.19 Current and Future Experiments 160
3.20 Conclusions 163
4. REGENERATION SYSTEMS ASSESSMENT 165
4.1 Introduction and Scope of Investigation 165
4.2 Regeneration Process Concepts 166
4.2.1 One-Step High-Pressure and Low-Pressure 166
Regeneration Concepts
4.2.2 Two-Step Regeneration Concept 169
4.2.3 Constant Load Concept 171
4.2.4 Once-through Operation 171
4.2.5 Wellman-Lord Stack Gas Cleaning Process 171
4.2.6 Alternative Concepts 173
4.3 Process Options and Variables 173
4.4 Base Case Designs 175
4.5 Economic Results for Base Case Designs 178
4.5.1 Process Capital Investment 178
4.5.2 Plant Energy Costs 178
4.5.3 Constant Load Concept Economics 180
4.5.4 Effects of Boiler Conditions 180
4.5.5 Process Cost Comparison 185
V4.6 Process Performance 188
4.6.1 Plant Heat Rate 188
4.6.2 Temperature Control 188
4.6.3 Process Turndown 190
4.6.4 Process Environmental Comparison 191
4.7 Assessment 193
4.7.1 Economic Factors 193
4.7.2 Environmental Factors 196
4.7.3 Base Design Feasibility 202
4.8 Combustion of Low-Sulfur Coals 208
4.9 Conclusions 210
-------
Page*
5. PLANT OPERATION AND CONTROL 213
5.1 Introduction 213
5.1.1 Plant Criteria 213
5.1.2 Control System Requirements 214
5.1.3 Pressurized Fluid Bed Boiler Power Plant Concept 214
5.2 Unit Start-up 221
5.2.1 Introduction 221
5.2.2 First Module Start-up 234
5.2.3 Second Module Start-up 236
5.2.4 Third Module Start-up 237
5.2.5 Regeneration System Start-up 237
5.2.6 Auxiliary Fuel Storage Requirements 239
5.3 On-Line Loading 239
5.3.1 Introduction 239
5.3.2 Sequential Loading 241
5.3.3 Predictive Loading 243
5.3.4 Rapid Loading 244
5.3.5 Loading Events Diagram 244
5.3.6 Dynamic Analysis 247
5.4 Unit Shutdown 249
5.4.1 Introduction 249
5.4.2 First Boiler Module Shutdown 249
5.4.3 Second Boiler Module Shutdown 254
5.4.4 Shutdown of Fourth Module 255
5.4.5 Emergency Shutdown 256
5.4.6 Dolomite Sorbent Regeneration System Shutdown 258
5.5 Trips 259
5.6 Runback, Run-ups, and Limits 263
5.7 Total Plant Control System 264
5.7.1 Introduction 264
5.7.2 Plant Master Control System 269
5.7.3 Steam Turbine 270
5.7.4 Boiler Feed Pumps 271
xii
-------
5.7.5 Gas Turbine 272
5.7.6 Flash Tank 277
5.7.7 Boiler Module 278
5.7.8 Boiler By-pass 280
5.7.9 Regeneration System 281
5.7.10 Stone Feed System 282
5.7.11 H2S Generator Control 284
5.7.12 CaSO. Reducer Vessel 285
4
5.8 Conclusions and Comments 286
5.8.1 Conclusions 286
5.8.2 Comments 287
ALTERNATIVE PRESSURIZED FLUID BED BOILER CONCEPTS 289
6.1 Fluidized Bed Adiabatic Combustor Combined Cycle Power Plant 289
6.1.1 Adiabatic Combustor Concept 289
6.1.2 Cycle Performance 290
6.1.3 Plant Turndown 293
6.1.4 Environmental Considerations 295
6.1.5 Fuel Processing Equipment 295
6.1.6 Power Plant Cost 300
6.1.7 Conclusions 300
6.2 Recirculating Bed Boiler Design 310
6.2.1 Deep Recirculating Fluidized Bed Boiler Concept 314
6.2.2 Cold Model Studies 316
6.2.3 Bed Tube Heat Transfer Coefficient 321
6.2.4 Conceptual Recirculating Bed Boiler Design 323
6.2.5 Boiler Operation and Performance 325
6.2.6 Economics 330
6.2.7 Conclusions 333
335
7. REFERENCES
-------
LIST OF TABLES
Page
1. Boiler Design 5
2. Economic Comparisons/600 MW Plants 10
3. Problem Areas 12
4. Pressurized Fluid Bed Boiler Development Studies 14
5. Pressurized Fluidized Bed Boiler Power Plant Operating 22
and Design Conditions
6. Cost of a 318 MW Pressurized Fluid Bed Boiler 24
7. Equipment Costs for Pressurized Fluidized Bed Boiler Power 25
Plant (635 MW)
8. Energy Costs for Pressurized Fluidized Bed Boiler Power 27
Plant (635 MW)
9. Efficiency Calculations at Variable Airflow Rates 49
10. Cases Evaluated for Determining Effect of Dust Loading 81
and Particle Size Distribution
11. Dust Loadings Leaving the Secondary Cyclones for Different 83
Cases Evaluated
12. Particulate Removal Clean-up Cost 84
13. Selection of First-Stage Cyclones at Different Gas Flow Rates 86
14. Selection of Second-Stage Collectors at Different Gas Flow 87
Rates
15. Performance of Pressurized Fluid Bed Boiler Plant as a 96
Function of a Gas-Turbine Inlet Temperature
16. Sensitivity Analysis Summary 98
17. Options Afforded by Calcium Carbonate/Sulfur Cycle 105
Basic Reactions
18. Pressurized TG Systems 107
19. Rate of Sulfation of Calcined Dolomite 1337, at 5% Calcium 127
Utilization, 1562°F
20. The Effect of Temperature on the Time Required to Achieve 132
a Given Level of Calcium Utilization
21. The Rate of Sulfation of Tymochtee Dolomite at Low SO- 135
Concentrations
xv
-------
LIST OF TABLES (Continued)
Page
22. Calcium Utilization in Fluidized Bed Experiments 140
National Coal Board Results with Dolomite 1337
23. Design Conditions for the Adiabatic Combustor Single 144
Bed, Design I
24. Experiments Using Tymochtee Dolomite 145
25. Experiments with Dolomite 1337 146
26. Regeneration Processes 149
27. Two-Step Regeneration at 10 Atmospheres 153
28. Two-Step Regeneration at 10 Atmospheres 154
29. Design Assumptions 176
30. Capital Investment for Regeneration Processes - Base Designs 179
31. Energy Cost for Power Plant - Base Designs 181
32. Credit from Sulfur Sales 181
33. Capital Investment for Constant Load Concept 182
34. Energy Cost for Constant Load Concept 182
35. Effect of Calcination Boiler Conditions on Two-Step 184
Regeneration Process
36. Comparison of Processes - Environmental Impacts 192
37. General Comparison 195
38. Comparison of Regeneration with Once-through System 199
39. Comparison of Regeneration with Stack Gas Cleaning System 200
40. Comparison of Stack Gas Cleaning on Conventional and FBB 201
Plants
41. Assessment of Base Designs 203
42. Low-Sulfur Coal Combustion - Investment 209
43. Low-Sulfur Coal Combustion - Energy Cost 209
44. Process Conditions and Performance 292
45. Environmental Impact 296
46. Design Basis 298
47. Adiabatic Combustor Designs 299
48. Particulate Removal System (200 MW) 301
49. Power Plant Equipment for Fluid Bed Adiabatic Combustor 303
600 MW
xvi
-------
LIST OF TABLES (Continued)
Page
50. Coal-Fired Power Plant Equipment Cost Comparison 304
51. Fluidized Bed Adiabatic Combustor Combined Cycle Power Plant 305
Equipment Cost Breakdown Based on Fluidized Bed Boiler Plant
Costs
52. Fluidized Bed Adiabatic Combustor Combined Cycle Power Plant 308
Equipment Cost Breakdown (Based on Modified PACE Plant)
53. Power Plant Capital Costs - 600 MW Plant 309
54. Design Parameters and Operating Conditions for the 318 MW 326
Fluidized Bed Boiler
55. Physical Dimensions of the Combustion Modules 327
56. Comparison of Heat Transfer Surface Requirements and Boiler 332
Costs
xvii
-------
LIST OF FIGURES
Page
1. Pressurized Fluidized Bed Boiler Power Plant 2
2. Pressurized Fluidized Bed Steam Generator Module 3
3. Boiler Cross-Section (Superheater) 4
4. Electrical Energy Costs 9
5. 30 MW Pressurized Fluid Bed Boiler Demonstration Facility 15
6. Pressurized Fluid Bed Boiler Power System 19
7. Pressurized Fluidized Bed Steam Generator for Combined 20
Cycle Plant
8. Steam Generator Cost Breakdown for the W-FW Basic Design 26
9. Effect of Bed Temperature on Power Generation 29
10. Heat Transfer Surface Requirement for One Module 29
11. Bed Depth Requirement at Different Bed Temperatures 30
12. Effect of Bed Temperature on Total Module Height 30
13. Effect of Bed Temperature on the Steam Generator Cost of 32
318 MW Plant
14. Dependence of the Steam Generator (318 MW) Cost on the 33
Maximum Allowable Bed Depth
15. Change in Net Plant Power Output Due to Change in Pressure 35
Drop across the Bed
16. Dependence of Bed Depth on the Fluidizing Velocity, Bed 37
Temperature 1750°F
17. Dependence of Bed Depth on the Fluidizing Velocity, Bed 37
Temperature 1636°F
18. Dependence of Bed Depth on the Fluidizing Velocity, Bed 37
Temperature 1522°F
19. Dependence of Bed Depth on the Fluidizing Velocity, Bed 37
Temperature 1407°F
20. Dependence of Module Diameter and Module Height on the 38
Bed Area
21. Shell Cost (4 Modules) at Basic Design Height 39
22. Dependence of the Steam Generator Cost on the Fluidizing 40
Velocity
xix
-------
LIST OF FIGURES (Continued)
Page
23. Dependence of the Shell Cost on the Fluidizing Velocity 42
24. Increase in Module Diameter and Decrease in Module Height 45
at Different Airflow Rates
25. Bed Depth Requirement at Different Airflow Rate 47
26. Cost Reduction in Heat Transfer Surface and Pressure Shell 48
at Different Airflow Rates
27. Cost Escalation in Gas-Turbine Equipment and Cost Reduction 48
in Steam-Turbine Equipment at Different Airflow Rates
28. Plant Heat Rate Versus Compressor Pressure Ratio 54
29. Tube Arrangement in Basic Design 55
30. Definition for Different Tube Arrangements 56
31. Bed Depth Multiplier for Preevaporator, Superheater, and 59
Reheater (tube diameter 1/2 inch O.D.)
32. Bed Depth Multiplier for Preevaporator, Superheater, and 59
Reheater (tube diameter 1 inch O.D.)
33. Bed Depth Multiplier for Preevaporator, Superheater, and 59
Reheater (tube diameter 1-1/2 inch O.D.)
34. Bed Depth Multiplier for Preevaporator, Superheater, and 59
Reheater (tube diameter 2 inches O.D.)
35. Bed Depth Multiplier for Preevaporator, Superheater, and 59
Reheater (tube diameter 3 inches O.D.)
36. Change of the Steam Generator Cost with Heat Transfer 62
Surface per Unit Bed Volume
37. Dependence of the Steam Generator (318 MW) Cost on the 64
Maximum Allowable Bed Depth (bed temperature 1750°F)
38. Dependence of the Steam Generator (318 MW) Cost on the 65
Maximum Allowable Bed Depth (bed temperature 1636°F)
39. Dependence of the Steam Generator (318 MW) Cost on the 65
Maximum Allowable Bed Depth (bed temperature 1572°F)
40. Dependence of the Steam Generator (318 MW) Cost on the 65
Maximum Allowable Bed Depth (bed temperature 1407°F)
41. Change of Bed-Tube Heat Transfer Coefficient with Pitch/ 67
Diameter Ratio
42. Change of Heat Transfer Coefficient with Tube Spacing 67
43. Heat Transfer Surface and Bed Depth Multiplier for 69
Preevaporator, Superheater, and Reheater
xx
-------
LIST OF FIGURES (Continued)
Page
44. Effect of Tube Pitch/Diameter Ratio on the Steam Generator 71
Cost
45. Effect of Tube Pitch/Diameter Ratio on the Steam Generator 71
Cost
46. Effect of Bed-Tube Heat Transfer Coefficient on the Steam 72
Generator Cost (bed temperature 1750°F)
47. Effect of Bed-Tube Heat Transfer Coefficient on the Steam 74
Generator Cost (bed temperature 1636°F)
48. Dependence of the Steam Generator Cost on the Plant Size 77
(bed temperature 1750°F)
49. Dependence of the Steam Generator Cost on the Plant Size 77
(bed temperature 1636°F)
50. Different Combinations of Elutriation Rates for a Dust 79
Loading Equal to that of ®-FW Basic Design
51. Different Combinations of Elutriation Rates for a Dust 79
Loading Double that of ©-FW Basic Design
52. Different Combinations of Elutriation Rates for a Dust 79
Loading Triple that of ©-FW Basic Design
53. Flow Diagram for Particulate Removal System 80
54. Flow Chart for Group 1 - Case 4 88
55. Flow Chart for Group 1 - Case 5 88
56. Particle Size Distribution for Different Gas Streams 89
57. Increase in First-Stage Cyclone Cost Due to Increase in 90
Gas Flow Rate (for 318 MW plant)
58. Increase in Second-Stage Cyclone Cost Due to Increase in 90
Gas Flow Rate (for 318 MW plant)
59. Cost Increments for Particulate Removal System at Different 92
Gas Flow Rates
60. Pressurized Fluidized Bed Combustion Power Plant with 95
Secondary Combustor
61. The Calcium Carbonate/Sulfur Cycle Basic Reactions 104
62. The DuPont 950 Thermogravimetric Balance 109
63. Diagram of the TG System 111
64. The Pressurized TG Apparatus 112
65. Temperature Effect on Rate for 1337 Dolomite Sulfation 118
66. Sulfation of Calcined Dolomite 120
xxi
-------
LIST OF FIGURES (Continued)
Page
67. The Effect of Calcination on Sulfation 121
68. Sulfation of Tymochtee Dolomite Calcined TG 90B 123
69. Effect of Sample Size on Sulfation of Calcined Dolomite 124
1337 in the TG
70. Sulfation of Calcined Dolomite 1337 126
71. Sulfation of Calcined Dolomite 1337 126
72. Effect of SO Concentration on the "Slow" Sulfation Regime 129
73. High CaO Utilization in the Sulfation of Calcined Dolomite 129
74. Tymochtee Dolomite Sulfation under Pressure: The Effect of 131
Temperature
75. Sulfation of Calcined Tymochtee Dolomite at Low S02 131
Concentration
76. Sulfation of Calcined Tymochtee Dolomite 133
77. Data Used in Koppel's Model for Sulfation 143
78. Regeneration of Dolomite 152
79. Reduction of Sulfated Dolomite 156
80. Regeneration of Carbonate from Sulfided Dolomite 159
81. Regeneration of Half-Calcined Dolomite 161
82. One-Step Regeneration Process Elements 168
83. Two-Step Regeneration Process Elements 170
84. Constant Load Concept 172
85. Plant Energy Cost (Disposal before Regeneration) 186
86. Plant Energy Cost (Disposal after Regeneration) 187
87. Plant Heat Rate with Regeneration Alternative 189
88. Comparison of Fluid Bed Power Plant with Conventional 194
Power Plant
89. General Comparison of Energy Costs 197
90. Effect of Capacity Factor 197
91. Fluidized Bed Combustor with Low-Sulfur Coal 198
92. 635 MW Fluidized Bed Boiler Combined Cycle Plant 219
93. Steam and Water Cycle 223
94. Chemical Recovery Plant 225
95A. First Boiler Module Start-up 227
xxii
-------
LIST OF FIGURES (Continued)
Page
95B. First Boiler Module Start-up 229
96A. Second Boiler Module Start-up 231
96B. Second Boiler Module Start-up 233
97. Sequential Loading Scheme 242
98. Module Operating Range 245
99. Events Diagram — Cold Start 246
100. Bed Temperature Change as a Function of Fuel Feed Policy 248
101. Load Response as a Function of Fuel Feed Policy 248
102. Shutdown of First Boiler Module 251
103. Trip Chart 261
104A. Control System Functions 265
104B. Control System Functions 267
105. Flow Diagrams for Plant Start-up 276
106. Coal Fired PACE Plant 291
107. Turndown Capabilities of Coal-Fired PACE Plant with 294
Unfired Heat Recovery Boiler
108. Adiabatic Combustor Designs 297
109. 200 MW Fluidized Bed Adiabatic Combustor Combined Cycle 302
Plant — Gas Piping from Particulate Collectors
110. Recirculation Bed Concept 311
111. Deep Recirculating Fluidized Bed Boiler 315
112. Detailed Schematic of the Two-Dimensional Cold Model 317
113. Two-Dimensional Cold Model with Modified Draft Tube Inlet 318
114. Dimensions of the Serpentine Copper Rods 320
115. Schematic Design of the Advanced Recirculating Bed Combustor 324
xxiii
-------
TABLE OF CONTENTS
Volume II
APPENDICES TO VOLUME I
Volume III
PRESSURIZED FLUID BED BOILER DEVELOPMENT PLANT DESIGN
Volume IV
FLUIDIZED BED OIL GASIFICATION/DESULFURIZATION
XXV
-------
VOLUME II
APPENDICES TO VOLUME I (PRESSURIZED FLUID BED
COMBUSTION PROCESS DEVELOPMENT AND EVALUATION)
ECONOMIC SENSITIVITY
A. Boiler Tube Specifications
B. Effect of Operating Conditions and Boiler Design on
Boiler Cost
C. Parametric Study of Elutriation Rate from a Fluidized
Bed Combustion Boiler
SULFUR REMOVAL
D. Base Design — High-Pressure One-Step Process
E. Base Design — Low-Pressure One-Step Process
F. Base Design — Two-Step Process
G. Temperature Control and Process Turndown
H. Effect of Boiler Conditions
I. Once-through Process
J. Constant Load Concept
K. Process Comparisons
L. General Study of One-Step Process
M. Low-Sulfur Coal
N. Limestone Wet-Scrubber Cost
PLANT OPERATION AND CONTROL
0. Part Load Operation
ALTERNATIVE PRESSURIZED FLUID BED BOILER CONCEPTS
P. Recirculating Bed Boiler Studies
xxv 11
-------
VOLUME III
PRESSURIZED FLUID BED BOILER DEVELOPMENT PLANT DESIGN
Page
1. SUMMARY 1
2. INTRODUCTION 3
3. BACKGROUND INFORMATION 7
4. DEVELOPMENT PLANT DESIGN 17
4.1 Design Basis 19
4.2 Flow Diagram 22
4.3 Material and Energy Balance 22
4.4 Boiler Module Design 22
4.5 Auxiliary Equipment 38
4,5.1 Dolomite and Coal Preparation and Feed 38
4.5.2 Water and Steam Supply 39
4.5.3 Dolomite and Ash Disposal 40
4.5.4 Steam Disposal 40
4.5.5 Particulate Removal 40
4.5.6 Combustion Gas Disposal 40
4.6 Gas-Turbine Test Facility 41
4.6.1 Test Turbine 43
4.6.2 Stationary Test Passages 48
4.6.3 Test Rig - Required Engineering 48
4.7 Regeneration/Sulfur Recovery 50
4.8 Instrumentation and Control 50
4.9 Perspective View 52
4.10 Plant Model 58
5. EXPERIMENTAL PROGRAM 59
6. COST ESTIMATE 63
7. IMPLEMENTATION 65
8. REFERENCES 71
APPENDIX: Pressurized Fluid Bed Boiler Development
Plant Designs and Estimates
xxviii
-------
VOLUME IV
FLUIDIZED BED OIL GASIFICATION/DESULFURIZATION
Page
1. SUMMARY 1
1.1 Introduction 1
1.2 Technology Assessment 2
1.3 Environmental Impact 5
1.4 Development 7
1.5 Conclusions 9
2. ATMOSPHERIC PRESSURE FLUID BED OIL GASIFICATION 11
2.1 Process Evaluation 11
2.1.1 Gasification/Desulfurization Concepts 11
2.1.2 Experimental Work 13
2.1.3 Design 14
2.1.4 Evaluation 17
2.2 Demonstration Plant Program 23
2.2.1 Scope 23
2.2.2 Location of Utility Partner 27
2.2.3 Conclusions 31
3. OIL GASIFICATION FOR COMBINED CYCLE POWER GENERATION 33
3.1 Introduction 33
3.2 Process Concepts and Options 34
3.2.1 Gasifier Air/Fuel Ratio 36
3.2.2 Limestone/Dolomite Regeneration Method 37
3.2.3 Once-through System 37
3.2.4 Other Options 39
3.3 Process Specifications and Design Basis 39
3.4 Material and Energy Balances 39
3.5 Capital Investment Evaluation 40
3.6 Performance 45
3.7 Comparison with Alternative Power Generation Systems 49
xxix
-------
3.8 Shell and Texaco Pressurized Oil Gasification Processes 52
3.9 Conclusions 55
4. SUPPORT STUDIES 57
A.I Introduction 57
4.2 Sulfidation of Calcined Limestone and Dolomite 58
4.2.1 Conclusions from this Study 59
4.2.2 Experimental Methods 59
4.2.3 Equipment Problems 60
4.2.4 Atmospheric Sulfidation of Tymochtee Dolomite 61
4.2.5 Further Isothermal Experiment Results 61
4.2.6 Calcined Dolomite 1359 62
4.2.7 Sulfidation at Pressure on Dolomite 64
4.2.8 Stone Substrates 65
4.2.9 Sulfidation: Interpretation 68
4.2.10 Calcium Utilization 70
4.2.11 Fuel Gas 70
4.2.12 Mass Transfer Limits 71
4.2.13 Comparison of TG Data and Fluidized Bed Results 72
4.2.14 Outline of the Apparatus 73
4.3 Waste Disposal 74
4.3.1 Conclusions from This Work 76
4.3.2 Previous Investigations 77
4.3.3 Experiments 78
4.3.4 Initial Investigation 79
4.3.5 The Oxidation of Sulfided Dolomite 81
4.3.6 The Oxidation of Sulfides Prepared from a 85
Series of Limestones
4.3.7 The Kinetics of Oxidation 85
4.3.8 Summary 87
4.3.9 Some Implications of the TG Study on 89
Calcium Sulfide Oxidation
xxx
-------
5. REFERENCES
APPENDICES
A. Performance of the Oil Gasification Process as a Function
of the Air/Fuel Ratio
B. Space Requirements and Gasifier Start-up
C. Evaluation of Oil Gasification Process Costs
D. Oil Gasifier Turndown-Variable
E. Presentation Document — Clean Boiler Fuel by
Fluidized Bed Gasification/Desulfurization
F. Utility Prospects for Demonstration Plant
G. Utilities Visited
H. Proposed Utility Agreement
I. Utility Manpower Requirement
J. Follow-up Meeting Agenda
K. Meeting Report — December 11, 1972
L. Letter of Intent
M. Invitation to Bid
N. Partial Oxidation Processes
0. Material and Energy Balances
P. Equipment Design
Q. Capital Investment
R. Performance and Energy Cost
S. Shell and Texaco Process Data
xxxi
-------
CONVERSION FACTORS — ENGLISH TO METRIC UNITS
Length
Area
Volume
Mass
Pressure
Temperature
Energy
Power
English System
in
ft
gal
oz
Ib
ton
lb/in2
in H20
°F
°R
Btu
Btu/min
Metric Equivalent
2.54 cm
0.305 m
6.45 cm
0.0930 m
16.39 cm3
28.32 1
3.785 1
28.35 gm
453,6 gm
907.2 kg
51.70 mm Hg
1.865 mm Hg
1.8 (°C) + 32
1.8 °K
252 cal
252 cal/min
xxxiii
-------
1. SUMMARY
Fluidized bed combustion technology can be applied with many
different configurations and can be utilized in many different steam and
power cycles. » » » > » Fluidized bed combustion systems, operating
at elevated pressure in a combined cycle power plant, offer the greatest
potential for producing electrical energy from fossil fuels within
environmental constraints and at a cost less than conventional power
plants using low-sulfur fuel or stack gas cleaning. This conclusion is
based on a previous evaluation of the fluidized bed combustion process.
Further development and evaluation of pressurized fluid bed combustion
systems have been carried out in order to assess more fully the concept
potential. Results from the following tasks are reported:
• Economic sensitivity study. The effectiveness and
economics of a pressurized fluid bed combustion boiler
combined cycle power plant (Figure 1) using state-of-the-art
power generation equipment (steam conditions 2400 psig/1000°F/
1000°F, gas-turbine expander operating at 10:1 pressure
ratio and 1600°F) were previously reported. The economic
sensitivity of the operating conditions and design
parameters selected for this design was determined. The
boiler design concept is presented in Figures 2 and 3,
and the design parameters are summarized in Table 1.
• Sulfur removal system data. A pressurized thermogravimetric
analysis system was used to collect kinetic data on sulfur
dioxide sorption by limestones and dolomites and spent stone
regeneration at projected operating conditions — e.g.,
temperature, pressure, stone size, gas composition. These
data provide information, not previously available, with
which the desulfurization of fuels in a pressurized fluid
bed system can be predicted and explained.
-------
I—
s
CO
CD
. -a
-a 33 .~
— CO O
QQ
ro
a.
o
Q_
i_
CD
'o
QQ
•a
CD
CQ
-a
CD
M
CD
ISI
00
IX)
CD
l_
Q.
t
i—I
CD
cn
Q_
-------
Reheated Steam
Refieater Bed
Superheated Steam
Superheater Bed
Superheater Bed
Pre-evaporator Bed
Feed Water
PUNT VESSEL
SIZE DIAMETER, 0
320 mw 12 Ft.
635 inw 17 Ft.
Dependent an spertting conditions.
excess air, bed depth, gas velocity
ELEVATION
PRESSURIZED FLUIDIZED BED STEAM GENERATOR
FOR COMIIED CYCLE PLANT
FOUR 14] REQUIRED
Figure 2 Pressurized Fluidized Bed Steam Generator Module
RM-59131
-------
Dwg. 6168A83
Superheater
Inlet
Header-
Primary
Combustor
1
with |\ Vertical
Serpentine Tubes
o
o
Gas Passage
Water Wai Is
Limestone
Inlet Pipe
Figure 3-Boiler Cross-Section (Superheater)
-------
TABLE 1
BOILER DESIGN
Design Parameter
Specification
Operating Conditions
Bed temperature, °F
Gas velocity, fps
Particle size, inches
Excess air, %
Bed depth, ft
Design Specifications
Construction
Heat transfer surface
Materials
Tube size, inches
Gas side heat transfer
coefficient, Btu/hr-
ft - F
Particle carry-over (grains/SCF)
Sulfur removal
Sorbent
System
1300-1750
5-10
1/4" x 0
10 or 100
10-15
maximum shop fabrication
conventional
1-1/2 and 2
50
10
dolomite
once-through and
regeneration
-------
Sulfur removal system economics. The operation of the
pressurized fluidized bed combustion process using
limestone or dolomite results in the production of
large quantities of dry, granular partially utilized
limestone or dolomite materials in the form of calcium
sulfate/calcium carbonate. This sorbent material may
be regenerated to a form suitable for repeated sulfur
dioxide removal and the sulfur recovered as sulfur or
sulfuric acid; or it can be disposed of in its partially
utilized form in a once-through system. Alternative
regeneration processes and once-through processes were
studied, environmental factors identified, economics
projected, performance projected, and the processes
compared.
Plant operation and control. The ability of the
pressurized fluid bed combustion boiler power plant
to start up, shut down, follow load demand, and handle
emergencies was conceptually presented in the previous
design and evaluation. A more detailed assessment
has been made to determine the ability of the concept
to meet electrical utility operating requirements. A
definition of procedures with estimates of key parameters
is presented.
Alternative concepts. As previously indicated, pressurized
fluid bed combustion can be utilized in numerous power
cycle concepts and can take on different configurations.
Two concepts were studied — an adiabatic combustor
combined cycle plant which represents a modification of
the previous power cycle and a recirculating bed boiler
which represents an alternative fluidized bed boiler
concept. Conceptual designs, performance, and economics
are projected and compared with the pressurized fluid bed
boiler combined cycle plant.
-------
The results from these studies of pressurized fluid bed
combustion boiler combined cycle power plants indicate that:
• Plant costs and performance are essentially invariant
with projected changes in operating and design parameters,
the costs and performance changes representing less than
a 3% change in energy cost.
• A once-through sulfur sorbent process is the most
attractive for first generation plants. This conclusion
is based on:
- Data which show that high utilization (1.2 to 1.5 Ca/S
molar ratio of the sulfur sorbent can be achieved, which
minimizes the quantity of solids for disposal or utilization
- Data which show that spent stone can be disposed of without
further processing
- Data which show that further development work is
required before demonstrating a regenerative system on
a commercial plant
- Improved plant efficiency with once-through operation
- Reduced complexity/greater reliability with once-through
operation
- Data which show that there are sufficient quantities of
sorbent for either once-through or regenerative operation
The high stone utilization results make the once-through
system more economical than previously reported, where three
to six times the stoichiometric Ca/S ratio was assumed.
• Plant operation and control can meet the requirements of the
utility industry; and the techniques proposed for load control,
start-up, shutdown, and emergencies are within the state-of-the-art.
Plant operation has been developed to provide the capability of
operating from 12-1/2% to 100% of rated capacity, change load
at the rate of 5% per minute, operate with some components out
of service for maintenance, start up or shut down the plant in
7
-------
several different sequences to ensure maximum usage of
plant equipment, and operate the plant as a swing or base
load plant.
• Alternative pressurized fluid bed combustion processes
should be developed. The adiabatic combustor concept
does not offer the performance or economic potential of
the pressurized boiler concept. The simplicity, however,
and the potential for shorter construction time and for
near-term application make the concept attractive. Advanced
pressurized boiler concepts such as the recirculating bed
boiler offer potential advantages of lower costs, simpler
control,and greater sulfur dioxide and nitrogen oxide control.
The advanced concepts should be developed as second
generation systems.
An assessment of the pressurized fluid bed combustion boiler
combined cycle power plant technology based on results from these and
other studies show:
1.1 ECONOMIC COMPARISON
Pressurized fluid bed boiler combined cycle plants using
state-of-the-art power generation equipment may have energy costs 15
to 20% less than those of conventional plants with stack gas cleaning —
Table 2 and Figure 4. The costs presented are based on the cost data
given in the Westinghouse Report of Contract CPA 70-9 and EPA data on
9
stack gas cleaning. The costs reported do represent a significant
increase from the capital-cost trend of the recent past. However, the
*
costs are considered real for the design basis specified and are
comparable with other contemporary power plant cost estimates. Cost
estimates for new systems must be compared with conventional plant
costs prepared using the same basis — thus the relative costs on the
same basis are important.
*
Factors affecting plant costs include the type of plant (e.g. base
load versus intermediate load), regional differences (e.g. labor
requirements and rates, site conditions), differences in plant
facilities (e.g. indoor or outdoor plant, thermal emission control)
and differences in what reported estimate includes (e.g. contingency,
year of operation, escalation rate).
8
-------
Curve 654050-A
16
15
14
f- 13
to
O
o>
Conventional Plant w/o
SCL Control
12
1!
Conventional Plant with
Wellman-Lord
pressurized Fluid Bed Boiler Plant
Once-throuqh Dolomite
Ca/S =2; Stone at $10/ton
1.0
2.0 3.0
Weight % Sulfur in Coal
4.0
Figure 4-Electrical Energy Costs
-------
w
rJ
H
CO
H
O rH
O Cfl
VO O
CD
§
a
o
c a
rH Cfl 0)
Cfl rH J-l 4-1
C PM 3 co
O *4-l >-i
•H 13 rH CO
4-t 0) 3
CJ !-l CO rH
CO iH Cd
J> [j^ o ^
G 1 -~- 0
0 rH £ B
C_> cfl 0>
o pi
u
3
4J IH
C rH B
rH Cfl 3 0)
SrH CO 4J
P-l CO
O 13 >-,
•H 13 VJ CO
4J 0) O
(3 r4 r4 rH
0) -H I cfl
C 1 CO 0
0 rH 0 S
U cfl rH 0)
O rH #
CJ 0)
S
H
0)
rH
•H
O 0) S
M 4-> co
•H 4-1
13 0) 0 CO
CO H 0 >>
pq a rH co
r^l O
13 0 Q rH
•H cfl
3 T3 JS >
H CO M) O
fc C 3 B
•H O 0)
T3 r^ ^-* PH
a) B .a
1*4 O 4-J 1"^
•H U | 3
^i a) n-t
3 OH
co C 3
CO O CO
CD
^t
PL.
CO r- H ^o
0 ...
CO !*•*• C5 "i^" C^J
, — |
Q r—l IT\ \O
O \D CM in ro
to ... .
O") 00 rH "sf "^
H
~5
c\
,1
X— ^
01
4J
3
o
o
Ns^X
m H o\ in o
CT> m r~- o r^ H
CN vO O *^" O CN
! 1
S
~~- ^
•c/> AJ
•^^.
" CO
•W H
CO H
O -H CO
u B <"
00
H •• U
cfl CO cfl
4J 4-1 &
•H CO O
ft O
cfl CJ 13
u cu H
>> K S a)
4J 60 -H •Ji 3 J
fl r-l pM O f*4 ^
cfl 0) H
rH C O
»J W H
4-1
C
cfl
H
ft
H
cfl
C
0
•H
4J
C
0)
C
o
o
MH
O
a
Q
•H
4-1
cfl
t-j
CO
ft
O
vO
X-N CT\
rH H
cfl
CO •"
O 4J
p>t pj
CO cfl
•H rH
\-{ en Vt
o a a)
h4 H H
1 ft -H
C O
cfl a) ,0
S co
H cfl 13
H ,C CO
CO O ,£)
« — 3 13
ft -H
3^3
)-. pq c u-i
O 3 O
4J 4-1 v£> 4-1 <4-<
O ±) &> O ~^~. O
cfl -H HO
14-t *^VO ^"^- rH £4
0) O -O -co- O
>-. VJ H O -H
4J O ~-» OO 4-1 4->
•H O CO cfl
o M in cs> ^
M cfl 3 sf 0) CO
cfl ft M-l 0) 4-1 ft
0) CO H C°) fi -H O
>-, O 3 cfl B
•~^ CO H 43 O m
S-S B-S cfl 4-1 H r^
m o o o o) o en
H r^ 2 u g Q H
CO
10
-------
1.2 ENVIRONMENTAL IMPACT
The ability to meet proposed sulfur dioxide, nitrogen oxide,
and particulate emission standards has been demonstrated. Thermal
emissions are reduced 15 to 25%. Solids from the plant operated with
once-through sorbent include coal ash and dry, granular calcium sulfate/
calcium carbonate — both suitable for disposal without further processing.
Spent stone may be regenerable,in which case the quantity of solids
would be reduced. Thus, regenerative systems should continue to be
studied and developed.
1.3 ENERGY RESOURCES
Fuel conservation is achieved with higher plant efficiencies —
'v 10% reduction in fuel requirement with once-through pressurized fluid
bed boiler plant using state-of-the-art power generation equipment when
compared with a conventional plant with stack gas cleaning. The
fluidized bed system also permits the use of low-grade fuels which are
not easily used or cannot be used in conventional systems — e.g.
solid wastes, wood chips, refinery bottoms, coal char from gasification
systems.
1.4 TECHNICAL UNCERTAINTIES
The economic sensitivity study on operating and design
parameters shows that the uncertainties for successful application are
minimized, since cost and performance are essentially invariant with the
projected design basis. However, back-up design and operating alternatives
have been specified where problem areas have been identified. An
illustration of the problem areas, proposed solutions, and back-up is
presented in Table 3. The design may be conservative in some areas.
For example, if deep beds are practical, only one superheat bed may be
required; and the carbon burn-up cell,which is desirable in a
pressurized system, may not be required with the deep beds and high
excess air.
11
-------
W
w
«3j
g
•H
4-1
Cfl
G
i-i
0)
4-1
rH
,
S-i
CO
a
•H
sj
PH
C
00
•H
CO
0)
«
13
CU
CO
o
p.
cr
rl
P-l
CO
4-1 C
Cfl -rH
CU ,0 1 00
X! S-l W C
CO 3 CO -H
)-l 4J 0) 00 4J
C cfl
13 CO -H C V4
C CO Xi O CU
cfl 00 S-i CX CO
3 ^ O C!
CO O 4-1 O O
CO 4J CO CO -H
00 >>Xi G 4J
}-l 4-1 iH -r-t
rH 0 -rl CXXI T3
O -H 13 O S-l C
O M O S-l 3 O
o ex a i3 4-1 a
• • •
CU
00
CO <4H
4-1 O
CO
4-1
-C) >1 C! •• rl
M 4-1 CO 8 CO
•H -H 4J CU 4->
f! O 0 4J rH
4-100 CO -H
rH a >^MH
s-i co to
O > CO T)
14_| CU T3 CU CU
to a cu > x>
C! CO -H CO tH
o oom M-I 4-» s-i
•H CO tfl
co a) co co a rH
•H CJ O tJ M 3
> 3 3 -H tu a
O Id 13 rH 4J Cfl
S-l CU •>
4-1
1
CU
a
>.
13
O
M
cu to
<3 rl
o
T) 4J
C! CJ
cfl CU
rH
CO rH
CU O
C CJ
0
rH 4-»
O CO
>^ 3
O *<3
rH
rH
Cfl
>
O
0
CU
&
cu
4-1
tfl
rH
3
CJ
•H
4-1
S-l
CO
CM
CU
a
o
4J
CO
5
§
C!
4J
O
CO
rH
CU
CO
•
CU
CO
3 to
4J
cu C
co cu
cfl x>
a) ^4
JJ O
o co
a
•H CO
S
o o
4J CO
CU S-l
> o
Cfl IW
X!
0)
>> 4J
cO oj
•
(N
O
4J
CN
,
rH
• *
13
a)
xi
a
•H
f\
CO
4J
•H
a
O c/3
H — .
O cfl
TJ 0
T-H
O
rl
4J
a
o
CJ
CO
13
•H
X
O
•H
Q
i-i
3
•4-1
rH
3
C/3
m
1 O
•iH
13 a
a o
O -H
O 00
CU
00 S-l
C
•H rl
o co
3 DS
13 O
CU i-4
rl
a
CU iH t)
4-1 CU
o co xi
e a
o o cu
S-t -rt J3
ex 4-1 4J
•
C
o
•H
4-1
CO
3
,Q
e
o
CJ
00
a
•H s-*.
}-t 13
3 CO
13 4J
cfl
T3 !-i
CO 4-1
N CO
•H fi
8 o
•H a
d S
•H 13
8 ^
cu
T)
•H
K
O
a
CU
00
o
S-i
4-1
•H
z
CO
rH
cfl
•iH
S-l
(U s~*
4-1 CU
Cfl rH
8 ^3
CO
CU H
-0 -rl
cfl cfl
S-l >
00 cfl
1
S-l CU
cu s-i
& cfl
00
•H J2
£1 0
•H
CU ,C
CO &
3 ^
•
cu
rO
3
4-1
S-l
co
rH
•H
O
,n
rH
CO
a
O CO
•H rH
4-1 tfl
fi iH
CU rl
F> CO
a 4J
0 cfl
a 8
CO
rH
Cfl
•H
S-i
CO
4J
tfl
S
00
o
rH
O «
C 13
,£! CO
O Q)
CU UH
4J CO
>-> rl
cu s-i cu
> S-l 13
•H 3 CU
4-1 rH CU
Cfl CO H-l
rl •> S
cu • cu
4J 00 S-l
rH • 0
tfl cu co
•
tJ
cu
cu ••
UH >>
4-1
IH -H
o a
cfl
rl CX
CU T3 cfl
XI CU 0
a ^
3 4J
a M in
cu a x:
CU CX 3 4J
co ex
co co cu cu
cu 4J a 13
rl fi 3
O -H 13 t)
C o co to
•H ex s-i ,a
• •
4-1
CO
>>
CO
13 /^
CO CO
CO S-l
MH CO
CX
xi ex
rl O
CO X!
o
0 ^J
rl 0
4-1 0
CU rH
(Xs-'
T3
CU
cu
PR
rH
cO
o
CJ
CO
rH
cfl
4-1
CU
8
o
4J -H
rH
4-1 CO
G ^
CU rH
X) cfl
rl
O CU
co >
0
•o e
13 CU
Cfl )H
•
S-i
cu
Xi
4-1
S-l
3
4H
CU
rl 8
3 CU
4-1 rH
Cfl XI
rl O
CU rl
CX CX
8
CU 13
4J -H
O
rl >
CU Cfl
S
0 0
rH 4J
•
13
•H
O
-a >
CU cO
a
•H 0
Cfl 4-1
4-1
a &
•H 0
Cfl rH
£2
>.
CU H
Vt 4J
3 a
4J CU
cfl -H
rl 0
CU -H
CX 4-1
CU 3
4-1 CO
rH
CO
rH
CO
4-1
CU
S
•rl
rH
CO
,b,
S-i
cfl
>
•
CU
!-i
3
4-1
Cfl
s-i to
CU S-l
CX CO
S rH
CU -H
4-1 O
XI
13
cu s-i
X> cfl
rH
>. 3
M 13
cfl O
> a
»>
1*
o
rH
H
O
Fn
13
CO
o
J
13
C
cfl
£
*J
O
T3
C
S-i
3
H
a
o
•H
4J
cfl
rt
•H
t^rH
rl -H
CO 4-1
> 3
O
CJ rH
CU CO
S-l -H
CJ
SH S-i
3 CU
rH G
3 O
CO U
• •
i-H
rH
•H
MH
13
CJ
CO
rH
rH
CO
CO
o
ex
CO
•H
Q
CU
cs
o
4-1
C/3
4-1
G
cu
CX
c/>
cu
IH
•H
rH
00
qj
o
r-j
13
CJ
CO
K»1
4.1
•H
rH
•rf
0
cfl
•H
, — |
CU
Vi
CO
g
•H
rt
S-l
3
4-1
1
CO
Cfl
00
CO
J>
cu
•H
J3
Q
cfl
-i i
,— 1
i.
rl
.[ j
PJ
QJ
rH
12
-------
1.5 DEVELOPMENT STATUS
Fluidized bed combustion for the purpose of power generation
has come under study, in various forms, in a number of countries during
the past twenty or more years. The first major development effort in
the United States was initiated in 1965 by the Office of Coal Research
(OCR) with Pope, Evans and Robbins to develop an atmospheric-pressure
industrial coal-fired packaged boiler. OCR has a continuing effort to
develop an atmospheric-pressure system and is supporting pressurized
system development. EPA became involved in the development of fluidized
bed combustion to achieve reduced air pollutant emissions in 1967. An
overall program has been under way by EPA to demonstrate the technology
on a full scale. The program, which has included eight contractors
and subcontractors, including Westinghouse; Pope, Evans and Robbins;
Esso; Erie City; Foster Wheeler; Argonne National Laboratories; the
Bureau of Mines; and the National Coal Board, represents a broad spectrum
of capabilities. Pope, Evans and Robbins, under contract to OCR, is
designing a 30 MW atmospheric-pressure fluid bed boiler. Under the
direction of the Office of Research and Development, EPA, pressurized
fluid bed combustion units have been built by Argonne National Laboratory
and Esso Research and Engineering. A 1 MW test facility is being built
by Esso; and a 30 MW development plant has been designed by Westinghouse.
The British Coal Utilization Research Association was the first to
build and operate a pressurized fluid bed combustor of approximately
4 10
1 MW equivalent capacity. ' The pressurized fluid bed boiler development
status is summarized in Table 4.
A demonstration plant program has been conceived under the
direction of EPA to commercialize the pressurized fluid bed boiler
concept. The 30 MW development plant will provide the capability to
study and demonstrate pressurized fluid bed combustion and heat transfer,
steam generation, solids feed and handling, particulate removal, gas-
*
Esso has since changed its name to Exxon. For purposes of this report,
the name Esso has been retained.
13
-------
turbine performance, boiler control, and sulfur dioxide and nitrogen oxide
emission control. Preliminary designs, cost estimates, experimental
program, schedule, and program alternatives are presented in Volume III.
A model of the plant, estimated to cost approximately $10 million, is shown
in Figure 5.
TABLE 4
PRESSURIZED FLUID BED BOILER DEVELOPMENT STATUS
Apparatus
• Pressurized combustion, sulfur removal and regeneration test
units operating (^ 1 MW equivalent)
• Pressurized fluid bed boiler pilot plant operating (^ 1 MW
equivalent)
• Second pressurized fluid bed boiler pilot plant to be opera-
tional in 1974 (^ 1 MW equivalent)
Designs
• Preliminary design of commercial plant complete (635 MW)
• Demonstration boiler plant preliminary design complete
(20-30 MW equivalent)
Results
• 95% sulfur removal demonstrated
• < 0.2 Ib N02/10 Btu demonstrated
• No erosion, deposition, corrosion of gas-turbine blades
after 200 hr tests
• Boiler tube materials adequate for commercial operation
• Horizontal tubes tested
• Coal feeding at pressure demonstrated (5 atm; 10 atm tests
under way)
• Continuous operation - runs up to 350 hours
• Economic sensitivity of design and operating conditions checked
14
-------
o
CO
o
"TO
O
E
CD
OJ
O
2Q
O
O>
O
CD
M
13
to
Z5
cn
15
RM-59130
-------
1.6 COMPARISON WITH ALTERNATIVE TECHNOLOGY
The pressurized fluid bed combustion boiler concept using
state-of-the-art power generation equipment is projected to meet
environmental requirements and to have energy costs as low as or lower
than any other competitive system — commercial or under development —
using state-of-the-art power generation equipment. This includes stack
gas cleaning, liquefaction/power generation, gasification, atmospheric-
pressure fluidized bed combustion, or low-sulfur fuel. The concept
also offers greater simplicity and versatility than competitive systems
using high-sulfur/low-grade fuels by utilizing only one fuel processing
step — combustion — and by controlling sulfur dioxide and nitrogen
oxide emissions during the fuel processing step. Thus, the system
avoids fuel pretreatment and processing of the stack gas. The system
can be demonstrated for use in the 1980s. Second generation pressurized
fluid bed combustion boiler systems offer the potential for increased
plant efficiency (up to ^ 45%) by increased gas-turbine temperature anu
128
increased steam temperature and pressure ' ' and the potential for
3 5
application in alternative and advanced power cycles. ' In conclusion:
• Pressurized fluid bed combustion combined cycle power
plants using commercial power generation equipment offer
the opportunity to generate electrical energy within
environmental constraints at lower energy costs than
competitive systems.
• No problems have been identified which preclude the
development of pressurized fluid bed combustion combined
cycle power plants.
• Demonstration of the pressurized fluid bed combustion
combined cycle power plant concept should be accelerated;
detailed design and construction of a 30 MW demonstration
facility should begin.
16
-------
2. ECONOMIC SENSITIVITY
A pressurized fluidized bed boiler power plant has been
1 2
designed using state-of-the-art power generation equipment, ' with
performance, costs, and pollution abatement projected for the system.
The results show the concept has the potential to meet S09, NO , and
Zf X
particulate emission standards; and could reduce energy costs 10%
below those of a conventional plant with stack gas scrubbing.
Operating conditions and design parameters for the pressurized
boiler were selected on the basis of an evaluation of available data,
power cycles, and alternative boiler concepts. It is important to know
how sensitive the operating and design parameters selected for the base
design are to the plant economics. An understanding of the effect of
changes in the proposed design on plant cost will provide a basis for
evaluating current pressurized fluid bed combustion pilot plant data,
planning experimental programs, designing the development plant, and
understanding the economic margin for solving technological problems.
The sensitivity analysis evaluates the effect of the following
variables on plant design, cost, and performance:
• Fluidized bed boiler operating conditions
- Bed temperature
- Fluidizing velocity
- Excess air
- Pressure
• Fluidized bed boiler design
- Heat transfer surface: configuration,
heat transfer coefficient, materials
- Module capacity
• Particulate carry-over from the boiler
- Loading
- Size distribution
17
-------
• Power plant equipment operating conditions
- Gas-turbine inlet temperature
- Steam temperature and pressure.
The evaluation is performed by considering each variable separately. It
is general rather than specific in order to permit the coupling of
different effects and thereby assess alternative designs; and it is
performed for the purpose of indicating the effects of variable changes
in plant costs relative to each other and to the total plant cost.
2.1 BASIS FOR SENSITIVITY ANALYSIS
The basis for the sensitivity analysis is the boiler and plant
1 2
design developed by Westinghouse under contract to EPA. ' The power
plant cycle is shown schematically in Figure 6. The plant subsystems
included in this sensitivity analysis are enclosed within the broken
lines. The pressurized boiler designed by Westinghouse and Foster Wheeler
is shown schematically in Figure 7. The preliminary boiler design was
for a nominal 300 MW plant. The boiler design consists of four modules
and provides for a maximum of shop fabrication and turndown requirements.
Each module includes four primary fluidized bed combustors, each containing
a separate boiler function — one bed for the preevaporator, two beds for
the superheater, and one bed for the reheater. Evaporation takes place
in the water walls. All of the boiler heat transfer surface is immersed
in the beds, except for baffle tubes above the bed to minimize particle
carry-over. Each module contains a separate fluidized bed or carbon
burn-up cell to complete the combustion of carbon elutriated from the
primary beds. The philosophy behind the boiler design was the maximization
of shop fabrication. Thus, the 300 MW plant utilizes boiler modules which
can be completely shop fabricated. From roughly 300 to 600 MW, the boiler
modules can be only partially shop fabricated since the pressure shell
is too large for rail transport. Larger plants utilizing the four-module
concept would be field erected.
18
-------
3
o>
o
OQ
-------
Reheated Steam
Reheater Bed
J^>
-------
The operating conditions and design parameters for the boiler
and the power cycle are summarized in Table 5. The power plant performance
and economics were determined on the basis of these specifications. The
cost breakdown for the fluidized bed steam generator is summarized in
Table 6. The fluidized bed boiler design was scaled to 600 MW capacity
and the costs estimated. These costs are summarized in Figure 8. A
breakdown of the power plant equipment costs for a 635 MW plant is
presented in Table 7. Limestone or dolomite regeneration is not included
in this analysis: the costs presented are for a once-through system.
The energy costs used for this analysis are also presented in Table 8.
The costs of a conventional plant with wet scrubbing on the same basis
are also indicated.
The following assumptions are made for the sensitivity
analysis:
• The plant maintains the four-module concept with two
modules per gas-turbine.
• The coal feed rate is maintained constant for each
variable analysis. Thus, the coal feeding and hand-
ling system is assumed to remain unchanged. This
may not be completely true if the bed area is changed
significantly and the number of feed points increased
or decreased. The cost of the coal feed system is
considered if the bed design is altered.
• Structural and erection costs are constant. The
structural steel and concrete costs for the boiler
plant equipment are ^ $2/kw. The maximum change in
these costs for the cases considered is ^ $0.20/kw
and, in general, will be significantly less. The
cost is thus assumed constant for this analysis.
Erection cost changes are negligible.
21
-------
TABLE 5
PRESSURIZED FLUIDIZED BED BOILER POWER PLANT
OPERATING AND DESIGN CONDITIONS
Item
Specification
Cycle
Steam system
Gas-turbine expander
Pressure ratio
Inlet temperature
Air cooling
Coal feed rate
Number of boiler modules
Boiler modules/gas turbine
Fuel/air ratio
2400 psia, 1000°F superheat, 1000°F reheat
10:1
1600°F
5%
53,910 Ib/hr/module for nominal
300 MW plant design
4
2
0.0919
Boiler Design
Bed area
Heat transfer surface
Walls
Bed
Gas side heat transfer
coefficient
Tube materials
Bed depth (expanded)
Gas temperature drop from
primary beds to gas -
turbine expander
35 ft (5' x 7') — for ^ 80 MW module
2" O.D. tubes on 3-1/2" welded wall spacing
1-1/2" O.D. tubes in preevaporator and
and superheater; 2" O.D. in reheater
(details in text)
50 Btu/hr-ft2-°F
SA-210-A1 — preevaporator
SA-213-T2 — lower superheater
SA-213-T22 — water walls; upper
superheater (lower loops); reheater
SA-213-TP304H — upper superheater
(upper loops)
11 to 14 ft
150°F
22
-------
TABLE 5 (Continued)
Item
Specification
Boiler Operating Conditions
Bed temperature (100% load)
Fluidizing velocity
Excess air
Particle carry-over
Carbon from primary beds
Auxiliaries
Coal feed system
Primary particulate removal
Secondary particulate removal
Stack gas coolers
1750°F
6 to 9 fps
17.5%
^ 7 grains/SCF
6% of carbon feed
petrocarb feed system
4 size 355 VM 8/0/150 Duclone
per module — nominal 300 MW design
2 model 18000 Type S collectors
per module — nominal 300 MW design
(quoted by Aerodyne Dev. Corp.)
conventional heat exchanger design
23
-------
TABLE 6
COST OF A 318 MW PRESSURIZED FLUID BED BOILER
Item
Cost
Pressure Parts
Shell
Subcontracted & Contracted
Equipment
Drafting
Home Office
Subtotal
Erection
TOTAL
$1,777,000
935,000
435,000
185,000
685,000
$4,017,000 field-erected
($3,856,000 shop-assembled)
500,000
$4,517,000
Pressure parts include tubing cost, headers, downcomers, risers,
tube bending, tube welding, and water-wall fabrication.
24
-------
TABLE 7
EQUIPMENT COSTS FOR PRESSURIZED FLUIDIZED
BED BOILER POWER PLANT1 (635 MW)
Component Cost($/kw)
Boiler Plant Equipment
Boiler 14.49
Particulate removal 12.76
Piping/ducts 4.43
Stack and foundation 0.47
Coal handling and feeding equipment 14.94
Ash and dust handling system 1.55
Instruments and controls 3.10
Miscellaneous equipment 0.94
Steam Turbine - Generator Equipment 44.14
Gas Turbine - Generator Equipment 14.80
Other (land, structures, electric plant 70.38
equipment, miscellaneous plant equipment,
undistributed costs)
SUBTOTAL 182.00
TOTAL CAPITAL COST (inc. escalation, IDC, etc.) 265.00
($340/kw for conventional plant with wet scrubbing on same basis)
25
-------
3.5
2.5
SO
O
x 2
o
CJ
1.5
0.5
Curve 647187-B
Pressure Parts -
Shell
Erection
Drafting
I
I
.Home Office (Engineering,
Contract Reserve, etc.)
Subcontracted &
Contracted Equipment
I
0 100200300400500600700800
Plant Size, MW
Figure 8: Steam Generator Cost Breakdown for the W-FW Basic Design
26
-------
TABLE 8
ENERGY COSTS FOR PRESSURIZED FLUIDIZED BED BOILER
Item Cost (mills/kWh)
Fixed Charges 6.44
Fuel 4.04
Dolomite 0.52
Operating and Maintenance 0.71
11.71
(13.45 for conventional plant)
27
-------
• Coal and stone feed size are assumed constant:
1/4" x 0. This parameter is important when con-
sidering particle carry-over, but insufficient
information is available to permit a quantitative
analysis.
2.2 OPERATING CONDITIONS
2.2.1 Bed Temperature
The full load design temperature ±s 1750°F. Lowering the
design bed temperature increases the total amount of heat transferred in
the bed and thus increases the steam-turbine power generation. At the
same time, lowering the bed temperature decreases the gas-turbine inlet
temperature — assuming there is no burning above the bed — and thus
decreases the total gas-turbine power generation. The decrease in gas-
turbine power is larger than the increase in steam power and results in
an overall decrease in plant power (Figure 9). Lowering the bed temperature
increases the total amount of heat transferred in the bed and thus
requires more heat transfer surface (Figure 10). Assuming the cross-
sectional area of the fluid bed (5 ft x 7 ft for a 300 MW nominal plant
size) and tube size/tube pitch are constant, the expanded bed depth for
each functional bed increases with a decrease in bed temperature, as shown
in Figure 11. The bed depth of the two superheater beds is assumed for
convenience to be the same. This will not affect the total heat transfer
surface requirement and the resultant module height shown in Figure 12.
The bed depth and module height can be reduced by enlarging the bed
area and module diameter. However, since the module diameter of 12 feet
is considered to be the largest size that can be shipped by railroad, an
increase in the module diameter to accommodate additional heat transfer
surface may not be economical for a 300 MW plant.
The effect of changing design bed temperature on the steam
generator cost is calculated on the basis of the boiler cost estimation
shown in Table 6 and on the assumptions that the module diameter is constant
28
-------
300
I 200
cu
c
0>
O
100
0
1300
9000
8000
7000
* 6000
tu"
8
I 5000
i_
1 4000
S 3000
2000
1000
0
1300
Curve 645271-A
Plant Power
.
Steam-Turbine Power
Assume Bed Depths are
Constant at Basic Design Values
ntrT..rhino Power
1400
1500 1600
Bed Temperature, °F
1700
1800
Figure 9: Effect of Bed Temperature on Power Generation
Curve 6W269-A
_L
_L
1400
1500 1600
Bed Temperature, °F
1700
1800
Figure 10.- Heat Transfer Surface Requirement for One Module
29
-------
25
20
I
1.
X
15
10
1300
150
100
a*
1
c
i75
50
0
1300
Curve 6^5270-A
1400
1500 1600
Bed Temperature, °F
1700
1800
Figure 11: Bed Depth Requirement at Different Bed Temperatures
Curve 64J273-A
I
1400
1500 1600
Bed Temperature, °F
1700
1800
Figure 12: Effect of Bed Temperature on Total Module Height
30
-------
at 12 ft and the total number of modules is four, based on turndown
consideration. The cost (not including erection) of the four-module
steam generator as a function of bed temperature with constant module
diameter (12 ft) is shown as curves la and Ib in Figure 13. Curve la
assumes the maximum allowable bed depth to be 20 ft. That means any
bed with an expanded bed depth larger than 20 ft will have to be split into
two beds with separate air plenums and freeboards. Curve Ib assumes
that there is no restriction on maximum bed depth. The choice of 20 ft
as the maximum allowable bed depth was arbitrary and was just to show
the effect of this variable on the cost of a steam generator. It is
doubtful that the bed depth of each fluid bed can be unrestricted without
creating undesirable bubble formation and slugging, poor bed-tube heat
transfer coefficient, and temperature gradients in the bed at some bed
depth. The maximum allowable bed depth at specific operating conditions
will have to be experimentally determined in a large unit. Without the
required experimental evidence, the steam generator cost (not including
erection) is plotted against the maximum allowable bed depth in Figure 14.
The bed temperatures were calculated by assuming the gas-turbine
temperatures of 1600°F, 1500°F, 1400°F, and 1300°F; and by assuming a
linear temperature loss between the boiler and the gas-turbine inlet.
The cost of the steam generator designed for 1636°F increases ^ 20%
0\> $2.8/kw) over that designed at 1750°F if the maximum allowable bed
depth is 15 ft. The cost increase is primarily due to the splitting of
beds with a bed depth greater than 15 ft. This bed splitting can be avoided
either by decreasing the boiler tube diameter and the spacing in the bed
or by increasing the module diameter. Decreasing the tube diameter and
spacing will change the bed-tube heat transfer coefficient, tube bending
and fabrication, and tube-wall thickness; increasing the module diameter
will not only change the cost of the pressure shell but will also affect
construction — complete shop assemblage versus partial field erection.
All these factors which have to be taken into account in designing an
optimal boiler are discussed in separate sections.
31
-------
Curve 647206-B
CD
i—H
X
O)
c
0>
o
00
0
1300
la
Curve
la
Ib
2
3a
3b
3a
Ib
Basic Design Point
Bed-Tube Heat Transfer Coefficient 50 Btu/ft2 - hr -°F
Max. Allowable Bed Depth
20ft
unrestricted
20ft
20ft
unrestricted
j j
Tube Arrangement
U"0. D. atH = 7"andV=3" in
preevaporator and superheater
2"0. D. at H =7" and V =4" in
reheater
same
1"0. D.atH-4"andV=?' in
all beds
?'0. D. atH=8"andV=4"in
all beds
same
I
1400
1500 1600
Bed Temperature, °F
1700
1800
Figure 13: Effect of Bed Temperature on the Steam Generator >Cost of 318 MW Plant
(not including erection)
32
-------
Curve 6*f7155-A
10
8
o
I—I
X
•w-
o
O
O
~fO
v_
-------
In the present cost estimation, the possibility of using
thinner-wall tubes for the designs at lower bed temperatures was also
taken into account by calculating the minimum tube-wall thickness
requirement. No allowance for corrosion was provided. The method for
calculating the minimum tube-wall thickness and the results of the
calculation are presented in Volume II, Appendix A.
A change in the operating bed temperature will also change the
gas temperature to both the primary and secondary cyclones and, thus,
change the actual volumetric gas flow rate. This will change the gas inlet
velocity to the cyclones which, in turn, affects cyclone collection
efficiency. This effect was estimated to be small compared to the effect
of the change in pressure drop across the bed due to a change in the
design bed temperature. Decreasing the design bed temperature increases
the heat transfer surface requirement in the bed. This requires an
increase in pressure drop due to an increase in bed depth if the bed area
and boiler tube configuration are constant. The decrease in net plant
power output as a function of the design bed temperatures is presented
in curve 1, Figure 15. The effect is small — a decrease of only ^ 0.3%
if the design bed temperature is reduced to 1407°F.
Bed temperature is one of the primary variables used for load
turndown in the present design. A 4:1 turndown can be met if the design
bed temperature is higher than 1600°F. The primary limitation on the
operating bed temperature is the sulfur removal efficiency of the sorbents
in the bed. At bed temperatures higher than 1750°F or lower than 1350°F,
the sulfur removal efficiency in the bed is too low. Thus, it is
concluded from the above bed temperature analysis that the design bed
temperature should be the highest temperature required for a desirable
degree of load turndown and sulfur removal in the bed.
2.2.2 Fluidizing Velocity
At a constant fuel feed rate and excess air, increasing the
fluidizing velocity will require a decrease in the bed area and in the
34
-------
o
crv
c.
.22
'o
o
O
i_
CD
CO
c
CO
CO
o
o
r—
X-
i
O.
o
i_
Q
CU
13
co
co
CU
CU
cr>
c
co
^
O
_o
cu
-o
cu
CD
^o
"co
CE
s_
cu
"cu
E
CO
Q_ u Q
00?-, ?
CM ^3- o^
II II of
> > 13
o2J oS ~
?J 00 ^
II II "O
_ cu
co
co
j-
CO
CO
CO
o
cu"
J3
"co
i_
cu
o.
E
cu
^-J CQ
CU CO CO
CNJ
i—I CM
O
O
L_
I
O
D_
^— >
C
CO
CU CU
^ CQ
C CD
CD *~
CT CO
e t/j
CO O
5 £
uK
I—I
cu
13
cn
UPV
O
O
CD-
OS'
3H|BA
J9MOd
35
-------
module diameter. Since, for a constant overall heat transfer coefficient
and a specific design bed temperature, the total heat transfer surface in
the bed is constant, a decrease in the bed area will require an increase
in the bed depth at constant tube size and tube spacing and, thus, an
increase in the module height. An economical design will depend on the
balance of these factors.
The bed area and bed depth requirements with respect to change
in fluidizing velocity at different design bed temperatures were calculated
and are shown in Figures 16 through 19. The corresponding module height
and module diameter are presented in Figure 20. The cost of the pressure
shell at different inside diameters is estimated on the basis of data from
Foster Wheeler Corporation and an independent estimation by Westinghouse.
This is shown in Figure 21. The discontinuity at a module inside diameter
of 12 ft is due to the cost difference between the shop-assembled and the
field-erected shell.
Additional fabrication costs must be considered for the shell
and for the concomitant change of water walls and tube bending. Taking
into account all the factors involved, the steam generator cost is plotted
against the superficial fluidizing velocity in the preevaporator as shown
in Figure 22 for a 318 MW and a 635 MW plant. The superficial fluidizing
velocity in the preevaporator was used here since it is the largest
velocity in all beds inside a single module. The superficial fluidizing
velocity in the superheaters and reheater can be calculated accordingly.
The results show that increasing the fluidizing velocity tends to increase
rather than decrease the total steam generator cost of a 318 MW plant.
Decreasing the fluidizing velocity in the preevaporator below ^ 8 ft/sec
will require a shift from shop assemblage to field erection and will
escalate sharply the steam generator cost. At a plant size of 635 MW,
a minimum cost does exist. Figure 22 shows costs for unrestricted
maximum allowable bed depth. If, for example, the maximum allowable
bed depth is limited to 10 ft or 20 ft, the disadvantage of increasing
the fluidizing velocity would be even larger at the 318 MW size. At
the 635 MW size, the minimum would shift to lower velocity.
36
-------
I 15
i
0
Superficial
Velocity
Bed Temperature 1750 °F
Tube Pitch/Diameter Ratio Basic Design
Bed-Tube Heat Transfer Coefficient
50 Btu/ftz - hr - «F
Superheater
Reheater
Preevaporator
»
o 10 a 30 «
Bed Area, ft'
Figure 16 Dependence of Bed Depth on the Fluidizmg Velocity, Bed Temperature, 1750°F
1 1
Bed Temperature 1636°F
Tube Pilch/Diameter Ratio W - FW Basic Design
Bed-Tube Heat Transfer Coefficient
50Btu/ftZ hr -f
Superficial
Velocity
Bed
Depth
Superheater-^
Preevaporator
0 10 20 30 40
Bed Area, ft2
Figure 17 Dependence of Bed Depth on the Fluidizmg Velocity, Bed Temperature. 1636°F
~ 15
.&
8
a 10 -
Bed Temperature 1522°F
Tube Pitch/Diameter Ratio W-FW Basic Design
Bed-Tube Heat Transfer Coefficient
50 Btu/ftZ hr - °F
Bed
reevaporator ^ Dep(h
lUperheater
-Superheater T
Reheater (Superficial
— Preevaporator } Velocity
Bed Area, ft
Fiaure 18 Dependence of Bed Depth on the Fluidizmg Velocity, Bed Temperature.
1522°F
> 15
t
a.
Preevaporator —
Superheater
Bed Temperature IdO?**
Tuoe Pilch/Diameter Ratio W -FW Basic Design
B^chTube Heat Transfer Coefficient
50Btu/ftzhr -«F
Figure 19 Dependence of Bed Depth on the Fluidizmq Velocity, Bed Temperature, 1
-------
Curve 6^7l86-B
300
Bed Temperature
1407°F
1522°F
250
200
150
-a
o
100
50
10
Maximum Allowable Bed Depth Unlimited
Tube Pitch/Diameter Ratio Basic Design
Bed-Tube Heat Transfer Coefficient
50Btu/ft2hr-°F
15
14
ou
E
fa
135
12
11
10
20 7
Bed Area, ft2
30
Figure 20: Dependence of Module Diameter and Module Height on the Bed Area
38
-------
en
s
o -->_ o
c: .0 c
fU _ fU
—— -^ — -^^
— ^ « .o .2 Jl
c ^ o
o
Q.
.£ TD "O
•*— d ro C n3
>^ CD C O C
UJ Q. ,._ Q. L_
LO "O -f—' "O H •
o c: c: c: c:
C_> i—i i—i i—i i—i
^
\
CSJ
CNJ
a
oo
vO
CXD
VO
CD
"CD
e
ZJ
o
CD
to
CD
03
QQ
"ro
to
CD
"5
TD
O
E
LO
O
O
CD
.C
on
i—i
CNJ
CD
^
CD
CNJ
o
CD
ra
LO,
CNJ
en
CN]
9
01 x $
39
-------
Curve 6^7201-A
10
I
8
sO
O
i—t
X
tst
O
O
CO
CD
C
CD
03
CD
1750T
635 MW
\ 6
Cost increases (~$ 160,000 for basic design
due to shift from 318 MW
shop-assembled to
field-erected vessel
at velocities <8 ft/ sec
1407
-------
It is important here to understand why increasing the
fluidizing velocity increases the steam generator cost at the 318 MW size
and produces a minimum at the 635 MW size. To illustrate the point better,
cost reduction due to a decrease in module diameter and cost escalation
due to an increase in module height for a four-module design were shown
in Figure 23 for the design bed temperature at 1750°F. Increasing the
fluidizing velocity escalates the steam generator cost almost linearly
from the basic design point due to an increase in module height (curve 2).
At the same time, the cost decreases because of a decrease in module
diameter; however, the decrease is much more gradual and levels off
at a higher fluidizing velocity (curve 1). This is because the bed area
alone occupies less than 40% of the total cross-sectional area of a
pressurized module. The remaining area is required for piping and
headers, and this space is relatively unchanged at a specific plant size
even though the bed area is reduced to increase the fluidizing velocity.
Thus, the steam generator cost increases with an increase in fluidizing
velocity. At the 635 MW size, however, the cost reduction due to a decrease
in module diameter is larger during the initial deviation from the basic
design point and thus creates a minimum (Figure 23).
Another approach for analyzing the effect of fluidizing
velocity would be to change the number of modules as well as the module
diameter. Based on turndown consideration, it is preferable to have a four-
module-design; however, if a three-module design showed a substantial
saving with negligible effect on turndown capability, it would be a
better choice. Changing the fluidizing velocity will change the total
bed area required for each functional bed, but the total bed volume for
each functional bed will remain constant once the tube size and spacing
are fixed. Thus, a design with a smaller number of modules would
require a larger module diameter at the same design fluidizing velocity.
An estimation can usually be performed to evaluate the relative economy
between these two designs. For example, consider a four-module and
41
-------
Curve 647185-A
2.5
1" 0. D. Tubes at
S 1-5
X
•w-
v/j
O
O
oo 1
0.5
0
Bed Temperature
i
1750°F
i
Curve
Tube Pitch/Diameter Ratio
Basic Design
Basic Design
635 MW <
Basic Design
318 MW
Curve 1
Curve 2
1
2
Cost reduction due to decrease in module
diameter
Cost escalation due to increase in module
height
0
5 10 15
Fluidizing Velocity in Preevaporator, ft/sec
Figure 23: Dependence of the Shell Cost on the Fluidizing Velocity
42
-------
a three-module design at the same design fluidizing velocity and with the
same bed depths. Since the bed volume of each functional bed is constant
at fixed tube size and spacing for both cases, we have
if the bed height is assumed constant for both cases and the bed area
occupies a fixed percentage of the total cross-sectional area of the
module, where D, and D are the respective module diameters for the
four-module and three-module designs. Since the shell cost is dependent
on the vessel diameter (Figure 21) , the relative advantage of these two
designs will depend on the plant size in question. For example, for
D = 12.5 ft, D can be calculated from Equation 1 to be 14.4 ft. The
fi
shell cost can be found from Figure 21 to be $0.94 x 10 for the
four-module design and $0.79 x 10 for the three-module design. If the
fabrication cost of the module internals is similar in both cases, the
three-module design will have a slight economic advantage; however, this
advantage becomes progressively smaller because of the rapidly increasing
shell cost at large module diameters and the rapidly decreasing portion
of the design which could be shop-fabricated. It is estimated that the
largest module diameter which is still economical for the three-module
design is ^ 17 ft. This conclusion is based on the assumption that
boiler turndown is not a problem. If the module diameter is in a range
where the module can be shop-fabricated and transported by railroad
(i.e., < 12 ft), designing for the maximum shippable module will
have definite advantages, provided that turndown is not a problem.
It is concluded from this analysis that for a plant size
around 300 MW, the module diameter should be the largest within
shipping limitations (12 ft for railroad transportation) ; at a 600 MW
plant size, an optimum fluidizing velocity exists and it should be
found for each capacity. However, the cost deviation from that of the
optimum design is less than $1.00/kw (Figure 22). Of course, decreasing
43
-------
the bed area may reduce the number of feed points, but the saving is only
^ $0.10/kw. Although the effect of fluidizing velocity on the steam
generator cost is presented primarily on the basis of the basic design
2
conditions, i.e., the bed-tube heat transfer coefficient = 50 Btu/ft -hr-°F,
the trends would be the same for a bed-tube heat transfer coefficient at
2
35 or 75 Btu/ft -hr-°F. Change in tube size and spacing may change the
slope of the curves or alter the minimum in Figure 17, but the conclusions
will remain the same.
The effect of changing the fluidizing velocity on the bed-tube
heat transfer coefficient, combustion efficiency, and total particulate
carry-over is not taken into account in this analysis because of a lack
of accurate quantitative data. The effect of dust loading and particle
size distribution on the cost of the particulate removal system is
evaluated separately.
2.2.3 Excess Air
A change in design excess air will affect the cycle efficiency
and the cost of the boiler module, the steam- and gas-turbine equipment,
and the particulate removal system. In order to quantify the effect of
excess air on the total boiler cost, the air/fuel ratio is allowed to vary
but the total fuel input is kept constant. To simplify the analysis,
other parameters — bed temperature, tube size and tube spacing, bed-tube
heat transfer coefficient, fluidizing velocity, and number of boiler
modules — are held constant at the basic design values.
Thus, when excess air is increased beyond the design value, the
bed area has to be increased if the fluidizing velocity is kept constant.
When the bed area is increased, the module diameter has to be increased
as well; however, the module height is decreased because of a decrease in
bed depth (Figure 24). Increasing air input to the bed will also increase
the amount of heat carried out from the bed by air and reduce the total
heat transferred in the bed. At a constant bed-tube heat transfer
coefficient, the total heat transfer surface requirement is reduced. At
44
-------
ainpow uj uojpnpay
c
O>
QJ
Z3
•o
o
I/)
ft?
o
o>
•4-^
o>
E
o
c
• —
0)
HH <
CNl
O>
Z3
O"
ia a|npoy\|
45
-------
constant tube size and tube spacing, the bed depth is also reduced, as
is the total module height. The bed depth requirement at different air-
flow rates is presented in Figure 25. At 100% excess air, a reduction
of > 40% in the heat transfer surface and a reduction of > 30% in the
module height is possible. The bed depths for different functional beds
are reduced to ^ 4 ft, which will decrease the pressure drop through
the beds and increase cycle efficiency. The module diameter, in turn,
is increased from the original 12-ft inside diameter to more than 16 ft
for an 80 MW module. Transferring all these changes into economics, an
increase in excess air can reduce the boiler cost up to ^ $0.60/kw, as
shown in Figure 26. Cost reduction due to the heat transfer surface
increases continuously with respect to excess air because of the decrease
in the total amount of heat transferred in the bed. Cost reduction due
to the pressure shell increases first because of a reduction in module
height and then decreases because of the increase in module diameter.
Increasing the excess air will increase the overall plant
efficiency as shown in Table 9. Larger gas turbines or additional gas
turbines are needed to handle the increased mass flow of gas. If
additional units are used, the increase in the gas-turbine equipment cost
is shown as a cost adder (Figure 27). This does not account for cost
reductions which can be realized by going to larger turbine capacities.
This cost increase in gas-turbine equipment is partially offset by a
decrease in the steam-turbine equipment cost also shown in Figure 27. The
major equipment items taken into consideration in this analysis include
gas-turbines with external manifolds, a steam-turbine system, circulating
water and condensing systems, a feedwater system including station
piping, and stack gas coolers.
Instead of keeping the fluidizing velocity constant, the bed
area and module diameter can be kept constant, allowing the fluidizing
velocity to increase with excess air. In this case, the cost reduction
in the heat transfer surface will be similar, but the cost reduction in the
46
-------
Curve 6^9018-A
16
14
12
10
I 6
0
I
I
I
0 10 20 30 40 50 60 70 80 90 100 110
Excess Air in the Bed, %
Figure 25: Bed Depth Requirement at Different Airflow Rate
47
-------
.7
.6
s .5
c.4
o
OS
3.2
.1
Curve 6^9019-A
CD
C
"T
_
I
o
t_3
10 20 30 40 50 60 70 80 90 100
Overall Excess Air (including air from CBC), %
110 120
0
Fiqure 27: Cost Escalation in Gas-Turbine Equipment and Cost Reduction in Steam-
Turbine Equipment at Different Airflow Rates
48
-------
00
w
H
pH
ts
O
pTj
JV*
M
W
! 1
CQ
<;
w
,-4 H
n <1
-<
u
£PJ
w
M
Q_)
M
tM
pL(
M
cu
C!
•H
H | o
O M
• H
O
3 w
cd
O
cd
o cu
•r-l 4-1
•U CO
cd >-.
& H
0
1
cu g
•u §
CO
cu
C
•H
J_J
3 &
H a
CO
ctt
o
CU
cd i§
P^ r^
•^-^
4-> 3
Cd 4J
CU pq
4-)
C
cd &
Pn
O
o
CO
•H
•H
cd
CO
w
cu
o
X
w
cd
49
-------
shell will continuously increase with increasing excess air and not go
through a maximum (Figure 26). The cost reduction in heat transfer
surface and pressure shell at 100% excess air in this case (with 10 to 15
fps fluidizing velocity) is estimated to be ^ $1.00/kw. However,
increasing the fluidizing velocity to larger than 15 ft/sec may be
impractical in this design approach.
Increasing the excess air will decrease the total heat transfer
surface required in the fluid bed until no boiler tube surface will be
required at an excess air of approximately 300%. In this case the power
system would become a combined cycle plant,with the gas to the turbine
expanders supplied from a coal-fired, adiabatic combustor. The heat
recovery boiler would probably be unfired. This system concept differs
significantly from a pressurized fluidized bed boiler power plant
concept. For example, the boiler becomes an adiabatic combustor,
particulate removal equipment costs increase significantly because of
the increased gas flow, gas piping costs increase, and the gas-turbine
power contribution increases from ^ 20% up to ^ 70%. An economic
analysis and assessment of this system has been made in Section 6. The
heat rate for the adiabatic combustor plant is projected to be 100 to
500 Btu/kWh (depending on the gas-turbine inlet temperature) greater
than for the pressurized boiler plant.
Increasing the excess air will provide more flexibility in
turndown. At 100% excess air, an additional ^ 10% load reduction is
possible, as compared with operation at 10% excess air. This means a
boiler designed at 1600°F and 100% excess air will be able to meet a
4:1 turndown requirement. An adiabatic combustor system should extend
the turndown capabilities.
More discussion on excess air- and gas-flow rate appears in the
section on particulate removal.
50
-------
2.2.4 Operating Pressure
The full load design pressure is 10 atmospheres. When the
design pressure level is changed with the other operating parameters
constant, the gas density will change in proportion to the pressure, and
the volumetric flow will vary accordingly. Therefore, the bed cross-
sectional area will have to be changed to maintain constant fluidizing
velocity, and the bed depth will have to be changed to maintain
constant bed volume. The heat transfer coefficient may change because
of changes in the quality of fluidization.
The changes in volumetric flow and in gas density will affect
the design of the particulate removal equipment. Gas-turbine cycle
efficiency is also dependent on the operating pressure. However,
analysis of the high-pressure fluidized bed boiler system indicates a
reverse direction in pressure level effect. Auxiliary equipment such
as the coal and dolomite feeding systems are pressure dependent as well.
The relative importance of these factors with respect to operating
pressure is analyzed in the following paragraphs.
First consider the boiler module alone. At constant fuel
feed rate and excess air, increasing the operating pressure will
decrease the gas volumetric flow rate. There are two approaches in
designing the boiler modules: keep the bed area constant and let the
fluidizing velocity change with the operating pressure; or keep the
fluidizing velocity constant and change the bed area according to the
operating pressure. The incremental module cost for the constant bed
area case is the cost of reinforcing the pressure shells, since for a
constant overall heat transfer coefficient and a specific design bed
temperature the total heat transfer surface in the bed is constant.
This amounts to $0.20/kw and $0.30/kw for operating pressures of 15
atmospheres and 20 atmospheres respectively at 300 MW nominal plant size.
At 600 MW plant size, the respective cost increments are $0.10/kw and
$0.20/kw. The credit of decreasing particulate carry-over by operating
at lower fluidizing velocities is not taken into account. If the fluidizing
51
-------
velocity is maintained constant, an increase in operating pressure will
require a decrease in bed area. Since the total heat transfer surface
is constant, a decrease in bed area will require an increase in bed
depth at constant tube size and tube spacing and thus an increase in
the module height. In this case, the effect of the operating pressure
on boiler design and boiler cost while keeping the fluidizing velocity
constant is the same as the effect of changing fluidizing velocity with
the operating pressure constant. Thus, Figures 16 to 19, 22 and 23 in
the section on fluidizing velocity can also be applied for this approach
if the coordinate for the fluidizing velocity is changed to (fluidizing
velocity at basic design) x (new operating pressure/basic design pressure).
For a design bed temperature of 1750°F, the cost increments are $0.80/kw
and $1.90/kw for operating pressures of 15 atmospheres and 20 atmospheres
respectively at 300 MW plant size. At 600 MW plant size, the cost
increments are $0.20/kw and $1.20/kw respectively.
Comparing these two design approaches, the constant area
case is the less costly one. Moreover, if basic design bed area and
fluidizing velocity are maintained, increasing operating pressure will
mean a higher capacity shop-fabricated module (i.e., module diameter
< 12 ft). At 15 atmospheres operating pressure, a module of ^ 120 MW
capacity can be shop fabricated; at 20 atmospheres pressure, the maximum
size module that can be shop fabricated is ^ 160 MW. However, the module
height will be considerably increased because of the increase in total
heat transfer surface required. The total increase in module height
will depend on the heat transfer surface arrangement in the bed.
Increasing the operating pressure will reduce the size of the
particulate removal equipment because of the decrease in volumetric
flow rate and will reduce the particulate removal efficiency as well
because of changes in gas density and viscosity. The increase in operating
pressure will also require reinforcement of the containment vessel and
the piping and ducting. The savings in the particulate removal equipment
by operating at 15 atmospheres and 20 atmospheres are estimated to be
$4.0/kw and $6.0/kw respectively for a 300 MW plant size. The major saving
52
-------
comes from the secondary collectors, where maximum single unit capacity is
assumed restricted to 30,000 ACFM. An increase in operating pressure
will result in fewer units. The cost reduction from operating at higher
pressure will be even greater for larger plant sizes or at higher
design excess air.
Cycle optimization calculation was performed to evaluate the
effect of operating pressure. The parameters studied are: intercooled
and nonintercooled compression; gas-turbine compressor pressure ratios
from 10 to 30; cycle gas side pressure drops of 3 to 8%. The results
are summarized in Figure 23 which gives the plant heat rate for the
intercooled and nonintercooled cases. For the nonintercooled case,
the best efficiency is obtained at a pressure ratio of 10; and for the
intercooled case, the optimum pressure ratio is 15 but with a higher
heat rate and a more complex gas turbine. Thus, an increase in operating
pressure higher than 10 atmospheres decreases the overall plant efficiency.
This decrease, however, is small: ^ 0.2% at 15 atmospheres.
Weighing the above discussions, the only distinct advantage
for operating at pressures higher than 10 atmospheres is the capability
of shop-fabricating a large capacity plant, especially at higher design
excess air.
2.3 BOILER DESIGN
2.3.1 Heat Transfer Surface Configuration
In the 300 MW design, the heat transfer surface is provided
by serpentine tubes where horizontal sections are spaced as shown in
Figure 29. Tubes of 2 in.O.D. are used at water walls and are spaced
3-1/2 in. apart. Tubes for preevaporator and superheaters are 1-1/2 in. O.D.
and tubes for the reheaters are 2 in.O.D. The tubes can usually be arranged
in staggered or rotated diamond arrays; or they can be arranged in a
square or rectangular pitch (Figure 30).
53
-------
Curve
8900
8700
10
Boiler Press. = 2400 psi
Boiler Effic. =88.6%
15 20 25
Overall Pressure Ratio
30
.c
1
CO
First Comp P/R = 3.0
15 20 25
Overall Pressure Ratio
Figure 28: Plant Heat Rate versus Compressor
Pressure Ratio
-------
Preevaporator & Superheaters
Dwg.
1_
3"
T~
~>e
O
O
O
o
o
o
o
o
o
s~*
o
a
o
o
y>
2"0. D. Water
Wall Tubes
1J" 0. D. Tubes
Re heater
1
0 1
'
111
4"
2" 0. D. Water
Wall Tubes
2" 0. D. Tubes
Figure 29: Tube Arrangement in Basic Design
55
-------
Dwg. 6171A50
Staggered or Rotated Diamond Arrangement
0 ^9^
/~\ n /~\
O l" O jvO
o o o
O O
o o o
o o
Square or Rectangular Arrangement
o-
1
1
©-
o
o
H
\\l
i "
---©
O
o
o
o
o
o
Figure 30: Definition for Different Tube Arrangements
56
-------
For a staggered or rotated diamond arrangement, the available
heat transfer surface per unit bed volume can be expressed as
27T
'
where
H = horizontal spacing as shown in Figure 30, inches
V = vertical spacing as shown in Figure 30, inches
D = outside diameter of tubes, inches
H/D = horizontal tube pitch/diameter ratio
V/D = vertical tube pitch/diameter ratio .
Keeping the tube diameter (D) constant, the available heat transfer
surface per unit volume is inversely proportional to (H/D) • (V/D) .
For example, decreasing the horizontal and vertical tube pitch/diameter
ratio by one-half increases the available heat transfer surface per
unit bed volume four times. Keeping the tube pitch/diameter ratios
(H/D and V/D) constant, the available heat transfer surface per unit
volume is directly proportional to the tube diameter (D) . Decreasing
the diameter of the tube by one-half, for example, doubles the available
heat transfer surface per unit bed volume.
For a square or rectangular arrangement, the available heat
transfer surface per unit volume can be expressed as
HV <§>£>•>
With tube diameter (D) and horizontal and vertical tube pitch/diameter
ratios (H/D and V/D) constant, the available heat transfer surface per
unit bed volume in a square or rectangular arrangement is just one-half
that in a staggered or rotated diamond arrangement.
With Equations 2 and 3, the effect of change in tube size and
tube pitch/diameter ratio can be easily evaluated. Since the present
analysis is based on constant fuel feed rate and excess air, the total
57
-------
heat transfer surface requirement is constant at constant design bed
temperature and constant heat transfer coefficient. Change in tube size
and tube pitch/diameter ratio will only affect the design in bed depth
and module diameter. When tube size and tube spacings are changed from
those in the basic design, the bed depth in every functional bed deviates
from that in the basic design as well. The amount of deviation for
different tube sizes and tube spacings was calculated and expressed as
multipliers in Figures 31 through 35. To obtain the new expanded bed
depth, simply multiply the bed depth in the basic design by the appropriate
multiplier. However, in estimating the cost, a small correction is
sometimes necessary because when the bed depth is increased, more and
more heat transfer surface at the water walls is immersed in the bed.
To keep the bed temperature constant, the heat transfer surface in the
bed has to be reduced proportionally to keep the total immersed heat
transfer surface constant.
The bed depth correction is estimated as follows: the heat
transfer surface at the water wall enclosure per ft bed height can be
expressed as
(H.T.S.)W = (±^A + i^B) • — £ . (4)
W SW SW 12
The heat transfer surface as immersed tubes per ft bed height is
(H.T.S.)B - (r ' > ' B ' TT
where
A,B = bed dimensions, ft
S = water wall tube spacing, in.
D = water wall tube diameter, in.
W
D = immersed tube diameter, in.
H = horizontal tube spacing as defined in Figure 30
V = vertical tube spacing as defined in Figure 30.
58
-------
8" ,7" -
0123456789
Vertical Spacing, inches
Figure 31 Bed Depth Multiplier for Preevaporator, Super-
heater, and Reheater I Tube Diameter I/?'0 D.)
34567
Vertical Spacing, inches
Figure 32 Bed Depth Multiplier for Preevaporator, Super-
heater, and Reheater I Tube Diameter 1" 0 D. I
1234567
Vertical Spacing, inches
Figure 33 Bed Depth Multiplier for Preevaporator, Super-
heater, and Reheater (Tube Diameter 1-112'
O.D.)
234567
Vertical Spacing, inches
Figure 34 8ed Depth Multiplier for Preevaporator Super-
heater, and Reheater (Tube Diameter 2" 0. D. I
234567
Vertical Spacing, inches
Figure 35 Bed Depth Multiplier for Preevaporator, Super-
heater, and Reheater (Tube Diameter 3" 0 D )
59
-------
Dividing Equation 5 by Equation 4 gives the ratio of heat transfer surface
as immersed tubes to heat transfer surface of the water wall enclosure
per ft bed height, R.
S D
R = 24 • ' ' (iT> <6>
w
Once R is known, the bed height correction can readily be made, but
the correction is usually less than 5% of the bed depth in question.
This redistribution of heat transfer surface also requires a
redistribution of heat duties because now more heat is transferred to
the water walls. Thus water walls, in addition to the evaporation duty,
may take up some superheating duty. The amount of redistribution of heat
duties depends on the amount of change of bed area from that in the
basic design. The smaller the bed area,the smaller the ratio of heat
transfer surface in the bed per unit bed height to immersed heat
transfer surface at water walls per unit bed height and, thus, the higher
the amount of heat duty redistribution.
The effect of the tube pitch/diameter ratio on the cost of the
steam generator (not including erection) for a constant 12-ft module
diameter was evaluated for three different tube sizes and tube spacings
with respect to change in the design bed temperatures; the results
are plotted in Figure 13. Curve 2 represents the estimated cost for a
staggered arrangement of 1-in. O.D. tubes at H = 4 in.and V = 2 in. (see
Figure 30 for definition) in all beds. Curves 3a and 3b are for a
staggered arrangement of 2-in.O.D. tubes at H = 8 in. and V = 4 in. in all beds,
Some interesting trends are present if these results are compared to
the steam generator cost for the basic design. Ignore curves Ib and 3b
for the time being because the assumption of unrestricted maximum
allowable bed depth is not considered reasonable. Then,
• Decreasing tube size and tube spacing increases the
steam generator cost at design bed temperatures
above ^ 1520°F (curve 2). A further decrease in bed
60
-------
temperature necessitates splitting the reheater
bed in the basic design into two beds and sub-
stantially increases the steam generator cost of
the basic design.
• Increasing tube size and tube spacing also increases
the steam generator cost (curve 3a) .
The reasons for these results are as follows:
• When tube size and tube spacing are decreased,
more heat transfer surface can be immersed in a unit
bed volume, which results in lower bed height and
module height; thinner wall tubes can be used, resulting
in lower tubing cost. These are positive advantages.
• Smaller tube size and tube spacing, however, increase
the amount of tube bending and tube welding required.
This is because more tubes of smaller diameter are
required to carry the same water/steam load at a
constant flow rate in the tube. Fabrication cost
as a function of tube wall thickness is not taken
into account because not enough information is
available. Pumping costs, a part of total operating
cost, are not included here.
The balance of these two factors determines the total steam generator
cost.
Since the tubing cost constitutes only about 20% of the cost
of the pressure parts, the increase in fabrication cost is more
important. This can best be illustrated in Figure 36 where the component
costs are plotted against the available heat transfer surface per unit
volume which relates to the tube pitch/diameter ratio. The shell cost
increases almost linearly with decreasing heat transfer surface per
unit bed volume. The cost of tubing, headers, downcomers, and risers
2 3
is almost constant at heat transfer surface larger than 6 ft per ft
61
-------
Curve 647190-A
X
•w-
o
O
0
Bed Temperature 175Q°F-, Max. Allowable Bed Depth 20ft
Bed-Tube Heat Transfer Coefficient 30 Btu/ft? - hr - °F
Total Steam
Generator Cost
Tube Bending,
Welding, Water
Walls Fabrication
Drafting,
Home Office etc
Shell
Tubing, Headers,
Down com e7s, Risers
02 4 6 8 10 12 14
Available Heat Transfer Surface/Unit Bed Volume, ft2/ft3
Figure 36: Change of the Steam Generator Cost with Heat
Transfer Surface per Unit Bed Volume (not
including erection)
62
-------
bed volume, but increases rapidly at heat transfer surface lower than
2 3
'v 6 ft /ft bed volume (which corresponds to the use of a tube diameter
of 1-1/2" O.D. or larger). When larger tubes are used, the minimum
wall thickness increases rapidly and so does the tubing cost which
contributes most of the cost escalation at lower heat transfer surface
per unit bed volume. However, the cost of the tube bending, tube
welding, and water-wall fabrication increases steadily with an increase
in heat transfer surface per unit bed volume. The balance of all these
2
factors create a minimum in total steam generator cost at about 6.5 ft
3
heat transfer surface per ft of bed volume. This can be achieved by
arranging 1-in. O.D. tubes at H = 4 in. and V = 3 in. Fortunately, the tube
size and tube spacing used in the basic design is very close to this
actual minimum. Clearly, there are different minimums at different
design bed temperatures. An optimum design for a specific operating
condition requires a separate evaluation. However, this optimum design
point is not as critical as may be generally conceived. For example,
at a design bed temperature of 1750°F (Figure 36), the difference in
the steam generator cost between the optimum design and other designs
2 3
is within $1.00/kw for available heat transfer surface of 3 to 11 ft /ft
bed volume, which covers the three tube sizes and tube spacings in the
current evaluation.
The maximum allowable bed depth is a more important variable.
The steam generator costs are also plotted against the maximum allowable
bed depth at constant design bed temperature and constant bed area in
Figures 37 to 40. Although the steam generator of 1-in. O.D. tubes at
H = 4 in.and V = 2 in.is not economical when compared to the basic design, it
becomes progressively more attractive at a lower maximum allowable bed
depth. At maximum bed depths lower than 14 ft (Figure 37), a steam
generator using 1-in. O.D. tubes is actually cheaper than the basic design
by up to ^ $3.1/kw. At a bed temperature of 1407°F, where more heat
transfer surface has to be immersed in the bed, the advantage of using
smaller tube size and tube spacing is even clearer (Figure 40). At a
63
-------
Curve 6^7191-A
10
Curve No. Bed Temperature Tube Pitch/Diameter Ratio
1 1750°F li"0. D. atH=7",V=3"in
preevaporator and S.H.
2"0. D. atH=7", V=4" in
op re heater
2 1750°F 1"0. D. atH=4", V=2"in
all beds
3 1750°F 2"0 D atH=8", V=4" in
^o all beds
p^*i
X , Bed-Tube Heat Transfer Coefficient 50 Btu/ft2 - hr - °F
o
8 3
1 I ^2
3
CO
CD
0, 1 1 1
0 10 20 30
Maximum Allowable Bed Depth, ft
Figure 37: Dependence of the Steam Generator (318 MW) Cost on
the Maximum Allowable Bed Depth (not including
erection)
64
-------
*"£
I *|
,i It
'a, "> =•
S 2 E
i r'i
S.-K-S
O> o £Z
01 x $ ')soo
.s
E S CD
s g< §
O -^^ '^ ^
?Q QJ l/l 3 u
I 11 «
< S " 1
'isoo J01BJ9U89 meats
^ e
5 Bf
E S 5
c JS
E
o 'x
.9
»j
»
01 X $ '1S03 JO(BJ3U89 U1B9)S
65
-------
maximum allowable bed depth of 10 ft, a $3.1/kw saving over the cost of
the basic design is realizable by using 1-in.O.D. tubes at a design bed
temperature of 1750°F and a $3.3/kw saving at a design bed temperature
of 1636°F.
In this cost evaluation, the credit for using the thinner wall
tubes in smaller tubes was also taken into consideration. Minimum wall
thickness was calculated for each design bed temperature and was
summarized in Tables A-l and A-2 of Appendix A for both 1-in.O.D. and 2-in.
O.D. tubes.
In conclusion, although the use of smaller tube spacings will
increase the amount of heat transfer surface immersed in a unit bed
volume, the economy is not always favorable because the accompanying
increase in cost for tube bending, tube welding, and fabrication some-
times overtakes the savings in tubing cost and module height. However,
smaller tubes and tube spacings do show more advantages when the maximum
allowable bed depth is limited. Moreover, there is a definite minimum
when cost is plotted against the heat transfer surface per unit volume.
This can be utilized to find the optimum tube size and tube spacings.
In some cases, the tube pitch/diameter ratio effect on the
bed-tube heat transfer coefficient has to be taken into account in
evaluating the optimum tube size and tube spacings. Data indicate
346
that smaller tube spacing tends to decrease the heat transfer coefficient ' '
as shown in Figures 41 and 42. Lowering the pitch/diameter ratio from
8 to 2 results in an increase in the heat transfer surface per unit
volume of bed by a factor of 16, and decreases the heat transfer
coefficient by only 18% (Figure 41). Thus, the major consideration in
determining the pitch/diameter ratio is not the heat transfer coefficient,
but the compactness and the cost of the boiler design. There is a gap
between tubes, however, depending on the particle size, below which a
sharp drop of heat transfer coefficient will occur (Figure 42).
Determination of this minimum gap requires further experimental studies.
66
-------
Curve 643505-A
50
c 40
ZJ
h—
CO
•5 30
20
10
-a
CD
co
60 mm
Tubes
o
A
- 0
35 mm
Tubes
•
T
A
No. of
Rows
2
2
3
3
4
Reference: Highley et al(3)
Pitch/Diameter
Ratio
2
4
4
6
Pitch/Diameter Ratio
Figure 41: Change of Bed-Tube Heat Transfer Coefficient with Pitch/Diameter Ratio
Curve 643509-A
50
40
CD
£ 3°
1
O
20
10
60 mm
Tubes
A
0
35 mm
Tubes
T
A
No. of
Rows
2
2
3
3
Reference: Williams et al m , (4)
Pitch/Diameter
Ratio
2
4
4
6
2468
Narrowest Gap Between Tubes, inches
Figure 42: Change of Heat Transfer Coefficient with Tube Spacing
10
12
67
-------
Although this conclusion is drawn from the experimental evidence at
atmospheric pressure, it should be qualitatively valid at pressure.
Heat transfer coefficients for the tubes at different bed depths may
also be different for the pressurized conditions, but unfortunately
there is no quantitative information available.
An added advantage to the use of smaller tube sizes and tube
spacings is that lower bed depth tends to decrease the pressure drop
through the fluid beds and thus increases the gas-turbine efficiency
(curve 2, Figure 15).
2.3.2 Heat Transfer Coefficient
In the basic design, the overall heat transfer coefficients
2 2
assumed are 47 Btu/ft -°F-hr for the preevaporator, 45 Btu/ft -°F-hr
2
for the superheater, and 43 Btu/ft -°F-hr for the reheater. The bed-
2
tube heat transfer coefficient is assumed to be 50 Btu/ft -°F-hr for all
beds. When the bed-tube heat transfer coefficient is changed, the
overall heat transfer coefficient will be changed as well. This leads
to a change in the total heat transfer surface requirement and the bed
depth. The heat transfer surface and the bed depth multipliers for
preevaporator, superheater, and reheater were calculated and are shown
in Figure 43. The bed depth thus obtained will have to be adjusted,
because when the bed depth increases, the heat transfer surface, as
water walls immersed in the bed, increases as well. To keep the bed
temperature constant, a corresponding decrease in heat transfer surface
in the bed (excluding water walls) is necessary in order to maintain
constant total heat transfer surface immersed in the bed. Equation 6
can be used for this correction.
A change in the heat transfer coefficient will also change the
tube metal temperature. A change of tube material may be necessary in
some cases. The design metal temperature is assumed to be the maximum
outside tube wall temperature based on the minimum permissible wall
thickness. The minimum wall thickness for bed-tube heat transfer
2
coefficients of 35, 50, and 75 Btu/ft -hr-°F is presented in Appendix A.
68
-------
-4-
vD
O
o
-»—>
CO
a
CO
CD
CD
i_
Q_
,0
i_
CD
13
S3
CD
O
O
CD
C
CD
CO
i_
CD
So
CD-
CD
CD
CO
CD
CO CD
CD -C"
O CD
co rv
*- T3
3 C
CO CO
CD v_
**— CD
OO ! .
C CO
CO CD
co
CD
CD
CD
CD
o
69
-------
Taking into consideration the aforementioned factors, the
total steam generator costs at bed-tube heat transfer coefficients of
2
35 and 75 Btu/ft -hr-°F were projected for different tube pitch/diameter
ratios at different design bed temperatures (Figures 44 and 45). At a
2
low bed-tube heat transfer coefficient (35 Btu/ft -hr-°F) where a large
amount of in-bed heat transfer surface is required, a boiler with
smaller tube size and tube spacing is much more economical than the one
with larger tube size and tube spacing. Savings up to 20% of the total
steam generator cost are realizable if the bed area is constant and the
maximum allowable bed depth is 20 ft. If the maximum allowable bed
depth is less than 20 ft, the saving will be even larger. At a bed-tube
2
heat transfer coefficient of 75 Btu/ft -hr-°F, where the heat transfer
surface requirement is substantially reduced, smaller boiler tubes and
spacings do not have a clear advantage (Figure 45). If the bed-tube heat
2 2
transfer coefficient is 35 Btu/ft -hr-°F, rather than 50 Btu/ft -hr-°F
as assumed in the basic design, the steam generator cost will increase
by $2.7/kw at 1750°F design bed temperature and by $5.4/kw at 1600°F.
If 1-in.O.D. tubes are used at H = 4 in. and V = 2 in.,the cost escalation
would be $2.0/kw at 1750°F, $3.4/kw at 1600°F, and $6.1/kw at 1400°F
(Figure 44). If the heat transfer coefficient is increased to 75
2
Btu/ft -hr-°F, the reduction in steam generator cost from that of the
basic design is only marginal, ^ $1.0/kw at 1750°F (Figure 45).
Figures 46 and 47 show the effect of the bed-tube heat
transfer coefficient on the steam generator cost at constant bed
temperature. An increase of the bed-tube heat transfer coefficient
2
from 50 to 75 Btu/ft -hr-°F (a 50% increase) decreases the steam
generator cost by ^ 10% (^ $1.40/kw). A decrease of the bed-tube heat
2
transfer coefficient from 50 to 35 Btu/ft -hr-°F (a 30% decrease)
increases the cost by ^ 20% (^ $2.80/kw) (Figure 46). The curves start
to level off at higher bed-tube heat transfer coefficients. Thus, a
further increase in the bed-tube heat transfer coefficient larger than
2
about 75 Btu/ft -hr-°F does not affect the cost substantially. However,
70
-------
010>"lf ll II
H S »a OB °a
i! .H «r so so
m £ M II II
£g'
f to t5 °
Q" Q ci |_
o
is
8.
nOI x *
B ra
, ci ci
5 O O *|
1 °
I 5
I— Oj
| l^
I ^ =
I* II
71
-------
Curve 647208-B
o
x 4
03
Curve
1
2
3
Tube Pitch/Diameter Ratio
Basic Design
1-0. 0. atH = 4"&V = ?'
2"0. D. atH=8"&V/4"
Bed Temperature 1750°F
IV\aximum Allowable Bed Depth 20 ft
Bed Area
35ft'
10 20 30 40 50 60 70 80 90 100
Bed-Tube Heat Transfer Coefficient, Btu/ft2 hr - °F
110 120
Figure 46: Effect of Bed-Tube Heat Transfer Coefficient on the Steam
Generator Cost (not including erection)
72
-------
a further decrease in the bed-tube heat transfer coefficient lower than
2
50 Btu/ft -hr-°F increases the steam generator cost rapidly, especially for
large tube sizes and tube spacings and at lower design bed temperatures
(Figures 46 and 47). A 40% increase in cost occurs when the bed-tube
2
heat transfer coefficient decreases from 50 to 35 Btu/ft -hr-°F at
design bed temperatures of 1636°F for 2-in,O.D. tubes at H = 8 in. and
V = 4 in.(curve 3, Figure 42). It is recommended that the steam generator
2
be designed at a bed-tube heat transfer coefficient of around 75 Btu/ft -hr-°F
if it is at all possible and avoid designing the steam generator at a
2
bed-tube heat transfer coefficient lower than 50 Btu-ft -hr-°F, especially
if large tubes and spacings are used and if lower bed temperatures are
employed.
To complete the evaluation, the cost information was also
prepared for other cases at different design and operating variables
to show the interacting effect of the tube pitch/diameter ratio, the bed-
tube heat transfer coefficient, the maximum allowable bed depth, and the
design bed temperature. The curves are presented in Appendix B.
Again, the maximum allowable bed depth turns out to be the
limitation of the steam generator design and cost, especially at a low
bed-tube heat transfer coefficient where a larger amount of heat transfer
surface is required in the bed. In this case, a smaller tube size and
tube spacing and a higher design bed temperature are preferred. At a
maximum allowable bed depth of 10 ft, the boiler designed at 35
2 2
Btu/ft -hr-°F costs $3.5/kw more than that designed at 50 Btu/ft -hr-°F
2
and $8.6/kw more than that designed at 75 Btu/ft -hr-°F at 1750°F
design bed temperature. At a design bed temperature of 1407°F, the
figures are $6.7/kw and $12.5/kw respectively. With 1-in.O.D. at H = 4 in.
and V = 2 in. the figures are $3.3/kw and $4.4/kw at 1750°F; and $2.8/kw
and $8.1/kw at 1407°F respectively.
Cost savings become smaller when the bed-tube heat transfer
2
coefficient is further increased over 75 Btu/ft -hr-°F. If the bed-tube
2
heat transfer coefficient is decreased to lower than 50 Btu/ft -hr-°F,
73
-------
Curve 647209-B
_- 4
o
o
o
•*— '
CO
o>
O
E 3
Curve
1
2
3
Tube Pitch/Diameter Ratio
Basic Design
1"0. D. atH=4"&V=2'
?'0 D atH=8"&V=?'
Bed Temperature 1636°F
Maximum Allowable Bed Depth 20 ft
Bed Area
35ft'
I
I
I
I
I
I
I
10
Figure 47:
20 30 40 50 60 70 80 90 100
Bed-Tube Heat Transfer Coefficient, Btu/ft2 - hr - °F
110 120
Effect of Bed-Tube Heat Transfer Coefficient on the Steam Generator
Cost (not including erection)
74
-------
the steam generator cost increases rapidly. Thus, it is recommended that
the bed-tube heat transfer coefficient be kept at higher than 50
Btu/ft2-hr-°F, and
design conditions.
2 2
Btu/ft -hr-°F, and preferably around 75 Btu/ft -hr-°F, at the current
2.3.3 Tube Materials
The basic design specifies conventional boiler tube material,
with SA-210-A1 for tubes in the preevaporator; SA-213-T2 for tubes in
the lower superheater; SA-213-T22 for tubes in the water walls, upper
superheater (lower loops), and reheater; and SA-213-TP304H for tubes in
the upper superheater (upper loops) . Changes in design bed temperature,
bed-tube heat transfer coefficient, or steam temperature may require
higher grade tube materials. This would not, however, substantially
affect the steam generator cost, because the tubing cost alone
constitutes only ^ 10% of the total steam generator cost (^ $1.40/kw).
Higher fabrication cost for higher alloy material may increase this
cost slightly. Nevertheless, the total boiler cost is not expected to
increase significantly because of a change of tube material, unless the
operating bed temperature and heat transfer coefficient are drastically
changed. The selection of tube materials at different design bed
temperatures and bed-tube transfer coefficients is presented in Appendix A.
2.3.4 Module Capacity
Boiler modules can be shop fabricated, partially shop fabricated,
or field erected, depending on the size. Modules up to 12 ft in diameter
can be shop fabricated; those up to 17 ft in diameter can be partially
shop fabricated (the boiler internals shop fabricated,the pressure shell
field erected). The plant concept for a given capacity can be either
multiples of shop-fabricated modules, partially shop-fabricated modules,
or field-erected modules.
75
-------
The boiler plant equipment cost is different for each case.
The steam generator cost will depend on operating conditions and design
variables such as design bed temperature, tube size and tube spacing,
maximum allowable bed depth, etc. Auxiliary equipment will also be
affected: coal feeding, limestone or dolomite feed and withdrawal,
particulate removal, steam piping, boiler feedwater system, etc.
The evaluation of these approaches is based on Figure 8 where
the cost variation of the pressure parts, shell, subcontracted
and contracted equipment, drafting and home office, and erection with
respect to the plant size is presented. The resulting costs (including
erection) are presented in Figure 48 for a design bed temperature of
1750°F, and in Figure 49 for a design bed temperature of 1636°F at a
maximum allowable bed depth of 20 ft. The results show that at a plant
size larger than ^ 340 MW, partially shop-fabricated four-module plants
with maximum shop fabrication of the pressure parts are more economical
than a collective multiple of the largest shop-fabricated modules of the
same plant capacity. Changing the tube size and tube spacing does not
affect this conclusion (Figures 48 and 49). A change in heat transfer
coefficient should not produce a different conclusion because the cost
escalation due to the addition of a single module is greater than it
would be if the existing modules were simply enlarged. However, this is
no longer true when the shell size is larger than ^ 17 ft, because the
degree of shop fabrication of the pressure parts decreases and the
steam-generator cost again increases at a much higher rate.
When the maximum allowable bed depth is decreased from 20 ft,
the cost saving of the four-module partially shop-fabricated plant is
expected to increase. The splitting of the beds in the shop-fabricated
module means an increase in module height. In the partially shop-
fabricated module, however, the splitting of beds can be avoided simply
by enlarging the module diameter. From the discussion on Figure 19, the
latter means a more economical alternative except when the module
diameter is increased beyond ^ 17 ft.
76
-------
01 x $ 'JSOQ jojBjeuao
o
o
CD
c
CD
O
e
ro
CD
c -^
CD C
•a ta
-
CD
t_
3
CD
OT x $ 'JSOQ
77
-------
The optimum design variables (fluidizing velocity, excess air
and pressure) in relation to module capacity are discussed in their
respective sections.
2.4 PARTICULATE REMOVAL SYSTEM ECONOMICS
The primary variables taken into consideration in analyzing
particulate removal system economics are dust loading, particle size
distribution entering and leaving the system, and gas flow rate. The
effects of boiler operating variables — bed temperature, freeboard
height, and superficial velocity; and design variables — tube pitch/
diameter ratio, and bed-tube heat transfer coefficient were also
evaluated.
2.4.1 Range of Dust Loadings and Particle Size Distribution
The cases evaluated are outlined in Table 10. Due to the
complexity of boiler design and operating parameters, a given dust
loading, such as one identical to that of the basic design, can be
produced by infinite combinations of these parameters (Figure 50).
The basic design is shown as a point which represents an 8.5% ash coal
with 6% carbon elutriation rate and 0.7% dolomite elutriation rate.
For the same coal, a 0% carbon elutriation rate will require a 5.3%
dolomite elutriation rate to produce the same dust loading. For a
coal of 13.4% ash, the elutriated ash alone will give the same dust
loading with no carbon or dolomite elutriated. Similar diagrams are
also prepared for the cases which double and triple the basic design
dust loading (Figures 52 and 53). The results of this parametric study
are presented in Appendix C.
2.4.2 Gas-Turbine Specification
A review of operating experience and an assessment of erosion
in gas turbines was prepared by Westinghouse under contract to the Office
78
-------
.2 c
•S'K
*8
51
.§£
1*)
E "^
o "f3
"f
c o
sr
& 3
JO % '
79
-------
Fluidized
Bed
Combustor
(FBC)
Ilg
IL
Carbon
Burn-up
Cell
(CBC)
IHr
Dwg. 6183A49
S.C. IIIg
II
P.C. = Primary Cyclone
S.C. = Secondary Collector
Fiqure 53: Flow Diagram for Particulate Removal System
80
-------
W
rJ
O
H
CO
fn O
O M
H
H P
CJ pq
W M
pq p^
Fn H
W CO
M
O P
M W
2! Csl
W W
H rJ
W O
P M
H
PM P-i
P
W
H
W
CO
W
CO
CJ
6
CU
4-1
cn
CO M
•H
cu cn
d a)
o p
rH
O
o
CU
N a
•H O
CO -rl
4-1
0) 3
rH J2
CJ -H
•H S-l
4J 4->
s-i cn
cfl -H
CM p
60
d cj
•H pq
rrj pt<
Cfl
o &o
rJ d
•H
4-1 >
cn cfl
3 cu
P r4
id
cu
cn 4J
0) cfl
co 3
Cfl rH
CJ Cfl
p»
W
CO
m
01
S-i
3
60
•H
Cl4
,_,
LO
CD
S-l
3
60
•H
Pn
O
4-1
S-l
cu
(4-1
cu
J_J
'
X_N
tn
CJ
d ^
60 CO
•H d
CO -H
CU cfl
*T3 i-<
60
O
•H P**
cn •
Cfl v£>
rH
CU
CO
Cfl
CJ
1
i — |
ft
3
O
S-i
CJ
00
LO
OJ
3
60
•H
CO
CU
rH CJ
0 pq
•H PM
4-1
r-l S
cfl O
ft M
14-1
j_(
O *&
M-t CU
4-1
rH Cfl
•H
cu s-4
> 4J
3 rH
0 CU
CU
3
cfl
d
60
•H
CO
cu
rrj
CU
rH
,0
3
0
CN
01
CO
CO
Q_}
|
rH
ft
3
0
S-l
CJ
00
in
CU
S-l
3
60
•H
pn
CO
CU
rH CJ
O pq
•H CJ
4-1
r-l g
CO O
ft S-l
14-1
S-l
0 T)
14-1 CU
4J
CM Cfl
•H
CU S-l
> 4-"
S-i 3
3 rH
O 0)
A^
CU
3
CO
d
60
•H
CO
0)
CU
rH
ft
•H
S-l
4-1
rO
CU
CO
cfl
CJ
1
rH
ft
3
O
^
O
^J-
LO
C1J
S-i
3
60
•H
fe
CU
3
CO
d
60
•H
CO
CU
0)
rH
ft
•H
S-i
4J
0)
cn
C8
CJ
I
rH
ft
0
S-l
o
m
m
CU
j-i
3
60
•H
Pt4
"s
CU
3
1
CO
d
60
•H
cn
cu
13
cu
H
ft
•H
S-l
4-1
in
cu
cn
cfl
CJ
I
rH
ft
3
O
S-i
CJ
CO
m
OJ
M
3
60
•H
P*4
CO
cu
rH O
0 pq
•H PH
4-1
H S
cfl O
ft r-l
<4-4
S-l
O T3
IM 0)
4-1
CN cfl
•H
CU S-l
^ 4-*
r-l 3
o cu
I
^
PH
CJ
CO
d ^
60 CO
•H d
cn -H
CU Cfl
T3 S-i
60
CJ
•H r^*
CO •
cfl vo
rH
CU
CO
cO
O
1
CN
ft
3
O
S-l
CJ
00
in
CU
S-l
3
60
•H
Pn
Cfl
cu
rH CJ
a pq
•H CJ
4-1
M £i
cfl O
ft S-l
14-4
S-i
o 'rd
14-1 0)
4-J
CO cfl
••H
CU S-l
> 4J
(-< 3
3 rH
o cu
s* —
CU
3
cfl
d
60
•H
cn
CU
CU
H
"3
o
Csl
CU
CO
cfl
CJ
1
CN
ft
3
0
S-i
o
00
in
cu
S-l
3
60
•H
Pn
-\
CU
3
cfl
d
60
•H
CO
CU
CU
rH
ft
•H
S-i
4J
00
cu
CO
cfl
o
1
CN
ft
O
S-l
o
81
-------
of Air Programs. Specifications for the fluidized bed combustion
system based on that study are:
• Dust loading less than 0.15 grains per standard
cubic foot
• Concentration of particles greater than 2 micro-
meters less than 0.01 grains per standard cubic
foot.
These design requirements will be updated as additional
laboratory test data and operating experience become available.
2.4.3 Effect of Dust Loading and Particle Size Distribution on Cost
The effect of dust loading and particle size distribution
leaving the boiler on the particulates going to the gas turbine for the
particle removal systems selected is shown in Table 11. The gas-turbine
specification is exceeded in several cases. In order to meet the speci-
fication, alternative particulate removal systems could be considered —
granular bed filters, electrostatic precipitators, ceramic filters, etc.;
additional mechanical collectors could be used in series; or the boiler
operation could be altered. The additional use of mechanical collectors
was the approach selected because the equipment is available and thus
provides the best cost data and because the effect of boiler operating
conditions on particulate emission is difficult to project. A summary
of the economic implications is presented in Table 12.
If the gas-turbine dust loading requirement is defined as
< 0.01 grains/SCF without reference to particle size, no centrifugal
separator presently available can meet this requirement within reason-
able cost in all cases discussed above. If such separators were available,
high temperature ceramic filters might have to be used. The possibility of
this application is being evaluated. In this respect, the particle size
distribution is a far more important parameter than the dust loading,
since the collection efficiency for the particle size smaller than 2y
82
-------
W
pq
Pi
o
fn
W
o
o
o
Q
>-i W
£ H
W W
C/3
co
W W
33 co
H W
< Pi
W W
C/l rH
O Q
H
to
P
O
CO
cu
u u
•H CO
4J \
cd 60
r\
^ 3-
0
m CM
60 C
c cd
•H 42
O ^
1-5 CU
60
4-1 J-i
05 cd
saving
, gr/SCF
h-3 CU
c
6U O
rj , |
•H O
cd u
J-I
4-1 Cd
CO T3
3 C
Q 0
a
rH CU
cd co
o cu
H fi
4-1
-a
cu
W 4-1
0) cfl
05 3
u co
W
VO
r^- <± o o O
rH CM ST rH Sfr
o o o o o
cd
rH CM Cn
O
<4H
4J
cd
4-1
CO
cd
cu
a
0
o
u
cd
•n
C
o
o
cu
CD
'c
o
0
cu
CO
cu
4-1
0
HH
o
a
cu
•rH
o
•H
UH
M— I
CU
fi
0
•H
4-1
O
cu
rH
r-H
O
O
i-H
cd
a
0
•H
4-1
O
cd
n
MH
cu
e
cd
CO
60
a
•rt
6
3
CO
CO
<
cd
o
•H
4-1
CO
•H
6
•H
ft
0
0
O
4-1
CO
_l
rl
c
o
4-1
A
e
CO
05
cd
05
•H
H
•
cu
c
o
4-J
en
•H
14_)
83
-------
W
e
QJ
4-J
cn
J^i Is
CO &
rH -CO-
CO
> QJ
O CO
01 QJ
C^ M
O
QJ d.
cfl
rH 4J
3 CO
O O
•H CJ
4-1
cfl
PM
C
O
•H
4-J
0 cfl
4-> CJ
•H
ft <4-(
3 -H
1 0
C OJ
CO ft
Q) CO
rH
C_> 0)
Cl
J-l 'H
QJ 43
p| j_j
4J 3
l-i H
3 1
E^ Cfl
(-1 CJ
o
y-i a)
-O QJ
Q) -H
QJ 43
a o
0)
CO
CO
U
1 1
1 1
CO
*T3
Q)
M
•H
3
cr1
0)
QJ
43
4J
O
C
£*•>
O cfl
C S
rH CM
0) QJ
CO CO
cO cfl
O U
1 1
rH rH
ft ft
3 3
O O
rl 1-1
a o
0 0
co in
• •
vO 00
CO
QJ
•H
OJ
cn
C
•H
QJ
C
0
rH
O
^
O
[>-,
S-i S-i
cfl QJ
T3 4-1
C rH
0 -H
0 4H
QJ
co -d
QJ
T3 43
C
O VJ
O CO
QJ rH
co d
CO cO
M
« 60
CO
QJ M
>. o
CO
QJ
CO
Cfl
U
1
rH
ft
3
o
M
o
o o
1 ro in
1 • •
vO 00
CO
0)
•H
^
OJ
CO
C
•H
OJ
fi
O
rH
CJ
r*"i
O
^*v
M V-i
ctf OJ
"^ 4-*
C rH
0 -H
a m
QJ
co 13
QJ
t3 43
a
0 M
CJ Cfl
CD rH
cn 3
d
cd cfl
}_i
•> 60
CO
O Q) M
a ?", o
o- m
QJ a)
CO CO
cfl cfl
U 0
1 1
rH H
ft ft
3 3
0 0
J-i M
o o
1
1
cfl
T3
QJ
}-i
•H
3
cr
QJ
QJ
43
4-1
O
c
K*"»
cfl
6
rH
QJ
CO
cfl
CJ
1
CN
ft
3
0
l-i
O
0 0
co in
• •
vO 00
CO
QJ
•H
£_j
QJ
CO
a
•H
QJ
c
o
rH
CJ
^»
CJ
^1
M M
CO QJ
T3 4->
C rH
O -H
O MH
Q)
CO T3
QJ
13 43
C
0 l-i
O CO
QJ rH
CO 3
a
cfl CO
}-i
»• 60
CO
QJ M
P"> O
CM
QJ
CO
cfl
U
1
CM
ft
3
O
M
O
o o
co m
• •
VD 00
CO
Q)
•H
^_j
QJ
CO
C
•H
QJ
C
O
H
O
£**i
O
t*^
1-4 1-4
cfl i O
CO
QJ
CO
CO
U
1
CM
ft
3
0
0
•
P!
0
•H
if
cfl
a
•H
LH
•H
a
0)
ft
CO
o
QJ
CO
0
rH
CJ
CO
•H
QJ
N
•H
CO
QJ
rH
o
•H
4J
J^
cfl
ft
4-1
3
43
fl
O
•rl
4J
Cfl
O
•rl
U-l
•H
a
QJ
ft
CO
a
cfl
4J
QJ
00
M
cfl
rH
cn
•H
60
pj
•rl
rrjj
CO
O
rH
jj
CO
j3
£j
CO
84
-------
II
decreases rapidly. Thus, the particle size distribution curves assumed
for the present evaluation (as shown in Figure 56) are conservative
because a large amount of fines is assumed to be present.
2.4.4 Effect of Gas Flow Rate on Cost
The effect of increasing the gas flow rate was evaluated for
four cases which correspond to the basic design flow rate, 30% excess
air, 50% excess air, and 100% excess air. The selection of first-stage
cyclones is based on the criterion of maximum efficiency at minimum
cost with a minimum cyclone efficiency of 85% (Table 13). The cost
increment for a higher gas flow rate is shown in Figure 57.
The first-stage cyclone cost includes not only the cost of the
first-stage separators supplied by Ducon but also the cost of the separator
pressure vessel and all of the gas piping from the steam generator outlet
to the secondary separator inlets. The costs for the pressure vessel
and gas piping from the steam generator to the first-stage separator is
lined with hard refractory, but the pressure vessel and the gas piping
from it are lined with stainless steel. In the other case, hard refractory
without an alloy liner was used throughout. At 100% excess air, the
increase is $1.70/kw and $2.40/kw for a hard refractory liner and a stain-
less steel liner, respectively. The major cost increase is from the
enlargement of pressure vessels and gas piping due to the higher gas
flow rate. The increase in separator cost alone constitutes only about
18% of the total cost increment.
The selection of the secondary collectors is shown in Table 14,
and the cost increase for the second stage is presented in Table 58.
The incremental cost for the second stage is more than twice that of
the first stage ($5.00/kw versus $2.40/kw at 100% excess air). This is
because model 18,000 is the largest cyclone now supplied by Aerodyne.
The capacity of model 18,000 with dirty gas as the secondary gas is
30,000 CFM. Any gas flow rate higher than that will require multiple
units with their individual pressure vessel. This would tend to increase
85
-------
n
rH
w
pp
"^3
H
w
H
^3
J-yij
JS
0
,-q
en
o
C-f
gj
Pd
P4
pn
fX4
M
Q
H
CO
PL]
&
O
t— 3
u
G
w
o
H
CO
1
H
CO
M
P^
O
55
O
M
H
QJ)
pLJ
h-J
w
CO
(-1
co
co
0)
o
w
pH
&"€
O
o
rH
CO
CO
0)
O
W
fr*S
o
in
j_,
•rl
a
o
o
o
o
n
SO
0
rH
O
O
O
*\
0
oo
o
o
o
*\
rH
O
o
o
M
CT>
„
0)
60
cfl
4-1
CO
4->
CO
•H
Pn
4-1
Cfl
S CU
O rH
pT] *^
Q
co 6
co
O !-i
0)
rH A
CO
4J pt)
a o
*
CJ
CT> rH
• •
o m
-sT CN OO
CN ro
• •
en in
<• rH 00
\o m
• •
oo m
-* rH 00
r^ m
• •
~* rH 00
Cfl
4-)
rC
•H
CU
cu !s
rH
3 >>
TJ ^
O
s ^
«_!••%
01 O >i
Q. • CJ
!3 C
9 a -H
a -H o
0 -H
rH "4-1
CJ P. MH
>. 0 W
O M
n c
14H O
o a> -H
J-l 4J
M 9 O
cu co cu
^3 CO rH
S cu T-H
3 M O
p^ p | QJ
•
rH
P.
3
O
0
14H
O
c
o
•H
4-1
3
_f*
•H
rJ
4-)
CO
•H
-a
cu
N
•H
CO
rH
O
•H
}_,
c8
0,
C
O
rrj
CU
CO
CO
.n
^
o
c
0)
•H
CJ
•H
4_l
4-1
CU
a
o
•H
4J
CJ
cu
rH
rH
O
Q^
cfl
86
-------
.-*-
rH
W
W
H
W
H
•«.
o
o
0
•s
CTl
m
I
a
o
o
0)
00
td
4J
CO
I
4-» 01
CO rH
rJ 3
•H ^3
pt< O
g
tJ
(d M
0)
& P-
o
rH |jT|
[*y.( ^
CJ
w ,
rH CD Cj
CJ "^ 0
r^S O 3
US £5
rH I-.
co cr«
rH f^
CO O\
rH P-
CO ON
H r->
ro CTi
td
4J
f^
00
•H
CU
rs
r*>
_rt
&Z
» »\
O r*^
O
s a
CU
d -H
•H 0
•H
• MH
O 'j j
o w
a d
o
CU -H
M ^ 1
3 0
co cu
CO rH
CU rH
VJ 0
PH O
P.
o
C
o
•H
4J
J3
•H
M
•U
CO
•H
CU
N
•r)
CQ
o
•H
td
P.
C
o
13
0)
03
n)
o
at
MH
cu
C
o
•H
J-l
o
01
o
u
87
-------
T
P.C.
Fluidized
Bed
Combustor
(FBC)
X
i
7
Carbon
Burn-up
Cell
(CBC)
i
\ ' '
P.C. =Pr
Dwg. 6183A82
s.c.
S.C. = Secondary Collectors of Same
Fractional Collection Efficient
Figure 54: Flow Chart for Group 1 - Case 4
Fluidized
Bed
Combustor
(FBC)
Primary Cyclones of Same
Fractional Collection Efficiency
Secondary Collector
Figure 55: Flow Chart for Group 1 - Case 5
88
-------
o
en
CO
E
O3
CD
uo
ns
O
"c
0>
CO
C
O
_ o **—
c
o
00
JOJ
o
CO
a.
CO
Q
03
M
CO
O>
\—
13
cn
LTv
LTV LTv LTv LTV LTS LTV LT\
r—I CVJ CO *sf Lf\ VO r-~
s
ca
O^
VO
O
OO
O
o
o
Aq% azj
89
-------
Curve 649027-A
o 0
o 2
"c
O)
E
o.
CD
o:
a
o
l
10 20 30 40 50 60 70
Excess Air in the Bed, %
80
90 100 110
Figure 57: Increase in First-Stage Cyclone Cost Due to Increase in Gas
Flow Rate (for 318 MW Plant)
Curve 649030-A
-w-
i_~
O)
to
o
o
"c
CD
9
o c
E
CD
o;.
o>
o
"•c
j_
j_
10 20 30 40 50 60 70
Excess Air in the Bed,
-------
the incremental cost of the second stage, but the rate of cost increase
slows down at excess air larger than 100%. At excess air larger than
100%, the rate of cost increase for the first stage speeds up. This is
because increases in pressure vessel size and gas piping diameter
increase the incremental cost rapidly at very high gas flow rates.
Taking into account the cost of gas piping from the second-
stage cyclones to the gas turbine, the total incremental cost at
different gas flow rates is shown in Figure 59.
Combining cost figures from Figures 26, 27, and 59, an increase
in total boiler cost of $9.6/kw to $10.2/kw is required to operate the
boiler at 100% excess air. This, however, does not take into consideration
the fact that the combustion efficiency in the primary beds will approach
100% at 100% excess air and that the carbon burn-up cell can be eliminated.
In addition, the particulate removal system would be much simpler and
the high excess air might also provide the necessary flexibility to
achieve plant turndown if the boiler were designed at lower bed temperatures.
An additional ^ 10% turndown capability is obtained by operating at 100%
excess air. The lower bed depths would also permit combining the two
superheater beds into one bed which would also reduce module height and
cost. Thus, it is concluded that if the carbon burn-up cell can be
eliminated, operating at 100% excess air may result in a lower energy
cost, with increased cycle efficiency and flexibility. This conclusion
is true at a 300 MW nominal plant size but is not necessarily true at a
600 MW plant size. This is because at the 600 MW plant size, the module
is 17 ft in diameter in the basic design. A further increase in excess
air requires either an increase in module diameter or in the total
number of modules. At 100% excess air, four modules of 23 ft in diameter
are required. Since the fabrication cost of the internals increases
rapidly at module diameters larger than 17 ft because of the rapidly decreasing
portion which can be shop-fabricated, an increase of $5.0/kw in boiler
module cost alone is conceivable. Adding the cost increase in the particulate
removal system and in turbine equipment, the total increase in boiler
cost is about $15.0/kw. In this case, operating at higher pressures may
be beneficial.
91
-------
I 7
03
8 5
o •>
S
-------
Instead of increasing the module diameter from 17 ft to 23 ft,
the number of modules could be increased but the module diameter kept
constant. At 100% excess air, seven modules are required. In this
case, in addition to module cost, individual particle removal equipment
has to be provided for each module; ducting and piping manifolds have
to be increased; the coal feeding system becomes more complicated; and,
above all, instrumentation and control have to be more sophisticated.
The total increase in boiler cost is estimated to be $20.0/kw to
$25.0/kw in this case. Consequently, even if the carbon burn-up cell
could be eliminated, the increase in cost and complication in control
do not clearly favor the operation at high excess air for large plant
size. A careful evaluation of overall design and control philosophy
should be done if operation at high excess air is to be attempted for
large plant size.
2.5 POWER PLANT
Alternative boiler design and operating conditions will have
three primary affects on the power generation equipment:
• Capacity of gas- and steam-turbine equipment
• Gas-turbine inlet temperature
• Ability to achieve higher steam temperatures and
pressures.
Since this analysis assumes a constant fuel rate, the capacity changes
are considered small except for the high excess air case. The effect
of capacity was considered with the excess air analysis.
2.5.1 Gas-Turbine Inlet Temperature
The design value for the gas-turbine inlet temperature was
1600°F for the base design of the pressurized fluid bed boiler. This
is well below the current state-of-the-art temperatures of 1800 to 1900°F
for utility intermediate load applications. The 1600°F level was
93
-------
established by assuming the flue gas leaves the bed at the 1750°F bed
temperature and that the temperature difference between the bed and the
gas-turbine inlet would be 150°F. Several factors may alter the gas-
turbine inlet temperature. These include:
• Temperature drop between boiler and turbine expander:
150°F was assumed but is probably excessive. A drop
as low as 50°F to 75°F may be achieved.
• Bed temperature. If the boiler design temperature
is changed, the turbine inlet temperature will
change. Sulfur removal considerations and ash
agglomeration will determine the maximum temperature.
• Combustion above the bed. Combustion has been
observed above the bed of a fluidized bed boiler
which increases the gas temperature 200 to 300°F.
Any combustion above the bed would increase the
gas-turbine inlet temperature. No combustion
was assumed in the base design.
• Modification of the cycle to provide for reheat
of the product gas from the boiler prior to the
gas turbine. One concept for doing this is
shown in Figure 60. Carbon carried out of the
primary beds would be gasified to produce a low-
Btu gas. The gas would be used in the second-
stage combustor.
Performance calculations were made to determine the effect of
a change in the gas-turbine inlet temperature on plant performance.
The results are summarized in Table 15. These results are for 17.5%
excess air and a boiler efficiency of 88.6%. A 200°F change in the
turbine inlet temperature will change the plant heat rate ^ 1% using
current technology.
94
-------
^D
•~D
<
Ol
O
-a
o
i_
n 1/1
co
.d °
E -53
O 3
O LL.
I
CD
4—'
03
"O
c
CO
to
-a
s ==
-»-• o
on on
r
i i
J*£
O
43
on
1
_o
i/i
J3
E
o
fU
o
o
1
u_
ol
CJ»
on>
1
•+-'
^
"c
CO
Q_
O)
o
Q_
_o
t/1
O
o
CD
GO
•a
M
S
•a
a>
M
13
1/1
CD
-------
TABLE 15
PERFORMANCE OF PRESSURIZED FLUID BED BOILER POWER PLANT
AS A FUNCTION OF GAS-TURBINE INLET TEMPERATURE
Gas-Turbine
Inlet Temperature
°F
1400
1500
1600
1700
1800
1900
2000
Plant
Output3
MW
625.6
634.9
644.1
644.8
645.3
641.8
636.2
Plant
Heat Rate3
Btu/kWh
9293
9157
9026
8921
8820
8773
8753
Fuel Burned ,
after Primary Beds
%
—
—
—
2.5
5.0
7.6
10.1
The decrease in performance at high gas-turbine inlet temperatures
is the result of increased bleed air required for turbine cooling
and the increase in gas-turbine waste heat which reduces steam
cycle extraction for regenerative heating. The heat rate at 2000°F
would be reduced ^ 270 Btu/kWh if turbine blade cooling was not
required.
Assumes any increase above the design value of 1600°F has to result
from burning fuel either above the bed or separately.
96
-------
2.5.2 Steam Temperature
Plant performance can be increased by increasing steam tem-
perature and pressure. The effect of higher steam temperatures on the
performance of the plant is shown in the following:
Steam
Temperature
°F
Steam
Pressure
psi
Gas-Turbine
Inlet Temperature
°F
Power
MW
Heat Rate
Btu/kWh
1000/1000
1100/1100
1200/1200
2400
3300
4500
1600
1600
1600
644.1
673.9
690.8
9026
8627
8417
An increase of 100°F in both superheat and reheat temperatures will give
a reduction of about 400 Btu/kWh in plant heat rate. The increased
performance and inherently less severe boiler tube corrosion in fluidized
bed boilers make high steam temperatures attractive.
2.6 ASSESSMENT
A summary of the sensitivity analysis is presented in Table 16.
Each parameter is indicated with a projection of what change might be
required as the result of experimental data. For example, the bed
temperature was set at 1750°F for 100% load. This temperature may
prove to be too high for economical sulfur removal and have to be reduced
to 1600 or 1650°F, which may be more favorable. The summary table
indicates what the effect of such a change would be on plant cost and
performance, assuming no other variable restrictions. In this case,
the plant cost would increase < $3/kw, plant performance would decrease
< 0.5%, and plant turndown to 25% load could still be achieved. This
would result in an energy cost reduction of < 0.3 mills/kWh. Since
the projected advantage over a conventional plant with stack gas
scrubbing is greater than 1.5 mills/kWh, the penalty for lowering the
temperature is not significant. This conclusion holds for all of the
variables considered and is generally valid for plant capacities of
200 to 700 MW. The results of the sensitivity analysis indicate that
the base plant design is relatively insensitive to changes in
operating conditions or design parameters.
97
-------
TABLE 16
SENSITIVITY ANALYSIS SUMMARY
Parameter Chan
Boiler Operating Conditions
Bed Temperature Reduction
1750°F to
Fluidiztng Velocity decrease
Increase
Excess Air Increase
Effect
Plant Coat
1 Power Ger
Auxiliaries Equic
eratlon
ment Performance
from < $3/kw Increase negligible negligible • • 0.5Z decrease
1600°F
to 5 fps < $0.50/kw increase (2)
to 15 fps T. $l/kw increase'3' (2)
In efficiency
• 4:1 turndown
requirement can
be met
.
-
to 100X < $l/kw decrease'*' -v $7.50/kw Increase < $4/kw increase'5' • "- 0.5X Increase
Partlculate Carry-over
Loading
Particle Size
Boiler Design
Heat Transfer Surface
Configuration
Heat Transfer
Coefficient
(a)
Bed Depth
Power Generation Kquipment
Gas Turbine Inlet
Temperature
Steam Temperature
11
Increase to 15 atm <
Increase loading
to 3 times the
design value
(T- 20 gr/SCF)
Increase fines:
particles • 10 um
increased from
15X to 25%
$0.50/kw increase
for constant area
Sl/kw increase for
constant velocity
Increase available 'v $l/kw Increase
heat transfer
surface per unit
volume from
5.5 ft2/ft3
to 11 ft2/ftj
decrease from
5.5 ft2/ft3 to
3 ft2/ft3
increase from
50 to 75
Btu/hr-ft2-°F
decrease from
50 to 35
Btu/hr-ft2-"F
assume tubing
cost 50Z greater
than base design
reduced to 10 ft
(7)
$l/kw Increase
$l/kw decrease
S3/kw increase
(7)
$2/kw Increase
(8)
+ 200°F from
1600°F
100°F Increase
In superheat
and reheat
> S3/kw Increase
$2/kw increase
$4/kw decrease
negligible
$6-8/kw increase
(6)
no Increase projected
(6)
plant efficiency
Improved turndown
capability
^ 0.2Z decrease
in plant efficiency
negligible
negligible
(10)
_UO)
negligible
+ IX in efficiency
•v 2X Increase in
efficiency
(1) Increase will depend on bed depth restrictions; S3/kw would correspond to a maximum allowable ht of -v-15 ft.
(2) The fluidiztng velocity will affect the partlculate removal equipment — see participate emission parameter for costa.
(3) Considerable savings may be realized for large capacity (> 600 mw) plants since higher velocities avoid the need for field erection.
600 mw) plants In order to avoid field erection. The $l/kw does not
Include a cost reduction which may result from elimination of the carbon burn up bed due to Increased efficiency.
(5) $4/kw assumes the larger capacity gas turbine has the same unit cost aa the base machine. Actual cost $l/kw of a larger machine
would be lower.
(6) Based on projected gas turbine requirement of < 0.01 gr/SCF of particles > 2 um.
(7) Assumes maximum allowable bed depth of 20 ft.
(8) Includes effect on fabrication.
(9) Assumes constant freeboard. (Change In freeboard requirement would have similar affect.)
(10) If the temperature is the result of equipment modification, the capital cost would be altered. Gas piping will be affected in any case.
(11) Does not include cooling tower cost which will b« reduced 9 higher efficiency. Cost effects are baaed on base design capacity and
efficiency.
98
-------
Operating conditions and design parameter changes can occur,
most likely as the result of additive problems, which would result in
a significant cost increase for the system. For example, if the bed
temperature had to be decreased to 1600°F, the heat transfer coefficient
2
was only 35 Btu/hr-ft -°F, and the dust loading from the boiler was
three times the design value, the plant cost could increase by 7 to 8%
and the efficiency decrease by 0.5%. This would result in an increase
in the energy cost of ^ 0.7 mills/kWh. This is a significant increase
but still within the economic margin. Caution must be exercised in
interpreting multiple changes in the variables. In the case above, the
cost effects were added. However, both lowering the bed temperature
and lowering the heat transfer coefficient increases the heat transfer
surface and, thus, the bed depth, if the bed area is maintained constant.
Any restrictions on bed depth would also have to be considered in the
evaluation. Parametric curves have been prepared to enable this type
of evaluation to be made.
The effect of boiler plant pressure drop and the steam-turbine
2
condenser pressure on plant performance has been presented. The
boiler plant pressure drop has a small effect on plant capacity and
heat rate: 1% increase in pressure drop results in a 0.1% increase in
plant heat rate. An increase in the steam-turbine condenser pressure
from 1-1/2 in. Hg. to 3 in. Hg. for a cooling tower results in a 2-1/2%
increase in heat rate.
The potential performance of the plant was evaluated by
assessing the effect of higher gas-turbine inlet temperatures and
higher steam temperatures and pressures. The results indicate that
plant efficiencies of ^ 45% can be achieved with gas-turbine inlet
temperatures greater than 2000°F with high-temperature blade material
to minimize cooling requirements and with steam temperatures of 1200°F.
99
-------
Advances in boiler plant subsystem concepts have not been
considered in this analysis. Cost reductions may be achieved by using
alternative concepts. The following components have been studied for
potential savings:
• Particulate removal. The projected system utilizes
four secondary collectors for final particle removal
before the gas turbine. Alternative systems are
being considered which may reduce the number and
size of the units. Cost estimates indicate that
reductions of $1 to $5/kw for the total particulate
removal system may be possible.
• High-temperature gas piping. The high-temperature
gas piping cost would be reduced if the particulate
removal system were simplified by using fewer
units per module. Additional savings might be
realized if refractory lined pipe could be used
between the secondary collectors and the gas
turbine. The present design uses a high-alloy
steel to assure protection of the gas turbine
from additional particulates.
• Coal feeding system. The coal feeding system
design is based on systems which have been built
and operated. The design provides a separate
coal feeding system for each fluidized bed in
order to assure control of the coal feed rate
to each bed. It may be possible, however, to
reduce the number of coal feed systems from 16
to 4 if independent control of solids flow to
each bed in a module can be achieved from a
single pressurized injector. The potential cost
reduction is estimated to be > $2/kw.
100
-------
• Stack gas cooler design. Cost estimates were
obtained for the stack gas coolers, but no
attempt was made to optimize the design or
consider nonconventional designs, such as those
using fluidized beds. Preliminary conceptual
evaluation indicates that the cost might be
reduced from the present $4.40/kw to ^ $3.40/kw.
2.7 CONCLUSIONS
• Base plant design is near optimal.
• The pressurized fluidized bed boiler power plant
maintains more than a 5 to 10% energy cost advantage
over a conventional plant with stack gas scrubbing.
- Effect of a potential change in bed temperature,
fluidizing velocity, heat transfer surface
configuration, gas side heat transfer coefficient,
boiler tube materials, bed depth limitations or
pressure will result in
• No significant change in plant operability
or performance
• < 2% increase in plant cost
• < 0.2 mills/kWh increase in energy cost.
- Effect of Increasing the dust loading to
three times the design value will increase
the plant cost ^ 4%, energy cost < 0.3 mills/kWh.
- Effect of increasing excess air will result in
• Increased turndown capability and performance
' 3 to 6% increase in plant cost
• No significant change in energy cost.
• Plant efficiencies of ^ 45% may be achieved for gas-
turbine inlet temperatures above 2000°F and steam
temperatures of 1200°F.
101
-------
3. LABORATORY SUPPORT STUDIES
3.1 INTRODUCTION
The evaluation of the fluidized bed combustion process for
power generation carried out by Westinghouse concluded that pressurized
combustion is the most promising technology from the point of view of
• Effective pollution abatement
• Efficient fuel utilization
• Economical power generation.
Among the key areas where further laboratory research and
development were recommended were sulfur removal, stone regeneration, and
sulfur recovery. In projecting possible sulfur removal, it was pointed
out that, owing to a lack of experimental data, the effect of combustion
pressure and of excess air on S0? reduction could not be predicted using
the currently available models of the sulfur removal process.
Thus, the economical and ecological considerations which lead
to the investigation of fluidized bed combustion as a clean source of
electricity require the support of thermodynamic and kinetic data to
translate the concept into a working system. Our extensive knowledge
of the thermodynamics of the system contrasts sharply with the available
kinetic data with which the system behavior can be predicted and inter-
preted.
The use of calcium carbonate (as limestone or dolomite) to
prevent sulfur gas emissions in fluidized bed fuel processing depends
upon a cycle of reactions, as shown in Figure 61. The conditions for the
process options afforded by this reaction cycle are given in Table 17.
Despite atmospheric pressure data obtained by Squires and his school
on the reactions of hydrogen sulfide with dolomite,and various studies
on sulfur dioxide sorption by limestones and dolomites, there is a
103
-------
Dwg.
H0S
CaS ^\
-~
H20,
co2
j
i
^
-"-•v. \
V \
H2, CO v
\
1
Ca C03
0,
so2, o2
so2, o2
Sulfur Removal
Stone Regeneration
Figure 61: The Calcium Carbonate/Sulfur Cycle Basic Reactions
104
-------
r>
rH
E4
,J
Cfl
SJ
£_{
CO
Q
1— 1
H
CJ
W
CJ
r- 1
CO
w
CJ
CJ
ryf
J— J
PM
j— j
CO
w
H
H?»
Q
p>
4-1 0)
CO rH
3 0)
r»
B i *
o 03
CJ "*-s
PM
o
0
O
vO
rH
V
H
/••x
00
o
x—\
O CM r^
CM O
•B CJ •>
VD
,
o
^3" ^3"
•*-s Tj
0)
o
-* rH
O H
CO O
CO 4H
CJ ^
•
^
C
o /-^
•H 0)
1 I t t
cfl 3
a co
•H CO
4H 01
•H M
CO ft
cfl
6CT3
^^ 01
G 4J
O cO
•H >
4-1 01
CO rH
3 01
J3
g 4-1
O tfl
o -^
PM
o B
O 4-1
O cfl
rH in
V A
H PM
CO
CM
03
-i.
CO
O
O
cfl
CJ
t
CM
O
CJ
-(-
0
CM
03
-L.
CO
cO
CJ
*
m
o G
rH O
•H
G O 4J
O 4-1 Cfl
•H O
4J 01 -H
Cfl 3 4M
a -T3 -H
•H }-i
4-1 T3 3
•H 0) ft
CO T3
tfl G S-t
60 01 01
— - B 4-*
G S cfl
o o &
•H 0
4-1 0) <&
CO !-i
3
J3 4-1 ft
SOB
0 C O)
O ^"^ 4-1
0)
S-i
3
CO
CO
01
£^l
ft
o
•H
PM !-4
o 0)
o ,n
O ft
CM CO
O
-J
CO ft
O
•H 4J
4-1 tfl
•H
CO •
cd o
60 G
-~- o
G o
0
•H CM
4J O
CO CO
3
& &
S o
0 rH
O *-*
PM
o
0
0
00
rH
A
H
CM
O
CO
-^-
O
cfl
CJ
t
CM
O
CM
• — ^
CO
-i-
co
cd
CJ
•
oo
/^
4-1
ft a
0) O
CJ -H
G 4-J
O cd
0 O
B "H
O 13 4-J
•H 01 -H
4J a cfl
co G cfl
3 cfl 60
n ^
g -0 rH
O Cfl rH
o *-s cd
01
3
CO
CO
a)
}-i
ft
o
•H
^j
01
r1.
ft
CO
0
B
4J
tfl
CO
CM
03
_j_
CM
O
CJ
-J-
co
O
U
cfl
CJ
f O
CO
O cfl
CM CJ
03
_i_
CM
CJ O
+ Cxi
O
rH
105
-------
clear need for kinetic data with which the desulfurization of fuels in
fluidized beds can be predicted and explained.
A thermogravimetric analysis system (TG) was modified so
that it could be used to study the reactions of corrosive gases with
solids at pressures above atmospheric. A program of studies on the
sulfur/calcium carbonate cycle was then undertaken with the following
objectives:
• To establish which reactions occur at proposed conditions
• To determine the reaction kinetics
• To determine the stone utilization achievable
• To study regeneration of the stone
• To probe reaction mechanisms
• To assess the influence of side reactions
• To recommend optimal operating conditions.
3.2 APPARATUS
The TG is an electrically recording balance which measures
the changes in weight of a solid suspended from the balance arm in the
electronically controlled furnace. The gaseous atmosphere surrounding
the solid and the temperature of the furnace may be adjusted or altered
as desired, within specifiable limits. Normal operating pressures are
atmospheric or subatmospheric, and provision for the use of corrosive
gases usually requires modification of commercial systems.
A number of investigators have reported work with a custom-
built pressurized apparatus, as summarized in Table 18. No work appears
to have been carried out with corrosive gases at pressures above atmospheric,
In our work, the DuPont thermal analysis system was modified
for use on this project for three reasons:
• The successful use of corrosive gases in this apparatus
by Squires, to study some of the reactions of interest
106
-------
CJ
•H
en
H
en
en
oo
rH O
3 H
PQ 3
"^3 t"-3
H H
C*
en
S
PH
CC
CJ
•rl
4J
0)
C
•H
Oj
a
•r-
T:
4-1
en
0)
b(
C
cc
p
cd
N
•H
rH
•H
4J
cd
rH
O
rH
•H
0
•rl
CO
ft
O
O
O
rH
(_,
0)
O
3
T3
CO
C
cd
°)-N
CM
I"*-.
Oi
rH
"*~s
0)
cd
pq
o
cc
JH G
0) O
ft -rl
ft 4-1
O Cd
G ^
TJ
• K ^,
C ^
O 0)
•H T3
4J
cd 01
S-l 4-1
o cd
ft m
Cd rH
> 3
o) ca
bo
•rl
CO
ft fu
o
O CM
O VD
CO vo
4-1
C
O
PM
3
Q
CM
rH
o>
rH
"^-^
G
cd
-r! -5
C ^! td
S co t-H
0 C rH
M 01 cd
PQ PH U
CO
0)
^H
0
CTs
rH
^— '
13 CO
G 6 co co
cd cd c cu
i-H -H O >H
13 rH g -H
C rH g 3
01 -H -H CT
;i 13 en en
co
cu
r^t
0
QJ
4J
en
^
CO
8n
cd
CJ
en
•H
CO
o
0 O
O
VO rH
CM
4-1
G
O
PH
3
O
CM
O\
1
I~~j
a)
CO
3
O
r1.
bO
G
•H
4-1
co
0)
107
-------
• The successful operation (under limited conditions)
of a pressurized modification of the DuPont system
by Brown and Penski
• The availability on-site of a DuPont system and
technical expertise in its use.
3.2.1 Modification of the TG Apparatus
A DuPont 951 balance assembly and furnace (Figure 62) was
mounted inside a pressure shell of stainless steel, as described by
2
Brown et al. Their design was modified to allow higher temperatures,
a flowing gas in the reaction zone, and corrosive constituents in the
gas.
High temperatures (2190°F) are possible because a web of
copper cooling coils inside the pressure shell presses around the glass
dome and metal block which houses the balance mechanism. A heat
reflector (asbestos with an aluminum surface) further insulates the
mechanism from the hot zone. External cooling coils wrapped around the
pressure shell maintain its temperature at around 100°F even when the
furnace is operating at high temperatures.
To protect the balance mechanism, a continuous flow of nitrogen
sweeps out the chamber containing the balance mechanism. Reactant
gases, which may contain H_S or SCL, are ducted into the quartz reactor
containing the sample and after reaction unite with the nitrogen purge
and are swept out of the reactor through a concentric outer chamber,
3 4
as in the design of Ruth et al. ' It is necessary to duct the gases
out of the pressure shell, since the furnace windings are insufficiently
protected by the insulating cement to prevent corrosion.
The gases are ducted from the modified quartz reactor through
a stainless steel flexible tube sealed to the quartz with a rubber
grommet. This seal is highly effective; and the grommet is replaced after
108
-------
Figure 62: The Du Pont 950 Thermograv(metric Balance
109
RM-59170
-------
each run. The exit gas line has provision for sampling, a trap for
water, a zinc oxide trap for excess hydrogen sulfide, and a stainless
steel filter, before it is depressurized through a pressure relief
valve. The system is outlined in Figure 63.
Reaction gases are supplied from cylinders through precision
stainless steel metering valves and flowmeters (five lines), and can
by-pass the TG system while flow adjustments are under way. The quartz
reactor tube abuts the end flange of the pressure shell and a thick
rubber gasket makes the seal.
Steam was introduced where required by introducing a metered
liquid stream through a fine steel capillary into a line heated by
heating tapes. A thermocouple beside the steam generation point
monitored the evenness of flow. A platinum wire heater mounted radially
in the reactor where the reaction gas entered the end flange prevented
condensation at this point,and the purge flow of nitrogen was sufficient
to prevent the accumulation of liquid water at the cold end of the
reactor.
The entire shell and TG was evacuable, and safety relief
valves were mounted on the pressure shell and in the gas-flow exit line.
Electrical connections for the furnace and the electronic controls
entered the system via conax connectors.
A DuPont 990 thermal analyzer controlled the furnace and
recorded the weight changes.
The system is shown in Figure 64.
3.2.2 Operation
Little modification in operating procedure is required.
(Mounting the sample is tedious, but the operator's skill reduced the
loading time to 15 minutes.) When the balance has been loaded, the end
flange is bolted on and the system pressurized using the purge gas.
110
-------
to
o
E
OJ
.1
O
en
o>
13
CD
8
o
o
o
111
-------
Figure 64: The Pressurized TG Apparatus
112
RM-59171
-------
Buoyancy effects are not a problem at temperatures greater
than 1100°F; changes in the reaction gas stream inevitably lead to minor
buoyancy effects which can be determined experimentally and compensated
for in data interpretation. Temperature stability is not as good as
during atmospheric operation: the Cr/Al thermocouple is of necessity
sheathed in a quartz capillary when corrosive gases are in use, and the
flow of reactant gas places a duty on the furnace which leads to
temperature inhomogeneities in the furnace. Changes in the reactant
gas composition result in changes in the isothermal temperature recorded
by the thermocouple, lending a degree of imprecision to the kinetic data.
Because of the heavier duty imposed on the furnace, resulting
in less precise temperature control, and the quartz shield on the
thermocouple, © checked the accuracy of the temperature recording. A
sample of specpure iron was heated in nitrogen at 5 K/min. The observed
Curie point was 1424°F, in satisfactory agreement with the literature
value, 1418°F.
3.2.3 System Performance
The performance of the apparatus may be assessed in terms of
the quality of the data produced and the reliability of the instrument.
The data obtained during atmospheric pressure operation proved
to be repeatable. An example may be cited where a cycle of reactions
was performed on an iron oxide substrate. The sample was reduced,
sulfided, oxidized, and reduced for seven cycles. Each cycle required
that the gas flow be reset. The mean deviation of the rate at 5%
utilization was 3.6%, from the second to the seventh cycle. It is
more difficult to assess the repeatability of the experiments involving
small samples of impure limestone and dolomite. Buoyancy effects were
not serious at temperatures above 1100°F; the precision of weight
113
-------
measurements is best demonstrated by the agreement obtained in weight
changes. For TG124 where the reactions
CaS + H20 + C02 — > CaC03 + H2S (1)
CaO + C02 (2)
occurred, the weight changes observed corresponded to:
-4
(1) 1.45 x 10 moles carbonate formed
-4
(2) 1.43 x 10 moles carbonate decomposed.
Two forms of breakdown restricted the number of experimental
runs. Corrosion of the balance mechanism requiring repair by DuPont
occurred twice. It is thought that corrosion was accelerated when the
pressure relief valve became blocked. When gas flow ceased in the
exit line, accumulation of hydrogen sulfide in the system overcame the
protection afforded by the purge gas. The material which accumulated
in the seat of the pressure relief valve was, apparently, solid sulfur.
A fine powder of sulfur accumulated in the valve despite the presence
of a stainless steel filter in the exit line. No sulfur accumulation
was noted elsewhere in the exit line. In an effort to overcome this
problem, a zinc oxide trap heated to 572°F was inserted in the exit
line, before the pressure relief valve. This has prevented any sulfur
deposition in the valve, when gases containing hydrogen sulfide are used.
Electrical faults developed on two occasions in the first
eight months of operation of the 990 console, necessitating replacement
of two circuit boards and a relay. In the second breakdown, additional
delay was caused by a faulty replacement circuit board.
3.3 PROCESS VARIABLES
The review of experimental results in the evaluation of the
fluidized-bed combustion process listed thirteen variables which
influence the reduction of sulfur oxide emission. Since some of those
variables are in part complex functions of more fundamental physical
variables, they can be associated with factors which influence the gas-
solid kinetics.
114
-------
Fluidized Bed Kinetic Study
Combustion Pressure Pressure
Bed depth
Ca/S mole ratio V S0~ concentration
Sulfur level of coal
Bed temperature Temperature
Fluidizing velocity ] Controlling regime (external
„ ,., -,.-,/ mass transfer, internal mass
Factors affecting gas-solids/ - , .
. . I transfer, chemical reaction
mixing ~s '
Stone type and source Stone type
Stone size Particle size
Excess air 02 concentration
Coal type
Location of coal feed Oscillation from oxidizing
to reducing atmosphere
The variables on the right have effects which can be studied
in the TG apparatus. The major consideration in this study is the
effect of pressure in combination with the other variables.
Some factors are of particular concern. The operating
temperature of the pressurized fluidized bed design is 1750°F (chosen
in consideration of desirable turbine inlet temperature and ash
softening characteristics). However, the retention of sulfur in lime-
containing fluidized beds has shown a maximum at lower temperatures
in some fluidized bed combustion experiments. ' ' The variation in
the rate of reaction of SO- and CaO with temperature is , therefore, of
great interest. The rate of reaction and the degree of stone utiliza-
tion achievable are also of primary importance. An increase in the rate
due to pressurized operation may lead to better sulfur retention:
increased utilization of the actual CaO content affects the criteria
for comparing once-through systems with regenerable sorbent processes.
115
-------
3.4 PREVIOUS LABORATORY STUDIES
The results of several laboratory studies are summarized
14-20
here.
• Sulfation is first order with respect to S0«.
• Sulfation is zero order with respect to 0_.
• The rate of reaction increases with temperature.
Various values of activation energy have been derived,
but there is no unambiguous determination of activation
energy.
• Reactivity depends upon the pore volume formed during
calcination.
- Small pores give a high initial reaction rate but,
if the particles are large, a low overall capacity
results.
- Large pores give lower rates but increased capacities.
- For small particles, capacity is determined by the
pore volume available for product accumulation.
- Other properties which influence the pore structure
formed on calcination may serve as general indices
to the capacity, i.e., temperature and conditions
of calcination and sodium content of the stone.
- The distribution of sulfur through a sulfated stone
depends on the stone: a coarse limestone shows
total permeation of the stone by sulfate, while
Iceland spar forms a rim of sulfate.
3.5 SUMMARY OF THE RESULTS
3.5.1 Pressure
Sulfation at 10 atmospheres was not significantly faster than
sulfation at 1 atmosphere. However, the utilization achieved before
116
-------
the rate fell to one-fiftieth of its initial value was 84% at pressure
as compared to 42% at 1 atmosphere, when the stone was calcined above
1560°F under an atmosphere of C0?.
3.5.2 Sulfur Dioxide Concentration
The sulfation was found to be first order with respect to
sulfur dioxide concentration. The increase in partial pressure of sulfur
dioxide in passing from 1 to 10 atmospheres at fixed gas composition does
not result in a ten-fold increase in observed rate. Changes in the
concentration of sulfur dioxide at a given pressure produce a proportional
change in rate.
3.5.3 Temperature
A slight increase in rate is observable with temperature in
the range 1350 to 1700°F for dolomite 1337 (Figure 65); for Tymochtee
dolomite initial rates (up to 25% utilization) are not distinguishable
in the range 1600 to 1850°F, but utilization is ultimately lower at the
higher temperature.
3.5.4 Agreement with Previously Reported Data
The kinetics observed on the TG apparatus were sensitive to
sample size: for small samples (10 mg raw dolomite) at atmospheric
pressure, the rates of reaction were in good agreement with those found
by Borgwardt and by the NCB investigators.
3.5.5 Stone Type
Both dolomite 1337 (high purity) and Tymochtee dolomite (10%
impurities) proved capable of yielding high utilization.
117
-------
c c
o o
•"••• ••" (/I
4— > •*— • ;!:
03 TO
KJ NJ 1— CM
.1.1 .!:: o> o
._ ._ .c 00
•+-> •*-> Q.
ZD ID to
8
'£•
o<
omite Sulfation
03
< CD O
«
a
E
CD
A
CD
13
CT
I I
Cf\
CM
Cf\
OO
CM
CM
CM —I
CM
oo
1_
U!W
IV
118
-------
3.5.6 Particle Size
Stone reactivity decreased as particle size increased, but the
rate of reaction of 4000 y diameter particles was faster than shell
diffusion would permit, indicating a contribution to reaction from
internal surfaces.
3.5.7 Oxygen Concentration
The reaction was zero-order in oxygen concentration in the
range 2 to 4%. Increasing the oxygen concentration to 11% gave no
increase in rate.
3.5.8 Stone Pretreatment
The result of stone pretreatment was crucial in its effect
on CaO utilization.
When calcination of the stone was suppressed with carbon
dioxide until temperatures of ^ 1560°F were achieved, subsequent
calcination yielded an extremely active stone, as shown in Figure 66.
Attempts to vary the calcination by slow calcination at lower
temperatures led to minor variations in the utilization.
The effect of calcination under C09 for Tymochtee dolomite is
shown in Figure 67. The active and inactive stone yielded 40% CaO
utilization within twenty minutes, although the active stone was exposed
to one-seventh of the S0? concentration used in the inactive stone
experiment.
3.6 THE EXPERIMENTAL PROGRAM
3.6.1 Sulfation of Dolomite
In the first phase of the experimental program, the sulfation
of two dolomites at 10 atmospheres pressure was studied. Analysis of
the rate data led us to undertake experiments at atmospheric pressure
to establish the relation between our results and those obtained by
other workers.
119
-------
o>
'i
_o
o
Q
-o
CD
C
O
*4—.
O
cz
o
"ro
>*—
ZJ
O)
r3
cr
CM •—i
papeay uoipejj uinjD|Bo
120
-------
121
-------
Samples of dolomite weighing 30 to 100 mg were calcined and
sulfated (30/40 mesh being used). The experimental details for
experiments TG86-TG98 are shown in Tables 24 and 25.
Calcined Tymochtee dolomite reacted at a rate dependent on
calcium utilization. At 1562°F, in 4% 09 and 1.250 ppm SO , 25% utilization
^ z
was reached within eleven minutes, but exposure to reactant gas for one
hour was needed to achieve 50% utilization.
The course of reaction followed a rate-law of the form iden-
18
tified by previous investigators:
da a -a/a
dt ~ t e
o
as shown in Figure 68. Basically, this rate-law describes a transition
between an initial fast rate and a much slower rate which eventually
controls reaction.
Calcined dolomite 1337 exhibited the same behavior on sulfation.
However, the initial fast rate was one-third of that reported by Borgwardt
for the same dolomite. The rate of reaction for half-calcined stone was
the same as that for fully-calcined stone, and increasing the gas-flow
rate from two to four liters per minute had no effect on the rate.
It was concluded that the geometry of large samples (30 to
100 mg) might prevent an adequate supply of reactant gas from reaching
the sample. Therefore, atmospheric pressure tests were carried out to
permit more direct comparison with previous work reported by other
investigators.
3.6.2 Atmospheric Pressure Tests (TG102-108)
Atmospheric pressure sulfation of calcined 1337 dolomite
showed that
• Sample size influenced the rate of reaction; large
samples exhibited slower rates than small samples
(Figure 69).
122
-------
CO
CD
O
h-
CD
-i—•
'E
"o
o
o
•a
Qi
C
03
O
c
o
o>
v_
3
en
pepeay
123
-------
VD
O
I oo
C
'E
oT
E
CSJ
CD
jr
c
r^.
oa
ca
i-H
CD
•*•-»
_O
O
o
-o
CD
C
'o
TO
o
'o
C
o
C
O
CD
CO
_CD
CL
E
(D
CO
"o
"tj
CD
'iunp|eo
124
-------
• The flow rate of reaction gas had only a minor effect on
the initial rate of reaction (Table 19).
• The rate of reaction observed at atmospheric pressure
was close to that estimated from the graphs published
by Borgwardt, when the sample size was kept to 10 mg
(Table 19).
• Although the smaller sample size showed a faster rate of
reaction, termination of the rapid phase of reaction was
more abrupt. Thus after thirty-four minutes, the same
degree of sulfation was achieved in the case of both
samples, and the larger sample proceeded from there to
react at a faster rate (Figure 70).
• A small sample (10 mg) was sulfated at 10 atmospheres
pressure, and it exhibited a high initial rate of sulfation,
marginally faster than the rate at atmospheric pressure
under corresponding conditions. However, it did not
undergo an abrupt transition to a slower rate of reaction
at 43% sulfation, and reached over 90% of theoretical
uptake (Figure 71).
The principal discoveries of the preceding experiments can
be summarized as :
• At atmospheric pressure, the rate of reaction on the TG
is similar to that observed by other workers.
• Large samples hinder the determination of maximum reaction
rate.
• At pressure (10 atmospheres), the rate of reaction does
not increase greatly. Utilization of the calcium may be
far greater at pressure (> 90% conversion of CaO to CaSO,,
in dolomite), than at atmospheric pressure (^ 50% conversion
of CaO to CaSO,, in dolomite).
125
-------
Curve 648708-A
-1
o 10 mg
D 100 mg
Atmospheric Pressure: 600 me/min *, 5000ppm
TG 108, 102 S02
4%02
10
4
c
o
*
CO
O
20
30
40 50
Time, minutes
60
70
80
90
Figure 70: Sulfation of Calcined Dolomite 1337
Curve 649800-A
12
16 20
Time/Minutes
24
28
32
36
40
Figure 71: Sulfation of Calcined Dolomite 1337
126
-------
w
„
CO
co
H
W PK
H o
H CM
S vo
hJ H
O
fl •
P o
PL] (— |
S3 H
H gsg
co in
Pn H
O o 10 oo co o I
O 0 0 O 0 HI
^H i~H t 1 r~H i~H r^ |
127
-------
• Since the calcination history of large samples is
different from that of small samples, the mode of
calcination may govern the utilization achievable.
3.6.3 Dolomite 1337 at 10 Atmospheres (TG137-150)
In these experiments, the following parameters were varied:
SO- concentration; temperature; calcination treatment.
Temperature had almost no effect on the rate of reaction in
the range 1350 to 1150°F as shown in Figure 65.
The change from fast reaction to slow reaction occurred at
30% utilization at 2500 ppm SO and at 40% utilization at 5000 ppm S0_,
as shown in Figure 71. The utilization in these experiments was similar
to that observed earlier in the atmospheric pressure experiments.
The S0_ concentration in the slow phase was varied from 4000
to 1000 ppm, and held at a given concentration until the rate was
established. The rate was first-order in S0_ concentration (Figure 72).
The calcination treatment prior to sulfation was varied from
rapid calcination at 1472°F in nitrogen to slow calcination by heating
from 1112°F at 2 K/min.
Slow calcination, by heating at 2 K/min, combined with reaction
at 1750°F decreased the utilization observed in the fast reaction period
30%. Lowering the temperature of sulfation to 1680°F failed to restore
utilization after slow calcination (a lower S0_ concentration was used).
Finally, a calcination treatment in which the magnesium
fraction was permitted to calcine but the calcium carbonate was held
under 1 atmosphere of CO^ was tried. Calcination of the carbonate
occurred rapidly at 1690°F. Sulfation of the stone so produced gave
high utilization in the fast rate phase and was comparable to the result
observed in TG110 (Figure 73).
Thus, minor differences in the utilization achieved during the
fast period of reaction result from S02 concentration and temperature
128
-------
Curve 6501^2-A
.02
CO
CT>
.01
CD
OC
0.9
IS 0-8
| 0.7
^ 0.6
-
.
o
<3o.3
0.2
0.1
-35 440 Calcined Dolomite 1337
4%02in N2
1600°F
10 Atmospheres
9 mg Sample Dolomite
TG149
123456
SCL Concentration/1000 ppm
Figure 72: Effect of SC>2 Concentration on the "Slow" Sulfation Regime
Curve 650143-A
TG
A 110
o 150
-35+40 Dolomite 1337
0.5% SCO
> 'n N2
4%o2 ;
1562°F
10 Atmospheres
O A
1 1.0 10 100
Time/Minutes
Figure 73.- High CaO Utilization in the Sulfation of Calcined Dolomite
129
-------
effects. The effect of calcination can produce differences of the order
of 10% utilization for variations in the calcination heating rate; or
differences of 40% when calcination is carried out at 1652°F under CO,,.
3.6.4 Sulfation of Tymochtee Dolomite at 10 Atmospheres
The remaining experiments were carried out using Tymochtee
dolomite. The calcination treatment which had been successful in
leading to high utilization was used. A larger particle size (16/18
mesh; 1000 y diameter) was used, and in three experiments the temperature
of sulfation was varied from 1600°F to 1850°F. Three features of the
course of reaction was notable (Figure 74).
• High utilization was achieved (> 60% in 20 minutes).
• The rates of reaction at the three temperatures were
not distinguishable in the early part of the reaction.
• After 20% utilization, the rate of reaction decreased
with temperature, leading to a 10% difference in
utilization between the 1600°F and the 1850°F sample.
This difference may be noted as an increase in the
time required to reach a given level of utilization
and is shown in Table 20.
3.6.5 Effect of Low SO Levels
In further experiments, the effect of high oxygen concentrations
and low SO levels on the reaction were investigated. Tymochtee dolomite
was used as substrate, and calcination was carried out at 1562°F to
attempt to get high utilization.
Three experiments at 660 ppm SO. and 1750°F were carried out
ifferent particle sizes, and one experiment ai
was performed on one particle size (Figures 75 and 76).
on three different particle sizes, and one experiment at 90 ppm S09
130
-------
Curve 650658-A
.70
.60
.50
.40
.30
S.20
.10
1850F
o 1700F
o 1600F
Tymochtee Dolomite
16-18 Mesh
Calcinedin9%C02
S02=5,000ppm
02=4%
P = 10 Atmospheres
24 6 8 10 12 14 16 18 20
Time/Minutes
Figure 74: Tymochtee Dolomite Sulfation under Pressure: The Effect of Temperature
.70
.60
,.50
c.40
o
Curve 650659-A
(O
O
.30
.20
.10
0
P =10 Atmospheres
T=1750F
16-18 Mesh Stone
11% 02
0 700 ppm S02
* 90 ppm S02
100 120 140 160 180 200
Time/Minutes
0 20 40 60
Figure 75: Sulfation of Calcined Tymochtee Dolomite at Low SC>2 Concentrations
131
-------
TABLE 20
THE EFFECT OF TEMPERATURE ON THE TIME REQUIRED TO ACHIEVE
A GIVEN LEVEL OF CALCIUM UTILIZATION
Utilization
% Ca
Time at
1600°F
Time at
1850°F
% Increase in
Reaction Time
25
35
45
55
65
1.99
3.4
5.2
7.95
11.55
2.5
4.4
7.0
11.3
19.45
25
29
35
42
68
132
-------
_ o
CD
•*-«
'i
_o
o
Q
0)
u
o
CD
C
O
*o
c
o
"TO
M—
"5
en
papeay
133
-------
The experiment at the lowest SO concentration showed that
the course of reaction was in reasonable accord with expectation, on the
basis of the reaction behavior at 5000 ppm. Table 21 shows the rate of
reaction relative to unit rate at 5000 ppm, at five values of calcium
utilization. The 660 ppm results were anomalous, in that the initial
rate was in agreement with first order predictions, but as reaction
proceeded, the rate fell off more sharply, as if high utilization was
not to be achieved. Since the calcination conditions may not have
yielded an active structure, the utilization achieved (Figure 75) is
pessimistic for the 1000 yd stone.
Comparison between the 500 y stones sulfated at two different
SO concentrations clearly illustrate that high utilization can be
achieved with the Tymochtee stone.
The results obtained in these experiments are relevant to the
design of the fluidized bed adiabatic combustor combined cycle power
plant. In this concept, excess air in the fluid bed ('v 300%) effectively
provides the heat transfer surface. Desulfurization in the fluidized
bed is affected both by higher oxygen concentrations (13%) and by lower
sulfur dioxide concentrations (^ 1000 ppm or less).
The results obtained on the TG show that no increase in rate
can be expected as a result of the increased partial pressure of oxygen.
This means that the step in the sulfation sequence which involves oxygen
is not rate—controlling under these conditions. The dilution of sulfur
dioxide reduces the rate of reaction, but clearly a sufficiently long
residence time of stone in the bed will lead to efficient calcium
utilization, and satisfactory sulfur removal will depend on the bed
depth chosen. In a later section, the TG data reported here is used
to program Koppel's model for fluidized bed conditions, and an estimate
of the bed height required is obtained.
134
-------
It is interesting to note that even at the 90 ppm S02 level,
reaction proceeds at a nonnegligible rate, and 30% calcium utilization
is noted after three hours' residence of a 1000 u diameter fraction in
such a stream. This is roughly equivalent to a 10% reduction in SO^
concentration per centimeter of bed height at the 90 ppm level, at
4 ft/sec through the bed.
TABLE 21
THE RATE OF SULFATION OF TYMOCHTEE DOLOMITE
AT LOW SO CONCENTRATIONS, 1750°F
% Ca Utilization
5000 ppm
660 ppm
90 ppm
5
10
20
30
35
S02 ratio
1
1
1
1
1
1
.132
.168
.072
.048
.030
.132
.025
.026
.017
.016
.024
.018
135
-------
3.7 THEORY
The effects of pressure on gas/solid reaction kinetics derive
from two fundamental and competing phenomena. Increasing the reactant
gas pressure increases the number of reactant molecules striking the
solid surface in unit time, but it decreases the diffusion of products
away from the reaction site. In addition, where reaction is part of a
chain that includes adsorption or desorption as a rate-influencing
step, saturation of the available sites can place a limit on the
influence of pressure on the reaction rate.
Studies on the reduction of iron oxide have demonstrated the
usefulness and limitations of pressurization. Thus, doubling the
hydrogen pressure was found to lessen the time required to reduce a
given sample of FeO, but it did not shorten the time for reduction to
12 13
50% of its former value. McKewan found that an increase in
pressure from 1 to 10 atmospheres increased the rate of reduction
of magnetite with hydrogen by a factor of 1.88.
These results may be interpreted as arising from convective
mass transfer control at the gas/solid interface or as an adsorption-
controlled phenomenon, limited by the number of available surface sites.
Distinction between the two mechanisms may reflect the interests of the
chemist or engineer studying the reaction rather than an intrinsically
different behavior at the molecular level.
136
-------
3.8 KINETICS OF DOLOMITE SULFATION
The kinetics of dolomite sulfation may be described by the
rates of individual steps as follows:
1) Arrival of S02 at the particle surface
2) Transport of SO into the dolomite
3) Transport of S0~ across a product layer of CaSO,
on the reactant crystallites of CaO
4) Chemical reaction between S0?, CaO and 0~
(itself a multistep process).
The rate of the fourth step represents the maximum achievable
rate; to attain it, the rates of the other three processes must be
maximized. Step 1 will depend on the gas/solid mixing or on the gas
velocity over the particle. Steps 2 and 3 depend on the physical
composition of the stone. Thus, if large pores are available to
transport the gas into the stone, then step 2 will not limit the rate
of reaction. Similarly, large pores will facilitate the expansion
required to accommodate the sulfate ion. Step 3 will begin to control
the rate at some stage as reaction proceeds, but maximizing the surface
area of CaO exposed to the gas (small CaO particles) will ensure a
large extent of reaction before solid product layers inhibit the rate.
The aim of this investigation is, in effect, to find the conditions
whereby the rates of steps 2 and 3 are maximized. Since the structural
features of the stone which are important in steps 2 and 3 are formed
during calcination, it is clearly important to know what happens during
calcination.
Dolomite calcines to either the half-burned state MgO-CaCO_,
or the fully-burned state MgO-CaO.
In the half-burned state, the external morphology remains the
o
same, but the surface is covered by minute 400 A cubic crystallites
of MgO.
137
-------
A
I
The start of calcite decomposition is independent of the
presence of magnesia (but only at the lowest pressures does the
decomposition of the calcite approach completion) since MgO covering
the calcite may sinter and grow in C0? at the higher temperatures. In
O
the half-burned state, the MgO crystallites are small - ^ 150 A, while
o
the calcite crystals are about 400 A. As the temperature is increased,
O
their size passes through a minimum near 1478°F and back to 400 A at
1650°F. However, in the fully burned state, misoriented microcrystals
of both oxides form directly, their dimensions rise continuously from
100 A at 1112°F to 1400 A above 1830°F, and a reduction in pore volume
o
also occurs over the same range, the smaller capillaries ^ 80 A
21
being eliminated first.
Sulfation will, therefore, exhibit kinetic behavior character-
istic of the stone sorbent structure laid down during calcination. The
effect of carrying out the reaction at 10 atmospheres pressure will
reflect the effect of pressure on steps 1 to 4. Thus the effect of
pressure on the transfer of S0_ from the gas phase to the solid will
increase it by a factor of two, given the experimental conditions used
here. The transport of S0_ into the dolomite will be largely unaffected
by pressure, since gas concentration is proportional to pressure and
the diffusion coefficient is inversely proportional to pressure. The
effect on transport of SO. across a layer of calcium sulfate depends
on the actual mechanism. The effect of pressure on the chemical
reaction step should be to increase the rate ten-fold. Thus the
combination of processes may result in acceleration of the rate of
10 atmospheres from one to ten times its atmospheric pressure value.
The experimental rate observed in these experiments is
slightly higher than those observed at atmospheric pressure; the overall
reaction is extremely similar (Figure 71), demonstrating that the
reaction is controlled by a process virtually independent of pressure.
138
-------
3.9 QUALITATIVE COMPARISON WITH FLUIDIZED BED RESULTS
The atmospheric pressure results reported here agree with
data reported by other laboratory investigators. It is, therefore,
instructive to compare this laboratory data with experiments on real
or simulated sulfur retention in fluidized bed combustors.
In our atmospheric pressure experiments on 500 y diameter
particles of dolomite 1337, the rate of reaction typically falls to
l/50th of its initial value at 42% utilization. Borgwardt's experiments
on similarly sized particles of the same dolomite showed that 55%
utilization occurred for the same decrease in rate. In the NCB experi-
ments, similarly sized particles gave 52% utilization within 1/4 hour,
and 54% utilization in 4 hours. Borgwardt's experiments on 100 y diameter
particles showed that about 84% utilization occurred before the rate
fell to l/50th of its former value.
These results may be compared with the utilization achieved
by Esso (R & E), Argonne National Laboratory, and the National Coal Board
(NCB) with dolomite 1337 in fluidized beds.
Argonne's test in a 6-in. fluidized bed combustor (BC 10), using
540 y dolomite 1337 showed 81 to 91% sulfur retention in the bed, with a
utilization of 37 to 41%.
The NCB results are summarized in Table 22. Utilization has
been calculated from the sulfur retention value and the Ca/S mole ratio
in the bed. In Task 1 experiments, the highest utilization observed
was 62.5%. This experiment was conducted with fines rate recycle about
five times the coal feed rate; the particle size distribution was such
that 50% of the dolomite lay in the range below 125 y diameter
particles. This high utilization may be associated with fine particles
having an effectively long residence time in the bed.
139
-------
1
1
CO r-
H cn
!3 cn
H I~l
Hr— "1
W
pi H
W M
X O
^^ ^^
w
PP IT}
EH
Q H
NJ
-------
In Task II experiments conducted at pressures of 3.5 and 5
atmospheres, the average utilization was 57%. This may be an upper
limit since allowance has been made for calcium in the coal. Strict
Ca/S ratios of the input stream lead to an average utilization limit
8% lower, or about 50%. Since the S/Ca ratio in the bed, primary cyclone,
and secondary cyclone increased in the order .324, .447, .625, the finer
particles may contribute significantly to the observed utilization.
The results obtained by Esso (R & E) apparently imply faster
rates of reaction at higher utilization than observed in the TG study.
Thus, predicted utilization for a residence time of 60 minutes in 2700
ppm SO in a fluidized bed is about 48% with 1100 y particles. Westinghouse
experiments at 2500 ppm SO- with 500 y particles predict 50% utilization
1 _1
in one hour but show a lower rate at this utilization, i.e., .002 min
in place of approximately .003 min shown by the Esso results. (Our
data relate to a constant 2500 ppm SO- flow over the solid.) However,
dolomite 1337 attrited rapidly in these experiments. Using 25/50 mesh
1337 stone, 50% of the material was collected in an adsorption/regen-
eration cycle in which 52.5 CaO utilization occurred.
Thus, the overall utilization achieved in fluidized beds falls
in the range 35 to 50%, and this corresponds to the end of the fast period
of reaction observed in our rate studies at atmospheric pressure, and
for the bulk of our pressurized runs (10 atmospheres). If this fast
reaction mode governs utilization in the fluidized bed, then the
possibility exists that the utilization may be doubled. From our
pressurized results on both dolomite 1337 and Tymochtee dolomite, it is
clear that a utilization of 80 to 90% can be achieved with 500 y particles.
There are data from fluidized bed investigations which support
this idea. Esso (R & E) calcined limestone 1359 under conditions where
the C0_ left the carbonate sites at a dissociation pressure above
atmospheric pressure. The resulting calcium utilization in reaction
with SO increased three-fold over that observed when calcination occurred
under low C0? pressures, (6 to 8%) as against (1.5 to 3.5%). Consoli-
dation Coal Co. obtained excellent S0_ removal (> 90%) in fluidized beds
of relatively coarse (16 x 28 mesh) Tymochtee dolomite, at 1800°F. In
141
-------
commenting on the results which are clearly superior to most of the
data which has been obtained by other workers, they suggest that rapid
but controlled high-temperature calcination is a factor which may con-
tribute to the high sulfur retention at high calcium utilization.
3.10 QUANTITATIVE COMPARISON WITH FLUIDIZED BED RESULTS
The data obtained on the TG apparatus may be used as input to
the model developed by Koppel for prediction of sulfur dioxide retention
in fluidized beds of limestone or dolomite.
Two cases were considered, one using conditions which obtained
during an ANL test with dolomite 1337 for which the experimental
result is known, and the other using design data for a projected
adiabatic combustor.
The most important parameter in Koppel's model is the
"equilibrium utilization," Y , achieved with a particular sorbent.
The model uses the expression Y = Y [l-e~kt] to describe the course
of sulfation. The rate constant k may be obtained by fitting the TG
data to the rate expression. Such curve-fitting involves an arbitrary
allocation of Y . Figure 77 shows the data for TG 158 curve-fitted in
this manner. The deficiency of this curve-fitting procedure is that
reaction is not permitted to proceed beyond Y , although in practice,
there is a finite but small rate of reaction beyond this point.
When the data of TG 108 are applied to Koppel's model, they
predict the retention which would be obtained under a given set of con-
ditions. Applying the conditions for the Argonne experiment gives a
predicted sulfur retention of 87%. The result obtained by Argonne was
81%, with a CaO utilization of 37%. When a maximum calcium utilization
of 37% was used, the model predicted 81% sulfur retention. Clearly the
choice of the equilibrium utilization fixes the sulfur retention, given
that other conditions remain constant. The quality of the results may
be judged on the agreement between the TG data and the course of
reaction as shown by the simulation.
142
-------
Curve 65113^-A
.4
•o
o>
M
Line Represents Equation Used
ATG 158 Data at 10 Atmospheres, 660 PPM S0«, 1750F
[1000-1200 Md Tymochtee]
A
A
3.3
c
o
"o
ro
.1
20
40 60 80
Time/Minutes
100
120
Figure 77: Data Used in Koppel1 s Model for Sulfation
143
-------
For the adiabatic combustor, two cases were examined.
Uniform SO generation in the fluidized bed was assumed. The normalized
bed height required for 90% sulfur retention in the bed was calculated,
leading to values for the bed height of 12.7 feet for the eight-module
2 2
case (64 ft each) and 4.62 feet for the four-module case (352 ft each).
The TG data used are shown in Figure 77. The data are pessimistic for
two reasons: high utilization was not achieved in this run; and the
model assumed no reaction occurs beyond 40% utilization.
TABLE 23
DESIGN CONDITIONS FOR THE ADIABATIC COMBUSTOR
SINGLE BED, DESIGN 1
Item Condition
Number of Modules 4
Number of Beds per Module 1
Module Diameter, ft 21
Module Height, ft 16
Bed Area, ft2 347
Fluidizing Velocity, ft/sec 6
30 mole % Ca utilization of the dolomite
90% removal of S02
4.3% S in coal
SO level without SO,, removal = 1070 ppm
Dolomite feed rate — 179,000 Ib/hr
Coal feed rate -- 140,000 Ib/hr
Bed height from Koppel's model and TG 158 = 4.6 ft
3.11 COMPARISON OF NCB DATA AND WESTINGHOUSE DATA
• NCB reports a rate constant for 548 p diameter 1337
3
dolomite of .112 m /kg sec at 4.6% calcium utilization
and 1510°F.
• Westinghouse atmospheric data at 1562°F for 450 p diameter
3
1337 dolomite at 5% calcium utilization is .45 m /kg sec.
144
-------
TABLE 24
EXPERIMENTS USING TYMOCHTEE DOLOMITE
TG
[so2]
P/atm
T°F
Calcination
Mesh
Comment
86
87
88
89
90
97
98
147
153
154
155
156
157
158
159
160
2500
2500
2500
2500
1250
5000
5000
5000
5000
5000
5000
5000
660
660
660
90
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
1472
1472
Scan
1562
1562
1562
1535
1562
1600
1600
1700
1850
1750
1750
1750
1750
N
N,
N2
N2
N2
i
CO,
CO,
5/6
16/18
16/18
16/18
5/6
16/18
35/40
16/18
high 02 (11%)
11% 02
11% 02
10% 00
Composition: 49.3% CaC03
36.6%
Argonne Monthly Progress Report, September 1972.
145
-------
91
2500
TABLE 25
EXPERIMENTS WITH DOLOMITE 1337
TG
[SOJ
P/atm
T°
F
j Calcination
Comment
10
1590
large sample, may hinder
kinetics
92
93
94
95
96
102
103
104
105
106
107
108
110
127
137
138
139
140
141
142
143
145
146
148
149
150
2500
1250
2500
5000
2550
5000
5000
5000
5000
5000
5000
5000
5000
5000
2500
2500
5000
5000
5000
5000
5000
5000
820
5000
5000
5000
10
10
10
10
10
1
1
1
1
1
1
1
10
10
10
10
10
10
10
10
10
10
10
10
10
10
1562
1562
1542
1562
1620
1562 "
1562
1562
1562 "
1562 "
1562 "
1562
1562
1562 "
1700
1661 "
1724
1600
1652
1652
1652
1750
it
1700
1353
1600
1562
high utilization noted
in sulfation
high utilization noted
146
-------
NCB reports that S0_ stagnation limited the rate to a maximum
3
possible value of .131 m /kg sec. It is, therefore, appropriate to
compare the rates at higher utilization.
For TG 107 at atmospheric pressure, the rate at 22% utilization
-1 3
is 1.1 x 10 m /kg sec.
_2
For NCB run 53 at 22.3% utilization, the rate is 9.77 x 10
3,.
m /kg sec.
3.12 REGENERATION
The objectives of sorbent regeneration systems are:
• To reduce the quantities of dolomite or limestone
required for sulfur removal, by recirculating regenerated
stone
• To reduce the quantity of waste solid rejected from the
process
• To recover sulfur, either in useable or storable form.
Assessment of the schemes proposed for reduction of calcium
sulfate to calcium oxide or calcium carbonate led to the conclusion that
two regeneration processes appeared likely to be economically and
technically successful.
In the first of these processes, high-temperature reduction
of calcium sulfate to calcium oxide yields sulfur dioxide for recovery:
H2~) H20)
+ coy" —> cao + so2 + co2 j- •
In the second two-stage process, a low-temperature reduction
of calcium sulfate to calcium sulfide is followed by conversion of
calcium sulfide to calcium carbonate,
4H21 4H20
+ 4CO V —^ CaS + 4CO.
147
-------
CaS + H2 CaC03 + H2S,
sulfur being recovered by further processing of the hydrogen sulfide
stream.
A summary of the apparent advantages and disadvantages of the two
processes is shown in Table 26.
The two-stage process was selected for investigation on the
TG apparatus. While it is difficult to choose between the two processes,
the two-stage process had less experimental data available for an
adequate assessment of its potential. Further, its applicability at
pressure made it a natural subject for investigation on the TG system.
Additionally, the second stage is also under investigation in connection
with sulfur removal and sorbent regeneration studies of fluidized bed
coal gasification.
Studies on the two-stage process have also been carried out
at Argonne. The second step has been investigated by Consolidation Coal
Company.
3.13 PREVIOUS REGENERATION STUDIES
Attempts to recover sulfur from calcium/sulfur salts were
22 23
investigated by Riesenfeld (1920), Rosenquist (1951), and more
24
recently by Squires and his school.
Squires has reported preliminary work on the pressurized con-
version of calcium sulfide to calcium carbonate, in which the feasibility
of H9S-rich off-gas streams has been demonstrated. A 24% H~S gas stream
24
was obtained at 1100°F and 220 psia.
Parallel with the work reported here, Consolidation Coal
Company have demonstrated multiple recycling of the reactions
CaCO + H S > CaS +
CaS
148
-------
CN
W
PQ
t
CO
w
CO
CO
w
u
0
Pi
PM
la
O
H
£H
<3J
pq
W
^
w
CJ
W
pi
CO
CU
bC
ct
4J
fi
c
£>
••£
n
u
C
CO
0)
M
cd
-*
M
fl
>
13
<3^
bO
fi
S iH
3 B
•H 3
CJ CO
H fi
cO 0
o a
13 "
cu a
4-1 O
fi iH
CO 4J
& 0
3 0)
0)
0 bO
3 fi
13 -H
O 4J
p. p.
B
m o c/3
6 ^d
d toO
C -H
O bO
•H 0) C
O -H O
3 M-I 3
13 i-H 13
0) 3 0)
M CO M
CO
CO
0)
O
o
M
P.
0)
bO
cd
4-1
CO
1
cu
.H
bO
fi
•H
CO
P.
CU
4J
CO
1
0)
(3
O
0)
cd
x: I fi
S 4J 13 H-( O CO
0) 0) iH i-H 3 CD CO 0)
.-H x! Is o> co td 0 M
Xl 4J -H 0) v-i iH
O co >, cd M 4-J 3
M 0) CU O 0"
p. 4-1 CO O4JO) 0)CO 0)
cd co 4J CO > 4J cd M 0)01
13 > 0) fi 0) 3 U ." 13M
•H iH M CU iH iH CU ^^ 0) -H iH
O4J O 4-icdXl si- CO i-H X3
>CJ 0) cdbOO >,OO X Ocr
cdco 13 M-icdcd xiwcj cd o cuco
0) rH cdcd rHCN^MC
Oi3 ti 3COO CUU -H fficd O
4J OCOCU4JO Cd XlCO-H
CU-H 13 -H •> B > -4JCN4J
13O 4J COCdO) 4-iOM cd CM ffi-H
cue! co CUMM cdcoo COM TJ
MO) M CJ4J3 B U <+H cdOUCUBC
•rl ,£5 4J 34-) M~~^ 4J-H XI30
3 C T30)cd OCNO cd4-iO)4->vHO
O4 ^ 4-1 CU 0) bO M ^H ^~^ 4-1 ^ cd M cd M
CUCJcOO McdCU W O3MX>0)
McdOIC 4J P, 4JO13 i-H -rl CO -HM
XIO CCS iH^-'O) cdmcoO)rH3
COM O OcdQ) OICdM 4-J-HO)4-J-H4J
M4-JCOCN4JT3 3 OJOlp-CCTM
3CJ3O ocdx! <4H w o4 g> ocuo)
4-1 -H CO CdCObO OBO) -rl XIX1 Pi
cd W rO CJ *rH *rH QJ ^-) ^j r~H bO V^ *T3 3
V-4 C S ^ TJ i-^1 ^> ^^ ^J ^3 "rH td C< flJ
QJ K^ CO J3 -U* CO CO rX 4-1 r^ (J ccj J-*
^, cd «p-| (jQ cfl ^-j x™ N *pH ^i QJ s^ p.
6B"^<1) C-HiHcOi-HCO bO 0)0) CO CO WbO
0) 4->X>M'HXl4-lO'H cd BOJOIOC
4-l"G'H3 4-> 4J {5 >rl X! <1) 4-1 Q)'H M CJ'H-H
OlOli-HCO 0)v-/Q)4-iT-lx! CO i-HM'H 34JCO
X!>X>lHCOP. , »rH 4—1 i~H QJ ^-i ^t _ v} Q
0) 0)
3 3
4-> CO bO 13
cd co C co iH
M 0) iH CN rH
CU !M M PS O
P. Pi CO 0) CO
B +J co
0) 14-1 " d M 13
4J O CN -H O iH
CO CO > O
^ cu u cu cd ^
O bO M 13 MH cd
0) H cd •> 3 -H
,H 4J CO 4J i-H >-, CO
Xi >, GCNCdOi-HO)
Cd Xl Cd pr{ £_) [0 i — 1 M
i-H > 0) cd 3
•H 13 T3"P.T3O4-i
cd o) cd co g -H -H td
> M COO01OBS-I
cd !^,O -HU4J>cdfl)
r- 1 > 13 Cd Pi PJ
cd 4J cfl " S >> B
4J CJ14-J O C^OC013CU
cd 0) •HOiHOIO4J
13 M CO B CO M B
•HO cd 4-> 3 B S
i-H 13 -H C4-icd4JCU O
cd 6 >> o cd xi I-H
4—1 rO Cd ""O ^Cd M 4—) fi
C 0) fi O fiCUQ) O13bO
0) B ^ B Ocop,o)-HcuC
B M 13 M •HCOBM4-)3-rl
•H O O 0) 4-ia)0)3OCM
M MH B XI cdM4-ic03-HO)
0) M 4-1 B P< COT34J4-1
p, Q o) M Pi J3 0) O fi fi
X n) Xl O O3OMMO-H
0) CJ 4J fi <4-ICO>H&P,acO
p.
0)
4J 13
CO 4J fi
1 CO P. OP.
O MO) CJ 0)
IS -rl 4-1 0) 4-1
H CK CO CO CO
149
-------
25
in a fluidized bed reactor system at 15 atmospheres. Their results
are noted in a later section.
o/r
Experimental work at Argonne has yielded results on both
the calcium sulfate reduction reaction and on continuous cycling from
sulfate to sulfide to carbonate. These results are also noted below.
The experiments on the TG apparatus carried out at 10
atmospheres pressure fell into two parts.
• Experiments in which dolomite was sulfated and the
calcium sulfate was reduced to calcium sulfide. In
some of these experiments, regeneration of the calcium
sulfide to calcium carbonate was attempted.
• Experiments in which calcium sulfide, prepared by direct
reaction of hydrogen sulfide with dolomite, was regenerated
to calcium carbonate, and experiments using reagent-grade
powdered calcium sulfide.
In all the experiments, 35/40 mesh stone was used, with the
exception of the cyclic regeneration experiment where 16/18 mesh stone
was tested, as noted in Figure 81.
3.14 CONCLUSIONS FROM THE REGENERATION EXPERIMENTS
Experiments on the two-step regeneration led to the following
conclusions:
• Calcium carbonate can be regenerated from sulfided
limestone and sulfided dolomite.
• Sulfated dolomite can be reduced to dolomitic calcium
sulfide at 1600°F.
• Sulfided dolomite produced by reduction of sulfated
dolomite is relatively inert to regeneration.
150
-------
• Inert calcium sulfide gradually builds up in dolomite
when it is directly sulfided with hydrogen sulfide and
regenerated in cyclic fashion.
• Future studies should concentrate on elucidating the
nature of this inactive calcium sulfide.
3.15 TWO-STEP REGENERATION EXPERIMENTS
The first experiments on sulfation regeneration are listed
in Tables 27 and 28. Dolomite, in either the calcined or half-calcined
state was sulfated on the TG apparatus. The sulfate was reduced to
calcium sulfide, and, in some cases, regeneration of the calcium sulfide
to calcium carbonate was attempted.
In the first experiment with fully calcined dolomite 1337,
sulfated to 33% calcium utilization, reduction in 10% CO at 1382°F gave
an apparent 98% yield of sulfide from calcium sulfate. Chemical analysis
yielded 87% of the expected calcium sulfide. The reduction time was
140 minutes. In the next experiment, reduction to calcium sulfide was
carried out in a CO/CO = 1/1 mixture. The rate at 1382°F was slow
(16% S reacted per minute), but it increased to (44% S reacted per
minute) on heating to 1472°F. To reduce the sulfate present would have
taken about four hours in this mixture, at this temperature.
In the next experiment, TG94, after reduction to sulfide under
C0« {thus maintaining the carbonate phase), the reaction with H^O and C02
was tried at 1200°F (650°C) (Figure 78). A rapid initial weight uptake
corresponding to 20% regeneration occurred, followed by an extremely
slow reaction (^ 5% carbonate formed per hour). Increasing the carbon
dioxide concentration had no effect. An attempt was made to test the
reactivity of the regenerated stone to sulfur dioxide and oxygen. However,
exposure of the stone to the gases resulted in parallel reactions, in
which the residual sulfide content also oxidized. This was demonstrated
by stopping the supply of sulfur dioxide; oxidation proceeded unchecked.
151
-------
8
CM
0)
>
*£
8
»/>
CD
l_
CD
^
O
O.
t/J
o
y- E
II SD^
|Sc,^ff^
B 31 O II " o
Q _- _ H- CL t=:
o
o
(f\
a
CD
'i
_o
o
O
c
o
to
v_
CD
cz
CD
cn
CD
CD
ZJ
cn
LA
m
152
-------
p^
CM
W
i_3
pq
H
CO
pa
W
K
PM
CO
o
H
<^
O
rH
H
z
o
M
H
W
^
W
W
Prf
PL|
W
H
CO
o
H
^
I_J
O
23
QJ
•H rH
UH 0)
rH -H
3 >>
CO
a w
0 fi
•r( O
4-1 -H
CJ 4->
3 -H
a) a
od o
O
a
0
, j
•rl
4->
cO
M-l
rH
~J
CO
0
4J
a
CD
,0
O
CO
4-1
a
a) M
8 > a)
cO £>-,
a
co B^S a)
t-H 00 -H
CO U-l
a T3 rH
•H 0) 3
£ S CD
(U O
o w o
B-S B-S
CO Is*
Oi CTi
b
rH
PM ||o
o H
CM i — ir»
oo CMJCT
CO O rH
rH U
L_I n
o
o II m
U r-»
r— i
B^S O
0 O
rH 1—1
sl- Is-
CO CM
T3 rO
CD at
•rl CO -H CO
o co a co
rH rH rH rH
Cfl CO
a a
CN CO
CTi CT»
B~S B~8 B-S
o oo oo
O CT> CTi
rH
PM PM PM
o o o
CM CO OO
v£> CO C1
in in m
rH rH rH
rH rH rH
1 — ~ ' — . -^.
CM CM CM
II B^ II 6-8 II 6^
in o o
CMN CMTM CMCM
O O O
o ii on UN
o o o o o o
U U CJ O U U
ȣ> O OO
CO CO CO
i — 1 F^ rH r--
cO co cO co cO co
a co o co o co
1 rH | rH I rH
II j l_[',_| l^t
rH rH rH
CO cO cfl
sf m vo
OX CT» C7\
B-S &-S
m in
• •
r^ r-*
ON ON
PM PM
0 0
CM CM
^O **O
in m
rH rH
rH rH
CM CM
II B^ II
O O
CMCM CMCM
0 O
U || U ||
o o o o
C_> U CO O
in
o
I-- VQ
pN.
*^ ^3
0)
-------
oo
CM
w
rJ
m
$
CO
S
HH
W
EG
C\
H-<
CO
_
C
coo
•H C -H
4-1
•a cu co
cu 4J TJ
> cfl -H
S-i MH !*!
CU rH O
co 3
XI CO CO
O cfl
CU U
C >
O -H >~,
•H 4-1 J3
4J a
O CO TJ
3 CU
13 O O CM
CU CO 3 O
M TJ CJ
» O
O O )-< C
C CJ ft -H
O
00
J
1^
6^ S-S
oo m
CM in
Pn Pu
0 0
O 0
O vD
CM in
rH rH
&"P
0^5.
0
rH
II
CM fr-S
O 0
U r-l
II II
O CM
CM O
EG 0
^j- m
CT\ C^
•rH
M-l
rH
3 CM
CO O
rH 6-S
cfl si-
3
T) C
•H «rl
CO
cu cu
M rH
,O
>4-l Cfl
O N
•H
CU "3
S -rl
o x
CO O
s~?
CO
rH
Pn
0
o
o
rH
^^J
T^
6"*9
o
rH
II
CM
0
U
II
o
CM
EC
vO
CTi
II
/-^
4-1
£ CO
CU
>, rH
£> O
^ a
1 CMJ-
W I
O
rH rH
CO
3 X
TJ
•H CM
CO CM
CU •
r-l rH
B-S
CO
VO
Pn
o
0
vO
m
ftsP
trx
o
rH
II
CM
O
C_>
II
O
CM
EG
i^
-------
Further experiments compared the reduction of sulfated dolomite
at different sulfate levels. The kinetics of reduction for samples
sulfated to 26% and 38% calcium utilization were similar, (Figure 79).
Whilie weight change measurements invariably indicated total reduction
of sulfate to sulfide, some contribution from the parallel reduction to
calcium oxide and sulfur dioxide may have been present. Because of the
high-volume gas flow maintained over the sample and the dilution effect
from the purge nitrogen, it was not feasible to determine the sulfur
lost in the gas stream analytically. However, after reduction of the
calcium sulfate at 1562°F,and regeneration of the sulfide to carbonate,
residual sulfide was determined iodometrically on a portion of the solid
-4
product. Sulfide expected from weight change was 1.22 x 10 moles.
-4
Sulfide by iodometric titration was 1.3 x 10 moles. It was concluded
that the predominant product in sulfate reduction was calcium sulfide
and that stone with a high level of calcium utilization in sulfation
could be completely reduced.
Attempts to regenerate carbonate from the calcium sulfide were
largely unsuccessful. Stone with a low original calcium utilization in
sulfation gave 28% regeneration at 1200°F. Stone with high utilization
gave 16% regeneration at 1200°F. The mole-fraction of regenerated
calcium carbonate was, however, similar in each case. This suggests
a physical mechanism in which a certain amount of carbonate sufficient
to seal off transport pores can form, after which reaction is limited
by slow diffusion processes. The possibility that the calcium sulfide
is reoxidized to calcium sulfate by carbon dioxide, reversing the
reduction step, was considered. However, if such sulfate is produced,
it is not itself reduced by a carbon monoxide stream. In addition,
regeneration at a temperature of 1560°F proceeded to 55 to 63% of the
calcium sulfide in TG95 and TG97 demonstrating the possibility of
regeneration. However, at such high temperatures, the concentration
of sulfur in the product stream would be far too low for recovery
purposes.
Additional work concentrated on the second-stage reaction.
155
-------
Curve 648709-A
90
70
"8 60
o
rs
18
oc
5 50
oo
CD
O
CO
IX
40
30
20
10
o TG 95 38% Sulfated
v TG 96 26% Sulfated
CO/C02 = 2/1
CO =20%
P = 10 Atmospheres
T= 1508°F
Dolomite 1337
CaS + 4 COo
10 20 30 40 50 60 70 80 90
Time/Minutes
Figure 79: Reduction of Sulfated Dolomite
156
-------
3.16 EXPLORATORY RUNS
In exploratory runs, a dolomite was sulfided and regenerated
at atmospheric pressure, and a limestone was sulfided and regenerated
at 10 atmospheres.
The atmospheric pressure experiment involved sulfidation of
Tymochtee dolomite in 15% H S. In fact, the magnesium content of the
stone sulfided, and exposure to nitrogen containing 10% HO resulted
in rapid hydrolysis of the magnesium sulfide. Some hydrolysis of the
calcium sulfide permitted resulfidation, and it appeared that sulfida-
tion was slower on cycles after the first one.
The pressurized experiment, at 10 atmospheres, was carried
out with three sulfidation runs and two regenerations in cyclic fashion.
The course of reaction observed was as follows:
• 92.5% of CaO reacted to CaS.
• 56.5% of CaS reacted to CaC03, then to CaO.
• 100% of the CaO converted to CaS.
• 47% of the CaS reacted to CaC03, then to CaO.
• 100% of the CaO formed converted to CaS.
The rates of sulfidation were the same in each cycle up to
30% utilization.
3.17 TESTS WITH POWDERED CALCIUM SULFIDE
At atmospheric pressure and 1112°F, 8% HO and 8% CO achieved
4.1% conversion to carbonate in four hours, and the rate of reaction over
the last hour was < 0.2 mole % hr . On heating the sample to 1472°F,
to measure the carbonate formed, the reaction revived, the rate
_i
increasing to 8% hr . Apparently, the rate of reaction is very
sensitive to temperature in this region. Regeneration was attempted
using 50% HO, 50% C09 at 1112°F. Within two minutes, 2% conversion
-1
was attained but the rate fell off to 1% hr . The temperature was
157
-------
then increased at a constant heating rate, and the rate of reaction
rose to 0.6% min at 1562°F. At this temperature, the calcium carbonate
decomposed to oxide, and continued conversion of the calcium sulfide to
calcium oxide was observed, the rate at 1652°F being 3% min . At the
start of this experiment, the powdered calcium sulfide was held in
steam/N_ for thirty minutes, and no weight change was observed.
A sample of powdered calcium sulfide was then reacted with
50% H 0/N at 1562°F; the rate of reaction was 0.8% min~ initially,
and 70% conversion was achieved in 133 minutes.
A test at 10 atmospheres on powdered CaS was then carried out
at 1292°F. Conversion to carbonate was ^ 10% in one hour.
It was concluded that the reaction should be studied at the
highest temperature consistent with the thermodynamic requirements for
the hydrogen sulfide off-gas concentration, or about 1292°F.
3.18 TESTS WITH SULFIDED DOLOMITE
TG 124 was an attempt to regenerate a previously sulfided
-4
batch of dolomite. The dolomite contained 1.450 x 10 moles of sulfide
according to iodometric analysis. The weight gain in the experiment
-4
during regeneration indicated that 1.452 x 10 moles of carbonate was
formed. Decomposition of the carbonate by heating showed that 1.432 x
-4 *
10 moles of carbonate had formed. It was concluded that 81.3% of the
calcium sulfide had been regenerated (Figure 80) .
At the end of the experiment, some of the dolomite particles
were embedded in plastic resin and ground to expose a cross section
through the interior of the stone. The sample was then stained with
lead acetate solution. The center of the particle stained heavily,
and the outside rim remained white, indicating that the bulk of the
residual sulfide was held in the core of the solid.
*
N.B. The remainder of the carbonate formed was the product of calcium
oxide recarbonation.
158
-------
Curve 650065-B
-30+40
86%Ca2+Sulfided
1292°F 10 Atmospheres
[H.O] = [C02]=20%
CaS . MgO + H20 + C02 = CaCOj . MgO + H
12 16
Time/Minutes
Figure 80: Regeneration of Carbonate from Sulfided Dolomite
159
-------
The rate of reaction in experiment TG 124 was reproducible, as
determined by repetition of the experiment. The rate was in reasonable
agreement with that reported by Pell at atmospheric pressure:
0.88 sec"1 Pell
1.23 sec"1 TG 124, TG 125.
For TG 124, assessment of the carbonate formed by reaction
was confirmed by decomposing the carbonate thermally. Agreement within
1% was noted.
3.19 CURRENT AND FUTURE EXPERIMENTS
Experiments have been commenced to determine the activity of
calcium carbonate formed on regeneration and the number of cycles in
which a stone sample is useful for desulfurization.
A sample was sulfided to ^ 60% sulfidation, resulfided, and
regenerated. Figure 81 shows the calcium sulfide content of the stone
after the sulfidation step and after the regeneration step. In all,
3.37 moles of sulfur are transferred from desulfurizer to regenerator
per mole of calcium in the sample. In each cycle, the regeneration
step was continued until the rate became extremely slow; choice of the
degree of sulfidation in each cycle was arbitrary. Experiments on this
system are continuing.
The utility of these experiments in relation to fluidized
bed combustion regeneration systems will depend on the degree of success
achieved in identifying inert calcium sulfide. There are three current
postulates as to why regeneration does not succeed after reduction of
the sulfate:
• The morphology of calcium sulfide produced on reduction
of calcium sulfate differs markedly from that produced
in the sulfidation of calcium carbonate. This could be
a combination of crystallite size and porosity effects.
160
-------
GO
5
O-J
rv\
u\
vO
01
>
i_
3
O
T K 1 I 1
i \
\ \
\ \
\ \
\ \
- D O,
\ \
\ \
\ \ 2
DO 15
\ \ i-
* ^ 1
\ \ CD
n o Q£
1 »
! \
i •
i i
i i
o ?
c 1 i
o . '
~ 1 1
»
"5 \ Nx
00 v \
\
\
D
\
\
^
1
I
1
D
1 1 1 1
°x oo r— S ir\
(SBO % JB|OLU) UO!}B2
| I | |
,>_ (/) O CNJ
f^ CSJ LL. O 31
i— ' LL. 0 O ^ ^
C3 ir\ i— i i— i i— < ^
-eg -a •— ' cs ,:-• £
*^i ^^ fc*~ QJ
„, o 'o .2 "ro S"
QJ ^^ "m *~" O
1 i " i S -^
^O S "TO 3 CD {-j
0^^3:1^ Q: i— i
V
\
\
vx
^
X
\
\
vo
1 1 1 1
§ ^ S S
nijjn LunpiBQ %
-
—
—
—
-
-
-
I— I
s
CD
'E
oo 2.
Q
CD
"<5
i
O (U
^0 2: 31
>^ C
( ^ ^5
*j"^
to
v_
CD
c
CD
CT1
^r
p-H
OO
CD
3
.2>
Ll_
CXJ
o
o
. — « ^*
^£
I/) f-H
^ ^-^
(/) t/1
c c
is is
c
-S2
-I-*
ai
c
-------
Examination of samples by electron microscopy and porosity
measurements on samples prepared in a fluidized bed should
help in determining the influence of this factor.
• Eutectic melting or incipient melting caused by a calcium
sulfate/calcium sulfide eutectic during reduction may
seal the internal surfaces of the stone. Careful deter-
mination of this eutectic and the conditions under which
it forms should help to estimate its effect.
• The oxygen activity over the stone during the reduction
may establish the structure of a regenerable or a
nonregenerable calcium sulfide. Studying the kinetics
of regeneration after different reduction treatments
would determine this effect.
The current status of other work on this process is summarized as follows;
• Experiments by Consolidation Coal Company on the second
stage of the regeneration process have demonstrated
regeneration in fluidized bed experiments over a number
25
of cycles (up to 18). The results are similar to
those shown in Figure 81, in that there is a continuous
build-up of inactive calcium sulfide in the stone.
• Argonne has demonstrated that cyclic two-step regeneration
yields only a small amount of regeneration after the first
cycle, whether judged by the calcium carbonate produced
or the concentration of hydrogen sulfide yielded in the
? f\
effluent gas stream.
Since TG experiments have yielded the same kind of results
that fluidized bed experiments have demonstrated, it is recommended that
current experiments to probe the fundamental nature of sorbent inertness
to regeneration be pursued.
162
-------
3.20 CONCLUSIONS
• A thermogravimetric analysis system (TG), modified
to operate at high pressures and high temperatures
to study the reactions of corrosive gases with solids,
- Produced quality data on sulfur removal/sorbent
regeneration
- Provided the capability to isolate chemical
reactions for study while maintaining close
simulation of plant conditions of temperatures,
pressure and gas composition
- Provided a reliable and efficient method for
obtaining design data.
• High utilization of calcium in limestones/dolomites
for sulfur removal can be achieved by adjusting the
carbon dioxide partial pressure during calcination.
The technique can be applied to a pressurized fluid
bed combustion system to approach the stoichiometric
requirement of sorbent stone in a once-through system,
thus reducing the tonnage of by-product solids and raw
stone requirements.
• The investigation of the regenerative process using
the pressurized TG system yields results in close
agreement with those obtained in large-scale test
units. None of the available data, to date, show
adequate regeneration under conditions suitable for
sulfur recovery.
• Use of the thermogravimetric analysis systems to
investigate the fundamental chemical problems in
the regeneration systems offers the possibility of
improving these processes for commercial application.
163
-------
• Data show that TG experimental results can be accurate
predictors for kinetic limits of sulfur removal in the
pressurized fluidized bed boiler system.
164
-------
4. REGENERATION SYSTEMS ASSESSMENT
4.1 INTRODUCTION AND SCOPE OF INVESTIGATION
The operation of the pressurized fluid bed combustion process
with coals presently available to most utilities in the U.S. results in
the production of large quantities of partially utilized dolomite or
limestone materials in the form of CaSO,. This sorbent material may be
either regenerated to a form suitable for repeated SO- removal in the
fluid bed boiler or disposed of in its partially utilized form in a
once-through system.
Regeneration processes are presently under investigation by
Esso Research and Engineering Co., Argonne National Laboratory, and
Westinghouse Electric Corporation to determine the specific nature and
behavior of various processes on a small scale. The question of whether
or not any of the regeneration processes currently being investigated
are economically and environmentally attractive on a commercial scale»
based on the results of small-scale experimental work, should be dealt
with before any process is operated on a large scale. Dealing with this
question should indicate the relative merits of the once-through sorbent
operation and should also indicate the need to pursue new regeneration
concepts.
In the present work, the one-step and two-step regeneration
processes currently being considered are studied. The design, performance,
and cost of a commercial utility power plant utilizing pressurized fluid
bed combustion with sorbent regeneration is projected. The potential problem
areas and the most sensitive cost factors are established for each of the
regeneration processes. The fluid bed power plants utilizing the regen-
eration processes are compared to the once-through operation of the fluid
bed combustion power plant, a fluid bed combustion plant utilizing the
Wellman-Lord stack gas cleaning process, and a conventional coal-fired
power plant utilizing limestone wet scrubbing or the Wellman-Lord stack
cleaning process for pollution control.
165
-------
The study selects base designs for each of the regeneration
processes from the available experimental information. Optimization or
detailed equipment design is not undertaken. Standard methods for
preliminary cost estimations are applied.
4.2 REGENERATION PROCESS CONCEPTS
The economics and performance of three regeneration processes
which function to regenerate the utilized SC^ sorbent produced in the
boilers of the pressurized fluid bed combustion power plant, have been
projected. These processes are a one-step process operated at 10 atmos-
pheres pressure, the same one-step process operated at 1 atmosphere
pressure, and a two-step regeneration process. A sensitivity
analysis has indicated the potential of the one-step process at both
high and low pressures (Appendix L). In addition to the consideration
of these continuous regeneration process operations which follow the
variable power plant load, the concept of a constant load regeneration
process has been examined. In this constant load case, storage capacity
is situated between the boilers and the regeneration process to allow
the regeneration process to operate continuously at a fixed sulfur load,
or with on-and-off control. Analysis of once-through operation in which
the utilized sorbent is disposed of without further processing has also
been carried out to weigh the economic advantages and disadvantages of
sorbent regeneration.
4.2.1 One-Step High-Pressure and Low-Pressure Regeneration Concepts
The one-step dolomite/limestone regeneration process consists
of a single fluidized bed reaction vessel in which utilized dolomite/
limestone (CaSO,) from the pressurized fluid bed boiler (coal-fired)
is reacted with a H_/CO reducing gas to produce CaO and S0«. The reaction
CaSO + {H2} t CaO + {H2°} + SO (1)
CO C02
166
-------
requires temperatures on the order of 2000°F to produce high levels
(1 to 15%) of S09 at reasonable rates. The S0« equilibrium concentration
is favored by reduced pressure, being inversely proportional to the
total pressure. The competing reaction
CaSO + 4{H2} £ CaS + 4{H2°1 (2)
CO C02
also occurs to produce the unwanted product CaS.
The regenerated dolomite/limestone is returned to the fluid
bed boiler with the required amount of fresh dolomite/limestone to
supplement the regenerated sorbent's reduced activity. The S0_ stream
from the regenerator is sent to a sulfur recovery plant in which sulfuric
acid or elemental sulfur is produced.
The one-step regeneration process is broken down into five
interrelated elements: a regenerator element, a reducing-gas producer
element, a sulfur recovery element, a Glaus plant tail gas handling
element, and a sorbent circulation element. These elements and their
relationship to one another are shown in Figure 82. The high- and low-
pressure processes are conceptually identical. The simple block flow
diagram indicates the basic streams in the process. Fuel, air and/or
steam is converted to a H-/CO reducing gas in the reducing—gas producer
element. This reducing gas is combined with tail gas recycled from the
sulfur recovery element to make up the reducing gas provided to the
regenerator. Utilized sorbent from the fluid bed boiler is transported
to the regenerator element to produce a regenerated sorbent and an S0_
stream. The regenerator product gas is processed in the sulfur recovery
element, with a portion of the sulfur recovery tail gas being recycled
to the regenerator and the remainder processed in the tail gas handling
element. The performance and design of these five regeneration process
elements are interrelated in a complex manner.
167
-------
.sj
LTV
CO
J-
csi
CD
O °°
t_3 fO
05 O
en
5
^5 03 C
«_ > O>
"^ 8 I
o~> 03 .2J
ct: QJ
i- ro
=3 I—
C/0
o
-•—» -*—*
03 <->
i_ —3
O> -T-)
cr o
O) 1-
03
O
^
03
O3
00
03 c:
o =
•- c
03 O
03
O3
a> *±
o cz
3 03
Q_ UJ
03
=3
o .
00 Q
to
ra
O 0
c
O3
E
cu
oo
CD
8
D_
O
"fn
CD
CD
en
0}
O.
CD
I
CD
C
o
CNJ
OO
CD
CD
-a
_
•a —
is
168
-------
4.2.2 Two-Step Regeneration Concept
The two-step sorbent regeneration requires two fluidized bed
reactors in series to carry out the conversion of calcium sulfate to
calcium carbonate. In the first step calcium sulfate is reduced to
calcium sulfide by the reaction
CaS04 + 4{£°} £ CaS + 4{R ^} (3)
at about 1500°F and 10 atmospheres. The second step converts calcium
sulfide to calcium carbonate by the reaction
CaS + H20 + C02 + CaC03 4- H2S (4)
at about 1250°F and 10 atmospheres. The regenerated sorbent is recycled
to the boilers while the H~S stream is sent to a sulfur recovery plant.
The two-step regeneration process provides a low temperature route to
regenerated sorbent.
Figure 83 illustrates the elements of the two-step process.
The process consists of seven elements: the calcium sulfate reducer
element, the H?S generator element, the reducing-gas producer element,
the C0~ recovery element, the sulfur recovery element, the tail gas
handling element, and the sorbent circulation element. Reducing gas for
the first-step sulfate reduction is provided by the reducing—gas producer
element, while CO for the second-step calcium carbonate production is
provided by recovering CO- from stack gas and recycled sulfur recovery
tail gas. Tail gases from the calcium sulfate reducer element and the
CO™ recovery element are incinerated or recycled to the fluid bed boilers
in the tail gas handling element. The sorbent circulation element
carries out the function of circulating the sorbent between the fluid bed
boilers and the regeneration process. Elemental sulfur is recovered in
the sulfur recovery element.
169
-------
o
CM
cn
"u
CO
CD
ta
CD
r- fO
la-
-*- o -•-
o 0-2
00 ^ UJ
o
CD
E
.2
UJ
(/I
CD
O
O
ta
L_
CD
C
CD
cn
CD
a.
CD
I
o
oo
CD
=3
cn
fO
O
CD
C.
J L
_CD
'o
CQ
fO
o
o
13 C
1 CD
CD _O
J*£ &_
fa o
170
-------
4.2.3 Constant Load Concept
In order to eliminate the problems involved with designing the
regeneration process to follow the power plant load, storage of the
utilized sorbent and the regenerated sorbent between the fluid bed boilers
and the regeneration process may be used to allow the regenerating process
to be operated at a constant load. This concept is shown in Figure 84.
Solid sorbent heat exchangers would be required to insure operability and
to avoid large thermal efficiency losses.
The concept could be applied to either the one-step or two-step
regeneration. If the power plant load fluctuates over a short time
period (say one week) then sufficient storage capacity can be supplied
so that the regeneration process can be designed at a capacity equal to
the average plant load rather than the maximum plant load required in the
variable load concept. Such capability can lead to a large cost reduction
in the cost of the regeneration process. Fluctuations in the power plant
load may occur to the extent that the regeneration process must be
designed for maximum plant capacity.
4.2.4 Once-through Operation
Once-through operation with pressurized fluid bed combustion is
a simplified operation because the calcium sulfate produced in the boiler
is in a stable form suitable for disposal without further processing. The
cost of the process is minimized by operating with maximum sorbent utiliza-
tion. Problems of control are minimized by going to the once-through
operation.
4.2.5 Wellman-Lord Stack Gas Cleaning Process
The pressurized fluid bed combustion power plant can also be
operated without including a sulfur sorbent in the fluid bed boiler and
instead using a final cleaning of the plant stack gas with the Wellman-
Lord process. The Wellman-Lord process is a regenerative process produc-
ing elemental sulfur and has been successfully operated on a large scale.
171
-------
vD
S
O
c
01
S
CD
^2
S-
O
CO
+
CO
CD
.
00
03
L_
o
ro
O.
O)
O
C
o
o
•a
ro
O
C.
03
t/>
C
O
OO
CD
rj
cr
55 a>
03 r-
O _ .=
o
CO
.
o
o
I
p
^ CO
172
-------
Sorbent regeneration and disposal would be eliminated by this concept
at the cost of the problems associated with stack gas cleaning.
Economic and environmental projections using this concept are summarized
in Section 4.7.
4.2.6 Alternative Concepts
The desulfurization concepts considered for pressurized fluid
bed combustion in this study are not meant to represent an exhaustive
collection of alternatives. Many other alternatives exist. Sulfur
sorbents other than limestone or dolomite might improve the process
performance, economics, and environmental impact. The possibility of
alternative regeneration processes for dolomite or limestone also exists.
These would include both high-temperature and low-temperature concepts.
Scrubbing systems other than the Wellman-Lord process could be considered
to operate with the pressurized fluid bed power plant. The scrubbing
process could also be operated at pressure to desulfurize the cooled gas
before it enters the gas turbine. Also, other plant cycles might be
considered. These alternatives, and others, should be dealt with in some
detail to assess their potential value.
4.3 PROCESS OPTIONS AND VARIABLES
The design options and process variables which are of importance
to the proposed concepts in Figures 82 and 83 are listed below.
PROCESS OPTIONS
• Reducing-gas producer fuel and process
• Source and process for CO production
• Recovery of elemental sulfur or sulfuric acid
• Recycle of process tail gas to boilers or incineration
• Limestone or dolomite as sorbent
• Disposal of spent sorbent before or after regeneration
• Number of regenerator modules
Reducing-gas for the reducing stages of the one-step or two-step
173
-------
regeneration processes can be produced by gasification; partial oxidation;
or reforming of coal, char, residual oil, distillate fuel oils, or methane.
These fuel options are examined in Appendix L. The source of CO for the
second stage of the two-step regeneration process is also a process
consideration having many options. Recovery of elemental sulfur or sulfuric
acid from the SO. or H_S streams produced in the processes are additional
options. The tail gas from the regeneration process must be either
incinerated and released to the atmosphere or recycled to the fluid bed
boilers. Incineration might be followed by energy recovery (turbine
expander or waste heat boiler) and final clean-up of S0~. Both limestone
and dolomite are candidates for sulfur sorbents, each one having possible
advantages and disadvantages. The spent sorbent could be disposed of
prior to or following the regeneration step. The choice of disposal
location (before or after regeneration) can greatly affect the cost
and performance of the regeneration process. The number of regenerator
modules in the plant depends on the plant turndown requirements and the
capabilities of the various elements to operate at reduced load.
The items in the following list represent a simplification of
the complete set of variables which describe the regeneration process.
Those variables having the greatest effect on the process design and
performance are included.
PROCESS VARIABLES
• Regenerator vessel temperature
• Regenerator vessel fluidization velocity
• Mole fraction SO or H S from regenerator
• Reducing-gas composition to regenerator
• Weight fraction sulfur in the boiler coal, W
• Sorbent make-up rate, m, moles Ca/mole S
• Boiler sulfur removal efficiency, e
B R
• Sorbent utilization in the boilers, X , and after regeneration, X
O O
• Sorbent purchase plus disposal cost, $/ton
174
-------
Boiler conditions: temperature, pressure, excess air,
calcination of the sorbent or carbonization
Process sulfur load: pounds of sulfur processed in the
regeneration process per pound of coal fed to the boilers
ID
Process sulfur load = W (e - mx ) for sorbent disposal
before regeneration
TJ
Process sulfur load = W (e - mx ) for sorbent disposal
after regeneration
4.4 BASE CASE DESIGNS
Parametric studies of the economics of one-step dolomite
regeneration have considered a wide range of operating and design
variables. These studies are included in Appendix L. In order to gain
clarity and obtain the most probable regeneration process configuration
and cost, base designs of the alternative regenerating processes described
for use with pressurized fluid bed combustion have been prepared on the
basis of the best existing experimental information.
The design specifications are outlined below:
DESIGN SPECIFICATIONS
• Fixed coal rate to boilers - 430,000 Ib/hr (^ 635 MW)
• Sorbent type - dolomite
• Boiler sulfur removal - 95%
• Variable process sulfur load - 0.01-0.06
• Recycle of process tail gas to boilers or incineration
• Recovery of elemental sulfur
• Reducing-gas production from coal only (methane for
sulfur recovery)
• Four regenerator modules (one for each boiler module)
The designs have been carried out for a fixed coal rate to the boilers
rather than a fixed plant MW capacity. Dolomite is considered rather
than limestone. High levels of sulfur removal are specified (though
they may not be required to meet S0_ emission regulations with all
coals). Variable process sulfur loads are considered (See listing in
175
-------
TABLE 29
DESIGN ASSUMPTIONS
Process
Case
10
Atm |
1
Atm
One-Step Dolomite Regeneration
Regenerator pressure, atm
Regenerator temperature, °F
S02 mole fraction, %
CO + H2 mole fraction in reducing gas, %
Sulfur recovery efficiency, %
Dolomite utilization in boiler, %
Dolomite utilization after regeneration,
Dolomite make-up rate, moles Ca/mole S
Fluidization velocity, ft/sec
Boiler conditions
Dolomite temperature to regenerator °F
Calcium sulfide in regenerated sorbent, >
10
2000
1 and 2
1.5 and 3
95
30
10
1
5
calcination
1200
0
1
1900
10
18
95
30
10
1
5
calcination
1200
0
Two-Step Regeneration
CaSO^ reducer pressure, atm
CaSO- reducer temperature °F
H2S generator pressure, atm
H^S generator temperature, °F
H2S mole fraction, %
CO + H2 mole fraction in reducing gas, %
Sulfur recovery efficiency, %
Dolomite utilization in boiler, %
CaS04 reduced in reducer vessel, %
Dolomite utilization after H«S
generator, %
Dolomite make-up rate, mole Ca/mole S
Fluidization velocity, ft/sec
Boiler conditions
Dolomite temperature to CaSO, reducer, °F
9
1500
11
1250
3
30
95
30
100
10
1
5
carbonation
1200
176
-------
Section 4.3). The options of recycling the process tail gas to the
boilers or incinerating the tail gas are considered. Recovery of elemental
sulfur is specified, and coal is assumed to be the only source of reducing
gas (except for the methane needed for sulfur recovery or incineration).
CO,, is recovered from stack gas and sulfur recovery tail gas in the two-step
process.
Design assumptions are listed in Table 29. For the one-step
regeneration process the SO- mole fractions and the reducing gas
requirements are projected from Esso's data. Two S0» mole fractions are
considered in the high-pressure case. A temperature of 1900°F is used
in the low-pressure regenerator based on the difficulty of temperature
control if the temperature were set at 2000°F. Boiler conditions have
been assumed to be calcination conditions with the one-step regeneration
process and carbonation conditions with the two-step regeneration process.
Either condition may exist in the boiler depending on the boiler pressure,
temperature, and excess air. The boiler conditions chosen for the two
regeneration processes are the optimum boiler conditions for the
respective process. If the one-step regeneration processes must handle
carbonated dolomite or if the two-step process must handle calcined
dolomite, then the complexity and cost of the processes will increase
drastically. The details of the base designs are presented in Appendices
D, E, and F.
The basis on which energy costs were computed is presented below:
COST BASIS
• Mid-1973 costs
• Capital charges at 15%/yr
• O&M at 5%/yr
• 70% capacity factor
• No credit for recovered sulfur
• Coal at 45C/MM Btu
• Methane at 80C/MM Btu
• Operating water at 8c/M gal.
177
-------
4.5 ECONOMIC RESULTS FOR BASE CASE DESIGNS
Capital and energy costs for the four variable load base
designs are considered and the effect of credit for recovered sulfur is
indicated. The costs of the constant load concept applied to each
of the regeneration processes are projected. The assumption of the state
of the sorbent in the boiler (calcined or carbonated) is shown to be a
critical factor in the cost and performance of a given regeneration
process. Finally, the plant energy costs with the regeneration processes
are compared to the plant energy cost with once-through operation, as a
function of the sorbent cost.
4.5.1 Process Capital Investment
Results of capital cost and energy cost estimates are shown in
Tables 30 and 31. Table 30 gives the capital investment in $/kw as a
function of the process sulfur load for the three regeneration processes.
The two-step regeneration process requires the greatest investment of
the processes considered. The sensitivity of the investment to the SO-
mole fraction in the regenerator product is indicated by the results
for the one-step process at 10 atmospheres. The investments are comparable
to investments for a limestone wet—scrubber system (See Appendix N).
4.5.2 Plant Energy Costs
Energy costs for the total power plant are given in Table 31
as a function of the process sulfur load. The table indicates that the
energy cost of the one-step processes at 1 atmosphere and 10 atmospheres
(2% S0?) are about the same and the energy costs of the one-step process
at 10 atmospheres (1% SO-) and the two-step process are about the same.
Again, the sensitivity of the energy cost to the SO mole fraction is
indicated by the 1 and 2% SO- costs for the 10 atmosphere one-step process.
The $2/ton dolomite cost assumed in the table may not be a realistic cost
178
-------
C/D
S3
O
W
P
w
C/3
O
CO
W
<:
H
I
co
CO
cu
o
o
M
PM
O O r-~
• • *
CM s± r~~
rH
<4H
CO
CO O
CU 4J
O
O CO
M tO
ft 60
4H rH
O -H
(3 4-1
O
•H *4H
4-1 O
CO
s-i cu
2J ^
CJ O
o o
C O)
I-H a;
3 o
179
-------
but is not of importance when comparing regeneration processes having
identical make-up rates. The figures in Table 31 do not include a credit
for recovered sulfur. Table 32 indicates what this credit would amount to
as a function of the process sulfur load and the sulfur cost.
Cost breakdowns for the regeneration processes are presented
and discussed in Appendices D, E, and F.
4.5.3 Constant Load Concept Economics
The capital investments for the constant load concept applied
to each of the regeneration processes is shown in Figure 90. It has been
assumed that the fluctuations in power plant load occur over a short time
period and are reproduced throughout the year so that the design capacity
of the regeneration process can be proportional to the plant capacity
factor. A capacity factor of 70% has been assumed. Details of the design
are shown in Appendix J.
The energy cost corresponding to each of the constant load
regeneration processes is shown in Table 34. The plant energy cost is
substantially increased over the variable load costs in Table 31 for the
one-step process at 10 atmospheres and the two-step process. The energy
cost of the low-pressure one-step process is only slightly increased
over the variable load concept energy cost. The advantages to be gained
from constant load operation may overweight this small increase in cost.
4.5.4 Effect of Boiler Conditions
The base regeneration process designs have been evaluated with
the assumption applied that the boiler conditions have provided the optimum
sorbent conditions entering the regeneration process. This means
specifically calcium oxide entering the one-step regeneration process (calcina-
tion in the boilers) and calcium carbonate entering the two-step process
(carbonation in the boilers). Since this optimum behavior may not occur
during actual plant operation, it is important to note the effect of the
opposite boiler conditions on the process cost and performance. Conservative
process assumptions are applied to obtain the results discussed.
180
-------
1
CO
W
t— I
pq
<
H
cfl
s~^>
^
CO
rH
rH
•H
3
CO
§
1— 1
CO
w
Q
H
CO
3
1
H
§
, ^
CLi
03
K
o
CM
<&
O
PM
H
CO
o
QJ)
>-l
O
P3
w
52;
w
a
cu
4J
Tf\
1
O
15
H
c
e o
4-1 CO
<3 ^0
B-S
O CN
rH
1 _
ft
0)
4-1 C
CO O
1 cn
cu
C 6-S
O H
s
4J
<3
rH O
1 CO
ft
o) s-s
•u o
CO rH
1
0)
C
O
TJ
rocess
fur Loa
PLI rH
3
CO
-o- in csi
CN co in
rH rH rH
si
O> 00 O
• # •
rH CN ^t
rH rH rH
J
CO in rH
CN co in
rH rH rH
v|
cri r^ oo
• • •
rH CN CO
rH i-H rH
rH CO VfJ
O O O
• * •
O O O
C
o
4J
fsl
o
Vi h >J
3 to >,
ft42 ^
O 6-«
a) in
4-> rH
•rH Cfl 14-1
0 4J 0
0 -H
rH ftS
O cfl pq
003
4J O 4-1 2
O o) pq g
Cfl V4 ~^.
<4-l S O
n S o
K^ O *^* 00
4J K-l O
•H m 4J
O 4-1 ^1- Cfl
cfl -H
ft 13 4-) 0)
cfl 0) cfl (3
O M cfl
0 rH jr-
6-S cfl 4J
O O O 0)
r- S3 U S
cd
/•—s
s
Ai
CO
rH
rH
•H
g
v_x
CO
w
CN J
co
C
o
4-1
— *.
0
r-l
-CO
4-1
Cfl
VJ
3
<4-l
rH
3
CO
i
TJ
cfl
CO O
CO i-l
0)
a >-i
0 3
VJ 4-1
PM rH
3
CO
o a\ oo
rH CN in
• • •
o o o
i£5 CT> CTi
O rH CO
• • »
O O O
co o a\
O rH rH
• • •
O O O
rH CO SO
O O O
• • •
O O O
•
>»
4-1
•iH
O
CO
ft
cfl
CJ
4J
a
cfl
rH
ft
/— \
0)
^x
^r!
^
m
CO
vo
TJ
a
cfl
>•.
o
c
CU
•rH
CJ
•rl
>4-l
IM
01
>~,
^
0)
>
o
o
01
I-l
M
3
4H
rH
3
CO
6-«
m
CTl
c
0
•X3
01
CQ
cfl
pq
cfl
181
-------
CO
CO
[V]
,-J
PQ
2;
M
^
H
1 |
P-i
O
i— 1
[5
**^_
CO-
.
4-1
C
OJ
e
K
0)
>
c
M
o
a
CO 4-1
a -H TJ
O O -rl
O cfl 3
O< • rH
00 Cfl tO >4-i
p! O CO
•H 00 O
3
T3
4-1
CO
(U
J-<
01
4J
c
•H
(U
T3
3
rH
O
C
•H
O
1^,
«\
CO
4J
co
O
O
CO
r>
, — |
I
T)
•H
B
*
rH
•H
CO
4-J
co
CO
0)
a
o
^i
ex
(4^
o
G
0
•H
4J
CO
tfl
00
1— 1
•H
CO
4-1
CO
to
0)
o
o
p.
M-l
o
O 4-1 M 0)
C -H ,
(1) P, CJ O
O n) C 0)
Q O M pr!
tfl ,£> O
ro
W
pq
H
CO
rH
•H
H
PH
W
CJ
5Z
O
c_>
o
^
H
H
CO
§
^
g
m
H
CO
o
u
o
w
P
0)
4-1
CO
1
o
&
EH
p
S 0
4-1 CO
a-
1 1 ^1
1—1 r-1 f— 1
i
-------
Some reduction in cost may be possible by improving the process
s chemes.
If calcination occurs in the fluid bed boilers and the two-step
regeneration process is used for sorbent regeneration,then extensive
carbonation will occur in the H_S generator vessel. The exothermic nature
of the carbonation reaction will disrupt the H_S generator temperature
control and the requirement for maximum H~S concentrations in the product
gas. Thus, the carbonation of the sorbent must be carried out in a
separate vessel prior to the sorbent regeneration. Carbonation may be
carried out by recycling plant stack gas. Table 35 shows the effects of
calcination boiler conditions on the energy cost of the two-step process.
The results are sensitive to the rate of sorbent circulation and the
extent of calcination. Details are discussed in Appendix H- Table 35
indicates an extremely large increase in the plant energy cost compared
to the results for carbonation boiler conditions in Table 31.
If carbonation of the sorbent occurs in the fluid bed boilers
and the one-step regeneration process is used for sorbent regeneration
(high-pressure or low-pressure operation), then extensive calcination will
occur in the one-step regenerator vessel. The endothermic nature of the
calcination reaction will make temperature control and maintenance of
high S0_ concentration impossible. Thus, calcination must be carried
out in a separate vessel prior to regeneration. Heat for the calcination
must be supplied by the combustion of coal in a separate coal combustor.
The fuel required for this process step will normally be some large
fraction of the fuel rate to the fluid bed boilers (depending on the
sorbent circulation rate and the degree of carbonation in the boilers)
so desulfurization of the hot product gas from the calcination vessel and
recovery of the hot gas energy is required. Due to the large scale of
this calcination process, the regeneration process would have to become
an integral part of the power generation system. This would require
major modifications to the present fluidized bed combustion plant
design concept and would greatly complicate the plant operation. Details
are considered in Appendix H.
Since the fluid bed boiler conditions (temperature, pressure,
excess air) change during turndown of the power plant, the sorbent may
183
-------
R)
CO
W
u
o
&
PH
a
o
H
H
W
w
w
§
PM
W
i
o
ts
H
t^f
in o
01
-I O
23 H
< H
H M
Q
|
Pi
W
1-4
O
pq
O
M
H
g
I |
CJ
J
O
^
£-H
U
w
fr ,
pr_)
w
V-i to
Q) H
C! rH
W -tH
cfl «
4-1 4-1
O CO
u
4-1
3
O O
*l PJ
4J M
CO
O H
t-I CD
3
P*^ f1-!
00
M M-)
0) O
C
^"> r^
603
M r^
0) ^~.
C CO
W rH
rH
a -H
•H fa
cu
cfl «
01 4-1
)-4 CO
0 O
PJ O
M
_
^
P! ~~~-
•H o CN
CN n
rH CO VO
o o o
* * •
0 O 0
•
0)
4J
•H
B
0
rH
0 .
TJ CO
CO
T) 0
0) rH
Cfl r^>
Vl O
0) fi
P: at
0) vH
00 O
CU -rH
>4-4
IH oi
0
CO
P! -H
o ^
•H -U
4-1
n) C
C 0
•H
O T3
rH Q)
cfl co
a cfl
&-S ^
O CO
0 • -H
i — 1 4-1
** o, ^g
O 4-1 ^
• O CO
0 rH
1 1 I
•*-* r~i
II C -H
cfl B
X! P, P!
•H
" M-t
O O 4J
ro CO
• rs o
O S O
II in t>^
01 00
pq c/>O M
>
-------
switch between calcined and carbonated states during turndown. This is
an additional operating complication to be considered for regeneration
processes. The boiler condition (calcination or carbonation) is a
critical factor to consider when comparing regeneration processes. The
state of the sorbent has a small effect upon the energy cost of once-
through operation (See Appendix I).
4.5.5 Process Cost Comparison
The energy costs of the four base regeneration process designs
and once-through operation are compared in Figures 85 and 86 as a function
of the dolomite purchase plus disposal cost. The figures consider 95%
sulfur removal of a 4 wt % sulfur coal. A dolomite utilization of 30%
in the boiler and a regenerative make-up rate of 1 mole Ca/mole S are
considered. Once-through make-up rates of 2 and 3 moles Ca/mole S are
shown. Figure 85 indicates the process costs for the case of sorbent
disposal from the process before the sorbent is regenerated, while
Figure 86 represents the case of sorbent disposal after the sorbent has
been regenerated. The energy cost basis listed in Section 4.4 has
been applied to the base conditions also listed there and in Table 30
to obtain the results.
The case requiring sorbent disposal after regeneration yields
higher energy costs than the case requiring disposal before regeneration.
Disposal before regeneration gives smaller process sulfur loads than
disposal following regeneration. The extent of this effect on the plant
energy cost depends on the sorbent utilization before and after regeneration
(See listing of variables in Section 4.3). The chemical nature of
the disposed sorbent wanted is a further consideration. The figures
indicate a large cost advantage for once-through operation over the
regeneration processes considered even for high dolomite costs. Once-
through make-up rates less than 2 moles Ca/mole S may be feasible,
leading to greater cost advantages. The effect of boiler conditions
must also be considered. The plant operation is also simplified with
185
-------
Burve 652052-B
T
T
15
Regenerative Process with Dolomite Disposal before
Regeneration; 30% Dolomite Utilization, m = lroQ|eCa
Once-through Process with Calcination _
in Boilers
14
Two-Step Process (Carbonation)
One-Step Process -10 atm , S02 = 1% (Calcination)
One-Step Process -10 atm , S02 = 2% (Calcination)
1 atm , S02 = 10% (Calcination)
One-Step Process-
4 wt.% Sulfur in Coal
95* Sulfur Removal Efficiency in Boiler
Disposal before Regeneration
12
11
I
4 6 8 10 12
Dolomite Purchase ^Disposal Cost, $/Ton
14
16
Figure 85: Plant Energy Cost (Disposal before Regeneration)
186
-------
Curve 652053-8
15
14
o
g;
I 13
UJ
c
.2
a.
12
11
I
I
——— Regenerative Process with Dolomite Disposal after
Regeneration; 30% Dolomite Utilization, m = 1™Q™ y
— — — Once-through Process with Calcination _
4wt. % Sulfur in Coal I
95% Sulfur Removal Efficiency in Boiler t
Disposal after Regeneration
mole S
I
I
I
I
I
4 6 8 10 12
Dolomite Purchase + Disposal Cost, $Aon
14
16
Figure 86: Plant Energy Cost (Disposal after Regeneration)
187
-------
once-through operation. Further details of the once-through process are
given in Appendix I and cost comparisons are considered in greater detail
in Appendix K.
4.6 PROCESS PERFOEMANCE
The fluidized bed plant heat rate, regeneration process
temperature control, regeneration process turndown, and the overall
environmental comparison of the regeneration processes are considered.
These performance factors are necessary items to examine in order to
understand the value of the various processes considered.
4.6.1 Plant Heat Rate
The plant heat rate is shown in Figure 87 for the four base
design cases and for once-through operation. The heat rate is expressed
as a function of the process sulfur load, W (e - mx ), for the regenera-
1 O b
tive processes or the factor W m x 10 for the once-through operation.
O
For the once-through operation the resulting heat rate for the cases of
calcination in the boiler and carbonation in the boiler are shown. Values
of the regenerative process sulfur load may approach 0.06 for very high
sulfur coals, while the factor W m x 10 for once-through operation would
O
probably never exceed 0.025 even with very high sulfur coals. Once-
through operation or the 1 atmosphere one-step process yield the lowest
plant heat rates. The 10 atmosphere one-step regeneration with 1% SO-
in the regenerator product yields the highest plant heat rate.
4.6.2 Temperature Control
The feasibility of temperature control of the regeneration
processes has been investigated and is discussed in Appendix G. The
major consideration of temperature control feasibility is whether or
not the required amount of energy can be supplied or withdrawn from the
regenerator vessels while maintaining the specified temperature and con-
centration of S02 or H2S in the regenerator product. It has already
been stated that temperature control cannot be obtained in the
188
-------
Curve 652054-B
12,000
4-f
00
a
re
OH
S 11,000
o
v>
3
U
1 10,000
o
•2
00
S
2 9,000
I
I
— One-Step Process - Calcination in Boilers
d) 1%S02, 10 Atmospheres
§2* S02, 10 Atmospheres
10* S02, 1 Atmospheres
•— Two-Step Process - Carbonation in Boilers,
— Once-through Process
(4) Calcination in Boilers
G) Carbonation in Boilers
_L
I
I
I
I
I
0.01 0.02 0.03 0.04 0.05 0.06 0.07
Wste -mXs) for Regenerative Processes
Ws mXlO-1 for Once-through Operation
Figure 87: Plant Heat Rate with Regeneration Alternatives
189
-------
regeneration processes if the sorbent is calcined in the boiler with the
one-step process. This problem was handled by using separate vessels
to process the sorbent before regeneration.
The two-step process is a low-temperature process compared
to the one-step regeneration process,and temperature control is easily
obtained in this process. With the one-step process,the temperature
of the reducing gas entering the regenerator vessel becomes higher as
the S0~ mole fraction in the regenerator product increases. For the 10
atmosphere case where SCL mole fractions are 1 to 2%,the temperature
control is feasible without excessively high temperatures; but with the
1 atmosphere case, where mole fractions up to 20% could be expected at
2000°F,temperature control is not possible. For this reason,the one-
step process at 1 atmosphere pressure requires a reduced vessel tempera-
ture of 1900°F and preheating of the incoming dolomite to 1900°F in a
separate vessel, yielding an S0~ mole fraction of about 10%. The highest
SO- mole fraction which would give feasible temperature control would be
about 12 to 13%, without having heat transfer surface in the reactors.
4.6.3 Process Turndown
The base designs have assumed that four regenerator vessels are
used by the regeneration processes (one for each boiler module) to
minimize the turndown required for each vessel. Though a variety of turn-
down concepts are possible, a most attractive concept appears to be for
the regenerator temperature and pressure to be maintained constant during
turndown to yield a constant SO- or H S concentration to the sulfur recovery
element during turndown. The sorbent circulation rate would be maintained
constant during turndown to provide simpler control of the sorbent circula-
tion equipment. The fluidization velocity in each vessel would be propor-
tional to the fractional load on each vessel during turndown. With four
regenerator modules »no more than 50% turndown of each vessel would be
required. Turndown of other elements in the regeneration process have
not been considered in detail, though the reducing-gas producer element
and the sulfur recovery element may be problem areas,and multiple modules
190
-------
may be required to obtain high performance during turndown. The option of
operating once-through during turndown at levels where the regeneration
process performance is low should-be considered. Some details of
regenerator turndown are presented in Appendix G.
4.6.4 Process Environmental Comparison
The four base designs for regeneration are compared with once-
through operation of the pressurized fluid bed combustion power plant on
the grounds of environmental factors in Table 36. For each base design
two columns are indicated: column A for disposal of dolomite before
regeneration and column B for disposal following regeneration. For
once-through operation two fresh dolomite feed rates are considered:
m = 3.0 and m - 1.5 moles Ca/mole S fed which corresponds to 31.6 and
63.3% dolomite utilization in the boilers, respectively. The design
basis is summarized on the table.
The first two factors, plant heat rate and plant energy cost,
indicate the overall impact of the processes on the energy picture. The
plant heat rate and the plant energy cost should be lower for once-
through operation than for the regenerative processes, even with a high
dolomite cost of $10/ton. Other factors of importance are the plant
raw materials and the plant outputs. Coal and methane requirements are
greater with the regenerative processes than they are with the once-through
operation at the cost of the increased dolomite consumption required for
once-through operation. Ash output is comparable to dolomite waste for
the regenerative processes. Sulfur removal efficiency is expected to
be greater for once-through operation than for regenerative operations,
as indicated in the table, because of reduced chemical processing.
If the total United States electrical utility industry were to
convert completely to pressurized fluid bed combustion, then the rate of
coal consumption would be greater for conversion to regenerative operation
191
-------
W
W
I
W
CO
CO
o
o
P-I
Pn
o
o
CO
§
o
u
i a
O H
o
§
a) I
s
B (3
iH rH
OH PL.
g a 2 g
ft O 4) -H
r-t -H > W
4-> d tO
. U O 3
cn 3 o .a
• *o B
3 O 41 O
. B B iH tn
-------
than for conversion to once-through operation. The relative rate of
coal consumption for regenerative and once-through operations is
shown in Table 36 and ranges from 4 to 18% greater for regeneration.
Consumption of the valuable fuel methane is another factor to be
considered. The price paid for this reduction in coal and methane
consumption (in terms of environmental impact) is indicated in the last
factor in the table. For complete conversion of the United States
electric utility industry to pressurized fluid bed combustion,about 10%
of the present United States limestone/dolomite production would be
required for regenerative operation and 15 to 30% for once-through
operation.
These preliminary considerations should add some perspective
to the advantages and disadvantages of regenerative and once-through
operations.
4.7 ASSESSMENT
4.7.1 Economic Factors
The plant energy cost for pressurized fluid bed combustion,
utilizing dolomite regeneration and once-through operation, is compared
to the plant energy cost for a conventional coal-fired power plant using
limestone wet scrubbing for pollution control in Figure 88. Curves for
the one-step process with 1% SO (10 atmospheres) and 10% SO (1 atmosphere)
are shown. The costs for the one-step process with 2% S0_ and the two-step
process are not shown because they have costs nearly identical to the
costs for the one-step with 10% SO (1 atmosphere) and the one-step
with 1% S0_ (10 atmospheres), respectively. The cases of dolomite
disposal before and after the regeneration are shown. Both limestone
and dolomite costs have been taken as $2/ton. The effect of dolomite
cost is shown in Figures 85 and 86.
Figure 88 indicates that the energy cost of the conventional
plant increases at a slower rate with increased coal sulfur content than
does the cost of the fluid bed combustion plant. The cost of the
193
-------
Curve 652051-B
15
14
d>
c
OJ
12
11
T
T
T
£) One-StepProcess at 10 atm, S02=l%
J) One-StepProcess at 1 atm
Disposal prior to Regeneration
— — Disposal after Regeneration / —
Once-through,; m=3, Calcination /
Convention Plant with Limestone /
Wet scrubbing
Regenerative Make-up = 1 Mole Ca/Mole S /
Sorbent Utilization in Boilers = 30% x
Sulfur Removal Efficiency = 95% / /
_Dolomite/Limestone at$2/Ton
\
\
01234567
Wt. % Sulfur in Coal
Fiqure 88: Comparison of Fluid Bed Power Plant with Conventional
Power Plant
194
-------
conventional plant is greater than the fluid bed combustion plant for all
cases except the one-step process with 1% S0_ with dolomite disposal
following regeneration. The costs for the one-step process with 1% S0_
with disposal before regeneration and the two-step process are comparable
to the conventional plant energy cost at high-sulfur contents. The
energy cost reduction with the one-step process with 10% SO (1 atmosphere)
and the one-step process with 2% SO is indicated to be about 1 mill/kWh
or greater. The cost reduction with once-through operation is shown to
be about 2 mills/kWh with a low dolomite cost of $2/ton. The once-
through cost reduction compared to the conventional plant will be about
1 mill/kWh with dolomite and limestone costs of $10/ton. The basis for
conventional plant costs is discussed in Appendix N.
TABLE 37
GENERAL COMPARISON
• FBCQM with Regeneration
Optimistic Case Pessimistic Case
Optimum boiler condition Optimum boiler condition
Ca/S molar ratio, m = 1/2 Ca/S molar ratio, m = 1.5
Sorbent at $10/ton Sorbent at $10/ton
Disposal before regeneration Disposal after regeneration
No sulfur credit No sulfur credit
• Once—Through
Optimistic Case Pessimistic Case
Ca/S molar ratio, m = 1.2 Ca/S molar ratio, m = 3.0
Carbonation in boiler Calcination in boiler
Sorbent at $10/ton Sorbent at $10/ton
• Conventional Power Plant with Limestone Wet Scrubbing
• FBCOM Plant with Stack Gas Cleaning (Wellman-Lord)
• Conventional Power Plant with Wellman-Lord Stack Cleaning
195
-------
A general comparison of energy costs between fluidized bed
combustion with dolomite regeneration, fluidized bed combustion with
once-through operation, fluidized bed combustion with Wellman-Lord
stack gas cleaning, a conventional power plant with limestone wet
scrubbing, and a conventional power plant with Wellman-Lord stack gas
cleaning is shown in Figure 89 . Table 37 indicates the conditions applied
in Figure 89, with costs ranging between optimistic conditions and pessi-
mistic conditions for regenerative and once-through fluid bed combustion
power plants. Figure 90 indicates the effect of the plant capacity factor
on the energy cost of regenerative and once-through fluid bed power
plants, while Figure 91 shows the break-even cost of clean fuel as a
function of the plant capacity factor. Evidently, the once-through
operation becomes more economically attractive than the regenerative
operation as the plant capacity factor decreases. The pressurized fluid
bed combustion plant using the Wellman-Lord stack gas cleaning process
has many potential improvements in efficiency, because no sorbent is
present in the fluid bed boiler, which may reduce the energy cost below
that indicated in Figure 91. The cost should be attractive compared to
a conventional plant.
In general, the potential exists for pressurized fluid bed
combustion power plants to reduce the cost of electrical energy generated
by conventional power plants by 5 to 15% if the regeneration and once-
through performance projected at the present time can be realized in
commercial application.
4.7.2 Environmental Factors
Tables 38, 39, and 40 compare environmental factors for the
alternative power plant concepts considered. Resources and wastes are
listed in the tables. Table 38 compares regenerative fluid bed combus-
tion and once-through operation. Table 39 compares once-through opera-
tion of the fluid bed combustion power plant to the fluid bed combustion
plant with Wellman-Lord stack gas cleaning. Table 40 compares the fluid
bed combustion plant with Wellman-Lord stack gas cleaning to a conventional
coal-fired plant with Wellman-Lord stack gas cleaning.
196
-------
MM)|/S||!UJ '
EC: o= zaCo o5
O
o
m, o
c
s 8
3 §_
o
O
03
c
0}
o
o>
l_
C71
e
C3
>*
O
rr\
o
cvl
197
-------
130
Curve 653^79-B
120
110
£ 100
_- 90
o>
3
C
ID
•S 80
o
u
0>
70
60
50
40
30
Base Cost of High-Sulfur
Coal (4.3% S)
20 30 40 50 60 70
Plant capacity factor, %
80
90
Figure 91: FBCOM with Low-Sulfur Coal
198
-------
00
co
W
w
H
t/j
PC
B
o
BS
H
W
S3
H
13
M O
4-1
CO
i
o
S
H
c
0
CO
c\
6
4J
cd
p.
a) o
4-1 rH
CO
1
(11
c
o
s
-M
cd
rH
§
4J
M
o mo
OO rH
O CM
n\
o»
o •* o
00 CO
r~ CM
rr\
CT>
vl
O co r^
CO CO VO
00 CM
•• CM
CTi
O CO vO
in CM -^
CO CM •
CM
cd
" "" —
o> CM
4-1 ^ 0
cd s-i co
S-I ,C
**^ s
4-1 CO g
cd /-x c o-'
CU rC O
,C! IS 4-1 0)
CO Al ^ d
(U 4J -~- Cd
0 C 3 rH ,£
S-I Cd 4-1 CO 4-1
3 rH pq o a)
O PL, *-s u S
CO
01
o
vO
in
1
J~~1
0
' 0
rH
0
o
ft
rH
o
o
rH
cd
TJ
— ^
CO
0
4-1
^*-s
cu
4-1
•H
o
rH
o
a
o
CO
rH
rH
in
rH
in
rH
^^
r>>
cd
13
CO
c
o
4J
•*~s
S-i
3
"4-1
rH
CO
o
vD
cri
O
I
m
o
O
a\
o
I
in
o
i
in
o
O CM
OO VO
o m
O
O1
O
m
o
i
in
o
o o
oo in
in
rH
CM
• O
O I
I m
rH O
oo
m
co
in
3
4-1
pq
CO
cu
4J
CO
cd
o
•H
4-J
3
1
r^
rH
0
ft
S-i
•H
-3
removal
3
4-1
pq
O
rH
CM CM
O
CO
rH
! 1
cd
S-i
cu
>
o
O
a
^Q
rH
•\
X
a
o
rH
i— I
CO
O)
4-1
cfl
J
^1
3
o
•H
4J
S-i
cd
P-l
lative
0)
S-J
rH
cd
e
S-i
0)
T-C-J
H
cd
CO
C
O
4-1
M
CO
-d
•H
rH
O
CO
0)
4-1
.p-|
O
iH
O
T3
,13
CO
cd
cd
i
cu
S-i
S-i
3
3
co
0)
rH
•H
O
,£>
cd
o
o
(3
•H
SJ
3
u-i
rH
3
CO
m
CO
CO
•H
CO
199
-------
w
1
Ifl
S
H
in
CO
O
Z
hH
3
O
t/5
O
rrt
M
H
S
1
O
erf
H
M
O
2*
O
t*4
O
23
o
CO
^H
05
1
o
4J
S
Id
Pi *O
a o
0 rJ
*H 1
4J d
to rt
3 J
43 rH
B rH
o *tf
r^ • *H •
9i
O rH O O O O
CD vO
O in
{^ rH
£
1^
-*s^
3
U >•
M Id >.
TJ (0
(U CB <4H "-^
4J 41 X TJ O CO
co j> id **^ co d
lJ -rt T3 CO O
4-1 n) co o S
CD rH d *->
4) 4) O M 4)
CO r( d -H
41 4J « « 3 (d B
U d iH a MH 43 O
t-l Cd Id O rH U rH
3 rH O Id 3 4) O
O fc U SS CO S Q
CO
4)
04
O IN
d
1
rH
•
O
* CM
9>
O
1
O
M
f<
^^
o
d
•H
y
>H
U-l
M-l
(U 3
4J
rH «
id
> vO
O O
§rH
p.-
m
rH
•
O
1
o
•
0
4J
vD
O
rH
rH
•t
^ CO
d rl CN 41
O O *J
•H CN Z Id
*J 0
3 CO 43
rH rH
•H rH
O rH
(3, Id !
(0 rl 41 S3
•H
4J
id
rH
41
M
M
rH
Therma
3
•H
O
CO 01
» 01
0) *J
id d
IM O
rH iH
3 «
CO 4J
I1?
•H (3.
tj
O -
M 01
^J
1 id
r* rH
O 3
O CO
o
• o
O iH
rH 43
^s 4J
O
O
m
o
x-\
id
TJ
^•^
CO
4J
l«H
^.^
(TJ
TJ
V)
d
o
4-1
M
TJ
cr
•H
_)
en
41
[ 1
rH
3
U
•H
)^
cd
o.
41
B
o
43
*H
3
d
o
•H
4J
3
rH
O
(0
1
1
0
in
4)
4-1
•H
B
o
o
i
^
O
Srd
• (0
iH
O *O
vo m
M
rH
x^
td
TJ
^^
CO
4-1
VM
VH'
id
TJ
en
a
o
4-1
n
CO
•H
rH
0
•
rH
^
0
§
U
sulfur
M
01
rH
•I-l
O
43
m
ft
i™" 4
rd
o
o
B
•H
tl
3
U-l
rH
3
0)
f>{
vD
•
~*
*
/- S
i
s
200
-------
c/>
H
PM
pq
p
o
M
H
Z
W
-o- o
W O
u
H
c/)
3
O
c/)
4J
c
cfl
-a
o c
PH cfl
rH rH
CO rH
C 0)
O 2£
01
O
U
4J a
CO cfl
43 rH
e H
O 0)
u rs
01 4-1
pq -H
•H
y
O CO VO
o o
o . o
m
CN
o
CTi
ro
o
o
o
•
o
O CO
co cO
I
r--
r^
in
O
O
ro
o
CTi
CN
CN
O
I
O
O
CO
A
CO
cO
O
vO
m
CO
0)
CJ
M
o
CO
01
Pi
3
4-1
pq
01
4J
cd
cfl
01
C
cfl
rH
PL,
cfl
O
U
co
C
O
4-1
EC
O
cfl
CO
a
o
3
<4H
H
3
cfl"
13
MH
O
CO
01
C
cd
CO
01
CO
CO
val efficienc
6 Btu
f
c
0
•H
4-1
3
rH
rH
O
fX
)-l
•H
<
O
e
0)
M
CN
O
C/3
rH
rH
CO
V-l
01
>
O
o
1-1
— ^
CN
O
a
42
rH
n
X
o
3
3
4-1
pq
D
0
rH
5
n
CO
01
4-1
CO
T-H
3
O
•H
4-1
)-l
CO
CW
O)
>
•H
4-1
cd
rH
cu
^
r\
rH
CO
e
r4
0)
42
H
>^
cfl
T3
CO
C
O
4-1
*,
id
•H
3
cr
•H
HJ
>.
cfl
13
CO
C
O
4-1
*
CO
73
•H
rH
o
00
cd
I
Cfl
!H
01
rH
•H
O
cd
o
o
C
•H
J-l
3
3
CD
ro
CO
•H
CO
CO
pq
CO
201
-------
Regenerative operation requires more coal and methane than
once-through operation while once-through operation requires more dolo-
mite disposal. The net environmental effect of these factors must be
considered with great care. Once-through dolomite waste levels may be
reduced to values nearly as low as with regenerative operation by obtain-
ing high utilization. CaSO, is an inert material having many potential
uses. Coal and methane are extremely valuable natural resources.
The Wellman-Lord process requires more coal and make-up NaOH
than the once-through process. The liquid disposal stream indicated in
Table 39 for the Wellman-Lord process can be eliminated by evaporation
to dryness using a low-pressure stream. NaOH manufacture requires the
expenditure of large amounts of energy and is equivalent to about a 2%
efficiency reduction in the fluid bed combustion power plant using the
Wellman-Lord process.
The fluidized bed combustion plant using the Wellman-Lord
stack gas cleaning process compares favorably with a conventional coal-
fired power plant using the Wellman-Lord process. Further improvement
in the performance of the fluid bed combustion plant using the Wellman-
Lord process is possible.
4.7.3 Base Design Feasibility
While both optimistic and pessimistic cost assumptions have
been made throughout the base regeneration process designs, it has been
optimistically assumed that all the process steps are commercially
feasible. Some process steps are questionable, as noted in
Table 41 where the major problems associated with each of the regenera-
tion processes are considered. The questionable areas in the regenera-
tion process are such as will lead to reduced power plant
availability and high maintenance costs compared to the once-through
operation. Higher energy costs than have been indicated for the process
base designs may also result from the factors discussed in Table 41.
202
-------
TABLE 41
ASSESSMENT OF BASE DESIGNS
LOW PRESSURE ONE-STEP REGENERATION PROCESS
Excessively High Reducing-Cas Inlet Temperature to Regenerator
• > 3000°F even with sorbent preheated to 1900°F for 10% SO
• Sorbent may be deactivated by high temperature near distributor -
fusion of impurities, increased crystal size.
• Materials problem
• Inlet temperature can be reduced by decreasing the S0? fraction below
10% by (1) increasing the regenerator pressure or (2) reducing the
regenerator temperature below 1900°F. Option (2) may lead to excessive
CaS formation and possible eutectic formation with CaSO .
• Partial combustion of coal in the regenerator rather than externally
may eliminate the high-temperature problem but may introduce other
problems,
Complex Sorbent Circulation Element
• Includes lock hoppers, valves and feeders, sorbent cooling and heating
vessels
• Represents > 50% of process capital investment
• Involves potential sorbent losses, attrition, deactivation, and
sulfur emissions
• Cost is sensitive to sorbent circulation rate; base design assumes
change in sorbent utilization of 20% (optimistic assumption)
• Similar heating and cooling equipment has been operated commercially -
moving bed shaft kiln at 1900°F with gas, oil, or coal as fuel; fluid
bed aluminum hydroxide calciner at ^ 2000°F; FluoSolids staged lime
kiln by Dorr-Oliver; Pebble bed heat exchanger.
• Dense phase sorbent transport rather than dilute phase could eliminate
lock hoppers, valves, feeders, and the necessity for cooling the sorbent.
This would require an extremely high elevation of the regenerator and
long stand legs.
203
-------
TABLE 41 (Continued)
Commercial Feasibility of Reducing^Gas Generation from Coal
• Lurgi and Wellman-Galusha are commercial coal gasifiers for atmospheric
pressure.
* May require size coal, steam; may be low efficiency with difficult
control of outlet temperature and CO + H_ concentration
• If coal can be partially combusted directly in the regenerator with the
same performance assumed in the base design, the capital investment would
be reduced by ^> 10%,assuming the regenerator vessel cost does not
increase (optimistic assumption).
Sorbent from Boiler in Carbonated Form
• Requires separate calcination system before regeneration
• Dolomite retains activity for CO pick-up (^ 60% still active after
10 cycles). Limestone is less active (^ 33% after 10 cycles).
• Increases complexity and cost (^ 60% increase in base design energy
cost for 100% recarbonation) - sensitive to sorbent circulation rate
• Increased potential for sorbent losses, attrition, deactivation, and
sulfur emissions
• Commerical calcining processes are well established.
• Also, conditions may change from calcining to carbonating in the boiler
during turndown.
Sulfur Recovery
• Allied Chemical has commercial process for sulfur recovery down to
4 to 5% S02 (^ 12 $/kw for 10% S02).
• Shell report on sulfur recovery used in base design has optimistic
costs (^4 $/kw for 10% SO ).
• Contaminants in regenerator product gas may cause problems (0?, HO,
tars, metals, particulates).
204
-------
TABLE 41 (Continued)
• Dilute or concentrated sulfuric acid may be recovered from dilute SO
gases - demonstrated by Parsons Co. on smelter's gas down to 2% SCL.
Complex Process Control
• Integration of the regeneration process elements to follow the power
plant load will be difficult. Each element has its own turndown
problems.
• The constant load concept is a possibility with this process. Once-
through operation during stages of turndown is a possibility.
HIGH-PRESSURE ONE-STEP REGENERATION PROCESS
Commercial Feasibility of Sulfur Recovery
• Base design assumes highly optimistic costs in Shell report. (3 to 4
$/kw not including incineration at 1 to 2% SO^)•
• Allied Chemical process would cost ^ 25 $/kw (0.95 mills/kWh), for
3% S02 and probably > 50 $/kw (1.8 mills/kWh) for 1% SO . Allied
Chemical process implies greater complexity than Shell design.
• Contaminants such as tars and particulates have been neglected.
• Sulfuric acid recovery is probably not commercial for < 2% SO .
High Reducing-Gas Inlet Temperature to Regenerator
• > 2300°F for 1% SO and > 2500°F for 2% SO . Increasing during
turndown - > 2800° F at 50% turndown and 2% SO .
• Possible sorbent deactivation near distribution
• If coal can be partially combusted directly in the regenerator, with
no change in performance and no increase in the cost of the regenerator
vessel, then the capital investment can be reduced by about 17% but with no
fuel saving. About ^ 10% reduction in regeneration process energy cost
205
-------
TABLE 41 (Continued)
or ^ 2% reduction in power plant energy cost. May eliminate high-
temperature problem - may introduce other problems.
Reducing Gas Generation from Coal
• Base design assumes a very low level of CO + H2 required in the
reducing gas (1.5% CO + H2 for 1% S02 and 3% CO + H2 for 2% S02>; it
has not been experimentally demonstrated.
• With low CO + H assumption the reducing gas can be generated using
a coal combustor operated slightly below stoichiometric combustion
rather than coal gasifier.
• Commercial coal combustors (slagging) are available, but Lurgi is the
only commercial gasifier at pressure.
• If higher CO + H levels must be used, then the coal consumption and
cost for reducing-gas production increases ^ linearly with the fraction
CO + H^. CaS production also increased. (Doubling the base design
fractions of CO + H increased the process energy cost by ^ 30%).
Sorbent from Boiler in Carbonated Form
• Requires separate calcination system before regeneration
• Dolomite retains activity for C0_ pick-up. Limestone is less active.
• Increased complexity and cost-sensitive to sorbent circulation rate
• Increased potential for sorbent losses, attrition, deactivation, and
sulfur emissions
• Commercial calcining processes are well-established
• Conditions may change from calcining to carbonating in the boiler
during turndown
Complex Process Control
• Integration of the regeneration process elements to follow the power
plant load will be difficult. Each element has its own turndown
problems.
206
-------
TABLE 41 (Continued)
TWO-STEP REGENERATION PROCESS
Commercial Feasibility of Reducing Gas Generation from Coal
• Lurgi is the only commercial coal gasifier at pressure.
• May require sized coal and steam; may be low efficiency with difficult
control of outlet temperature and CO + H concentration.
Commercial Feasibility of CO Recovery
^
• Commercial processes for C0_ recovery exist (such as the Hot Carbonate
process).
• The effects of contaminants which may be present in Claus plant tail
gas or power plant stack gas and the turndown characteristics have
not been examined.
High CaS Content in Regenerated Sorbent
• CaSO, reduction step gives high conversion to CaS in laboratory
studies. H S generating step gives poor conversion to CaCO in batch
tests.
• High CaS content in regenerated sorbent may cause eutectic melt to
form in boiler.
• Large circulation rates will be required with utilization changes of
5 to 10%. High utilization in the boiler will require a large quantity
of reducing gas.
Sorbent in Calcined Form in Boiler
• Requires separate carbonation system before regeneration
• Increases complexity and cost of process. Cost increase is highly
sensitive to the sorbent circulation rate.
207
-------
TABLE 41 (Continued)
• Increased potential for sorbent losses, attrition, deactivation, and
sulfur emissions. Disposal product may contain CaS.
• Conditions in boiler (calcination or carbonation) may change during
turndown.
Sulfur Recovery
• Recovery of sulfur from low H S gas (>_ 3%) streams is probably
commercial, but also probably more costly than indicated in the
Shell report used in the base design.
Complex Process Control
• Integration of the regeneration process elements to follow the power
plant load will be difficult. Each element has its own turndown
problems.
4.8 COMBUSTION OF LOW-SULFUR COALS
Considerations have so far been with the fluid bed combustion
power plant operated regeneratively or once-through with coals having
sulfur contents greater than 1 wt % and requiring high sulfur removal
(^ 95%). The case of low-sulfur coals which require little or no
desulfurization is also of importance since such coals are attractive
fuels and are still found in some areas of the United States.
The capital costs of the pressurized fluid bed combustion
plant and the conventional power plant with limestone wet scrubbing
are shown in Table 42. A 1 wt % sulfur coal is considered with 70, 50,
and 0% desulfurization required. The plant energy cost is shown in
Table 43 for the same cases. As was indicated in Figure 89, the cost
differential between fluidized bed combustion and conventional power
generation becomes larger as the coal sulfur content is reduced. An
11% reduction in energy cost is projected for the case of no desulfurization.
With no desulfurization required, many improvements in the design of
the fluidized bed combustion processes are possible. These are discussed
in Appendix M.
208
-------
CN
W
n . .
M H
a
§ S
c_)
d M
o ca
H 0
N M
H 3
>-i M-4
3 -H
3
CO
CO S^
CTi
CN
ro
CN
oo
00
in
CN
ro CN
o
in
CM
O
CO
CD
OJ
CO
o
4J
n)
CO
co
cu
o
o
a
o
cu
CJ
60
01
a
cu
4J
CO
cu
a
o
c
o
01
CO
cfl
pq
209
-------
4.9 CONCLUSIONS
Pressurized fluid bed combustion combined cycle power
generation plants with limestone/dolomite sulfur
removal have lower energy costs and offer environmental
advantages over conventional plants with stack gas
cleaning.
A once-through sulfur sorbent process is the most
attractive — economically and environmentally — for
first-generation pressurized fluid bed boiler combined
cycle power plants. Energy costs could be reduced more
than 10%, air emission criteria met, and the spent
sorbent utilized or disposed of without further
processing and with no environmental pollution.
Once-through dolomite/limestone operation should be
demonstrated on the pressurized pilot plants now being
A
built and the pressurized fluid bed combustion boiler
demonstration plant.
Further development of regeneration processes is
required before they can be demonstrated on a
commercial scale.
Experimental work should continue to
- Achieve high stone utilization for once-through
operation
- Improve present regeneration process concepts
- Develop new regeneration processes
- Study alternative sorbents to limestones/dolomites
The cost advantage for a pressurized fluidized bed
combustion plant over conventional plants with stack
gas cleaning is achieved with high-or low-(e.g. 1%)
sulfur content in the fuel.
*EPA contracts to Esso and Argonne National Laboratory.
210
-------
The cost advantage for pressurized fluidized bed combustion
plants over conventional plants may be greater using
low-sulfur fuels not requiring sulfur removal.
211
-------
5. PLANT OPERATION AND CONTROL
5.1 INTRODUCTION
A philosophy of operation and recommendations for controlling
the pressurized fluid bed boiler power plant have been developed. In
addition, an assessment has been made of the ability of the concept
to meet utility operating requirements. This study presents a definition
of procedures with estimates of key parameters.
It is assumed that the reader is familiar with the operation
and control of large once-through boiler power plants and fluidized bed
combustion boilers. Many normal power plant operating procedures and
control techniques are referred to but not detailed or defined. The
operating procedures contained in this report are based on the normal
operation of conventional boiler and steam-turbine power plants. The
control system recommendations are based on a large, hybrid control
system utilizing digital computers and analog control capability.
5.1.1 Plant Criteria
A power plant that is to operate as a swing plant must be
capable of starting up, changing load, and shutting down easily. In the
plant design, the equipment configuration should be selected to provide
considerable flexibility for plant operation.
Plant operation is based on the criteria that the plant
• Operate as a swing or base load plant
• Be capable of operation from 12-1/2% to 100% of rated capacity,
at design conditions of 2400 psi and 1000°F superheat and
1000°F reheat
• Be capable of changing load at the rate of 5% per minute
• Operate with some of the plant components out of service for maintenance
213
-------
• Be capable of starting up or shutting down components in
several different sequences to ensure maximum usage of the
plant equipment.
5.1.2 Control System Requirements
The recommended control system is intended to provide the
total plant automation and control required to operate the plant effi-
ciently, safely, and with the maximum amount of plant availability. To
accomplish this, the following features must be provided:
• The plant automation required to start up and with the various
plant components required for load changes
• A coordinated control system to operate all of the plant
components properly
• An interface system so the plant can be operated remotely
by the load dispatcher
• A loading scheme that will produce the best heat rate
at all times
• An emergency loading scheme to load the plant in the
minimum amount of time
• A predictive loading program to determine the plant
configuration and loading sequence if the desired load
and time requirements are known in advance
• An emergency or safety system to handle contingency conditions.
Since the operation of a plant must be understood before a
control system can be recommended, the first portion of this report
deals with the plant operation and the later portion with control system
recommendations.
5.1.3 Pressurized Fluid Bed Boiler Power Plant Concept
Details of fluidized bed combustion boilers and the base 635 MW
;d power pi;
the concept follows:
pressurized power plant design are presented elsewhere. A review of
214
-------
In the fluidized bed combustion boiler, sufficient air is added
to the fuel to virtually complete the transformation of its carbon to CO-
and its hydrogen to H?0:
H2 + 1/2 02 •> H20 .
Both these reactions produce large quantities of heat which
are absorbed by generating steam.
Sulfur in the fuel is converted primarily to SO . A limestone
or dolomite sorbent is utilized to remove this pollutant from the com-
bustion gases:
\
CaCO
CaO
S02 + 1/2 02 -»• CaS04
The CaSO, can either be disposed of as a solid — once-through
operation — or be processed to regenerate CaCO,. (or CaO) sorbent
and to recover sulfur as a useful product — elemental sulfur, S, or
sulfuric acid, H SO,. Once-through operation could be advantageous if
high utilization of the sorbent can be realized in the boiler. This
would provide a simpler plant design and operation. Two processes appear
most attractive for regenerating the stone:
• Reduction of calcium sulfate to calcium sulfide and sub-
sequent regeneration with steam and CO,
/4H\ /4HO
CaSO, + / -*- CaS +
y4 co/ U co,
CaS + HO + CO -»• CaCO + H S
• Direct reduction of calcium sulfate to calcium oxide and
SO- employing partially oxidized fuel as the reducing agent:
(H2\
CaSO, + + CaO + SO,, +
215
-------
These alternatives are currently being studied to determine which should
be used for commercial plants. The present evaluation assumes the first
regeneration reaction system is used.
A fluidized bed combustion boiler is formed by an enclosure
usually consisting of water walls — abutting boiler tubes in which water
and steam flow. Coal and sorbent are fed to the bed by pneumatic
feeders extending through the air distributor or the water walls. From
60 to 70% of the heat released in burning the fuel with air is transferred
to the water/steam in the tubes surrounding and submerged in the bed.
Various tube configurations have been proposed — horizontal tubes
extending through the bed, horizontal tubes with serpentine bends in
vertical platens or planes, and vertical tube walls passing up through
the distributor plate, slicing the bed into narrow segments.
Operating at elevated pressure, a fluidized combustor requires
a compressor to pressurize the air and to overcome the pressure loss
over the fluidized bed combustor. The power cycle schematic utilizing
coal (or oil) fluidized bed combustion at elevated pressure is illus-
trated in Figure 6, p. 19. At an operating pressure of 10 to 15 atmospheres,
excess air of 10 to 15%, and a fluidizing velocity of 8 to 12 ft/sec, a
depth of 8 to 15 feet is required to accommodate the heat transfer
surface in the bed. Energy is recovered from the high-temperature gases
by passing them directly into a gas turbine and expanding them to
atmospheric pressure. This expansion lowers the temperature of the gases
by 600 to 800°F, thereby reducing the amount of surface required to
recover heat from the combustion gases leaving the fluidized bed. The
pressurized system can be operated at higher excess air ratios. This
increases the fraction of gas-turbine power, reduces combustible losses
from the boiler, and increases the waste heat recovery after the gas
turbine. This may allow for improved plant efficiencies and provide an
effective means for achieving continuous, rapid plant turndown.
The boiler design consists of four modules. The modularized
design provides for a maximum of shop fabrication and standardization and
216
-------
assists in meeting the turndown requirements for the plant. Each
module includes four primary fluidized bed combustors, each in turn
containing a separate boiler function — one bed for the preevaporator,
two beds for the superheater, and one bed for the reheater. Evaporation
takes place in the water walls. All of the boiler heat transfer surface is
immersed in the beds, except for the water walls. There is no convection
heat transfer surface since the maximum allowable bed temperature is less
than the state-of-the-art gas-turbine temperature. The fluidized bed
combustors are stacked because of the inherent advantages in
vertical gas circuitry, steam circuitry, and pressure vessel design.
Each module contains a separate fluidized bed or carbon burn-up
cell to complete the combustion of any carbon elutriated from the primary
beds. A separate bed may not be required since carbon losses may be low
enough in the proposed pressurized boiler design with deep beds. A carbon
burn-up cell (CBC) is definitely not envisaged if the system is operated
with high excess air. This also has the advantage of increasing plant
performance by increasing the fraction of gas-turbine power. A simplified
drawing of a module is shown in Figure 2, p. 3. The 318 MW plant module
can be shop fabricated and shipped by rail. The 635 MW plant module is
designed to be shipped in sections, each shop-fabricated. The primary
beds for a 318 MW plant are approximately 5 ft x 7 ft. The bed depths
are approximately 12 feet — sufficient for the required heat transfer
surface. The carbon burn-up cell (CBC) is approximately 2 ft x 7 ft.
The CBC contains no submerged surface in the bed. The submerged tube
bundles are formed by vertical tube platens or planes, each platen a
continuous boiler tube in a serpentine arrangement. The heat transfer
surface can be viewed as horizontal tubes. The preevaporator and reheater
contain 1-1/2-in. diameter tubes; the evaporator water walls and reheater
bed contain 2-in. diameter tubes.
A preliminary design plant layout, operating philosophy, and
a cost estimate were prepared for the 635 MW plant. A composite flow
diagram for the base design is shown in Figure 92. A summary of the boiler
2
design and the plant layout has been presented. A summary of part load
operation is presented in Appendix 0.
217
-------
\'
\
1
» X
is
H-
1
!'t
i
T
>"
E^
•^
LI
i
H
uJl
L lil_
35
J
Ct
i
i. i
H^t?
^ .
•o
c
to
13
cn
-------
W-g-
-------
CL
7
k_
ro
1/1
°o
CO
7 X33HS "2- 133HS "O
NJO c] 1NO} NO a,i-NOD C—
A A *
1 on
H31»3HM3dfi5 (TC33S
3w noaj aft3i ijiino j OQJI
SAO 01 ouva 13PJ Haw* SHI isnrcu
i
Missffld 3AWA n 3111 -c IMOJ IBS
3HJ. 5NI15IKOV AS J-.W J.lti » NO
ISd Q03i 01 3tinSS3da SIllOiMl 3St3rf]NI
01 03s.l1 si'iJ L, '^iO £1 1A1VA *
SHI IB.™ K c,s d au s.iuun
™ vjj * I'l r^iv"n
isrs?^rs,sir, ^s^.s^
*
d«31
d3'W33J an 3HTSS]dJ K'Wl HSVld
IDdlMi: 01 JJrUBili 3l»d3dO iJAHIA
D Oi« 0 3MVA 3 IU!» b3lD3ri I OW
ci rtujis ir«iv QW jnivA v mi*
3tK!SS;Ud Vtii HSVlJ ISd OD9 IQdlHO
WU *01J Ml* S5
SMIJIOimj nwaaJ Ol a3dmu 3NIN3 Xa D3S(3H3NI SI U31I
A8 Q36 dVA3 3dd 01 »C"!J dIV 3SD3il3M
t
drt3i il31lr»U13J
ONK IJIJI lOai'DJ 01 U3KJ71D1
31«3dO 3A1VA 0 [J,( 3 3K1 3MVA 3
3H1 Hll* H31^ 341 01 OlilliW
3H1 SCK^U al\31 HFW1 KiOlj 3H! HdhW
CM liA31 HSIiatlSB S3HSU1J M]i(l S9
Q33J leca DMisnriw *a
Q3NH1MV/1 SI d.i31 Nln SIHI d C0:l 01
*3l 0=3 3ilVd 0,'W Q33i n«3 Ji*15
j^osi 01 diwi oia
3Slta ^V W1IN01 039 dVA3 3Md 1HV1S
S
O^
£
^J
.S1
u_
a 3HI3IlQ3a *B
la 01 KCTU a iv
a DNI3n03iJ Xfl
a 01 wnj ui*
IdAB 3HI3Aa3ti
a at imj uiv
idAe DMiancBa
a 01 «rnj uiv
vdAa 3Ni3na3d
a o; «oij HIV
U3dA« 131WI 038 '
a.Nis3
-------
SS2Jd S30V3H
7.
1
i LN03 NO C
f ^
3S013 I'll* S3A1",\ A
OKI '3 'V 3K1 (NV MIBdllt 3H1 Cl QJ TIM WJ3.5
n* 's3Ai»A ADD ]HI NO ioaiN03 3diss3;id mm osi*
ss,S';
Sd 0001 NI*lNI*rt ONV
3A1W UONajAQD 33*1d
ISc
OVD1 4 5 ddtf dP NOld 9to
did 3H1 '*lll SIK1 IV
oooi Oi sansssad MNKI
3033N SV 13CU 3SJ3U3NI
noa imam 3d* swai^m ms»
3NISUni UJV SMJI1IQN33 IW31S N3H1
BNITHB BOd NOIlva
I
S3NI1 IW31S
H31*3Ka3dns~OMJ33S
"" D1 W" T ^
01 *01JMl«l»l3S*3a3NI 03SV310NI SI
I
k
033J 1*03 DNIlSHfO* A8
Q3NIWIWI SI dill H A Sim i 00 il 01
KLmK"*™m"mma
J 051 Dl d«31 038 3SWB (J.V
MOllrffll 038 M31V3HH3dns CMJ31S ld*lS
3m 01 03UIIWK
039 «
MTOW1"1£MKi-1£,J
dig 0]3 3m ONINJdQ U)1J Ul* S
a a3iv3Ha3dns omcss xa 03Svia»ii si
53A1M nan N NSJO
-, t.">l,t U SJQ38 «n^ «l"
01 MfflJ HI* 3SV3a3NI « 03SV3H3NI S
i
Q33J 1K03 HNIiSnfa( 18
03Nl*lN](rt SI dH31 N H SIK1 JJ1QEI 01
dTQi 038 3SIVa &•» 033J 1*10 1W1S
1
J (
DNIH3dO AllvtibVd AB
3m 01 Q311lrtC» S
1QMU03 0^ S:
3 fffl * 3Hi
t 4M QL fW
3A1
ISd 009£ Oi 3UnSS3a
d«31 b3mo
01 a3H13301
3Ai(A 3 mi* ans
UW» lifKSid W
d j a anoi
dWU 0333 IJ31IOG
( Oi dlSli MS 5SIW
vjm,3ons issiu mns
339 s31»3riH3drs 1SMIJ »01J UIV S
an OiJi;UHTlJ NIH AS Q3S*3OII SI
CL
r
133HS *2
\ j>
3
'o
CQ
•K
u!!
§
ssvdAe wiaraM la
U31I08 01 «Q1d UIV
SVdAS 3m ONI3T1C3H
H31ID8 01 MOId UIV
svdia 3x1 DNiDfraa
an ioa 01 lOij ui*
131103 01 K01J HI*
U32D1 31VM3JO S3A1« Ql 3SV3naN 1V1Q1 !1NINI»1NI»«
is urn o.* iM^ » xa anita ai »mj »« 3sva»ii
HSVIJ iSa 009 noai)J03
J 1
h
•« # H1IM 10«i«J3 OJ*
""
3d OW 13A31 lOaiICO
I*a3d0 3A1VA Q Urt 3
H 9 ON Dl H3K* 10H
3 SV3 MV1S dOlS DW
N3MHQ 3NI8H01 iaViS
1 1
1
IfNbON Oi KldM 039
3SOJ3HI OJ* 0333 3N01S 1HVIS
G33J 1(03 3Hiisnrm IB
Q3NinNIVH SI dNHl NlN EIH1 j OOtl
01 dffll 033 3SIW OW 033J 1103 1HK1S
3SIVU Olt MOJIfCI 038
WS
1TINI £338 SNIN3JO
01 «51J ONIZIQI
3A1IA 1&U1N03 dVfU (
ions mm und a
U01VMOd(A3 3Ud 1HV1S
3dd«J2 H31COOQ iaVIS
003 01 3unSS3Hd TUf*
QTfll fGdO 3/1«A 8 INK
i
iwisfca M013 ai* ivioi loaiKri
13H3 aOlSnaMH SSVdJ.8
HI « ovoi Nin wi t loaiuro
1
H3N1S Itb 00% IV
no 13 uoiETBwn ssvdAa iNiauru sw iuns
UOiQtt 9: 1UV1S 3NI3tfU SVS ItttlS
H3dO S3A1VA B (Mf 'fl '• J QNV 03SD13 S3Am
HV33 aNltiJU W 3Nl6Bf\l 33*1.1 ON* 3 OfJt * '3 'A 'N 'A '0 'I IU»d 3NIMQVHU cV.ii,. V
crvovfi wna*is3 'BJSSJOCO iwis tenoam n*3i: mm «is*s asivw nsmj u^ SSAI** SSH
1
BII*» iinroaia 01 ASTOBSN
xnv nliStf B3iV*C33d WaKiO O«
ajiioen* asmo
3JJIS3A1 S*3 NBdO
® ($
rv' ;
c I — I1" '
o a. |p- |jj j
Sh r-'^1^
O !-'• t~ ?
CJ u. S
Sg S
s^ : i
03 5 S
"-1 a - s c S
1 'a j : i |
s " ' Uii s
Is
^,.™, I-
10HVH3 il
1 | , . ^H
1 o | ' '< '
-------
Irrt
-------
q
u
-
>-*•
£ ^
0
O
UJ
cO
g
£
K 2
tsmoi
ZL/J
M S a
— co S
DO
— OC
IDLLIQC
.
f— n: a,
3 CD UJ — —
UZQ
R 3
£
a
«2j
ot/iC
"Z.
— uj o
-
=e vi t— E
O — -etK
U. P I U."
— «t CL U
—
j u- «: to o
O O > UJ I— J
;> i/i
..
4COO, L30O-
« F D H
CD
'o
QQ
in
CD
gg
U4 • go
r?S a ^ h-
, CQUJ — Z
UJ— -S -
t— m x cc i
^ 5 S S m
229
-------
CONT'D
SHEET 1
AIR FLO* TO BOILER IS INCREASED
BY REDUCING THE DYPASS AIR FLO*.
AIR FLO* TO THE BOILERS SHOULD
NO* EQUAL THE TOTAL PRODUCED
BY THE GAS TURBINE AND THE
BYPASS DAMPER IS CLOSED.
CONT D
ADMIT STEAM TO THE TURBINE BY OPENING
THE I VALVE AND CONTROLLING PRESSURE
KITH THE U VALVE.
ADMIT COLD REHEAT STEAM TO THE
REHtATER TO HARM THE BED.
WIN. FLU1DIZIN3 AIR IS ADMITTED TO
THE REHEAUR BED BY PARTIALLY OPENING
THE BED AIR DAMPER.
i
START REHLATtR BiD IGNITOR AND RAISE
BED TEMP. TO 750"F.
START CCAL FFEO AND RAISE BED
TEW. TO I-.OO'F. THIS WIN. IMP,
IS MAINTAIN1 fl BY AHIUJTIN;; COAL FEED.
INCREASE AIR FLCVV BY OPENING THE BED
AIR DJ!,?ER. THIS INCREASE SHOULD NOT
PERMIT 7,iE REHEAT TEVP. TO
EXCEED THE a'PESHEAT TEk'P. BY
MOPE THAN 50'F.
T-
ADJUST *ATES FLOft - FUEL FLOW RATIO IN
EA3-: BED TG k'AINTA'N THE DESIRED
H'ATeS OR STEAM TEVPERATURES.
BALANCE STEAV FLOW TO REHEATERS TO
STEAM FLCW FROM SUPERHEATERS SO
THE ROLLER MODULE ARE BALANCED.
THE SECGtC BOILER MODULE IS NO*
ON LINE READY TC INCREASE LOAD.
Figure 96B: Second Boiler Module Start-Up
233
-------
5.2.2 First-Module Start-Up
The start-up procedure for the first module assumes that the
plant is cold and completely off-line. Only the starting of the gas
turbine, boiler module, steam turbine, and sulfur recovery system are
covered in this description. Starting the condenser, getting the
steam turbine on turning gear, filling the system with water, flushing
the water system, and other conventional start-up procedures are not
covered in this description as they are the same as those for a conventional
subcritical once-through steam plant.
The gas turbine will be started in a manner similar to that for
starting a conventional gas turbine, except that this unit utilizes a
separate combustion chamber. The combustion system will be considerably
smaller than normal. This will reduce the initial temperature overshoot
and make the start-up smooth, with little danger of surge or high turbine
inlet temperatures. The gas-turbine start-up will proceed in the normal
manner with all of the compressor airflow passing through the by-pass
combustor until the turbine has reached synchronous speed and has been
synchronized. The minimum load of approximately 4 MW will be controlled
during the boiler module start-up by adjusting the gas-turbine combustor
fuel flow. The stack gas cooler boiler feed pump will be started and a
flow of 40% of boiler module capacity established. For the first module,
the flow will be controlled with the stack gas cooler boiler feed pump
valve, and the B valve on the module will be open. Air will be admitted
to the preevaporator module by opening the module inlet damper, starting
the booster air compressor, and throttling the by-pass airflow. The flow
rate will be controlled by positioning the by-pass damper. The bed will
be filled with fresh stone at a reduced depth, and the airflow will be
at a minimum value so that the heat required by the igniters will be held
to a minimum. The preevaporator igniters will be started and the bed
temperature increased to approximately 700°F, the temperature required
to ignite the coal.
Coal feed will be started and bed temperature increased to
1300°F. The bed depth and airflow are then increased to the normal level.
234
-------
When the water wall outlet temperature has increased sufficiently,
air will be admitted to the first superheater bed and the igniters started.
The first superheater bed will also be filled with clean stone at a reduced
depth. Coal feed will be started at ignition temperature and the bed
temperature increased to 1300°F. The bed depth and airflow are then
increased to the normal levels.
During this time, the gas-turbine combustor fuel will be
modulated to maintain a minimum load of 4 MW on the gas-turbine generator.
Airflow to the boiler modules will be controlled by positioning the
by-pass valve; and total airflow will be controlled by throttling the
inlet guide vanes on the gas turbine as required.
The start-up progresses as described previously in "Evaluation
of Fluidized Bed Combustion,"Vol.Ill, Westinghouse Research Laboratories
report to Office of Air Programs, EPA, Nov. 1971.
Several operating points are not mentioned in the start-up
procedure of that report. These include the fact that when the
first-stage superheater outlet temperature is sufficiently high, a
turbine-driven boiler feed pump will be started and the feedwater passed
through the feedwater heaters. This permits heat recovery in the no. 6
and 7 heaters. The stack gas cooler boiler feed pump will be shut down
at this time.
The reheater will be fired prior to admitting steam to the
turbine, with the bed temperature limited to 880°F. When steam is
admitted to the reheater, the reheat temperature set point will be set
equal to the superheat steam value. By this means, the reheat and
superheat steam temperature will be the same and the temperature shock
to the turbine will be minimized. The boiler module outlet temperatures
are sufficient to permit shutting off the gas-turbine combustor when
the reheater is fired.
A desuperheater will be interlocked with the U valve so steam
can be sent to the condenser.
235
-------
5.2.3 Second-Module Statt-Up
The second boiler module start-up is similar to that for the
first boiler module except that the first boiler exit temperature will be
sufficient to permit the by-pass combustors to be shut down and the
feedwater will be hot. These differences will cause the following
changes to the procedure described for the first unit:
• Start-up of a gas turbine will not be required.
• The depth of the preevaporator and first superheater
beds will not need to be reduced for ignition.
• The stack gas cooler boiler feed pump will not be used,
as water flow will be through the feedwater heaters;
feedwater flow will be controlled by positioning the
B valve.
Since the feedwater temperature will be considerably higher
than the 80°F of the first module, the heat-recovery portion of the
start-up cycle will be reduced. The E and C valves will be interlocked
so they will not admit water or steam from the flash tank until the
flash tank temperature exceeds the output temperature from the no. 6
heater for the E valve and the no. 7 heater for the C valve.
The I valve cannot be opened until the superheater pressure
reaches 2600 psi and the superheater outlet temperature is 1000°F.
Therefore, the system must be pressurized by use of the U valve. This
process will proceed faster than that for the first unit because no steam
temperature changes to the turbine will be involved.
The reheater light-off can be delayed because the steam can
be diverted to the first boiler module reheater. Also, this module can
be preheated with steam from the turbine prior to light-off. These
features make the light-off of this reheater module less critical than
that of the first module reheater.
The turbine load will be increased to equal the water flow
through the B valve by opening the I valve. The superheater outlet
236
-------
pressure will be maintained by modulating the U valve. After the I
valve is open, the pressure set point for the U valve will be increased
to serve as a relief valve.
When the I valve is 100 percent open, the load will be increased
further by increasing the flow through the B valve. The water flow must
now be controlled by modulating the B valves and holding the feedwater
heater pressure constant by varying the boiler feed pump speed.
The steam flow through the reheaters must be balanced or
maintained proportional to the flow through the B valves by positioning
the reheater inlet valves.
5.2.4 Third-Module Start-up
The start of the third module will be the same as for starting
the gas turbine and airflow system for the first boiler module and for
starting the second boiler module , the difference being the balance
and comparisons of three boiler modules instead of two.
5.2.5 Regeneration System Start-up
The plant will be started up by operating the stone feed once-
through rather than regenerating. This will be accomplished by starting
with clean stone in the beds, with the surge vessel full of clean stone,
and with the hold vessel empty. No stone will be added to or removed
from the beds until S09 is sensed at the boiler outlet. Clean stone
will be added to the bed producing the SO from the surge vessel. The
spent stone will be removed and placed in the hold vessel. The boiler
module can operate in this mode until the hold vessel is full. This
provides sufficient time for the plant operation to stabilize and permits
starting the chemical regeneration system without any adverse effects on
the remainder of the plant.
237
-------
When starting the specific regeneration system is desired, as
shown in Figure 94, the CaSO, reducer vessel and the H_S generator
vessel will be filled with stone from the hold vessel and the two stone
isolation valves closed. The gas producer will be started at atmospheric
pressure and the gas passed through the CaSO, reducer vessel and dis-
charged through the by-pass valve to the stack. The gas throttle valve
to the expander will be closed. The expander will be started with air
from the gas turbine and the air throttle valve used for control. Air
to the compressor will enter through bleed valve no. 1, which modulates
to control the required flow and is discharged through bleed valve no. 3,
which modulates to control the desired discharge pressure. The C0?
isolation valve and bleed valve no. 2 will be closed. The C09 scrubber
will be started by admitting the desired flow of stack gas. This will
be controlled by modulating the stack gas control valve. The stripper
flow will be controlled to maintain the desired temperature in the C0_
scrubber. As CO flow increases, bleed valve no. 1 will be closed and
bleed valve no. 2 opened to maintain the desired minimum flow to the
compressor. Cooling water will be started to the condenser to maintain
the desired CO. temperature. When bleed valve no. 1 is closed and all
of the air has been purged from the system, the surge control will be
transferred to bleed valve no. 2 and bleed valve no. 3 closed. The CO
flow will be started to the H S generator by opening the CO. isolation
valve and controlling the system pressure to approximately 2 atmospheres
with the discharge valve. The system pressures will be increased with
the discharge valve and by-pass valves to maintain the H.S generator
vessel pressure approximately 2 atmospheres above that of the CaSO,
reducer vessel. Pressures will continue to increase until the CaSO.
4
reducer vessel pressure is 2 atmospheres below those of the boiler module.
The HJD to CO- ratio will be established and maintained by admitting
steam to the CO compressor discharge. When the proper pressures are
attained and the system stabilized, the two stone isolation valves will
be opened to start the stone flow. The expander can now be transferred
to the CaSO, reducer vessel discharge gas by transferring from the air
throttle valve to the gas throttle valve.
238
-------
The control systems will be transferred to the on-line mode
from the start-up mode.
5.2.6 Auxiliary Fuel Storage; jlequirements
The estimated auxiliary fuel requirements for bringing four
boiler modules on-line from a cold start are:
Gas turbine 15,500 Ib
Igniters in beds 20,000 Ib
Total 35,500 Ib
For a hot start where three additional boiler modules are brought on-line
from a minimum load condition with one boiler module, the estimated
auxiliary fuel requirements are:
Gas turbine 6,000 Ib
Igniters in beds 12,000 Ib
Total 18,000 Ib
Assuming no. 2 fuel oil with a specific gravity of 0.85, 5,000 gal. of
fuel oil would be consumed in a cold start and 2,500 gal.would be con-
sumed in a hot start.
Assuming intermediate load operation of the plant with one hot
start per day and one cold start per week, the fuel storage capacity
required for once a week delivery and 100% reserves would be 40,000 gal.
5.3 ON-LINE LOADING
5.3.1 Intro due t ion
The start-up of the individual boiler modules is discussed in
Section 5.2, but the discussion does not include when or in what order
the boiler modules should be started.
The plant will contain four boiler modules, with a turndown
range of approximately 2:1. That is, each can operate from 50% maximum
load to maximum load. By operating the plant with various combinations
239
-------
of the four boiler modules, the plant load can be varied from 12-1/2% to
100% of maximum capacity.
The expected time to load the plant will be consistent with
that of a large conventional fossil fired once-through boiler power
plant. The time required for flash tank operation and pressurizing is
the same for either type once-through boiler. For a cold start, the
steam turbine requires a soak time of two to three hours at approximately
2200 rpm. This will permit time for the chemical process to stabilize
and the auxiliary equipment to be placed in operation. The steam turbine
requires a soak time of two to three hours at approximately 2200 rpm.
This will permit time for the chemical process to stabilize and
auxiliary equipment to be placed in operation. The steam turbine requires
approximately 90 minutes to increase the load to 100% rated load from
the time the generator is synchronized to the line. Proper manipulation
of the second gas turbine and the three remaining boiler modules will
permit the fluidized bed boiler to increase steam production as required
to meet the steam-turbine needs for the 90-minute loading.
The plant load can be varied by adjusting the boiler modules
simultaneously or sequentially. Sequential operation of the modules
will require starting and stopping equipment. Simultaneous operation
of the modules limits the load variation range to approximately 2:1.
The fastest load changes are accomplished by adjusting the equipment in
service. Five percent per minute load changes can be achieved by
adjusting fuel and water to the modules. Time is required to start the
modules. Load increases beyond the operating range of the equipment in
service, therefore, will be slower than 5% per minute. Load decreases
can be accomplished rapidly by tripping portions of the plant that is
operating.
Three basic approaches to load changes will be discussed.
• Sequential loading
• Predictive loading
• Rapid loading.
240
-------
5.3.2 Sequential Loading
The sequential-by-module loading scheme will be used for normal
small increases or decreases as received from the load dispatcher. Figure
97 shows the sequential loading scheme on a load versus time (no time
scale),for the total plant at a uniform rate and for the portion delivered
by each module. When an increase or decrease request is received but
the eventual end point is not known, the load on the first boiler module
will be increased as required to 80%. If an additional load is requested,
the second boiler module will be started. The steam flow from the first
boiler module will be reduced as required to place the second boiler
module on-line and not by-pass steam to the condenser. Additional load
increases will be accomplished by increasing the first boiler module
and keeping the second boiler module at minimum until the first boiler
module is fully loaded. The second boiler module is then loaded as
required to 80% load. If an additional load is requested, the third
boiler module is started and placed on-line in the same manner as the
second. The procedure is then repeated for the fourth boiler module.
This procedure will give the most efficient (but not the
fastest) loading of the plant. In general, the best plant heat rate is
obtained by operating with the least number of modules required to
produce the desired load and to operate with all of the modules but one
at high load.
The start-up of boiler modules requires the use of the flash
tank, so the boiler modules must be started one at a time. The shutdown
does not require the flash tank, so more than one boiler module can be
taken off the line at the same time. The start-up or shutdown of the
gas turbine can be done simultaneously with the starting or stopping of
the boiler modules, if reduction of the total time is desired.
241
-------
O
o
c
'E
8
x
O
O
GO
01
.5
T3
3
15
CT
CD
1/1
t—
CD
ro
0>
"D 0)
C —
O 3
-
MW
242
-------
5.3.3 Predictive Loading
If the operator knows to what load he must have the plant at
a given time, a program can determine the most desirable configuration
of plant equipment and the best loading and starting sequence with which
to achieve it. The program can operate for either an increasing or a
decreasing load. The program will use but not be limited to the following
items to determine the best loading and starting sequence:
• The best heat rate for each boiler module will be at
high load.
• The reheaters can be started or stopped independently
of the rest of the boiler modules.
• The boiler module temperature changes should be held to
a minimum.
• The rate of change of the steam turbine is limited.
• The gas turbines can be started prior to being required
and held at minimum load.
• The gas turbines must be run after module shutdown to
regenerate the stone or lower the bed.
• The plant will contain only one flash tank and it can be
used with only one boiler module at a time.
• The flash tank will not be needed for module shutdown,
but the desuperheater will have limited capacity.
• Stone can be stored, made up, or operated once-through so
the regeneration system will not be operated at low loads.
• Any configuration can be operated over its entire range,
if necessary.
243
-------
The predictive program will be used only when called for by
the operator and when supplied with the proper information by the
operator; otherwise, the sequential loading program will be used.
5.3.A Rapid _Loading
Plant load can be increased or decreased faster by operating
the modules simultaneously or in parallel rather than sequentially.
This method of operation can be used for the 2:1 operating range without
by-passing steam to the condenser, or over a wider range by sending steam
to the condenser.
Figure 98 shows the total operating range of each of the four
configurations, including that portion accomplished by sending steam to
the condenser. This figure shows that the operating range with four
modules in service is from 265 MW to 620 MW; with three modules in
service, 165 MW to 465 MW; with two modules in service, from 90 MW to
320 MW; and with one module in service, from 30 MW to 145 MW.
Load changes within these ranges can be accomplished at rates
faster than 5% per minute. The plant will not be operating at its best
heat rate under these conditions, so changes at such rates should be made
only in an emergency and the load distribution adjusted slowly when the
operation stabilizes to the more economical configuration.
5.3.5 Loading Events Diagram
The attached events diagram for a cold start of the fluidized
bed boiler power plant shows the relation of pressures, temperatures,
and flows at several points of the steam and water cycle and the air and
gas cycle. The valves on the events diagram are not meant to be exact
design data but to show graphically the simultaneous relation of the
variables of the two plant cycles as the plant is started up.
244
-------
Curve 65^053-A
600
500
400
1 300
Maximum
200
Maximum
Maximum
Maximum
100
Minimum
By-pass
Condenser
Minimum
"*By-pass to Condenser
Minimum
By-pass to Condenser
Minimum
- By-pass to Condenser
0
One
Module
Two
Modules
Three
Modules
Four
Modules
Figure 98: Module Operating Range
245
-------
7S
fe 25
CD
0
£ 3000
h-
^ 2000
•o '55
c CL
E 1000
s
K 0
f 1000
Curve 65-W56-B
4 5 6
c
(O
500
Water Temo
o
o>
V.
13
13
k~
s.
100
50
0
100
50
0
1500
1000
500
To Gas Tu
Turbine
Temp
Inlet
Total
To
Preevap.
Preevap,
Bed Temp.
Water Wall
Wate
Flash Tank and
'Superheater
Sup
Flasl
2nd Superheate
1st Superheate
Cold R.H-r--0 R.H.
reevap< ^ Water Wall
Compressor Dis-
^charge Flow
-Byrpass
2nd S.H.
'Flow to Beds
Tank 3^
R.H.
456
Figure 99-Event Diagram - Cold Start
246
-------
The six events are as follows:
• The time the plant start-up begins
• The gas turbine has reached 3600 rpm and the first
boiler bed ignitor is started.
• The second superheater is placed in service by closing
the A valve, opening the N valve, and opening the V valve.
This causes the steam to flow through the second super-
heater and then to the condenser, rather than from the
flash tank to the condenser.
• The boiler is made once-through by opening the V valve
and closing the N and P valves.
• The pressure ramp is started.
• The pressure ramp is completed and the boiler is at
2400 psi and ready to increase load.
5.3.6 Dynamic Analysis
An analysis was made to determine the ability of the fluidized
bed boiler to change load. Two types of changes were made to the heat
input: step changes of 5%, 10% and 15%; and ramp changes of 5% per
minute, 10% per minute, and 15% per minute. These changes were imposed
at the equivalent of 70% plant load. Figure 100 shows the bed temperature
versus time for the two changes described above. Figure 101 shows the
power output versus time for the two fuel input changes.
These curves show the expected effect on temperature and load
caused by the stored energy in the beds. They also show that the unit
is sensitive to step changes and to overfiring. The control system
will utilize this information and overfeed fuel in predictable amounts
to increase bed temperature as required to increase load, and the
reverse to decrease load. If a nearly constant 5% per minute load increase
is desired, the fuel rate will have to be increased by approximately 15%
per minute for the first minute. This rate of change in fuel rate would
247
-------
1800
Curve 65W5't-B
1700
£1600
1500
1400
,sw change-"
^F_uelSLepChajae^%ofFullLoad_
60
120
180 240
Time, sec
300
360
Figure 100: Bed Temnerature Change as a Function of Fuel Feed Policy
Curve 65M152-A
100
90
70
X^ueljtep Chan^Wof Full J-oad
60
120
180
Time, sec
240
300
360
Figure 101: Load Response as a Function of Fuel Feed Policy
248
-------
be decreased to 10% per minute by the end of the second minute, and to
7-1/2% per minute by the end of the fourth minute. Manipulating the
fuel feed in this manner will permit the unit to change load at the
desired 5% per minute.
5.4 UNIT SHUTDOWN
5.4.1 Introduction
Large load decreases will be accomplished by removing boiler
modules and gas turbines from operation. Because of the complex con-
figuration of the plant, several operating procedures are required:
• Shutting down the first boiler module associated with a
gas turbine. The first or third modules to be shut down.
• Shutting down the second boiler module and one gas turbine
o Shutting down the fourth boiler module, one gas turbine,
the steam turbine, and the chemical recovery plant.
Figures 93 and 94 are process schematics showing the control
valves referred to in this section and their functional location in the
plant.
Figure 102 shows, in a simplified form, the steps required to
shut down the first boiler module. The shutdown procedure is essentially
parallel in the various parts of the plant.
The shutdown procedures that follow are based upon starting
with the plant fully loaded and slowly decreasing the total plant load
until all of the units have been shut down and the plant is off-line.
5.4.2 First Boiler Module Shutdown
The load will be reduced as described in section 5.3, on-line
loading, by reducing the fuel input and water flow to a level that would
permit one boiler module to be taken off the line.
Steam flow will be further reduced by positioning the airflow
dampers to the boiler module and thereby permitting a reduction of water
249
-------
IE!
- Ssi
=ii
Sil
p
s§
.2!
3
•§
o
CO
-gfp
^-SS
Sil
des
is
te'
C-J
o
I—I
CD
III
i
si—I
^S=s _ES
igg So
-Sh-
IIL
-------
o
tj
c
3
c.
o
o
o t,
e-z.
t!
B°
(U K
lisa
X ^
££
S5
£g 5
2;sc
sSS
i »
a l|siali^sgl|§
i
£L
MODULt
FLOW < TEM
t-ONTV,OL
S^
S&gri*
lilli
-------
'§
k
illlilsil:;
55555555
c
o
o
o :~
O.O
Sj
1 1
Us
ilL
E 5
ad
aj$
i? ri
i i1!^
-'"-.sg^e
zSSSjtaa
5«££3aS
ill.
»-
ill
35 95
VS«3|SS* = =:
43!iSib-SlSS»H
isiiiil:-:::::
!!ili I
illll I
!
.
ill :s:s
^a;;;
^^aS
IL
-------
ii1*4**,
irss
o
o.
eo
L-
o
la* s
is
-------
5.6 RUNBACKS, RUN-UPS, AND LIMITS
A complex plant requires a system of runbacks, run-ups, and
limits to maintain plant load at a safe value in case something unforeseen
occurs. This system will override load demand to adjust the operating
level as required by such a contingency. The system must contain three
parts; they are defined as follows:
• Runback — a reduction of load initiated by an alarm
condition. The load demand will be reduced until the
alarm clears. The runback can occur at any load level
and reduce the load demand to any value required to clear
the alarm condition. Minimum limits cannot be exceeded in
clearing the fault. The control items that will be included
in the runback scheme are low excess air, high bed temperatures,
insufficient feedwater flow with both pumps in operation,
approaching the surge condition on a gas turbine, and others.
• Run-up — an increase of load initiated by a low alarm.
The load demand will be increased as much as needed to
clear the alarm. The operating value at which this
condition can occur varies with the fault and cannot be
defined as an absolute value. Maximum limits cannot be
exceeded to clear faults. Runbacks or high limits will
always have precedence over run-ups because it is usually
safer to reduce the load than to increase it. If
necessary, a portion of the plant will be tripped to clear
the alarm condition. Examples of run-ups are low bed
temperature, minimum water flow to a boiler module, minimum
load on a gas-turbine generator, and others.
• Limits — high or low operating restrictions placed on the
plant based on the equipment in service. Limits are different
from runbacks and run-ups in that they have absolute known
values. If a high limit exists and the plant is operating
below that value, it does not affect operation as long as
263
-------
the load demand does not reach that value. The load demand
cannot be increased above the value of the limit. If the
value exceeds the limit value at any time, the load will
be reduced to the limit value as rapidly as possible.
Control system limits are imposed by critical equipment
such as pumps, fans, and others. These pieces of equipment
are essential for safe plant operation above some value,
as, for example, with two 60% capacity boiler feed pumps:
the plant cannot operate satisfactorily and safely above
approximately 55% of maximum capacity with only one pump
in service.
The runback, run-up, and limit system will be a part of the
master control and will be operational only when the plant is on
automatic control and fully coordinated. When operating in manual or
not coordinated, the alarm system will warn the operator of contingency
conditions, and, if necessary, the trip system will trip part or all of
the plant to prevent damage.
5.7 TOTAL PLANT CONTROL SYSTEM
5.7.1 Introduction
The total plant control system will be a hybrid, integrated
system made up of a combination of digital and analog control and informa-
tion equipment. The control functions that will be performed by this
system are shown on Figure 104. This drawing shows ten systems, each
designated by a block on the drawing. The inputs and outputs to each
system, as well as the control functions performed within that system,
are shown on the drawing.
A total analog control system, with a digital information system
of the type currently used on conventional power plants, could be applied
to the fluidized bed boiler control; but the advantages of the digital
system for large automated and adaptive control applications could not
be incorporated. The fluidized bed boiler will lend itself to control
264
-------
flow and fuel flow. Total airflow to the pair of modules will be
controlled by positioning the gas-turbine inlet guide vanes and by-
passing air from the compressor outlet to the gas-turbine inlet. When
the boiler module reaches its lowest controllable value, the module is
ready to be shut down. Steam flow to the turbine will be further reduced
by closing valve I and controlling the boiler superheater outlet pressure
with the U valve. This requires adjusting the pressure set point down
from the setting used for the relief valve function.
The coal flow to the second superheater will be stopped and
the I valve closed enough to maintain a 1000°F superheater outlet
temperature. This will reduce the heat in the bed and the amount
and temperature of steam sent to the condenser. As the I valve closes,
the pressure will be maintained by the U valve spilling steam to the
condenser and by closing the B valve. A minimum temperature will be
maintained in the preevaporator bed and the first-stage superheater bed.
The I valve will close when the superheater outlet temperature reaches
approximately 70°F below set point. This indicates that the maximum
heat has been recovered from the secondary superheater bed. All of the
steam from the boiler module will now be diverted to the condenser through
the U valve, thus minimizing the steam-turbine temperature shock.
When the I valve starts to close, the coal feed to the reheater
will be stopped and the reheater outlet steam temperature held at 1000°F
by closing the reheater valve. This will permit the coal to be burned
up in the reheater bed and the maximum amount of heat to be recovered.
The steam will be directed tc the other reheaters as the coal burns out.
The reheat steam temperature will be held constant and the thermal shock
to the steam turbine will be minimum.
The coal to the preevaporator bed and to the first superheater
bed will be turned off when the I valve has been 80% closed. The pressure
set point to the U valve will be lowered proportional to the water-wall
outlet temperature, the V valve will close, and the W valve pressure set
point will be adjusted downward as the water wall outlet temperature
decreases because of the bed burn-out.
253
-------
After all of the coal has burned out of the beds, the stone
will be recirculated until it has the correct chemical composition for
the next start. It is assumed that this module would be the first of
a pair to be restarted, so it will be necessary to reduce the levels of
the preevaporator and first superheater beds. The stone recirculation
equipment will then be shut down.
When the beds of the module have been restored to the conditions
desired for the next start of the module, the air will be diverted by
means of the gas-turbine by-pass damper and the boiler module air dampers.
Water flow to the module will be stopped when the conditions are correct
by closing the B valve.
The temperature control to the carbon burn-up cell will be
adjusted so the fuel feed is stopped as the carbon discharge from the
beds stops.
After all of the above have taken place, the boiler module is
shut down but ready for the next start-up, when the load requirements
demand it.
5.4.3 Second Boiler Module Shutdown
The description of the shutdown of the second boiler module
will cover the shutdown of the second module of the first pair to be
shut down. This shutdown will be followed by the shutdown of the first
gas turbine. The second gas turbine with at least one boiler module
and the steam turbine will still be in service.
The second boiler module will be shut down in the same manner
as the first module except that none of the bed levels will be reduced
and the gas turbine will be shut down. The bed levels will not be
reduced since it is assumed that this will be the second module of a pair
to be restarted.
254
-------
The control system will maintain a minimum load of 4 MW on each
gas-turbine generator. When one boiler module is not in service and the
other has a minimum bed temperature of 1300°F, the gas-turbine inlet
temperature will be approximately 1000°F. This temperature, with design
airflow, will produce 4 MW. If the gas-turbine inlet temperature or the
airflow decreases, the by-pass combustor must be ignited to maintain the
minimum load. As the sections of the second boiler module are shut down,
fuel to the by-pass combustor will be increased to maintain minimum load.
When the bed stone has been regenerated for the next start-up, the gas
turbine will be shut down and placed on turning gear ready for the next
time the load requirements call for it to be started.
The booster fan and other auxiliary equipment will be shut down
as the equipment is no longer needed.
5.4.4 Shutdown of Fourth Module
The shutdown of the fourth or last boiler module will require a
different procedure because the steam temperature and reheat temperature
cannot be maintained by the other boiler modules.
When the boiler module is at its lowest controllable level, the
gas turbine should be at a minimum load and the steam turbine at approxi-
mately 10% of total load. The boiler module cannot be operated lower than
this and still maintain bed temperature. The steam-turbine generator must
be separated from the line and the governor valves closed quickly to
shut down the boiler module. When this occurs, the fuel to all beds must
be shut off and the boiler pressure controlled by the W and U valves.
As in the case of the first boiler module start-up, the feedwater flow
will now be controlled with the feed pump and the pressure controlled
with the boiler valves.
The gas turbine may be kept in service to permit regeneration
of the bed in preparation for the next start-up.
The shutdown of the steam turbine will follow the normal shut-
down procedures for the steam plant. As the auxiliary equipment such as
condensate pumps, boiler feed pump, etc., are no longer required, they
will be shut down.
255
-------
If no electrical load is required, the gas turbine can be placed
on compressor discharge pressure control and the airflow controlled as
required by varying the gas-turbine speed, positioning the gas-turbine
compressor inlet guide vanes, and throttling the gas-turbine by-pass valve.
5.4.5 Emergency Shutdown
The plant will usually be shut down using the normal shutdown
procedures described in the preceding discussion. Emergency conditions
can occur, however, which will require emergency shutdown procedures.
The emergency shutdown of a boiler module can result from several different
types of failures. The plant trips are explained in Section 5.5. The
major categories are as follows:
• Loss of airflow to the module caused by loss of the gas
turbine or plugging of the distributor plate
• Failure in the water side of the module caused by failure
of a pump or burn-out of bed tubes
• Loss of coal or stone feeding equipment
• Loss of other auxiliary equipment such as stone regeneration
• Steam-turbine trip as a result of loss of regenerator load
• Failure of particulate removal equipment causing particle overload
• Loss of load on gas-turbine generator.
5.4.5.1 Loss of Airflow
The loss of a gas turbine will result in the complete loss of
airflow to two boiler modules. The plugging of the distributor will
cause the loss of reduction of airflow to only one bed. If airflow is
lost, the bed cannot be fluidized and the coal cannot be burned. The
coal and stone feed will be stopped when the airflow stops. The steam
temperature will drop because of the loss of heat input. When the
superheater outlet temperature reaches 30°F below set point, the I valve
will be closed. The U valve will be opened to maintain pressure in the
256
-------
module. The water flow will be reduced to minimum by throttling the B
valve. The water wall-pressure will be lowered as the temperature
decreases. The water wall flow will be stopped when the water outlet
temperature equals the feedwater temperature. This indicates that the
beds are cool.
The reheater valve will be closed when the reheater outlet is
30°F below set point.
The coal cannot be burned out, the stone regenerated, or the
preevaporator bed lowered because the beds are not fluidized. The beds
will be emptied when cold and prepared for the next start-up.
5.4.5.2 Failure in the Water Side
Water side failures can be caused by an external problem such
as failure of a boiler feed pump, feedwater heater, condenser, or other
auxiliaries; or by an internal failure such as a tube rupture.
Failure of the feedwater supply will result in the tripping of
the module to prevent burning out the tubes.
Tube rupture can be severe or minor, depending on the number of
tubes and the size of the failure. If the rupture is small, the module
will be shut down in the normal manner. However, a large failure can
result in an increase in bed pressure or carry-over of the bed caused
by jetting. If the emergency shutdown is required, the I valve and B
valve will be closed and the U valve opened to release water and steam
to the condenser. The coal feed will be stopped and the bed left in
operation to burn out the coal, regenerate the stone, and lower the
bed so repairs can be made. The reheater will be shut down in the normal
manner if the rupture is not in the reheater. If the problem is in the
reheater, however, the reheater inlet valve will be closed and the coal
stopped to the reheater bed only. The reheat steam flow will be diverted
to the other reheaters and the remainder of the module shut down in the
normal manner.
257
-------
5.4.5.3 Loss of Coal or Stone Feeding Equipment
If the coal feeding equipment to the reheater or to the final
superheater beds fails, the boiler module can be shut down in the normal
manner because these are the first coal feeds to be turned off. If the
failure is to the preevaporator or first superheater bed, all of the
coal feed must be stopped at the same time because the water wall or
first superheater temperature will drop and prevent normal boiler
operation.
If the stone feeding equipment malfunctions, the module will
be shut down in the normal manner, except that the beds will not be
regenerated.
5.4.5.4 Loss of Auxiliary Equipment
The loss of auxiliary or other associated equipment can result
in a normal shutdown, a load reduction, or an emergency shutdown. All of
these conditions cannot be determined until the plant design is completed.
5.4.5.5 Steam-Turbine Trip
A steam—turbine trip will call for a trip or emergency shutdown
of all of the boiler modules. This emergency will require the same
procedure as described above for a water-side problem.
5.4.5.6 Failure of Particulate Removal Equipment
The failure of the particulate removal equipment will require
removing the associated gas turbine from the line to prevent erosion of
the turbine blades. The boiler modules associated with this gas turbine
will also have to be shut down and, if this is the only gas turbine on
the line, the steam turbine will have to be shut down because no steam
will be available.
5.4.6 Dolomite Sorbent Regeneration System Shutdown
Since the sorbent regeneration system requires a supply of C0_
in order to operate, it will normally be shut down before the last boiler
module. Therefore, the stone can be clean in only three of the four
258
-------
boiler modules. However, it may be possible to obtain sufficient C0_
when the gas turbine is operating on supplemental fuel oil. In this
event, all of the boiler modules can be shut down with clean sorbent.
Stone recirculation continues until after the CaSO, reducer
vessel and the H_S generator vessel have been shut down. The CaSO,
reducer vessel will be shut down by stopping the feed of the coal, air,
and steam to the gas producer.
As the vessel pressure drops, the expander throttle valve from
the gas-turbine air compressor to the expander will be opened to maintain
C0~ flow to the H_S generator. The exit gas throttle valve will be closed
so all of the gas will be sent to the stack through the by-pass valve.
The C0? scrubber stripper flow will be stopped, the cooling
water to the condenser stopped, the spray flow stopped, and the bleed
valve opened as needed to prevent surge on the compressor. The expander
will be stopped by closing the throttle valve.
The stone feed will be stopped when the stone has been trans-
ferred from the hold vessels to the surge vessel so the stone will be in
the proper place for a start-up.
The last boiler module must be shut down after the stone feed
has stopped.
5.5 TRIPS
A complex power plant requires a plant trip scheme to interlock
the various plant components. Figure 103, trip chart, shows the effect
on e'ach of the plant components if one plant component is tripped. The
vertical columns represent the contingency condition and the horizontal
rows, the effect. For example, if the no. 1 gas turbine trips, the
no. 1 and no. 2 boiler modules must be tripped with no effect on the
no. 3 and no. 4 boiler modules or the no. 2 gas turbine, and so on.
259
-------
optimization more readily than the conventional boiler because of the
ability to control each module or portion of the boiler independently.
Analog control equipment will be used in certain critical
control loops and in the safety systems. Digital control equipment will
be used for plant automation, plant coordination, and adaptive control.
Analog and digital equipment will be combined in the conventional boiler
and turbine control loops. The total system will be arranged to utilize
the advantages of the two types of control techniques and will be arranged
to permit total plant automatic operation, system automatic operation,
and manual operation.
The following section briefly describes the function performed
by the ten control systems.
5.7.2 Plant Master Control System
All complex power plants need a master control system to
direct the control systems for the various plant components, to provide
central automation, to accept demand signals from the load dispatcher,
and to provide a common point for operator interface.
The master system controls total plant load from a remote
set point or an operator set value. It distributes the load to the
various plant components to provide the desired type of operation as
described above.
It will also provide the automation commands to start or stop
plant components as required to meet the desired load. Plant contingency
situations will be handled from the master with the necessary action
commands sent to the control systems and subsystems.
A system of panel interface with permissive and reject logic
will be provided to permit the operator to select safe operating modes
and configurations.
The master system will consist mostly of control logic systems
rather than feedback systems.
269
-------
The input signals to the master will be plant load and load
rate of change from the dispatcher or operator, limits from the plant
components or the operator, runback or run-up demands from the plant
components, mode selection requests from the operator, and operating
status information from the operator.
The output signals will be the various control set points to
the plant systems and subsystems, and automation commands to the plant
components.
5.7.3 Steam Turbine
A conventional single reheat steam turbine utilizing an
electrohydraulic control system and automatic turbine start-up will be
employed.
This control system is different from that of the standard
digital electrohydraulic system because the plant is operated in tne turbine-
following rather than the usual boiler-following manner. In the turbine-
following system, the plant load is changed by adjusting fuel input, and
the turbine inlet pressure is controlled by adjusting the turbine
governor valves. In the boiler-following system, the plant load is
changed by adjusting the governor valves, and the turbine inlet pressure
is controlled by adjusting the fuel input. The on-line control system
will be of the feed-forward type, with the master control requesting
valve area and this signal trimmed to maintain desired throttle pressure.
The safety systems will be the same as for a conventional turbine except
for the throttle pressure control which is included in the base control
system. Valve management, valve testing, overspeed protection control,
and speed error controls will be included.
The calculated rotor stress will provide a maximum limit for
the rate of change of load as permitted by the turbine metal temperatures
at that instant.
The turbine acceleration program will contain all of the features
of the standard programs and is not described further here.
270
-------
The pressure ramping of the first boiler module will be done with
floating governor valves position rather than fixed valve position. This
will accommodate the turbine-following control approach and make auto-
mation at the start-up easier, as the valves will always be in the proper
position at the end of the ramp.
The speed error can be limited in the increasing direction so
the plant can be operated in a base loaded manner without frequency
participation. The speed control will function only as overspeed pro-
tection during this mode of operation.
The analog and digital inputs and outputs to the steam-turbine
control system will be the same as for a standard 500 MW single reheat
machine except that the master control will provide a demand for valve
area (PI/PT) and the throttle pressure set point. Also, mode selection
logic can be initiated by the master control as well as by the operator.
5.7.4 Boiler Feed Pumps
The boiler feed pump control system will perform several
functions in addition to those of the electrohydraulic speed and the
safety control provided with the pumps:
• Feedwater flow control during the first boiler module start-up
• Maximum pressure override during the first boiler module
start-up
• Feedwater header pressure control after the first boiler
module is off the flash tank
• Feedwater pump flow balancing when two pumps are in operation
• Feedpump recirculation control
• Automatic start-up and shutdown of the feedpumps
The above listed control circuits are standard and will not
be described here.
271
-------
The master control will provide set points for flow and pressure
control plus automatic signals requesting start-up and shutdown of the
feedpumps. Mode selection logic can be initiated by the master control
as well as by the operator.
The input signals to the feedpump control, in addition to those
provided with the electrohydraulic control system, will be feedwater
header pressure, individual feedpump flow, individual feedpump temperature
rise and recirculation flow. The output signals will be individual
feedpump speed demand and individual feedpump recirculation valve
position demand. The logic signals provided will be those required by
the turbine pump manufacturers for automation.
The input signals for the start-up feedpumps will be the
individual pump flows and temperature rise. The outputs from the
start-up pump controls will be control valve position demand and recir-
culation valve position demand.
5.7.5 Gas Turbine
The gas turbine and gas-turbine control system will be different
from a W501 in the following areas:
• A separate combustion chamber will be used.
• The fuel system will be smaller in capacity because only
minimum load fuel will be required.
• Air can flow through the boiler modules, the by-pass
combustor, or both.
• Inlet guide vanes will be used to control airflow.
• The by-pass combustor will not be needed after the first
boiler module has reached approximately minimum load.
• Turbine inlet temperatures can be measured.
• Compressor airflow will be measured.
• Peak temperature limits will not be required because of
the low (1750°F) maximum bed temperature.
272
-------
• After start-up, only minimum load is controlled.
• The by-pass valve will be used to control airflow to the
boiler modules.
• The gas turbine can be controlled as an air compressor
providing air to the boiler modules and no electrical
power.
The following normal control functions will be performed, but
not necessarily in the same manner as in the W501:
• Start-up
• Overspeed
• Surge
• Overtemperature
• Minimum load
• Minimum and maximum fuel limits
• Automatic synchronization
• Safety protection for vibration, bearing temperatures,
generator temperatures, lubrication oil system, and others
as required
• Fuel transfer, oil-gas, if required
• Normal shutdown
The gas turbine will be started up and accelerated to 3600 rpm
by using the starting motor and by-pass combustor. Fuel is scheduled to
give the proper acceleration rate to maintain the turbine within the
temperature limits and the compressor away from the surge line. The
start-up will be smooth and with little danger of surge or overtemperature
because of the smaller combustors, which permit flame stability at a
lower fuel flow. This will cause less temperature rise at low airflow.
When the unit has reached 3600 rpm, the combustor will be automatically
adjusted to maintain a minimum load of approximately 4 MW. From this
273
-------
point on in the loading schedule, the gas-turbine fuel will provide only
enough heat to prevent the gas-turbine load from dropping below 4 MW.
As the heat input to the boiler module increases, the gas-turbine fuel
will decrease and then stop until it is again needed for unit shutdown,
at which time it will be restarted.
Surge protection will be provided by operating the fuel valves
(if in service), positioning the inlet guide vanes, and opening the
compressor bleed valves.
Control of overtemperature will be accomplished by reducing
fuel flow to the combustor if it occurs during start-up, and by running
back the fuel to the boiler modules if it occurs on load control.
Since total airflow control will be required by the boiler
modules and accomplished by positioning the compressor inlet guide vanes,
several of the control variables normally associated with gas turbines
will be eliminated.
The by-pass airflow valve will be an addition to normal gas-
turbine controls and will be used to control airflow to the boiler
modules during start-up by permitting air to by-pass the boiler module.
This by-pass air will flow through the combustor and mix with
the airflow from the boiler modules.
One other major difference will be that turbine inlet
temperatures can be measured and limited directly rather than estimated
by using blade path temperature, exhaust temperature, and combustor
shell pressure.
A variable pressure drop will exist between the compressor
discharge and the turbine inlet, and both values will be used in control
and protection of the gas turbine. This pressure drop varies because
of the boiler module operation.
The gas turbine can act as an air compressor rather than an
electrical generator for boiler module shutdown if no electrical power
is required. This will be accomplished by controlling the fuel to the
274
-------
combustor to adjust the turbine speed as needed to maintain the desired
pressure. The inlet guide vanes will be used in conjunction with
variable speed to provide the desired airflow and to keep the gas turbine
in the proper speed range.
Overspeed protection for the gas turbine will be provided
by interlocking the gas—turbine overspeed safety system with the boiler
module fuel valves, operating the compressor bleed valves, and positioning
the inlet guide vanes. In order to prevent excessive overspeeding of
the gas turbine in the event of the loss of its electrical load, pressure
relief valves are necessary in the gas-turbine piping loop. The gas
turbine is designed to withstand 10% overspeed. Therefore, the pressure
relief valves should be sized so as to give zero acceleration of the
turbine rotor at 10% overspeed with the relief valves open and a gas
temperature of 1600°F. The overspeed trip valve can be located in the
air line between the compressor and the fluidized bed boiler module
as shown in Figure 105, or it can be located in the hot products line
between the fluidized bed boiler and the gas turbine.
For cold side installation, the calculated valve sizes are
35 in.for one valve and 25 in. for two. Since the diameter of the
four air pipes between the gas-turbine manifold and the boiler module
is 28 in., the use of a 25-in,valve in each of two 28-in. diameter pipes
is indicated to be the optimum arrangement.
For hot side installation, the calculated valve sizes are
45 in.for one valve and 32 in.for two valves. Since the diameter of
the two hot gas pipes between the boiler module and the gas turbine is
54 in,, the use of a single pressure relief valve would be feasible.
From the standpoint of reliability of operation and ability
to restore service quickly, the use of pressure relief valves in the
cold side piping is definitely preferred.
The control system inputs required in addition to those
normally used for W501 control will be turbine inlet pressure and total
airflow, by-pass airflow, and combustor fuel flow. Blade path
275
-------
r
(O
^H'
CO
TO
cn
TO
Q
s
CD
1_
^
CD
276
-------
temperature and exhaust temperature measurements are not as critical as
usual and will be reduced in the number of measurements.
The outputs required in addition to the normal W501 will be
inlet guide vane position demand, bleed valve position demand, boiler
module fuel runback, on-line fuel stop and restart signals, and by-pass
valve position demand.
The master control will provide start and stop signals, total
airflow set point, compressor discharge pressure set point (when operating
as an air compressor), and by-pass airflow set point. Mode selection
logic can be initiated by the master control as well as by the operator.
5.7.6 Flash Tank
One flash tank will be provided for the four boiler modules
and used on only one boiler at a time. The A, C, D, and E valves are
associated with the flash tank and will be common to all of the modules.
The I, U, N, P, V, W, Y, and B valves will be associated with each boiler
module and four sets will be provided, one for each boiler. The boiler
by-pass valves are discussed in Section 5.7.8.
The flash tank will be used during clean-up, start-up, and
pressure ramping of the once-through boiler modules. When the boiler
has reached rated pressure and the P valve is closed, the flash tank is
available for use by the next boiler.
The flash tank will function in the normal manner for a sub-
critical once-through boiler. The level will be controlled by the D
and ,E valves, the pressure by the A and C valves, and heat recovery by
the C and E valves.
The set points for level, pressure, and feedwater temperature
will be received from the master controller, since the flash tank is
common to the boiler modules and must be coordinated with the plant
operation for start-up.
277
-------
The heat recovery of feedwater temperature control will be
interlocked with the actual feedwater temperature so water or steam from
the flash tank will not be admitted to the no. 6 or no.7 heaters until
the water or steam temperatures are higher than the corresponding heater
outlet temperature.
The inputs to the flash tank control system, other than the
automation and set points received from the master, will be flash tank
level, flash tank pressure, no. 6 feedwater heater outlet temperature,
and no.7 feedwater heater outlet temperature.
The output control signals will be valve position demand for
valves A, C, D, and E.
5.7.7 Boiler Module
The boilers will be subcritical once-through and controlled
in the conventional manner for a plant operating in the turbine-following
configuration. That is, a feed-forward system with pressure controlled
by the steam-turbine governor valves, load controlled by regulating water
and fuel flow, and steam temperature controlled by adjusting the water
flow rate to fuel flow rate ratio. Desuperheating sprays between the
primary and secondary superheaters will be used for superheat steam
temperature control during transient conditions. Reheat steam temperature
will be controlled by varying the fuel flow to the reheater bed and by
inlet sprays, if the bed is at minimum fuel input. These sprays can only
lower the reheat temperature.
Airflow will be held constant to the boiler modules and the
individual beds so fuel-air ratio will not be controlled. However, the
maximum fuel permitted to any bed will be limited by the airflow to
that bed. The bed fuel-air ratio will be determined by measuring the
oxygen in the exhaust from each bed.
Since the fuel will be controlled to the four beds of a boiler
module individually, the water to fuel ratio to each bed must be main-
tained. Also, the total water to fuel ratio for the module must be
maintained because the fluid flows through the four beds in series.
278
-------
Since the water and steam flow through the beds will be in
series, they cannot be adjusted individually, and the amount of steam
produced will be determined by the water flow. The water/fuel ratio
must be adjusted by varying the fuel flow. This correction will be done
slowly because the pressure will also be effected by changes in water
flow or fuel flow.
The desired water or steam outlet temperature of each module
will be a function of load and water inlet temperature. These values
are known and will be used by the control system to trim the water/fuel
ratio of each module by adjusting the fuel.
The upper and lower limits of the fuel flow to each bed will
be a function of the bed temperature, 1300°F minimum and 1750°F maximum.
The control system will also be adaptive, in that it will
adjust the outlet water or steam temperature, if necessary, to keep the
correction of each module near the minimum. Also, the heat absorption
will be adjusted between the modules if any of the beds approach the
1300°F minimum or 1750°F maximum. This will be done by adjusting the
individual water or steam outlet temperatures so that all of them
approach the high or low bed temperature limit at the same time.
The steam temperature control used for transients will be a
cascaded control system that maintains a differential temperature by
varying the spray flow. The spray flow will be held at approximately
3 to 5% of boiler capacity for normal operations so the temperature can
be controlled for either increasing or decreasing loads and/or
temperatures.
Water flow to the boiler in other than the first module will
be controlled by the B valve; on the first boiler module it will be
controlled by the feedpump.
Coal will be fed to the bed by positioning four control valves
in parallel.
279
-------
The stone will be recirculated as needed for regeneration, and
added or removed as required to maintain the desired bed depth.
The inputs to the control system for each boiler module will
be water flow, water-wall outlet temperature, first superheater outlet
temperature, second superheater outlet temperature, reheater outlet
temperature, desuperheater inlet and outlet temperature, 0~ content of
each bed exhaust gas, temperature of each bed, depth of each bed, coal
flow to each bed, igniter fuel flow to each bed, S0_ of each bed exhaust
gas, and particulates leaving the collectors.
The outputs from the control will be four fuel demands, four
igniter fuel demands, four bed air valve position demands, reheat valve
position demand, B valve position demand, superheater spray valve position
demand, and reheat spray valve position demand.
5.7.8 Boiler By-Pass
Each boiler module will have a set of valves for start-up.
These will be in addition to the flash tank valves described in Section
5.7.6 and will receive their set points from the master control. The
module start-up will be coordinated with the rest of the plant. The
valves will perform the following functions:
• I valve. Isolates the boiler from the turbine and also
controls steam flow while bringing the boiler onto and
off the line
• U valve. A back pressure controller. It will maintain
the steam pressure in the second superheater by spilling
steam to the condenser. It can also act as a relief valve
while the boiler is on-line.
• V valve. A stop valve between the first and second super-
heaters. It will be positioned to transfer from the flash
tank to on-line.
280
-------
• N valve. An open-closed valve used to admit steam from the
flash tank to the second superheater. The valve will be
held closed until the steam pressure in the flash tank is
sufficient to be admitted to the second superheater.
• P valve. Admits water or steam from the primary superheater
to the flash tank. It will be positioned with the V valve
to transfer from the flash tank to on-line.
• W valve. Used to control water-wall pressure during start-up.
• Y valve. A large valve in parallel with the W valve and
opened when the unit is on-line.
• B valve. Used to control water flow as described in
Section 5.7.7.
• H valve. The H valve will be an isolation valve.
The input signals to the control system will be feedwater flow,
water-wall pressure, flash tank pressure, second superheater outlet
temperature, second superheater output temperature, turbine header
pressure, turbine header temperature, water-wall outlet temperature, and
feedwater temperature.
The outlet signals will be valve position demands for the W, Y,
P, N, V, U and I valves.
5.7.9 Regeneration System
The regeneration system described is assumed to provide a
basis for an evaluation of a total system. Alternative operating
features and systems are being studied.
Typical areas which require further evaluation are:
• Reducing gas, producer-gas composition, reliability
and flexibility
• Temperature control
281
-------
• Interface with variable pressure in boiler
• Turndown-single unit versus separate unit for each
boiler or turbine module
• Stone circulation system operability.
For control purposes, the specific regeneration system being
considered will be divided into three subsystems:
• Stone feed system
• H9S generation system
• CaSO, reducing system.
Each of these subsystems will be controlled essentially
independently of the others, provided the other systems are in operation.
The interaction between the systems will be primarily the feed of stone
from one subsystem to the next and will be primarily affected by the
amount of sulfur contained in the stone. This cannot be measured but
can be estimated from empirical information by knowing the amount of
coal feed and the sulfur content of the coal. For this reason, the
feed-forward system will be used with corrections made to each sub-
system to properly balance the regeneration system with itself and with
the remainder of the plant. For steady-state operation, load demand will
be proportional to coal feed and used as the feed-forward signal. Boiler
modules can operate once-through during start-up and for short periods
on-line. The feed-forward signal will be corrected for the number of
boiler modules shut down or operating once-through and their portion of
the load.
5.7.10 Stone Feed System
The stone will be fed to the 20 beds by pneumatic conveyors
from a common surge vessel. Make-up stone will be fed to the surge
vessel at the rate of approximately 15% of the stone feed rate by means
of a double bell feeder. If the level drops below a predetermined value,
282
-------
the make-up rate will be increased. Overrides will be provided to
permit raising the level in the surge vessel for start-up. The make-up
feed will be stopped if the surge vessel should fill to a predetermined
high level.
Each boiler module will have a hold vessel which will be
controlled to a maximum level by removing spent stone. This control
will be of the on-off type operating between two levels. Overrides
will permit the hold tank level to be lowered to a minimum value for
start-up.
Stone will be fed to the beds at a fixed rate by utilizing
variable time control of the pneumatic feed system. SO- will be measured
in the exit of each boiler module. If the S02 content increases to a
preset value, the sampling point will be moved to each bed to determine
which bed is producing the SO-, and the stone feed will be increased to
that bed. When the system is again in balance, the S09 measurement
will be returned to the boiler module exit.
Stone will be removed from each bed by means of a weir which
keeps the level constant. Each bed will also contain a valve which
can be used to lower the bed depth for start-up.
The master signals to the stone feed system will be stone feed
automation commands, number of beds in service, and number of beds
operating once-through.
The measured inputs will be surge vessel level, four hold
vessel levels, four SO percentages leaving boiler or bed, and 20 bed
levels.
The outputs will be control signals to surge vessel make-up,
four spent stone removal valves, 20 stone feed to bed valves, 20 stone
from bed valves, and S09 sampling point selection.
283
-------
5.7.11 HgS Generator Control
The H.S generator control system will be a feed-forward
system that controls the C0? and H90 flows as a function of load demand
with corrections for percent H^S in the exit gas and humidity of the
CO to the H_S generator vessel. Pressure in the H?S generator vessel
will be controlled constant, and a surge control system will be provided
to protect the compressor during start-up.
The load demand signal will be corrected for the number of
modules shut down or operating once-through. This corrected signal
demands a position of the expander throttle valve after being trimmed
by a signal proportional to the deviation of compressor suction pressure
from the desired set point. There will be two expander throttle valves:
one to regulate air from the gas-turbine compressor and one to regulate
gas from the CaSO, reducer vessel. The air throttle valve will be used
only during start-up. The suction pressure deviation trim signal will
be used during start-up.
The CO- flow demand will be corrected to produce the desired
H~S in the discharge gas from the H S generator vessel. C09 production
will be greater than that required by the process and controlled by
positioning the stack gas control valve and the stripper control valve.
Actual CO flow to the H?S generator vessel will be controlled by
positioning the CO bleed valve. This configuration will permit an
accurate control of CO^ flow and permit the system to handle transient
conditions. The percent H S trim is not used during start-up.
The H20 to C02 ratio will be controlled by maintaining a
desired condenser discharge temperature. The condenser will produce
saturated gas. The remainder of the H-0 required to produce the desired
ratio will be provided by adding steam to the CO. at the compressor
discharge. This permits accurate control during load transients and
steady state operation. It also keeps the power required to operate
the compressor to a minimum.
284
-------
The pressure in the H S generator vessel will be maintained at
the desired value by modulating the discharge valve.
Three bleed valves will be required to protect the compressor
from surge during all operating conditions. Bleed valves no. 1 and no. 3
will be used during start-up, but cannot be used on-line because air must
be kept out of the system. Bleed valve no. 2 will provide the recircula-
tion of CO- to maintain minimum flow without admitting any air to the
system. Minimum flow for safe operation will be determined from
compressor pressure ratio and speed.
The master signals to the H~S generator system will be auto-
mation commands, load demand, beds in service, and beds operating
once-through.
The measured inputs will be exit gas percentage C0_, HO and
H?S, H_S generator vessel pressure, C0_ flow, compressor pressure ratio,
compressor speed, and compressor inlet pressure.
The outputs will be control signals to the expander air throttle
valve; expander gas throttle valve; stripper flow valve; condenser
cooling water valve; steam valve; H.S generator vessel discharge valve;
compressor bleed valves no. 1, no. 2, and no. 3; stack gas control valve;
C0_ bleed valve; CO- isolation valve; vent valve; and two stone isolation
valves.
In order to maintain fluidization in the vessels during reduced
load, multiple regeneration modules may be required rather than the
single module concept discussed here.
5.7.12 CaSQ, Reducer Vessel
The CaSO, reducer control system will be a feed-forward
system that controls the gas from the gas producer and steam to the
CaSO, reducer vessel as a function of load demand, with corrections for
percentage combustibles in the exit gas and exit gas temperature.
Pressure in the CaSO. reducer vessel will be controlled constant.
4
285
-------
The corrected load demand signal discussed in the H S
generator system will be used in this system. It positions the air and
coal feed after being corrected to maintain the desired percent com-
bustibles in the exit gas. The load demand signal to the steam valve
will be corrected for exit gas temperature.
The pressure in the CaSO, reducer vessel will be maintained
constant by positioning the expander by-pass valve.
The master signals to the CaSO, reducer system will be auto-
mation commands, load demand, beds in service, and beds operating once-
through.
The measured inputs will be exit gas percentage combustibles,
exit gas temperature, reducer vessel pressure, gas producer steam flow,
gas producer airflow, and gas producer coal feed.
The outputs will be control signals to gas producer steam
flow, gas producer airflow, gas producer coal feed, and expander by-pass
valve.
5.8 CONCLUSIONS AND COMMENTS
5.8.1 Conclusions
• The plant should be capable of control for base load or
swing operation.
• By changing the number of boiler modules and gas turbines
in operation, the plant can operate over a load range
or from 12-1/2% to 100% plant capacity. This is comparable
with conventional power plants.
• Load changes of 5% per minute over the load range of the
equipment in service can be accomplished, that is from
50% load to 100% load if all four modules are operating.
This rate of change is comparable with other types of
large power plants.
286
-------
5.8.2 Comments
New procedures for start-up of the once-through boilers
can be developed instead of using the conventional
systems. This is possible because of the multiboiler
configuration and the ability to by-pass steam or water
to the condenser. Also, since the minimum water flow
per bed is approximately 40%, the start-up valves can be
simplified.
The control system must be of the feed-forward type with
corrections for pressures, temperatures, and flows.
Since the fuel to each portion of the boiler can be
controlled independently (that is preevaporator, primary
superheater, the secondary superheater), the temperature
and pressure controls should be capable of preventing
excursions normally associated with load change transients,
At least two different loading procedures are required.
One for known final conditions and one for ramp-type
increases without known endpoints. This is necessary
to prevent transients on individual modules.
More automation will be required than for the conventional
plant because of starting and stopping modules to change
load.
The chemical recovery control and operation must be
coordinated with the total plant operation. Once-through
operation of the stone permits the plant to operate at
low loads without the chemical recovery plant.
Since the fuel can be controlled to each portion of the
boiler independently, the temperatures within the boiler
can be controlled. This feature will permit better steam
temperature control.
287
-------
The short resident time of the water in the boiler
minimizes the pressure and temperature transients
normally associated with once-through boilers.
Airflow measurement to each bed will be difficult. The
damper position of the various beds can provide an
approximate measurement. Gas analysis can be used to
determine excess air so airflow measurement may not
be critical.
The sampling systems for the 0,, and SCL measuring
equipment will require high maintenance because of the
dirty gas at the point of measurement.
A gas turbine hot gas by-pass will be required to provide
for a loss of load on the gas-turbine generator. This would
permit the boilers to operate momentarily with the gas-
turbine generator out of service.
The condenser and desuperheater design and control should
be investigated to assure that steam can be by-passed to
the condenser for boiler module start-up and shutdown.
The stone flow control and bed depth measurements will
be a problem. The stone feed problem is also effected by
the variable pressure operation of the beds caused by the
changes to gas-turbine inlet pressure as a function of
load.
A combustor may be required to start the expander used
by the chemical recovery plant.
Superheater and reheater spray valves will be required
to increase the operating range at constant steam temperature
and to provide better temperature control during transients.
Surge control will be required for gas-turbine compressors.
This system will have to be more complex than the open-loop
system normally used on gas turbines. It should utilize
pressure ratio, inlet volume,and speed for control.
288
-------
6. ALTERNATIVE PRESSURIZED FLUID BED COMBUSTION CONCEPTS
Fluidized bed combustion technology can be applied with many
different configurations and can be utilized in many different steam and
power cycles as previously indicated. Two alternative pressurized fluid
bed combustion concepts were studied — an adiabatic combustor combined
cycle plant which represents a modification of the base power cycle and
a recirculating bed boiler which represents an alternative fluidized
bed boiler concept. Conceptual designs, performance, and economics
are projected and compared with the pressurized fluid bed boiler combined
cycle plant.
6.1 FLUIDIZED BED ADIABATIC COMBUSTOR COMBINED CYCLE POWER PLANT
6.1.1 Adiabatic Combustor Concept
A pressurized fluidized bed boiler combined cycle power plant
has been designed using state-of-the-art power generation equipment.
Performance, costs, and pollution abatement were projected for the system.
The results show the concept has the potential to meet S00, NO , and
^ X
particulate emission standards and may reduce energy costs 10% below a
conventional plant with stack gas cleaning. This is achieved by
effectively combining the combustion, heat transfer, and pollution
control processes. The fluidized bed boiler operates at excess air
values from 10 to 100% with up to 70% of the heat released in burning
the fuel with air transferred to the water/steam in the tubes surrounding
and submerged in the bed.
Increasing the design point excess air with constant bed
temperature will decrease the total heat transfer surface in the fluid
bed until no boiler tube surface will be required at an excess air of
approximately 300%. In this case, the power system is a combined cycle
plant with the gas to the turbine expanders supplied from a coal-fired,
2
adiabatic combustor.
289
-------
Combined cycle plants of this type, which burn natural gas
and/or heavy distillates,are now being marketed to electric utilities
for intermediate and base load applications. One embodiment of such
a plant is the Westinghouse Power at Combined Efficiencies (PACE) plant.
In the PACE plant, the exhaust from the gas turbine is reheated to a
temperature of about 1200°F ahead of the heat recovery boiler which has
steam conditions of 1250 psia/950°F.
A modified PACE configuration, shown in Figure 106, was
selected for evaluation of the adiabatic combustor system. In this
configuration, the heat recovery boiler is unfired and desulfurization
is achieved by feeding limestone or dolomite to the adiabatic combustor
with once-through sorbent utilization.
6.1.2 Cycle Performance
Process conditions for four alternative configurations are
summarized in Table 44. Case IV presents the generating and performance
data for the unfired heat recovery boiler case selected for evaluation.
As in the high-pressure fluidized bed boiler, the desulfuri-
zation process limits the maximum temperature in the fluidized bed to
about 1750°F, which means that the maximum gas-turbine design temperature
is about 1700°F. With a 1700°F gas-turbine inlet temperature, the
temperature of the exhaust gas to the heat recovery boiler is about
850°F. This precludes the use of the standard PACE plant steam conditions.
The steam conditions selected for the adiabatic combustor system are
approximately 700 psia/800°F. Also, since the gas temperature to the
boiler will decrease with the gas-turbine load, it will be necessary
to use a sliding pressure boiler in which the steam conditions will
follow the exhaust gas temperature at part load. The estimated heat
rate at the design point for this modified PACE plant is 9096 Btu
(HHV)/kWh.
The standard PACE plant consists of two W501 gas turbines
with a modularized heat recovery boiler for each gas turbine. With
290
-------
cn
S
a.
UJ
TO
a>
o
O
CD
r—4
0>
CD
291
-------
TABLE 44
PROCESS CONDITIONS AND PERFORMANCE
Case
:
t^ - °F 59
W . - Ib/sec 1535
al
p.. - psia 14
p« - psia 146
t2 - °F 614
W _ - Ib/sec 1450
az
t0 - °F 450
2a
P2 - psia 146
W0 - Ib/sec 81
za
Po - psia 135
t3 - °F 1600
W 3 - Ib/sec 1485
p- - psia 15
t5 - °F 798
W 5 - Ib/sec 1567
t, - °F 1200
o
p, - psia 15
W , - Ib/sec 1583
p6
t? - °F 359
P - MW 136
s
PGT 126
PQ 263
PN 259
HR(HHV) 9281
Wfl - Ib/sec 35
Wf2 - Ib/sec 16
[ II
59
1535
.65 14.65
.5 146.5
614
.0 1429.5
450
.5 146.5
.9 102.3
.5 135.5
1700
1468.5
.2 15.2
860
1570.6
1200
.06 15.06
1584.3
359
.5 136.8
.9 141.2
.4 277.9
.0 273.6
9003
39
13.7
III
59
1535
14.65
146.5
614
1450.0
450
146.5
81.9
135.5
1600
1485
15.06
795
1567
795
15.06
1567
411
49.6
128.1
177.6
175.4
9440
35
0
IV
59
1535
14.65
146.5
614
1429.5
450
146.5
102.3
135.5
1700
1468.5
15.06
857
1570.6
857
15.06
1570.6
408
60.5
142.4
202.9
200.4
9095
39
0
292
-------
the gas turbine and steam conditions described above, this modified
PACE plant would have net power output of 200.4 MW. In order to compare
this concept with the high-pressure fluidized bed boiler with a capacity
of 635 MW, a plant consisting of three coal-fired PACE systems was
assumed with a total capacity of 601.2 MW.
A comparison of heat rates on the same basis for the various
coal burning power plants is as follows:
Type of Plant Heat Rate - Btu (HHV)/kWh
Conventional Steam 9186
High-Pressure Fluidized 8990
Bed Boiler
Coal-Fired Modified 9096
PACE Plant
6.1.3 Plant Turndown
The turndown capabilities of the coal-fired modified PACE
plant should be very wide. Figure 107 shows that a reduction of about
40% is possible by variation of bed temperature from 1750 to 1400°F
(corresponds to turbine inlet temperatures of about 1700 and 1350°F).
Modulation of compressor air provides another 20% turndown. Beyond
this point, compressor air can by-pass the combustor (see Figure 106)
operating at constant bed temperature to reduce the gas-turbine inlet
temperature to the idle level of about 950°F. It would probably not
be practical to operate the boiler all the way to this point. There-
fore, the boiler could be shut down and the plant power reduced to
near zero. A by-pass around the boiler might be necessary. In order
to carry out the by-pass technique described above, the design
superficial velocity would have to be high enough to permit about 30%
reduction in flow at a bed temperature of 1400°F.
A general discussion of plant turndown for pressurized fluid
bed combustion boiler power plants is presented in Section 5.3 of
this report.
293
-------
to
L_
o
I
8
q>
'o
QO
cr
•o
o>
e 'I
I =
I 2
8
oo
CM
O
•E
I
C
IX.
o
1—4
o>
13
01
o
u6|saa 10
294
-------
6.1.4 Environmental Considerations
The environmental impact of an adiabatic fluidized bed
combustor power plant is projected in Table 45 . Sulfur dioxide emission
standards can be met based on the available pressurized fluidized bed
boiler data and the kinetic data at the adiabatic combustor operating
conditions (see Section 3.10). Data are required to determine if the
nitrogen oxide emission standard can be met. Nitrogen oxide limits can
be achieved in a pressurized fluid bed boiler. However, the high excess
air for an adiabatic combustor may affect the NO formation. For example,
if NO is reduced by reaction with CO, the high excess air may reduce the
available CO and increase the NO emission. Particulate emission standards
can be achieved in an adiabatic combustor system based on fluid bed
combustion emission and particulate removal equipment data. Data on
particulate control for an adiabatic combustor system are required to
confirm this projection. The gas flow for an adiabatic combustor system
is ^ 40 Ib gas/lb fuel compared with ^ 12 Ib gas/lb fuel for a pressurized
fluid bed boiler system. This requires a more extensive particulate
removal system for the adiabatic combustor system than for a pressurized
fluid bed boiler system.
Heat rejection with an adiabatic combustor system will be
reduced ^ 35% below a conventional coal-fired plant. This results
from the increased fraction of gas-turbine power (see Table 44). Solids
from the system are projected from the pressurized boiler and TG data
(see Section 3). Low-grade fuels and solid wastes can be effectively
utilized in a fluidized bed adiabatic combustor system.
6.1.5 Fuel Processing Equipment
A preliminary evaluation was made of the adiabatic combustor
fuel processing system. The design basis is summarized in Table 46.
Two design concepts — a single fluid bed module and a stacked fluid
bed module — were considered as illustrated in Figure 108. A summary
of the respective design features is presented in Table 47. The partic-
295
-------
TABLE 45
ENVIRONMENTAL IMPACT
Item
Impact
Air Emissions
S02, lb/106 Btu
NO , Ib N09/106 Btu
x / f
Particulate, lb/10 Btu
Heat Rejection to Cooling Water
Solids
Resources
< 1.2
to be determined
< 0.1
^ 35% less than conven-
tional coal plant
dry, granular CaSO,
multifuel capability,
e.g., low-grade coals,
oils, solid wastes
296
-------
i £
Air Plenum
Air
Air
Air
Flue Gas
Flue Gas
Gas
Air
Single Bed Design Stacked Bed Design
Figure 108: Adiabatic Combustor Designs
297
-------
TABLE 46
DESIGN BASIS
• 30 mole % Ca utilization of the dolomite
• 90% removal of SO- from burning Pittsburgh #8 coal of
4.3% S
Case IV 1070 ppm -*• 107 ppm
• 3 hr average residence time of dolomite
• Refractory insulation
1700°F normal 2000°F design
AP across wall is minimum (say 0.5 psi)
{3-1/2" insulating "A.P. Green VSL50" plus
2-1/2" "A.P. Green Lo Abrade"
• Dolomite feed rate 179,000 Ib/hr (as CaCO- MgCO-)
Coal feed rate 140,000 Ib/hr
298
-------
TABLE 47
ADIABATIC COMBUSTOR DESIGNS
Single Bed Design
Design I
Design II
Stacked Bed Design
Design I
Design II
Number of Modules
Number of Beds per Module
Module Diameter, ft
Module Height, ft
Bed Depth, ft
2
Bed Area, ft
Fluidizing Velocity, ft/sec
Equipment Cost, $/kw
4
1
21
16
6.5
347
6
2.5
2
1
30
16
6.5
695
6
2.1
4
3
12
50
6.6
113
6.2
2.0
2
6
12
100
6.6
113
6.2
2.1
299
-------
ulate removal system design is specified in Table 48. A preliminary
fuel processing equipment layout is presented in Figure 109 for a single
bed module concept.
6.1.6 Power Plant Cost
The fluidized bed boiler power plant cost breakdown presented
by Westinghouse was reviewed and compared with an adiabatic combustor
power plant. Table 49 presents a comparison of the power plant equipment
for the adiabatic combustor plant with the pressurized boiler. Equip-
ment costs for a 600 MW adiabatic combustor system have been projected
for two cases:
• Pressurized boiler basis. Assumes six gas turbines,
one adiabatic combustor per turbine, and one, 180 MW,
steam turbine. Costs are based on the pressurized
fluid bed boiler plant costs.
• Modified PACE plant. Assumes three PACE plants but
also assumes same style plant as in the pressurized
boiler plant, e.g., buildings, stack, etc. Does not
account for potential cost increases in undistributed
costs due to increased engineering and field supervision
as the result of the coal processing system.
An equipment cost summary is presented in Table 50. Equipment costs
for the fluidized bed combustion boiler and a conventional plant with
stack gas cleaning are also presented. Component costs are presented
in Tables 51 and 52 for the adiabatic combustor system. Total power
plant capital costs for each system are presented in Table 53.
6.1.7 Conclusions
• Equipment costs for an adiabatic combustor power plant
will be less than 4% below those of a conventional plant
with stack gas cleaning using the same plant design basis.
• Total capital cost will be 5 to 15% less than that of a
conventional plant with stack gas cleaning using the same
plant design basis and construction time from 2-1/2 to 3 years.
300
-------
TABLE 48
PARTICULATE REMOVAL SYSTEM (200 MW)'
1st Stage
2nd Stage
Actual Gas Flow, ACFM/module
Collector Selection
Number of Collectors per Module
Pressure Drop, in. W.G.
Collection Efficiency,
% by weight
124,500
size 635 Model 810
Duelone
5
20.9
85.1
124,500
model 18,000
Tornado Cyclone
4
31
97
This table is for the four-module design. For the two-module design,
double the number of cyclones per module.
301
-------
^-GA£> TURE.IN1
O" 6' It*
I I I I [
SCAUt
Figure 109: 200 MW Fluidized Bed Adiabatic Combustor Combined
Cycle Plant - Gas Piping from Participate Collectors
302
-------
TABLE 49
POWER PLANT EQUIPMENT FOR FLUID BED
ADIABATIC COMBUSTOR 600 MW
Structures and Improvements
Turbine Plant Equipment
Electric Plant Equipment
Station Equipment Transmission
• structural steel for adiabatic
combustors could be reduced -
ship with structural steel attached
• lighting and painting in boiler
module area reduced
• turbine room and heater bay: turbine
room may be more expensive if 3 steam
turbines used - optimum arrangement
may be two modules with two steam
turbines to have best steam piping
arrangement; heater bay essentially
eliminated, although there are 3 deaerators
• intake and discharge structures cost
will be lower since water requirements
are ^ 65% less
• circulating water system will be
proportionately higher since there
are 3 subsystems, one with each
PACE module
• condensing system requires less
area and concept does not have
feedwater heaters which reduces cost
and condensate piping system cost
• steam turbine: could use one unit or
one unit per module; with only one
turbine must interface with 6 heat
recovery boilers — tradeoffs not clear
• feedwater pump, extraction steam piping
and steam piping costs are reduced. No
hot and cold reheat, piping + lower steam
conditions
• water treatment system less complex due
to lower steam conditions
• switchgear will be higher since there are
more generators
• generator voltage
303
-------
o
m
fa
rJ
W
H
£5
o
CO
I-H
^3
p ,
^
O M
4-1
H CO
co 3
O Xl
U B
O
H U
W O
g "ri
CM 4-i
H td
O" cd
is -H
H <
L-l 01
PM PP
Pi 13
W iH
3 3
O rH
PM PM
8
oi
M
Pn
1
<3J
o
u
* /**N J3
4-1 CO • 00
13 B 00 0 3 01
CU Cd T) 4J O 4J 3
•H H H 0! M -H g
MH PL, xi xl B
iH - 4J O O
13 K • ^A 1 rH O
O U O 0 U O vo
S 4-1 o in
CO -r-l H 1 H CO
CO 3 -H 0) O vO
01 H O 0 Q
ri fa cq G
PJ O
00
cT
•H
rH G
cd xl td
B 4-1 01
O -H H 3
B cd CD in
01 4-1 Cd CO
> G CJ vO
B cd
O H .id
CJ Pi O
td
4-1
CO
rrj
in co
OS vo
• •
O OS
in o
OS CO
co r*-
rH
in CN
os in
o oo
CO 00
H CM
H r^
CM
CO
4-1
B
01
3
>
en o
4J M
xl ex
oo E
•H M
PH
-o "a
B cd
cd
r5 CO
0)
B 3
Cd 4-1
o
13 3
cd *j
rJ CO
O
^^
.
vo
I-.
o o
1 — • rH
vO vO
F^» -sf
00 O
vo 00
OS H
<4H
0
vo
CO
CM
H
vo m
m co
CO 00
CM
vo m
m co
CO 00
CM
H H
CM O
- O
O H
01
CO >,
3 •• (-1
OB td
XI CU 13
00 4J B
B co >, 0
tH !>,H 0
4-1 en H oi
en cd en
d) rl B
3 0) -H M
H 00 0)
; -H -H oo
" O rl rl
co xi o cd
CO H
11 13 3
O 0) S O
O XI &
M CM 4-1
PM T3 CN
•H XI
G 3 tH 4-1
OH O *H
•H MH S
4-1 M
CO 01 O 13
3 X! 4J 01
XI 4-1 cd O
B rl Cd
0 B 01 H
cj tn G ex
01 O)
13 01 00 rl
OI 13
PQ td >, en
BOM
13 GO
N M 60 O
•H 01 M 01
13 3 01 H
•H f B H
3 en 01 o
H G O
Pn 0 "
•H B $*•
MH 4J 01 M
O CJ 4-1 td
3 co 13
a 13 >•, G
O 0) CO O
•rl M O
4-1 Xl 0!
td 4-i oo en
3 co 3
H o O M
ClJ O W _3
> XI 0
Pt3 00 4-1 MH
— • B 1
•H 01 "
B & o n
•H o e o
H O 4-1
•OH td
01 O M M
4J MH O 0)
fi MH B
0) 0) 01
CD XI 13 00
01 H 0!
M 03
Ok S ^
cd H 01 O
4-1 f>- M O
cd os in
13 rH 4-1 CM
en
4J M 0 XI
co 01 o 4-1
O XI -H
o s co &
S 4J
B > X! 13
O O 00 01
Z •* U
•a M cd
01 . H
co , H
M 3
3 en
01
G M
td
i-> C ^
o >.
CO -H M
•H 4J Cd
O CO
oi 3 co
CO 13 01
cd oi o
X> MO!
B
4-1 4J
co en 4-i
o o o
0 U B
0 "*3
•
4-1
CO
r^l
CO
00
B
•H
co
co
01
o
O
M
Cu
H
cd
o
0
XI
4J
•H
3
CQ
4J
CO
O
O
CU
4J
3
M
4-1
en
•H
-a
B
3
13
01
CO
cd
01
o
B
•H
0
CU
3
13
13
01
4-1
td
o
•H
13
B
•H
B
cd
XI
4-1
rl
01
00
•H
XI
00 vll
01 H
H H
XI -H
td S
H
4-1
01 co
01 O
CO CJ
CU MH
304
-------
*
rH
W
h-3
CQ
"^
H
2
§
g
9
3
w
pq
H
OT
O
1
§
M
|~i /~\
O1 CO
W 4J
en
H o
J2j O
h-3 4J
CM c
ct)
W P-i
O ^
P-i 0)
W tH
(J O
u pq
u 13
41
n pq
H
J2J *"O
H 0)
pq N
S "H
o *o
U -H
3
P4 rH
O fe
H
CO (3
to o
pQ
J2J 'O
O 0)
o to
0)
0 A
M — '
H
^
PQ
^
M
9
Q
W
PQ
Q
Nl
h— 4
o
M
i~3
|ij
CO g
W«^,^
^^^
O
c_>
T3
0)
rH
to
4J
M
M
a
0
*rH
4J
td
rH
CO
4J
CO
c
I-H
ij
4-1
C
o) ca
0 •IJ
•H O
3 0
a1
W
ooO(T>oo coi voo oo vo oo m-*mcscoo i>-
HrHiOrH-*iHi-HrH r~»l COr^ 00 O CO OulpHrHrOrHvD
rH CO rH I-»
ooooooooo oo o o o ooooooo
ooooooooo oo o o o ooooooo
ooooooooo oo o o o ooooooo
oooiooiooLom oo o o o ooooooo
-d-OOOr^.CM-3-rHr-IOO VOCM 00 VO 00 HmOlTlOOrH
rHvOrHl^mr-OOvOm rHCM CM CO CM COCM(y>f~»OvOrH
rHcOCM OCMCMIO CMVOCO CMvO
rH CM
rH rH CO CM
OOOOOOOO OO O O
OOOOOOOO OO O 0
OOOOOOOO OO O O
cM-*oooma\o oo o o
COrHOOOOI~~COOOO> VOO 00 Sf
-*CMCOCOOcOCMrH CMOO 00 CM
«\ M M f\ *
rH rH rH 00 CM
rH
m
a
00
•H -a
co (3
CU tfl
a i
CO 0) 4J O
0) rH CO -U
M 3 rH
3 'O CO -H
w o - a) a
u B e G o
3 i a) a) -H -i-i IH
CO !-l 0) O 4-1 M) 43 O, 4J-H
4-1 OO 4-1 CO |3 CO 00 td OO i-) 3 Cfl 4J
G G co 0 -U ^i G 4-J o) 3 1 oo *H 4-J G
a) -HTJ oicoocoG-'-'a) coboooca)
0 *r^ 0) M 4-* « »H -H ,1^ 4-J G G 0) 0
Q) rHOOCU COMrH4JT3J3CO n) GOiHgfli
O CO 603n)-rl t04-)>(UCM360 Tt TJ4JTJM3
t-i a) ooC^a^co rOcoorH +j TSGO coG-wo'
ft rH G "H o rO 330 rH 0 0 G G *cJ 'O (d co fl)
0 3 -HTSGco co^aSootoo cfl-H GGJ3G
M 1343 T3rHO'H>, 8 >-i U M cd 1-1^3 rH 60 tfl 3 -H CO
O 3 r-l iH i-( T3 M OOO IHOOM-ICOOOTJC O4J 3
T3 00-H34J CU GOCUO) MGGiHCDMHCO'OO
G O34itO'd> -H 4J6000OW)OtHtdrHQ) 3CO)
VlO C04Jn)OrH|CO'Hr-H4J-H -HCOQ) CrHG (OrH
CO O 4J 0) rH CUOD CU CO|Q)4-'3CO ft 0) ft 3 CU 4J CO MH CO *X5 rH rH CO
QJ^COCOO-HCUM •W|OtOO|-rlC'HrOUHG^! COrH -U
I-i 3-H>-i'rlCr*! OIOra-Hi3ftOft0 Q> 4JJsJt05-iTHc04J 4J|>-IC04JG rH OrHJii-HUHa 4JU -U
-------
TABLE 51 (Continued)
* i!
CD --^
o rH-<|-COCNCOrHr- r-H CO O O CN -* CO jj
-tf rH CO rH ^
o
s
O OOOOOOOO OOOOOOO o
O OOOOOOOO OOOOOOO K
O OOOOOOOO OOOOOOO &
O OOOmOLOOO Or-HOOOOrH M,
VO Or^rHCOCNLn ^
CM CN rH rH rH rH w
cu
rH
0
a
OOO OOOOOOO OOOOOO n
OOO OOOOOOO OOOOOO M
OOO OOOOOOO OOOOOO 'J
OOO OOOu^tOOO iOvDCT»O *^ *^ 00 n) r^
•O-rHlX'cOVDCMrHrHrH rH 2 °
CN CM CU !-.
!> 3
>-l 4-t
0
4-1 CU
O • CO
O cd co • 3
0 g C 2 «4H
°-o H g 41 'H
oo a g § 3 £
o ° S^ "^tiS^
•>O COC 4J3LO
~*O i^CU G & >, -
?* coS M cucuS^
•co- a ftcr 4-1 co S CL m ?
csj CUB-HCO c MOO § S
C2J 4JCU3rH 0) CUC MCU OCU
vo ^nj-ucro a C-H -HS OM
^^VD COCOCOCUM ft -HJ-I 33 OtS
^ BrtCtx, 4J -H4Jcd-H rfJ 2m
CO OJCOCUCOCUC 3C34-1^ CUQ £
aWCO 4JrHBT3 CO 0-dJC u£ 4JO
CU-HT3 CD-HMCr)0'HO B CU BOrH .rtfi
4Jflt-H l^iOOOCUB^ CU ftOO rHt^4JCU'H
C03O co^)4-i04JCU^'rj 4-1 CU -H M mcrt3^
>,i4H ,0 cd cd4J3C co a 3T3-U S^ft-o
COrH'H 3^MT3IiC04JCd ^-» »H CT^C^C^ 3_rJMi 1
,aoa concuc >.i co >cocucdo WUMOS
3mcd cuCcdbocoBco ja MT) a Tjaj-iujn.
co&B 0)>cu a cd*J SMCUMCUCO 3 4J o u r-
CUTJrHrH -HO 0)4JrH 0) O-rlVlpirH 4J(U^C4JOG^34J3cd cd coxiw o ca cu
•H-iHfS-U H^CCrHcd 34-i cdJ3O43O4J 4J "^ Ji S rH
^s«4HMO 3 -Hcu3>r
-------
13
cu
G
C
O
U
ughput.
lexity.
o
M
JS
4J
T)
CU
CO
cfl
CU
n
a
a
M
60
ft
0
O
O
•n
cu
CO
Cfl
cu
M
CJ
a
H
,C
igh-pressure fluidized bed boiler.
hn Kelly of Power Generation Services, Lester, Pa.
,G
8
o
M
14H
13
0)
4-»
4-1
•H
0
O
•H
0
<->
>>
ft
TJ
CU
•H
rH
ft
ft
3
CO
•n r
ecovery boiler
M
4J
cfl
CU
J3
13
CU
!-j
•H
<4H
a
p
*r
VO
CO
4J
•H
§
tH
Cfl
4-1
O
H
Fn
o
m
o
00
~cfl
•H
CO
ft
in
• «
4-1
CO
O
0
4J
•H
ti
1=1
CO
Q
>>
^Q
TJ
cu
•H
•H
ft
ft
3
CO
rH
60
«
c
•H
m
CN
P-l
0
O
O
OO
~cfl
•rl
M CO
0 ft
4-1
cfl O
(-4 0
cu r^
PJ
0)
60 •• S
co §
13 C
P! 0 0
cfl -H 00
CU
0
•H
,0
S-i
3
4J
0
cfl
CU
4J
CO
0
4-1
•H
13
ti
O
CJ
3
rf
CU
4J
CO
rH
• «
>,
4-1
•H
O
cfl
ft
cfl
CJ
iguration: tandem compound
two flow 25 in. ends
G. Taylor, Power Generation Systems.
14H
pi
0
a
cu
p!
•H
f>
rl
3
H
•
*
,Q
13
CU
•H
rH
ft
ft
3
CO
a
ers eliminated except for deaerating heater. Capacity and head for feedwater pumps reduced.
ity reduced due to elimination of feedwater heaters . Water treatment requirements reduced due
onditions.
4->
cfl
CU
,C
M
0)
4-1
CO
>
13
0)
CU
Fn
O
X
cu
tH
ft
0
0
a
60
a
•rl
ft
•H
P-I
ft
O
0
CO
CU
4J
CO
£
O
rH
O
4-1
n gas-turbine power. Increase in number of gas turbines.
o
•H
4-1
O
cfl
t-t
MH
$-1
CU
J3
60
•H
33
CT
ty of emergency gas-turbine driven generator. Reduced capacity of transformer for start-up
•H
O
cfl
ft
cfl
O
13 Q)
(U Pi
U -H
3 r£l
13 »-i
CU 3
OS 4J
M
307
-------
CN
m
W
pq
H
S23
Es
Q
Q
E2
"^4
M
p4
pq
H
to
o
c_>
P^
£^
r-rl
§
PM
H
o*
H
H 13
U CU
•rl
P t-l
W -H
S3 -0
H O
pn S
s
0 G
o o
&< T)
O CU
H CO
C/l cfl
] * ^
PQ ^^
^•*
O
U
u
M
H
<
W
M
Q
<
P
W
pq
Q
NJ
M
P
M
f~~i
1 "]
fn
col
CO
o
u
tJ
cu
rH
rH
CO
4-1
CO 6^
G
H
c
O
•H
4J
CO
rH
rH
CO
4-1
CO
G
4-1
CJ
CU CO
0 *-i
a, co
•H 0
3 o
Cf
W
o
o
rH
O
O
O
o"
O
vO
O
0
O
ft
vO
oo
CO
o
0
o
fS
rH
CM
CO
4-1
C
cu
0
s
> CO
o ca
t-i cu
l"3
M T3
0
*rj g
a
cO >-t
O
CO 4J
CU CO
M 3
3 rQ
4-> 0
0 O
3 0
M
4J
C/)
O O O O
vO O 00 v£)
co r»* oo O
CO
o o o o
o o o o
O O 0 O
O O 0 O
VO CN 00 vD
i— 1 CM CM CO
ft ft ft
CM CM in
CM
o o o o
00 0 O
o o o o
ft Cl *\ *X
o o o o
O CM O CN
CO & CO cfl fl M
0 4J >> CO 4J -H 3
CU CO )-i CO ,Q 4-1
4J ^0^4
CO Vt rH 4-1 Tj 3 CO
>, O Cfl O (3 4-1 Cfl
CO 4-1 > CU CS| 00
^3 CO O rH CO
3 3 0 H 0 cfl 0
W ^Q CU O O 00 O
g ^ a s-i M
60 o m o iw
C a cu cu w
•H 4-1 60 00 60
CO O Cfl Cfl PJ )-4 G
CO -H rH 4J -H O iH
CU 4J 3 CO Qt 4-1 Pi
CJ cfl O I -H O -rl
0 ^Q -H T3 P< CU ft
M CO 4-1 C rH
T3 cfl cfl o -H
CU
3
fu
O
oo
CO
o
o
o
*t
o
oo
CM
A
CM
ft
3
cu
cfl
0
T3 60
G G
Cfl -H
rH
CO 60T3
M G G
o -H m
0) (U
3 Q) 4-1
fn 14.) pj
0 CU
O rH ,0
O cfl M
O O
O O
m CM
0 rH
rH
o o
o o
o o
00
rH in
co r^N
ft
vO
60
C
•H
rH
T3
G
cO
JS
60
G 4J
rH 3
pj
tij *i3
r* rt
cfl
rH
cfl JM
o co
0 0
co oo
CO OX
vO
o o
o o
0 0
o o
^D vO
o a*
•\ M
CM rH
pj
O
•H
4-1
Cfl 4J
4-1 G
G cu
CU 0
0 ft
2-3
4J cr
S "
•H W
3
*"O O
C CU
cfl G
cfl
rH rH
O rH
M CU rH
4-1 0 Cfl
C CO 4-1
O -H O
u a w
1}
CO
uoo
o
o o
O "vT
O ft
o o
CN VO
r^- rH
*,
^^
P-.
O
o
o
ft
m
o\
o
o
0
n
m
CN
VO
rQ
CO
4-1 CO
•H TJ
C rH
3 0
M-l
W -H
CO U G
0 < cd
CU PL) 0
4-1
CO TJ rH
>, CU CO
CO -H C
3 -H CU
CO TJ 4J
O
PM
^1
O
cfl
CU
G
0)
60
^>
O
G
cu
60
M
CU
0 ,J
H
O
•
CO
0)
rH
3
*"O
0
0
M
O
4J
CQ
3
*§
O
CJ
o
•H
4-1
cfl
JQ
cfl
•H
rO
cO
S-4
O
14-1
CO
C
0
•H
4J
cfl
"G
3
O
H cfl ,
M
0)
rH
*rH
0
4J
^
M
CU
^
O
a
cu
M
4-1
cfl
cu
rC •
CU
m C
O -H
60 M
G 3
•H 4J
^4
•H 0
M-l Cfl
0)
0 4J
G CO
r §
60
• O
CU vD
•• CO
CO
C 4-1
0 O
•H
4-1 CU
•H CO
•0 3
C
o cu
0 fi
4-1
TJ
cu "O
•H C
IH cO
•H
T3 CO
0 CD
•H
J3 4J
4J -H
•H *T3
15 cJ
O
4J O
G
CO 0
•H cfl
M CU
CO 4-1
> CO
G
•H M
CU
W Js
U O
-------
CO
m
W
O
O
\o
I
—•
PH Ctf
01
1 1
r— t r— i
cfl 43 CJ
C 4J
O -H CO
•H |3 cfl
4-1 O
a
Cxi
VO
Ov
VO
00
rH
CO
VO
rH
CO
CM
rH
cfl
4-1
O
4-1
•§
CO
CO
m
CO
CO
/•-x
m
O
^*
r^
o
vD
CO
O
m
CO
CM
rH
CM
m
n
a
o
•rl
4J
Cfl
rH
Cfl
O
CO
w
vO
O
r--
m
CM
/— \
rH
0
rH
00
CM
rH
m
VO
CM
CM
CM
O
CM
CM
CM
m
p*.
•
CO
oo
CM
rH
cfl
4-1
O
4J
3
CO
o
rH
in
CM
rH
Oi
CM
00
oo
f^.
•xt
60
a
•H
to (3
3 0
13 -H
4-1
4J CJ
CO 3
CU M
t 1 ^J
0) CO
•w C
G 0
M O
CJ
O vO
m vo
m vo
ao
CM
s-\
CM
rH
oo
rH
CO
U
O CO
m cr>
m i^-
o
CO
o vo
m vo
m vo
in
CM
O CO
U-| rH
m r-»
co
CO
H
w
0
o3 O
CO [ 1
Q) 60 H
4-1 fi M
M -H PH
to
cfl
H
P*
•H
""0
0>
4-1
c
cu
CO
cu
to
cj
o
^
cfl
0)
r4
pq
CO ,
rH
r~-
a\
rH
CO
S-i
cfl
0)
CO
r4
cfl
01
>>
CM
"-^
>,rH
>-.
rH
3
l-D CO
O 13
r^ C o
O» O iH
rH -H H
4J a)
• 0 CX
C! 3
CO t-l (H
"-5 4J O
S^H
4J
CO O
O
O
1
CO
t-t
1 O
4-1
to ca
ai 3
rH ,n
•H B
0 0
J3 CJ
13 O
CU -rl
,Q 4J
cfl
13 43
•H cfl
II) *H
rH 13
PH ^
CO
to
Cfl CO
01 M
t>i cfl
0)
CO >.
1 CM
CM rH
rH~ CM
1
CM O
4-1
1
CM
CO
•H 1
CO
cd co ci
43 -H 6-S O
CO CM -H
to cfl \ 6-S 4-1
01 43 rH CM O
rH | "^^ 3
•H W r- H to
O CJ 1 4J
43 >
rH 0 Cfl 0)
cfl 01 ~^
a 4-1 rH
co C I
MM CN
O
G
o
•H
4J
0
3
to
4J
a
o
CJ
to
ca
01
CO
T)
309
-------
• The capital cost of the pressurized fluid bed boiler
plant is up to 20% less than that of the adiabatic
combustor plant.
• Plant heat rate can be reduced 'x- 1% compared with con-
ventional coal-fired steam plant using state-of-the-art
technology.
• Environmental constraints can be met, heat release to
cooling water is reduced 'Vi 35%.
• The simplicity of the adiabatic combustor system, the
potential for a shorter construction time and the
potential for near term commercial application may make
the concept commercially viable.
6.2 RECIRCULATING BED BOILER DESIGN
A deep recirculating fluidized bed combustion boiler was
conceived as an alternative to the proposed pressurized fluid bed
combustion boiler concept which is evaluated in Section 2. A recirculating
bed boiler may offer advantages over the proposed design but will require
more development and is thus considered a second generation concept.
The recirculating fluid bed concept is illustrated in Figure
110. Gas is fed to the base of an open draft tube section. The super-
ficial velocity of the gas flowing up the riser may be 10 to 60 ft/sec.
The solids are picked up pneumatically in the draft tube. The effective
overall density of solids and gases is less in the draft tube section
than in the downcomer, which creates a solid circulation pattern upward
through the draft tube and downward in the downcomer. Solids and gases
from the draft tube pass into a fluid bed which may be of expanded
cross-section. Solids from the fluid bed above the draft tube flow into
the downcomers and enter the base of the draft tube. Gas is introduced
at the base of the downcomer at a rate necessary to permit the down-
ward flow of solids.
310
-------
Uwg. 6207A3L
Gas Out
*
I
-L^~
* v.
\
Alternative
Solids Feed — - c
\
VJ
»
i
i
r
t
•
^
T
i
. --
r^
^
1
i
i
r
---•
^—
^ Solids Out
^- Draft Tube
• Downcomer
Gas Distributor Plate
*• Snlirl
Gas and Solids Feed
—*• Gas Flow
Figure 110-Recirculation Bed Concept
311
-------
The potential advantages of a recirculating fluid bed boiler
concept are:
• Feeding reactants to the bed. Oil and coal have been fed
to recirculating beds without distribution and agglomeration
problems.
• Boiler turndown and start-up. Locating heat transfer surface
in the recirculating solids permits turndown by regulating the
solids recirculation rate. Boiler start-up and shutdown are
also simplified by so locating heat transfer surface.
• S00 removal and NO reduction. Limestone or dolomite can
*£ X
be used as the primary bed material in order to remove S02
and to catalyze the decomposition of NO. The riser portion
of the bed can be operated under reducing conditions; the
main bed and recirculating sections, under oxidizing conditions.
In this way the formation of NO is reduced and the activation
X.
of the sorbent is promoted. It is postulated that fluctuation
between oxidizing and reducing conditions promotes penetration
of the limestone sorbent by the sulfur in the forms of sulfate,
3
sulfite, or sulfide.
• Decreasing the cost of fluidized bed boilers further. A
system study of pressurized fluidized bed boiler power plants
indicated that they are lower in cost, more efficient in
operation, and more effective in pollution abatement than
conventional coal- or oil-fired power plants with stack
gas scrubbing systems. The deep recirculating beds promise
to reduce this cost further because of potentials in
decreasing coal feed points, in increasing the heat transfer
coefficient in the downcomer, and in simplifying the control
and design of the boiler.
Several recirculating fluid bed systems have been developed
456
and operated. The British Gas Council ' ' developed deep recirculating
fluidized beds for oil and coal gasification. In the gasifier this was
312
-------
used to smooth out the local high temperatures at the point of oxygen
entry and in the hydrogenator to allow for the quick removal of particles
wetted by the oil in the vicinity of the oil inlet and thus avoid the
formation of agglomerates as well as closely control temperature for
process purposes. Experiments were performed in large-scale models up
to 5 feet in diameter and operated at pressures and temperatures up to
70 atmospheres and 750°C (1382°F).
The same concept was also used for vertical pneumatic transpor-
tation of sticky and bridging powders. A variation of this concept was
utilized to promote solid mixing, circulation, heat and mass transfer in
8 9 10
the bed , and for pretreatment of caking coal . Taskaev and Kozhina
utilized a recirculating bed for the low-temperature carbonization of
coals. Product gas preheated to a temperature of 600° to 700°C (1112 to 1292°F)
was recirculated to entrain the coal particles of sizes up to 1/4 in. through
a vertical draft tube 1 in.in diameter and 20 in«in length into a
concentric tube 6 in, in diameter and 40 in. in length. The coal or char
particles fell through this annular space (a downcomer) in a dense
descending bed and were then reentrained by the gas and recirculated
until carbonization was complete.
A recirculating fluid bed reactor is utilized in the multistage
fluidized bed coal gasification process for production of low-Btu gas
for power generation being developed by Westinghouse. A recirculating
bed operating near 1600°F is used in this concept to devolatilize the
coal and potentially to simultaneously desulfurize the fuel gas.
A preliminary design and evaluation of a pressurized recirculating
fluid bed boiler for combined cycle power generation was made. The
available literature on recirculating fluid beds was reviewed, cold model
studies performed to establish initial operating and design criteria,
a preliminary design developed, costs projected, and the concept
evaluated.
313
-------
Based on the above discussions, a deep recirculating fluidized
bed appears to be an attractive approach to the design of a boiler. The
subsequent sections illustrate the concept of a deep recirculating
fluidized bed boiler, describe the cold model studies, and discuss
criteria for design and operation.
6.2.1 Deep Recirculating Fluidized Bed Boiler Concept
The concept of a deep recirculating fluidized bed boiler is
illustrated in Figure 111. Primary air along with coal or oil fuel is
fed to the boiler at the base of an open riser section. This section is
bounded by metal or ceramic walls which may contain water-filled tubes
for wall cooling and/or steam generation. The superficial velocity of
the air and combustion gases flowing up the riser is 10 to 60 ft/sec
at the operating temperature and pressure — 1300°F to 1950°F and 4 to
30 atmospheres. Coal fuel may either be pulverized (less than 200 mesh)
or coarsely crushed (perhaps 1/4" x 0). Oil may be introduced to the
boiler so as to wet the stream of sorbent after it flows down the down-
comer, as it enters the base of the riser and jets upward in the riser.
The fuel combines with the primary air in the riser. The heat produced
increases the temperature of the combustion products, fuel ash, and
sorbent; some heat may be removed through the walls of the riser. The
quantity of sorbent entering the base of the riser is controlled by choice
of design and operating parameters in order to limit the temperature
increase of the gases and solids flowing up the riser.
The solids and gases emerging from the riser pass into a bed
of expanded cross-section and decreased superficial gas velocity. Baffles
and/or heat transfer surface may be located in the bed section in order
to obtain smooth fluidization and recirculation of the sorbent, to
improve sulfur removal and NO reduction by the sorbent, and possibly to
recover some heat. The gas velocity in the bed is in the range 1 to 15
ft/sec, sufficiently large to fluidize the sorbent but sufficiently
small to minimize entrainment. The sorbent may either be coarsely
crushed (1/4" x 0) or finely ground (perhaps 100 to lOOOy).
314
-------
Dwg. 723B343
Cyclone
Freeboard
Down comer
Heat
Transfer
Surface
Particle
Separator
Secondary Separation
Air Air Primary Air
Air
Fuel Injector
Fig. Ill-Deep Recirculating Fluidized Bed Boiler
Heat Transfer
Surface
315
-------
Hot sorbent from the bed flows into the downcomer, passes
downward through the heat exchange surface (water/steam filled tubes which
may, for example, be platens of horizontal or long vertical tubes), and
enters the base of the riser section. The exchange surface extracts the
heat energy transferred to the sorbent by the combustion process
occurring in the riser and bed. Secondary air is introduced at the base
of the riser at a rate necessary to permit the downward free flow of the
sorbent and to prevent reducing conditions in the riser due to any
residual fuel accompanying the sorbent. Flow of the sorbent in a section
of the downcomer can be reduced or halted by reducing or cutting off the
flow of secondary air to that section. This capability makes it possible
to adjust independently heat removal and heat production in the boiler.
Control of the primary air can be used to adjust the recirculation rate
of the solids and the degree of oxidation in the riser and above the bed.
Combustion after the bed increases the range of gas-turbine operation.
Start-up and turndown are thus facilitated.
6.2.2 Cold Model Studies
In order to study the feasibility of using deep recirculating
fluidized beds for coal combustion boiler power plants, a two-dimensional
cold model was constructed from the transparent acrylic sheets to
investigate solid circulation, operation characteristics, and design
criteria. The detailed dimensions are shown in Figure 112. Only the
bed cross section, 8.5 in. x 1.5 in.is fixed; the other dimensions
(A, B, C, D, E, F, and °0 in Figure 112 can be adjusted to study the
effect of these design variables. Pressure taps are also provided for
measuring differential pressure drops between points 1, 2, 3, and 4.
Some experiments were performed in a slightly modified two-
dimensional bed as shown in Figure 113. Two separate fluidized beds
(or "boots") are created by inserting two aluminum bars 3/4 in. thick
underneath the draft tube. The objective of this modification was to
study the effect of the draft tube inlet design on the solid circulation
rate and the air distribution between the downcomers and the draft tube.
316
-------
Downcomer
Draft
Tube
Air— —
Plenum
8.5"
X
F
4
D
n
Airf
Air Jet
Nozzle
1
Air
Dwg. 6197A35
1.5"
Solid Aluminum
Bar
Downcomer
A, B, C, D,E, F, and a
are adjustable
= Pressure Tap
Positions
Air
Plenum
Figure 112-Detailed Schematic of the Two-Dimensional Cold Model
317
-------
Dwg. 6197A36
Downcomer
Draft Tube
Air Plenum
Downcomer
Solid Aluminum
Bar
Air Plenum
Jet Nozzle
Figure 113-Two-Dimensional Cold Model with Modified Draft Tube Inlet
318
-------
All the dimensions in the modified model are similar to the original
model figure except now D1 = 5 in, F1 = 7 in, and G = 3.0 in.
Serpentine solid copper rods as shown in Figure 114 were also
inserted in the downcotner sections to simulate the imbedded heat transfer
surfaces. The serpentine rods are supported by inserting the ends of
the rods into closely sized holes on the aluminum bars. Two serpentine
rods were arranged in parallel 4.5 in. from the top of the draft tube
with equal distances between the rods and between the rods and the walls.
Ottawa sand was used as the bed material. During operation,
air is injected into the draft tube through the air jet nozzle to provide
an air velocity up to ^ 43 ft/sec in the draft tube while the downcomer
sections are "minimally fluidized." This creates a solid circulation
pattern upward through the draft tube and downward in the downcomers.
The minimum fluidizing condition in the downcomers is assured first by
turning up the air supply to the downcomers until air bubbles appear in
the downcomers and then turning down the air supply until air bubbles just
disappear.
The solid downward velocity in the downcomer was estimated by
following a tracer particle for 16 in.with a stop watch. The tracer
particles are silica gel dyed with red pigment and of similar size. For
every operating condition, at least ten particles in each downcomer were
traced and arithmetic average velocity was taken to be the solid particle
velocity in the downcomer. The solid recirculation rate was then
calculated by assuming plug flow in the downcomer.
A mathematical model has been developed to predict solid
12
circulation rate. The model is good to within + 20% if experimental
values of slugging pressure losses are used. The detailed development
of the mathematical model and the experimental conditions and results are
presented in Appendix P. The critical design variables for a recirculating
fluidized bed boiler are identified as the draft tube inlet design, the
draft tube height, and the downcomer/draft tube area ratio.
319
-------
Curve 653461-A
1 "
copper rod
"4
2
1"
1"
9"
Figure 114-Dimensions of the Serpentine Copper Rods
320
-------
6.2.3 Bed-Tube Heat Transfer Coefficient
No heat transfer experiment has been performed to obtain the
bed-tube heat transfer coefficient for the boiler tubes immersed in the
downcomer; however, literature data on heat transfer coefficient in a
moving bed were utilized to project a reasonable value for the present
recirculating bed boiler design. The aspects of heat transfer in a gas
fluidized bed has been reviewed previously. The present focus is on
literature data which are relevant to heat transfer in a moving bed.
If radiative heat transfer can be ignored, the dominating
factor in the heat transfer is the high heat capacity of solid particles
relative to that of gas. Heat is primarily transferred from the bulk of
the bed by solid convection, and convective heat transport through the
gas is relatively unimportant. Because of the high surface area exposed
by the particles, the solids effectively act as a local heat source or
heat sink. Consequently, the thermal gradient between the immersed
transfer surface and the bed is restricted to the zone immediately
adjacent to the surface. Thus highest overall heat transfer coefficient
will occur under conditions which limit particle residence times at the
surface. It is difficult to rely on the ordinary bubbling processes in
a fluidized bed for controlling the heat transfer in the bed. A forced
particle circulation such as in a recirculating bed design will be able
to control the particle flow past the heat transfer surfaces in a
regulated manner.
1 *^—i f\
Betterill et al, examined the relation of heat transfer
efficiency with particle residence time on the heat transfer surfaces
both experimentally and theoretically and found that high rates of heat
transfer could be obtained if particle residence time at the transfer
surface was sufficiently limited. The experiments were carried out in a
13 14
stirred packed bed, ' and a flowing bed with a small exposed wall
heat transfer surface. The bed-tube heat transfer coefficients obtained
2
ranged from 70 to 280 Btu/ft -hr-°F depending on particle sizes, particle
properties, and particle residence times at the heat transfer surface.
321
-------
Actual heat transfer coefficients for various densities of fluidized coal
2
were also measured. The results ranged from 80 Btu/ft -hr-°F for a
2
residence time of ^ 50 milliseconds to 190 Btu/ft -hr-°F for a residence
time of % 7 milliseconds.
Experiments were also carried out by mounting arrays of tubes
on suitable jigs in the straight section of the continuous channel
containing a flowing fluidized bed. ' Both horizontal and vertical
tube bundles were used in the experiments. Local heat transfer coefficients
between elements of the surface and the bed in cross flow increased with
an increasing solids flow rate, as expected. The local heat transfer
2
coefficients are between 100 and 130 Btu/ft -hr-°F. However, when the
solids flow rate was increased, a wake developed from the downstream —
facing surfaces, and there was a marked tendency for the bed to defluidize
locally at the upstream-facing surface. Consequently, the overall
bed-tube heat transfer coefficient was not very sensitive to solids flow
2
rate and was of the order of 55 Btu/ft -hr-°F. There was no attempt to
align the axes of the tube bundles in the flow direction of the flowing
fluidized bed.
Another similar heat transfer phenomenon is that of heat
transfer between submerged heat transfer surfaces and the bed in a
spouted bed. The heat transfer coefficient measured in the annular
section between the spout and the containing wall were found to change
2 2
from ^ 30 Btu/ft -hr-°F close to the wall to ^ 50 Btu/ft -hr-°F close
18
to the spout. The only difference from the heat transfer in the
downcomer section of a recirculating bed is that the downcomer section
is minimally fluidized in the former design while there is no gas input
to the annular section in the latter case. Due to this added agitation,
the heat transfer coefficient in a recirculating bed is expected to be
higher than that in a spouted bed.
19
Zenz and Othmer reviewed the heat transfer in a moving bed.
The typical heat transfer coefficient obtained for coal and limestone
2
in a horizontally flowing bed of solids ranges from 40 to 100 Btu/ft -hr-°F.
322
-------
For a moving bed with little relative movement between particles, the
thermal conductivity of the solid particles becomes the controlling
factor. Equations are available for these calculations.
Based on this literature review, a commercial recirculating
bed boiler is designed by assuming the overall heat transfer coefficient
2
in the downcomer section as 50 and 75 Btu/ft -hi
are then compared with that of the basic design.
2
in the downcomer section as 50 and 75 Btu/ft -hr-°F. The boiler costs
6.2.4 Conceptual Recirculating Bed Boiler Design
A conceptual recirculating bed boiler design is prepared on
the basis of the boiler plant design developed by Westinghouse under
contract to EPA. The operating conditions and design parameters for
a 318 MW boiler and the power cycle have been summarized in Table 5 of
Section 2.1. The conceptual recirculating bed boiler design is evaluated
in two cases. Case I assumes the overall heat transfer coefficient in
2
the downcomer section to be 75 Btu/ft -hr-°F; and Case II assumes 50
2
Btu/ft -hr-°F. The coal is assumed to combust mostly in the draft tube
and partially in the fluid bed above the draft tube. The heat transfer
surfaces in the fluid bed keep the bed temperature essentially at 1750°F.
The heat transfer surfaces in the downcomer section reduce the solids
temperature from 1750°F at the top of the downcomer down to 'v 1000°F at
the bottom of the downcomer. The solids are then picked up in the draft
tube to increase their temperature to 1750°F.
The design selected for a 318 MW recirculating bed boiler
consists of six individual modules of 12-ft inside diameter with one
module for preheater, two modules for evaporator, two modules for super-
heater, and one module for reheater. This design was selected to permit
standard shop-fabrication of each vessel. A four-module design could
also be developed, which would increase the diameter of two of the
reactors. The conceptual design is depicted in Figure 115. Coal is fed
concentrically with 90 to 95% stoichiometric air into the draft tube.
Most of the coal is expected to be consumed in the draft tube. The
323
-------
Dwg. 62lWf6
Steam Headers
Heat Transfer Tubes
Downcomer
Draft Tube -
Air Plenum
~i
\
I
I I
i i
r
/ /
* /
\ \
\
A
^
-111
J1_L_
D
Figure 115-Schematic Design of the Advanced Recirculating Bed Combustor
'.'24
-------
remaining coal is carried into the expanded fluid bed above the draft tube
where additional air is available from the downcomer section. Thus the
draft tube is operated under a slightly reducing condition while the
fluid bed above the draft tube is operated under an oxidizing condition
20
with 10% excess air. High excess air can be utilized in this design.
However, high excess air (e.g. 100%) has not been considered in the
present evaluation. The effect on vessel design and performance (e.g. turndown,
NO ) would have to be considered. The bed cross-section expands at the
outlet of the draft tube to reduce the superficial gas velocity to 3 to
4 ft/sec to prevent slugging and to reduce carry-over of solid particulates.
The heat transfer surfaces are vertical tubes of 2 in.in diameter and
extend through the downcomer section and the fluid bed above it. The
horizontal bends into the steam headers just above the fluid bed will
help prevent slugging and reduce elutriation. The design parameters
and operating conditions are summarized in Table 54. The resulting
physical dimensions of the modular combustors are presented in Table 55.
The largest vessel diameter was restricted to 12-ft inside diameter
so that the vessels can be shop fabricated and shipped by rail.
6.2.5 Boiler Operation and Performance
The coal-burning recirculating fluidized bed boiler would be
operated somewhat differently from the basic pressurized fluidized bed
boiler because of the differences in boiler design and modular arrange-
ment. In the recirculating fluidized bed boiler design, each individual
module performs only one function, i.e., it is either for preheating,
evaporating, superheating, or reheating. Thus, this concept requires
a minimum of four modules to complete the total boiler function from
preheating to superheating and reheating. In the pressurized fluidized
bed boiler design, however, each module contains one preheater bed, two
superheater beds, one reheater bed, and walls for evaporating. Thus,
each module is capable of generating 280 MW electrical power. Due to this
difference in modular arrangement, the philosophy for cold start-up,
hot restart, turndown, and load control would be different.
325
-------
-*
in
w
1-3
PQ
^£
E™t
JS
JE^
00
H
en
w
PH
H
P5
O
P^4
C/]
£5
O
M
H
M
P
2
o
o
0
M
H
<3l
PM
td
PM
o
Q
f£\
•^
CO
3
a
5
3
<^
PM
2<
O
M
CO
PL]
P
O
P^
W
. -j
H
O
pq
P
W
pq
P
W
M
|
fc-3
E
o
H
H
J
1-3
O
PH
H
O
8
J_l
CO
4J
CO
i-C
co
}_)
CD
4J
cd
01
^1
OJ
P
3
M
O
4-J
cd
^
o
p^
J>
w
M
O
4-)
cd
o
ft
cd
^
CO
CO
PM
in
en
O rH
m *
rH t- O
rH in
fN^
rH
O
in
en
O rH
m 1
CM r~. o
rH in
rH
O
m rH
CM t»- f
H O
in
rH
O
o 8
m rH
H r- T
rH O
m
rH
S-4
§
O
CJ
^H 3
0 O
P
co a
M 'H
3
4-J CO
co cd oo
co M a
rH CO td
3 P- J3
TJ e u
O 0)
S H CO
M-l T3 3
O CO 4J
pfl cd
CO TJ CO
fN «p- 1 f"V.
^ 3 a
3 rH 0) PH
23 fn H o
vO
O vo
rH O
H
X!
X!
O
• en
en CM
CM
en en
\O
O vo
rH CO
rH
X
X
o
• oo
VO rH
en .
vo en
vO
O vo
rH O
rH
X
^
O
• in
vO vO
rH .
VO rH
VO
O VO
rH O
X H
X
m
• m
CM en
m •
CM rH
M
^•v^
,Q
rH
«\
co
4-1
cd
M C
jC O
"""•^i- *H
3 4-1
4-1 Cd
Pq rH
3
>• r4
4J -H
P U
4-> -rt
Cd rH
S £
o
O O -i -H
Cl) P 4J
o ^w ^*>
C{J Q) Clj 4-1
u-i X M -H •
(-I4-I P O 0
3 0 O 0)
M C CO -H rH CO
J_l . 4J 4-J > 4J
co >s o co m
MH 4-1 CO O S rH
C 0 -^. ^ -H tj
td O 4-1 'Vj ,£! O CO
HO) 0) -H >W
4-1 CO 4-> CO *rl
td co ,0 M co PI 3
CO td 3 -H 3 rH
PC O H m I
•H a 4:
cd cd i
r«M
CO H 4J
O cd 3
M mm
0
• •
H
o
rH
• *
rH
0
rH
• *
rH
0
rH
*•
rH
I-l
QJ
0
0
a
g
a
a
CO
p.
o
-d
§
?{ ^
3 4J
H«3
4-J
cd Q)
s^
m
CM
in
CM
m
CM
0
CM
M
a)
IM
CO
a
co cd
0) t-t
^«
!-i cd
CO CO
0
u a)
i "^
P t*>
M-l ^
O TJ 6*5
CO
4J iH «
cj Qi CO
o o .a
-1 0 fn
326
-------
m
in
w
H-l
PQ
H
C/D
W
|~)
Q
*
p^
O
H
O
U
W
DC
H
PM
O
en
j55
O
H
CO
H
P
-j
cd
til
f~]
M
P.
3
.
O
cti
j^i
O
a
cd
W
^
o
4J
rd
S-i
O
P
a)
>
a)
ai
S-i
(3
•H
CO
C
o
•H
CO
c
cu
a
•H
P
T3
•H
00
•
CM
•
T3
•
•H
00
CM
«
Id
•
•H
00
CM
^
t3
•
•rl
f^
•
CM
rH
• X-N
00 !-i
•H CU
TT i 1 1
O G
4-1 C8
•rl
(-4 T)
QJ ^— '
U-l
0) <
M
" —
13
•H
CO
*
O
•
tJ
•
•rl
CM
O
r-l
•
T3
•
•H
O
O
rH
fc
TJ
*
•rl
CTi
•
CT^
-^s
cu
4->
B
n)
•H
v*^
P5
id
•
•rl
O O
• •
CM O
rH rH
•
13
•
«(-|
0 0
CM O
rH rH
•
T3
•
•H
0 0
CM O
rH rH
•
T3
*
•H
O O
• •
CM O
rH rH
s~**
M
CU /~N
4J 4-)
Q) 43
6 00
Cd -rl
•H O
rH en
vo o
o\ o
rH CO
0 0
• •
in o
rH CO
CM O
* •
oo o
rH CO
/— N /^V
4-1 4J
(JO 60
•H iH
CU 0)
43 43
^^ \— '
EtJ P4
327
-------
The results from the cold model studies in the two-dimensional
model indicate that the solid circulation rate in a recirculating bed
can be easily controlled by turning down the air input into the downcomer
section. The circulation can be effectively stopped by completely
shutting off the air input into the downcomer. Hence, theoretically,
the recirculating fluidized bed boiler will be able to follow the load
change continuously from no load to full load. Subdivision into modules
for turndown consideration Is not necessary. The design of the recirculating
bed boiler is thus based on other considerations, such as ease in assemblage
and erection, control and turndown, and economics. To be conservative, the
318 MW recirculating fluidized bed boiler design consists of six modules:
one for preheating, two for evaporating, two for superheating, and one
for reheating. The downcomer section of preheater and reheater modules
will be separated into two sections by water walls to facilitate turndown.
Air plenums for evaporator and superheater will be sectioned into several
sections so that the circulation rate of each section can be individually
adjusted.
The bed temperature in the fluid bed above the draft tube is
the primary variable for control: too high a temperature will cause ash
to fuse and dolomite to lose reactivity; too low a temperature will decrease
substantially the reactivity of dolomite and possibly decrease the combustion
efficiency of coal in the bed. If carbon is elutriated from the bed, it
can be recycled and fed at a point just above the downcomer section so that
the elutriated carbon particles can be trapped in the descending bed in
the downcomer to increase residence time of the particles. There is no
fresh coal feed in the downcomer section to compete for the air, and there
is an additional chance for the unburned carbon particles to be picked up
in the draft tube and carried into the fluid bed above the draft tube.
Thus, the carbon burn-up cell in the pressurized fluidized bed design*
probably will not be needed in the recirculating fluidized bed design,
which would simplify the control scheme and boiler design.
*
A carbon burn-up unit is not considered necessary in either design with
high excess air.
328
-------
Air Input into the downcomer section will be used to control
the solid recirculation rate and thus the heat transfer rate to the heat
transfer surfaces. Because of this great flexibility, cold start-up, hot
restart, and load control are much simpler. To start up, auxiliary fuel
can be fed into the fluid bed through the draft tube. When the bed
temperature reaches the coal ignition point (^ 1000°F), coal feed can be
started. The solid circulation can be started when the bed temperature
reaches the design temperature 1750°F. For hot restart, the recirculating
bed design would be even more convenient, because the bed is not completely
shut off during turndown as in the case of the pressurized fluidized bed
design. The bed is essentially maintained at the operating temperature
at least for some sections. A hot restart will simply mean an increase
in the coal feed rate and the solid circulation rate, while in the
pressurized fluidized bed design, the bed has to be brought up to the
coal ignition temperature by auxiliary firing before coal feeding can
be initiated. This means an additional time delay in following the load
change. This potential capability of shortening the load response time
is an important advantage for the recirculating fluidized bed design.
The performance characteristics of the recirculating fluidized
bed are not expected to be too different from those of the basic
pressurized fluidized bed design. The overall boiler efficiency is
expected to be comparable, with slightly higher pressure drop through the
recirculating fluidized bed due to deeper bed design; however, the air
pollution control capability in the recirculating bed is potentially
greater. The two-stage combustion of the coal in the recirculating bed —
reducing conditions in the draft tube section and oxidizing conditions
in the downcomer section and the main bed — promises to reduce the
formation of NO . If fluctuation of the sorbent between oxidizing and
X
reducing conditions promotes penetration of the sulfur into the sorbent
in the form of sulfate, sulfite, or sulfide, then increased activity and
higher utilization of the sorbent can be achieved.
329
-------
6.2.6 Economics
In addition to the air pollution control capability, the
recirculating fluidized bed boiler possesses the following potential
advantages:
• In decreasing coal feed points. Coal can be fed and combusted
in the draft tube through a single coal feed point. The
dilute pneumatic transport of coal particles in the draft
tube increases gas/solid contact. The air by-passing
through bubbles which adversely affect the gas/solid contact
in a fluidized bed does not occur in the draft tube.
• In preventing agglomeration. High solid circulation rate
in the recirculating bed prevents caking and agglomeration
of coal in the vicinity of coal feeding points.
• In capability of designing for deeper bed. The increased
circulation and mixing in a recirculating bed eliminates
the problems of particle segregation and temperature
gradient in the main bed. This permits design for a deeper
bed.
• In control and load response. Continuous turndown is
possible in a recirculating bed. For cold start-up or
hot restart, a recirculating bed is easier to handle.
The response time is shorter in following load change
in a recirculating bed as discussed in the preceding
section.
• Possibly in increasing the heat transfer coefficient
in the downcomer section over that in a fluidized bed.
Because of fast solid circulation through the downcomer
section, there is a possibility of increasing the heat
transfer coefficient in the downcomer section to higher
than that in a fluidized bed (50 Btu/ft -hr-°F). This
possibility has to rely on experimental verification.
330
-------
All the advantages mentioned here can be directly or indirectly
transferred into economic savings in designing a recirculating fluidized
bed boiler.
The current design for the 318 MW recirculating fluidized bed
boiler is a six-module plant with one module for preheating, two for
evaporating, two for superheating, and one for reheating. Each module
consists of only one deep recirculating bed as compared to four stacked
beds in the pressurized fluidized bed boiler design. The pressurized
vessels are designed for operation at maximum operating conditions of
150 psig and 650°F with 6-in.refractory insulation (4-in.insulation plus
2-in.hard face). Draft tubes are constructed from either 316 stainless
steel or Incoloy 800 and are to be watei?- or steam-cooled. Vertical tubes
of conventional boiler tube materials are used for heat transfer surfaces.
The total cross-section of the reactor vessel is used for the recirculating
bed area as compared to < 40% utilization of vessel cross-section area in
the pressurized fluid bed design. The cost of the six-module boiler is
evaluated for the heat transfer coefficient in downcomer the section of 75
o
and 50 Btu/ft -hr-°F. The heat transfer coefficient in the fluid bed
2
above the draft tube is assumed to be 50 Btu/ft -hr-°F. The heat transfer
surface required in the recirculating bed design and the boiler cost are
compared with those of the basic design in Table 56. Although the heat
transfer surface requirements for the recirculating bed design are 6%
to 38% higher than those for the basic pressurized fluidized bed design,
the overall boiler costs may be lower by 5% to 13%. The increase in heat
transfer surface requirements is due to a decrease in temperature difference
between the descending bed in the downcomer section and the heat transfer
surface; however, the decrease in module height from four modules of 110
ft to six modules of 30 ft substantially reduces the boiler costs. This
reduction in module height is primarily due to elimination of a stacked
bed design. When fluidized beds are stacked one on top of the other as
in the basic design, freeboard and air plenum have to be provided for
each bed, thus increasing the module height.
331
-------
co
H
CO
o
M
O
w
i—I
PQ
g
•H
CO
cu
p
13
CU
PQ
M
c
•H
4J
rt
rH
O
Vl
.1
*ri
o
cu
p!
a
»_j
•n
CO
0)
p
CJ
•H
CO
rt
PQ
o o o o
CO vO
. 1 ^J
rn i— i
vo
o\ CM m oo
rH H
5
4J
•i-i m
C P^
0) o -
•H 1 W
U Vl CU
•H rfl 0
M-l 1 rt
I4W CM MH Vl
CU 4-1 Vl O
0 MH 3 *•» d Vl
u -^ cn rt o cu
3 Vl iH 4-1
Vi -u vi o 4-i rt Vi
CU PQ CU P< rt 0) CU
<4H UH rt Vi J3 4J
co •> co > o vi rt
c vi g cu p. cu cu
cd cu rt cu rt D. J3
Vi 6 Vi M > 3 »i
H O H CM W CO pj
CJ
J 1 f4 11
+J H +J
rt S rt
CU 0 0)
PC P M
0
O
o
o
f*^.
vO
CO
•co-
1
O
0 0
CO O
•> r>
CO CO
vo in
M
CO
•CO-
,0
O
o
o
•t
O 00
VO 00
oo -*
ocT co
^rf" ^/J"
1
o
o
o
o
r*»
co
A
CO
•co-
rt
•*
VO O
o o
o
vo ^
vf vO
m
00
CO
•CO-
4J
i-j co
< 0
H 0
0
H Vl
CU
rH
•H
O
PQ
•
4-1
•H
C
3
13
CU
4-1
cd
o
•H
Vl
rO
cd
(4H
1
a
o
.c
co
cu
rH
3
T3
o
a
1
sd-
^
^^
oo
rH
CO
cd
Vi
0
14-1
CO
•H
4-J
CO
O
U
4-1
co
o
u
CU
fi
4-1
CO
cu
Jl
4-*
rt
CJ
•H
TJ
a
•H
CU
M
C
rt
Vl
cu
o
•H
Vl
a.
cu
fi
H
•
4-1
•H
a
3
13
cu
4J
cd
u
iH
Vi
A
rtJ
rt
<4H
1
a.
o
,c
CO
CU
rH
3
TJ
O
s
1
VO
^
§
00
rH
CO
rt
Vl
o
MH
CO
•H
4-J
CO
3
rsus
cu
>
rH
rH
rt
?
13
CU
H
0
O
O
CU
rQ
3
4-1
1
(U
,£>
3
4-1
11
4-J
MH
rt
Vi
T3
1
1~~|
CU
cu
4J
CO
ca
CO
cu
rH
c
•H
rt
4->
ca
vo
rH
CO
a
CU
,
o
rH
O
0
Pi
M
rt
332
-------
The cost of the particulate removal system is expected to be
comparable in both designs because the airflow rate is similar and the
dust loading is expected to be similar, as well. Lower velocities in
the freeboard area in the recirculating bed design should result in
slightly lower dust loading if slugging does not set in.
The structural cost for six modules, each 30 feet high, should
be lower than that for four modules, each 110 feet high. The manifolding,
however, would be more expensive for the 6-module design. The instrumentation
and control are projected to be easier and less costly for the recirculating
bed design. Only 6 beds are involved, as compared to 16 individual beds in
the pressurized fluidized bed design. It requires turning on and off a
complete fluidized bed or a complete module in the pressurized fluidized
bed for turndown purposes, while in the recirculating bed design it only
requires adjustment in downcomer airflow rate. The coal feeding points
are reduced from 64 to 6, which also substantially reduces the boiler
costs. Thus, it is probably conservative to assume that the cost of
auxiliary systems are comparable for both the recirculating bed and the
pressurized fluidized bed designs.
6.2.7 Conclusions
The overall boiler costs of a recirculating bed design are
estimated to be lower by 5% to 13% than those of the pressurized
fluidized bed basic design, depending on the construction of the draft
tube and on the heat transfer coefficient in the downcomer section. The
costs are arrived at by assuming that the costs of the auxiliary systems
are comparable. This is a conservative assumption, in view of the
simplicity in operating a recirculating bed boiler. In addition to the
cost advantage, the recirculating bed design has the following potential:
• In preventing agglomeration. High solid circulation rate
in the recirculating bed prevents caking and agglomeration
of coal in the vicinity of coal feeding points so that low-
grade coal can be used.
333
-------
• In capability of deep bed operation. The increased circulation
and mixing eliminates the problems of particle segregation
and temperature gradient in the main bed. This permits design
for a deeper bed.
• In control and load response. Continuous turndown without
shutting off the bed is possible. This promises easier
start-up and faster response in following load changes.
• In S09 removal and NO reduction. The two-stage combustion —
£• X
operating the riser portion of the bed under the reducing
conditions and the main bed and downcomer section under
oxidizing conditions — reduces the formation of NO and
X
promotes activation of the sorbent.
The recirculating boiler concept offers many potential advantages
over the base pressurized fluid bed boiler design and should be developed
further. However, the recirculating bed concept has not been demonstrated
on a sufficient scale to permit commercialization prior to the base
design. Further information on heat transfer coefficients and testing
of the heat transfer surface design, air distribution design, solid
circulation control, and fluidization characteristics are required. Thus,
a recirculating fluid bed boiler offers economical and environmental
advantages as a second generation pressurized fluid bed boiler.
334
-------
7. REFERENCES
1. SUMMARY
1. "Evaluation of the Fluidized Bed Combustion Process." Final Report,
GAP Contract 70-9, Westinghouse Research Laboratories, Nov. 1971,
(three volumes, NTIS PB 211-494, PB 212-916, PB 213-152).
2. International Fluidized Bed Combustion Conferences: Proceedings
First International Fluidized Bed Combustion Conference, Hueston
Woods, Ohio, 1968; Proceedings Second International Fluidized
Bed Combustion Conference, Hueston Woods, Ohio, 1970; Proceedings
Third International Fluidized Bed Combustion Conference, Hueston
Woods, Ohio, 1972.
3. Fraas, A. P., "Potassium-Steam Binary Vapor Cycle with Fluidized-Bed
Combustion." Paper presented at annual AIChE meeting, New York,
Nov. 1972.
4. Hoy, H. R. and Stantan, J. E. Amer. Chem. Soc. Div., Fuel Chem.
Prepr., Vol. 14, No. 2, May 1970, p. 59.
5. Squires, A. M. "New Systems for Clean Power." Int. J. Sulfur
Chem.. Part B, 7_, No. 1, 1972, p. 85.
6. Keairns, D. L. "Fluidized Bed Gasification and Combustion for
Power Generation." Proceedings Frontiers of Power Technology
Conference, Oklahoma State University, October 1972.
7. "Development of Coal-Fired Fluidized Bed Boilers." Final Report
Volume I, Office of Coal Research, R&D Report No. 36, Contract No.
14-01-0001-478, Pope, Evans and Robbins, 1970.
8. Elliott, D. E. and Healey, E.M. "Some Economic Aspects of High
Temperature Steam Cycles." Proceedings of Second International
Conference on Fluidized Bed Combustion, 1970.
335
-------
9. Burchard, J. K., Rochelle, G. T., Schofield, W. R., Smith, J. 0.
"Some General Economic Considerations of Flue Gas Scrubbing
for Utilities." Control Systems Division, E.P.A., 1972.
10. "Reduction of Atmospheric Pollution." Final Report, GAP Contract
with National Coal Board, London, England (June 1970-June 1971),
Sept. 1971.
11. O'Donnell, J. J. and Sliger, A.G. "Availability of Limestones
and Dolomites." Proceedings of the Second International Lime/Limestone
Wet Scrubbing Symposium, Vol. II, (June 1972), p. 799.
336
-------
2. ECONOMIC SENSITIVITY
1. "Evaluation of the Fluidized Bed Combustion Process." Vols. I-III,
submitted to the Office of Air Programs, Environmental Protection
Agency, by Westinghouse Research Laboratories, Pittsburgh,
Pennsylvania, Nov. 1971.
2. Keairns, D. L., Hamm, J.R.,and Archer, D.H. Paper presented at
Annual AIChE Meeting, San Francisco, Nov. 1971; AIChE Symposium
Series Volume "Air Pollution and Its Control," 126, Vol. 68, p. 267,
1972.
3. Highley, J., Chandrasekeva, D. and Williams, D.F. National Coal
Board, Coal Research Establishment, Fluidized Combustion Section
Report, No. 20, April 1969.
4. McLaren, J. and Williams, D.F. J. of Inst. of Fuel, 303,
August 1969.
5. Smith, S. Private communication, Boiler Tube Division, The Babcock
& Wilcox Co.
6. Genetti, W. J. and Bartel, W.E. Paper presented at AIChE 72nd
national meeting, St. Louis, Missouri, May 22-24, 1972.
7. ASME Boiler & Pressure Vessel Code, Power Boilers Section I, 1971.
337
-------
3. SULFUR REMOVAL SYSTEMS - LABORATORY STUDIES
1. Archer, D. H., Keairns, D.L. , et al. "Evaluation of the Fluldized
Bed Combustion Process." Summary Report, Contract CPA 70-9, submitted
to EPA, November 1971.
2. Brown, H. A., Jr., Penski, E.G., and Callahan, J.J. Thermochim.
Acta. ^, 271 (1972).
3. Ruth, L. A., Squires, A.M., Graff, R.A. Paper presented at Meeting
of American Chemical Society, L. A., California, March 1971; Environ.
Sci. Technol. 12, 1009 (1972).
4. Ruth, L. A. Thesis Ph.D., "The Reaction of Hydrogen Sulfide with
Half-Calcined Dolomite 4." City University of New York (1972).
5. Rabatin, J. G., and Card, C.S. ANAL. CHEM. 31., 1689, 1959.
6. Baker, E. H.. J. C. S.,464, 1962.
7. McKewan, W. M. TRANS AIME 224, 387, 1962.
8. Johnson, J. L. Paper presented at A. C. S. Meeting, Dallas, Texas,
1973 [Rev. Sci. Instr. _39_, 1227 (1968)].
9. Ho Bae, Jae, Rev.,Sci. Instr., 43, 983 (1972).
10. Wendlandt, W. W., J. R. Williams, E. L. Simmons. Thermochim. Acta.
1, 101, 1972.
11. Squires, A. M. Private Communication, April 1973.
12. Tarentbaum and Joseph, TRANS AIME, 135, 59 (1939).
13. McKewan, W. M. "5th International Symposium on the Reactivity of
Solids." ed G. M. Schwab, Elsevier, N. Y., 1965, p. 623.
14. Cox, D. G., Highley,J., et al. "Reduction of Atmospheric Pollution,"
Final Report to EPA by the National Coal Board, London, England,
September 1971.
339
-------
15. Jonke, A. A., et al. "Reduction of Atmospheric Pollution by the
Application of Fluidized Bed Combustion." Annual Report to EPA by
Argonne National Laboratory, ANL/ES-CEN-1002, June 1970.
16. Esso Research Center (England): results quoted in Ref. 1, Vol. 2,
p. 17.
17. Hatfield, J. D., Y. K. Kim, R. C. Mullins, and G. H. McClellan.
"Investigation of the Reactivities of Limestone to Remove Sulfur
Dioxide from Flue Gas." Prepared for the Air Pollution Control Office
by TVA, 1971.
18. Borgwardt, R. H. Env. Sci. Tech. 4., 59, 1970.
19. Borgwardt, R. H., and R. D. Harvey, Env. Sci. Tech. j>, 350, 1972.
20. Coutant, R. W., B. Campbell, R. E. Banelt, and E. H. Lougher (Batelle).
"Investigation of the Reactivity of Limestone and Dolomite for
Capturing SO- from Flue Gas," Summary report to NAPCA, June 1969.
21. Young, D. A. "Decomposition of Solids," Pergamon Press, London,
1966.
22. Riesenfeld, E. H. J. Prakt. Chem. 100, 142, (1920).
23. Rosenquist, T. J. Metals Trans. J^, 535 (1951).
24. Squires, A. M. in "Fuel Gasification," Advances in Chemistry Series
No. 69, p. 205, American Chemical Society, Washington, D. C. 1967.
25. Curran, G. P., B. Pasek, M. Pell, E. Gorin, Monthly Progress Reports,
EPA Contract No. EHSD-71-15, 1972. Consolidation Coal Company.
26. G. J. Vogel, E, Carls, A. A. Jonke, et al. Proc. 3rd Int. Conf. on
Fluidized Bed Combustion, Nov. 1972.
340
-------
5. PLANT OPERATION AND CONTROL
1. "Evaluation of the Fluidized Bed Combustion Process," Vol. I, II,
and III, submitted to Office of Air Programs, Environmental Protection
Agency, by Westinghouse Research Laboratories, Pittsburgh, Pennsylvania,
(1971).
2. Keairns, D. L., W. C. Yang, J. R. Hamm, D. H. Archer. "Fluidized Bed
Combustion Utility Power Plants - Effect of Operating and Design
Parameters on Performance and Economics," Proceedings of Third Inter-
national Conference on Fluidized Bed Combustion, Hueston Woods, Ohio,
(1972).
341
-------
6. ALTERNATIVE PRESSURIZED FLUID BED BOILER CONCEPTS
1. "Evaluation of the Fluidized Bed Combustion Process," Vol. I, II,
and III, submitted to Office of Air Programs, Environmental Protection
Agency, by Westinghouse Research Laboratories, Pittsburgh, Pennsylvania,
(1971).
2. Keairns, D. L., W. C. Yang, J. R. Hamm, D. H. Archer, "Fluidized Bed
Combustion Utility Power Plants - Effect of Operating and Design
Parameters on Performance and Economics." Proceedings of Third Inter-
national Conference on Fluidized Bed Combustion, Hueston Woods, Ohio,
(1972).
3. ANL/ES/CEN-F026, Monthly Progress Report #26, December 1970, Argonne
National Laboratory, Chemical Engineering Division.
4. Horsier, A. G., Lacey, J. A., and Thompson, B. H. Chem. Eng. Prog.,
j>5 (10), 59 (1969).
5. Dent, F. J. "Methane From Coal." 9th Coal Science Lecture, BCURA (1960),
6. Horsier, A. G., and B. H. Thompson, "Fluidization in the Development
of Gas Making Processes." Tripartite Chemical Engineering Conference,
Montreal, Canada (1968).
7. Decamps, F., G. Dumont and W. Goossens, "Vertical Pneumatic Conveyer
with a Fluidized Bed as Mixing Zone," Powder Technol., _5 299 (1971/72).
8. Buchanan, R. H. and B. Wilson, "The Fluid-Lift Solids Recirculator,"
Mech. & Chem. Eng. Trans. (Australia), 117, May (1965).
9. Curran, G. P. and Gorin, E. "Studies on Mechanics of Flow Solids
Systems." Report prepared for Office of Coal Research by Consolidation
Coal Company (1968).
10. Taskaev, N. D. and Kozhina, M. I. Trudy Akad. Nank Kirg. 2 S.S.R.,
7, 109 (1956).
343
-------
11. Archer, D. H., Vldt, E. J., Keairns, D. L., Morris, J. P. and
Chen, J. L. P. "Coal Gasification for Clean Power Production."
Proceedings of the Third International Conference on Fluidized Bed
Combustion, Hueston Woods, Ohio, October 1972.
12. Yang, W. C. and Keairns, D. L. "Recirculating Fluidized Bed Reactor
Data Utilizing a Two-Dimensional Cold Model." Paper presented at
the 75th National Meeting, AIChE, Detroit, June 3-6, 1973.
13. Botterill, J. S. M., Butt, M. H. D., Cain, G. L. and Redish, K. A.
Proceedings of the International Symposium on Fluidization.
Netherlands University Press, Amsterdam, 1967.
14. Botterill, J. S. M., Butt, M. H. D., Cain, G. L., Chandrasekhar, R.
and Williams, J. R. Proceedings of the International Symposium on
Fluidization, Netherlands University, Amsterdam, 1967.
15. Botterill, J. S.M., Chandrasekhar, R. and Van der Kolk, M. Brit.
Chem. Eng. 15_ (6), 769 (1970).
16. Botterill, J. S. M. Powder Tech. _4, 19 (1970/71).
17. Botterill, J. S. M., Chandrasekhar, R. and Van der Kolk, M. Chem.
Eng. Prog. Symp. Series 66, No. 101 (1970).
18. Mathur, K. B., Chap. 17, Fluidization, edited by J. F. Davidson
and D. Harrison, Academic Press (1971).
19. Zenz, F. A. and Othmer, D. F. Fluidization and Fluid-Particle
Systems. Reinhold Publishing Corporation, N. Y. (1960).
344
-------
5.2 UNIT START-UP
5.2.1 Introduction
Plant start-up, as well as large-scale load increases, requires
placing different parts of the total plant in operation. The normal
start-up of the plant is sequential; that is, the plant components are
placed in operation in a definite order as follows:
• Gas turbine
• Boiler module
• Steam turbine
• Boiler module
• Gas turbine
• Boiler module
• Boiler module
This sequence consists of three different start-up
procedures:
• Gas turbine, first boiler module, and steam turbine
• The second and fourth boiler module start-up. These two
are the same.
• The second gas turbine and third boiler module.
Figure 93, steam and water cycle, and Figure 94, chemical
recovery plant, are process schematics showing the control valves referred
to in this section and their functional location in the plant.
Figures 95A, 95B, 96A, and 96B show in a simplified way the steps
required to start up the first and second boiler modules.
The start-up procedures that follow are based on sequentially
starting the entire plant from a cold condition.
221
-------
I^LIOGRAPHIC DATA '• K°port No. 2-
HEET EPA-650/2-73-048a thru -d
i4,u .,n,i suinuir Pressurized Fluidized-Bed Combustion Process
evelopment and Evaluation, Volumes I, II , HI, and IV
fol II, Appendices; Vol HI, Boiler Development Plant Design;
ol IV Oil Gasification/Dosulfurization)
% uthor(s')
. L. Keairns, D. H. Archer et al.
Performing Organisation N.imc and Address
"estinghouse Research Laboratories
ittsburgh, Pennsylvania 15235
Sponsoring Organi/ation Name and Address
PA , Office of Research and Development
ERC-RTP, Control Systems Laboratory
esearch Triangle Park, North Carolina 27711
3. Rmbined-cycle power generation. It identifies no problems which preclude the
jvelopment of pressurized FBC combined-cycle power plants and FB oil gasification
)wer plants which can generate electrical energy within environmental goals at lower
lergy costs than competitive systems. Work reported here, a continuation of earlier
3C process evaluation efforts , is aimed at the development and demonstration of
ese FB fuel processing systems.
Key Words and Oocument Analysis. 17o. IVscriptors
r Pollution
Dmbustion
isification
.uidized Bed Processing
jsulfurization
Is
>ssil Fuels
istes
ectric Power Generation
Identifirrs/Open-Ended Terms
r Pollution Control
itionary Sources
uidized-Bed Combustion
C-OSATl Held/C.roup
? 21B
vailability Statement
Unlimited
19. Security Class (This
Report)
UNCLASSIFIED
20. Security Class (This
Page
UNCLASSIFIED
21. No. of Pages
380
22. Price
N TIS-35 (REV. 3-72)
345
USCOMM-DC I4952-P72
------- |