EPA-600/7-78-039
U.S. Environmental Protection Agency Industrial Environmental Research EPA-600/7-/
Office of Research and Development Laboratory
Research Triangle Park, North Carolina 27711 MaTCh 1978
REGENERATION OF
CALCIUM-BASED SO2 SORBENTS
FOR FLUIDIZED-BED COMBUSTION
ENGINEERING EVALUATION
Interagency
Energy-Environment
Research and Development
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-600/7-78-039
March 1978
REGENERATION OF CALCIUM-BASED SO2
SORBENTS FOR FLUIDIZED-BED
COMBUSTION: ENGINEERING
EVALUATION
by
R. A. Newby, S. Katta, and D. L. Keairns
Westinghouse Research and Development Center
1310 Beulah Road
Pittsburgh, Pennsylvania 15235
Contract No. 68-02-2132
Program Element No. EHE623A
EPA Project Officer: D. Bruce Henschel
Industrial Environmental Research Laboratory
Office of Energy, Minerals and Industry
Research Triangle Park, N.C. 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 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 systems under contract to the Industrial Environmental
Research Laboratory, U. S. Environmental Protection Agency (EPA), at
Research Triangle Park, NC. The contract scope includes atmospheric
and pressurized 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
system studies (e.g., sorption kinetics, regeneration, attrition,
modeling), alternative sulfur sorbents, nitrogen oxide emissions, par-
ticulate emissions and control, trace element emissions and control
spent sorbent and ash disposal, and system evaluation (e.g., the impact
of new source performance standards on fluidized-bed combustion system
design and cost).
The report contains the results of work, defined and completed
under the environmental control task using calciumbased sorbents
that was carried out from December 1975 to January 1977. Results from
work carried out by Westinghouse or reported by other investigators
after January 1977 are not assimilated into this task report. The
work reported represents an extension of prior work completed by
Westinghouse under contract to EPA. Results from this prior work
on fluidized-bed combustion include:
Assimilation of available data on fluidized-bed combustion,
including sulfur dioxide removal, sorbent regeneration,
sorbent attrition, nitrogen oxide minimization, combustion
efficiency, heat transfer, particle carry-over, boiler tube
corrosion/erosion fouling, and gas-turbine erosion/corrosion
deposition
iii
-------
Assessment of markets for industrial boilers and utility power
systems
Development of designs for fluidized-bed industrial boilers
Development of designs for fluidized-bed combustion utility
power systems: atmospheric-pressure fluidized-bed combustion
boiler-combined cycle power systems, adiabatic fluidized-bed
combustion-combined cycle power systemsincluding first- and
second-generation concepts
Preparation of a preliminary design and cost estimate for a
30 MW (equivalent) pressurized fluidized-bed combustion
boiler development plant
Assessment of the sensitivity of operating and design parameters
selected for the base pox^er plant design on plant economics
Collection of experimental data on sulfur removal and sorbent
regeneration using limestone and dolomites
Preparation of cost and performance estimates for once-through
and regenerative sulfur removal systems
Evaluation of alternative sulfur sorbents
Collection and analysis of data on spent sorbent disposal
utilization and environmental impact of disposal
Projection and analysis of trace emissions from fluidized-bed
combustion systems
Analysis of particulate removal requirements and development
of a particulate control system for high-temperature, high-
pressure fluidized-bed combustion systems
Construction of a high-pressure/temperature particulate
control test facility
Development of plant operation and control procedures
Construction of a corrosion/erosion test facility for the
0.63 MW Exxon miniplant
Continued assessment of fluidized-bed combustion power plant
cycles and component designs to evaluate environmental impact.
iv
-------
The results of these surveys, designs, evaluations, and exper-
imental programs provide the basis for the work being carried out under
the current contract. Seven are available that document the prior
contract work*.
*See References 1, 12, 57, 62, 65, 66,
v
-------
.ABSTRACT
Projections of the economics of fluidized-bed combustion (FBC) power
plants, operated with regeneration of calcium-based sorbents, have been
developed on the basis of current estimates of sorbent regeneration process
performance. Both Atmospheric-Pressure Fluidized-Bed Combustion (AFBC)
and Pressurized Fluidized-Bed Combustion (PFBC) are considered. Coal-
fueled reductive-decomposition is evaluated for AFBC, and three sorbent
regeneration schemes are evaluated for PFBC (two reductive-decomposition
schemes and a two-step regeneration process). Economic comparisons with
FBC power plants operated with once-through sorbent systems and with
conventional power plants using limestone wet-scrubbing are presented.
The sulfur recovery process for regenerative FBC is identified as the
dominant cost component of the regeneration process. Regenerative FBC
performance requirements for economic feasibility are projected and
critical development needs are discussed.
vii
-------
CONTENTS
Page
1. INTRODUCTION 1
2. CONCLUSIONS 3
3. RECOMMENDATIONS 8
4. STATUS OF TECHNOLOGY 11
Regeneration Reaction Schemes 11
Experimental Status and Performance Projections 15
Reductive Decomposition 15
Two-Step Regeneration 17
Commercial Performance Projections 17
Sulfur Recovery from Regenerator Off-Gass 18
Conclusions 18
Fluidized-Bed Combustion Power Plant
Sulfur Production Rates 19
Recovery of Elemental Sulfur From H2S
Gas Streams 22
Commercial Process Operating Factors 24
Economics 26
Recovery of Elemental Sulfur From S02
Gas Streams 29
Economics 30
Sulfuric Acid Production From S02 Gas
Streams 32
Assessment 34
Carbon Dioxide Recovery 36
Conclusions 36
Rate of Carbon Dioxide Consumption 38
Commercial C02 Recovery Options 40
Economics of Carbon Dioxide Recovery 40
Sorbent Circulation 45
Sorbent Circulation System Requirements 45
Techniques for Transporting Solids 50
Plant and Transport System Layouts 53
Evaluation of Transport Techniques 57
Sorbent Circulation Rates 59
Cost Projections 61
5. REGENERATION FOR ATMOSPHERIC-PRESSURE FRC 64
Process Description 64
Heat and Material Balances 66
Equipment Description 66
Regeneration Element 69
Sorbent Circulation Element 69
Sulfur Recovery Element 69
IX
-------
CONTENTS (Cont'd)
Page
Design Parameters L~
Regeneration Temperature and Pressure 72
Feed Location and Reducing Gas Composition 73
Process Options
Regenerator Vessel Location 75
Sorbent Size 76
Spent Sorbent Removal 7R
Performance Projections 78
Ca/S Makeup Ratio 78
Thermal Efficiency 79
Environmental Impact ""
Reliabilitv BO
Cost Estimate 81
Recycling of RESOX Tail-Gas to Boiler 88
Assessment 94
Conclusions 94
Development Requirements 95
Reliability 96
6. REGENERATION FOR PRESSURIZED FBC 97
Process Options 97
Evaluation Basis 97
Process Performance Projections 101
Capital Investment 104
Energy Cost 107
Economic Comparison with Limestone
Wet-Scrubbing 108
Environmental Comparison 113
Assessment 113
7. NOMENCLATURE 116
Abbreviations 117
8. REFERENCES 118
APPENDICES
A. EFFECT OF VARIOUS FACTORS ON REGENERATOR SO
CONCENTRATION 125
Results 125
Conclusions and Recommendations 128
References 132
-------
FIGURES
Page
1. 'Required Combustor Sulfur Removal Efficiency 21
2. Sulfur Production Rate 23
3. Capital Investment for Sulfur Recovery with Steam
and CCL Regeneration Process 27
4. Sulfur Production Cost for Steam and CC^
Regeneration Process 28
5. Capital Investment for Sulfur Recovery with One-Step
Regeneration Process 31
6. Sulfur Production Cost for One-Step Regeneration
Process 33
7. Concentrated Sulfuric Acid Production for One-Step
Regeneration Product Gas 35
8. Two-Step Regeneration Process Options 37
9. Carbon Dioxide Consumption Rate 39
10. Carbon Dioxide Absorption Process 41
11. Carbon Dioxide Recovery Process Investment 43
12. Carbon Dioxide Recovery Process Investment 44
13. Power and Investment for Stack-Gas Compression for
Recovery 46
14. Solids Transport Techniques 55
15. Dense-Phase Solids Transport Techniques 56
16. Sorbent Circulation Rate Projection 60
17. Capital Investment for Sorbent Circulation System 63
xi
-------
FIGURES (Cont'd)
Page
18. Atmospheric One-Step Regenerative Process Flow Diagram 68
19. Atmospheric One-Step Regeneration Schematic Flow
Diagram of One Module **2
20. Capital Cost of Regenerative Process as a Function of PSL 83
21. Energy Cost of Regenerative Process as a Function of PSL 84
22. Comparison between Cost of RESOX and Allied Chemical
Processes as a Function of PSL 85
23. Process Investment for Different Concentrations of SO
with Allied Chemical Process 89
24. Cost of Sulfur Recovery and Regeneration Elements for
Different % SO with Allied Chemical Process 90
Al. Equilibrium for 1/4 CaS + 3/4 CaSO = CaO + SO at
101 kPa 42
A2. Effect of Heat Losses and Air Preheat Temperature on
Concentration of S0? - Coal as Fuel 127
A3. Effect of Heat Losses and Air Preheat Temperature on
Concentration of SO - Methane as Fuel 129
A4. Effect of Change in Sorbent Utilization across Regenerator
on Concentration of SO-, Sorbent Circulation and Fuel
Input 130
A5. Effect of Heat Losses and Fuel Input on Concentration
of S02 131
xii
-------
TABLES
Page
1. Summary of Equipment Recommended for Different Types
of Study 10
2. Summary of Sulfur Recovery Costs for Fluidized-Bed
Combustion Sorbent Regeneration 20
3. A List of CO Recovery Processes 42
4. Economic Projections for CO Recovery from Stack Gas 47
5. Required and Desired Characteristics 48
6. Solids Transport Technologies 50
7. Directional Capabilities of Transport Methods 54
8. Comparison of Transport Technologies 58
9. Design Specifications and Assumptions 67
10. Heat and Material Balances 70
11. Energy Cost 87
12. Cost Sensitivity of Atmospheric Regenerative Process 91
13. Energy Cost 92
14. Energy Cost 92
15. Energy Cost 93
16. Process Options 98
17. Power Plant Basis 99
18. Process Sulfur Load 100
Xlll
-------
TABLES (Cont'd)
Page
101
19. Selection of Process Options
-i r\O
20. Operating Conditions and Performance Projections
103
21. Performance Projections
22. Investment for Pressurized Reductive Decomposition
Process - 1 Percent S02
23. Investment for Pressurized Reductive Decomposition
Process - 2 Percent S02 10
24. Investment for Low-Pressure Reductive Decomposition
Process - 10 Percent S02 106
25. Investment for Two-Step Process 106
26. Energy Cost for Pressurized Reductive Decomposition
1 Percent S02 109
27. Energy Cost for Pressurized Reductive Decomposition
2 Percent S02 109
28. Energy Cost for Low-Pressure Reductive Decomposition
10 Percent S02 110
29. Energy Cost for Two-Step Regeneration Process 110
30. Cost of Dolomite Required to Give Equal Once-through
and Regenerative Costs 111
31. Comparison of Regenerative Pressurized Fluid-Bed
Combustion with Conventional Coal-Fired Power
Generation
32. Comparison of Environmental Impacts 114
xiv
-------
ACKNOWLEDGEMENT
We want to express our high regard for and acknowledge the contribu-
tion 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 contributions
through discussions and support of the program.
The program consultation and continued support of Dr. D. H. Archer,
Manager, Chemical Engineering Research, at Westinghouse, are acknowledged.
xv
-------
SECTION 1
INTRODUCTION
Fluidized-bed combustion systems are being developed for a variety
of applications (e.g., steam generation, electric power generation,
process heat, ...) utilizing a number of concepts (e.g., atmospheric
and pressurized combustion, temperature control by fuel-air ratio, heat-
transfer surface, particle circulation). These systems have the poten-
tial for achieving lower costs, improved resource utilization, and
reduced environmental impact compared to conventional combustion systems.
Developmental facilities for fluidized-bed combustion power genera-
tion are presently based on once-through sorbent (limestone or dolomite)
operation. Although research facilities are addressing the area of sor-
bent regeneration, the technical and economic feasibility of regeneration
is not yet known. Regeneration of sorbent for the purpose of reducing the
rate of spent sorbent production, and possibly reducing its environ-
mental impact, faces trade-offs in the areas of economics, environmental
impact, plant complexity and reliability, and general technical
performance.
The achievement of current New Source Performance Standards for
large coal-fired steam generators (>250,000 Ib/hr) is the basis for this
study: SO- emission 516 ng/J (1.2 Ib/MBtu), Particulate emission
43 ng/J (0.1 Ib/MBtu), NO emission 301 ng/J (0.7 Ib/MBtu).
X
The current status of technology in four areas has been covered:
regeneration reaction schemes, sulfur recovery, C0_ recovery, and
sorbent circulation. Various processes or methods in use for each
area relevant to fluidized-bed combustion sorbent regeneration are
discussed. The advantages and disadvantages of each process, when
-------
applied to regeneration, are outlined. Capital and operating costs as
a function of key parameters for each of the above technologies have
been projected.
Section 5 contains an evaluation of the economic potential, the
technical feasibility, the problem areas, and the development require-
ments of the atmospheric-pressure one-step regeneration process
(reductive decomposition) as applied to an atmospheric fluid-bed boiler
(AFBC). Other regeneration processes have been studied previously.
The importance and effect of various design parameters, such as solids
residence time in the regenerator, gas velocity and composition, tem-
perature and pressure, are outlined. A discussion of process design
factors, such as the concentration of SO- in the regenerator off-gas,
reactivity of the regenerated sorbent, fresh sorbent makeup rate, and
the thermal efficiency of the regenerator, are presented. The various
process options for the sorbent, regenerator, fuel, and sorbent particle
size distribution are discussed. Capital and energy costs of both once-
through and regenerative options as a function of the coal sulfur con-
tent are projected. A cost sensitivity analysis of the process has been
performed. Problem areas and development requirements are identified.
The economics and performance of three regeneration processes that
function to regenerate the utilized S00 sorbent produced in the boilers
£-.
of the pressurized fluid-bed combustion (PFBC) power plant have been
reported. These processes are a one-step process (reductive decompo-
sition) operated at 1013 kPa pressure, the same one-step process oper-
ated at 100 kPa pressure, and a two-step regeneration process. A sensi-
tivity analysis indicates the potential of the one-step process at both
high and low pressures. An update of the work reported in 1973
appears in Section 6. Revised performance and economics are presented
for all three concepts. The projections reflect current performance
expectations and revised component cost data.
-------
SECTION 2
CONCLUSIONS
The following conclusions can be drawn from a technical and
economic evaluation of the regenerative processes involving the
"one-step" reductive decomposition of the sulfated sorbent at atmos-
pheric pressure and elevated pressures, and the elevated-pressure two-
step reaction scheme:
An integrated regeneration system for both atmospheric
and pressurized fluidized-bed combustors has yet to be
technically demonstrated. Information on critical per-
formance factors for commercial operation is not yet
available (e.g., regenerative sorbent activity, regen-
eration kinetics, desulfurizer kinetics with regenera-
tive sorbents, sorbent attrition behavior, ash
agglomeration behavior, sulfur recovery performance,
environmental impact of regenerative spent sorbents).
The sulfur recovery system is the dominant subsystem
in the regenerative process. The pressurized one-step
regeneration (reductive decomposition) results in low
S0_ concentrations in the regenerator off-gas (1-2 vol %)
requiring significant amounts of coal for reductant,
complex energy recovery, and sulfur recovery systems.
The atmospheric regeneration yields 10 to 12 vol % of
S0« and, hence, has a substantial advantage over the
pressurized regeneration.
In the case of atmospheric-pressure regeneration
applied to pressurized fluidized-bed combustion
-------
processes, the major uncertainty lies in the solids
transport system (transporting hot solids between,
for example, a 1013 kPa combustor and a 100 kPa
regenerator).
Assuming 2 and 12 vol % of S02 off of the regener-
ator for pressurized (PR) and atmospheric regener-
ation (AR), respectively, Ca/S makeup ratio of 1.0,
a process sulfur load of 0.03 (mass of sulfur
handled by the regeneration process per unit mass
of coal fed to the fluid-bed boiler), and sulfur
recovery in the form of elemental sulfur, the following
capital costs for regeneration (635 MWe plant) in terms
of $/kW have been projected:
AR for
AFBC
32.2
PR for
PFBC
66.8
AR for
PFBC
57.2
Two-Step
Regeneration
for PFBC
80.1
Regeneration process investment cost includes the cost
of the regeneration system, sulfur recovery system, and
sorbent circulation system.
Using the same bases as above, the following energy costs
in terms of mills/kWh have been projected:
AR for
AFBC
2.9
PR for
PFBC
4.94
AR for
PFBC
4.94
Two-Step
Regeneration
for PFBC
5.04
-------
For atmospheric regeneration applied to an atmospheric-
pressure fluidized-bed combustion power plant, the
following energy costs in terms of mills/kWh for a
process sulfur load of 0.025 have been projected for
three different fresh Ca/S makeup ratios in the case
of regenerative operation, and for a Ca/S ratio of
2.2 in the case of the once-through option:
Regenerative Option Ca/S
Makeup Ratio, mills /kWh
0.2
0.6
1.0
Once-through
Option,mills/kWh
1.9
2.2
2.5
1.92
In other words, for the condition assumed, a regenerative
system would have to be able to reduce the fresh sorbent
makeup rate to Ca/S = 0.2 or less in order to compete
with a once-through system.
If sulfur is recovered in the form of sulfuric acid rather
than elemental sulfur, the capital cost of the regenera-
tion process can be reduced to the following extent:
AR for AFBC
45%
PR for PFBC
24%
AR for PFBC
11%
If a sulfur recovery process is developed specifically
for fluidized-bed combustion sorbent regeneration, the
regeneration potential may be.considerably improved.
If sulfur vapor rather than SO,, can be produced in the
regenerator the cost of regeneration may be reduced
drastically. The scope and the need for process inno-
vations in sulfur recovery are thus evident.
-------
The only regenerative PFBC power plant that is economic-
ally attractive, compared to a conventional power plant
with limestone wet-scrubbing, is based on the low-
pressure reductive decomposition. As indicated pre-
viously, however, the application of atmospheric regen-
eration to PFBC systems requires resolution of the
potential difficulties associated with hot solids
transport between the combustor and regenerator.
The overall environmental performance (power plant
efficiency, fuel consumption, spent solids quantity,
plant reliability, etc.) of the low-pressure reduc-
tive decomposition is superior to the other regenera-
tion processes.
The once-through sorbent operation is superior to the
regenerative operations in all environmental aspects
except for the quantity of spent sorbent produced
(i.e., coal consumption, plant efficiency, plant reli-
ability, waste ash and sulfur, methane consumption).
A review of three critical support systems necessary for the evalua-
tion of the regeneration processes shows the following:
An assessment of commercial or near-commercial technology
for the recovery of elemental sulfur from SO- streams
shows that the Allied Chemical process is the most
commercially developed. The RESOX process under develop-
ment by Foster Wheeler appears to have several advantages
over the Allied Chemical process. The most critical
factor influencing the performance and cost of sulfur
recovery is the concentration of SO- in the regenerator
off-gas.
-------
e An evaluation of the various solid transport techniques
reveals that the dilute-phase pneumatic transport is
the most suitable for the presently conceived fluidized-
bed combustion power plant.
From a review of the technology for the recovery of carbon
dioxide, the hot carbonate processes appear to be the most
promising.
-------
SECTION 3
RECOMMENDATIONS
The following recommendations are made after reviewing the present
technology applicable to regeneration and the development effort that
has been carried out so far:
Studies should be continued on particle attrition, sorbent
deactivation due to the presence of fly ash or to sinter-
ing, or to particle agglomeration due to eutectic forma-
tion and gas-particle contacting in fluidized beds.
Studies should be continued on the change in activity and
the regenerability of the sorbent with repeated cycling,
and the separation of sorbent and ash in the regenerator.
The maximum percentage of SO- in the regenerator effluents
that can be achieved in a continuous operation of the
combustor-regenerator system at commercial operating con-
ditions needs to be demonstrated.
The development of sulfur recovery processes suitable for
different regeneration schemes under consideration should
be initiated.
Exploratory work should be conducted on new schemes, such
as the production of sulfur vapor rather than sulfur
dioxide in the regenerator.
The low-pressure reductive decomposition for PFBC appears
to have greater potential than pressurized regeneration.
The sorbent circulation system for the low-pressure
regeneration for PFBC, the area of greatest uncertainty,
should be evaluated in greater detail.
-------
The present developmental effort on regeneration should
be directed to correspond to the operating conditions
envisaged for commercial operation. Much of the past
effort on regeneration appears to have no relevance to
industrial practice.
A regeneration system modeling study is needed to assess
the regeneration technology and process economics in
greater detail and to permit the assessment of the experi-
mental data that is being accumulated.
0 The development of optimum methods for the disposal/
utilization of the spent sorbent, that meet environmental
constraints, is necessary.
Environmental emissions from regeneration/sulfur recovery
systems, including air and liquid emissions, and includ-
ing the leaching characteristics and other environmental
impacts of the solid residue from the regeneration system,
should be estimated and compared with the impacts of residue
from once-through systems.
The following table shows the scale of equipment recommended for each
type of study that needs to be either continued or initiated for estab-
lishing the feasibility and the commercialization of the regeneration
processes considered in this study:
-------
TABLE 1. SCALE OF EQUIPMENT RECOMMENDED FOR
DIFFERENT TYPES OF STUDY
Nature of
Study
TGA*
Bench-
Scale
,. i
Pilot
Plant
Demonstration
Plant
Development of regeneration
processed with in-situ par-
tial combustion of coal in
regenerator (demonstrates
S02 concentration, sorbent
activity)
Demonstration of integrated
process (must be coupled
with sulfur recovery
process and include com-
mercial plant operating
demands)
Development of existing
sulfur recovery processes
for regeneration
Development of novel sul-
fur recovery processes
for regeneration
Exploratory work on new
schemes such as the pro-
duction of sulfur vapor in
regenerator
Change in the sorbent
activity and regenerability
with repeated cycling
Sorbent deactivation due to
sintering and the presence
of fly ash
Sorbent attrition
Separation of sorbent and
ash
Utilized sorbent disposal/
utilization
Reconstitution of utilized
sorbent for use in
regeneration
Environmental emissions
from regeneration/sulfur
recovery systems
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
X
*TGA - Thermogravimetric Analysis
10
-------
SECTION A
STATUS OF TECHNOLOGY
The primary regeneration concepts for fluidized-bed combustion
(2 3}
using calcium-based sulfur sorbents have been identified. ' The
regeneration concepts being evaluated as part of the current task objec-
tives are reviewed. Regeneration data were previously reviewed,
and performance conclusions based on recent data are incorporated
in this report to provide an updated assessment of current informa-
tion. In addition to the regeneration process itself, three critical
support systems are reviewed in order to analyze regeneration pro-
cesses for AFBC systems and to update the previous Westinghouse
evaluation of regeneration for PFBC systems. These three support
systems are: commercial and developing sulfur recovery technology;
carbon dioxide recovery technology; and pneumatic transport technology.
Performance and economic projections are presented to provide a basis
for evaluating alternative sorbent regeneration processes.
REGENERATION REACTION SCHEMES
Several concepts have been proposed for the regeneration of calcium-
based sorbents for use in fluidized-bed combustion systems. The follow-
ing regeneration reaction schemes have the general thermodynamic potential
indicated:
Thermal decomposition: CaSO, ^=*CaQ + S02 + 1/2
Potential: severely limited by thermodynamics
11
-------
'-\
fC02
Reductive decomposition: CaSO + J I = CaO + S02 + «H f
.s
Potential: possible for both atmospheric and pressurized oper-
ation; favored by lower pressures
Two-step reaction scheme with sulfide generation followed by
steam reaction:
. JH2°1
CaS + H20 = CaO + H2S
Potential: second step severely limited by thermodynamics
Two-step reaction scheme with sulfide generation followed by
steam and CO. reaction:
M jVl
CaS + 41 i
\
Potential: second step thermodynamically limited to pressurized
operation
Two-step reaction scheme with sulfide generation followed by
oxidation:
CaSO, + 4c ^ CaS
2) CaS + 3/2 0 ^= CaO + SO.
^ £
Potential: thermodynamically identical to reductive decomposi-
tion (i.e., the second step is really the sum of the reactions
CaS + 2 02 ^ CaSO^
3 CaS04 + CaS ^± 4 CaO + 4 S02 ,
and the latter reaction is the one that characterizes reductive
decomposition thermodynamics) .
12
-------
o Two-step reaction scheme with partial sulfide generation followed
by solid-solid sulfide-sulfate reaction:
1) CaSO, + 4C - cas
2) 3 CaSO, + CaS ==: 4 CaO + 4 SO
Potential: High S0» concentrations thermodynamically possible
for atmospheric or pressurized operation,
o Two-step reaction scheme with sulfide generation followed by
slurry carbonation (Glaus-Chance reaction) :
(H2\
1) CaSO, + 4|c'j CaS
2) CaS + H20(£) + C02
or, in separate stages,
2a) 2 CaS + C02 + H20(£) =± Ca(HS>2
2b) Ca(HS)2 + C02 + H20(«,) =2 CaC03 + 2 H2S
Potential: high H2S concentration
Two-step reaction scheme with sulfide generation followed by
(4)
Shaffner and Helbig reaction :
JH2\ IH2°\
1) CaSO, + 4|CJ) = CaS + 4^
2) CaS + MgCl2 + 2 H20 Z^± H2S + CaCl2 + Mg(OH)2
Mg(OH)2 + CaCl2 + C02 = CaC03
Potential: high IUS concentration.
13
-------
Of the eight regeneration reaction schemes shown, only two con-
cepts were evaluated or tested: the reductive decomposition at atmospheric
pressure and elevated pressures (see Figure 18) and the two-step reaction
scheme with sulfide generation followed by steam and C0£ reaction for
pressurized operation (see Figure 8) . The reductive decomposition scheme
is a high-temperature route (typically about 1100°C), while the two-step
scheme is a relatively low-temperature route to regenerated sorbent (the
maximum temperature in the sulfide generation step is approximately
800°C and about 700°C in the H2S generation stage).
Argonne (ANL)^ is studying the kinetics of the solid-solid sulfide-
sulfate reaction but has not proposed an integrated process concept. The
two two-step regeneration concepts with aqueous-phase reaction steps
would result in fine precipitates of CaCO and would require a reconsti-
tution process step to generate a suitable sorbent particle structure.
Esso (UK) has demonstrated the second step of the two-step reaction
scheme with sulfide generation followed by oxidation on a pilot plant
scale, but the performance of this reaction scheme shows no advantage
over reductive decomposition.
The reductive decomposition concept and the two-step regeneration
concept with steam and CO,, reaction are evaluated for AFBC (Section 5)
and PFBC (Section 6) power generation systems. The reductive decompo-
sition regeneration scheme consists of reducing CaSO, to CaO in a
single fluidized-bed vessel at a temperature of about 1100°C. This
reduction can be carried out at atmospheric pressure or at pressures
as high as 1013 kPa. The two-step regeneration scheme consists of
reducing CaSO, to CaS in the first step at about 815°C and then con-
verting CaS to CaCO in the second step at about 675°C. The first
step is thermodynamically favored by low temperature and unaffected
by pressure, while in the second step increased H^S concentration is
favored by low temperature and high pressure (1000 to 1200 kPa),
which also suppresses competing side reactions. The two-step regen-
eration scheme is thus suitable for high-pressure operation.
14
-------
Several specific processing concepts are proposed and evaluated for
these regeneration schemes in Sections 5 and 6.
EXPERIMENTAL STATUS AND PERFORMANCE PROJECTIONS
An integrated fluidized-bed combustion-sorbent regeneration system
has not yet been constructed or operated. The only truly continuous
combustion-regeneration simulation available is the Exxon miniplant
(pressurized combustor), which does not represent a totally integrated
facility (i.e., no coal feeding to regenerator, no sulfur recovery, no
spent sorbent processing). Operation of the Exxon miniplant regen-
eratively is in the early stages. Other large-scale fluidized-bed com-
bustion systems (e.g., the Rivesville 30 MWe atmospheric-pressure
fluidized-bed boiler) are being based initially on once-through sorbent
operation.
Most of the experimental information on sorbent regeneration has
resulted from the operation of batch or semicontinuous small-scale
fluidized-bed units or from thermogravimetric apparatus. Several
organizations have been involved in studies of reductive decomposition
regeneration and two-step regeneration. Some key findings are reviewed
and interpreted.
Reductive Decomposition
A comprehensive atmospheric-pressure regeneration program is under
way on a bench-scale process development unit at ANL. The results
obtained to date for atmospheric-pressure reductive decomposition are
encouraging and report a maximum S09 concentration of about 8.5 vol %
in the dry regenerator off-gas in a ten-cycle combustion-regeneration
experiment at a pressure of 152 kPa and a temperature of 110Q°C with
slightly oxygen-enriched air. The extent of sorbent regeneration was
about 70 percent. In another experiment at a pressure of 124 kPa,
the maximum concentration of S0_ was about 10.4 vol %. Sorbent losses
per cycle due to attrition followed by elutriation were about 10 wt %.
15
-------
Argonne presented a correlation for the extent of CaO regenera-
tion as a function of temperature and sorbent residence time:
ln(l - X) = A T + B ' T
s &
where
X = extent of CaO regeneration
T = sorbent residence time, min
s
A, B = function of temperature, represented by quadratic equations
obtained in the temperature range of 980°C (1800°F) to 1100°C
(2000°F).
/Q\
Exxon conducted batch reductive decomposition regeneration
studies with the partial in situ combustion of natural gas at a pressure
of 910 kPa and a temperature of 1040°C. They obtained S02 levels
in the regenerator off-gas in the range of 0.7 to 1.0 vol %, while
the equilibrium concentration was about 2.5 to 3 percent. In a
recent 100-hour continuous conbustion-regeneration shakedown run of the
miniplant, they obtained an average SO^ concentration of about 0.5 per-
cent with regenerator conditions of 770 kPa and 1010°C. The possible
reasons for these low levels of S09 are the heat losses and the large
diameter of the regenerator.
(9)
Exxon had reported in earlier studies that a concentration
of S0_ in the regenerator off-gas of about 7.8 percent could be obtained
at 1100°C and atmospheric pressure. Their batch absorption-regeneration
experiments suggest that the sorbent could retain sulfur absorption
activity after five to six cycles. The data Exxon obtained indicate
that coal fly ash also deactivates the sorbent.
The reductive decomposition regeneration work by Wheelock sup-
ports the concept of a two-zone reactor where oxidizing and reducing
zones were created to minimize the formation of CaS. The reduction of
CaSO^ (gypsum) with CO or H,, was studied in the temperature range from
1150 to 1260°C, with partial combustion of natural gas. The reaction
16
-------
rate reached a peak around 1200°C but reduced drastically around 1260°C,
possibly due to the sintering of the material that closes up the pore
structure. The reaction rate was observed to be first order with respect
to the reducing-gas concentration. The diffusion of the reducing gas
through the pores appears to be a controlling resistance. Although this
study was conducted on gypsum, it has a strong bearing on the regenera-
tion studies involving sulfated limestone or dolomite. Wheelock points
out the impurities coming from coal or sorbent may react with calcium
oxide, leading to sintering and slagging, thus reducing the ability of
the material to adsorb SO
Two-Step Regeneration
Argonne performed TGA and batch fluidized-bed studies on the two-
step regeneration scheme for PFBC. The first step of this reaction
(reduction of calcium sulfate to calcium sulfide) was found to proceed
to high levels of conversion. The second step (the carbonation of the
sorbent and generation of H2S) was found to be the limiting step.
Cycling the sorbent resulted in rapid loss of activity, reduction in
the H^S concentration generated, and reduction in the extent of regen-
eration. No programs are now actively pursuing this two-step reaction
s cheme.
Commercial Performance Projections
The present state-of-development of modeling gas-solid reactions
and fluidization phenomena does not permit the comprehensive assessment
and scaling of the small-scale data to the commercial conditions. Even
on an elementary level of modeling, the material and energy balance
limitations and process design constraints that influence the regenera-
tion system must be determined in order to project process performance.
Phenomena such as particle attrition, sorbent deactivation, particle
agglomeration due to eutectic formation, gas-particle contacting in
fluidized beds, and so on, are not well understood and may result in
unreliable and optimistic projections of system performance. The limi-
tations of projections that have been presented in the literature must
17
-------
be understood and applied with caution. For example, Argonne estimated
a sorbent makeup calcium-to-sulfur ratio of 0.35 for atmospheric-
pressure reductive decomposition based on their cyclic tests, and Exxon
estimated a ratio of 0.55 from early atmospheric-pressure cyclic
(5 9)
results. * Exxon also projected very low levels of sorbent makeup
on the basis of the miniplant shakedown tests, which applied noncom-
mercial operating conditions. These projections could not account for
the variety of sources for sorbent deactivation, attrition, or other
losses that will occur in a commercial plant.
Argonne applied their cyclic results in a mass and energy constrained
process model to predict concentration of SCL in the regenerator off-gas
beyond the experimental range, both for atmospheric and pressurized oper-
ation. They predicted that maximum S09 concentration will be obtained
for a solids residence time of about 2.5 minutes, for a temperature of
both 1040 and 1100°C. Short sorbent residence times in the regener-
ator have been recommended, but no attempt has been made to interpret
the batch-cyclic data in terms of a continuous process with wide dis-
tributions in sorbent residence times or to investigate the process
implications of such short particle residence times.
Tentative projections of regeneration performance made for the pur-
pose of process feasibility evaluation are presented in Sections 5 and
6.
SULFUR RECOVERY FROM REGENERATOR OFF-GAS
The commercially available options for sulfur recovery were identi-
fied by means of the open literature and vendor contacts. Preliminary
economic projections and technological constraints for the sulfur
recovery processes were compiled to show the influence of major variables
upon process feasibility.
Conclusions
In general, commercial technology is available for the recovery of
elemental sulfur from H2S and S02 gas streams with concentrations down
18
-------
to less than 1 mole percent and up to 100 percent. The economics of
sulfur recovery with acid-gas concentrations less than 10 percent
H2S or S02 will have a significant impact upon the overall power plant
economics.
The environmental impact of sulfur recovery will also be signifi-
cant, especially for the cases of low SC>2 or H2S acid-gas concentrations.
Commercial sulfur recovery techniques require the consumption of clean
fuels for S02 reduction, tail-gas incineration, and so on. Purge
streams of environmentally unattractive by-products will be generated.
The sulfur recovery efficiency drops as acid-gas H9S or S0? concentra-
tion is reduced.
Technological constraints will be characteristic of each of the
potential sulfur recovery processes. Acid-gas cleaning, cooling, and
water removal requirements prior to sulfur recovery may be stringent
with some processes. Process fluctuations in acid-gas composition or
flow rate may seriously limit sulfur recovery performance. The overall
power plant process logic will depend on the sulfur recovery process
energy requirements, operating conditions, steam generation capabilities,
and the ability or need to recycle the sulfur recovery tail-gas.
Numerous developing or experimental sulfur recovery techniques have
been proposed that might improve the feasibility of sulfur recovery,
but these processes are not presently available or proved. Development
efforts must consider commercial technology as a design basis, but must
continue to factor in new developments. Specific sulfur recovery tech-
nology has been recommended for alternative sulfur removal systems for
fluidized-bed combustion with calcium-based sorbents (Table 2).
Fluidized-Bed Combustion Power Plant Sulfur Production Rates
The rate of production of sulfur in a fluidized-bed combustion
power plant is dependent upon the sulfur content of the coal. Figure 1
19
-------
TABLE 2. SUMMARY OF SULFUR RECOVERY COSTS FOR FLUIDIZED-
BED COMBUSTION SORBENT REGENERATION
Regeneration Expected
Process (vc
Recommended
S02 (H2S) Sulfur Recovery
1 %) Process
Two step: 3 (H,S) Stretford or Claus
C02 and steam
One step: 2-3
at pressure
(1000 kPa)
One-step: 10
atmospheric
pressure
process with Wellman-
Lord SO recycle
Allied Chemical with
Wellman-Lord prelimi-
nary concentration
Conventional
sulfuric acid
plant
Allied Chemical (or
Foster-Wheeler
RESOX process when
commercially available)
conventional
sulfuric acid
plant
Sulfur in
Coal (wt %)
2
3
5
2
3
5
2
3
5
2
3
5
2
3
5
Sulfur
Production Rate
(Mg/day)
60
120
220
60
120
220
60
120
220
60
120
220
60
120
220
Capital
Investment
($/kW)
8-12
12-18
18-26
16-20
24-29
35-43
10-12
15-19
21-27
8-13
12-19
17-27
5
7
10
Production
Cost
(mills/kWh)
0.40
0.70
1.10
1.0-1.
1.6-2.
2.4-3.
0.4-0.
0.7-0.
1.0-1.
0.4
0.6
0.9
0.2
0.4
0.5
2
0
0
5
8
3
Basis: 1976; 600 MWe power plant; 90% sulfur recovery efficiency; no credit for sale of recovered sulfur or sulfuric acid;
for further details see Figures 1 through 7.
-------
Curve 684848-A
o
c
.
'o
I
CD
IS)
O
CD
S"
0.9
0.8
0.7
IS 0.6
0.5
0.4
0.3
0.2
0.1
0
Standard = 0.516 kg S02/GJ
Coal Heating Value,
MJ/kg{Btu/lb)
23.3 (10,000)
35.0 (15.000)
0.01 0.02 0.03 0.04 0.05
Weight Fraction Sulfur in Coal
0.06
Figure 1. Required Corabustor Sulfur Removal Efficiency
21
-------
shows the power plant sulfur removal efficiency required to meet the
present EPA SO- emission standard. Two coal heating values that
cover the range of heating values typical to Eastern coals are
considered.
Figure 2 shows the required rate of sulfur production (Mg/day) from
a 600 MWe power plant as a function of the coal sulfur content. Again,
two coal heating values are considered. Due to the large variation in
coal properties, the sulfur production rate required by a 600 MWe power
plant could conceivably range from 20 up to 350 Mg/day.
Recovery^ of Elemental Sulfur From H S Gas Streams
A two-step regeneration process could be utilized by the pressurized
(^1013 kPa) fluidized-bed combustion system to regenerate sorbent in
(12)
the form CaCO_. H?S is produced by the reaction of H^O and C0_,
with CaS (formed in the first stage from CaSO, reduction) at about 700°C
and about 1000 kPa pressure. Ratios (by volume) of steam to CO- of
about 1:1 are expected for the reaction system, although this ratio is
a critical process variable presently under study. Elemental sulfur
will be recovered from the H S-H-O-CO- gas stream. Particular atten-
tion must be accorded the utilization of the steam and C09 components
because replacement or recovery is expensive.
Commercial technology for H2S conversion to elemental sulfur is
represented by both vapor-phase and liquid-phase oxidation processes:
The Glaus process (designed and operated by numerous firms)
(14)
The Stretford process (Ralph M. Parsons Co. and Union Oil Co.
of California)
The Takahax process (Tokyo Gas Company, licensed to Ford, Bacon
and Davis)
The Giammarco-Vetrocoke process (Powergas Corp.).
The Glaus process operates at atmospheric pressure (up to 200 kPa).
Though Glaus plants have not been built to operate at elevated pressures
22
-------
Curve 685302-A
CO
o
en
CD
a:
c
o
'd
13
o
O
1_
Q_
s
13
CO
Basis: 600MWe power plant
Plant heat rate of 9500kJ/kWh
(9000Btu/kWh)
Coal heating value, MJ/kg (Btu/lb)
23.3 (10,000)
35.0(15,000)
IMg = 1.102 Ton
100 -
12345
Weight Percent Sulfur in Coal
Figure 2. Sulfur Production Rate
23
-------
(vLOOO kPa), it apparently would be possible to design and operate an
elevated pressure Glaus plant, possibly with improved economics.
This would be advantageous in terms of recycling CO^ to the regenerator,
since the regenerator operates at about 1000 kPa, although purification
of the CO stream may be required.
The Glaus sulfur recovery performance is very sensitive to the
water content of the regenerator acid-gas and other contaminants.
Because the regenerator gas will probably contain very high water con-
tents (up to VJO mole percent) , the water level must be reduced to
about 10 percent HO by cooling and condensation. The regenerator gas
must also be cleaned of particulate material prior to sulfur recovery
in order to avoid operating problems and to produce a pure sulfur
product.
The Stretford process does not adsorb CO from the regenerator acid-
(17)
gas, so CO- recycle to the regenerator is simplified. The adsorption
step will operate at the regenerator pressure to produce a clean C0_
gas requiring low recompression to return it to the regenerator. The
Giammarco-Vetrocoke process also operates at pressure and, like the
/T Q\
Stretford process, produces an environmentally unattractive by-product.
A liquid-phase Glaus process, such as that investigated on a laboratory
(19)
scale by Consolidation Coal, would be advantageous in ways similar
to the Stretford process because (1) no CO. would be adsorbed, (2) steam
need not be removed from the acid-gas prior to sulfur recovery, and
(3) the pressurized operation permits simplified recycle of the CO gas
to the regenerator without purification.
Commercial Process Operating Factors
The most critical factor influencing the performance, equipment
selection, and cost of sulfur recovery is the volume fraction of II S
24
-------
(and H20) In the regenerator gas. The sulfur recovery efficiency is
reduced as the H2S concentration is reduced (and the HO concentration
increases).
Greater than 15 percent. H S in the dry gas (vLO percent HO)
(equivalent to 9.1 percent H S in the wet regenerator gas)
A conventional Glaus plant can be used with a resulting sulfur
recovery efficiency ^90 percent with two or three Glaus reaction stages.
Four basic process variations are used, depending upon the H~S concentra-
(13) 2
tion. Lower H^S concentrations require a tail-gas cleaning or pre-
liminary concentration of the regenerator gas.
Greater than about 6 percent H S (dry basis) (equivalent to
about 3.5 percent H.S in the wet regenerator gas)
A conventional Glaus process followed by a tail-gas cleanup process
(14)
such as Beavon, Shell Scot, Cleanair, Sulfreen (low-temperature
Glaus Reaction), or IFF process (liquid-phase Glaus) must be used in
order to achieve 90 percent recovery efficiency.
Greater than about 1.5 percent H?S (dry basis) (equivalent to
about 0.85 percent H?S in the wet regenerator gas)
The Stretford process, or a Glaus process followed by incineration,
and Wellman-Lord, Haldor Topsoe, Chiyoda, or some other S02-concentrating
step with S0_ recycle to the Glaus plant will achieve sulfur recovery
efficiencies of 90 percent or greater.
Less than about 1.0 percent H~S in the wet regenerator gas
A preliminary H_S concentrating process such as Selexol, ADIP,
Benfield, Catacarb, Sulfinol, or MEA must be used. Because all of the
H S absorption processes identified also absorb large amounts of CO-
(25 to 90 percent), the concentrating equipment will be large and the
resulting concentrated H S will still be relatively dilute. The pre-
liminary concentrating process will be followed by one of the previous
25
-------
three options, depending upon the resulting H_S concentrations. The
purification steps operate at pressures of 1000 to 2000 kPa and produce
an essentially atmospheric-pressure, concentrated H,,S stream.
All of the sulfur recovery options require incineration of the tail-
gas before releasing it to the environment. The processes will all have
some solid or liquid waste associated with them. The Glaus process will
produce minor amounts of spent catalyst. The tail-gas cleanup processes,
the Stretford and Wellman-Lord processes, and the H2S-concentrating
processes will all produce some form of liquid waste requiring treatment.
All of the sulfur recovery processes require the utilization of clean
fuels for partial combustion of the H_S stream or incineration of
tail-gas.
Economics
Capital and operating costs have been projected for the recovery
of sulfur from H«S for the sulfur recovery options described. The basis
for costing is a 600 MWe power plant with a sulfur production rate of
181 Mg sulfur/day. The regenerator reactant gas is assumed to be
50 percent HO and 50 percent C0?.
Capital investments in $/kW (direct costs - i.e., do not include
freight, insurance, taxes, construction overhead, engineering, contin-
gency, and contractor fee - usually a factor of about 30 percent of
direct costs) are shown in Figure 3 as a function of the regenerator gas
H_S volume percent. The costs include condensation of the water to a
level of about 10 percent in the regenerator gas and incineration of the
final tail-gas. The costs are probably accurate to about ±30 percent.
For other sulfur production rates the capital investments should be
scaled with a 0.6 power factor.
The cost of producing sulfur in mills/kWh is estimated in Figure 4.
The basis of 181 Mg sulfur/day with no credit for sulfur or steam gener-
ation was applied. Scaling to other sulfur production rates requires a
power factor of about 0.7.
26
-------
60
Curve 685311-B
50
25
20
15
10
(A) Conventional Claus process
(B) Claus process with tail-gas cleanup (Beavon)
(£) Stretford process
§ Claus with tail-gas recycle (Wei I man-Lord)
Preliminary concentration (Sulfinol)
Followed by Claus and Beavon
Basis: Direct costs; 1976 basis
600 MWe plant; 181 Mg sulfur/day
90% sulfur recovery efficiency
Regenerator gas includes equal parts of
C02 and H20
Costs include reduction of HJ3
content to 10% (vol.); incineration
Sulfinol process absorbs 25% of the
regenerator gas C09
10 15
H-S Volume Percent from Regenerator
20
Figure 3. Capital Investment for Sulfur Recovery with Steam
and CO- Regeneration Process (H2S Generated)
-------
Curve 685310-B
Basis: 1976
2.0
to
oo
LO
600 MWe plant; 181 Mg sulfur/day
90% sulfur recovery efficiency
Capital investment from Figure 3
Capital charges - 18%/yr; 100% capacity factor
Maintenance - 5%/yr
Taxes and insurance - 1.5%/yr
Labor and overhead - $6/Mg sulfur produced
Variable costs (steam, fuel, cooling water, power,
chemical and catalysts) - based on $6/Mg
sulfur produced with 15% H-S and scaled with
power factor of 0.85 for HLS percent
No credits for sulfur, steam, etc.
Expected accuracy + 30%
10
H?S Volume Percent from Regenerator
15
20
Figure 4. Sulfur Production Cost for Steam and CO,
Regeneration Process (H-S Generated)
-------
Recovery of Elemental Sulfur from SO Gas Streams
Calcium sulfate may be reductively decomposed to calcium oxide and
S02 by contacting the utilized sorbent with a variety of reductants
(H2, CO, CH^, carbon, etc.) at about 1100°C. The regeneration scheme
may be applied to either the AFBC or the PFBC systems. Oxidation of
CaS produced by oil or coal gasification will produce a similar SO- gas
stream. The product gas from the regenerator will con-tain SO (and
traces of other sulfur compounds), low levels of reductants (H- and CO),
C°2* H2°' anc* N2 anc* "*~ow -*-eve^-s °f oxygen (probably less than 0.5 vol %.
If coal is used as a reductant in the regenerator, the gas may also con-
tain traces of hydrocarbons, tars, coal ash, and various trace elements.
Commercial or near commercial technology for the recovery of ele-
mental sulfur from SO. streams is represented by
Direct reduction of S0_ to sulfur -
(21)
Allied Chemical process (using methane or alternative reductants)
(22)
Foster Wheeler RESOX process (using coal)
(23)
ASARCO-Phelps Dodge process
Bureau of Mines citrate process
Westvaco activated carbon process
Stauffer Aquaclaus process
(24)
Other technology has been summarized.
Generation of H_S for Claus reaction with the regenerator S02
stream -
Reduction of SO- to H_S using methane
(25)
Generation of H-S by sulfur-methane reaction
Numerous processes are also available for the initial concentration
of S0_ streams (Wellman-Lord, Bergban-Forschung, ASARCO DMA, Cominco
process, Haldor Topsoe, Chiyoda, etc.) in the case where the regenerator
gas is very dilute in S0_.
29
-------
The Allied Chemical process is the most commercially developed of
/ 9fi^
all of these processes. The process can be applied directly with
SO volume fractions of 1.0 to 0.04. Below 4 percent SCL a preliminary
concentration process must be applied because of thermodynamic and heat
balance limits. The Allied Chemical process is also sensitive to the
regenerator gas oxygen content since this oxygen must also be reduced;
but because low oxygen levels are expected, this factor will not be a
concern. Only low-pressure plants have been developed. The process
requires that the regenerator gas be cooled and cleaned of particulates
and impurities, such as arsenic and selenium oxides, in order to produce
3
high-quality sulfur. About 400,000 dm methane/Mg of sulfur (13,000 scf
methane/ton of sulfur) is required for S0« reduction with no oxygen in
the gas stream. Allied Chemical is also developing the utilization of
( 2fi^
alternative liquid reductants from propane to middle distillates.
The Foster-Wheeler RESOX process presently in the pilot stage, has
several apparent advantages over the Allied Chemical process for this
application. The regenerator gas does not require as much cooling and
cleaning before reduction, and coal is used as the reductant. The
sulfur recovery efficiency of the process, however, may be lower than
that of the Allied Chemical process, and preliminary concentration of
the regenerator gas or tail-gas cleaning may be required at higher S0_
mole fractions than with the Allied Chemical process. A pressurized
version of the RESOX process may result in much improved economics when
compared to the low-pressure design considered here.
Economics
Capital investments and sulfur production costs have been estimated
for the Allied Chemical sulfur recovery process and for the Foster-
Wheeler RESOX process (atmospheric-pressure operation). Figure 5 shows
the capital investment (direct cost) in S/kW as a function of the SO
volume fraction in the regenerator gas. The direct Allied Chemical
30
-------
Curve 685309-B
U>
60
50
40
30
20
10
(A) Allied Chemical direct reduction
^B) Wei I man-Lord preliminary concentration
followed by Allied Chemical
^Q) Resox (Foster Wheeler) (80% sulfur recovery
efficiency)
CD) Resox with Beavon tail-gas cleaning
Basis: Direct costs; 1976basis
600MWe plant; 181 Mg sulfur/day
90% sulfur recovery efficiency
No free oxygen in regenerator gas
6 8 10 12 14
SCL Volume Percent from Regenerator
16
18
Figure 5. Capital Investment for Sulfur Recovery with the
One-Step Regeneration Process (S02 Generated)
-------
process is represented for SO contents down to A vol %, and for lower
SO contents the Wellman-Lord preliminary concentration process is
applied.
The RESOX process is shown over a limited range of S02 concentra-
tions because of limited cost information. A curve for the RESOX process
alone, which results in an estimated 80 percent sulfur recovery effi-
ciency, and a curve with the RESOX process followed by the Beavon tail-
gas cleaning process, which results in a sulfur recovery efficiency of
about 95 percent, are both shown in the figure.
Estimates of capital investment at other sulfur production capac-
ities may be obtained by using a 0.6 power factor.
Figure 6 shows the estimate of sulfur production cost as a function
of the S02 volume fraction in the regeneration gas. The Allied Chemical
and RESOX processes require about the same auxiliary power and the cost
of methane and coal will be nearly identical (with a coal rate of about
1 Mg coal per Mg sulfur produced for a 12 percent S0? stream). The
RESOX process requires much less cooling water than the Allied Chemical
process. Overall sulfur production costs are expected to be about the
same for the two processes in the SO. vol % range where cost data are
available for the RESOX process.
The sulfur production cost (mills/kWh) may be scaled for different
sulfur production rates with a 0.7 power factor. No credits have been
taken for the sulfur product or steam production.
Sulfuric Acid Production from SO Gas Streams
The technology to produce concentrated sulfuric acid from a variety
of S0? sources is well developed and should be applicable to the product
(27 28}
gas from the one-step regeneration process. * While the recovery
of elemental sulfur may be advantageous when considering storage, dis-
posal, or marketing of the product, the economics of sulfuric acid pro-
duction is estimated in order to provide a basis of comparison.
32
-------
Curve 685308-B
3.0
2.0
Basis: 1976
600MWe plant, 181 Mg sulfur/day
90% sulfur recovery efficiency
Capital investment from Figure 5
Capital charges - 18%/yr, 100% capacity factor
Maintenance - 5%/yr
Taxes and insurance - 1.5%/yr
No credits for sulfur, steam, etc.
Expected accuracy ± 30%
(A) Allied Chemical process
(D Wellman-Lord followed by Allied Chemical process
Resox expected to be about the same as (A)in the
range of 10-15% S(X,
1.0
5 10 15
S0? Volume Percent from Regenerator
Figure 6. Sulfur Production Cost for One-Step Regeneration
Process (SO- Generated)
-------
Figure 7 gives the estimated capital investment and production cost
for sulfuric acid as a function of the SO vol % in the regenerator gas
taken from the open literature and vendor quotes. The basis applied in
Figure 7 is identical with that used for the elemental sulfur production
costs (1976 costs, 181 Mg sulfur (equivalent) per day, etc.).
As se ssment
The cost information developed is summarized in Table 2 for fluidized-
bed combustion with calcium-based sorbents along with estimates of the
regenerator gas composition expected for the various regeneration schemes.
Capital investments and production costs are given for a range of coal
sulfur contents (or sulfur production rates) for elemental sulfur and
sulfuric acid production. Specific sulfur recovery processes are recom-
mended for each of the regeneration cases.
The high-pressure one-step regeneration requires the highest sulfur
recovery costs, while the atmospheric-pressure one-step regeneration and
the two-step regeneration require sulfur recovery costs that are com-
parable. The recovery of sulfur as sulfuric acid is about one-half as
costly as the recovery of elemental sulfur. The updated sulfur recovery
costs are higher than the cost projected in the 1973 regeneration
(12)
study by about a factor of 7 for the one-step regeneration at pres-
sure, 4 for the atmospheric-pressure one-step, and 2.5 for the two-step
regeneration.
While these cost estimates do not provide sufficient information to
permit the selection of the preferred sorbent regeneration scheme and
do not provide grounds for the evaluation of elemental sulfur recovery
versus sulfuric acid production, they do provide a basis for the evalua-
tion of sorbent regeneration when coupled with process studies involving
the entire power plant. It appears that the cost of sulfur recovery
will have a significant effect on the-power plant economics. The
environmental performance of the power plant will also be significantly
affected by the sulfur recovery process.
34
-------
Curve 685303-A
20
LO
CD
on
cu
g- 10
o
0
\
\
\
I
Basis - 1976
600MWe Plant, 181 Mg sulfur/day
No credit for sulfuric acid sales
I
5 10 15
S02 Volume Percent from Regenerator
1.3
1.1 1
0.9
0.7
O
o
c
o
t3
Z3
"8
Q.
T3
<
0.5 .a
0.3
oo
20
Figure 7. Concentrated Sulfuric Acid Production for the
One-Step Regeneration Product Gas
-------
Vendor contacts should be continued to develop more specific design
requirements and performance information for the selected sulfur recovery
processes.
CARBON DIOXIDE RECOVERY
A two-step sorbent regeneration scheme under evaluation involves
reacting the sulfided sorbent with steam and CO to generate the
carbonate form of the sorbent and hydrogen sulfide gas:
r*
CaS + CO + HO :^± CaCO + H,S .
£. £, J £
The sulfided sorbent is produced in the sulfate reduction stage of the
two-step regeneration process.
The consumed CO- must be replaced by recovering C0« from one of
several potential CO- sources in the power plant. Figure 8 shows the
C0? recovery options for the fluidized-bed combustion two-step regenera-
tion process. The carbon dioxide in the regenerator acid-gas stream
that is recycled from the sulfur recovery process to the regenerator
may also require purification, depending upon the nature of the sulfur
recovery process.
Commercial C0? recovery processes have been surveyed and economic
projections generated to provide perspective on the process requirements,
options, and potential problem areas. The results provide a basis for
the preliminary evaluation of regeneration processes and also the poten-
tial of alternative sorbents.
Conclusions
Well-developed technology is available for the recovery of CO-
from various sources within the power plant. The hot carbonate
processes appear to be the most promising.
The C0_ recovery investment appears to be acceptable if only
the stoichiometric C02 usage plus minor losses must be recovered.
The sulfur recovery process must not extensively contaminate the
circulating CO- stream.
36
-------
Dwg. 1682B90
C103, C-104
Particulate Control
E-101, E-102
Steam generator & pressure reduction
or Turbine expander & waste heat boiler
Utilized 1st S
Sorbent ^ CaS
Redu
C-l
Sulfided
Sorbent
v
tep
°4
cer
01
n
Regenerated 2nd- Step
Sorbent HoS
Generator
C-102
1
I
l
Cim
1UJ j
m r ifti
^ h 1U1
S
Possible
Heat Exchange
m Coal J
/*»>
XI
* C-104
Air
* *E
; "
Possible
Heat Exchange
Jl
10 MdCK
»» +A Drt
TO BO
COp Reco
Steam A
i
l
i
/.
m -
-102 y-' R
F-103
Recirculating
C(L Stream
) Jsteam
V
K- 101, 102, 103,104
Compressor
/r
XI
K-102
or Kecycie
ler or to
,ery Process $tack.Gas
or Fluid-Bed Boiler
Combustion Gas
or 1st - Step Regenerator
Tail -Gas
A Sulfur
sulfur
ecovery
B-100
E-1040
_/r^
XI
K-103
^
\7K-104
^ !
C02
Recovery
B-101
Tail Gas
(possibly to steam
u CaSOy, reducer)
f
E-105
E- 103, 104, 105,106
Condensor or Cooler
Figure 8. Two-Step Regeneration Process Options
-------
e The power requirements and investment for stack-gas compression
(the most likely source of CO.) is largely dependent upon the
stack-gas CO- content and the coal sulfur content and may be a
very significant cost.
An overall power plant optimization is required to minimize
investment and maximize energy utilization. Vendor contacts
should be initiated to gather more detailed information.
Rate of CO,, Consumption
The rate of CO,, consumption in the two-step regeneration process
for fluid-bed combustion is proportional to the sulfur content of the
coal fueling the power plant. This rate, based on reaction stoichiom-
etry, is shown in Figure 9 as a function of the coal sulfur content
and the coal heating value. For 3 mole percent of H^S gas produced
in the H^S-generator step (the expected value based on kinetics and
thermodynamics), the rate of CC- circulation to the sulfur recovery
process is about 16 times the C0? recovery rate based on the reaction
stoichiometry. Thus, any contamination of the recirculating C0_ stream
in the sulfur recovery step or any losses of CO- during recirculation
could greatly increase the required C0_ recovery rate.
The assumptions applied for this study are that losses result in a
10 percent increase in the CO- recovery rate, while contamination to the
recirculating CO stream is negligible (i.e., a Stretford process or
liquid-phase Glaus process is used for sulfur recovery).
C02 could be reclaimed from the power plant stack gas, from the
fluid-bed boiler hot combustion gases and/or from the first-stage regen-
erator tail-gases (Figure 8). The fluid-bed boiler hot combustion gases
and the first-stage reactor tail-gases are under pressure (1000-1500 kPa)
and might not require compression prior to C0? recovery, but the ambient
stack gas would require compression. It is assumed that CO- will be
recovered from the power plant stack gas. The applicability of other
sources depends upon overall material balances. The cost of stack-gas
38
-------
Curve 68530^-A
Basis - 600MWe Plant
90% Sulfur recovery efficiency
10% losses in COL
300
_o>
O
E
01
CD"
ro
c
O
"o.
E
00
o>
JO
X
O
200
100
0
Coal Heating
Value of
23.3MJ/kg
(lO.OOOBtu/lb)
Coal Heating
Value of
35.0 MJ/kg
(15,OOOBtu/lb)
0
1234567
Weight Percent Sulfur in Coal
Figure 9. Carbon Dioxide Consumption Rate
39
-------
compression and CO recovery are considered separately. Other options
will be considered in more detail with the evaluation of the overall
process design.
The CO. content of the stack gas for the pressurized fluidized-bed
combustion systems under study are assumed to be: pressurized boiler,
8.5 to 15 vol %; adiabatic combustor, 4 vol % CO
Commercial CO,, Recovery Options
A general process flow diagram for a C0_ recovery process is shown
in Figure 10. Other process arrangements are also used. Numerous com-
mercial processes of this type have been developed, both of the chemical
and physical absorption categories. The most widespread methods have
been based on water scrubbing (physical absorption - not commonly applied
today), MEA scrubbing, and hot potassium carbonate (Benfield process,
(29)
Giammarco-Vetrocoke, Catacarb, Carsol, etc.). Commercial processes
are listed in Table 3. It appears that the hot carbonate processes are
best suited for this application.
Economics of Carbon Dioxide Recovery
The capital investment (direct cost) for CO recoverv from stack
gas has been projected for a recovery rate of 181 kg-moles C09/hr. This
is a medium recovery rate for a 600 MWe power plant, depending on the
coal sulfur content (see Figure 9). Operating costs and utility charges
(power, chemicals, steam, fuel, and cooling water) are not projected.
Figure 11 shows the investment for a hot carbonate-type CO^ recovery
process in $/kW as a function of the CO- recovery efficiency. The CO,
" £
partial pressure to the absorber is a parameter in the figure. Invest-
ments are on the order of $1 to 2/kW, and a 0.6 power factor should be
used to scale the costs to other C0? capacities.
Figure 12 relates the investment to the C0_ partial pressure for a
fixed C0« recovery efficiency of 90 percent. The investment increases
steeply as the C09 partial pressure drops below 200 kPa.
40
-------
Purified Gas
1000-2000 kPa)
CCL Source
(1000-2000 kPa)
-d
-------
TABLE 3. A LIST OF CO ^RECOVERY PROCESSES
(30)
Licensor
(Process)
Probable
Absorbent
Allied Chemical
(Selexol)
BASF
BASF
Benfield
Eickmeyer
(Catacarb)
Carbochimique
(Carsol)
Giammarco
(Vetrocoke)
Linde-Lurgi
(Rectisol)
Lurgi
(Purisol)
Fluor
Shell
(Sulfinol)
Union Carbide
(U-CAR)
Dimethyl ether of polyethylene gycol plus an
alkanolamine
Alkazid
Triethanolamine
Hot potassium carbonate solution plus an
activator
Hot potassium carbonate solution plus an
activator
Hot potassium carbonate solution plus an
activator
Potassium arsenite solution plus an organic
activator
Methanol
N-methyl-2-pyrrolidone
Propylene carbonate
Tetrahydrothiophene plus an alkanolamine
Monethanolamine plus an activator
42
-------
1.0
Curve 685306-A
S o.io
>~,
i_
o>
o
o
CM
O
O
CD
C~
O
0.01
0.001
\
Basis - $ 1976 Direct Costs
181kg-moles (400 Ib-moles) C02/hr
recovered; use 0.6 power factor
to scale capacity
236kPa(2.33atm)C02
Partial Pressure
472kPa(4.66atm) C09
Partial Pressure
708kPa(6.99atm) CO
Partial Pressure
L
1.0
2. 0 3.0
Direct Investment, $/kW
Figure 11. Carbon Dioxide Recovery Process Investment
43
-------
Curve 685305-A
3.0
c
o>
£
CD
>
C
2.0
Basis -$ 1976 Direct Costs,
181 kg-moles (400 Ib-moles) COJhr recovered
90% CCL recovery efficiency
1.0
I
0
200 400 600 800
Carbon Dioxide Partial Pressure. kPa
Figure 12. Carbon Dioxide Recovery Process Investment
-------
The investment and power requirement for stack-gas compression is
considered independently in Figure 13 as a function of the stack-gas
pressure and the stack-gas CO content. The investment for stack-gas
compression may exceed that for the CO recovery process. For the
adiabatic combustor case and the coal gasification case, the power
requirement may be between 5 and 10 percent of the total power plant
output, an unacceptably high power usage unless this energy is recovered
at some point in the process.
Some optimum combination of stack-gas compression ratio and CO,,
recovery efficiency must exist in terms of total capital expenditure
and power requirement, but is not determined in this study. Table 4
lists the investment projections for C0~ recovery and the investment
and power requirement for stack-gas compression for three coal sulfur
contents and the stack-gas CO contents representative of fluid-bed
boilers, adiabatic combustors, and coal gasification. The adiabatic com-
bustor case and the coal gasification case (4 percent of C0~ in stack
gas) could require a large investment and huge power usage.
SORBENT CIRCULATION
The regenerative operation of fluidized-bed combustion power plants
calls for the circulation of sorbent material between the fluidized-bed
combustor and the regenerator vessel. The sorbent (limestone or dolomite
based) must circulate at a rate that satisfies the desulfurization and
regeneration reaction rates. The circulation system must meet all of
the power plant requirements. Numerous techniques are available or have
been proposed for transporting solids that may be applicable to
fluidized-bed combustion plants. These techniques are reviewed and
assessed in terms of the system requirements. Recommendations and cost
projections are developed.
Sorbent Circulation System Requirements
Table 5 lists the required characteristics and desirable charac-
teristics that may be used to judge alternative transport techniques.
45
-------
Curve 685307-B
4.0
w-
3.0
c
_o
'(/)
t/)
OJ
u.
O-
£
o
O
2.0
1.0
Basis
$ 1976 Direct costs
181kg-moles (400lb-moles) C02/hr recovered
Use 0.6 power factor to scale costs
80% CO? recovery efficiency
70% compressor efficiency
40
15% C02 in stack gas
8. 5% C0
Adiabatic
Combustor
[(D 4% C0?
OJ
30I
a:
20
0)
CD
1000
2000
10
Pressure of Stack Gas, kPa
Figure 13. Power and Investment for Stack Gas Compression
for CO- Recovery
46
-------
TABLE 4. ECONOMIC PROJECTIONS FOR C02 RECOVERY FROM STACK GAS
Coal Sulfur
Content (wt %)
C02 Recovery Rate
(kg-moles/hr)
C02 Recovery Investment
($/kW)
Stack Gas Compression Investment
($/kW) and Power (MWe)
Stack Gas CO- Content
Stack-Gas C02 Content
2 100
3.5 220
5 . 340
15%
1.3
2.0
2.7
8.5%
1.4
2.3
3.0
4%
1.8
2.8
3.7
15%
1.2 $/kW
(5.5 MW)
2.0(12.2)
2.5(18.9)
8.5%
1.8(11.6)
3.0(25.5)
3.8(39.4)
4%
3.1(27.6)
5.0(60.8)
6.4(94.0)
Basis:
600 MWe power plant
80% C02 recovery efficiency
15% CO- stack gas compressed to 1500 kPa
8.5% C02 stack gas compressed to 2000 kPa
Fluid-bed boiler
4% C0_ stack gas compressed to 2500 kPa \ Adiabatic combustor or coal gasification
-------
TABLE 5. REQUIRED AND DESIRED CHARACTERISTICS
Required System Characteristics
1. Must transport sorbent material of specific size distribution at
the required rate and at the required process conditions (tempera-
ture, pressure, environment) between the combustor and regenerator
modules
2. Must distribute the regenerated sorbent uniformly to the multiple
beds of the combustor
3. Must maintain specific bed depths within the combustor and regen-
erator (in combination with the spent sorbent withdrawal system
and the fresh sorbent feeding system) within acceptable limits
4. Must be operable over a range of sorbent flow rates permitting
plant turndown, operation with alternative fuel or sorbents, and
plant start-up, shutdown, and maintenance
5. Must respond to load changes sufficiently fast as not to limit
the power plant performance (power demand, environmental impact,
etc.) .
6. Must not reduce the power plant reliability - occurrences such as
agglomeration, plugging, erosion, corrosion, thermal stress, valve
malfunctions, etc., must have an acceptably low rate
7. Sorbent attrition or other losses occurring in the circulation
system must be significantly less than the sorbent attrition and
losses which occur in the balance of the plant
8. The operation and functions of the combustor and regenerator must
not be disrupted by the circulation system - excessively high
transport gas rates (causing large bubbles, erosion to internals,
heat balance overloads, dilution of product gases, etc.) or
periodic transport rate fluctuations (causing periodic fluctua-
tions in reactor performance, etc.) must be absent.
9. Must comply with plant safety requirements
10. Must result in acceptable cost - capital investment, operating
cost, power requirement, transport gas rate, maintenance, etc.
48
-------
TABLE 5. (Continued)
Desirable System Characteristics
1. Should provide a gas seal between the combustor and regenerator
2.* Should not require accurate control of combustor and regenerator
pressure difference
3. Should not dictate the vessel design, arrangement, and plant
layout - should not require minimum vessel separation
4. Should not dictate plant turndown, shutdown, start-up procedures
5. Should permit injection of the sorbent into the beds at any speci-
fic point (or multiple points if required) rather than dumping
sorbent on the bed surface
6. Should result in minimum sorbent attrition
7.* Should not require inert transport gas other than plant stack gas
or other available tail-gases
8. Should not require separation of transport gas from sorbent before
injection of the sorbent into the combustor or regenerator
9.* Should eliminate high maintenance moving parts - high-temperature
valves, high-temperature feeders
10. Should minimize number of components such as hold vessels, lock-
hoppers, cyclones, solids coolers, etc.
11. Should minimize length and diameter of high-temperature piping
12. Should permit inspection and flow rate measurement
13.* Should represent commercial technology
14.* Should result in minimum cost
*Most important.
-------
The ten requirements represent items that must be satisfied for a cir-
culation technique to be considered a possible candidate. The 14 desir-
able characteristics represent items that are not absolute musts but
are used to judge between the candidate techniques that satisfy the
system requirements. The desires considered most important are so
indicated.
Techniques for Transporting Solids
Numerous techniques for transporting solids have been developed or
proposed. These techniques are listed in Table 6. While some of these
techniques are capable of realizing a complete circuit between the com-
bustor and regenerator, most are physically restricted only to horizontal
or to vertical transport (upward or downward) and must be used in com-
bination with several techniques.
TABLE 6. SOLIDS TRANSPORT TECHNOLOGIES
Mechanical (bucket elevator, conveyor belt, screw conveyor, etc.)
Vibrational
Dilute - Phase Pneumatic
Induced
Forced
Dense - Phase Pneumatic
Dense-phase riser
Dense-phase lateral
Horizontal dense-phase
Dense-phase standleg
Plug flow
Extrusion flow
Mass continuous flow
Pulsed flow
Fluidized conveyor
Fluidized lateral lift
Bulk flow (non-air-assisted)
50
-------
Mechanical methods of transport, either vertical or horizontal,
(31)
at ambient temperature and pressure, are widely practiced. They
were applied to the circulation of high-temperature solids (^649°C) in
early versions of catalytic crackers but were entirely abandoned in
('oo'S
favor of the more reliable pneumatic transport methods.
Solids transport by vibration is a relatively new technique finding
wide application for transporting and distributing materials under
(33)
ambient conditions over relatively short horizontal distances.
Particle attrition and dusting are very low, and feeding is very uniform
over a wide operating range. Vendor contact has indicated that vibra-
(34)
tional methods may be utilized at high temperature and pressure.
Dilute-phase pneumatic transport methods are based on the ability
of a dilute gas-solid suspension to be transported with ease throughout
complex circuits. Operating velocities are high (>15 m/s), gas rates
are high, and pressure drops are low. Particle attrition and transport
line erosion are expected to be more of a concern than with alternative
techniques. Dilute-phase pneumatic transport is the most highlv
developed and applied method presently used for transport at ambient
conditions. The use of the technique at high-temperature and -pressure
conditions does not face any practical restrictions. The phenomenon has
/o-| OC 'Ifi}
been described, and practical design methods are available. ' '
A variety of dense-phase pneumatic transport methods is available.
Dense-phase pneumatic transport is defined as pneumatic transport of
solids at gas-to-solids ratios too low for the stable, entrained trans-
port characteristic of dilute-phase transport. By inference, this
results in lower velocities, less transport gas, and higher pressure
drops with less particle attrition and transport-line erosion.
The dense-phase standleg is simply a vertical pipe in which a
dense gas-solid suspension flows downward by the influence of gravity.
The suspension is aerated to the degree required to provide fluidized
51
-------
behavior (i.e., ease of flow, pressure drop per unit length approxi-
mately equal to the suspension bulk density). Industrial experience
at high temperature and pressure (mostly with cracking catalyst materials)
("32 ^719^ i. j (40)
is abundant; ' lateral standlegs have also been studied.
The dense-phase riser is a vertical transport line that carries
solids upwards and operates at a bulk density higher than would be sta-
ble in dilute phase vertical transport. The line may operate in a slug-
ging or a more stable fashion due to the presence of packing in the
line.C38,41,42)
Dense-phase lateral and horizontal transport may be utilized to
connect standlegs and risers or to connect fluidized beds directly. The
flow may be smooth for materials that maintain an aerated state or
periodic or nonstable for materials that deaerate quickly. In either
(43-45)
case, periodic injection points for aeration gas are required.
Transport techniques that utilize the tendency for dense gas-solid
suspensions to form stable plugs transportable over long distances have
been called plug flow transport. " Such systems are commercially
available at relatively small capacities and ambient conditions. A
high-temperature system (atmospheric pressure) has been operated on a
pilot-plant scale and designed for a demonstration plant. The tech-
nique normally requires lockhoppers and high-temperature valves.
Extrusion flow is a special case of dense-phase transport restricted
(49)
to solids having a special set of properties. Extrusion flow per-
mits the stable transport of a gas-solid suspension throughout a complex
circuit at bulk densities very near the packed-bed value without special
design considerations. Extrusion flow has not been commercially demon-
strated for any material.
Extrusion flow behavior may be achieved for any material by the
application of a special hardware design resulting in mass continuous
flow. The method is described in the patent literature and requires
complex lockhopper systems; it has not been applied industrially.
52
-------
A method that utilizes aspects of the dense-phase standleg and
horizontal dense-phase transport in combination is identified as the
pulsed-flow technique. In this method the rate of solids transport
through a dense-phase horizontal line that is fed by a dense-phase
standleg is controlled by the rate of pulsing of horizontal transport-
gas jets. The method has been utilized at high temperature on pilot-
plant scale and is being studied on a cold model facility of about
25 Mg/hr capacity.(48)
The fluidized-bed conveyor is a commercial system (Fuller Co. Air-
slide Fluidizing Gravity Conveyor) that uses the fluidlike behavior of
a fluidized bed to promote horizontal (about 5° angle with horizontal)
transport of solids over long distances at high solid-to-gas ratios.
While only applied to ambient situations, the high-temperature and pres-
sure application should be technically feasible.
Other techniques are being explored on the laboratory scale and
have no large-scale experience. For example, a fluidized lateral lift
that utilizes the splash behavior of fluidized beds to transport solids
laterally upward has been studied.
Finally, bulk flow techniques have been applied generally for dump-
ing and draining bins and hoppers. These methods are not of significance
to transport systems except as possible holding or surge stages of the
system.
Table 7 lists the transport characteristics of the various trans-
port techniques with respect to their directional capabilities. Fig-
ure 14 illustrates the transport techniques conceptually to represent
their key points. Figure 15 illustrates a variety of dense-phase trans-
port methods.
Plant and Transport System Layouts
A number of layouts could be proposed that would simplify the
sorbent circulation system. For example, the combustors could be placed
directly beneath the regenerator vessel or the regenerator directly
below the combustor to simplify the sorbent flow path. The regenerator
53
-------
TABLE 7. DIRECTIONAL CAPABILITIES OF THE TRANSPORT METHODS
Vertical
Flow
Upward
Mechanical X
Vibrational
Dilute Pneumatic X
Vertical
Flow
Downward
X
X
Horizontal
Flow
X
X
X
Lateral
Flow
Upward
Lateral
Flow
DOT m ward
X X
X X
Dense Pneumatic
Horizontal dense
Dense riser X
Standleg
Mass continuous X
Fluidized conveyor
Bulk flow
Fluidized lateral
lift
Extrusion flow X
Plug flow X
Pulsed flow
X
X
X
X
X
X
X
X
X
X
X
X
X
X
vessel might be separated into numerous individual vessels, one for each
combustor bed, and placed in close proximity to each combustor bed to
simplify sorbent distribution and control complexities. Concepts that
place the regenerator bed as an integral part of the combustor bed and
utilize an internal circulation scheme might also be applied. The
petroleum industry has shown historically that optimization of the
catalytic cracker-regenerator layout is very important to the economics
(32)
of catalytic cracking. For the specific case of fluidized-
bed combustion it appears that such layout optimizations will
be less effective: the combustor economics require multiple modules
with four or five separate beds stacked vertically, resulting in a
tall pressure shell. The regenerator vessel is expected to be very
small, compared to the combustor, and much lower in elevation. Thus
it seems that the most economic plant would be based upon a highly
flexible sorbent circulation system that carries sorbent between a
54
-------
Dwg.6397A74
Mechanical
Vibrational
Dilute-Phase
Pneumatic
General
Dense-Phase
Pneumatic
Bulk Flow
Figure 14. Solids Transport Techniques
55
-------
Dwg. 6397A73
\
Dense-Phase
Stand leg
Dense-Phase
Riser
Horizontal or
Lateral
Dense-Phase
Extrusion of
Mass Continuous
Flow
Pulsed Flow
Fluidized
Conveyor
Figure 15. Dense-Phase Solids Transport Techniques
56
-------
combustor and a regenerator whose locations are based on minimum
capital investment and maintenance cost. Such a sorbent circulation
system must be capable of transport in all directions (downward,
upward, laterally) and must consist of a combination of the techniques
listed in Table 7.
Evaluation of Transport Techniques
The sorbent transport techniques considered are ranked with respect
to their ability to meet the system requirements and desirable charac-
teristics in Table 5. This ranking is shown in Table 8. Five of the
techniques do not satisfy the basic system requirements and are not con-
sidered further. The mechanical method is expected to be too unreliable
for use. The mass continuous and bulk flow techniques are also
believed to be too unreliable for use as elements of the sorbent
circulation system. Fluid lateral lift is expected to provide slow
response to power plant demand, and the extrusion flow technique is
not physically applicable to solids having the properties of the sorbent
materials.
The ranking is based on the desirable characteristics for the
remaining eight techniques. Ranked as most attractive are the follow-
ing: dilute-phase pneumatic transport, dense-phase riser, and dense-
phase standleg. Only the dilute-phase pneumatic transport technique
can be applied alone to provide a complete sorbent circulation system.
The dense-phase pneumatic techniques must be used in combinations and
would require complex equipment to act as interfaces between the various
system elements (e.g., high-temperature feeders, lockhoppers, valves,
etc.). Realistically, the dilute-phase pneumatic transport technique
is the most suitable for the presently conceived fluidized-bed combus-
tion power plant. Optimization of fluidized-bed combustion after ini-
tial commercialization may result in the application of a dense-phase
pneumatic transport method.
57
-------
TABLE 8. COMPARISON OF TRANSPORT TECHNOLOGIES
Mechanical
Vibrational
Dilute Pneumatic
Dense Pneumatic
Horizontal
Riser
Standleg
Mass continuous
Fluid conveyor
Bulk flow
Fluid lateral lift
Extrusion
Plug
Pulsed
Requirements (Table 5)
123456789 10
P Y Y Y Y N
YYYYYYYYYP
YYYYYYPPYY
YYYYYYYPYY
YYYYYYYYYY
YYYYYYYYYY
Y Y Y Y Y N
YYYYYYYYYY
Y Y Y Y Y N
Y Y Y P N
N
YYYYYPYPYY
YYYYYYYPYY
Desirable Characteristics (Table 5)
1 2 3456 7 8 9 10 11 12 13 14
PYPYPYYYNY PY P P
YYYYYPYPPP PN Y P
PPPYPYYYNY PN N P
YYYYYYYYYY PN Y P
YYYYYYYYYY PN Y P
PPPYPYYNNY PY P P
YYYYYYYYNN PN P P
PPNYYYYYYY PN N P
Rank
2
1
4
1
1
3
4
Y = Satisfies item
N = Does not satisfy item
P = Possibly satisfies item
-------
Sorbent Circulation Rates
In order to estimate probable sorbent circulation rates, simple
material balances may be developed to give
W
T? - _§.
R " 32
where R is the rate of sorbent circulation in moles of calcium per unit
mass of coal fed to the combustor, W is the weight fraction of sulfur
s
in the coal, n is the sulfur removal efficiency of the combustor, m is
T3
the molar makeup rate of fresh sorbent to the combustor, X<, is the cal-
cium utilization of the sorbent in the regenerator, and X_ is the calcium
O
utilization in the combustor.
The sulfur removal efficiency is dependent upon the sulfur content
of the coal, and the sulfur emission standard which must be satisfied.
The sorbent makeup rate is dependent upon the required sulfur removal
efficiency, the combustor and regenerator deactivation performance and
attrition bases, etc. For the present projections M is assumed to be
equal to 0.75 moles Ca/mole sulfur fed for a sulfur removal efficiency
of 0.85 and is assumed to be directly proportional to n:
m = 0.882 n .
The rate of coal fed to the combustor is determined by the plant
size (MWe), the coal heating value, and the plant heat rate. For a
600 MWe power plant with a 9500 kJ/kWh plant heat rate and a coal
heating value of 35 MJ/kg, the coal feed rate is 163 Mg/h. The
molecular weight of sulfated limestone is assumed to be 84 g/g-mole of
calcium, and the molecular weight of sulfated dolomite is assumed to
be 136 (assuming the presence of no inerts and a molar ratio of calcium
to magnesium of 1).
Figure 16 shows the rate of sorbent circulation as a function of
coal sulfur content, with Xg - Xg = 0.1 and XR = 0.25 for limestone and
59
-------
Curve 687829-A
250
200
1 1 1 r
Basis-
600 MW Combined-Cycle Plant
XD-X =0.1
D b
Plant Heat Rate = 9500 KJ/kWh
Coal Heating Value = 35 MJ/kg
(15,OOOBtu/lb)
150
CD
o
c.
o
+-
tg
n
o
100
O)
O
LO
50
0
w
/
*
/Dolomite
1
1
1
Limestone
0 0.01 0.02 0.03 0.04 0.05 0.06
Coal Sulfur Content, Wt. Fraction
Figure 16. Sorbent Circulation Rate Projection
60
-------
Xg = 0.40 for dolomite. The rate is inversely proportional to the dif-
TJ 13
ference Xg - Xg. The rate will be slightly greater for an atmospheric-
pressure system than for a pressurized system due to the lower plant
heat rate in the pressurized case (all other factors being identical).
Ash will also be present in the circulating sorbent stream but
should be a small contribution and is neglected.
Cost Projections
The capital investment for a dilute-phase pneumatic sorbent circu-
lation system is estimated as a function of the sorbent circulation rate
and based on the following assumptions:
600 MWe power plant with four fluidized-bed combustion modules,
four combustion beds per module and a single regenerator vessel
per module located at grade elevation 30 m from the combustor.
One hold vessel per module is included in the circulation system
to handle surges, for system control, and for start-up and shut-
downs. The vessel is assumed to be 1.5 m by 3m tall. One
distribution vessel per module is placed after the regenerator
to distribute the sorbent flow uniformly to the four streams
returning to the four combustor beds.
Four dense-phase standlegs carry sorbent from the four combustor
beds to the hold vessel at 0.3 m/s. A total piping length of
80 m is assumed for the dense-phase lines per combustor module.
e A single dilute-phase pneumatic transport line per combustor
module carries the sorbent 50 m from the hold vessel to the
regenerator.
Four dilute-phase pneumatic transport lines per combustor module
carry the sorbent from the distribution vessel back to the four
combustor beds. A total of 130 m of piping per module is
required.
Two compressors are required per module, one for each circula-
tion step (combustor-to-regenerator and regenerator-to-combustor).
61
-------
Air may be used for transport gas, and no separation stage is
required for the sorbent and transport gas. The gas-solid sus-
pension may be fed directly to the regenerator and combustor.
The transport lines are refractory-lined carbon steel pipe,
with 20 cm of refractory and insulation required. Piping
costs are taken from a previous study of high-temperature
low-Btu gas piping.
The equipment included in the sorbent circulation system cost esti-
mate are the air compressors, the hold vessels and distribution vessels,
and the piping system. The piping system includes carbon steel pipe;
refractory flanges; hangers; supports; and structures and labor for erec-
tion, fabrication, and testing.
Costs are reported as direct costs in 1976 dollars. Results are
shown in Figure 4 for both the atmospheric-pressure and pressurized
systems in $/kW. The piping cost represents about 70 percent of the
total capital investment. About 1000 m of high-temperature piping
is required because of the need to account for thermal expansion and
vibration. The main difference between the atmospheric-pressure and
pressurized systems is in the cost for transport air compressors. Power
requirements are not large as long as the transport air is efficiently
used for coal combustion in the combustor and regenerator.
The sorbent circulation system cost is not strongly sensitive to
the separation between the combustor and regenerator vessels. Decreas-
ing the separation distance from 30 m to zero separation would only
reduce the total piping length by about 35 percent. Based on Figures 16
and 17, for a typical high-sulfur coal the capital investment for the
sorbent circulation system should be about $12/kW for the pressurized
system and $10/kW for the atmospheric-pressure system. This cost is
higher than that previously estimated for the sorbent circulation
system.
62
-------
Curve 687828-A
15
14
13
12
11
10
9
8
CO
o
CD
i_
Q
0
1 1
Basis -
Direct Costs
1976 Costs
Dilute Phase Pneumatic Transport
~30 m Separation between
Combustor and Regenerator
600 MWe Power Plant
Pressurized System
Atmospheric-Pressure System
I
I
100
200 300
Sorbent Circulation Rate, Mg/hr
400
500
Figure 17. Capital Investment for Sorbent Circulation System
-------
SECTION 5
REGENERATION FOR ATMOSPHERIC-PRESSURE FLUIDIZED-BED COMBUSTION
The one-step reductive decomposition of calcium sulfate is recom-
mended as the most attractive regeneration process for fluidized-bed
combustion systems operating at atmospheric pressure. An evaluation
was completed to develop performance projections, cost estimates, and
critical development requirements. A 635 MW plant was selected as a
reference design.
PROCESS DESCRIPTION
The following reaction takes place in the one-step regenerative
process involving the reductive decomposition of CaSO,:
JVl
CaS04+icoJ Ca0
The undesirable competing reaction involving the formation of CaS is
the following:
CaSO, i- 4
An oxidizing zone has to be provided in the regeneration vessel to convert
CaS to CaSO, by the following reaction:
CaS + 200 = CaSO. .
2 4
In the regeneration process coal is introduced into a fluidized-
bed vessel or regenerator for in-si-tu partial combustion to provide the
reducing gas and the heat necessary for the reduction of CaSO, to CaO.
64
-------
The regenerated sorbent is returned to the fluid-bed boiler, where fresh
sorbent will be introduced to make up for reduced activity and losses of
the sorbent by attrition and elutriation. Fresh sorbent could be intro-
duced into the regenerator rather than to the combustor. Although the
heat load on the regenerator increases, the possible advantage is a
reduction in the trace element release (i.e., trace elements may be
captured in the sulfur recovery process). The effect of regenerator
operating conditions on sorbent calcination has to be examined. If
the makeup ratio of Ca/S is small, as Argonne projected, the increase
in heat load may be moderate. Part of the sulfated sorbent is discarded
for disposal or utilization. The regenerator off-gas containing about
12 percent S02 at a temperature of 1100°C passes through primary and
.secondary cyclones and then exchanges heat with the incoming air to
the regenerator before being processed in a sulfur recovery plant for
the production of elemental sulfur. The circulation of the sorbent
between the boiler and the regenerator is carried out pneumatically.
The important variables of the process are the process sulfur load
(PSL) and the concentration of S0? in the regenerator effluent. The
process sulfur load is the ratio of the amount of sulfur handled by the
regenerative process to the amount of coal fed to the boiler and, hence,
determines the scale of the process. It is given by the following
equation:
PSL
= W (n - m X*) ,
where W is the;weight fraction of sulfur in the coal, n is the boiler
s
sulfur removal efficiency, m is the Ca/S molar makeup ratio to the
boiler, and X8 is the mole fraction of Ca as sulfate in the sulfated
S
sorbent. The concentration of SO- in the regenerator effluent determines
the size of the equipment and depends on the type of fuel used in the
regenerator, temperature, pressure, heat losses, and the change in the
utilization of calcium across the regenerator (i.e., the extent to which
65
-------
sulfur is removed from the sorbent in the regenerator). In the present
study only atmospheric pressure was considered, and coal was chosen as
the fuel to be burned in the regenerator. The other variables of
importance are the regeneration temperature and the extent of sorbent
regeneration.
HEAT AND MATERIAL BALANCES
The design specifications and the design assumptions for this study
are given in Table 9.
The technical and economic evaluation of the regeneration process
is based on the available experimental data, the desirability and the
feasibility of a particular option, the results of earlier studies on
regeneration, and on the state of the art. The partial in-situ combustion
of coal in the regenerator is an attractive option and, on the basis of
preliminary experiments, seems to be feasible. The extent of regeneration
of the sulfated sorbent is also assumed on the basis of the experimental
data. * The experimental basis for many of the assumptions applied is
limited and no directly applicable, steady-state data have been produced.
Regarding the sulfur recovery system, it is generally agreed that
the production of elemental sulfur is more attractive for the power
(58)
plants than is the production of sulfuric acid. The RESOX process,
under development by Foster Wheeler Energy Corporation (FWEC), appears to
be more attractive than various other processes and, hence, has been
chosen for the basic design.
A process flow diagram showing the various streams is given in
Figure 18. For the sulfur recovery system a supplementary sulfur
recovery process, such as the Beavon process developed by the Ralph M.
(59)
Parsons Co., is required since the sulfur removal efficiency of the
RESOX process is only about 80 percent. The reasons for selecting these
two processes for the basic design are given later.
66
-------
TABLE 9. DESIGN SPECIFICATIONS AND ASSUMPTIONS
Design Conditions;
Boiler coal rate
Emission standard for SO,
t
Basis for boiler design
Sorbent type
Process sulfur load
Sorbent disposal
Plant capacity factor
Sulfur recovery
Number of regenerator modules
Operating pressure and temperature
of AFBB
In-situ partial combustion of coal
in the regenerator
Design Assumptions;
Regenerator temperature
Dolomite makeup rate (including
attrition losses)
Dolomite utilization in boiler
Dolomite utilization after regenerator
Percent SO. in the regenerator
effluent
No CaS is formed
240,408 kg/hr (635 MWe)
516 ng/J; current NSPS
Westinghouse report of 1971
Dolomite (material and
energy balances and plant
costs almost identical for
the case of a limestone
sorbent)
0.026
Before regeneration
70%
Elemental sulfur by the
RESOX Process, with tail-
gas cleanup by Beavon
101 kPa and 870°C
(57)
1100°C
1 mole Ca/1 mole S
35%(1,5)
10%d,5)
12%
(1,5)
67
-------
Own. 6392A28
O3
Coal
Air
Makeup
Sorbent
Sulfur
To Stack
Flu id -Bed
Boiler
870°C
101 kPa
S.R. -
Spent Stone
Disposal/Utilization
Sulfur Recovery
Figure 18. Atmospheric One-Step RegeneratI
-------
The process is divided into three elements: the regeneration element,
the sorbent circulation element, and the sulfur recovery element. Heat
and material balances are given in Table 10.
EQUIPMENT DESCRIPTION
Regeneration Element
The regeneration element consists of four vessels constructed of a
carbon steel shell, refractory lined and insulated with a refractory grid.
Four blowers supply air to the vessels. The air is preheated in four heat
exchangers by the regenerator off-gas. The regenerator off-gas passes
through four primary cyclones and then through four secondary cyclones
before going to the RESOX process. All of the cyclones are refractory
lined and insulated.
Sorbent Circulation Element
The sorbent circulation element consists of four modules like the
regeneration element. Each module has a hold vessel that receives
sulfated sorbent from six different lines from the six fluidized beds
of the AFBB. The sulfated sorbent is then pneumatically transported to
the regenerator vessel via a single line. The regenerated sorbent passes
to a distribution vessel and then to the boiler through six different
lines to the six fluidized beds of the boiler. Four blowers supply air
for transporting the sulfated and regenerated sorbent streams. All of
the piping and the hold vessels are refractory lined.
Sulfur Recovery Element
The regenerator off-gas passes through a heat exchanger to the sulfur
recovery plant. The RESOX process consists of a reducing vessel where
SO is reduced to elemental sulfur by coal and a condenser where the
sulfur is condensed. Since the sulfur recovery efficiency is around
80 percent, the RESOX tail-gas can- either be recycled to the boiler or
treated in a secondary sulfur recovery step such as a Beavon process.
The latter was chosen for the base design. The Beavon process consists
69
-------
TABLE 10. HEAT AND MATERIAL BALANCES
Stream
No.
(Fig-
ure 18)
1
2
0
J
4
5
6
7
8
9
10
11
12
13
Fl
Process Stream Temp (°C) /Pressure (kPa) (kg
owrate Enthalpy
moles/hr) (kJ/kg mole) Comments
Coal to regenerator 93.3/137.8 5602 kg/hr 86.5 kJ/kg
Air to regenerator 704.4/158.5 1207
[T.i-j-1 -1 r,f-.f^ orsvi*ian t- ft71/1O^A 771 n
utilized, s or Dent Q/J./IUJ*** / / j ^T^
h
21,074
moles of Ca , MgO - 50%, CaS04 - 17.5%
r CaO - 32.5%
Regenerated -orbcnt 1093 Vl03 4 -"73 kg moles of Ca 110 1-0 MR° ~ 50%! CaS°4 ~ 5%
Regenerator off-gas 1093.3/137.8 1558
Air to heat exchanger 121/165.4 1207
Regenerated off -gas to SRP 537.8/130.9 1558
39,890 C02 - 18.6%, H20 - 7.8%
S02 - 12.4%, N2 - 61.2%
2,888
18,319
Coal to SRP 93.3/137.8 3549 kg/hr 86.5 kJ/kg
Sulfur (80% recovery) 121/103.4 155
Tail-gas from SRP 148.9/110.2 1636
TT 1 f- -L - - |- -LT.IJ.T P^T/IHTA OO1*^K
Waste stone to cooler o/i/iuj.4 j££ -"
nr
ke
TT|- I r If nl O1T/1A1.A O'>O*X«
Waste stone lor disposal yj.J/iuj.H jzz r
Sulfur (-vlOO% recovery) 148.9/110.2 38.7
moles of Ca i^rt oo/
'
moles of Ca
-------
of a catalytic reactor where all of the sulfur in the tail-gas is converted
to H,,S either by hydrogenation or hydrolysis and an absorber where H S
is absorbed in sodium raetavanadate and eventually converted to sulfur.
DESIGN PARAMETERS
Experimental data obtained on batch regeneration units indicate that
a gas residence time of 1 to 2 s is required to approach the pseudo-
gas equilibrium. The solids residence time is then estimated by
V ,, N (XR - XR)
T /T = -a . ll=£i . Is si .
s g V e y
Using typical values of 0.25 for (XB - XR), 0.1 for y, 0.6 for e, and
1800 for V /V , the solids residence will be about 3000 times the gas
^ S (8)
residence time. Foster Wheeler estimated the required actual gas
residence time to be about 0.4 s. Based on this value, the solids
residence time can be expected to be about 19 min. Argonne's
experimental results on regeneration of sulfated Tymochtee dolomite
indicate that an increase in the solids residence time decreases the S0_
concentration but increases the extent of regeneration. At a temperature
of 1100°C and a solids residence time of 5 min., an S0_ concentration
(5)
of about 10 percent is predicted. According to their predictions,
decreasing the solid residence time to 2.5 min. increases the SO con-
centration to about 11.5 percent but decreases the regeneration from
60 to about 30 percent. Solids residence time has to be carefully
chosen to maximize the SO concentration at acceptable regeneration
levels. The extent of sorbent utilization in the combustor also has to
be taken into account. Argonne reported that a solids residence time of
about 12 min. is needed for complete regeneration at a temperature of
1100°C and a nominal gas residence time of about 0.5 s. The nominal and
actual gas residence times may be'defined as h /U and l^e/U, respectively.
71
-------
The fluidizing velocity may be expected to be in excess of 1.2 m/s
because of the fairly large size of particles used O500 to 1000 urn).
The maximum value for the fluidizing velocity may have to be limited to
about 2.4 m/s because of excessive particle elutriation at higher veloc-
(Q\
ities. Assuming a gas residence time of 0.5 s, the static bed depth
would vary from 0.6 to 1.2 m. Deeper beds result in increased pressure
losses and are, therefore, undesirable for atmospheric-pressure
operation.
Regeneration Temperature and Pressure
The reduction of the sulfated sorbent to oxide is favored by increased
temperature. The maximum temperature is limited by the ash-softening
temperature. A temperature of 1100°C is generally considered as a safe
temperature in the reducing atmosphere from the point of view of agglom-
eration in the bed. It was stated that for certain coals, an air-
/g\
operating temperature of 1315°C may be possible. Since the combustor
operates at a temperature of 870°C, the effect of thermal shock due to
repeated cycling on the attrition and reactivity of sorbent has to be
considered. Wheelock found a sharp reduction in the reaction rate
for the reduction of gypsum around 1260°C, where the material attained
a glassy appearance, indicating sintering. This suggests that the regen-
eration temperature should be less than 1260°C.
The equilibrium concentration of S0? is inversely proportional to
the pressure. Thus, it is advantageous to carry out the regeneration at
low pressures. At atmospheric pressure and a temperature of 1100°C,
the equilibrium concentration of S0? appears to be about 24 percent.
Heat and material balances, however, would limit the concentration of SO-
to about 10 to 12 percent when coal and air are the input streams to the
regenerator. At a pressure of 1000 kPa and 1100°C, the equilibrium
concentration of S0» appears to be about 2.5 percent. Heat and material
balances would not be a limiting factor at high pressures.
72
-------
Feed Location and Reducing-Gas Composition
In the regeneration of CaS04, CaS is also formed from a side reac-
tion. In order to minimize the amount of CaS in the bed, Exxon
created an oxidizing zone near the top portion of the bed by auxiliary
air injection above the fluidizing grid in their batch studies at
VLOOO kPa with propane as the fuel. They obtained about 40 percent CaS
without auxiliary air injection but obtained no significant CaS with
auxiliary air injection, possibly because of conversion of sulfide to
either sulfate or oxide. In Argonne's experiments on regeneration at
atmospheric pressure using coal, an oxidizing zone is established at the
bottom of the fluidized bed and a reducing zone near the top by inject-
ing coal well above the grid. They obtained about 5 and 1.1 wt % CaS
when the ratios of the nominal heights of the reducing zone to the
oxidizing zone were approximately 5 and 1.3, respectively. In the
latter case much better sorbent regeneration was obtained (71 percent,
compared to 41 percent). The substantial difference between Exxon's
and Argonne's results may be due to the difference in fuel used (which
affects the distribution of the reducing gas throughout the bed) and the
difference in the pressure (which changes the equilibrium concentration
of S0_, which, in turn, is likely to affect the concentration of CaS
in the bed). These two studies illustrate the importance of controlling
the oxidizing and reducing zones.
{
In addition to minimizing the formation of CaS and increasing the
regeneration, the two-zone reactor provides a higher concentration of SO
and is less sensitive to operating conditions. An oxidizing zone at the
top of the bed may provide more complete consumption of the coal and lower
the H and CO content of the outlet gas.
The experimental data on regeneration of CaSO supported on inert
(5)
material, obtained in TG apparatus, was analyzed by Argonne, and the
following equations were presented:
d[CaS04]/dt = -A [R.G.]°*8 [CaSO^] exp (-14,900/RT) ,
73
-------
where [CaSO ] and [R.O.1 represent the molar concentrations, t is the
regeneration time in seconds, T is the temperature in K, and R.H. is the
reducing gas. The value of A was reported as 3.36 for J^ or CH^ and
1.08 for CO. The regeneration rate, therefore, is three times lower for
CO than for H2 or CH^.
Process Options
The various process options that are important to the one-step
regeneration process are:
o Fuel for reducing gas
o Sorbent
o Recovery of sulfur
o Location of regenerator
o Sorbent size
o Disposal of spent sorbent
Coal/CH^/fuel oil
Limestone/dolomite
Elemental sulfur/sulfuric acid
Integral with 7BC/external
Single- vs double-screened feed
Before or after regeneration
The fuel required to supply the heat and the reductants to carry out
the regeneration of sulfated sorbent can be either coal or methane. The
gasification of coal or partial oxidation of methane can be performed
either integrally with regeneration or separatelv. Heat and material
balance calculations show that SO. concentrations of up to 1.5 percent
higher can be obtained in the case of coal than in the case of methane.
Natural gas, being in short supply, would not be a preferred option.
Argonne obtained better regeneration results with coal than with
methane in their studies with partial in-situ combustion of the fuel.
The concentration of the reducing gas in the regenerator is likely to be
more uniform for coal than for methane. The separate partial combustion
of the fuel to produce the reducing gas is not a preferred option
because it is less efficient, owing to higher heat losses and costs, but
it may have to be considered if the integral operation is not feasible.
Roth dolomite and limestone are equally effective sorbents for the
AFBB operation. The difference in regenerability between these two has
yet to be shown. In the case of limestone, a slight reduction in the
74
-------
cost of the sorbent circulation element can be expected because of its
lower molecular weight. The selection of a particular stone would depend
upon the availability, cost, attrition resistance, and the reactivity of
the stone for sulfation and regeneration with repeated cycling.
The recovery of sulfur can be carried out either by the production of
elemental sulfur or sulfuric acid. Production of elemental sulfur is a
preferred option due to its ease in handling, storage, and transport;
and it can be performed either by the Allied Chemical process or by the
RESOX process. If the concentration of SO- in the regenerator off-gas is
too low (<4 to 5 percent), a preliminary step for concentrating SO , such
as a Wellman-Lord process, is needed before either of the above processes
can be applied.
The RESOX process appears to be better suited to the regenerative
process than are other sulfur recovery processes. Its advantages are
the following:
Coal is used as the reducing agent.
No particulate removal from the sulfur-bearing gases is needed.
Conversion of a part of S0_ to H_S needed in other processes
is not required.
The disadvantages are:
e The process is uncertain since it is not yet commercially
developed.
o Sulfur recovery efficiency (about 80 percent) is lower.
e A particular type of coal may possibly be required, thus
necessitating a separate coal-handling system.
e A secondary sulfur recovery process is needed.
The sulfur produced may not be of commercial grade.
Regenerator Vessel Location
Several possibilities for the location of the regeneration vessel in
relation to the combustor have been conceived in the past. Pope, Evans
75
-------
and Robbins experimented with the regenerator attached to the boiler,
with an orifice cut between them for transferring solids. It was not
successful because, due to agglomeration near the orifice, the flow of
solids between the vessels could not be controlled. They also tried
regeneration in the carbon burn-up cell without much success. The SO^
concentration was too low because of the need for maintaining high
fluidizing velocity.
In the present study a single regeneration vessel was chosen for all
of the six beds where the transfer of solids is by means of dilute-phase
transport using hold vessels. The boiler and regeneration vessels can not
be located close to one another. Another possibility is to locate a
regeneration vessel close to each bed of the combustor and carry out the
solids transfer by means of dense-phase transport via two inclined legs
connecting the two vessels with pulses of gas controlling the flow of
solids. This arrangement does not appear to be suitable for the present
study since control and turndown would be complex and the economics would
be unfavorable.
Sorbent Size
The selection of a narrow or wide particle size distribution for the
sorbents would depend on the rate of sulfation in the boiler and the rate
of reduction of the sulfated sorbent as a function of the particle size,
as well as such factors as attrition and elutriation of particles in the
boiler and the regenerator. The results reported to date on the sulfa-
tion and regeneration of limestones and dolomites indicate that the
particle size is not a critical factor within the size range (^500 to
1000 ym) normally used in fluidized-bed combustion. Since the
residence time of the coarse particles is greater than that of elutriate
fines, the reduced activity of the coarse particles is compensated for
by the longer residence time. If the fines are recycled, they can be
expected to increase the utilization of the sorbent. Argonne
found no significant difference in sulfur removal with limestone
particles of 50 to 600 urn diameter.
76
-------
The amount of coal ash in the bed should be minimal in order to keep
the bed chemicallv active. Coal with a size distribution different from
that of the sorbent could help in removing the ash. The sorbent size dis-
tribution may be an important factor in controlling the ash content of the
(9)
bed. Exxon ' observed that more slag ash was retained in the regenerator
with smaller sorbent sizes. In interpreting the experimental results on
sorbent deactivation as a function of the particle size, this factor must
be considered.
In view of the above considerations, a single-screened feed is pre-
ferred to a double-screened feed that would increase the operating cost
considerably. Fines should be screened since thev will increase the rate
of elutriation from the regenerator and could cause agglomeration.
The elutriation of sorbent fines present in the feed and those
resulting from attrition of the sorbent is an important consideration in
the particulate control of the regenerator off-gas and also in estimating
the Ca/S makeup ratio. It can be expected that the sorbent elutriation
is additive to the elutriation of ash and carbon fines. The elutriation
rates of the sorbent are verv difficult to determine in the case of cyclic
absorption-regeneration and are dependent on the operating conditions of
the combustor and the regenerator. ANL reported a sorbent loss of 6 per-
cent per cycle because of attrition followed by elutriation from the
reactors in a ten-cycle experiment. Exxon found in their miniplant
studies that the sorbent losses through attrition and entrainment were
very low during an integrated operation of the combustor and the
regenerator. This result shows that regeneration may increase the
attrition resistance of the sorbent. Additional work is needed to estimate
the particle carry-over from the regenerator.
Coal ash present in the regenerator may react with the fine sorbent
particles, placing an added burden on the particulate removal equipment.
This reaction may lower the slagging tendencies of the ash and prevent
deposition of corrosive solids on cyclone surfaces. An agglomerating
77
-------
tendency among the sorbent fines can be expected above the ash fusion
point because the molten ash may wet the sorbent particles, causing
them to become sticky. Coals with low ash-fusion points may not be
suitable as fuels for the regenerator. ANL reported that the ash-fusion
temperature of Arkwright and Sewicklev coals under reducing conditions
is close to 1100°C, which is a desirable temperature level for
regeneration.
Spent jforbent Removal
Part of either the sulfated or the regenerated sorbent must be dis-
carded, since their activity is reduced by repeated cycling, requiring
a certain makeup sorbent rate. It is better to discard the sulfated
sorbent than the regenerated sorbent, from the point of view both of the
reduced load on the regenerator and of the greater environmental accept-
ability of the sulfated stone. In the present studv it is assumed that
the spent stone is not processed but passes through a cooler/conveyor
to reduce its temperature before disposal.
PERFORMANCE PROJECTIONS
The single most important variable of the process is the concentra-
tion of SO in the regenerator effluents, which depends on the type of
fuel used in the regenerator, temperature, pressure, heat losses, and the
change in the utilization of calcium across the regenerator. The concen-
tration of SO in the regenerator effluent was estimated to be about
12 percent. The effect of various factors on the maximum concentration
of SO that can be achieved has been studied from material and energy
balance considerations, as reported in Appendix A.
Ca/S Makeup Ratio
The required calcium/sulfur molar feed ratio depends on the activity
of the sorbent in the boiler and the regenerator and the rate of circula-
tion of solids between the two processing steps. Through analysis of TG
data, from atmospheric-pressure operation of once-through systems,
Ca/S makeup ratios of 2.8/1 and 2.2/1 have been projected for calcined
78
-------
limestone and dolomite, respectively, for a temperature of 816°C. At
temperatures above R43°C, however, sulfur removal efficiency was drasti-
cally reduced. At a temperature of 954°C, Ca/S nakeup ratios of 5/1 and
about 4/1 were projected for calcined limestone and dolomite, respectively.
The projections on Ca/S makeup ratio for regenerative systems have been
very few. Pope, Evans and Robbins^" '' estimated a ratio of 1/1, and
from recent bench-scale experimental data Argonne projected a
ratio of about 0.35. This latter value is based on limited small-scale
data on inadequate analyses. Additional data are needed for a reliable
estimate of this ratio.
Argonne presented an analysis to estimate the Ca/S nakeup ratio
based on (a) experimental data on total CaO/S mole ratio required at
75 percent sulfur retention as a function of the sulfation cycle, and
(b) an analytical expression for the age (number of cycles) distribution
of the sorbent in the boiler at steady-state as a function of the makeup
rate. Sorbent utilization for nth sulfation cycle was represented by
0.92 exp (-0.139n), where n is the number of the sulfation cycle. A
(9)
similar analysis was performed by Exxon.
'Thermal Efficiency
The combustion efficiency in AFBB systems can be expected to be about
90 percent without a carbon burn-up cell, the inefficiency resulting
mainly from the carry-over of carbon fines. In an atmospheric
regenerator higher carbon losses can be expected because of (a) the
reducing atmosphere, (b) the absence of any internals, and (c) the reaction
of carbon with CO , resulting in the formation of CO near the top
portion of the beds and its subsequent loss through the regenerator off-
gas. The compensating factors in favor of regeneration are the highest
operating temperature and lower fluidization velocity (^1.2 to 2.4 m/s
as opposed to -VL.8 to 3.6 m/s in the combustor). Thus, it is desirable
to operate the fluidized bed of the regenerator at a low fluidization
velocity from the point of view of reducing the carbon loss. The heat
losses of the regenerator have been estimated to be in the range of 0.5
79
-------
to 1 percent. The energy requirement of the regenerator for a 635 MW
plant has been estimated to be ^3 percent of the energy requirement of
the boiler (Appendix A). Assuming the carbon loss in the regenerator is
about 15 percent, this represents about 1/2 percent of the fuel input to
the boiler, which may not be a serious loss. The effect of CO and
carbon fines present in the regenerator off-gas on the sulfur recovery
system should be considered.
Environmental Impact
The environmental impact of the regeneration process depends on the
particulate emission rate, trace element release, CO emission, combustion
efficiency, NO emission, S0? emission from the sulfur recovery plant,
X L,
the rate at which spent sorbent is produced, and the method of spent
sorbent processing. Since the fuel requirement of the regeneration process
is about 3 percent of the fuel required by the boiler, the increase in the
emission of particulates and other pollutants can be expected to be small
with, hence, no significant impact on the power plant emissions. There
is evidence that the regeneration process increases the attrition resist-
ance of the sorbent and, thus, reduces the loss of sorbent fines. The
emission of trace elements will in all likelihood be lowered because of
the reduced sorbent requirement of the process. The data available on
the release of trace elements by the sorbent in the boiler are limited.
The emission of sulfur compounds from the regeneration process
depends on the sulfur recovery process. For the basic design presented
here, the emission of S0«, COS, and other sulfur compounds can be
expected to be less than 100 ppm, of which less than 1 ppm is present as
(14)
H S. If the Allied Chemical process is chosen, the incineration
of the tail-gas from the sulfur recovery plant (SRP) must be accomplished
to reduce the H_S emissions.
Reliability
The reliability of the regeneration system will be affected by fac-
tors such as the need for a special type of coal; the problem of hot
80
-------
spots; and agglomeration of the sorbent due to fusion of coal ash, deac-
tivation of the sorbent, attrition, and decrepitation of the sorbent;
particle carry-over; and so on. The operating conditions of the regener-
ator nay have to be modified if a particular type of coal is desired for
regeneration. The problem of hot spots may not be as severe in an
atmospheric regenerator as in a pressurized regenerator. There is no
adequate information on the attrition and the deactivation of various
sorbents. The ability to minimize the formation of CaS by adequate con-
trol of the lengths of oxidizing and reducing zones is an important
factor.
The circulation of solids between the combustor and the regenerator
at high temperature is another problem area that will affect the reliabil-
ity of regeneration. With the modular design solid feeding is simpler,
because of the possibility of side feeding, but the distribution of solids
is complex. Piping and controls are complex, but start-up is easier,
since one module can be started at a time. The modular design has better
turndown capabilities.
It is expected that the AFBB will be operated at turndown ratios of
up to 4:1. With the system design based on four regeneration modules,
each unit may be shut down completely, depending on the turndown require-
ment, while other units continue to operate at full capacity. Thus, the
performance of the regeneration system can be maintained at any turndown
ratio, increasing the plant reliability.
COST ESTIMATE
The equipment of one of the four modules considered in an economic
evaluation of the process is shown in Figure 19. Estimates for the
investment cost were prepared on the basis of the equipment costs given
in a 1975 Westinghouse Report on oil gasification and the informa-
tion obtained from manufacturers. "The process investment cost and the
energy cost as a function of the process sulfur load are shown in Fig-
ures 20 and 21, respectively. Figure 22 shows a comparison of the process
81
-------
00
Utilized
Sorbent
from
Boiler
Sulfur Recovery Plant
Resox and Beavon Processes
Spent Stone
Cooler/Conveyor 1 s s D
C.F. -Coal Feeding System
C. W. - Cooling Water
S. S. D. - Spent Stone Disposal
U.S. -Utilized Sorbent
R. S. - Regenerated Sorbent
Figure 19. Atmospheric One-Step Regeneration Schematic Flow
Diagram of One Module
-------
Curve 687115-A
cu
E
1/1
o>
o
o
48
42
36
30
24
18
12
i r
635 MW Plant
Cone, of SCL = 0.12
T
PSL - Ib of S Handled by Regeneration
per Ib of Coal to Combustor
0.01 0.02 0.03 0.04
Process Sulfur Load
0.05
0.06
Figure 20. Capital Cost of Regenerative Process as a
Function of PSL
83
-------
Curve 687042-A
«/) O
o :>
CP
OJ
c
635 MW Plant
Cone, of SO =0.12
PSL - Ib of S Handled by Regenerator per
Ib of Coal toCombustor
0.01 0.02 0.03 0.04 0.05
Process Sulfur Load
Figure 21. Energy Cost of Regenerative Process as a
Function of PSL
84
-------
60
50
-w-
CD
CO
CD
uo
in
CD
U
O
40
30
20
10
Curve 687118-A
635 MW Plant
Cone, of S02 = 0.12
PSL - Ib of S Handled by Regenerator per Ib of Coal
to Combustor
Resox with Beavon Process for
treating Resox Tail Gas
Allied Chemical Process with
Tail-Gas Incineration
1
0.01 0.02 0.03 0.04
Process Sulfur Load
0.05 0.06
Figure 22. Comparison between Cost of EESOX and Allied
Chemical Processes as a Function of PSL
85
-------
investment and the energy costs of two options of the sulfur recovery
system namely: the RESOX process, with the Beavon process for treating
the RESOX tail-gas, and the Allied Chemical process with incineratin of
the tail-gas.
The sulfur recovery element is by far the most expensive system.
For a process sulfur load of 0.025, its cost accounts for more than
60 percent of the total investment cost. The sorbent circulation element
is the least expensive of the three elements. The effect of PSL on the
cost of the sulfur recovery element is substantially higher than on the
cost of the regeneration element or the sorbent circulation element.
Hence, it is desirable to have as low a load as possible on the sulfur
"D
recovery element. From the effect of m and X_ on PSL, one can see
that for low values of m<>0.3), the effect of X on PSL is relatively
small, but at high values of mO^l.O) the effect is very significant.
It is obvious that if the sorbent is to be disposed of before regenera-
tion, it is desirable to have a high value of X^; if the sorbent is
£">
to be disposed of after regeneration, it is desirable to have a high
T)
value of X .
O
The capital cost/capacity exponent for the regenerative process can
be taken as 0.65. the capital cost given here includes the indirect costs,
contingency, and fee, but not the interest during construction.
The energy cost is computed on the following cost assumptions :
Cost corresponds to the end of 1976.
Capital charges plus operation and maintenance are 20 percent of
the total cost.
Contingency is 20 percent and contractor fees are 3 percent of
the base cost.
No interest during the construction period is included.
Coal is $2.0 per Mg and dolomite is $5 per Mg.
Waste stone disposal cost is $3 per Mg.
Electricity is 23 mills /kWh.
3
Process water is $0.10 for 3.8 m .
86
-------
There is no credit for recovered sulfur.
- The capacity factor is 70 percent (6132 hours of operation in a
year).
If credit for sulfur is given at $25/Mg for a process sulfur load
of 0.03, the total energy cost will be reduced bv 0.28 mills/kWh.
The energy cost of the once-through and regenerative options for
three regeneration Ca/S makeup ratios is given in Table 11. For dolomite
TABLE 11. ENERGY COST
(mills/kWh)
Once-through
Regeneration Ca/S Ratio
0.2
0.6
1.0
Capital Charge,
plus O&M
3.79
4.72
4.74
4.76
Coal
Dolomite
Total
8.35
1.54
13.68
8.80
0.14
13.66
8.80
0.42
13.96
8.80
0.70
14.26
Assumptions:
The:Ca/S ratio is 2.2 for the once-through option, while meeting
current NSPS.
The cost of boiler equipment only based on a 1971 Westinghouse Report
is included."''
An escalation index of 1.6 is used to update the cost figures to
correspond to the end of 1976. This results in a boiler plant equip-
ment (includes all solid handling systems) cost of $90.5/kW.
The percent of sulfur in coal is about 4. This corresponds to a PSL
of 0.025 in the case of regeneration.
Other assumptions are the same as previously.
87
-------
purchased at $5/Mg, a Ca/S ratio of 0.2 is required by the regeneration
process in order to break even with the once-through process. Higher
sorbent costs will reduce the break-even Ca/S ratio.
Recycling of RESOX Tail-Gas to Boiler
The RESOX tail-gas can be recycled to the boiler instead of being
treated in a secondary sulfur recovery process such as the Beavon process.
The investment for cost recycling as a function of SO. concentration in
the regenerator off-gas is as follows:
% so2
Investment
cost, $/kW
4
3.06
8
1.91
12
1.45
The above costs are estimated on the assumption that the operation of
the boiler is not affected by the recycling of the RESOX tail-gas.
They compare favorably with those of the Beavon process, which has an
investment cost of about $4/kW for a 12 percent SO- concentration.
A cost sensitivity analysis of the process is shown in Table 12.
The important design options to be considered are: a separate coal
gasifier versus an integrated operation to produce the reducing gas;
a change in the number of regenerator modules; and the production of
elemental sulfur versus sulfuric acid and the process to be adopted.
The energy cost of the once-through and regenerative options is
compared for three different sorbent costs in Table 13.
The energy cost of the regeneration option for production of both
elemental sulfur and sulfuric acid is given in Table 14.
The energy cost of the once-through and regenerative options is
compared for three different coal costs in Table 15.
88
-------
Curve 687117-A
CD
oo
CD
to
CD
O
O
635 MW Plant
PSL - Ib of S Handled by Regenerator per Ib
of Coal to Combustor
20 -
10
0.01 0.02
0.03 0.04
Process Sulfur Load
0.05
0.06
Figure 23. Process Investment as a Function of PSL for
Different Concentrations of S02 With Allied
Chemical Process for Sulfur Recovery
89
-------
Curve 687116-A
50
CD
E
CD
CD
O
40
30
co
'o
O
o
20
10
I I I I
635 MW Plant
PSL - Ib of S Handled by Regenerator per Ib of
Coal to Combustor
Sulfur Recovery Element
Regeneration Element
1
I
20
16
0.01 0.02 0.03 0.04
Process Sulfur Load
0.05
-w-
c
CD
E
12-
*' 1 1 1
ro
i_
CD
O
o
0.06
0
Figure 24. Cost of Sulfur Recovery and Regeneration
Elements for Different % SC>2 with Allied
Chemical Process as a Function of PSL
90
-------
TABLE 12. COST SENSITIVITY OF THE ATMOSPHERIC REGENERATIVE PROCESS
Variables
Sorbent-Type Lime-
stone vs Dolomite
Decrease in Regener-
ator Modules from
4 to 1
Separate Coal
Gasifier
Sulfur Recovery
Process
Sulfur vs Sulfuric
Acid
Decrease in Sorbent
Utilization
Decrease in Regenera-
tor Temperature
Recycling of RESOX
Tail-Gas to Boiler
Regeneration
Element
Reduction in
cost of about
$2.50/kW
Increase in
cost of about
$2.70/kW
Sorbent Sul
Circulation Reco
Reduction of $0.70/kW for
limestone
fur Total Regenerative
very Process
Reduction of $0.70/kW for
limestone
Reduction in cost of Appreciable reduction, $3.75/kW reduction plus
$1.25/kW not estimated. that due to sulfur
recovery
Becomes necessary when
the integrated operation
is not feasible
No significant difference RESOX appears to be
between the Resox and potentially advantageous
Allied processes
Drastic reduction in the Large impact on the
cost process
May not be
significant
Negligible
Additional Factors
to Be Considered
Difference in regener-
ability has yet to be
shown
Feasibility, effect on
turndown ratio, shop vs
field fabrication
Performance
Needs further R&D effort
Marketability, should be
studied in detail
Significant increase in Cost increases due to Appreciable effect at Change in reactivity and
cost. A minimum of 0.2 lower levels of SO, lower levels of attrition characteristics
is recommended. utilization
Cost increases due to Cost increases due to Appreciable effect Process may not be
lower levels of lower levels of SO feasible below a
utilization certain level
Reduction of
$2.50/kW Reduction of S2.50/kW Effect on boiler
operation
-------
TABLE 13. ENERGY COST
(nills/kWh as a function of the cost of fresh sorbent plus
cost of disposal of spent stone per Mg)
(Ca/S = 1 for regeneration, Ca/S = 2.2 for
once-through)
$4.00
Once-
through
Regenera-
tion
$8.00
Once-
through
Regenera-
tion
$12.00
Once-
through
Regenera-
tion
Capital
Charge
+ O&M
Coal
Dolomite
Total
3.79
8.35
0.77
12.91
4.76
8.80
0.35
13.91
3.79
8.35
1.54
13.68
4.76
8.80
0.70
14.26
3.79
8.35
2.31
14.45
4.76
8.80
1.05
14.61
TABLE 14. ENERGY COST
(units/kWh, showing effect of regenerative
systems by a product)
Once-through
Regeneration
Production of
Sulfur
Capital Charges 3.79 4.76
+ O&M
Coal 8.35 8.80
Dolomite 1.54 0.70
Total 13.68 14.26
Production of
Sulfuric Acid
4.33
8.80
0.70
13.83
92
-------
TABLE 15. ENERGY COST
(mills/kWh as a function of the cost of coal per Mg)
$15
Once-
through
on
Regenera-
tion
$20.00
Once-
through
Regenera-
tion
$25.00
Once-
through
Regenera-
tion
Capital
Charges
+ O&M
Coal
Dolomite
Total
3.79
6.26
1.54
11.59
4.76
6.60
0.70
12.06
3.79
8.35
1.54
13.68
4.76
8.80
0.70
14.26
3.79
10.44
1.54
15.77
4.76
11.0
0.70
16.46
From an examination of Tables 11, 13, and 14, the regenerative
option appears economically less attractive than does the once-through
option, in general. A comparison of energy costs of both options for
different Ca/S ratios shows that at a low makeup rate of 0.2, regenera-
tion becomes competitive. Options such as recycling the RESOX tail-gas
or producing sulfuric acid rather than sulfur are likely to reduce the
cost of regeneration and make it more economical.
For a Ca/S makeup ratio of 1.0, the break-even point between the
regeneration and once-through options is likely to be in the range of
$12 to 16/Mg of sorbent used. On the basis of a market study on the
cost of limestones and dolomites, the cost of fresh sorbent
delivered can be expected to be in the range of $4 to 8/Mg in most
areas of the nation. The cost of disposal may be expected to be in the
range of $2 to 4/Mg of waste stone, at which costs a regeneration Ca/S
makup of 0.2 would be necessary to break even. An examination of
Tables 13 and 15 shows that the cost of sorbent is very important in
comparing the two options, but the cost of coal is insignificant.
93
-------
For a Ca/S makeup ratio of 1.0, the option of producing sulfuric
acid rather than sulfur does not make the regeneration cheaper than
once-through, even though the cost of sulfur recovery decreases by more
than half.
ASSESSMENT
Conclusions
The cost of the regenerative process is based on the
assumption that a concentration of SO- of about 12 per-
cent can be obtained. This has yet to be demonstrated
experimentally.
The sulfur recovery system is the dominant subsystem
in the regenerative process and costs more than three
times that of the regeneration system. The sorbent
circulation system is of minor importance.
* A comparison of the RESOX process and the Allied
Chemical process for the sulfur recovery system shows
that the latter appears to be cheaper for the particular
regenerative process studied here. Since the cost is
based on preliminary estimates, this difference may not
ultimately be significant.
A comparison of the once-through option with the regen-
erative option shows that the process investment cost is
about 20 percent higher and the energy cost about 4 per-
cent higher for the latter option. This comparison is
based on the assumption that the spent stone does not
need further processing before disposal.
Recycling RESOX tail-gas to the boiler appears to be more
attractive than treating it in a secondary sulfur recovery
process.
If sulfur is recovered in the form of sulfuric acid rather
than elemental sulfur, the capital cost of the regenerative
option can be reduced by about 10 percent.
94
-------
The regenerative option might become competitive with the
once-through option if the makeup ratio of Ca/S can he
reduced to about 0.2 for a sorbent cost of about (fresh
stone plus disposal) $8 per Mg.
For a Ca/S makeup rate of 1.0, regeneration is likely
to break even at a sorbent cost (fresh stone plus dis-
posal of $12 to 16/Mg.
The need for spent stone processing because o£ possible
future environmental constraints is not known. The
potential environmental benefits of sorbent regeneration
resulting from a reduced quantity of spent sorbent must
be weighed against increased coal consumption and auxiliary
power, and must consider the environmental nature of the
spent sorbent in addition to costs.
If sulfur vapor rather than sulfur dioxide can be produced
in the regenerator, regeneration is likely to become
attractive. The scope and the need for process innova-
tions in sulfur recovery are thus evident.
Research and development should be continued to demon-
strate the technical feasibility of the regenerative
process and to make improvements in the sulfur recovery
process.
Development Requirements
Areas in which work needs to be continued and that are important to
the regenerative process can be identified as the following:
Development of a sulfur recovery process suitable for the
atmospheric one-step regeneration process
Determination of the maximum percent of SO- in the
regenerator effluent that can be achieved in a contin-
uous operation of the combustor-regenerator system
Change in the activity and the regeneratility of the
sorbent with repeated cycling
95
-------
Separation of sorbent and ash in the regenerator
Investigation of the possibility of hot spots in the
regenerator bed that will contribute to fusion of ash
and sorbent fines
Disposal/utilization of the spent sorbent
Effect of coal fly ash on the deactivation of the sorbents.
Reliability
The reliability of the regeneration process is affected by several
factors, such as the type of fuel, the sorbent characteristics, and
operating problems, all of which are being investigated. The present
development effort is in a preliminary stage and it is, therefore, dif-
ficult to make accurate projections on capital and operating costs,
technical feasibility and performance, environmental impact, and reli-
ability = The analysis and various projections, based on the limited data
available made in the current study, should be viewed as preliminary.
A study on alternative sorbents being conducted at Argonne, Exxon,
and Westinghouse might give rise to new regeneration systems. The
possibility exists that new and advanced concepts are under consideration
at several research organizations. It is hoped that adequate attention
will be paid in the future to developing a sulfur recovery process suit-
able for regeneration. It does not appear sound to try to solve problems
of regeneration without first making it an economical process; to do so
requires a concerted effort in all areas of regeneration.
96
-------
SECTION 6
REGENERATION FOR PRESSURIZED FLUIDIZED-BED COMBUSTION
Process evaluations for pressurized fluid-bed combustion sorbent
regeneration systems were reported in 1973. ' An update of that
work is presented here, based on current performance expectations and
an improved understanding of major component costs. The economics and
performance of two reductive decomposition schemes, one operated at about
1000 kPa and one at about 100 to 200 kPa, and a two-step regeneration
scheme were estimated. Costs and performance are compared with
once-through PFBC and conventional power plants with stack-gas
cleaning. The designs are conceptual in nature and were not based
on sensitivity analysis or optimization.
PROCESS OPTIONS
Major process options are listed in Table 16. Several of these
options have been previously described.
Numerous process options associated with the regeneration process
operating conditions (reaction temperatures, fluidization velocities,
bed depths, sorbent particle size, etc.), process flow logic (heat
exchange between streams, temperature control, energy recovery, etc.)
and minor equipment selection also exist.
Evaluation Basis
The power plant basis listed in Table 17 has been applied in the
assessment. The process sulfur load, reflecting in part the sulfur con-
tent of the coal, is varied from 0.01- to 0.06. Important process charac-
teristics are given in Table 18 as a function of the process sulfur load.
97
-------
TABLE 16. PROCESS OPTIONS
Regeneration Reaction Scheme: reductive decomposition, two-step
process (calcium sulfide formation followed by H-S generation
by steam and C0? reaction)
Regeneration Reactant Gas Generation: external generation, in-situ
generation or purchase
Fuel for Reactant Generation: methane, fuel oil, coal, etc.
Sources for C02 Recovery: stack gas, sulfur recovery tail-gas,
pressurized combustor gas, external combustion gas, purchase
Process for External Reductant Generation: gasification, partial
oxidation, reforming, etc.
Sulfur Recovery Form: elemental sulfur, sulfuric acid
Sulfur Recovery Process: Allied Chemical process, Foster Wheeler
RESOX process, etc. (see Section 4)
Process for C02 Recovery: Benfield Hot Potassium Carbonate
process, etc. (see Section 4)
Spent Sorbent Processing: none, dry sulfation and carbonation,
etc. (see Volume V)
Handling Sulfur Recovery Tail-Gas: recycle to combustor tail-gas
cleaning, incineration and exhaust to environment
Regenerator Pressure: 100 to 1500 kPa
Sorbent Type: dolomite or limestone
Layout: number of modules, number of parallel trains, etc.
Type of Sorbent Circulation System: dilute pneumatic transport,
dense-phase pneumatic transport, etc. (see Section 4)
98
-------
TABLE 17. POWER PLANT BASIS
Plant Capacity - 635 MWe (based on once-through sorbent power plant
performance)
Plant Heat Rate - 9040 kJ/kWh (based on once-through sorbent
performance)
Combustor Excess Air - 17.5%
Combustor Pressure - 1013 kPa
Process Sulfur Load - 0.01 to 0.06
S02 Emission - meets current EPA standard of 516 ng/J
Sorbent Type - dolomite
Layout - four pressurized boiler modules, four parallel regeneration
trains, single sulfur recovery plant
Spent Sorbent Processing - none; sorbent is disposed of following
regeneration
Sorbent Circulation System - dilute pneumatic transport
Sulfur Recovery Tail-Gas - incinerated and exhausted
Three regeneration schemes are evaluated: a pressurized reductive
decomposition scheme, an atmospheric-pressure reductive decomposition
scheme, and a two-step process. The specific process options for
each of these regeneration processes was selected on the basis of the
results of previous engineering assessments and are presented in
Table 19. Selected regeneration process operating conditions and pro-
jected performance levels are summarized in Table 20. Sulfur dioxide
concentrations of 1 and 2 vol % from the regenerator are examined
for the 1000 kPa reductive decomposition process because the achievable
level for this critical performance factor has not been demonstrated.
These SO levels are suggested by small-scale experimental work and
thermodynamic predictions. A level of 10 vol % is assumed for the
low-pressure reductive decomposition process. An I^S level of 3 vol %
is assumed for the two-step regeneration process. The combustor
99
-------
TABLE 18. PROCESS SULFUR LOAD
Coal
P
0.06
Sulfur, wt % 7.2
Combustor Sulfur Removal
Efficiency, %
-------
TABLE 19. SELECTION OF PROCESS OPTIONS
Reductive Decomposition Processes
Reductant gas generation - in situ with regenerator
Fuel for reductant - coal
Sulfur recovery form - elemental sulfur and sulfuric acid evaluated
Sulfur recovery process - Allied Chemical process with methane
reductant (see Section 4)
Two-Step Regeneration Process
Reductant gas generation for sulfide generation step - in situ
with first-step reactor
Fuel for reductant - coal
Source for CO- recovery - stack gas
Sulfur recovery form - elemental sulfur
Sulfur recovery process - Stretford process (see Section 4)
C02 recovery process - Benfield Hot Carbonate process (see
Section 4)
Recarbonation of sorbent by stack gas contacting prior to
regeneration
Process Performance Projections
Some key performance characteristics of the PFBC regeneration sys-
tems evaluated are summarized in Table 21 as a function of the process
sulfur load. Auxiliary power requirements (for the sulfur recovery
process, for the compression of air and stack gas, and for sorbent circu-
lation), the rate of coal consumption for regenerator reductant, the rate
of methane consumption for sulfur recovery, and the rate of steam consump-
tion are estimated. The regeneration processes are large power and fuel
consumers, and the process designs must be concerned with maximum energy
recovery. The energy content of the regenerator product gas is used to
provide the regeneration process auxiliary power requirements. No energy
101
-------
TABLE 20. OPERATING CONDITIONS AND PERFORMANCE PROJECTIONS
Reductive Decomposition
Regenerator pressure, kPa
Regenerator temperature, °C
S09 mole percentage produced
Sulfur recovery efficiency, %
Dolomite utilization in
boiler, %
Dolomite utilization after
regeneration, %
Dolomite makeup rate, Ca/S
Fluidization velocity, m/s
Boiler conditions
Calcium sulfide in sorbent,%
Two-Step Regeneration
Atmospheric Pressure Pressurized
150
1050
10
90
30
10
0.5-1.0
1.5
Calcining
0
CaSO, reducer pressure, kPa/temperature, °C
H-S generator pressure, kPa/temperature, °C
ELS mole percentage produced
Sulfur recovery efficiency, %
Dolomite utilization in boiler, %
CaSO, reduced to CaS in reducer, %
Dolomite utilization after regenerator, %
Dolomite makeup rate, Ca/S
Fluidization velocity, m/s
Boiler conditions
CaO recarbonated in precarbonator, %
Ratio of steam to CO. in reactant gas
1000
1100
1-2
90
30
10
0.5-1.0
1.5
Calcining
0
900/815
1100/680
3
90
30
100
10
1.0-2.0
1.5
Calcining
50
1
102
-------
TABLE 21. PERFORMANCE PROJECTIONS
Process Sulfur Load
Reductive Decomposition
1% S02
0.06 0.03 0.01
27, S0?
0.06 0.03
0.01
10% SO
0.06
0.03
0.01
Two-Step Process
3% H2S
0.06
0.03
0.01
o
U)
Auxiliary Power, MWe
Coal Consumption,
% of boiler coal
input
Methane Consumption,
kJ/hr, 10*1
Steam Consumption,
% of power plant
fuel input
Technical
Uncertainties
70 37
69 35
272 136
0 0
15
12
41
0
Enerpy recovery,
sulfur recovery
41 22 10
34 17 6
272 136 41
000
Energy recovery,
sulfur recovery
21 12 6
631
260 130 40
000
Temperature con-
trol, solids
circulation sys-
tem operability
69
29
0
5
Energy
39 15
15 5
0 0
3 1
recovery,
sorbent deacti-
vation
-------
is exported from the regeneration process to the plant power cycle in
this evaluation, although this may be called for in an optimized power
plant. The sulfur added to the regeneration process in the reductant
coal has been neglected, though it will have a very significant impact
on the processes in most cases.
Several technical uncertainties exist for each of the regeneration
schemes. The high-pressure reductive decomposition processes (1 and
2 percent SO ) require very large coal inputs and auxiliary power con-
sumption. The efficiency and operability of energy recovery is tech-
nically uncertain, along with the operability and controllability of sulfur
recovery with such low SO concentrations. The two-step regeneration
process is characterized by the additional technical uncertainty that the
regeneration process itself mav be effective in producing an active sor-
bent material. The low-pressure reductive decomposition process consumes
power and coal at a lower rate, but its operability and reliability is in
question because of the complexity of the solids circulation system.
Capital Investment
Estimates of capital investment for the regeneration processes (not
for the total PFBC power plant) have been developed with the following
basis:
Mid-1977 costs
635 MWe power plant
Interest during construction, general items, and engineering are
not included.
All other direct and indirect cost items are included.
The estimated investments are presented as a function of the process sul-
fur load in Tables 22 through 25.
The most expensive process section for the pressurized reductive
decomposition process is the sulfur recovery or sulfuric acid recovery
section. The sorbent circulation section is the most expensive section
104
-------
TABLE 22. INVESTMENT FOR PRESSURIZED REDUCTIVE
DECOMPOSITION PROCESS - 1 PERCENT SO $/kW
Process Sulfur Load
Process Section
Regeneration
Sorbent Circulation
Sulfur Recovery
(Sulfuric
Total
0.06 0.03
15.6 10.6
16.9 16.1
92.3 60.8
Acid Recovery (55.3) (36.5)
124
.8 (87.8) 87.5 (63.2)
0.01
4.8
15.1
29.5
(17.6)
49.4 (37.5)
TABLE 23. INVESTMENT FOR PRESSURIZED REDUCTIVE
DECOMPOSITION PROCESS - 2 PERCENT S02, $/kW
Process Section
Regeneration
Sorbent Circulation
Sulfur Recovery
(Sulfuric Acid Recovery)
Total
Process Sulfur Load
0.
10
16
68
(44
95.6
06
.5
.9
.2
.2)
(71.6)
0.03
5.7
16.1
45.0
(29.2)
66.8 (51.0)
0.01
3.2
15.1
21.8
(14.2)
40.1 (32.5)
105
-------
TABLE 24. INVESTMENT FOR LOW-PRESSURE REDUCTIVE
DECOMPOSITION PROCESS - 10 PERCENT SO., $/kW
Process Section 0.06
Regeneration 11.4
Sorbent Circulation 31.9
Sulfur Recovery 27.7
(Sulfuric Acid Recovery) (18.4)
Process Sulfur Load
0.03 0.
7.8 3
31.1 30
18.3 8
(12.2) (5
Total 71.0 (61.7) 57.2 (51.1) 42.0
01
.1
.1
.8
.9)
(39.1)
TABLE 25. INVESTMENT FOR TWO-STEP PROCESS, $/kW
Process Section
Regeneration
Sorbent Carbonation
Sorbent Circulation
C02 Recovery
Sulfur Recovery
Total
Process Sulfur Load
0.06
7.
41.
16.
10.
36.
113.
3
7
9
7
9
5
0.03
5.1
27.1
16.1
7.5
24.3
80.1
0.01
2.8
12.5
15.1
3.8
11.8
46.0
106
-------
for the low-pressure reductive decomposition process, requiring complex
lockhoppers with water-cooled valves. Although sulfuric acid recovery
is considerably cheaper than sulfur recovery, the market and storage
questions must be resolved for each specific location. Sulfur recovery
has generally been selected as the preferred option.
Sulfur recovery and sorbent recarbonation (by stack-gas carbonation
of the sorbent in fluidized-bed reactors) represent the most expensive
process sections for the two-step regeneration process.
While the low-pressure reductive decomposition process has the lowest
capital investment, the major technical uncertainties associated with the
sorbent circulation system must be acknowledged.
Energy Cost
Energy costs associated with each of the regeneration processes have
been projected using the following basis:
Costs for the regeneration process only do not include
the cost of coal, capital charges, and operating and
maintenance charges for the balance of the power plant.
Interest during construction included at 7-1/2%/yr, 3-1/2 yr con-
struction time
Mid-1977 costs
Capital charges 15%/yr
Operating and maintenance cost 5% of investment per year
70% plant capacity factor
Sulfuric acid recovery not considered
No credit for sulfur produced
Coal at $0.80/GJ Cv$80/Mg)
Methane at $1/GJ
Dolomite at $10/Mp (purchase plus disposal)
Sorbent Ca/S ratio of 1.0 for all three process sulfur loads
to achieve current new source pollution standards.
107
-------
For a once-through sorbent operation with dolomite, the required Ca/S
ratios as a function of the process sulfur load are given as follows,
based on a once-through sorbent utilization of 50 percent:
Process Sulfur Load Once-through C
0.06 1.7
0.03 1.5
0.01 1.2
Tables 26 through 29 give the projected energy costs for the regen-
eration processes and compare them to the once-through operation energy
cost (energy cost of regeneration process minus cost of sorbent for the
once-through process are assumed to be identical with the regenerative
systems, within the accuracy of these cost estimates.
The energy costs of the regeneration processes are considerably
greater than the energy costs of once-through sorbent operation using
the basis applied in this study. For the assumption that the $8/Mg
regenerative processes may be operated with a Ca/S ratio of 0.5, the cost
to which dolomite must rise in order to result in a once-through energy
cost identical with that of regenerative energy is shown in Table 30.
For comparison, a typical cost of sorbent (fresh sorbent plus disposal)
is currently $8/Mg.
If the cost of dolomite (delivered cost plus disposal cost) should
reach levels as high as are shown in Table 30, the regeneration processes
will still have to demonstrate the operability and reliability required
by an electric utility in order to be acceptable.
Economic Comparison with Limestone Wet-Scrubbing
The pressurized fluidized-bed combustion power plant with regenera-
tive sorbent operation must compete economically with commercial power
generation systems such as a conventional coal-fired power plant with
limestone wet-scrubbing of the plant stack gases. The investment cost
of a conventional plant with limestone wet-scrubbing is estimated to be
108
-------
TABLE 26. ENERGY COST FOR PRESSURIZED REDUCTIVE
DECOMPOSITION - 1 PERCENT SO mills/kWh
0.06
Capital Charges 3.57
Operating and Maintenance 1.19
Coal
5.14
Methane 0.45
Dolomite 1.64
Total
11.99
Regeneration Energy Cost
minus Once-through Energy
Cost 9.20
Process Sulfur Load
0.03 0.01
2.50 1.42
0.83 0.47
2.57 0.77
0.23 0.07
0.90 0.35
7.03 3.07
5.68 2.65 -
TABLE 27. ENERGY COST FOR PRESSURIZED REDUCTIVE
DECOMPOSITION - 2 PERCENT S02, mills/kWh
Process Sulfur Load
0.06
0.03
0.01
Capital Charges
Operating and Maintenance
Coal ..
Methane
Dolomite
Total
Regeneration Energy Cost
minus Once-through Energy
Cost
2.74
0.91
2.54
0.45
1.64
8.28
5.49
1.91
0.63
1.27
0.23
0.90
4.94
3.59
1.15
0.39
0.38
0.07
0.35
2.34
1.92
109
-------
TABLE 28. ENERGY COST FOR LOW-PRESSURE REDUCTIVE
DECOMPOSITION - 10 PERCENT S02, mills/kWh
Process Sulfur Load
0.06
0.03
0.01
Capital Charges
Operating and Maintenance
Coal
Methane
Dolomite
Total
Regeneration Energy Cost
minus Once-through Energy
Cost
2.04
0.68
0.47
0.45
1.64
5.28
2.49
1.64
0.55
0.23
0.23
0.90
3.55
2.20
1.21
0.40
0.07
0.07
0.35
2.10
1.68
TABLE 29. ENERGY COST FOR TWO-STEP REGENERATION
PROCESS, mills/kWh
Process Sulfur Load
0.06
Capital Charges 3.25
Operating and Maintenance 1.09
Coal 2.17
Dolomite 1 . 64
0.03
2.29
0.76
1.09
0.90
0.01
1.32
0.44
0.32
0.35
Total
Regeneration Energy Cost
minus Once-through Energy
Cost
8.15
5.36
5.04
3.69
2.43
2.01
110
-------
TABLE 30. COST OF DOLOMITE (PURCHASE PLUS DISPOSAL)
REQUIRED TO GIVE EQUAL ONCE-THROUGH AND
REGENERATIVE COSTS, $/Mga
Regeneration Process 0.06
Reductive
Reductive
2% SO
Reductive
10% S02
Decomposition with 57
Decomposition with 38
Decomposition with 23
Two-Step Regeneration 37
Process Sulfur Load
0.03 0.01
73 118
50 88
34 79
51 92
: Regenerative Ca/S = 0.5
about $570/ktT ' (mid-1977 dollars, 635 MWe capacity) and the energy cost
about 23.7 mills/kWh ($10/Mg for limestone, coal at $0.80/GJ, steam at
3 3
$2.2/10 kg, process water at $0.05/10 £, capital charges at 15%/yr,
process sulfur load of 0.03).
A pressurized fluidized-bed combustion power plant is estimated to
represent an investment of about $424/kW (once-through with no spent
sorbent processing), with an energy cost of about 18.4 mills/kWh, not
including the cost of sorbent. Costs for the conventional power plant
and the PFBC power plant are taken from a previous Westinghouse study,
are on a constant basis, and have been scaled to mid-1977 dollars. The
investment costs and energy costs of regenerative pressurized fluidized-
bed combustion (with elemental sulfur recovery) are compared with a
conventional power plant in Table 31 based on a process sulfur load
of 0.03 (4.0 wt % sulfur coal) and a dolomite cost of $10/Mg. The
111
-------
TABLE 31. COMPARISON OF REGENERATIVE PRESSURIZED
FLUID-BED COMBUSTION WITH CONVENTIONAL
COAL-FIRED POWER GENERATION
Conventional Plant
Once-through PFBC
Regenerative PFBC
Reductive decomposition
with 1% S00
£.
Reductive decomposition
with 2% S02
Reductive decomposition
with 10% SO-
Two-step regeneration
Capital Investment
($/kW)
570
424
526
502
491
518
Energy Cost
(mills/kWh)
23.7
19.8
25.4
23.3
22.0
23.4
regenerative PFBC costs are produced by adding interest during con-
struction, general items, and engineering to the investments in Tables 22
through 25 (about 17% increase) and adding those costs to $424/kW.
Energy costs are obtained by adding the cost in Tables 26 through 29 to
18.4 mills/kWh.
While these are always uncertainties associated with cost estimate
comparisons, these estimates are on a sufficiently consistant basis to
conclude that the only regenerative PFBC power generation system that
compares favorably with the conventional power plant with limestone
wet-scrubbing is the system based on low-pressure reductive decomposition.
The pressurized reductive decomposition with 2 percent SO,, and the
two-step regenerative PFBC power generation systems are comparable to
the conventional power plant. The pressurized reductive decomposition
with 1 percent SO results in a PFBC power plant that is not competitive
with the conventional power plant.
112
-------
Environmental Comparison
The environmental performance of the regeneration processes for PFBC
is compared with that of once-through PFBC and conventional coal-fired
power plants with limestone wet-scrubbing in Table 32. All of the
power generation systems are assumed to satisfy the current EPA
new source emissions standards (SO-, NO , particulates) for large coal-
<£. j\
fired plants.
The low-pressure reductive decomposition process is the most environ-
mentally satisfactory of the regeneration processes. The once-through
PFBC operation is environmentally superior to the regeneration processes
in all aspects except spent sorbent production. The environmental impact
of the regenerative spent sorbent versus the once-through spent sorbent
due to differences in chemical nature is not known. The conventional
power plant with limestone wet-scrubbing requires coal consumption at a
greater rate than all of the PFBC power plants except for the pressurized
reductive decomposition with 1 vol % S02. The limestone wet-scrubbing
produces a spent sludge material requiring large land usage for a pond.
ASSESSMENT
An integrated PFBC regeneration system has yet to be demonstrated.
Most performance data have been generated on small-scale, batch, and
semicontinuous apparatus, and reliable information concerning the critical
performance factors for commercial operation is not available.
The technical performance of the three PFBC sorbent regeneration
schemes evaluated is uncertain. The pressurized reductive decomposition
will result in such low SO concentrations (1 to 2 vol %) that huge
amounts of coal for reductant will be required, and complex energy recovery
and sulfur recovery systems will be necessary. The low-pressure
reductive decomposition appears technically favorable except for major
uncertainties in the solids transport'system. Sulfur recovery will be
nearer conventional practice with the two-step regeneration since HS
113
-------
TABLE 32. COMPARISON OF ENVIRONMENTAL IMPACTS FOR PFBC AND
CONVENTIONAL POWER PLANTS(a>
PFBC with
Pressurized Reductive
Decomposition
17, S02
2% S02
PFBC with
Low-Pressure
Decomposition,
10% S02
PFBC with
Two-Step
Regeneration,
3% H2S
PFBC
Once-through
Operation
Conventional
Power Plant ^
Plant Heat Rate,
Raw Materials
(c)
KJ/kWh
12,400
10,800
9,600
10,400
9,040
VLl.OOO
Coal input, v ' Mg/hr 263
Sorbent input/e^ MG/day 588-1175
Methane input/f) 106 kJ/hr 136
Plant Exports
Spent sorbent/g) Mg/day 435-870
Ash(h) , Mg/day 631
Sulfur/1^ Mg/day 141
228
588-1175
136
435-870
547
141
201
588-1175
130
435-870
482
141
224
1175-2350
0
870-1740
538
141
195
1,763
0
1,900
468
0
237
840
0
1,850
569
0
(a) Basis: 635 MWe power plant capacity, 4 wt 7, sulfur coal, current emission standards for SO , NO , and particulates
satisfied - 516 ng/J, 43 ng/J, and 30 ng/J, respectively. x
(b) New plant with limestone wet-scrubbing.
(c) Includes auxiliary coal and methane input.
(d) Includes coal for regeneration reductant.
(e) Ca/S (dolomite) of 0.5-1.0 for reductive decomposition, 1.0-2.0 for two-step regeneration, 1.5 for once-through
PFBC, and 1.2 (limestone) for limestone wet-scrubber.
(f) Methane used in sulfur recovery system only.
(g) Dry, granular for PFBC, limestone sludge for wet-scrubber.
(h) 10 wt % ash in coal.
(i) Sulfur in auxiliary coal is neglected.
-------
rather than SO,, is generated, but large amounts of power, coal, and
steam are consumed, again requiring complex energy recovery. The ability
of the two-step regeneration process to produce an active sorbent
material is also questioned. All of the regeneration processes are
complex and their operability and reliability are major concerns.
The overall environmental performance of the low-pressure reductive
decomposition is superior to the other regeneration processes. Both the
pressurized and the low-pressure reductive decomposition processes
require the consumption of clean fuels such as methane in the sulfur
recovery system. The once-through sorbent operation is superior to the
regenerative operations in all environmental aspects except the quantity
of spent sorbent produced.
The only regenerative PFBC power plant that is economically attrac-
tive when compared to a conventional power plant with limestone wet-
scrubbing is based on low-pressure reductive decomposition. The once-
through PFBC power plant has a considerably lower energy cost, based on
a dolomite cost of $10/Mg, than any of the regenerative power plants.
A regenerative system modeling study is being initiated to assess
the regeneration technology and process economics in greater detail. The
computer study will permit data being generated by facilities such as the
Exxon miniplant to be evaluated quickly and comprehensively in order to
aid the experimental program direction.
115
-------
SECTION 7
NOMENCLATURE
m = Ca/S mole ratio, moles of calcium in the fresh sorbent per mole of
sulfur to the boiler
Mg = megagrams (10 grams)
n = sulfation cycle number
N = number of regenerator modules
PSL = process sulfur load, W (n-mXZ)
S o
R = rate of sorbent circulation in moles of calcium per unit mass of coal
fed to the combustor
U = regenerator superficial fluidization velocity
V = molar volume of regenerator off-gas
n
V = particle volume per mole of calcium
S
W = lb of sulfur per Ib of coal
5
X = extent of CaO regeneration
ID
X = mole fraction of calcium as sulfate in the sulfated sorbent
O
XV = mole fraction of calcium as sulfate in the regenerated sorbent
J
j = mole fraction of S0_ in the regenerator off-gas
h = height of the static bed
S
h = height of the expanded bed
a = fraction of calcium in the boiler rejected after each cycle
e = regenerator bed expanded void fraction
n = boiler sulfur removal efficiency
T = residence time of gas in the regenerator
8
T = residence time of solids in the regenerator
116
-------
Abbreviations
AFEB atmospheric-pressure fluidized-bed boiler
AFBC atmospheric-pressure fluidized-bed combustion
ANL Argonne National Laboratory
FWEC Foster Wheeler Energy Corporation
O&M operating and maintenance
PER Pope, Evans and Robbins
PFBB Pressurized fluidized-bed boiler
PFBC Pressurized fluidized-bed combustion
RESOX Registered trademark of the sulfur recovery process under development
by FWEC
SRP sulfur recovery plant
117
-------
SECTION 8
REFERENCES
1. Keairns, D. L., et al., Evaluation of the Fluidized-Bed Combustion
Process Vol. I, Office of Research and Development, EPA, Westing-
house Research Laboratories, Pittsburgh, Pa. 15235, December 1973,
EPA 650/2-73-048a, NTIS PB 231-162.
2. Proceedings of the Third International Conference on Fluidized Bed
Combustion, Session on Sorbent Regeneration, October, 1972,
EPA 650/2-73-053, NTIS PB 231-977.
3. Proceedings of the Fourth International Conference on Fluidized Bed
Combustion, Session on Sorbent Regenerations, December, 1975, pub-
lished by MITRE Corp., May, 1976.
4. Encyclopedia of Chemical Technology, Second Edition, Kirk-Othmer,
New York: Interscience Publishers, 1963.
5. A Development Program on Pressurized Fluidized Bed Combustion,
Annual Report, Office of Fossil Energy, ERDA, EPA, Argonne National
Laboratory, Argonne, Illinois, 1976, ANL/ES-CEN-1016.
6. Craig, J. W. T., et al., Chemically Active Fluid-Bed Process for
Sulfur Removal During Gasification of Heavy Fuel Oil-Second Phase,
EPA, Esso Research Centre, Abingdon, England, Nov. 1974,
EPA 650/2-75-109, NTIS PB 240-632.
7. Studies of the Pressurized Fluidized-Bed Coal Combustion Process,
Office of Research and Development, EPA, Exxon Research and Engi-
neering Co., Linden, N.J., September 1977, EPA 600/7-77-107, NTIS
PB 272-722.
118
-------
8. Proceedings of a Workshop on Regeneration of Sulfated Limestone/
Dolomite for Fluidized Bed Combustion, Foster Wheeler presentation,
ERDA, Washington, D.C., March 1975, Prepared by Gilbert Associates,
Reading, Pa.
9. Hamtnons, G. A., and A. Skopp, A Regenerative Limestone Process for
Fluidized Bed Coal Combustion and Desulfurization, Exxon Research
and Engineering Co., Linden, N.J. Final Report for Contract
No. EPA 70-19 (Feb., 1971).
10. Swift, W. M., and T. D. Wheelock, "Decomposition of Calcium Sulfate
in a Two-Zone Reactor," Ind. Eng. Chem. Process Des. Dev., 14(3),
323 (1975).
11. Shang, J. Y., and R. A. Chronowski, Comparison of AFBC with PFBC,
Proceedings of the Fourth Intern. Conf. on Fluidized Bed Combustion,
December, 1975, Published by Mitre Corporation.
12. Keairns, D. L., et al., Evaluation of the Fluidized-Bed Combustion
Process, Vol. II - Fluidized-Bed Boiler Combined-Cycle Power Plant
Development, Vol. I Appendices, Report to EPA, Westinghouse
Research and Development Center, December 1973, EPA 650/2-73-048b,
NTIS PB 231-163.
13. "Characterization of Glaus Plant Emissions," Processes Research,
Inc., 1973, EPA-R2-73-188.
14. Beavon, D. K., and R. N. Fleck, "Beavon Sulfur Removal Process for
Glaus Plant Tail Gas," in Sulfur Removal and Recovery from Industrial
Processes, J. B. Pfeiffer, ed., American Chemical Society, Washing-
ton, DC, 1975, p. 93.
15. "Glaus Technical and Economic Study," Shell Development Co., 1972.
GAP Contract EHSD-71-45.
16. Goar, B. G., "Feed Gas Impurities Can Create Glaus Plant Problems,"
Gas Conditioning Proceedings, p. C1-C18, 1974.
119
-------
17. Richardson, F. P., "Applications of the Stretford Process to Large
Coal Gasification Plants," Gas Conditioning Proceedings, p. 11-17,
1974.
18. "Characterization of Sulfur Recovery from Refinery Fuel Gas,"
Battelle Columbus Laboratories, EPA, June 1974, NTIS PB 239-777.
19. Consolidation Coal Co., Inc., "Production of Clean Fuel Gas from
Bituminous Coal," EPA 650/2-73-049, December 1973.
20. Semrau, K., "Controlling the Industrial Process Sources of Sulfur
Oxides," in Sulfur Removal and Recovery from Industrial Processes,
J. B. Pfeiffer, ed., Washington, DC: Americal Chemical Society,
1975, p. 1.
21. "Applicability of Reduction to Sulfur Techniques to the Development
of New Processes for Removing S0? from Flue Gases," Final Report,
Volume I, Allied Chemical Corporation, 1969, PH-22-68-24, NTIS
PB 198-407.
22. "Glaus Technical and Economic Study," Shell Development Co., 1972,
GAP Contract EHSD-71-45.
23. Henderson, J. M., and J. B. Pfeiffer, in Sulfur Removal and Recovery
From Industrial Processes, J. B. Pfeiffer, ed., Washington DC:
American Chemical Society, 1975, p. 35.
24. "Sulfur Dioxide: Its Chemistry as Related to Methods for Removing
it from Waste Gases," U.S. Bureau of Mines, June 1973, NTIS
PB 221-899.
25. "Evaluation of the Regenerative Pressurized Fluidized Bed Combustion
Process," EPA, M. W. Kellogg Co., 1974, EPA 650/2-74-012, NTIS
PB 232 012.
26. Hunter, W. D., Jr., et al., "The Allied Chemical Sulfur Dioxide
Reduction Process for Metallurgical Emissions," in Sulfur Removal
and Recovery from Industrial Processes, Washington, DC, 1975,
p. 23.
120
-------
27. Connor, J. M., "Economics of Sulfuric Acid Manufacture, in Sulfur
and S02 Developments," CEP Technical Manual, AIChE, 13_ (1971).
28. "Engineering Analysis of Emissions Control Technology for Sulfuric
Acid Manufacturing Processes," Chemical Construction Corp., 1970,
NTIS PB 190-393.
29. Katell, S., J. H. Faber, "An Economic Studv of the Hot Carbonate
Process for Removing Carbon Dioxide," U.S. Bureau of Mines Informa-
tion Circular 7952, 1960.
30. Stzelzoff, S., "Choosing the Optimum CO -removal System," Chemical
Engineering, 115, September, 1975.
31. Perry, R. H., and C. H. Chilton, eds., Chemical Engineering Handbook,
5th Edition, New York: McGraw Hill Book Co., 1973.
32. Sittig, M., "Catalytic Cracking Techniques in Review," Petroleum
Refiner, 31., 263 (1952).
33. Carroll, P. J., H. Colijn, "Vibrations in Solids Flow," Chemical
Engineering Progress, _71_ (21, 53 (1975)).
34. Discussions with Eriez Magnetics, Erie, Pa. at Westinghouse R&D
Center, Pgh., PA, on June 23, 1976.
35. Yang, W. C., "A Unified Theory on Dilute Phase Pneumatic Transport,"
Presented at Int. Powder and Bulk Solids Handling and Processing
Conference, Chicago, IL, May 11-13, 1976.
36. Stoess, H. A., Jr., Pneumatic Conveying, New York: Wiley-
Interscience, 1970.
37. Leung, L. S., L. A. Wilson, Downflow of Solids in Standpipes, Powder
Technology, _7, 343 (1973).
38. Matsen, J. M., "Flow of Fluidized Solids and Bubbles in Standpipes
and Risers," Powder Technology,'_7, 93 (1973).
121
-------
39. Matsen, J. M., "Some Characteristics of Large Solids Circulation
Systems," Fluidization Technology, Vol. II, D. L. Keairns, Ed.,
New York: McGraw-Hill Book Co., 1976, p. 135.
40. Kunii, D., et al., "Flow Characteristics of Circulation Systems
with Two Fluidized Beds," Int. Chem. Eng., JL4(3), 588 (1974).
41. Capes, C. E., "Dense Phase Vertical Pneumatic Conveying," Canadian
Journal of Chem. Eng., 49_, 182 (1971).
42. Capes, C. E., et al., "Method and Apparatus for Conveying Par-
ticulate Material Upwardly in a Gas Stream," U.S. Pat. 3,776,601,
December 1973.
43. Wen, C. Y., and H. P. Simons, "Flow Characteristics in Horizontal
Fluidized Solids Transport," AlChe, Journal, _5(2)» 263 (1959).
44. Fricke, H. D., et al., "Flow Studies of Dense-Phase Gas/Solid Sus-
pensions with a Positive Expulsion Fluidization Tank," presented
2nd Symposium on Storage and Flow of Solids (ASME), Chicago, 111.,
September 1972.
45. Chari, S. S., "Pressure Drop in Horizontal Dense Phase Conveying
of Air-Solid Mixtures," AIChE 63rd Annual Meeting, Chicago, 111.,
November 1970.
46. Petersen, D. H., "Plug Conveying - An Economic, Pneumatic Transport
System," Aufberitungs-Technik, Nr. 1, 35 (1973).
47. Flain, R. J., "Pneumatic Conveying: How the System is Matched to
the Materials," Process Engineering, 88 (Nov. 1972).
48. .Keairns, D. L., et al., "Fluidized Bed Combustion Process Evalua-
tion," Vol. II, Westinghouse Research Laboratories, Pittsburgh,
Pa., 15235, EPA, March 1975, EPA 650/2-75-027-b, NTIS PB 241- 835.
49. Zenz, F. A., and P. N. Rowe, "Pa-rticle Conveying in Extrusion Flow,"
Fluidization Technology, Vol. II, D. L. Keairns, Ed., New York:
McGraw-Hill Book Co., 1976, p. 151.
122
-------
50. Berg, C. H. 0., "Conveyor of Granular Solids," U.S. Patent 2,684,868,
1954.
51. Berg, C. H. 0., "Conveyance of Granular Solids," U.S. Patent 2,684,872,
1954.
52. Berg, C. H. 0., "Conveyance of Granular Solids," U.S. Patent
2,684,870, 1954.
53. Muskett, W. J., et al., "The Fluidized Transport of Powdered Mate-
rials in an Air-Gravity Conveyor," Pneumo Transport 2, 2nd Inter-
national Conference on Pneumatic Transport of Solids in Pipes,
September 1973.
54. Siemes, W., and L. Hellmer, "Die Messung der Wirbelschichtviskositat
mit der Pneumatischen Rinne," Chemical Engineering Science, JL7_, 555
(1962).
55. Verteshev, M. S., et al., "Horizontally Moving Fluidized Beds,"
Int. Chem. Eng., 9/3), 505 (1969).
56. Shinohara, K., and T. Tanaka, "A New Device for Pneumatic Transport
of Particles," J. Chem. Eng. Japan, _5(3) , 2?9 (1972).
57. Archer, D. H., et al., "Evaluation of the Fluidized Bed Combustion
Process," Vol. II, Report to EPA, Westinghouse Research and Develop-
ment Center, Pittsburgh, PA, November 1971, Contract 70-9, NTIS
PB 212-916.
58. Steiner, P., et al., "Removal and Reduction of Sulfur Dioxides from
Polluted Gas Streams," Advances in Chemistry Series, 139, 180
(1975).
59. Proceedings of a workshop on Regeneration of Sulfated Limestone/
Dolomite for Fluidized Bed Combustion, Pope, Evans and Robbins,
R&D Presentation, ERDA, Washington, DC, March 1975, prepared by
Gilbert Associates, Reading, PA.
123
-------
60. Criteria for the Selection of S0_ Sorbents for Atmospheric Fluid
Bed Combustors, Task 1 Report, Electric Power Research Institute,
Inc., Westinghouse Research Laboratories, Pittsburgh, Pa., 15235
September 1976, Contract No. RP 721-1.
61. Jonke, A. A., et al., Reduction of Atmospheric Pollution by the
Application of Fluidized-Bed Combustion, Annual Reports Argonne
National Laboratory, July, 1968-July, 1969 and July, 1969-June
1970, NTIS ANL/ES-CEN-1001 and 1002.
62. Keairns, D. L., et al., "Fluidized Bed Combustion Report to EPA,
Phase II-Pressurized Fluidized Bed Coal Combustion Development,
Westinghouse Research Laboratories, Pittsburgh, PA, September,
1975, EPA- 650/2-75-027c, NTIS PB 246-116.
63. A Development Program on Pressurized, Fluidized-Bed Coal Combus-
tion, Argonne National Laboratory, Argonne, Illinois, July 1975-
June 1976, NTIS ANL/ES-CEN-1016.
64. A Development Program on Pressurized Fluidized Bed Combustion,
Argonne National Laboratory, July 1976-June 1977, NTIS
ANL/CEN/FE-77-3.
65. Archer, D. H., et al., "Evaluation of the Fluidized Bed Combus-
tion Process," Vol. Ill, Report to EPA, Westinghouse Research and
Development Center, Pittsburgh, Pa., November 1971, Contract
70-9, NTIS PB 213-152.
66. Keairns, D. L., et al., "Evaluation of the Fluidized-Bed Combus-
tion Process, Pressurized Fluidized-Bed Combustion Process Develop-
ment and Evaluation" - Vol. I; "Pressurized Fluidized-Bed Boiler
Development Plant Design," Vol. Ill, Report to EPA, Westinghouse
Research and Development Center, Pittsburgh, Pa., December 1973,
EPA-650/2-73-048a and c, NTIS PB 231-162, 232-433.
124
-------
APPENDIX A
EFFECT OF VARIOUS FACTORS ON REGENERATOR SO CONCENTRATION
A brief study has been made to estimate the maximum concentration
of SO^ that can be obtained in the regenerator effluent gas and to iden-
tify the factors that limit that concentration. The effect of heat
losses, air preheat temperature, and type of fuel on the SO level was
studied from the point of view of material and energy balance. Design
and operating conditions selected for the study are the same as those
given earlier. The following assumptions are made:
No CaS is formed.
Two cases were considered. In the first case, no CO or H was
assumed to be present in the exit gas. In the second case, the
water-gas shift reaction was assumed to be at equilibrium.
Sulfur in the fuel to the regenerator is neglected.
In-situ partial oxidation of coal or CH, for the reductant.
RESULTS
The equilibrium concentration of SO-, as experimentally determined
by Curran et al., is shown in Figure 25 for atmospheric pressure.
For any other pressure the equilibrium concentration of SO is equal to
this value divided by the pressure in atmospheres. The equilibrium con-
centration obtained from standard thermodynamic data is also shown. The
substantial difference between the two curves may be due to the formation
of a solid solution from the two reactants.
Figure Al shows the SO- concentration as a function of heat losses
when coal is the fuel to the regenerator. The substantial increase in
SO level with the increase in air preheat temperature to 1038°C can
125
-------
Curvs 5846;9-;
CNJ
o
c
o
o
E
13
Standard
Thermodynamic Data
4 =r
1040
Temperature, °C
Figure Al. Equilibrium for 1/4 CaS + 3/4 CaS04 = CaO + SOo
at 101 kPa
126
-------
Cu-ve 634611-A
14 -N
CNJ
O
- 8
- 4
10 20 30
Heat Losses, % of Fuel Input
Figure A2. Effect of Heat Losses and Air Preheat Temperature
on Concentration of SC>2 - Coal as Fuel
-------
be noted. Heat input as a percentage of heat input to the fluid-bed
boiler is also shown. No CO or H was assumed to be present in the exit
gas. Figure A3 shows the same results when methane is the fuel to the
regenerator. The concentration of SO- is slightly higher in the case
of coal. The above results would be valid even when the fuel is burned
in a separate vessel to produce the reducing gas except for the additional
heat losses since the results are based on material and energy balances0
The effect of the change in sorbent utilization across the regenera-
tor on SO,, level, sorbent circulation rate, and the fuel input is shown
in Figure A4.
Figure A5 shows the concentration of S09 as a function of fuel input
in the case of coal, where water-gas shift reaction is at equilibrium.
Two curves are shown for air at room temperature and a preheat tempera-
ture of 1038°C when heat losses are negligible. A minimum concentration
of SO is indicated only when air is preheated; below this concentration,
isothermal conditions in the regenerator cannot be maintained. The mini-
mum fuel input corresponds to the case where no CO or H is present in
the exit gas. Any excess fuel reduces the concentration of S0? and
results in the formation of CO and EL. The effect of heat losses is
also shown in curves 1 to 4 of Figure A5. The maximum S0? concentration
in each case is shown by a vertical dotted line, and the percentage heat
loss at this point is also indicated.
CONCLUSIONS AND RECOMMENDATIONS
Based on material and energy balances, it is possible to achieve
SO. levels greater than 10 percent at atmospheric pressure by
preheating the air. At a temperature of 1100°C, the equilibrium
concentration of SO- as obtained by Curran et al. * appears
to be about 24 percent. Hence, material and energy balances
would be the limiting factor-at atmospheric pressure.
128
-------
Curve 684613-A
12
10 -
OJ
-------
Curve 68k6]k-B
2.5
to
o>
1.5
o>
-t
to
UJ
o
-2 1
-- *
O
£0.5
11 -
- 5
10 -
9 -
8 _
7 -
- 6
Utilization of Ca after
Regeneration = 0.1
0.2 0.3 0.4 0.5 0.6
Change in Utilization of Ca across Regenerator
Figure A4. .Effect of Change in Sorbent Utilization Across
Regenerator on Concentration of S02 Sorbent
Circulation and Fuel Input
-------
Curve 684612-1
o
CO
CD
Z5
50 -
40
30
20
10
I
Fuel -Coal
Air at 25°C
Air at 1038°C
Heat Losses
GJ
(1)
21
42
64
382
8.2
14.5
18.3
28
Assumption: Water-Gas Shift Reaction
at Equilibrium
(1) At max. SCL Concentration
0
6 8
Concentration of SO
10
12
14 15
2'
Figure A5. Effect of Heat Losses and Fuel Input on
Concentration of SCL
-------
Since the equilibrium concentration of S09 is inversely propor-
tional to the total pressure, equilibrium would limit the SO
concentration for pressure higher about 200 kPa.
Heat losses have a large impact on the SO level. The low con-
centrations of SO,- obtained in the work then possible, carried
out at Exxon and Argonne National Laboratories, indicate that this
may be because of the heat losses that can be expected to be higher
in a small unit. Efforts should be directed toward preheating
the air and insulating the regenerator to reduce the heat losses.
The change in utilization of the sorbent across the regenerator
also has a significant effect on the S0_ concentration, particu-
larly at lower levels of utilization. Considering the fuel input
to the regenerator and the sorbent circulation rate in addition
to the S0« level, the utilization should be greater than about
0.2.
REFERENCES
1. Curran, G. P., Fink, C. E. and Gorin, E., "C02 Acceptor Gasification
Process," Advances in Chemistry Series 69, 141-165 (1967).
132
-------
TECHNICAL REPORT DATA
(Please rcad Instructions on i/ic rcrcrsi before completing)
. REPORT NO.
EPA-600/7-78-039
2.
3. RECIPIENT'S ACCESSION-NO.
TiTLEANDSUBT1TLE Regeneration o.f Calcium-Based SC2
Sorbents for Fluidized-bed Combustion: Engineering
Evaluation
5. REPORT DATE
March 1978
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
8. PERFORMING ORGANIZATION REPORT NO.
R.A.Newby, S.Katta, and D. L. Keairns
. PERFORMING ORGANIZATION NAME AND ADDRESS
Westinghouse Research and Development Center
1310 Beulah Road
Pittsburgh, Pennsylvania 15235
10. PROGRAM ELEMENT NO.
E HE 82 3 A
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 P!
Annual; 12/75-1/77
PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/13
15.SUPPLEMENTARY NOTES]ERIj_RTp project officer fc D. Bruce Henschel, Mail Drop 61,
919/541-2825. Earlier report on this work is EPA-850/2-75-027c.
16. ABSTRACT
The report gives results of an engineering evaluation of regeneration of
calcium-based SO2 sorbents (limestone and dolomite) for application in both atmos-
pheric and pressurized fluidized-bed combustion (FBC) processes. Economics of
FBC power plants, operated with regeneration, are projected based on current esti-
mates o.f regeneration process performance. Coal-fueled reductive decomposition is
the regeneration process considered for atmospheric FBC; three regeneration
schemes (two reductive decomposition processes and a two-step process) are evalu-
ated for pressurized FBC. Estimated costs of FBC power plants with regeneration
are compared with costs of FBC plants using once-through sorbent (no regeneration).
The economic feasibility of the regenerative system depends on several variables ,
including in particular the sulfur concentration achievable in the regenerator off-gas,
the reduction in fresh sorbent feed rate possible through regeneration, and the cost
o.f fresh sorbent and of solid residue disposal. The performance required for the
regenerative FBC system to achieve economic feasibility is projected, and critical
development needs are discussed. An integrated regeneration system for both atmos-
pheric and pressurized FBC, capable of achieving the performance necessary, has
yet to be demonstrated experimentally.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Calcium
Regeneration
(Engineer in
Sulfur
b.lDENTIFIERS/OPEN ENDED TERMS
COSATI Field/Group
Air Pollution
Coal
Combustion
Electric Power Plants
g)
Air Pollution Control
Stationary Sources
Sulfur Recovery
Reductive Decomposition
Fluidized Bed Processing
Sulfur Dioxide Limestone
Sorbents Dolomite
13B
2 ID
2 IB
10B
07A
07B
11G
08G
13. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
145
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
133
------- |