v>EPA
United States Industrial Environmental Research EPA-600/7-80-015d
Environmental Protection Laboratory January 1980
Agency Research Triangle Park NC 27711
Experimental/ -
Engineering Support
for EPA's FBC Program:
Final Report
Volume IV. Engineering
Studies
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-80-015d
January 1980
Experimental/Engineering Support
for EPA's FBC Program:
Final Report -
Volume IV. Engineering Studies
by
J.R. Hamm, D.F. Ciliberti, R.W. Wolfe,
R.A. Newby, and D.L Keairns
Westinghouse Research and Development Center
1310 Beulah Road
Pittsburgh, Pennsylvania 15235
Contract No. 68-02-2132
Program Element No. INE825
EPA Project Officer: D. Bruce Henschel
Industrial Environmental Research Laboratory
Office of Environmental Engineering and Technology
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
PREFACE
The Westinghouse R&D Center Is carrying out a program to provide
experimental and engineering support for the development of fluidized-
bed combustion (FBC) systems under contract to the Industrial Environ-
mental Research Laboratory (IERL), U. S. Environmental Protection Agency
(EPA), at Research Triangle Park, NC. The contract scope includes atmo-
spheric (AFBC) and pressurized (PFBC) fluidized-bed combustion processes
as they may be applied for steam generation, electric power generation,
or process heat. Specific tasks include work on calcium-based sulfur
removal systems (e.g., sorption kinetics, regeneration, attrition, mod-
eling), alternative sulfur sorbents, nitrogen oxide (NOX) emission, par-
ticulate emission and" control, trace element emission and control, spent
sorbent and ash disposal, and systems evaluation (e.g., impact of new
source performance standards (NSPS) on FBC system design and cost).
This report contains the results of work defined and completed
under technical directives issued by the EPA project officer. Work on
these tasks was performed from January 1976 to January 1979 and is docu-
mented in the following EPA contract reports:
• The present report, which presents the results of four
technical directives on systems evaluation
• Report on an engineering assessment of intimate coal/
sorbent mixtures for S02 control in FBC applications which
is reported in our 1978 EPA report, EPA-600/7-78-0051
• Report on the "Effect of SC>2 Emission Requirements on
Fluidized-Bed Combustion Systems: Preliminary Technical/
Economic Assessment," issued in August 1978 (EPA-600/7-78-
163, NTIS PB 286 871/7ST).2
iii
-------
Work on the other tasks performed under this contract is reported
in:
• Experimental/Engineering Support for EPA's FBC Program:
Final Report Volume 1, Sulfur Oxide Control, EPA-
600/7-80-015a, January 1980
• Experimental/Engineering Support for EPA's FBC Program:
Final Report Volume II, Particulate, Nitrogen Oxide, and
Trace Element Control, EPA-600/7-80-015b, January 1980
• Experimental/Engineering Support for EPA's FBC Program:
Final Report Volume III, Solid Residue Study, EPA-
600/7-80-015c, January 1980
• Alternatives to Calcium-Based S02 Sorbents for Fluidized-
Bed Combustion: Conceptual Evaluation, EPA-600/7-78-005,
January 1978
• Regeneration of Calcium-Based S02 Sorbents for Fluidized-
Bed Combustion: Engineering Evaluation, EPA-600/7-78-039,
NTIS PB 218-317, March 1978
• Disposal of Solid Residue from Fluidized-Bed Combustion:
Engineering and Laboratory Studies, EPA-600/7-78-049 (NTIS
PB 283-082), issued in March 1978, which presented the
results of work performed from January 1976 to January
1977
• Evaluation of Trace Element Release from Fluidized-Bed
Combustion Systems, EPA-600/7-78-050, NTIS PB 281-321,
March 1978.
iv
-------
ABSTRACT
Engineering studies addressing several aspects of fluidized-bed
combustion (FBC) system design and performance are reported. An evalua-
tion on the impact of SC>2 emission requirements on FBC system perfor-
mance and cost is reviewed (EPA-600/7-78-163). Stringent SC>2 emission
requirements can be satisfied economically if proper selection of design
and operating parameters is made. Another study on the feasibility of
feeding coal/sorbent mixtures to FBC units is also reviewed (EPA-
600/7-78-005). Critical data gaps exist for this concept. Moreover,
general economic feasibility would not be expected. An alternative S02
control concept for pressurized fluidized-bed combustion (PFBC), that is,
pressurized scrubbing of the products of combustion with water, is evaluated.
The concept is not economically competitive because of the requirement for
recuperative heating and reduced plant efficiency. A potential reduc-
tion in solid waste is realized with the concept, but„the SC>2 control
efficiency may be limited.
An evaluation of PFBC examining the technical and economic trade-
offs between the level of particulate control achieved and the frequency
of gas-turbine blade replacement is described. The evaluation incorpor-
ates models of PFBC particulate carry-over, particulate control device
efficiency, and turbine erosion. Also, an indirect air-cooled PFBC con-
cept is evaluated and compared with other PFBC concepts. The indirect
air-cooled concept provides significant particulate control advantages
over the adiabatic coiabustor PFBC concept, while resulting in about a
4 percent lower plant efficiency and a 1 percent higher cost of
electricity.
-------
TABLE OF CONTENTS
Page
1. INTRODUCTION 1
2. CONCLUSIONS 2
Effect of Emission Requirements on FBC Systems 2
Intimate Coal/Sorbent Mixtures for S02 Control 2
Feasibility Evaluation of Fluidized-Bed Combustion Using
Pressurized-Water Scrubbing 2
Particulate Control Trade-off for PFBC Systems 3
Indirect Air-Cooled Fluidized-Bed Combustion Concept
Systems Evaluation 3
3. RECOMMENDATIONS 4
4. SULFUR OXIDE CONTROL 6
5. INTIMATE COAL/SORBENT MIXTURES FOR S02 CONTROL IN
FLUIDIZED-BED COMBUSTION 12
6. FEASIBILITY EVALUATION OF FLUIDIZED-BED COMBUSTION
USING PRESSURIZED-WATER SCRUBBING 14
Introduction 14
Concept and Process Options 14
Selection of Base Case Design Concept 18
Plant Basis 20
Material and Energy Balances 21
Equipment Design 24
Capital Investment 26
Cost of Electricity 31
Environmental Comparison 31
Conclusions 35
7. PARTICULATE CONTROL TRADE-OFF FOR PFBC SYSTEMS 37
Overview 37
Background 38
Estimation of Particle Loading/Size Distribution 39
Estimating Gas Turbine Impact 48
Results 59
vi
-------
TABLE OF CONTENTS (Continued)
Page
8. INDIRECT AIR-COOLED PRESSURIZED FLUIDIZED-BED COMBUSTION
CONCEPT SYSTEMS EVALUATION 63
Introduction 63
Background 63
Results of Study 70
Particulate Control/Gas Turbine Expander Erosion
Considerations 79
Environmental Considerations 84
Conclusions 87
9. REFERENCES 89
APPENDIX
A. GRADE EFFICIENCIES FOR PARTICULATE REMOVAL EQUIPMENT 92
B. PARTICLE SIZE DISTRIBUTION AT PERTINENT STATIONS IN
PARTICULATE REMOVAL SUBSYSTEM FOR ALTERNATIVE CASE I 96
vii
-------
LIST OF FIGURES
Page
1. Concept and Process Options 15
2. Material and Energy Balances 22
3. Diagram of Particulate Removal System 39
4. Particulate Removal Equipment Arrangement 39
5. Grade Efficiency of Primary Cyclone Separator 42
6. Grade Efficiency Curve for Tan Jet 42
7. Schematic of Tan Jet 43
8. Granular-Bed Filter Module 43
9. Various Filter Performance Assumed for Final Cleanup
Stage 45
10. Particulate Concentration at Outlet of First-Stage
Granular-Bed Filter 48
11. Particulate Concentration at Outlet of Second-Stage
Granular-Bed Filter 48
12. Projected Outlet Size Distribution Based on Rexnord
Commercial Unit (Dolomite Particles) 49
13. Projected Outlet Based on Rexnord Commercial Unit (Ash
Particles) 49
14. Projected Outlet Based on Rexnord Commercial Unit (Char
Particles) 50
15. Projected Granular-Bed Filter Outlet Size Distribution
Based on Westinghouse Bench-Scale Experiments (Dolomite
Particles) 50
16. Projected Granular-Bed Filter Outlet Based on Westinghouse
Bench-Scale Experiments (Ash Particles) 51
17. Projected Granular-Bed Filter Outlet Based on Westinghouse
Bench-Scale Experiments (Char Particles) 51
18. Projected Granular-Bed Filter Outlet Size Distribution
Based on Conventional Fabric-Filter Unit Performance
(Dolomite Particles) 52
19. Projected Granular-Bed Filter Outlet Based on
Conventional Fabric-Filter Unit Performance
(Ash Particles) 52
viii
-------
LIST OF FIGURES (Continued)
20. Projected Granular-Bed Filter Outlet Based on
Conventional Fabric-Filter Unit Performance
(Char Particles) 53
21. Blade Leading Edge Erosion Rates 53
22. Projected Turbine Life for a Particulate Removal
System with Two Stages of Granular-Bed Filters 57
23. Projected Turbine Life for a Particulate Removal
System with One Stage of Granular-Bed Filters 57
24. Projected Turbine Life for a Particulate Removal
System with One Stage of Granular-Bed Filters 58
25. Cost of Electricity Increments due to Turbine Blade
Replacement Using Granular-Bed Filters (performance
based on granular-bed filter efficiency) 61
26. Cost of Electricity Increments due to Turbine Blade
Replacement Using Granular-Bed Filters (performance
based on fabric-filter efficiency) 61
27. Combined-Cycle System Utilizing Fluidized-Bed Combustion
with Indirect Heating of Part of the Working Fluid
(no CBC) 66
28. Combined-Cycle System Utilizing Fluidized-Bed Combustion
with Indirect Heating of Part of the Working Fluid
(Alternative Case I - CBC) 68
29. Combined-Cycle System Utilizing Fluidized-Bed Combustion
with Indirect Heating of Part of the Working Fluid
(Alternative Case II - no CBC) 70
30. Plot Plan of Single Gas Turbine Module for Base Case 72
31. Plot Plan of Single Gas Turbine Module for Alternative
Case I 72
32. Plot Plan of Single Gas Turbine Module for Alternative
Case II 73
33. Summary of Particulate Loading and Size Distribution for
Base Case 82
34. Summary of Particulate Loading and Size Distribution for
Alternative Case I 83
35. Summary of Particulate Loading and Size Distribution for
Alternative Case II 85
ix
-------
LIST OF TABLES
1. Material and Energy Balances 23
2. Gas-Cleaning Auxiliaries 24
3. Process Equipment 25
4. Capital Investment for Pressurized Water Scrubbing 27
5. PFBC Boiler Plant Equipment Costs 28
6. PFBC Power Plant Cost Breakdown 29
7. Comparison of Cost of Electricity 32
8. Environmental Comparison 34
9. Projected Particulate Concentration Levels 46
10. Particulate Emission Levels 47
11. Thermal History of Particles Entering a 0.05-in. Thick
Boundary Layer 55
12. Summary of Plant Performance 71
13. Summary of Plant Design Configurations 74
14. Capital Cost Estimate for Base Case 75
15. Capital Cost Estimate for Alternative Case I 76
16. Capital Cost Estimate for Alternative Case II 77
17. Specific Cost Comparison 78
18. Cost of Electricity Summary 78
19. Expander Inlet Particle Loading 81
20. Environmental Comparison of the Configuration 86
-------
NOMENCLATURE
AFBC - atmospheric-pressure fluidized-bed combustion
Ca/S - calcium-to-sulfur ratio
CBC - carbon burnup cell
CF - capacity factor
COE - cost of electricity
DOE - Department of Energy
EGAS - Energy Conversion Alternatives Study
EPA - Environmental Protection Agency
ERDA - Energy Research and Development Agency
FBC - fluidized-bed combustion
GBF - granular-bed filter
HHY -
HRSG - heat recovery steam generator
IERL - Industrial Environmental Research Laboratory
ISO - International Standards Organization
NOX - nitrogen oxide
NSPS - New Source Performance Standard
O&M - operating and maintenance
PFBC - pressurized fluidized-bed combustion
SOX - sulfur oxide
TDC - total direct costs
TGA - therraogravimetrie analysis
xi
-------
ACKNOWLEDGMENT
We want to express our high regard for and acknowledge the
contribution of Mr. D. B. Henschel who served as the EPA project officer.
Mr. P. P. Turner and Mr. R. P. Hangebrauck, Industrial Environmental
Research Laboratory, EPA, are acknowledged for their continuing contri-
butions through discussions and support of the program.
We gratefully acknowledge the contributions of the following
Westinghouse personnel: K. D. Weeks for his assistance in evaluating
the air-cooled PFBC cycle and Dr. D. H. Archer, Manager, Chemical Engineering
Research, for his program consultation and continued support.
xii
-------
1. INTRODUCTION
This volume documents five systems evaluation tasks that were per-
formed during the contract as technical directives from the EPA proiect
officer. One study, the effect of sulfur dioxide (SO ) emission require-
ments on fluidized-bed combustion (FBC) systems, was issued as a sepa-
rate report in 1978 (EPA-600/7-78-163). Another study, an engineering
assessment of intimate coal/sorbent mixtures for S02 control by FBC
applications, was also issued in 1978 (EPA-600/7-78-005). Results from
three studies, the feasibility evaluation of pressurized-water scrubbing
for S02 emission control with PFBC (completed in 1977), a particulate
control/turbine life trade-off study for PFBC systems (completed in
1977), and an evaluation of indirect air-cooled PFBC concepts (completed
in 1978) were not previously issued as separate contract reports.
A summary of the sulfur oxide (SOX) control report is presented in
Section 4, the intimate coal/sorbent mixture study is summarized in Sec-
tion 5, and results of the remaining three studies are reported in Sec-
tions 6 through 8.
-------
2. CONCLUSIONS
The primary conclusions from the five studies follow.
EFFECT OF EMISSION REQUIREMENTS ON FBC SYSTEMS
• AFBC and PFBC systems can economically meet the New Source
Performance Standards (NSPS) for utility power plants:
90 percent sulfur removal, 12.9 ng/J (0.03 Ib/MBtu) par-
ticulate emission, and 258 ng/J (0.6 Ib/MBtu) nitrogen
oxide (NOX) emission.
• The selection of FBC design and operating parameters to
minimize the sorbent feed requirement is critical for
realizing economical systems.
INTIMATE COAL/SORBENT MIXTURES FOR S02 CONTROL
• Sufficient technical data on the performance of intimate
coal/sorbent mixtures, such as pellets consisting of coal
and limestone powders, do not exist to project FBC perfor-
mance reliably.
• The performance of intimate coal/sorbent mixtures in FBC
is unlikely to be economically competitive with conven-
tional FBC concepts.
FEASIBILITY EVALUATION OF FLUIDIZED-BED COMBUSTION USING PRESSURIZED-
WATER SCRUBBING
• A PFBC plant using the cold pressurized-water scrubbing
concept for S02 control is not economically competitive
with calcium-based PFBC or conventional steam power plants
with stack-gas cleaning.
-------
PARTICULATE CONTROL TRADE-OFF FOR PFBC SYSTEMS
• A methodology has been developed and is available for
evaluating trade-offs between fluid-bed corabustor, gas-
cleaning, and turbine design and operating parameters.
INDIRECT AIR-COOLED FLUIDIZED-BED COMBUSTION CONCEPT SYSTEMS EVALUATION
• Indirect air-cooled PFBC concepts will have
- Lower performance (higher heat rates) than PFBC boiler
concepts
- Similar performance with PFBC adiabatic combustor
concepts.
• The cost of electricity for indirect air-cooled PFBC con-
cepts is essentially the same as for PFBC adiabatic cora-
bustor concepts.
• Indirect air-cooled AFBC concepts will have lower perfor-
mance (higher heat rates) than indirect air-cooled PFBC
concepts. The AFBC concept provides for turbine reliabil-
ity using a clean gas.
• Environmental emissions standards can be met with all
indirect air-cooled FBC concepts.
-------
3. RECOMMENDATIONS
The primary recommendations from work carried out under the techni-
cal directives are presented here, followed by recommendations for
extended systems studies to evaluate and guide the development of eco-
nomical FBC systems operating within environmental constraints.
• Investigate the ability of AFBC and PFBC processes to
achieve more stringent emission standards, with specific
focus on the relationship of performance to combustor
design and operating parameters.
• Evaluate and develop advanced FBC sulfur removal concepts
(e.g., sorbent pretreatment, sorbent regeneration, alter-
native regenerable sorbents). This will be particularly
important when solids procurement or disposal represents a
constraint.
• Develop understanding of NOX minimization alternatives,
perform system evaluation to select economic options, and
demonstrate capability.
• Carry out experimental test programs to obtain performance
data on particulate control equipment applicable to AFBC
and FFBC systems. The primary need is an understanding of
high-temperature, high-pressure particulate control equip-
ment performance - e.g., cyclones, granular-bed filters,
fabric filters, and other advanced filter and cyclonic
concepts. Data will be important for process constraints
(e.g., turbine erosion/deposition) and environmental con-
straints (leading and potential fine particle emission
criteria).
-------
• Carry out experimental test programs to obtain understand-
ing of turbine tolerance - e.g., erosiveness of particu-
late, response characteristics of turbine materials, effect
of turbine design and operating parameters.
• Extend and apply methodology for evaluating fluid-bed
combustor/gas-cleaning/turbine design and operating param-
eter trade-offs to identify optimal fluidized-bed combus-
tor systems for given application and environmental
requirements.
Additional FBC systems studies are recommended to
• Assess the effects of potential NSPS on industrial FBC
systems to aid EPA in developing standards.
• Project and evaluate the environmental performance of FBC
system designs currently proposed by Department of Energy
(DOE) contractors or commercial vendors to understand the
status of these designs.
• Evaluate the impact of variable coal sulfur content and
variable sorbent properties on the control of S02 emis-
sions from FBC systems in order to quantify the effect of
variable properties on sorbent consumption and system
economics.
• Evaluate FBC unit start-up and turndown techniques with
respect to environmental performance in order to identify
superior techniques and performance sensitivity.
• Assess the technical/environmental performance of alterna-
tive FBC operating regimes (e.g., turbulent fluidization,
circulating fluidized bed, fast fluidization, multtsolids
systems) to understand their potential and limitations.
-------
4. SULFUR OXIDE CONTROL
Westinghouse evaluated the impact of up to 90 percent sulfur
removal on the capital and energy costs of conventional dense-phase,
fluid-bed, AFBC and PFBC power plants in a previous report as part of
this contract. A brief summary of that report is presented here. The
full study is presented in EPA-600/7-78-163.
Two levels of emissions standards were considered:
• The current (1978) EPA NSPS for large coal-fired boilers:
SOX, 516 ng/J (1.2 Ib S02 MBtu); particulates, 43.0 ng/J
(0.1 Ib/MBtu); and NOX, 301 ng/J (0.7 Ib N02 MBtu)
• A set of more stringent degrees of control: SOX> 90 per-
cent removal of coal sulfur content; particulates,
12.9 ng/J (0.03 Ib/MBtu); and NOX, 258 ng/J
(0.6 Ib/MBtu).
These levels were selected for the study because they represent one set
of values considered during the planned revision of the NSPS for utility
boilers.
Projections of AFBC and PFBC power plant performance and economics
have been developed through the assimilation of previous FBC power plant
design studies, FBC performance models, and data assessments. The key
parameters in the evaluation are the sorbent Ca/S ratio, the coal sulfur
content, and the fluid-bed combustor design and operating conditions.
The projections of FBC power plant energy costs indicate that for
both the existing SOX emission standard and for 90 percent sulfur
-------
removal FBC is potentially cost competitive with conventional coal-fired
power plants using lime-slurry scrubbing. The competitiveness of FBC
depends upon proper selection of fluid-bed combustor operating condi-
tions—i.e., sufficiently long gas residence in the bed (sufficiently
low gas velocity and sufficiently deep beds) and sufficiently small sor-
bent particle size. This selection of variables will result in a larger
combustor, but the cost savings resulting from decreased sorbent
requirements would more than compensate for increased combustor costs.
In the design of FBC power plants one should emphasize maximization
of fluid-bed combustor performance rather than minimization of the com-
bustor cost through compact design. The corabustor cost represents a
small portion of the FBC power plant investment and is also relatively
insensitive to changes in design and operating conditions. On the other
hand the overall FBC power plant cost of electricity is strongly depen-
dent on the combustor performance.
The Ca/S molar ratio—that is, the moles of sorbent calcium fed to
the fluid-bed combustor divided by the moles of sulfur fed in the coal—
is the single, most important performance factor relative to FBC power
plant cost and performance for high-sulfur eastern coals (2 to 5 wt %
sulfur). The Ca/S ratio has a dramatic impact on the FBC power plant
thermal efficiency, capital investment, and cost of electricity. An
increased Ca/S ratio, if required for lower SOX emissions, results in
increased auxiliary power consumption for solids handling and signifi-
cant sorbent calcination energy losses. The resulting reduced net plant
efficiency and slightly increased equipment costs for solids handling,
crushing, drying, feeding, and spent solids disposal lead to increased
capital investment and energy costs. In addition, the increased cost of
raw sorbent at increased feed rates significantly increases the energy
cost.
-------
The all-important projection of sorbent feed requirements was
accomplished in this study by using a kinetic model for SOX capture
Westinghouse had developed. This model—using rate constants measured
in laboratory thermogravimetric analysis (TGA) equipment and confirmed
where possible by using available data from experimental fluidized-bed
combustors—is capable of projecting sorbent requirements, where TGA
data have been generated, as a function of key corabustor operating/
design conditions.
While the cost and performance of several subsystems in the FBC
power plants are uncertain (for example, solids feeding and particulate
control), these are expected to be resolved through proper design and
specification of materials and operating conditions and maintenance and
operating procedures. The overall financial impact of these cost/ per-
formance uncertainties will probably be small relative to the uncertain-
ties in such site factors as sorbent availability, sorbent cost, coal
cost, solid waste disposal feasibility or utilization markets, local
emission standards, and so on.
For low-sulfur western coals and lignites the impact of an
increased Ca/S ratio is greatly reduced because of the relatively small
quantities of sorbent involved. Uncertainties associated with sorbent
selection and cost are also less significant.
Projections of particulate control and emissions of NOX for FBC
power plants indicate that the more stringent emission requirements con-
sidered here of 12.9 ng/J (0.03 Ib/MBtu) and 285 ng/J (0.6 Ib/MBtu),
respectively, are economically feasible and of lower cost impact than
the more stringent SOX requirement. Conventional fabric-filter (bag-
house) techniques should permit achievement of this requirement, depend-
ing on particle size and future environmental standards. We expect PFBC
plants to require two stages of particulate control equipment operating
at the combustor temperature and pressure: e.g., conventional cyclones
-------
followed by a filter system. Nitrogen oxide levels from the assessment
of FBC experimental results have been shown to be generally lower than
258 ng/J (0.6 Ib/MBtu) without special control efforts.
On the basis of available Information, the projections developed
indicate that both AFBC and PFBC should be able to achieve the higher
levels of control considered in this evaluation economically if proper
combustor design and operating conditions are selected. Development
programs should focus on developing large-scale information on the rela-
tionship between corabustor operating conditions and FBC plant emissions,
while engineering evaluation should assess FBC pollution control
capabilities.
The detailed conclusions and recommendations developed in this
report are as follows:
• On the basis of available information, the more stringent
emission requirements considered in this study (SOX, 90%
sulfur removal; particulates, 12.9 ng/J (0.03 Ib/MBtu);
NOX, 258 ng/J (0.6 Ib N02/MBtu) should be economically
achievable for both AFBC and PFBC power plants.
• The proper selection of fluid-bed corabustor design and
operating conditions is critical to the economical reali-
zation of these environmental goals. The gas residence in
the bed, in particular (as determined by gas velocity and
bed height), should be sufficiently long, and sorbent par-
ticle size should be sufficiently small. In this assess-
ment residence of 0.67 to 2.0 s (gas velocities of 1.5 to
1.8 m/s) and particle sizes averaging 500 um appeared to
offer effective SOX removal performance, although these
conditions are not necessarily optimal.
• The high level of SOX emission control considered has a
greater Impact on the FBC power plant energy cost than do
-------
the revised particulate and NOX standards considered. The
most critical process parameter with respect to FBC power
plant cost and performance Is the Ca/S ratio.
• The fluid-bed combustor cost does not depend strongly on
changes In design and operating conditions. The fluid-bed
corabustor should be designed to minimize the cost of plant
energy rather than cost of the combustor. For example,
low - rather than high - fluldlzatlon velocities will
probably result in lower FBC power plant energy cost.
• Particulate control to levels as low as 12.9 ng/J
(0.03 Ib/MBtu) should be economically achievable for AFBC
using commercially available techniques. Baghouses seem
most suitable for this duty. No testing of any type of
final-stage particle control device on an AFBC unit, how-
ever, has yet been conducted.
• Particulate control to levels below 12.9 ng/J (0.03 lb/
MBtu) may be dictated for PFBC by turbine protection
requirements, depending on particle size. Projections
indicate that 0.03 Ib/MBtu should be achievable, but the
technology to meet this control at high temperature and
pressure has not yet been demonstrated.
• Oxides of nitrogen will generally be emitted by FBC at
levels below the 258 ng/J (0.6 lb N02/MBtu) requirement
considered in this evaluation. No direct control tech-
niques for NOX have been clearly demonstrated on
fluidized-bed combustors to date, although several options
are under study.
• The greatest FBC power plant uncertainties presently
involve reliability questions - e.g., solids feeding, par-
ticulate control (especially for PFBC), material erosion/
corrosion/deposition, and process control. The impact of
emission standards averaging time basis and system reli-
ability has not been evaluated.
10
-------
• AFBC and PFBC development programs should focus on more
stringent emission standards and their relation to combus-
tor design and operating conditions.
• Advanced FBC sulfur removal concepts, for example, sorbent
precalcination, sorbent regeneration, sorbent fines recon-
stitution, additives for improved sorbent utilization,
alternative metal oxide sorbents, should be evaluated with
respect to more stringent emission standards.
11
-------
5. INTIMATE COAL/SORBENT MIXTURES FOR SC>2 CONTROL
IN FLUIDIZED-BED COMBUSTION
Westinghouse performed a conceptual evaluation of the use of inti-
mate coal/sorbent mixtures (e.g. , pellets consisting of powdered coal
and limestone) as part of this contract. This evaluation has been
reported previously, in EPA-600/7-78-005, and a brief summary of that
report is included here.
The study was carried out to investigate the technical and environ-
mental feasibility and economic potential of "intimate coal/sorbent mix-
tures" when used in an FBC system for power generation. Various classes
of intimate coal/sorbent mixtures were first qualitatively screened for
feasibility on the basis of their probable performance assessment.
Intimate coal/sorbent mixtures selected as potentially feasible in the
initial screening were then subjected to an engineering assessment of
technical and environmental performance. Areas such as SOX and NOX con-
trol, trace metal and particulate control, solid waste and plant effi-
ciency, and design factors for the fluidized-bed combustor were consid-
ered. Because no actual performance or kinetic data exist for the inti-
mate coal/sorbent mixtures, only potential performance could be
addressed and problem areas identified.
Economic potential was examined by using optimistic performance
assumptions for the intimate coal/sorbent mixture. Process alternatives
for the preparation of the mixtures were identified and cost projections
for the preparation systems were generated.
12
-------
The major conclusions reached are as follows:
• The only technically feasible intimate coal/sorbent mix-
ture that could be identified for the current fluidized-
bed combustion design concept is the consolidated coal/
sorbent particle concept.
• Attrition of the consolidated particle is the most criti-
cal factor influencing the performance and feasibility of
the concept. Modifications to the corabustor design would
probably be required in order to apply the consolidated
particle concept.
• The performance (technical and environmental) cannot be
estimated without initiating a test program. The overall
technical and environmental performance of the consoli-
dated particle concept could conceivably by worse than or
better than the conventional fluid-bed combustor, but it
is highly unlikely that any significant improvement in
performance is to be realized.
• Except under very extreme conditions, the consolidated
particle concept will not be economically competitive with
conventional FBC concepts.
• Washing the pulverized coal during consolidated particle
preparation could reduce trace elements, ash, sulfur, and
the sorbent requirement. The economics of this option
have not been investigated.
• The most attractive consolidated coal/sorbent particle
from the standpoint of technical and environmental impact
would utilize a binder to maintain the coal-ash and sor-
bent particles in discrete, consolidated particles follow-
ing combustion. A binder that will effect this behavior
has not been identified.
13
-------
6. FEASIBILITY EVALUATION OF FLUIDIZED-BED COMBUSTION
USING PRESSURIZED-WATER SCRUBBING
INTRODUCTION
The national and private development efforts for fluidtzed-bed com-
bustion are based on SOX absorption by calcium-baaed sorbents (limestone
or dolomite) at high temperatures. Both AFBC and PFBC concepts are
being pursued with either regenerative or once-through sorbent opera-
tion. We believe that once-through sorbent operation represents only
the first-generation of FBC systems, but even with sorbent regeneration,
if it is eventually realized, FBC will produce significant quantities of
dry, granular, sulfated limestone or dolomite that must be disposed of
or utilized in an environmentally satisfactory manner.
We have evaluated an alternative FBC concept that may be applicable
to PFBC. This concept is a cold gas-cleaning scheme that uses water to
scrub the pressurized combustion products without additives for control-
ling S02- The potential advantage of this alternative is the reduction
of solid-waste emissions.
A feasibility study has been conducted to better define the concept
and to estimate its cost and performance. Approximate material and
energy balances, conceptual equipment designs, and process economic
estimates were performed in order to determine concept problem areas and
process economic and environmental feasibility.
CONCEPT AND PROCESS OPTIONS
The basic PFBC concept with pressurized water scrubbing is shown in
Figure 1 with various process options indicated. An understanding of
these options Is Important if one is to select the best process to be
evaluated.
14
-------
To
Gas Turbine
Absorber
Gas (Temperature)
Combustion
Products
Part icu late
Control
(Type, Number
of Stages)
Coal
Dwg. 1687B33
Tail Gas «
(Recycle, Exhaust)
Pressurized
Combust or
(Pressure,
Temperature,
Ash or Inert
Bed)
Particulate
Control
(Type, Number
of Stages)
Air
(excess air rate)
Sulfur or*
Acid
Sulfur or
Sulfuric Acid
Plant
Stripper
Gas
Makeup
Water
Pressurized
Absorber
(Type
internal
heat
transfer,
temper-
ature)
Pressure ^k
Reduction-*]
Stripping Gas
(Air, Steam,
Stack Gas)
Filter Cake
Figure 1 - Concept and Process Options
-------
Coal is combusted with air in the pressurized fluidized-bed corabus-
tor. The bed consists of either coal ash or an inert bed material such
as alumina. The bed temperature (760-1040°C) and pressure (620-
1600 kPa) are important process variables relating to the corabustor per-
formance and the cycle efficiency. The excess air rate, also, is a
critical process variable since it defines the quantity of gas that must
be handled by the pressurized water scrubbing system. The excess air
may range from 10 to 100 percent for fluidized-bed boilers and may be
about 300 percent for an adiabatic fluidized-bed corabustor (no heat
transfer surface in the bed). The corabustor fluidization velocity and
heat transfer rates are assumed to be very similar to those of the
calcium-based corabustor, as are the attrition and elutriation rates,
although they could be lower with proper selection of the inert bed
material.
High-temperature particulate removal equipment (cyclones, filters)
could be situated so as to operate before the combustion products are
cooled, and/or low-temperature removal equipment (filters, scrubbers,
electrostatic precipitators) could be placed to operate after the cool-
ing step. Captured bed material (coal ash, inert material) could be
recycled to the corabustor or removed from the system.
Cooling and reheating the combustion products would be a critical
step. A recuperator or a convection-type steam generator followed by a
recuperator would cool the combustion products to a level suitable for
the absorber (<150°C) and would reheat the absorber gas to a temperature
resulting in an economical, combined-power cycle. Various types of
recuperators could be used: a shell-and-tube heat exchanger constructed
from high-alloy tube materials (bare or finned) and designed for high
thermal expansion conditions, the more conventional, stove-type or
packed-bed-type heat exchanger requiring cyclic heating and cooling of
parallel vessels containing refractory material (packed-bed or checker
16
-------
structure) with gas flow controlled by high-temperature valves, or a
circulating pebble-bed heat exchanger requiring continuous circulation
of a refractory heat transport medium between parallel vessels.
Conventional countercurrent absorber and stripper towers would be
used to remove the SOX from the combustion products at pressure and to
generate at atmospheric pressure an SC>2 gas suitable for elemental sul-
fur or sulfuric acid (l^SO^) recovery. Packed columns or a plate-type
design could be applied with proper construction for the highly corro-
sive environment. An internal heating or cooling surface could be
placed in the columns to control the column temperatures. Mist elimina-
tors might be required in order to protect downstream equipment from
corrosion.
The stripping gas could be either air, steam, or stack gas. Each
would have advantages in terms of power requirements, oxygen content,
and capital investment.
Elemental sulfur or H2S04 could be recovered from the stripper gas.
The composition of the stripper gas is critical to this step. A commer-
cial sulfur recovery process such as Allied Chemical's could be applied.
The Allied Chemical process requires the use of a clean fuel, such as
methane (City), for S(>2 reduction. Alternatively, a developmental pro-
cess such as the Foster Wheeler RESOX Process, which uses coal as a
reductant, could be applied. The tail gas from the sulfur plant could
be exhausted or recycled to the absorber.
The circulating solution system requires heat exchange, cooling and
heating with conventional devices in order to control the absorber and
stripper temperatures. In addition to a pump to circulate the solution
a means of pressure reduction such as a pressure reduction valve or a
power recovery turbine would be required, since the absorber is operated
at elevated pressure and the stripper is at low pressure.
17
-------
Partlculate material trapped In the absorber must be removed In
order to maintain the absorber performance. Various commercial devices
that will permit the filtration of a side stream of the circulating
solution are available.
Makeup water would be fed to the system to account for filter cake
losses and evaporation losses.
SELECTION OF BASE CASE DESIGN CONCEPT
A selection of a base design concept for the PFBC with pressurized
water scrubbing from the options presented in the previous section has
been made. We have judged, on the basis of preliminary considerations,
that the selected base-concept would probably be the most successful of
all of the concepts presented.
A previous cycle study for a PFBC concept that used low-temperature
venture scrubbing as an alternative to high-temperature particulate con-
trol was applied to reach the following conclusions:^
• A cornbustor temperature of about 927°C (1700°F), resulting
in a combustion product temperature of about 871°C (1600°F)
to the recuperator, arid a corabustor pressure of about
1034 kPa (150 psia) are suitable combustor operating con-
ditions for this concept.
• A recuperator effectiveness of at least 0.86 (resulting in
a turbine inlet temperature of about 760°C (1400°F)) and an
excess air rate of no more than 20 percent should be used
for economic feasibility. These conditions will yield a
plant heat rate of about 10,000 kJ/kWh (9,500 Btu/kWh),
including the boiler efficiency increase due to the elimi-
nation of sorbent calcination energy losses and energy
losses in the gas cleaning system. Using a steam generator
prior to using the recuperator or using the high excess air
fluidized-bed boiler or adiabatic combustor will not be
economically feasible with this concept.
18
-------
Other equipment and process selections are as follows:
• An inert ceramic bed (alumina) in the corabustor because it
should result in superior combustor performance in terms of
particle elutriation and potential ash fusion
• Two stages of high-temperature particulate removal equip-
ment (cyclones) located directly after the corabustor. The
first stage would recycle coarse material (alumina and car-
bon) to the combustor. The second stage would remove fines
from the corabustor products (coal ash and alumina) in order
to protect the recuperator from erosion and deposition.
These fines would be removed from the system. No low-
temperature particulate control equipment would be used
before water scrubbing, and we assumed that the absorber
and stripper could tolerate a relatively high particulate
content.
Because the recuperator is a critical process component, both the
shell-and-tube recuperator and the cyclic stove recuperator have been
evaluated. The packed bed concepts were not considered because of the
possibility of particle elutxiation and plugging. An effectiveness of
0.90 was selected with a turbine inlet temperature of 788°C (1450°F).
Valve-tray columns were selected for the absorber and stripper to
improve performance under conditions of high particulate content and to
permit simplified periodic maintenance of the columns. Plastic lining
was specified to protect against corrosion. Preliminary calculations
indicate that internal heating or cooling would not be required in the
columns.
Stack gas would be used for stripping rather than steam or air.
Steam consumes a large quantity of power and results in a large water
loss. Stack gas contains a lower oxygen content than does air and
19
-------
results In less reductant consumption for sulfur recovery. Preliminary
cleaning of the stack gas would be required in order to protect the
blower.
Elemental sulfur would be recovered by using the Allied Chemical
Process with City as the reductant. Developing technologies such as the
Foster-Wheeler RESOX process have not yet been demonstrated and may not
be very efficient or capable of producing commercial-grade sulfur. The
sulfur recovery process tail gas would be exhausted on the basis of an
assumed 90 percent sulfur recovery efficiency. On the basis of economic
projections, an S02 content of at least 4 mole % would be required in
the stripper off-gas.
A hydraulic turbine would be used for power recovery from the cir-
culating scrubber solution. The solution would be cooled by a cooling-
water exchange and heated by clean fuel (heating oil) combustion. Low-
grade steam was considered for heating the solution, but the steam
requirements exceeded the availability in the plant.
A typical design philosophy for large fluidized-bed combustion
plants calls for modular design with four parallel combustors in a
600 MWe power plant. This philosophy has been followed in this design
evaluation that specifies parallel gas-cleaning trains.
PLANT BASIS
The following plant basis was selected for the evaluation:
• 594 MWe power plant net output (635 MWe conventional PFBC)
power plant net output
• Four boiler modules
• 17.5 percent excess air in primary combustors
• 4 wt % sulfur coal with 10 wt % ash and a heating value of
30 x 106 J/kg (13,000 Btu/lb)
• S02 emission controlled to 0.5 kg S02/GJ (1.2 Ib S02/
Btu), equivalent to about 81 percent sulfur removal
20
-------
• Single sulfur recovery plant with 90 percent sulfur
recovery efficiency
• Absorber 89.5 percent efficient in removing SOX.
This basis provides direct comparison with previous PFBC designs using
calcium-based, high-temperature gas cleaning.
MATERIAL AND ENERGY BALANCES
Material and energy balances were performed for the base case
described and are summarized in Figure 2 and Table 1.
An iterative approach was used for the absorber and stripper system
material and energy balances in order to provide reasonably near optimum
designs for these columns. The minimum absorber operating temperature
possible [based on normal cooling water temperatures of 27-30°C (80-
85°F), and considering the absorber inlet gas temperature of 121°C] is
about 38°C (100°F). This temperature was selected for the design in
order to yield the most efficient SOX absorption. An operating tempera-
ture of 66°C (150°F) was selected for the stripper on the basis of max-
imizing the S(>2 concentration in the stripper gas and minimizing evapo-
rative water losses. The maximum SOX content of the stripper gas for
this process operated with realistic temperature conditions is about
6 mole %. A value of 5 mole % was selected for the design in order to
give reasonable column dimensions.
Energy balances around the absorber and stripper indicate that the
solution circulation rate would be so great (361,725 kg-moles/hr) that
heat of absorption effects, heat of evaporation effects, and sensible
heats of entering gas streams would be negligible and the columns would
operate isothermally.
A small amount of C02 would also be absorbed from the combustion
products and released into the stripper gas (about 180 kg-moles/hr).
The particulate content of the circulating solution was assumed to build
21
-------
Dwg. 1 68783'*
To Turbine
to
N5
Coal
©
Recuperator
AAA/Vr-n
I
Particulate
Control
Fines
8
Air
Sulfur
Recovery
Process
Figure 2 - Material and Energy Balances
-------
Table 1
MATERIAL AND ENERGY BALANCES
(594 MWe Power Plant)
Dwg .1712 SkO
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
IS
19
20
Stream
Combustion
products
Combustion
products
Absorber
gas
Absorber
gas
Stripper
stack gas
Stripper
air
Stripper
gas
Sulfur
Tail gas
Makeup water
Filter cake
Solution
Solution
Solution
Solution
Solution
Solution
Solution
Filtrate
Filter
solution
Flow Rate,
kg-moles/hr
89,255
89,255
88,847
88,847
3.454
3,454
4,543
204
-5,260
1,200
2, 722 kg/ hr
361,725
361, 725
361,725
361,725
361,725
361.725
333,773
27,952
27, 952
Composition, mole %
15% C02, 0. 284% S02
15%C02, 0.284%S02
15%C02. 0.030%S02. 0.5%H20
15%C02, 0.030%S02
4%02,C02.H20,N2
4%02> C02. H20, N2
5%S02, 16%C02, 15%H20, 3%02,
61% N2
Commercial grade sulfur
0.43%S02
H20
95 wt%particulate, 5 wt% acid solution
0. 004% S02 in water ( 0. 5 wt% solids)
0. 004% S02 in water ( 0. 5 wt% solids)
0. 004% S02 in water ( 0. 5 wt% solids)
0. 004% S02 in water ( 0. 5 wt% solids)
0.067%S02in water (0.5 wt% solids)
0. 067% S02 in water ( 0. 5 wt% solids)
0. 067% S02 in water ( 0. 5 wt% solids)
0.067%S02(Owt%solids)
0. 06 7 S02 in water ( 0. 5 wt% solids)
Temperature,
871(1600)
121(250)
38(100)
788(1450)
121 (250)
177 ( 350)
66(150)
38(100)
38(100)
27 (SO)
38(100)
35(95)
43(110)
66(150)
66(150)
68(155)
60 (140)
38(100)
38(100)
38(100)
Pressure,
kPa (psia)
1034(150)
1014 ( 147)
965(140)
952(138)
103(15)
172(25)
110(16)
103(15)
103(15)
965(140)
103(15)
965(140)
1000(145)
1296(188)
172 ( 25)
110(16)
117(17)
310(45)
310(45)
1034 ( 150)
23
-------
up to about 0.5 wt % of partlculate material. This would require that
8 percent of the circulating solution be continuously filtered in order
to maintain a steady particulate level in the solution. The effects of
particulate levels of this order on the absorber, stripper, hydraulic
turbine, pump, and heat exchangers requires further investigation.
For the 4 wt % sulfur coal and the 90% sulfur recovery efficiency
assumed for the sulfur plant, the overall process sulfur removal
efficiency would be 80.5%, if we assume an absorber efficiency of
89.5 percent. Coals with a higher sulfur content would require higher
sulfur removal efficiencies and greater solution circulations rates.
Coals with less sulfur would result in less S02 in the stripper gas.
The auxiliaries (power, fuel, water) required for the pressurized
water scrubbing process are listed in Table 2. Methane was used as the
reductant in the sulfur recovery process and fuel oil for the circulat-
ing solution heater.
Table 2
GAS-CLEANING AUXILIARIES
Makeup Water 365 £/min (96 gal/rain)
Cooling Water 72,000 A/min (19,000 gal/min)
Methane 5,200 m3/hr (183,200 scf/hr)
Fuel Oil 5,000 kg/hr (11,000 Ib/hr)
Power 3,870 kW
Pump 2,190 kW
Blower 2,520 kW
Turbine -840 kW
EQUIPMENT DESIGN
The major equipment items for the gas-cleaning system are described
in Table 3. Following the philosophy of maximum use of shop fabrica-
tion, we used a modular design which resulted in a single recuperator
-------
Table 3
Dwg. 1712Bl*l
tsJ
Equipment
Absorber
Stripper
Recuperator
Number per
Combustor
3
3
1
Description
Plastic-lined vertical shell column, 17 m (12ft) diameter, 32m (105ft) tall, 40 valve trays,
0. 76 m ( 1 5 ft) tray spacing, 1400 kPa (200 psia) design pressure, with mist eliminators
Plastic-lined vertical shell column, 4m (12ft) diameter. 37 m (120ft) tall, 76 valve trays,
0. 46 m ( 1. 5 ft) tray spacing. 345 kPa (50 psia) design pressure, with mist eliminators
Shell-and-tube horizontal heat exchanger, floating-head design, 4 m (13ft) diameter, 23m (75ft) long;
Alternative Recuperator
Circulating Solution 3
Filter
Circulating Solution 3
Turbine
Circulating Solution 3
Pump
Circulating Solution Heat 3
Exchanger
Circulating Solution 3
Cooler
Circulation Solution 3
Heater
Stripper Stack-Gas Blower 3
Stack-Gas Recycle System 1
8,000. 2.5cm din) OD finned tubes with 0.089cm (0.035 in) wall thickness, 15m (50ft) tube length,
finned-tube-area-to-base-tube-area ratio =10, Inconel or Incoloy construction, 1400 kPa (200 psia) .
design pressure
Two parallel vessels with internal refractory checker structure connected by high-temperature valves
(4 per vessel); each vessel 6.7 m (22ft) in diameter and 46 m (150 ft) long; each vessel containing
2.9 x 10 kg (6.5 x 10 Ib) of refractory checker structure; 1-hr cycle time assumed
Continuous-pressure drum filter, handles 42.000 kg (93.000 Ib) of solution/hr, 29 m (310 ft )
filter area, 1000 kPa (150 psia) inlet pressure, 690 kPa (100 psi) pressure drop
Hydraulic turbine, handles 7,600 4/min (2.000gal/min). recovers 67 kW (90 HP)
Centrifugal pump, handles 7,6001/min (2.000gal/min). consumes 182 kW (244HP)
Shell-and-tube heat exchanger, heat duty of 1.4 x 10 W (4.78x10 Btu/hr), tube surface of
1800m2 (19,000 ft2)
Shell-and-tube heat exchanger, heat duty of 5.26 x 106 W (1.79 x 107 Btu/hr). tube surface of
560m
2 (6,000 ft2)
Shell-and-tube heat exchanger, oil-fired, heatduryof 1.4x10 W (4.8x10 Btu/hr), tube
surface of 1860 m2 (20,000 ft2)
Centrifugal unit, stack-gas rate of 11.000 kg/hr (24,200 Ig/hr), 216 kW (290 HP) motor power
Baghouse, screw conveyor, airlock, lockhopper, valves, and booster fan-, 570.000 j/min (20,000acfm)
stack gas.
-------
(4 boilers per 594 MWe plant) and three parallel absorber/stripper gas
cleaning trains per fluidized-bed boiler. A single sulfur recovery
plant was used for the 594 MWe power plant.
The absorber and stripper columns were designed by following design
techniques and recommendations presented In the literature.4~6 For sim-
plicity we applied design relationships for dilute gas mixtures that
assumed the validity of Henry's Law. These assumptions should be excel-
lent for the absorber and reasonable for the stripper. Henry's Law con-
stants of 40 and 80 were assumed for the absorber and the stripper,
respectively.
The greatest uncertainty In design concerns the recuperators.
Designs for two types of recuperators (shell-and-tube and refractory
stove) were developed and are described.
The remaining items are essentially conventional devices modified
for the corrosion protection required and for the partlculate content of
the acid solution they must handle. The large number of modules
required for the process indicates potential economic limitations.
CAPITAL INVESTMENT
Equipment and total gas-cleaning process capital investments were
estimated on the basis of the descriptions in Table 3. The results of
these estimates are presented in Table 4 and based on mid-1977 dollars.
Again, the greatest uncertainty surrounds the recuperator costs. We
estimate that the total direct cost for the pressurtzed-water scrubbing
system would be $172/kW with the shell-and-tube and $144/kW with the
refractory stove recuperator design.
A breakdown of the boiler plant equipment costs for the dolomite-
based PFBC plant and the water-scrubber-based PFBC plant is given in
Table 5. A breakdown of the total power plant cost Is given In Table 6.
Both plants have the same coal-feed rate, Identical combustor designs,
26
-------
Table 4
CAPITAL INVESTMENT FOR PRESSURIZED WATER SCRUBBING8
Owg.77l8A30
Equipment
Purchased Equipment,
$x!06
Cost of Installed Equipment,
*x!06
Absorbers 2. 9 8. 9
Strippers 2.2 6.6
Recuperators
Shell-and-tube 21. 6
Refractory stove 16.8
Solution Filters 1. 5
Solution Turbines 0.7
Solution Pumps 0. 3
Solution Heat
Exchangers 3.3
Solution Coolers 1. 2
Solution Heaters 0. 6
Stack Gas B lowers & Recycle System 1. 2
Sulfur Recovery
Plant
TOTAL DIRECT COST
'Basis: mid-1977 dollars; 594 MW plant
G
49.8
31.2
3.4
1.8
0.7
7.6
2.8
1.3
2.2
17.0
102. 2 (shell-and-tube), 83. 6 (refractory stove)
and identical combustion product flow rates. The dolomite-based PFBC
plant produces 635 MWe of electrical energy, but the water-scrubber-
based PFBC plant produces 594 MWe of electrical energy because of lower
plant efficiency. Costs have been taken from previous Westinghouse PFBC
cost studies and updated to include particulate cleaning equipment and
escalation.^>7
We estimate that the total power investment for the PFBC with
calcium-based, high-temperature gas cleaning would be $423/kW. This
27
-------
Table 5
PFBC BOILER PLANT EQUIPMENT COSTS
Dolomite-Based
Equipment | System, $/kW
Water-Scrubber
Based Systems,
$/kWa
Steam Generator 20.89
Draft System
Partlculate removal 45.78
Draft flues and ducts 2.39
Piping 3.99
Stack and foundation 0.68
Coal- and Sorbent-Handllng
and Feeding Equipment 21.54
Ash- and Dust-Handling Systems 2.23
Stack-Gas Cleaning System
Instrumentation and Controls 4.47
Miscellaneous Equipment 1.36
103.33
Net Plant Electrical Output 635 MWe
22.33
19.66
2.55
4.27
0.73
13.03
1.23
172.05 (140.74)
4.78
1.45
242.08 (210.77)
594 MWa
aSystem with refractory stove recuperator is in parentheses; system with
shell-and-tube is shown to its left.
cost is based on mid-1977 dollars, 635 MW plant capacity, once-through
operation with dolomite, 17.5 percent excess air, and three stages of
particulate control equipment (final-stage, granular-bed filter).
The PFBC with pressurized-water scrubbing for S02 control would
cost about $639/kW with the shell-and-tube recuperator and $594/kW
with the refractory stove recuperator. This estimate is based on mid-
1977 dollars, 594 MW plant capacity, 17.5% excess air, two stages of
particulate control equipment (high-temperature cyclones), a combustor
28
-------
Table 6
PFBC POWER PLANT COST BREAKDOWN
Item
Land and Land Rights
Structures and Improvement
Boiler Plant Equipment
Gas Turbine-Generator Equipment
Steam Turbine Generator Equipment
Electric Plant Equipment
Misc. Plant Equipment
Undistributed Costs
Other Plant Costs
Subtotal
Normal Contingency
Subtotal
Escalation
Subtotal
Interest during Construction
General Items and Engineering
Limestone-Based
System, $/kW
1.63
28.21
103.33
21.33
63.62
22.93
5.13
40.86
4.19
291.23
17.47
308.70
57.88
366.58
48.13
7.93
Water Scrubber
Based-System,
$/kWa
1.74
30.16
242.00 (210.77)
22.80
68.01
24.51
5.48
43.68
4.48
442.94 (411.63)
26.58 (24.70)
469.52 (436.33)
88.03 (81.81)
557.55 (518.14)
73.20 (68.03)
7.93
TOTAL CAPITAL COST
422.64
638.68 (594.10)
aSystem with refractory stove recuperator is in parentheses; system with
shell-and-tube is shown to its left.
29
-------
cost identical with the high-temperature, gas-cleaning case, and a
reduction of $10/kW to account for the elimination of dolomite-handling
equipment.
A conventional, coal-fired steam power plant with limestone scrub-
bing for S(>2 control would probably cost from $500 to 570/kW.
The alternative PFBC system using venturi scrubbing for particulate
control is estimated to cost between $459 and 502/kW based on 17.5 per-
cent excess air, two stages of high-temperature particulate removal
equipment, and mid-1977 dollars (Reference 3, Appendix A).
The option of applying the RESOX process for sulfur recovery to the
PFBC with water-scrubbing for S02 control in place of the commercially
available Allied Chemical process would probably increase the capital
investment further because of the low sulfur-recovery efficiency
expected with a 5 percent SC>2 gas. The basic RESOX plant would cost
about the same as the Allied Chemical process, but the tail-gas cleaning
plant for the RESOX process (Beavon process, for example) could easily
cost another $20 to 30/kW based on a sulfur recovery efficiency of 50 to
60 percent.
An estimate of the most optimistic case for the pressurized-water
scrubbing concept for PFBC was also developed. If the minimum modular
design is used (a single gas-cleaning train per coinbustor module) with
no increase in plant construction time, the refractory stove recuperator
design, a coal-fired solution heater, and a RESOX sulfur recovery plant
(assumed to have 90 percent sulfur-recovery efficiency), the total power
plant capital investment would be reduced to $577/kW instead of the more
realistic case of $59l/kW to 635/kW.
In these capital cost estimates we have assumed off-site disposal
of waste solids. In the dolomite-based PFBC system, the bed overflow
and collected fly ash would be conveyed dry to on-site storage silos. A
similar system for collected fly ash would be used in the water-
scrubber-based PFBC system, where no accumulation of coal ash in the
30
-------
inert combustor bed was assumed. The filter cake is also disposed of
off site, handled with slurry techniques similar to those used on FGD
sludge. Disposal cost is counted as an operating cost, accumulated
within the cost of electricity.
COST OF ELECTRICITY
The costs of electricity generated by PFBC with calcium-based S02
control and with pressurized-water scrubbing S02 control are developed
and compared in Table 7. The energy cost of a conventional coal-fired
power plant with limestone scrubbing is also shown.3.7 The basis on
which these costs are derived is listed in the table.
The energy cost associated with the pressurized-water scrubbing
concept is projected to be 3.8 to 5.0 mills/kWh greater than the
calcium-based fluidized-bed combustion power plant energy cost and
2.7 mills/kWh greater to 0.4 less than a conventional power plant energy
cost. For the most optimistic case previously defined the total energy
cost would be 23.7 mills/kWh.
The cost of disposing of waste solids and liquids does not con-
tribute significantly to the cost of electricity for the disposal costs
assumed. If higher costs should occur in the future (say >$20/Mg) the
water-scrubber PFBC concept could result in competitive costs of
electricity.
Also, should the cost of sorbent increase significantly (say to
>$20/Mg) then the water-scrubber PFBC concept could provide economic
incentive for development.
ENVIRONMENTAL COMPARISON
The environmental performance of calcium-based PFBC and of PFBC
with pressurized-water scrubbing for S02 control are compared in
Table 8. The concepts are expected to be comparable with respect to
31
-------
OJ
Table 7
COMPARISON OF COST OF ELECTRICITY3
I7HB93
Item
PFBC: Calcium-Based
SO.Control
PFBC: Pressurized -Water
Scrubbing S02 Control
Conventional Power
Plant: Limestone Scrubbing
Capital Investment, $/kW
Energy Cost, mills/kWh
Capital charges
O&M
Fuel (coal)
Sorbent
Auxiliary fuel
Makeup water
Cooling water
Solid/ liquid waste disposal
TOTAL
a
423
10.4
1.1
6.8
1.5
0.6
20.4
Shell-and-Tube
Recuperator
639
15.6
1.7
7.2
0.6
0.1
<0. 1
0.2
Refractory Stove
Recuperator
594
14.5
1.6
7.2
0.6
0. 1
<0. 1
0.2
Basis: Capital charges 15% of capital investment per year
Capacity factor 70%
0 & M 2. 36% of capital investment per year
Sorbent (dolomite and limestone) at$10/Mg
Ca/ S of 2. 0 for fluid-bed combustion, 1
Coal at $0. 80/GJ($0.76/106Btu)
Cooling water at 0. 5*/1031 (24« /M gal)
Process water at 5tf 1031 (20(1 M gal)
Methane and fuel oil at $ IIGJ ($ 0. 95/106 Btu)
Dry solid disposal $4/Mg
Sludge disposal at $ 10/Mg
25.4
24.2
500-570
12. 2 -13. 9
1.5- 1.7
7.2
0.6
<0.1
1.2
22. 7 - 24. 6
-------
NOX, particulates, and heat rejection. The high-temperature, calcium-
based, gas-cleaning process, however, has the greater potential for SC>2
emissions lower than the current standard of 0.5 kg/GJ (1.2 Ib/lO^ Btu).
The PFBC water-scrubber concept is probably limited to sulfur removal
efficiency less than 90 percent because of limited S02 solubility in
water and the limited efficiency of sulfur recovery in commercial and
developmental sulfur recovery processes. NOX emissions from the inert-
bed combustor could conceivably be less than or greater than the
calcium-based combustor NOX emissions; factors such as catalytic effects
from calcium compounds or alumina particles and the Influence of high-
versus-low combustion gas SOX content may affect the formation/decompo-
sition of NOX in the combustor. Differences in particulate emissions
between the two cases are also possible because of the elutriation of
sorbent fines from the corabustor in the dolomite-based PFBC. Particu-
late standards should be achievable with both concepts with properly
selected equipment.
The solid wastes associated with the high-temperature, calcium-
based, gas-cleaning process would be larger in mass than those for the
pressurized-water-scrubbing concept by a factor of 2.4 if all forms are
considered, by a factor of 6 to 20 if the waste sorbent is compared to
the filter cake and attrlted alumina only. The difficulty of handling
the waste solids, however, and the environmental impact of these wastes
would not necessarily be directly proportional to mass, and further pro-
cessing of the filter cake waste would be required. The environmental
impact and exact nature of the filter cake material is unknown, but we
expect that this material could be handled by methods applied for cor-
rosive wastes in the chemical industry, with neutralization being poten-
tially acceptable.
With respect to waste liquids, makeup water consumption, clean fuel
consumption, and the plant heat rate the high-temperature gas-cleaning
process appears superior to the pressurized-water-scrubbing concept for
33
-------
Table 8
ENVIRONMENTAL COMPARISON (600 MWe Power Plant)
Dwg. 171!892
CO
-p-
S02. kg/GJ(lb/106
6
N0x, kg/ GJ ( Ib/ 10
Partial late, kg/GJ(
PFBC with C<
so2<
Btu) <0. 5«
Btu) <0. 3«
ilcium-Based PFBC with Pressurized -Water-
Control Scrubbing S0? Control
1.2) 0.5(1.2)
0.7) Probably <0. 3«0. 7)
lb/106Btu) <0. 04KO. 1) <0.04«0. 1)
Heat Rejection. % less than . ~ 15
-14
conventional plant
Waste Liquids None
Waste Solids, Mg/hr/MW {fraction of coal feed mass)
Total 0. 142 (0. 458)
Coal ash 0. 031 (0.10)
Sorbent 0.111(0.358)
Sulfur None
Others None
Makeup Water Con sumption, J/min None
Clean Fuel Consumption
Methane, 103
-------
S02 control. The amount of process water circulated in the water-
scrubber concept is about ten times as much as in a conventional plant
with limestone scrubbing.
For the most optimistic case the consumption of clean fuels would
be reduced to zero, significantly more coal would be consumed (~7%) and
the emission of particulates and S02 would be expected to increase
slightly. The solid waste generation would increase by about
0.003 Mg/hr/MW due to increased coal ash.
The high energy consumption of the PFBC water-scrubber concept is
of particular concern.
CONCLUSIONS
The PFBC power plant utilizing the cold gas-cleaning (water scrub-
bing) concept for S02 control is not economically competitive with
calcium-based PFBC or conventional steam power plants unless the cost of
sorbents and/or waste solids disposal is substantially increased above
the costs assumed in this study. There appears to be no alternative
that could significantly improve the cold gas-cleaning concept eco-
nomics. The recuperative heat exchanger is technically an item of great
uncertainty and may limit the concept feasibility.
Environmentally, the pressurized-water-scrubbing concept could
eliminate the massive amount of sulfated dolomite waste generated by
PFBC, but the nature of the environmental effect of the waste filter
cake produced in the process is uncertain. The energy conversion effi-
ciency of the water-scrubber concept is very poor, and the sulfur
removal efficiency associated with the concept is limited to less than
90 percent.
35
-------
7. PARTICULATE CONTROL TRADE-OFF FOR PFBC SYSTEMS
OVERVIEW
Work is reported in Volume 2 (EPA-600/7-80-015b) that provides per-
spective on determining the impact of emission requirements, fluidized-
bed combustor design and operating conditions, and turbine performance
constraints on PFBC particulate control requirements and plant econom-
ics. An EPA technical directive was performed in 1978 to assess the
effect of final-stage filter performance and the effect of filter stag-
ing on particulate loading and size distribution emitted and the result-
ing implications on turbine life. This work provided a basis for the
subsequent analysis presented in Volume 2.
The results from the technical directive study are reported since
they illustrate the methodology used to project turbine blade life and
electrical energy costs as a function of different particulate loadings
to the turbine using the Westinghouse turbine erosion and particle pro-
file models. The projections of the particulate loadings and size dis-
tributions used here were based on methods employed prior to the devel-
opment of the particle profile model described in detail in Section 5 of
Volume 2. The model originally used for particle profile projections
was not capable of including the effects of a recycle cyclone. The PFBC
configurations discussed in this analysis, therefore, are limited to the
carbon burnup cell (CBC) concept for high carbon utilization. The tur-
bine erosion model has similarly been extended with additional under-
standing and experimental determination of the important model
parameters.
This work is important in that it documents the early analysis
and represents the basis for developing the tools that can provide per-
spective on important trade-offs and permit the design of reliable PFBC
plants that operate within environmental constraints.
36
-------
BACKGROUND
A substantial amount of work has been performed at the Westinghouse
R&D Center to analyze the flow trajectories and erosion effects of par-
ticulates entrained in the flow stream of gas turbines. Reference 8
predicts quantitative turbine erosion rates for a typical 65 MW utility
gas turbine expander by combining three major complementary analyses.
These are the calculation of the inviscid dimensional flow stream
through the blading; the calculation of the trajectories of particles
entrained in the blading flow stream; and the calculation of the erosive
effects of those particulates whose calculated trajectories result in an
impact with the blade surfaces.
We must emphasize that the analysis in Reference 8 did not include
the effect of particulate deflection and velocity reduction due to
profile-boundary layers on the blading surfaces. Note also that the
author cautioned that the erosion model used was based on data available
in the literature, which at best is sketchy.
On the basis of the observations of actual erosion patterns on
experimental coal-burning gas turbines,9,10 we have concluded that sec-
ondary flow phenomena have a substantial effect on the trajectories of
small (i.e., <10 Mm) particles. Accordingly, an analysis was carried
out**- that includes the effects of viscous boundary layers on the tra-
jectory of small particles. The results of this analysis confirmed
qualitatively the tendency of the boundary layer flows to concentrate
the particulates in certain regions of the gas turbine. Because of the
complexity of the problem, however, quantitative erosion rates were not
calculated.
Another phenomenon that Is of interest, particularly for very high-
temperature turbine applications, is the temperature reduction of the
particle as it penetrates the boundary layer next to a cooled blade sur-
face. To determine the temperature history of a particle as it passes
through the boundary layer one must also calculate the trajectory veloc-
ity history.
37
-------
ESTIMATION OF PARTICLE LOADING/SIZE DISTRIBUTION
The works reported above are related to the erosion caused by par-
ticles after they enter the turbine. One must, of course, determine the
concentration and size distribution of the particles that enter the tur-
bine. These parameters are determined by the partlculate removal equip-
ment Installed in the hot gas stream between the turbine and the pres-
surized fluidlzed-bed boiler in which are generated both the hot gas
stream and the entrained partlculates. (Essentially the same particu-
late removal equipment would be used If the hot gas stream source were
an adlabatic fluldized-bed combustor or a fluldized-bed gasifler.) The
performance of the partlculate removal equipment depends upon the con-
centration, density, and size distribution of the entering particles,
and the detailed design of the equipment would depend upon these fac-
tors. The general approach, however, is to remove the bulk of the large
particles in the early stage(s) and remove the fines in the later
stage(s). The performance and cost of a particulate removal system in a
plant with a pressurized fluidlzed-bed boiler have been described in
some detail.12
Figure 3 is a flow diagram of the separation equipment arrangement
designed for the present study. Figure 4 shows a typical arrangement of
the particulate removal equipment and the piping connecting the pres-
surized fluidized-bed boiler to the gas turbine. As Figure 3 indicates,
the effluent from the pressurized fluidized-bed boiler first enters a
cyclone separator, a relatively inexpensive component that can handle
high particulate loadings and has a much higher removal efficiency for
the larger particles than for the smaller.
The particles collected by the primary cyclone separator contain a
large portion of char (carbon), which is fed to a CBC to recover the
chemical heat of combustion. Additional combustion air is fed to the
CBC to complete the reaction. The volume of flow through the CBC is
small compared to that of the main flow stream, so the exit stream is
38
-------
Dwg. 6407A85
From PFBB
Cyclone
Separator
Cyclone
Separator
Carbon
Burnup
Cell
\
Air
T
Particulates
Particulates
Air
First-Stage
Granular-
Bed Filter
Particulates
Second-Stage
Granular-
Bed Filter
Particulates
Figure 3 - Diagram of Particulate Removal System
Gas Turbine Generator
Cyclones
Carbon
Burnup Cell
Pressurized f^\-^f^\ Pressurized
Fluidized- I /Sv Jpluidized-
Bed Boiler \ X \—/ Bed Boiler
B A
Figure 4 - Particulate Removal Equipment Arrangement1^
39
-------
passed through a small cyclone separator before being introduced into
the main flow. This cyclone is designed to accommodate the very high
particulate loadings leaving the CBC and is equipped with secondary
inlet air jets that prevent fouling of the unit and provide angular
momentum for the particle separation process.
The combined flow stream then enters the granular-bed filter, which
is capable of high removal efficiency for particles smaller than 10 vim.
Two stages of granular-bed filters are shown in Figure 3. Since the
granular-bed filters are by far the most expensive component of the par-
ticulate removal system, this study will investigate whether the second
stage is economically justified.
An analysis made specifically for the present study has been used
to calculate the particulate concentration and size distribution at each
location within the particuate removal system. The analysis considers
the sorbent, the ash, and the char separately, since each of these con-
tituents has a different effective density, an important factor in the
efficiency of a particulate removal component.
The design conditions of the fluidized-bed boiler used to calculate
the rate and size of the elutriated particles is as follows:
Superficial bed velocity 1.52 m/s (5 ft/s)
Bed depth 2.4 m (8 ft)
Pressure 1013 kPa (10 atm)
Bed temperature 1010°C (1850°F)
Sorbent Dolomite
Sorbent feed size 3.2 x 0 mm (1/8 x 0 in)
Coal feed size 6.4 x 0 mm (1/4 x 0 in)
Excess air Primary bed, 6%; CBC 36%
Ca/S atom ratio 1.5:1
Carry-over from the bed was estimated by calculating the size of the
particle whose terminal velocity was equal to the superficial velocity
in the combustor. All of the feed material below this size was assumed
40
-------
to be elutriated. Efforts to include sorbent attrition in the calcula-
tion involved the use of an average attrition rate that was a function
of bed velocity and residence time. A more detailed account of the cal-
culations involved in computing the carry-over characteristics is pre-
sented elsewhere.^
The performance of the primary conventional cyclone is shown in
Figure 5 as a function of species density. These estimates were made on
the basis of a commercial vendor's estimates of performance at pressure
and temperature. The cyclone pressure drop is estimated at 165 cm wg at
an inlet velocity of 17 m/s.
The performance of the CBC cyclone (Figure 6) is similarly based on
the manufacturer's estimates for the elevated temperature and pressure.
The specific device considered was a Donaldson Tan Jetv^y cyclone oper-
ating with a primary flow pressure drop of about 180 cm wg. This device
also employs a high-pressure, secondary flow of clean gas to impart the
rotary motion to the dust-laden gases entering the device. A schematic
of this device is presented in Figure 7.
The granular-bed filter concept used to establish the costs of the
final filter system is basically the same Ducon filter that had been the
basis of cost estimates for the previously published ECAS^2 report.
Figure 8 presents a schematic of the filter system. In the Westinghouse
design relatively few, but large (7-8 m dia), pressure vessels were used
to effect a cost savings over a large number of small modules.
Although the performance of cyclone separators is known to a rea-
sonable degree of accuracy, the performance of the granular-bed filter
- because it is still in an early stage of development — is poorly def-
ined. In the long range granular bed filters may achieve a removal
efficiency equal to that currently achieved by conventional low-
temperature fabric filters. At present, Westinghouse has bench-test
results of a granular-bed filter whose performance is not as good as
41
-------
Curve 694490-A
Specific Gravity
2.8 1.35.8
0.01
0.1
D Ipm)
Figure 5 - Grade Efficiency of Primary Cyclone Separator
Specific Gravity
2.88 .88.64 ^'
10
30
F 90
- 70
90
98
99
99.9
99.99
Tan Jet
TM
i ill
T= 927 °C I1700°F)
P= 1013kPa(10Atm)
i ill
i iii
.10 1.0 10.0 100
D . Mm
P
Figure 6 - Grade Efficiency Curve for Tan Jet
42
-------
6214816;
1 Clean Gas Out
Secondary
Gas
Dust
-Dirty Gas In
-Body
Tangential Gas
Jet Assembly out
Gas Jets
Figure 7 - Schematic of Tan Jet
Cwg. 6Z24A1Z
Catalyst- <
Laden
Oust
, I
Filter Element Internals
Clean Gas
Operating Cycle
Cleaning Cycle
Figure 8 - Granular-Bed Filter Module
43
-------
that of a fabric filter. As a third benchmark, results are available
from the EPA^ on an existing commercial unit (Rexnord) whose perform-
ance is substantially inferior to that of fabric filters.
In an attempt to bracket the actual situation, the calculations
presented here have been based on the grade efficiencies of each of the
three cases mentioned above. The actual grade efficiencies used for the
fabric filter and the Westinghouse granular-bed filter performance are
pre-sented in Figure 9 along with the Rexnord data.
Using these three efficiency criteria, we have calculated the con-
centration of particulate at the outlets of the first and second stages
of granular-bed filters, and these are shown in Tables 9 and 10. A more
graphic illustration of these performances is shown in Figures 10 and
11. Figures 12 through 20 show the size distribution of the particles
leaving these filters. Similar information at the discharge of the
cyclone separators has not been included since preliminary analysis
showed that the turbine erosion rates corresponding to these particulate
loadings were so high that a system not including granular bed filters
would be impractical.
ESTIMATION OF GAS TURBINE IMPACT
On the basis of the information available in all of the above ref-
erences, a relatively simple method was conceived with which to calcu-
late the erosion rate of gas turbine blading as a function of particle
distribution and concentration. A parametric, quantitative assessment
of power generation cost penalties has been determined from those data
on erosion rate, practical wear limits, and the cost of blading replace-
ment (compared to the cost of particulate removal equipment as a func-
tion of its removal efficiency).
Reference 8 shows that maximum erosion occurs at the leading and
trailing edges of rotors and at the trailing edge of stators. Although
44
-------
Curve 694489-A
\ I I I I U
1 - Fabric Filter, Ref. W
2 -® Granular-Bed Filter Data
3 - Rexnord Granular-Bed
Filter Ref. 13
99
0.1 1.0
Particle Diameter,
Figure 9 - Various Filter Performances Assumed
for Final Cleanup Stage
10.0
these rates are roughly comparable in magnitude, the calculations
neglect the effect of blade profile boundary layers. Since the boundary
layer thicknesses are larger at the trailing edge than at the leading
edge, the actual erosion rates will be reduced a greater amount at the
trailing edge. This is particularly true for the particulates that pass
through a granular-bed filter since they have a large proportion of very
small particles that are readily deflected and slowed by the boundary
layer velocity profile. This general reasoning is supported by the
experimental evidence^, which shows the maximum erosion at the leading
edge of the rotor blades. On the basis of these considerations, we have
made calculations for this analysis only for the rotor leading edge.
The calculation procedure is essentially a stepwise integration of
the erosion of each particle size over the range of particle sizes
entering the gas turbine for each of the three types of particulates.
45
-------
Table 9
PROJECTED PARTICULATE CONCENTRATION LEVELS
Projected Based
Upon
Location,
Outlet of
Dolomite Concentration,
g/sm3 (gr/scf)
Ash Concentration,
g/sm3 (gr/scf)
Char Concentration,
g/sm3 (gr/scf)
Rexnord Granular-
Bed Filter
Rexnord Granular-
Bed Filter
J) Bench Tests
Granular Bed Filter
J) Bench Tests
Granular Bed Filter
Conventional
Fabric Filter*
Conventional
Fabric Filter*
First GBF
Second GBF
First GBF
Second GBF
First GBF
Second GBF
0.1326
(0.05795)
0.0866
(0.03783)
0.00787
(0.00344)
0.000214
(0.0000935)
0.00103
(0.000449)
0.000000
(0.000000)
0.2179
(0.09518)
0.1243
(0.0543)
0.0138
(0.00601)
0.000290
(0.000127)
0.00203
(0.000886)
0.000000
(0.000000)
0.0398
(0.01737)
0.0244
(0.01065)
0.00240
(0.00105)
0.0000559
(0.0000244)
0.000353
(0.000154)
0.000000
(0.000000)
*Particle concentrations indicated are
a conventional low-temperature fabric
based upon assumptions
filter.
that GBFs perform as effectively as
-------
Table 10
PARTICULATE EMISSION LEVELS
Performance No. of (
Characteristics Stage:
Rexnord 1
Rexnord 2
(w) Bench Tests 1
(W) Bench Tests 2
Fabric Filter 1
Fabric Filter 2
Total Particulates Emitted
3BF
5 g/snH (g/scf)
(0.1705)
0.390
(0.1028)
0.235
(0.0105)
0.0240
(0.00025)
0.000572
(0.0015)
0.00343
nil
g/MJ
(0
0
(0
0
(0
0
(0
0
(0
0
(lb/106 Btu)
.137)
.136
.191)
.0820
.0195)
.00838
.000465)
.000200
.00279)
.00120
nil
-------
oo
0.229
(a 10)
0.206
-------
Curve 713968-A
Curve 713969-A
0.01
0.1
0.5
1
2
5
10
a 20
i/t
| 40
1 60
99 -
99.9 -
99.99
0.2
I I I I II ITJ
Dolomite Particles
Exit of Second GBF
I i I
0.4 0.6 0.8 1 2
Particle Diameter, t
3 4 5 678 10
0.01
0.1
0.5
1
2
5
10
a 20
*/>
f 40
I 60
»
£ 80
90
99
99.9
99.99
~nn—n '—'—' MINI
Ash Particles
Exit of First GBF
I
Exit of Second GBF
I I
I ill
0.2 0.4 0.6 1 2 3 4 5 678 10
Particle Diameter, urn
Figure 12 - Projected Outlet Size Distribution
Based on Rexnord Commercial Unit
(Dolomite Particles)
Figure 13 - Projected Outlet Based on
Rexnord Commercial Unit
(Ash Particles)
-------
Curve 713967-A
0.01
0.1
0.2
1
2
5
10
40
c 60
s
£ 80
90
99
99.9
99.99
I
Char Particles
Exit of First GBF
Exit of Second GBF
0.2 0.4 0.6 1.0 2 4
Particle Diameter, urn
6 8 10
0.1
0.5 -
i
2-
5
10
S 20
e
| 40
3
I"
£ 80
90
99
99.9
99.99
Curve 713973-fl
I I I I I | I
Dolomite Particles
Exit of First GBF
Exit of Second GBF
I I
I
i i i i i i
0.2 0.4 0.6 0.8 1 2
Particle Size, u m
6 8 10
Figure 14 - Projected Outlet Based on Rexnord
Commercial Unit
Figure 15 - Projected Granular-Bed Filter Outlet
Size Distribution Based on
Westinghouse Bench-Scale
Experiments (Dolomite Particles)
-------
Curve 713966-A
0.01
0.1
0.5
1
2
5
10
20
12
o>
•o
0)
Q.
c 60
8
80
90
99
99.9
99.99
T —i -T | -l r
Ash Particles
Exit of First GBF
Exit of Second GBF
i I i 11
I i
i ill
.2 .4 .6 .8 1 2
Particle Size, M m
3 456 8 10
0.01
0.1
0.5
1
2
5
10
J 20
! 40
)
60
- 80
90
99
99.9 -
Curve 713965-A
99.99
I I I I I I I I I
Char Particles
T I i i i i l
-Exit of First GBF
-Exit of Second GBF
l i
0.2 0.4 0.6 0.8 1 2
Particle Size. iam
34 6 8 10
Figure 16 - Projected Granular-Bed Filter
Outlet Based on Westinghouse
Bench-Scale Experiments
(Ash Parties)
Figure 17 - Projected Granular-Bed Filter
Outlet Based on Westinghouse
Bench-Scale Experiments
(Char Partcles)
-------
Curve 713971-A
0.01
Curve 713972-A
Ul
0.0 -
0.5
1
2
5
10
i 2°
I
I 40
: 60
»
! 80
95
99
99.9
99.99
1 2 3
Particle Diameter, \i m
4 5 6 7 8 10
Figure 18 - Projected Granular-Bed Filter
Outlet Size Distribution
Based on Conventional Fabric-
Filter Unit Performance
(Dolomite Particles)
a 01
0.1
0.5
1
5
10
20
40
60
0)
a.
90
95
99
99.9
99.99
I 1 I I I I I I | I I I I I M |
Ash Particles
I I i i i i i 11
i i i
1 2 3 4 5 6 8 10
Particle Diameter, \im
Figure 19 - Projected Granular-Bed Filter
Outlet Based on Conventional
Fabric Filter Unit
Performance (Ash Particles)
-------
Curve 713970-A
Ui
U)
99.
1 2 3 4 5 678 10
Particle Diameter, mn
35x6
(14)
30x5
(12)
£ 25x4
CE
3 ID
il
3> E
ee —
20x3
(8)
15x2
(6)
10x2
(4)
5X1
(2)
' I ' I
Curve 71 3t-64-A
' r
Particle Density
=2.5g/cm
op = 1.5g/cm
Particle Concentration = (0.00023g/sm
i i i i i i i i i i
10
12
Particle Diameter, (D ),
P
14
Figure 20 - Projected Granular-Bed Filter Outlet
Based on Conventional Fabric-Filter
Unit Performance (Char Particles)
Figure 21 - Blade Leading Edge Erosion
Rates
-------
1. Choose a step size (e.g., 1 Mm).
2. Enter Figure 11 (or appropriate subsequent figure) and read off
the percentage undersize at each side of the step.
3. Take the difference between these two values. This gives the
percentage of the particulate size at the mean of the step.
4. Multiply this percentage by the total particulate concentration
entering the turbine.
5. Enter Figure 21 (taken from Reference 8) and read from the
curve the metal recession (for 10,000 hours of operation) at
the rotor leading edge for the particular particle size of this
integration step. (Note the density parameter. The erosion
rate reduces slightly for lower density particles since they
deviate less from the gas streamlines. Note also that the ero-
sion of Figure 2L is for a given particle concentration.)
6. Multiply the erosion read from the curve by the ratio of the
actual concentration (from Step 4) to the concentration for
which Figure 21 has been calculated.
7. From Table 11, which traces the trajectory of a given size par-
ticle in a specified boundary layer thickness, read the velo-
city of the particle at impact with the blade surface.
8. Calculate the square of the ratio of the velocity at impact to
the velocity entering the boundary layer.
9. As mentioned in Reference 8, the erosive effect of a particle
is proportional to the square of the impact velocity. There-
fore, multiply the recession rate calculated in Step 6 by the
factor calculated in Step 8.
10. Repeat Steps 1 through 9 over the complete range of particle
sizes.
11. Add the erosion rates of all the particle sizes.
12. Based on the estimate of the Westinghouse Gas Turbine Division
that the maximum allowable amount of erosion of a turbine rotor
blade would be 254 mm (0.100 in), multiply 10,000 hours by the
54
-------
Table 11
THERMAL HISTORY OF PARTICLES ENTERING A 0.005 INCH
THICK BOUNDARY LAYER WITH 10 DEC. INCIDENCE, AT Rl-L.E.
GAS VEL. = 1200 FT/S, TO = 2060.DEC. R, TW = 1460.DEC.R, THER.EMISS.
0.9
.94115-26
.11111-08
.22222-06
.33333-06
.44444-06
.55556-06
.66667-06
. 00000
.32000+00
.63997+00
.95989+00
.12797+01
.15995+01
.19191+01
.99154+00
.94377+00
.88793+00
.83243+00
.77733+00
.72263+00
.66827+00
.11980+04
.11862+04
.11716+04
.11568+04
.11411+04
.11245+04
.11071+04
.15990+04
.15931+04
.15359+04
.15784+04
.15705+04
.15623+04
.16635+04
.12000+04
.12000+04
.11998+04
.11996+04
.11992+04
.11988+04
.11932+04
-.21156+03
-.21013+03
-.20672+03
-.20732+03
-.20592+03
-.20453+03
-.20315+03
.15994+04
.15994+04
.15993+04
.15992+04
.15990+04
.15988+04
.15986+04
.10401+02
.10393+02
.10430+02
.10523+02
.10679+02
. 1 0903+02
.11205+02
.11773+05
.11741+05
.11713+05
.11696+05
.11691+05
. 1 1 699+05
.11721+05
.49687+05
.33002+04
.15458+04
.99565+03
.72610+03
.56570+03
.45902+03
ijl
-------
ratio of 254 mm (0.100 in) to the total recession calculated in
Step 11. This gives the time in hours required to erode away
254 mm (0.100 in).
13. Repeat the above procedure for each of the particulate
constituents — i.e., dolomite, ash, and char.
RESULTS
The boundary layer near the leading edge of a rotor blade is very
thin, of course, since it has had little distance in which to develop.
The thickness is a function of the Reynolds number and can be estimated
from simple flat plate theory. Because of the rapid acceleration of the
main stream flow around the leading edge of a blade, the boundary layer
growth tends to be retarded in this region compared to a flat plate. In
Figure 144 of Reference 15, an experimental determination of boundary
layer thickness is given for an airfoil at approximately the right
Reynolds number and thickness chord ratio. The thickness shown in the
figure near the leading edge is approximately 127 urn (0.005 in). Accor-
dingly, calculations of the erosive life of the blading have been made
for boundary layer thicknesses of 0.0, 127, 254, and 508 pm (0.0, 0.005,
0.010, and 0.020 in) to determine sensitivity to this critical
parameter.
Another parameter of importance, which is not well defined at this
point, is the erosiveness of the particulates. On the basis of evidence
available from the literature, Reference 8 assumed that the erosiveness
of the ash and dolomite particles was 1/25 of that of silcon carbide
(SiC) particles.
Figures 22 through 24 show the results of the calculation for the
three levels of performance of the granular bed filters. Blade life, in
hours, is plotted against boundary layer thickness with parameters of
erosiveness. The range of erosiveness chosen is from twice to half what
was indicated in Reference 8.
56
-------
Curve 689722-A
20,000
18,000
16,000
14,000
£ 12,000
g
•=_ 10,000
CD
3 8000
6000
4000
2000
Curve 6897Z1-A
Erosiveness Compared
to SiC Particles .
0 0 (2) (4) (6) (8) (10) (12) (14) (16) (18) (20) (22)
51 102 152 203 254 305 356 406 457 508 559
Boundary Layer Thickness, \i m (mils)
Figure 22 - Projected Turbine Life for a
Participate Removal System with
Two Stages of Granular-Bed
Filters (performance based on
Rexnord, commercial unit)
24,000
22,000
20,000
18,000
16,000
e HOOO
1 12,000
oT
2 10,000
8000
6000
4000r-
2000
Erosiveness Compared
to SiC Particles
1/50,
0 (2) (4) (6) (8) (10) (12) (14) (16) (18) (20) (22)
1 102 152 203 254 305 356 406 457 508 559
Boundary Layer Thickness, urn (mils)
Figure 23 - Projected Turbine Life for a
Particulate Removal System with
One Stage of Granular-Bed
Filters (performance based on
Westinghouse bench-scale
experiments)
-------
Figure 22 shows the life of the turbine with two stages of granular
bed filters, each with a performance equal to that of the Rexnord com-
mercial unit. As the figure indicates, for a boundary layer thickness
in the expected range of 0 - 127 ym (0 to 5 mils), the life of the tur-
bine is very short — no more than six months, even assuming an optimis-
tic level of erosivity. One stage of Rexnord filter was found to give
inadequate life.
Figure 23 shows the life of the turbine with one stage of granular
bed filters with a performance equal to that of the Westinghouse bench-
test unit. As the figure indicates, this level of performance increases
the turbine life appreciably, although it is still much shorter than is
usual for utility equipment.
Not shown on a figure is the life resulting from the use of two
stages of granular-bed filters with the Westinghouse measured
30,000
28,000
26,000
24,000
22,000
20,000
18,000
e
§ 16,000
§ 14,000
12,000
10,000
8000
6000
4000
2000
0
1/25
Erosiveness Compared
to SiC Particles
i
i
i
i
Curve 689733-A
0 12) (4) (6) (8) 110) 112) (14) (16) (18) 120) 122)
51 102 152 203 254 305 356 406 457 508 559
Boundary Layer Thickness, pm (mils)
Figure 24 - Projected Turbine Life for a Particulate Removal System
with One Stage of Granular-Bed Filters (performance
based on conventional fabric filter efficiency)
58
-------
performance. This Is because the partlculate concentration is so low
that even if a zero boundary layer thickness is assumed, the turbine
life is 19 years. In other words, the erosion rate is negligible.
Figure 24 shows the life of the turbine with one stage of granular-
bed filters with a performance equal to that of a conventional low-
temperature fabric filter. The turbine life here is appreciably
increased over that of the previous two cases. The curves are rela-
tively steep, especially so near the zero boundary layer thickness, but
with the erosiveness that Reference 8 assumed, the turbine life is
approximately two years. With two stages of granular-bed filters, how-
ever, the erosion rate is essentially zero since the particulate concen-
tration is zero when calculated to six decimal places.
If we assume that replacing the turbine blading would be con-
sidered an operating and maintenance (O&M), expense, we can calculate
it in the form of a cost of electricity (COE) in units of mills/
kWh. Based on information from the Westinghouse Gas Turbine Division,
the installed cost of a complete change of turbine blading is approxi-
mately $2 million per turbine. (No charge is included for plant down
time. We have assumed for purposes of this study that blade changes
would be made during normal maintenance periods and not as a result of
forced outages.) As shown in the EGAS report,H the power output of a
pressurized fluidized-bed power plant with two W501 gas turbines is
679,000 kW. Thus, the equation for the cost of electricity in mills/kWh
due to a blading change in terms of the time in hours between blading
changes is as follows:
COE » $2 x 106/(turbine-change) (2 turbines) (1000 mills/$)
CF kW rating (679>°00 ™ rating) (A hr/change)
Note that as the capacity factor (CF) goes down, the kW output goes down
so the COE goes up. On the other hand, as the CF goes down, the coal
59
-------
input and resultant particulate loading goes down so the life, A, goes
up. Since the life was calculated on the basis of full load, the
sensible approach is to take CF = 1.0. Thus, the equation becomes:
With this equation the lifetimes shown in Figures 22 through 24 can be
converted to COE. This result is shown in Figures 25 and 26 for two of
the three assumed granular bed filter performances. The calculation was
not made for the Rexnord performance since the turbine lives would be
too short to have practical application in a utility power plant. The
COE is plotted versus the boundary layer thickness, with parameters of
erosiveness for one stage of granular-bed filters. The A COE associated
with a two-stage, granular-bed system has been calculated and is also
shown in Figures 25 and 26.
Calculations have been made to estimate the cost of granular-bed
filters for the gas flow conditions of this power plant. Each module
handles a volume flow rate of 15.72 m3/s (33,300 acfm) , which results in
a filter pressure vessel diameter of 7.6 m (25 ft), if we assume a
design face velocity of 15.2 m (50 ft/mi n) , a face area per filter ele-
ment of 0.37 m^ (4 ft^), and a plan area per element of 0.26 m^
(2.8 ft^) (which includes the open flow area around each element). The
cost of a 7.6 m (25 ft) diameter, granular-bed filter module for opera-
tion at 982°C (1800°F) and 1013 kPa (10 atm) has been estimated on the
following basis :
Cost base - mid-1975 $
Field labor - 51% of direct installation costs
Professional services - 10%
Escalation - constant dollars
Interest during construction - 10%
Construction time - 5 years
60
-------
\ I I 1 1 1 1
Erosiveness Compared
to SiC Particles
^ Cost Due to
Second Stage
Curve 689720-A
(2) (4) (6) (8) (10) (12) (14) (16) (18) (20) (22)
1 102 152 203 254 305 356 406 457 508 559
Boundary Layer Thickness, urn (mils)
Figure 25 - Cost of Electricity Increments due
to Turbine Blade Replacement Using
Granular-Bed Filters (performance
based on granular-bed efficiency)
3.0
2.8
2.6
2.4
2.2
2.0
"
1.6
o
~
o 0.8
0.6
0.4
0.2
I I I I
^ Cost Due to
-*— Second Stage
Erosiveness Compared
to SiC Particles
I
0 (2) (4) (6) (8) (10) (12) (14) (16) (18) (20) (22)
51 102 152 203 254 305 356 406 457 508 559
Boundary Layer Thickness, M m (mils)
Figure 26 - Cost of Electricity Increments due
to Turbine Blade Replacement Using
Granular-Bed Filters (performance
based on fabric-filter efficiency)
-------
The resulting cost for a 7.6 m (25 ft) single-stage, granular-bed
filter module is $1,850,000.
The equation for the cost of electricity due to capital expense is
COE - (yearly cost of capital, $/$-yr) (capital cost. $) (1000 mills/$)
(capacity factor) (8760 hrs/yr) (plant power, kW)
The incremental COE due to the incremental cost of two stages compared
to one stage is calculated to be 1.08 mills/kWh.
CONCLUSIONS
1. This analysis provides a basis for evaluating the economic
advantage of improving filter performance rather than replacing
turbine blades.
2. More accurate experimental information on particulate erosive-
ness and staged-bed filter performance is required before a
definitive comparison can be made.
3. Although the cost of staged filters is substantial, the differ-
ence between this cost and the cost of blade replacement in the
case of least frequent blade replacement (Figure 25) is only
about one half mill/kWh, which indicates that rather large
capital costs can be tolerated for efficient filter systems.
RECOMMENDATIONS
1. Accelerate development of granular-bed filters and alternative
filter concepts designed for gas turbine applications.
2. Continue experimental studies to investigate particulate ero-
sion rates of gas turbine blading.
62
-------
8. INDIRECT AIR-COOLED PRESSURIZED FLUIDIZED-BED COMBUSTION
CONCEPT SYSTEMS EVALUATION
INTRODUCTION
A gas turbine cycle combustor using PFBC with indirect heating of
pressurized air in immersed tubes is being investigated by Curtiss
Wright under DOE funding.1° We compared the performance and cost of
energy of a combined-cycle plant using this configuration and those of a
combined-cycle plant using an adiabatic fluidized-bed combustor. We
considered two configurations of the partially indirectly heated system:
one with a CBC and one without. In each case we selected the amount of
excess air that would give an overall carbon loss equivalent to 1 per-
cent of the energy in the coal.
Since environmental concerns are of primary importance to EPA, this
study included an assessment of the effect of the partially indirectly
heated concept on pollutant emissions (particulates, SOX, NOX, products
of incomplete combustion, and solid wastes) compared to the adiabatic
PFBC System.
This technoeconomic study was carried out in late 1976 and early
1977 and reflects the FBC technology and the economic situation that
obtained in that time period. No attempt has been made to update the
results.
BACKGROUND
The products of combustion from PFBC of coal are passed through a
gas turbine expander. The cost of the particulate removal equipment
required to clean the combustion products well enough to avoid problems
with erosion of, corrosion of, and deposition on the expander parts is
expected to be significant. This is particularly true of gas turbines
63
-------
with an adiabatic fluidized-bed combustor, where air equivalence ratios
of approximately 3 prevail (200% excess air), since the cost of the par-
ticulate removal equipment is roughly proportionate to the volume of the
combustion products and the ratio of combustion products to coal is
high.
In the partially indirectly heated gas turbine combustor concept
the volume of the combustion products that must be cleaned is reduced
substantially by using the minimum amount of combustion air and heating
the balance of the gas turbine airflow indirectly with tubes submerged
in the fluidized bed. This indirectly heated air is then mixed with the
combustion products after they have been cleaned, and this mixed stream
constitutes the flow to the gas turbine expander. There is a trade-off,
therefore, between the cost of the particulate removal equipment and the
cost of the heat transfer surface required for indirectly heating the
air that bypasses the combustor.
Description of Systems Evaluated
The coal-fired power systems that were evaluated and compared in
this study are:
• Base Case - A combined-cycle system with an adiabatic
fluidized-bed combustor and in situ desulfurization
• Alternative Case I - Partially indirect heating with a CBC
• Alternative Case II - Partially indirect heating without a
CBC.
Performance calculations were made for each of these configurations
with Ohio Pittsburgh No. 8 seam coal with 3 percent moisture.
The gas turbine design conditions were as follows:
Ambient air conditions - International Standards Organization (ISO)
Compressor airflow - 345 kg/s
64
-------
Compressor pressure ratio - 10
Compressor isentropic efficiency - 0.853 (polytropic efficiency
W 0.89)
Expander isentropic efficiency - 0.927 (polytropic efficiency
W 0.90)
Temperature drop due to heat transfer between combustion products
and combustion air - 8°C.
The fluidized-bed combustor design conditions common to all cases
were as follows :
Primary bed temperature - 1010°C (value used in EGAS17)
Coal size - <6.4 mm
Dolomite size - <4.8 mm
Ca/S atom ratio - 1.5
Superficial velocity -^1.5 m/s
Maximum bed depth - 4.6 m
Pressure loas (including particulate removal equipment) - 7.5%.
The design conditions for the heat recovery steam generator (HRSG)
were as follows :
Type - unfired
Pinch - 22.2°C
Offset - 2.8°C.
The Base Case was previously treated in Reference 18 in a somewhat
different configuration. The configuration used in this study is shown
in Figure 27. A Ducon cyclone separator was used for the first stage of
particulate removal, a Ducon granular-bed filter for the second stage.
Grade efficiency plots for these components are given in Appendix A.
The fluidized-bed design conditions specific to the Base Case were as
follows:
Excess air - 237 percent
Superficial velocity - 1.5 m/s
Bed depth - 2.0 m
65
-------
7
Back-flushing
.,
Air
RuMized-
Bed
Combustor
Spent
Sorbent
5. )681M5
Granular-Bed
l—, Filter
Generator
Heat Recovery
Steam Generator
Steam Turbine
Generator Set
Figure 27 - Combined-Cycle System Utilizing Fluidized-Bed
Combustion (Base Case)
Combustion losses
Incomplete combustion - 0.4
Losses to atmosphere - 0.8
Sensible heat in solids - 1.25
Desulfurization reactions - Q.i
Total 2.55 percent
Overall combustion efficiency 97.45 percent.
With allowances for thermal losses from the ductwork, the temperature of
the gas at the gas turbine expander inlet was 996°C, and the gas turbine
expander cooling airflow was 8.3 percent.
66
-------
The Alternative Case I configuration is shown in Figure 28. A
Ducon cyclone separator was used for the primary bed effluent, a Tan Jet
cyclone separator for the CBC effluent, and a Ducon granular-bed filter
for the final separator. Grade efficiency plots for these separators
are given in Appendix A. The fluidized-bed design conditions specific
to Alternative Case I are as follows:
Primary bed
Excess air - 0 percent
Bed depth - 4.6 m
Superficial velocity - 1 m/s*
Carbon losses - 10 percent of equivalent energy in coal
CBC bed
Excess air - 0 percent
Bed depth - 4.6 m
Superficial velocity - 0.85 m/s
Carbon losses - 10 percent of input
Temperature - 1010°C
Combustion losses
Incomplete combustion
Losses to atmosphere
Sensible heat in solids
Desulfurization reactions •
Total
Overall combustion efficiency - 96.85 percent.
The effectiveness of the submerged heat exchangers in the primary
and CBC beds for indirectly heating part of the gas turbine working
fluid were assumed to be 85 percent. This assumption gives an air
outlet temperature of 907°C from both the primary and CBC bed heat
exchangers. After the combustion products and the indirectly heated air
*Superficial velocity reduced below 1.5 m/s nominal design value to
satisfy limits on maximum bed depth.
67
-------
Dwa. 1701B60
Primary
Bed
Cyclone \i
Fluidized Bed
Combustor
Back-flushing
Air
Granular -
Bed
Filter
Coal&
Sorbent
Ash-Char.
& Sorbent
Fines
Steam Turbine/
Generator Set
Booster
Compressor
Generator
Heat
Recovery
Steam
Generator
Figure 28 - Combined-Cycle System Utilizing Fluidized-Bed
Combustion with Indirect Heating of Part of
the Working Fluid (Alternative Case I - CBC)
68
-------
have been mixed, and radiation losses and heat transferred between the
hot products and the combustion air have been allowed for, the gas tur-
bine expander inlet temperature is only 928°C.
The Alternative Case II configuration is shown in Figure 29. This
configuration is obviously considerably less complicated than that with
the CBC shown in Figure 28. A Ducon cyclone was used for the first-
stage separator, a Ducon granular-bed filter for the second stage.
Grade efficiency plots for these separators are given in Appendix A.
The fluidized-bed design conditions specific to Alternative Case II were
as follows:
Excess air - 60 percent
Superficial velocity - 1.5 m/s
Bed depth - 3.3 m
Combustion losses
Incomplete combustion - 1.00 percent
Losses to atmosphere - 0.80
Sensible heat in solids - 1.25
Desulfurization reactions - 0.10
Total 3.15
Overall combustion efficiency - 96.85 percent.
The effectiveness of the heat transfer surface submerged in the bed
was again assumed to be 85 percent, giving an air temperature out of the
heat exchanger of 907°C. After the combustion products have been mixed
with the indirectly heated air, and allowances for losses to the atmo-
sphere and heat transferred from the hot combustion products to the com-
bustion air have been made, the temperature at the gas turbine expander
inlet is 942°C.
Alternative Case I is probably not a practical configuration
because of the likelihood of severe corrosion of the immersed air heater
tubes in a bed with zero excess air. It does, however, represent one
boundary of the design spectrum for partially indirectly heated systems
(the other boundary being the Base Case).
69
-------
Dug. 7719A58
Back-flushing
Air
Fluidized-Bed
Combustor
Spent
Sorbent
Coa1& I
Sorbent I
Granular-Bed
—, Filter
Booster
Compressor
Steam Generator
Steam Turbine/,
Generator Set
Figure 29 -
Combined-Cycle System Utilizing Fluidized-Bed Combustion
with Indirect Heating of Part of the Working Fluid
(Alternative Case II - no CBC)
RESULTS OF STUDY
A summary of the plant performance for the cases studied is given
in Table 12.
All three of these configurations have heat rates that are appre-
ciably better than that for a conventional steam plant with FGD (i.e.
10,475 kJ/kWhl?) with the admittedly optlmistic bed teraperature Qf
1010°C. If limitations on bed temperature per se and/or hot gas duct
temperature require the bed temperature to be reduced significantly, the
heat rates for all configurations will be increased uniformly. If lim-
itations on the maximum metal temperature of the in-bed air tubes
require the teraperature of the indirectly heated air to be reduced sig-
nificantly, the heat rates of the partially indirectly heated configu-
rations will increase relative to that for the Base (adiabatic) Case.
70
-------
Table 12
SUMMARY OF PLANT PERFORMANCES
Case
Gas Turbine
Output,
MW/G.T.
Module
Steam
Turbine
Output,
MW/G.T.
Module
Total
Electrical
Output,
MW/G.T.
Module
Coal Feed
Rate,
ton/hr/
G.T.
Module
Heat Rate,
kJ/kWh
?ase
Adiabatic)
Alternative 1
Alternative II
73.8
66.3
67.8
34.1
27.4
28.7
107.9
93.7
96.5
36.2
32.7
33.4
9148
9587
9504
Plot plans of a single gas turbine module for the Base Case, Alter-
native Case I, and Alternative Case II are shown in Figure 30 through
32. Table 13 summarizes the plant design configurations for plants with
a nominal capacity of 400 MW.
Estimates of the capital cost of nominally 400 MW plants for each
of the three configurations were made on the basis of the following
assumptions:
Cost base - 1st quarter of 1976
Construction time - 4 years
Indirect construction costs - 13.5 percent of the total direct
costs*
Professional services - 10% of the direct plus indirect costs
Contingency - 10 percent
Escalation rate - 6-1/2 percent
Interest during construction - 10 percent
Expenditure rate - S curve supplied by NASA for use in the
EGAS studyl7
*Equivalent to ~50 percent of direct installation costs.
71
-------
Dwp. 7681 13
Compressor FBC
o
Air
Granular-Bed
Filters
Generator
Figure 30 - Plot Plan of Single Gas Turbine Module for Base Case
Compressor FBC
Granular-Bed
Filters
0 ". 'S31A1?
Generator
Expander
Carbon
Burnup
Cell
Figure 31 - Plot Plan of Single Gas Turbine Module for
Alternative Case 1
72
-------
Dwr. 7681A14
Granular-Bed
Filters
Compressor
FBC
Cyclones
Oi
<>
Expander ,
Generator
Figure 32 - Plot Plan of Single Gas Turbine Module for
Alternative Case II
Cost escimates for equipment manufactured by Westinghouse, such as
the gas turbines, heat recovery steam generators, and steam turbines,
were obtained from cost correlations supplied by the pertinent Westing-
house divisions during EGAS.1'' Cost estimates for high-temperature par-
ticulate removal equipment were based on information obtained from
equipment suppliers during EGAS. Cost estimates for FBC equipment,
solids feeding equipment, and in-bed heat transfer surface were made
using procedures that originated in the Evaluation of the Fluidized Bed
Combustion Processes^ and were used in EGAS.
Summaries of the capital cost estimates for the Base Case, Alter-
native Case I, and Alternative Case II are given in Tables 14 through
16. The only costs that vary significantly with system configuration
are those for the FBC modules and the gas-cleaning equipment. While the
variations in the cost of these two components are opposite (e.g., the
73
-------
Table 13
SUMMARY OF PLANT DESIGN CONFIGURATIONS
Configuration
Case
Base
Alternative I
Alternative II
Capacity, MW 431.6
No. of Gas Turbines 4
Combustion Modules
Number 8
Diameter, m 3.8
Height, m 25.6
Beds/module 4
Bed depth, m 1.98
CBC Modules
Number —
Diameter, m —
Height, m
Beds/module —
Bed depth, m
Ist-Stage Separator Modules
Volumetric flow, am-Vs 29.3
Number 16
Diameter, m 0.76
Height, m 3.0
CBC Separator Modules
Volumetric flow, am^/s
Number
Diameter, m
Height, m
2nd-Stage Separator Modules
Volumetric flow, am-Vs 14.6
Number 32
Diameter, m 7.77
Height, m 8.84
No. of HRSG Modules 4
No. of Steam T-G Modules 1
374.8
4
8
3.65
39
4
4.57
4
3.65
10.97
1
4.57
16.1
8
1.0
4.0
3.21
4
2.6
3.05
17.9
8
8.23
9.3
4
1
386.0
4
8
3.65
41
5
3.35
25.3
8
1.3
5.25
12.7
16
7.01
8.23
4
1
74
-------
Table 14
CAPITAL COST ESTIMATE FOR BASE CASE
(4 gas turbines)
Item 106 $
1.00 Land and Land Rights 4.600
2.00 Structures & Improvements (on-site waste disposal) 13.845
3.00 Heat Rejection System 3.605
4.00 Material Handling and Storage 16.445
5.00 Energy Conversion
PFBC 2.381
Combustion air piping v 0.827
Transport air subsystem 1.546
Gas cleaning 41.341
Refractory-lined pipe 1.608
Refractory- and metal-lined pipe 2.368
Gas turbine/generator 36.542
Steam turbine/generator 7.606
HRSG 6.366
Subtotal 100.585
6.00 Auxiliary Mechanical Equipment 5.037
7.00 Auxiliary Electrical Equipment 9.614
8.00 Station Transmission Equipment 2.403
Total direct costs 156.134
Indirect construction costs (13.5% of total direct costs) 21.078
Subtotal 177.212
Professional services (10% of direct and indirect costs) 17.721
Subtotal 194.933
Contingency (10% of above) 19.493
Subtotal 214.426
Escalation during construction (6 1/2-4) (15.8% of above) 33.879
Interest during construction (10-4) (21.4% of above) 45.887
Total capitalization 294.193
75
-------
Table 15
CAPITAL COST ESTIMATE FOR ALTERNATIVE CASE I
Item 106
1.00 Land and Land Rights 4.144
2.00 Structures and Improvements (on-site waste disposal) 12.473
3.00 Heat Rejection System 3.108
4.00 Material Handling and Storage 14.815
5.00 Energy Conversion
PFBC 12.614
Combustion air piping 0.836
Transport air subsystem 1.452
Gas cleaning 13.408
Refractory-lined pipe 0.926
Refractory- and metal-lined pipe 3.032
Gas turbine/generator 36.540
Steam turbine/generator 6.400
HRSG 5.357
Subtotal 80.565
6.00 Auxiliary Mechanical Equipment 4.342
7.00 Auxiliary Electrical Equipment 8.288
8.00 Station/Transmission Equipment 2.072
Total direct costs 129.807
10.0 Indirect construction costs (13.5% of total direct costs) 17.524
Subtotal 147.331
11.0 Professional services (10% of direct and indirect costs) 14.733_
Subtotal 162.064
12.0 Contingency (10% of above) 16.206
Subtotal 178.370
13.0 Escalation during construction (6 1/2-4) (15.8%) 28.183
14.0 Interest during construction (10-4) (21.4%) 38.171
Total capitalization 244.724
76
-------
Table 16
CAPITAL COST ESTIMATE FOR ALTERNATIVE CASE I
Item 106 $
1.00 Land and Land Rights A.A55
2.00 Structures and Improvements (on-site waste disposal) 13.A08
3.00 Heat Rejection System 3.372
4.00 Material Handling and Storage 15.926
5.00 Energy Conversion
PFBC 10.09A
Combustion air piping 0.836
Transport air subsystem 1.517
Gas cleaning 18.583
Refractory-lined pipe 0.688
Refractory- and metal-lined pipe 3.200
Gas turbine/generator 36.SAO
Steam turbine/generator 6.AOO
HRSG 5.608
Subtotal 83.A66
6.00 Auxiliary Mechanical Equipment A.711
7.00 Auxiliary Electrical Equipment 8.992
8.00 Station/Transmission Equipment 2.2A8
Total direct costs 136.578
10.0 Indirect construction costs (13.5% of total direct costs) 18.A38
Subtotal 155.016
11.0 Professional services (10% of direct and indirect costs) 15.501
Subtotal 170.517
12.0 Contingency (10% of above) 17.052
Subtotal 187.569
13.0 Escalation during construction (6 1/2-A) (15.8%) 29.636
1A.O Interest during construction (10-A) (21.A%) A0.1AO
Total capitalization 257.3A5
77
-------
adiabatic system has the highest gas cleaning cost and the lowest com-
bustion cost), they are not equal, and the specific costs of the the
three configurations vary significantly, as shown in Table 17. The
adiabatic system is estimated to have the highest specific capital cost
and Alternative I the lowest.
Table 17
SPECIFIC COST COMPARISON
Case
Base
Alternative I
Alternative II
Specific Cost-$/kW
682
653
667
The cost of electricity was calculated for the three cases studied
on the basis of the following assumptions:
Life of plant - 30 years
Annual charge - 18 percent
Capacity factor - 65 percent
Fuel cost (1st quarter of 1976) - $0.95/GJ
Fuel escalation rate - 5%/year
O&M (including dolomite) -2.5 mills/kWh
Table 18 summarizes the energy costs for the three cases.
Table 18
COST OF ELECTRICITY SUMMARY, mills/kWh
Item
Capital
Fuel
O&M
Total
Base
21.6
21.5
2.5
45.6
Alternative I
20.6
22.6
2.5
45.7
Alternative II
21.0
22.3
2.5
45.7
78
-------
This table shows that the variations in specific capital costs of the
three systems are nearly balanced by the variations in fuel consumption
and that the spread in cost of electricity among the three cases is
about 0.2 percent. Since the uncertainties in estimating the cost of
the cost-variable components (the fluidized-bed combustor modules and
the gas cleaning equipment) are considered large as compared to the
spread in cost of energy, no significance can be attributed to this
spread in cost of energy.
The in-bed heat transfer surfaces used in this study were plain
tubes placed in horizontal array. No attempt was made to evaluate the
finned/finned-tubing concept that was proposed for the partially indi-
rectly heated cycle by Curtiss Wright.
PARTICULATE CONTROL/GAS TURBINE EXPANDER EROSION CONSIDERATIONS
The requirements for particulate removal from the combustion prod-
ucts of PFBC for economical gas turbine expander life are predicted to
be in excess of those for meeting emission limits.21 For that reason
this assessment of the merit of the partially indirectly heated concept
emphasizes the control of gas turbine expander erosion and deposition
rather than particulate emissions.
Erosion of and deposition on gas turbine expander parts are func-
tions of the concentration, size distribution, and physical properties
of the particles entrained in the working fluid. (They are also func-
tions of the size and design of the gas turbine expander, but considera-
tion of the latter aspect is outside the scope of this study.)
The particulate in the products of combustion from fluidized beds
with in situ desulfurization has three components: ash from the coal,
fines from the desulfurization sorbent, and unburned carbon.
The ash from FBC of coal consists of friable platelets that have an
erosivity substantially less than the fused cenospheres generated in the
combustion of pulverized coal. All of the ash in the coal is assumed to
79
-------
be entrained in the products of combustion, either as free ash or asso-
ciated with unburned carbon. Actually, small quantities of ash are car-
ried out with the coarse spent sorbent that is removed directly from the
fluidized bed. The size distribution of the free ash is based on data
given in Reference 22.
For this study we have assumed desulfurization using once-through
dolomite. The dolomite feed is single screened so it contains a sig-
nificant amount of fines that are elutriated from the bed almost immedi-
ately after the dolomite is injected into it. The amount of excess sor-
bent is small (50%) and the bed volume is large, so the residence time
of the coarse dolomite is long (~10 hr). As a result, there is a sig-
nificant amount of attrition and decrepitation of the coarse dolomite
and subsequent elutriation of the sorbent fines generated thereby. The
size distribution of the attrited and decrepitated sorbent fines was
also based on data given in Reference 22.
The quantity of unburned carbon (char) entrained in the combustion
products from a fluidized bed is primarily a function of bed temperature
and the amount of excess air. Bed depth, feed particle size and distri-
bution, and superficial velocity also are factors. Fluidized-bed com-
bustion efficiency is based on data given in Reference 23. The size
distribution of the char is assumed to be the same as that for the coal
feed for those particles having diameters smaller than the particle
whose terminal velocity is equal to the bed superficial velocity. The
char composition is assumed to be that of devolatilized high-volatility
bituminous coal.
The concentration and size distribution of particles in the dis-
charge of the particulate removal equipment are based on separate calcu-
lations for each of the three constituents. This individual calculation
is necessary because of the differences in the density of ash, spent
sorbent, and char, which have a significant effect on the performance of
centrifugal separation equipment.
80
-------
Appendix B contains detailed information on the concentration and
size distribution of ash, sorbent, and char particles at various stages
in the particulate removal subsystems for Alternative Case I. Summaries
of this information are given in Figures 32 through 34 for the Base
Case, Alternative Case I, and Alternative Case II. Table 19 compares
the particulate loadings at the gas turbine expander inlet for each of
these cases.
Table 19
EXPANDER INLET PARTICLE LOADING
Loading, g/sm^
Sorbent
Ash
Char
Base
0.0050
0.0030
0.00004
0.00804
Alternative I
0.0033
0.0031
0.00073
0.00713
Alternative II
0.0038
0.0025
0.00006
0.00636
This table shows that the amount of char is less than 1 percent of
the total particulate for the Base Case and for Alternative Case II and
about 10 percent for Alternative Case I, which has the CBC. Since there
is a good possibility that the fine carbon particles in the products
of combustion will be oxidized after the bypass air is added, the
amount of char shown for Alternative Case I is considered to be of no
significance.
The concentrations of ash in the alternative cases are within
about 15 percent of the value for the Base Case so the variance is
insignificant.
The only variation considered to be significant is that in the sor-
bent concentration, with the values for the alternative cases being 25
to 35 percent lower than that for the Base Case. Since the sorbent is
the most erosive constituent of the particulate, this difference might
81
-------
0.01
oo
to
1 -
•Z 10 -
I
90 -
99 -
99.99
Participate loading at Expander Inlet
Sorbent ttOOSOg/sm3 (a 0022 gr/scfl
a0030g/sm3 (0.0013gr/scf)
0.00004 g/sm3 (0.00002 gr/scft
0.00804 g/sm3 (0.00352 gr/scf)
° Char
o Ash
a Sorbent
10
10"
L I I I
1st Stage
Separator
Inlet
Particulate Loading at 1st*
Stage Separator Inlet
Sorbent & 17 g/sm3 (2.26 gr/scf)
Ash 3.02 g/sm3 (1.32 gr/scf I -
Coal 0.14q/sm3 (a 06 gr/scf I
8.33g/sm3 (3.64 gr/scf)
i i i r i i i i
101
102
10"
Particle Size-urn
Figure 33 - Summary of Particulate Loading and Size Distribution for Base Case
-------
Curve 68951 5-B
00
u>
0.01
o>
10
50
90
99 -
99.99
"I 1—TTT
a Char
o Ash
A Sorbent
Partlculate Loading
at Ejqpander Inlet
Sorbent 0.00332 g/sm3 (0.00148 gr/scf)
- Ash 0.00309 g/sm'
-------
be expected to significantly affect the life of the gas turbine expander
vanes and blades. Examination of the size distribution plots for the
sorbent particles entering the expander inlet for the three cases (Fig-
ures 33 through 35) shows, however, that 100 percent of the particles
are smaller than 10 pm and 50 percent are smaller than 2 ym. A recent
survey^ of turbine manufacturers indicated that 0.009 to 0.045 g/sm^ of
particles larger than 10 vm is tolerable. We conclude, therefore, that
none of the cases would have an erosion problem if, in fact, a granular-
bed filter or other device having the performance assumed becomes a com-
mercial reality.
If the performance assumed for the granular-bed filter cannot be
achieved in a practical separation device with the feed sorbent size
distribution used here, an excessive amount of particle larger than
10 ym may be present at the expander inlet. Since a substantial frac-
tion of the sorbent elutriated from the beds are fines present in the
sorbent feed, the use of double-screened sorbent would be expected to
reduce substantially the amount of sorbent in the combustion products
going to the gas turbine expander. This suggests that there may be a
trade-off between the cost of double-screened sorbent and the cost of
replacing gas turbine expander parts. Consideration must be given, how-
ever, to the effect of the use of double-screened sorbent on the effec-
tiveness of its desulfurization.
ENVIRONMENTAL CONSIDERATIONS
We have estimated the emission performance of the three configura-
tions using available process models and emission data for PFBC. Using
a Ca/S atom feed ratio of 1.5 for all three configurations, based on an
average activity dolomite such as Tymochtee, we determined the sulfur
removal efficiency. In addition, we estimated the NOX, CO, and particu-
late emissions and projected the resulting solid waste product rate.
These estimates are shown in Table 20.
Table 20 indicates that the three configurations would satisfy all
of the current NSPS for coal-fired boilers : all S02 emission of
84
-------
Curve 690376-6
00
Particulate Loading
at Expander Inlet:
Sorbent 0.00375g/sm3 (0.00164 gr/scf)
Ash 0.00247 g/sm3 (0.00108 gr/scf)
Char 0.00006 g/sm3 (0.00003 gr/scf)
0.00630 g/sm3 (0.00275 gr/scf)
Ist-Stage
Cyclone Inlet
0 Char
Ash
Sorbent
Sorbent & 874 g/sm
(3.877 gr/scf)
(Z 581 gr/scf)
(0.301 gr/scf)
99.99
10
10*
Particle Size - \m
103
Figure 35 - Summary of PartiCulate Loading and Size Distribution for Alternative Case II
-------
Table 20
ENVIRONMENTAL COMPARISON OF THE CONFIGURATIONS
Configuration
Base Case
Alternative
Case I
Alternative
Case II
S02,
ng/J
(Ib/MBtu)
284
(0.66)
116
(0.27)
116
(0.27)
NO
ng/J
(Ib/MBtu)
215
(0.5)
86
(0.2)
150
(0.35)
CO,
p£m
50
300
100
Particulate,
ng/J
(Ib/MBtu)
7.3
(0.017)
7.7
(0.018)
6.5
(0.015)
Solid kg/kg
Waste, kg/kg
(Ib/lb coal)
0.38
(0.38)
0.38
(0.38)
0.38
(0.38)
516 ng/J (1.2 Ib/MBtu), an NOX emission of 301 ng/J (0.7 Ib/MBtu), and a
particulate emission of 43 ng/J (0.1 Ib/MBtu). The base configuration
would achieve 90 percent sulfur removal at the selected Ca/S feed ratio
of 1.5, while the two options would achieve 96 percent sulfur removal
because of significantly longer gas residence in the fluidized-bed com-
bustor. Sulfur losses from the CBC have been accounted for in the esti-
mate for Alternative Case I. The desulfurization efficiency for the
Base Case could be increased at a modest cost, however, by increasing
the bed depth.
The oxides of nitrogen would vary considerably among the configura-
tions because of the variation in excess air levels. The low excess air
level in Alternative Case I would result in a low NOX emission but a
relatively large CO emission. Unburned hydrocarbon emissions may also
be significant in Alternative Case I, but little information is avail-
able to make such a projection.
Particulate emissions would be comparable for the three configura-
tions and much less than the environmental limit if the particulate con-
trol system selected for turbine protection were to be used. The solid
86
-------
waste produced by the three configurations would be similar in rate of
production and in physical/chemical properties. Small differences could
exist in the particle size distributions of the waste materials, but the
resulting environmental factors (e.g., leaching characteristics) would
be expected to be very similar. The solid waste production could be
reduced slightly in the alternative cases by operating at the smaller
Ca/S feed ratios required to yield the 90 percent sulfur removal effi-
ciency of the base configuration, or a Ca/S ratio of about 1.2 for
Alternative Case 1 and 1.3 for Alternative Case II. The solid waste
would thus be reduced to 0.34 and 0.35 kg/kg coal, respectively. This
reduction in ratio could also result in a reduction in particulate emis-
sions for the alternative cases.
The emissions from the three configurations would also be sensitive
to the properties of the coal and sorbent selected for operation. The
coal ash and sulfur content directly affect the fluidized-bed combustor
control requirements, while dolomites vary significantly in sulfur
removal activity, attrition resistance, and trace element content.
CONCLUSIONS
As a result of this evaluation of partially indirect heating of the
working fluid for a gas turbine with a pressurized fluidized-bed com-
bustor we conclude that:
• With the projected granular-bed filter performance, gas
turbine expander erosion problems are not anticipated for
either the base (adiabatic) configuration or the two par-
tially indirectly heated configurations. If, however, the
particulate removal performance projected herein cannot be
attained on a commercial basis, both of the partially
indirectly heated alternatives would have a potential for
significantly larger expander life because of lower sor-
bent fines concentration in the gas entering the expander.
87
-------
• With the replacement cost of gas turbine expander parts
assumed to be uniform, the estimated costs of electricity
for the three configurations considered are essentially
equal. The variations in capital costs among the three
cases would be balanced by variations in heat rate*
• Acceptable environmental performance is indicated for all
three configurations. The partially indirectly heated
alternatives indicate a potential for environmental per-
formance significantly better than that of the Base Case.
• All three of the configurations considered have a poten-
tial for heat rates appreciably better than that of a
conventional coal-fired steam plant with flue gas
desulfurization.
88
-------
9. REFERENCES
1. Newby, R. A., and D. L. Keairns, Alternatives to Calcium-Based 862
Sorbents for Fluidized-Bed Combustion: Conceptual Evaluation,
report to EPA, Westinghouse Research and Development Center,
Pittsburgh, PA, January 1978, EPA-600/7-78-005.
2. Newby, R. A., N. H. Ulerich, E. P. O'Neill, D. F. Ciliberti, and
D. L. Keairns, Effect of S02 Emission Requirements on Fluidized-Bed
Combustion Systems; Preliminary Technical/Economic Assessment.
Report to EPA, Westinghouse Research and Development Center,
Pittsburgh, PA, August 1978, EPA-600/7-78-163, NTIS PB 286 811/
7 ST.
3. Keairns, D. L., et al. Fluidized Bed Combustion Process
Evaluation - Phase 11 - Pressurized Fluidized Bed Coal Combustion
Development, Report to EPA, Westinghouse Research Laboratories
Pittsburgh, PA, September 1975, EPA-650/2-75-027c, NTIS PB 246-116.
4. Treybal, R. E., Mass-Transfer Operations, New York: McGraw-Hill
Book Co., Inc.; 1955.
5. Riesenfeld, F. C., A. L. Kohl, Gas Purification, Second Ed.,
Houston: Gulf Publishing Co.; 1974.
6. Perry, R. H., C. H. Chilton, Chemical Engineer's Handbook, 5th Ed.,
New York: McGraw-Hill Book Co.; 1973.
7. Hoke, R. C., Emissions from Pressurized Fluidized Coal Combustion.
Proceedings of the 4th International Conference on Fluidized Bed
Combustion, December 1975, McLean, VA: The Mitre Corporation;
1976; and Monthly Reports to EPA.
8. Advanced Coal Gasification System for Electric Power Generation,
Quarterly Progress Reports for ERDA, Westinghouse Electric Corpora-
tion, Pittsburgh, PA, October 1975 and January 1976, Contract No.
E(49-18)-1514.
89
-------
9. Smith, J., et al., Bureau of Mines Coal-Fired Gas Turbine Project,
RI 6920, U. S. Department of Interior, Bureau of Mines, 1967.
10. Atkin, M. L., and G. A. Duke, The Operation of a Modified Ruston
Hornsby Gas Turbine on a N.S.W. Bituminous Coal, Aeronautical
Research Laboratories Report 133, Department of Supply, Australian
Defense Scientific Service, 1971.
11. Ulke, A., and W. T. Rouleau, The Effect of Secondary Flows on Tur-
bine Blade Erosion, ASME Paper 76-GT-74.
12. Beecher, D. T., et al., Energy Conversion Alternatives Study -
Westinghouse Phase II Final Report, Vol. Ill, Summary and
Advanced Steam Plant with Pressurized Fluidized-Bed Boiler,
November 1, 1976, Contract NAS 3-19407.
13. McCain, J. D. , Evaluation of Rexnord Gravel Bed Filter, Report to
EPA, EPA-600/2-76-164, NTIS PB 255 095.
14. Spagnola, H., Operating Experience and Performance at the Sunbury
Baghouse, Symposium on Particulate Control in Energy Processes,
San Francisco, CA, May 11-13, 1976.
15. Goldstein, S., Modern Developments in Fluid Dynamics, Volume II,
New York: Oxford University Press; 1938.
16. Moskowitz, S., and G. Weth, Pressurized Fluidized Bed Pilot Plant
for Production of Electric Power using High Sulfur Coal. Proceed-
ings of the 12th Intersociety Energy Conversion Engineering Confer-
ence, Washington, DC, August 28 - September 2, 1977, pp. 696-703.
17. Beecher, D. T., et al., Energy Conversion Alternatives Study -
Westinghouse Phase II Final Report, Vol. I; Summary and Combined
Gas-Steam Turbine Plant with Integrated Low-Btu Gasifier, Contract
NAS 3-19407, November 1, 1976.
18. Keairns, D. L., D. H. Archer, R. A. Newby, E. P. O'Neill,
E. J. Vidt, Evaluation of the Fluidized Bed Combustion Process,
Vol. I. Report to EPA, Westinghouse Research Laboratories, Pitts-
burgh, PA, December 1973, EPA-650/2-73-048a, NTIS PB 231-162.
19. Coal-Fired Power Plan: Capital Cost Estimates, Report to EPRI,
Bechtel Power Corporation; January 1977, EPRI AF-342.
90
-------
20. Archer, D. H., D. Lt Keairns, J. R. Hamm, et al., Evaluation of the
Fluidized Bed Combustion Process, Vol. I, II, and III. Report to
the Office of Air Programs, EPA, Westinghouse Research Labora-
tories, Pittsburgh, PA; December 1971, NTIS PB 211-494, 212-116,
AND 213-152.
21. Menguturk, M., and E. F. Sverdrup, Tolerance of a Large Electric
Utility Turbine to Erosion Damage by Coal Gas Ash Particles, ASTM
Symposium of Erosion, Vail, CO, October 24-26, 1977.
22. Merrick, D. , and J. Highley, Particle Size Reduction and Elutrta-
tion in a Fluidized Bed Process, Recent Advances in Pollution Con-
trol, AIChE Symposium Series No. 137, Vol. 70; p. 366-78.
23. Hake, R.C., et al., Studies of the Pressurized Fluidized-Bed
Combustion Process, Report to EPA, Exxon Research and Engineering Co.,
Linden, NJ, September 1977, EPA-600/7-77-107.
24. CFCC Development Program - Advanced Clean-up Hardware Performance
Guidelines for Commercial Plant (Task 4.1.2), March 1978,
FE-2357-37.
91
-------
APPENDIX A
GRADE EFFICIENCIES FOR PARTICIPATE REMOVAL EQUIPMENT
92
-------
Dug. 6429A58
Final Stage
Coal
Sorb
Primary
Bed
V
Oucon
Primary
Cyclone
GBF
Ash and Spent
-"-* Sorbent
Air
C
B
C
T
Air
I Tan Jet
ICBC Cyclone
Ash&
Attrited Sortent
Figure A-l - Schematic of Particle Control System
Curve 694490-A
Specific Gravity
a 01
0.1 -
181.35.88
i
1
I
99.9
99.99
Figure A-2 - Grade Efficiency of Primary Cyclone Separator
93
-------
Curve 687623-B
0.01
0.1
1
10
30
30
70
90
98
99
99.9
Specific Gravity
2.88 .88.64
99.99
Tan Jet
TM
T= 9Z7°C(1700°F)
P= 1013kPa(10Atm)
i ill
i i r
i i i
.10
1.0
10.0
100
Dp.
Figure A-3 - Grade Efficiency Curve for Tan Jet
-------
Curve 687622-A
0.01
o.it-
10
30
*
d 50
o
1 70
90
98
99
99.9
99.99
0.10
1 1 IT
i i I
LO
Particle Size, pm
10.0
Figure A-4 - Grade Efficiency of Granular-Bed Filter
95
-------
APPENDIX B
PARTICLE SIZE DISTRIBUTION AT PERTINENT STATIONS IN
PARTICULATE REMOVAL SUBSYSTEM FOR ALTERNATIVE CASE I
Analyses of particle size distributions and concentration were made
at each pertinent station in the particulate removal subsystem for each
case studied. Plots of the size distribution at each station for
Alternative Case I are included in this appendix. Figure B-l gives the
size distribution plots for sorbent particles, Figure B-2 gives those
for ash particles, and Figure B-3 gives those for char particles.
The concentrations of sorbent, ash, and char particles at each of
the pertinent stations are given in Table B-l.
Table B-l
Solids Concentrations, g/sm^
Station So
rbent Ash
Primary Bed Exit 10.35 10.92
1st Stage Separator Exit 0.55 0.69
Carbon Burnup
CBC Separator
GBF Inlet
GBF Exit
Cell Exit 9.52 8.86
Exit 1
.72 1.31
0.67 0.76
0.011 0.0085
GT Expander Inlet 0.0034 0.0026
Char
8.38
0.21
1.08
0.021
0.20
0.0025
0.0007
Total
29.64
1.44
19.46
3.04
1.63
0.022
0.0067
96
-------
Curve 689518-B
0.01
VO
.
£
10
50
90
99
99.99
—i—r—r~i 1 1—r—r
Case I - Sorbent
o Feed to Primary Bed
a To ht-Stage Separator
o From Ist-Stage Separator
? To CBC
o From CBC to CBC Separator
& From CBC Separator
a To Granular Bed Filter
o From Granular Bed Filter
10
Particle Size, urn
Figure B-l - Size Distributions of Sorbent Particles for Alternative Case I
-------
Curve 689516-B
VO
oo
0.01
.s- 10
2 50
I
•E
i »
99
99.99
—i—r—r-| 1
Case I - Ash
o To Ist-Stage Separator
a From Ist-Stage Separator
o From C8C Cyclone
a To Granular Bed Filter
o From Granular Bed Filter
~i~ n
I I I I I I I I I 1
«-l
10
10'
Particle Size, Mm
10
io3
Tr-
I 1 I
Figure B-2 - Size Distributions of Ash Particles for Alternative Case I
-------
Curve 689517-6
0.01
VO
.2*
o>
10
50
90
99 -
99.99
I I I I I I
Case I - Char
o Coal Feed to Primary Beds
D To Ist-Stage Separator
o From Ist-Stage Separator
a To CBC
o From CBC to CBC Separator
o From CBC Separator
h To Granular Bed Filter
» From Granular Bed Filter
10
10
Particle Size,
Figure B-3 - Size Distributions of Char Particles for Alternative Case I
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-80-015d
2.
3. RECIPIENT'S ACCESSION NO.
AND SUBTITLE Experimental/Engineering Support for
EPA's FBC Program: Final Report
Volume 4. Engineering Studies
S. REPORT DATE
January 1980
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
J.R.Hamm, D. F.Ciliberti, R.W.Wolfe,
R.A.Newby, and D. L.Keairns
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
10. PROGRAM ELEMENT NO,
Westinghouse Research and Development Center
1310 Beulah Road
Pittsburgh, Pennsylvania 15235
INE825
11. CONTRACT/GRANT NO.
68-02-2132
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PER
Final; 12/75 - 12/78
ERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES JERL-RTP project officer is D. Bruce Henschel, Mail Drop 61,
919/541-2825. EPA-600/7-78-163 also relates to this work.
16. ABSTRACT The report gives results of engineering studies addressing several aspects
of fluidized-bed combustion (FBC) system design and performance, as applied to
coal. It reviews an evaluation of the impact of SO2 emission requirements on FBC
system performance and cost. Stringent SO2 emission requirements can be satis-
fied economically if design and operating parameters are properly selected. An
alternative SO2 control concept for pressurized FBC (PFBC), pressurized scrub-
bing of the products of combustion with water, is evaluated. The concept is not eco-
nomically competitive because of reduced plant efficiency and the need for recuper-
ative heating. A potential reduction in solid waste is realized with the concept, but
the SO2 control efficiency may be limited. An evaluation of PFBC, examining the
technical and economic trade-offs between the level of particulate control achieved
and the frequency of gas-turbine blade replacement, is described. The evaluation
incorporates models of PFBC particulate carry-over, particulate control device
efficiency, and turbine erosion. Also, an indirect air-cooled PFBC concept is com-
pared with other PFBC concepts. The indirect air-cooled concept provides signifi-
cant particulate control advantages over the adiabatic combustor PFBC concept,
resulting in about 4% lower plant efficiency and 1% higher cost of electricity.
7.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
Pollution Dust
Combustion Gas Turbines
Fluidized Bed Processing
Coal
Sulfur Oxides
Scrubbers
Pollution Control
Stationary Sources
Particulate
13B 11G
21B 13G
13H,07A
21D
07B
131
8. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
111
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
100
------- |