CD A U.S. Environmental Protection Agency Industrial Environmental Research
•»• •» Office of Research and Development Laboratory
Research Triangle4Park. North Carolina 27711
EPA-600/7-78'005
-*t\-ro
JSRUSr 1978
ALTERNATIVES TO
CALCIUM-BASED SO2 SORBENTS
FOR FLUIDIZED-BED COMBUSTION
CONCEPTUAL 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 seven series.
These seven broad categories were established to facilitate further
development and application of environmental technology. Elimination
of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields. The seven 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
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-agehcy Federal Energy/Environment
Research and Development Program. These studies relate to EPA's
/iission to protect the public health and welfare from adverse effects
/of pollutants associated with energy systems. The goal of the Program
is to assure the rapid development of domestic energy supplies in an
environmentally—compatible manner by providing the necessary
environmental data and control technology. Investigations include
analyses 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 environmental issues.
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 recommen-
dation for use.
This document is available to the public through the National Technical
Information Service, Springfield, Virginia 22161.
-------
EPA-600/7-78-005
January 1978
ALTERNATIVES
TO CALCIUM-BASED
SO2 SORBENTS
FOR FLUIDIZED-BED COMBUSTION:
CONCEPTUAL EVALUATION
by
Richard A. Newby and Dale 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, particulate emissions and
control, trace element emissions and control, spent sorbent and ash
disposal and systems evaluation (e.g. impact of new source performance
standards on fluidized-bed combustion system design and cost.
This document contains the results of work, defined and completed
under the alternative sorbent task, from December 1975 to December 1976.
Results from work carried out by Westinghouse or reported by other
Experimental work at Exxon (L.A. Ruth, "Regenerable Sorbents for
Fluidized-Bed Combustion," Proceedings of the Fluidized-Bed Combustion
Technology, Exchange Workshop, Vol. II, 301, NTIS No. CONF-770447-P-2,
April 1977), and Argonne National Laboratory (R.B. Snyder et al, Synthetic
Sorbents for Removal of Sulfur Dioxide in Fluidized-Bed Combustors",
June 1977, NTIS No. ANL/CEN/FE-77-1) has been reported since this
evaluation. Their results complement this report by extending the
experimental base for the evaluation.
iii
-------
investigators after February 1977 are not integrated 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
• 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 combinedrvcycle power systems, adiabatic fluidized-bed
combustion combinedrcycle power systems—including 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 para-
meters selected for the base power 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
IV
-------
• 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.
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 reports are available that document the
prior contract work (see references 16-18).
-------
ABSTRACT
A conceptual engineering evaluation has been performed to screen
supported metal oxides as alternatives to natural calcium-based sorbents
(limestones and dolomites) for S0£ control in fluidized-bed combustion
processes. Ranges of applicable operating conditions for atmospheric-
pressure fluidized-bed combustion (AFBC) and for pressurized fluidized-
bed combustion (PFBC) have been defined. Acceptance criteria for
thermodynamic performance, melting point, and material-and-energy balance
limitations of metal oxides in the combustor and in the sorbent regene-
ration process have been proposed and applied in the evaluation.
Three regenerative schemes for the alternative sorbents have
been considered,and ranges of operating conditions resulting in accept-
able performance for the sorbents have been identified. Fourteen
potentially acceptable sorbents have been identified for AFBC and eleven
for PFBC.
Estimates of regeneration system economics, sorbent cost
(supported on alumina),and sorbent availability have been developed in
order to project maximum acceptable loss rates for the sorbents.
Sorbent loss rates, due to attrition or deactivation, must
be small relative to calcium-based sorbents (less than 0.1 percent of bed
content per hour) in order to compete economically with natural Qalcium-
based sorbents, even if maximum thermodynamic performance is realized.
Sorbent availability will be of extreme importance for many metal oxide
materials. Recommendations are made for further engineering assessment
and experimental sorbent screening.
vii
-------
CONTENTS
1. INTRODUCTION 1
2. CONCLUSIONS 3
Alternative Sorbents 3
Intimate Coal/Sorbent Mixtures 4
3. RECOMMENDATIONS 5
Alternative Sorbents 6
Intimate Coal/Sorbent Mixtures 7
A. SCOPE AND METHOD OF ASSESSMENT 8
System Concepts 9
Operating Conditions and Acceptance Criteria 11
Sorbent Materials Considered 15
5. PROJECTIONS OF DESULFURIZATION PERFORMANCE 18
Equilibrium Desulfurization Performance 18
Sorbent Melting Data 19
Metal Carbonate Stability 25
Sulfite Stability 31
Summary of Desulfurization Potential 31
6. PROJECTIONS OF SORBENT REGENERATION POTENTIAL 34
Thermal Decomposition 34
Reductive Decomposition 34
Two-Step Regeneration Schemes 35
Summary of Regeneration Potential 43
7. MATERIAL AND ENERGY BALANCE LIMITATIONS 48
Reductive Decomposition 49
Two-Step Regeneration Schemes 50
Summary 51
8. AVAILABILITY 56
Sorbent Demand Model 64
9. ECONOMICS 68
Sorbent Material Cost 68
Maximum Acceptable Total Sorbent Cost 70
Maximum Acceptable Sorbent Loss Rate 76
10. ASSESSMENT OF ALTERNATIVE SORBENTS 81
11. REFERENCES 83
ix
-------
CONTENTS (Continued)
APPENDICES
IIIA. PROGRAM PLAN 85
Selection of Support Material 85
Selection of Sorbent Materials 86
Sorbent Testing 86
Attrition Tests 86
Desulfurization Tests 86
Regeneration Tests 88
IIIB. ENGINEERING ASSESSMENT OF INTIMATE COAL/SORBENT 92
MIXTURES FOR S02 CONTROL IN FLUIDIZED-BED
COMBUSTION APPLICATIONS
Purpose and Scope of Study 92
Basis for the Study 92
Initial Screening of Concepts 93
Alternative Intimate Coal/Sorbent 93
Mixtures
Alternative Coal/Sorbent Consolidated 96
Particle Characteristics
Alternative Coal/Sorbent Consolidated 97
Particle Preparation System
Projections of Environmental and Technical 97
Performance
Economic Projection 102
Maximum Allowable Costs 103
Consolidated Particle Preparation System 112
Cost Estimate
Assessment 125
References 126
x
-------
FIGURES
Number
1
2
3
4
5
6
7
Bl
B2
B3
B4
B5
B6
B7
Equilibrium Constant for MeSO, ^r? MeO + SO-
Equilibrium Constant for MeSO, iZ? MeO + SO,
Metal Carbonate Stability
One-Step Reductive Decomposition Equilibrium S0_
Projection - No Sorbent Carbonation 3 MeSO, +
MeS «""*" 4 MeO + 4 S09
One-Step Reductive Decomposition Equilibrium
Projection - Sorbent Carbonation (15% C02)
3 MeSO^ + MeS + 4 C02 5± 4 MeCO., + 4 S02
MeS + ft O^Z. MeO + H2S Reaction Equilibrium
MeS + H20 + C02 i± MeC03 + H2S Reaction Equilibrium
Coal/Sorbent Consolidated Particle Preparation
System
Combustor Sulfur Removal Efficiency Required to
Achieve U.S. EPA Emission Standard
Maximum Difference in Sorbent Feed Rates
Maximum Sorbent Cost Savings
l
Maximum Break-Even Capital Investment
Break-Even Power Requirement
Coal/Sorbent Consolidated Particle Preparation
Page
21
22
26
37
38
42
44
100
109
110
111
113
114
116
System
B8 Coal Pulverization Power Requirement 119
B9 Sorbent Pulverization Power Requirement 120
BIO Cost of Size Reduction System for Coal and Sorbent 121
xi
-------
TABLES
Number Page
1 Fluidized-Bed Combustor Conditions and Criteria 12
2 Regeneration Conditions and Criteria 13
3 Initial List of Alternative Sorbents Evaluated 16
(Metal Oxides)
4 Metal Oxides Not Considered Because of Previous 17
Judgments
5 Limiting Values for the Metal Sulfate Decomposition 20
for Sorbent Acceptance
6 Metal Oxides Acceptable Based on Thermodynamically 23
Predicted Desulfurizer SOX Concentration
7 Mel ting-Point Data 24
8 Metal Carbonate Stability 27
9 Metal Carbonate Desulfurization Potential 30
10 Summary of Acceptable Sorbents Based on 32
Desulfurization Performance
11 Assessment of Acceptable Sorbents Based on 33
Desulfurization Performance
12 Regeneration by Reductive Decomposition 36
13 Metal Sulf ide Formation 3 MeS04 + MeS i± 4 MeO + 40
4 S02
14 H2S Generation from Metal Sulfide by 41
MeS + H0;^
15 H2S Generation from Metal Sulfide by 45
MeS + H20 + C02 ^±L Me C03 + H2S
Xll
-------
TABLES (Continued)
Number page
16 Thermodynamic Screening of AFBC Sorbents 46
17 Thermodynamic Screening of PFBC Sorbents 47
18 Results of Material and Energy Balance Considerations 52
for Regeneration by Reductive Decomposition
19 Results of Material and Energy Balance Considerations 53
for Regeneration by the Two-Step Process - AFBC
20 Results of Material and Energy Balance Considerations 54
for Regeneration by the Two-Step Process - PFBC
21 BaCO Statistics 57
22 SrCO- Statistics 58
23 Ti02 Statistics 59
24 Fe2°3 statistics 59
25 A1203 statistics 60
26 Li2C03 statistics 61
27 Na2C03 statistics 62
28 CaC03 Statistics 63
29 Sorbent Availability Comparison 66
30 Sorbent Raw Materials Costs 69
31 Maximum Acceptable Alternative Sorbent Cost for 73
Reductive Decomposition Regeneration (S)
32 Maximum Acceptable Alternative Sorbent Cost for 74
Two-Step Atmospheric-Pressure Regeneration (S)
33 Maximum Acceptable Alternative Sorbent Cost for 75
Two-Step Pressurized Regeneration (S)
34 Maximum Sorbent Loss Rates for Reductive 77
Decomposition Regeneration
xiii
-------
TABLES (Continued)
Number Page
35 Maximum Sorbent Loss Rates for Two-Step ^8
Regeneration - AFBC
36 Maximum Sorbent Loss Rates for Two-Step ^9
Regeneration - PFBC
Al Alternative Sorbent Selection
87
A2 BaC03 Test Conditions 89
A3 CaAl204 Test Conditions 90
A4 BaTiC- Test Conditions 91
Bl Projections of the Calcium-to-Sulfur Feed Ratios in 9^
Fluidized-Bed Boilers with Conventional Once-through
Sorbents
B2 Coal/Sorbent Consolidated Particle Characteristics 98
B3 Consolidated Particle Preparation System Options
B4 Speculative Performance of Coal/Sorbent
Consolidated Particles
B5 Description of Selected Preparation Equipment
B6 Cost Estimate for Consolidated Particle Preparation
B7 Comparison between Break-Even Factors and Estimates
xiv
-------
ACKNOWLEDGMENT
We want to express our high regard for and acknowledge the
contribution of Mr. D. B. Henschel who served as the EPA project officer.
Mr. P. P. Turner and Mr. R. P. Hangebrauch, Industrial Environmental
Research Laboratory, EPA, are acknowledged for their continuing contri-
butions 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.
We specifically acknowledge the direct consultation and assistance of
Dr. E. P. O'Neill of Westinghouse in performing this study.
xv
-------
SECTION 1
INTRODUCTION
Conventional steam-turbine cycles and combined steam- and gas-
turbine cycles are bring utilized with fluidized-bed coal combustion
techniques, (e.g., atmospheric-pressure fluidized-bed boilers, super-
charged fluidized-bed boilers, and adiabatic combustors) to develop
power generation systems having superior environmental control and
economics. Current development programs primarily emphasize the use
of calcium-based sorbents (limestones and dolomites) to provide in situ
control of SO in the combustor.
X.
First-generation fluidized-bed combustion power plants will operate
with once-through utilization of the calcium-based sorbents to generate
calcium sulfate containing, dry, granular, spent sorbent material. The
quantity of spent sorbent produced will be approximately one to four
times as much as is produced by an equivalent limestone wet scrubber (on
a dry basis) for a conventional power plant but will be easier to handle
and may be environmentally more acceptable than the sludge produced by
wet scrubbing.
For second-generation fluidized-bed combustion power plants regen-
eration processes using calcium-based sorbents are being studied for
the purpose of reducing the quantity and environmental impact of spent
sorbent material. The state of development of regeneration processes
for calcium-based sorbents is characterized by bench-scale experimentation
and uncertain performance and economic feasibility.
An alternative to limiting calcium-based spent sorbent production
by sorbent regeneration is utilizing other granular sorbent material
in the place of calcium-based sorbents. Alternative sorbents are
expected to be much more expensive than natural calcium-based sorbents
-------
in the place of calcium-based sorbents. Alternative sorbents are expected
to be much more expensive than natural calcium-based sorbents because of
raw material preparation. In order to be successful they must be highly
regenerable and free from loss in physical integrity. The sorbent must
result in acceptable plant performance and economics.
This report on alternative sulfur removal sorbents presents the
results of the work scheduled to identify alternative sorbents, for
atmospheric and pressurized fluidized-bed combustion systems, to evaluate
their technical and economic potential, and to identify development
requirements. Recommendations for experimental and alternative sorbent
system development work are presented.
In this report potentially attractive alternative sorbents for
fluidized-bed combustion (simple metal oxides and mixed metal oxides)
are identified by screening evaluations of performance and economic
factors. Performance criteria are developed for utilization in experi-
mental data evaluation.
In conducting this screening, thermodynamic, melting point, and
material and energy balance limitations for simple and double metal oxides
have been projected. Estimates of sorbent availability and economic
limitations are developed and translated into maximum tolerable rates of
alternative sorbent losses from all causes (attrition, elutriation
deactivation, agglomeration, etc.). Optimistic assumptions are applied
in the screening evaluation in order to project maximum sorbent potential.
Critical design and testing areas are identified and a program plan is
developed to evaluate attractive alternative sorbents in more detail
(Appendix A).
An alternative once-through calcium-based sorbent concept for
fluidized-bed combustion has also been evaluated (Appendix B). The
concept, "Intimate Coal/Sorbent Mixtures," applies methods of bringing
coal and calcium oxide into intimate contact during coal combustion to
improve calcium utilization and SO emission control, and to reduce the
3t
quantity of spent sorbent. Techniques for realizing intimate contact
are screened and the most promising technique,pelletization of coal and
sorbent into consolidated particles, is evaluated with respect to
formance potential and economic limitations.
-------
SECTION 2
CONCLUSIONS
ALTERNATIVE SORBENTS
Fourteen alternative sorbents have been identified as
possibly suitable for atmospheric pressure fluidized-bed com-
bustion: Na2C03, CaO/CaCO , SrC03, BaC03, LiA102, LiFe02,
Li2Ti03, NaA102, NaCOg • Fe^, CaAl^, SrAl^, SrTi04
4 and
Eleven sorbents are possibly suitable for pressurized fluidized-
bed combustion: CaO/CaCO-, Sr2C03, BaC03, LiA10_, LiFe02,
Li2Ti03, CaAl20^, SrAl^, SrTiO.,, BaAl^ and BaTi03.
The availability of the active sorbent material or the support
material will be a very important factor in justifying the
development of fluidized-bed combustion with alternative
sorbents.
If maximum sorption/regeneration performance levels could be
achieved, it is expected that alternative sorbent loss rates
(due to attrition, deactivation, etc.) must be less than 0.1 per-
cent of bed content per hour in order to compete economically
with regenerative limestone/dolomite-based fluidized-bed
combustion (reductive decomposition regeneration) . This
criterion must be translated into experimental performance
goals for cyclic-kinetic studies and attrition studies.
Attrition behavior under simulated process conditions should be
characterized in order to select promising support materials,
although availability may determine the support material.
-------
• The kinetics of desulfurization and regeneration should be
characterized within the applicable range of operating conditions
(temperature, pressure, fuel, etc.) and under conditions which
simulate the maximum performance levels.
• Unless the high levels (near maximum) of S0? or H~S projected
in this report can be achieved in the regeneration process,
the alternative sorbents will be economically unfeasible.
• At present, too little is known about the cost of sorbent prepa-
ration. Feasible process technology for sorbent preparation
should be proposed and the economic feasibility assessed.
• The availability of the sorbent materials and the support mate-
rials require more modeling in order properly to assess all of
the market factors. The area of industrial expansion into
alternative sorbent preparation and distribution must also be
explored.
• Alternative sorbents may require alternative fluidized-bed com-
bustion design concepts to satisfy their limitations and promote
their advantages. The impact of alternative designs that would
reduce sorbent attrition/deactivation should be evaluated (e.g.,
low fluidization velocities, fine particles, shallow beds, low
thermal shock control).
• The environmental impact of alternative sorbents may lead to
elimination of certain candidate materials (spent sorbent dis-
posal, trace metals release, impact on NO , CO, particulates,
2t
etc.). The complex sociological interactions of these materials
must be explored before development is committed.
INTIMATE COAL/SORBENT MIXTURES
• The only intimate coal/sorbent mixture that could be identified /
as being technically feasible for the current fluidized-bed com-
bustion design concept is the pelletization of coal and sorbent
into consolidated particles.
-------
Attrition of the consolidated particle is the critical factor
influencing the performance and feasibility of the concept.
Modifications to the combustor design would probably be required
in order to apply the consolidated particle concept.
The performance (technical and environmental) cannot be estimated
without initiating a test program. The overall technical and
environmental performance of the consolidated particle concept
could conceivably be worse than or greater than the case of con-
ventional coal and sorbent injectiot into a fluid-bed combustor,
but it is probable that no significant improvement in performance
will be realized.
Except under very extreme conditions, the consolidated particle
concept will not be economically competitive with conventional
fluid-bed combustion concepts.
Washing the pulverized coal during consolidated particle
preparation could reduce trace elements, ash, sulfur, and the
sorbent requirement. The economics of this option have not
been investigated.
The most attractive consolidated coal/sorbent particle from a
technical and environmental impact standpoint would utilize a
binder that would maintain the coal-ash and sorbent particles
in discrete, consolidated particles following combustion. A
binder that will provide this behavior has not yet been
identified.
-------
SECTION 3
RECOMMENDATIONS
ALTERNATIVE SORBENTS
• The three most promising alternative sorbents, BaCO_, CaAl~0,,
and BaTiO., are recommended for further evaluation.
• The following tasks are recommended for the next phase of the
current program to be performed with the selected sorbents:
1. Further evaluation of the availability, cost, and properties
of selected sorbents and support materials through
contacts with industrial suppliers
2. Evaluation of alternative sorbent commercial preparation
methods and preparation costs
3. Preparation of the selected sorbents, using commercial
preparation techniques, and evaluation of the prepared sorbents'
physical and chemical properties
4. Collection of data on desulfurization kinetics and regenera-
tion kinetics on small-scale TGA and fluid-bed units under
high-performance conditions. Tests on sorbent attrition
at process conditions.
5. Assessment of plant performance and cost with the selected
sorbents
6. Proposal of development plan for the best alternative
sorbent if, based on the above evaluation, one appears
commercially advantageous compared to the naturally
occurring calcium-based sorbent materials.
• On the basis of availability considerations, the support material
should be alumina. The prepared alumina properties to be
-------
utilized must provide a trade-off between low cost, high
attrition resistance, high sorbent kinetics promoted by
large pore diameters, and the amount of active sorbent that
can be loaded. Alumina suppliers should be contacted.
INTIMATE COAL/SORBENT MIXTURES
• The concept of intimate coal/sorbent mixtures should not be
pursued unless a binder can be identified that will maintain the
structure of the consolidated coal/sorbent particles following
combustion, or unless the economics of washing the pulverized
coal to remove sulfur and ash during consolidated particle prepara-
tion is found to be attractive
• The areas of improved binder identification and pulverized coal-
washing evaluation are recommended for further study by
engineering assessment or small-scale experimentation.
-------
SECTION 4
SCOPE AND METHOD OF ASSESSMENT
The possibility that there are alternative sorbents for SO absorp-
tion that are superior to naturally occurring calcium-based sorbents
(limestones and dolomites) in terms of performance and economics has
been the subject of several previous screening studies. Two of the
studies have specifically considered the application of regenerative
alternative sorbents to atmospheric-pressure fluidized-bed combustion. '
These studies, in general, are devoted to the discussion of the thermo-
dynamic potential, with additional but limited comment on sorbent
cost and availability.
In the present screening evaluation of alternative sorbents for
fluidized-bed combustion, both atmospheric-pressure and pressurized com-
bustion systems are considered. Operating conditions applicable to the
combustion systems and sorbent regeneration systems are defined, and
acceptance criteria used to judge thermodynamic, material, and energy
balance limitations are developed and specified. Alternative sorbents
satisfying these criteria are evaluated by simple models for availability
and cost limitations in order to project maximum acceptable rates of
sorbent loss due to attrition, elutriation, deactivation, agglomeration,
and the like. The results provide a qualitative perspective, rather than
quantitative design factors, on the required sorbent behavior.
The purpose of the study is to identify acceptable alternative
sorbents for fluidized-bed combustion and to recommend testing areas
that will further assess their potential. Critical process evaluations
are also recommended.
-------
On the basis of screening results, an experimental testing
program has been devised to evaluate the critical performance factors of
selected alternative sorbents (Appendix IIIA) . Specific test
conditions and acceptance criteria for the tests are proposed.
In addition, a second fluidized-bed combustion process concept is
evaluated (Appendix IIIB) . This concept is based on "intimate coal/sorbent
mixtures" and is an alternative once-through sorbent concept for
calcium-based sorbents.
SYSTEM CONCEPTS
The general concept applied is analogous to the current fluidized-
bed combustion concepts based on calcium-based sorbents (limestone or
dolomite) : in situ coal combustion and in situ desulfurization within
a fluidized-bed environment using granular sorbent as the bed material.
Three variations of the fluidized-bed combustion process are represented:
The atmospheric-pressure fluidized-bed boiler (AFBC) , the pressurized
fluid-bed boiler with combined-cycle power generation, and the
adiabatic fluidized-bed combustor, also based on combined-cycle
operation (PFBC) . In addition, each of these concepts may utilize either
current or advanced turbine technology, leading to changes in the
applicable operating conditions of the fluidized-bed combustor.
Four concepts for the regeneration of the sorbent are evaluated :
1. Thermal decomposition:
5=^ MeO + S0 + 1/2 0
where MeO represents a metal oxide and MeSO, is a metal sulfate.
2. Reductive decomposition:
-------
3. A two-step process based on sulfate reductipn to sulfide
followed by H_S generation by steam reaction:
H°
MeS + H20 7—* MeO
where MeS is a metal sulfide.
4. Finally, a two-step process in which the H_S generation
step is based on steam/carbon dioxide reaction:
MeS + C02 + H20 ^H j ^
where MeCO., is a metal carbonate.
It is assumed that all energy and reductant requirements for regeneration
are supplied by coal utilization rather than by clean fuels.
The concept of a once-through operation with an alternative sorbent
is rejected on the grounds that no commercially manufactured alternative
sorbent could compete economically with a limestone or dolomite
in once-through operation. No inexpensive, naturally occurring
alternative sorbents can be identified whose performance can compete with
that of limestone and dolomite. Thus, the use of alternative sorbents
must be based on the concept of preparing an artificial sorbent by chemical
processing from the raw sorbent constituents and a suitably prepared
support material,and must be characterized by improved regenerability and
reduced attrition and deactivation relative to naturally occurring calcium-
based sorbents.
10
-------
OPERATING CONDITIONS AND ACCEPTANCE CRITERIA
The operating conditions selected for the three fluidized-bed com-
bustion concepts are based on the conditions required for optimum
power generation. Table 1 lists the ranges of conditions for the
fluidized-bed combustion systems with current and advanced turbine
technology. Fluidized-bed combustor pressure, the normal operating
temperature range, turndown temperature, and the excess air levels
in the combustor are listed. It is assumed that the combustor
temperature should not exceed the agglomeration temperature of the coal
(selected as 1040°C) or the sorbent melting point. The resulting flue gas
composition (percent oxygen, carbon dioxide, and SO emission
2t
required to meet the EPA emission standard for coal-fired power
plants) is also presented. The equilibrium SO level (SO,, plus
A. ^
SO,,) of which the alternative sorbent must be capable at combustion
conditions is selected as 20 percent of the emission standard value in
order to provide a reasonable reaction driving force in the combustor.
The selection of this criterion is rather arbitrary, but the general
results do not depend critically upon it.
The regeneration system operating conditions must be compatible
with the combustor operating conditions. The conditions are listed in
Table 2 ,for each of the power cycle concepts with regeneration,by thermal
decomposition, reductive decomposition, or the two-step processes. Again,
the regenerator temperature should not exceed the agglomeration point
of the bed or the melting point of the sorbent, taken here as 1400°C to
account for the possibility of high softening-point coals. The minimum
equilibrium value of S02 (4 mole %) or H-S (3 mole %) generated in
the regenerator is listed,based on a knowledge of the requirements
for sulfur recovery processes.
Material and energy balance constraints will be limiting in many
cases of high equilibrium potential (e.g., the energy required for the
reaction system may require the addition of energy combustion of coal
such that the combustion products dilute the reaction products to levels
below equilibrium values).
11
-------
TABLE 1. FLUIDIZED-BED COMBUSTOR CONDITIONS AND CRITERIA
K)
System
Bed
Pressure,
kPa (atm)
Bed Temi
Normal
°C
jerature
Minimum3
°C
Flue Gas Composition
Excess
Airb
%
02, %
COy, %
SOX
Emission
Ppmc
Equilibrium
Criteria
for S0xd
ppm
Atmospheric
Boiler (AFBC)
Current turbine
technology^6)
Advanced turbine
technology"
Pressurized
Boiler (PFBC)
Current turbine
technology
Advanced turbine
technology8
Adlabatic
Comb us tor (PFBC)
Current turbine
technology
100 (1)
100 (1)
1000 (10)
1000 (10)
650-1040
760-1040
600 10-25
700
2-4 15-13
600-530 120-106
900-1040
900-1040
760
760
10-100
2-10 15-8.5 600-330 120-66"
1000 (10) 900-1040
760
300
16
4.25
166
33
^Estimated minimum bed temperature during turndown.
Based on complete combustion of the fuel carbon and hydrogen to C02 and H20; neglects 02 and
N2 in the coal and combustion of the fuel sulfur.
'TBased on current EPA S02 emission standard of 0.516 kg S02/GJ.
Based on equilibrium SOX (S02 + 803) equal to 20% of S02 standard to provide reasonable kinetic
driving force in the desulfurizer.
^Current steam-turbine technology 540°C (1000°F) steam superheat temperature.
Advanced steam-turbine technology 650eC (1200*F) steam superheat temperature.
^Refers only to steam-turbine technology; although advanced gas-turbine inlet conditions of up to 1540°C
(2800°F) and 250 kPa (25 atm) are forecast, the maximum bed operating temperature of 1040°C (1900°F)
(based on ash softening) restricts gas turbines to current technology of 870-1040°C (1600-1900°F) and
800-1000 ifcPa (8-10 atm). Coals having higher ash-softening points or alternative fuels (residual fuel
.oils, etc.) could conceivably utilize advanced gas-turbine technology but are not considered.
Alternatively, a level of about 10 ppm could greatly improve turbine corrosion performance, and
sorbents having this capability should be so noted.
-------
TABLE 2. REGENERATION CONDITIONS AND CRITERIA
U>
Combustion
System
Bed
Pressure,0
kPa (atm)
Bed
Temperature
Fuel
Typec
Reducing
Gas
Systemd
Equilibrium
S02 or H2S
Criteria^
(mole Z)
Energy and
Material
Balance
Maximum
S02f
(mole %)
Atmospheric Boiler
TAFBC)
Thermal decomposition
Reductive decomposition
Two-step
1st step
2nd step
Pressurized Boiler or
Adiabatic Combustor
(PFBC)
Thermal decomposition
Reductive decomposition
Two-step
1st step —
2nd step
100 (1)
100 (1)
100 (1)
1000 (10)
1000 (10)
1000 (10)
Less than coal
agglomeration
(•v-1400ec) or
sorbent melting
point
r
Less than coal
agglomeration
(^IWO'C) or
sorbent melting
point
Coal
Coal
Coal
Coal
Coal
Coal
>4
in situ >4
-2 8
in situ <10
>3 (H2S)
— >4
in situ >4
-2 8
in situ <10 *
>3 (H2S)
>4
>4
— -
>3
>4
>4
—
>3
pressure should be close to the combustor pressure based on sorbent circulation feasibility;
some difference can be tolerated - expected to be about 100 kPa difference for the atmospheric
boiler case and up to 500 kPa for the pressurized cases.
The maximum regeneration temperature may be based upon bed sintering (agglomeration) , sorbent
deactivation, energy restrictions, economics, etc.
Coal will be used to supply energy in all cases except where the amounts required are trivial.
If reducing gas Is generated by in situ coal partial combustion, then the ash agglomeration point
may limit the operating temperature or an agglomerating bed with ash-sorbent separation may ba
used.
SLower S02 or H2S concentrations require large equipment, energy requirements, and difficult
sulfur recovery. In practical terms sorbents providing S02 greater than 10 mole % will be
more realistic. The generation of MeS can be tolerated as long as the 502 content meets the
specified criteria and the sorbent- circulation rate does not reach unreasonably high rates.
The maximum S02 concentration based on stoichiometry and energy balance assumptions may be
estimated and Is related to the rate of MeS generation. In practical terms, 10 mole % S02 is
more realistic.
nd other sulfur products generated in the first step may be released if they meet
environmental requirements or may be recycled to the combustor if necessary.
-------
Minimum values of SCL or H?S based on material and energy
balances are also selected at 4 and 3 mole % for S0_ and H-S, respectively.
The extent of sorbent regeneration (fraction of sulfate converted to oxide)
is a process variable which is not directly restricted. Any value
is acceptable if it leads to acceptable sorbent activity and system
economics.
The rate of sorbent losses due to attrition or deactivation is not
constrained by thermodynamics or material and energy balance considera-
tions but by sorbent availability and cost considerations. Sorbent losses
are evaluated parametrically later in this report to provide guidance for
sorbent attrition testing.
Of course, in a practical system design many other sorbent character-
istics will be of great importance. The kinetics of desulfurization and
regeneration are obviously important and depend upon such factors as the
structure of the sorbent and the operating conditions. Such factors as
environmental impact (e.g., trace element release in the combustor and leaching
properties of the spent sorbent), particle agglomeration, fluidization
quality, bed mixing behavior, heat transfer, tube erosion and corrosion,
bed pressure drop, turbine protection, and safety must eventually
be addressed to find the attractive alternative sorbents.
In summary, the initial sorbent screening is based upon the follow-
ing acceptance criteria:
• Thermodynamically predicted SO control capabilities in the
combustor
• Thermodynamically predicted SO^ or H_S concentrations achievable
from the regenerator
• SO or H2S concentrations from the regenerator based on material
and energy balances
• Phase-change limits (melting), based on sorbent physical proper-
ties, relative to the acceptable combustor and regenerator operat-
ing conditions.
14
-------
The approach is to eliminate any sorbent from consideration
once it is determined that it does not meet any one of the acceptance
criteria. This method of elimination quickly reduces the number of
candidates and reduces to a minimum the effort required at each
stage of the study. Optimistic assumptions are applied in order
to determine the maximum possible sorbent performance.
The assessment based upon each of the acceptance criteria is dis-
cussed in Sections 5, 6, and 7 of this report. The sorbents which meet
all four of the criteria are then evaluated further in Sections 8 and 9
on the basis of availability and economics.
SORBENT MATERIALS CONSIDERED
Only metal oxides have been previously identified as potential S0_
sorbents. Both simple metal oxides and double oxides are evaluated.
The initial list of sorbents to be considered and those not to be
evaluated are selected from previous studies and are listed in
Tables 3 and 4. The explanations provided in the original studies for
the sorbents not considered are also included in Table 4 and were
accepted without further assessment.
Alternative support materials for the sorbent structure are dis-
cussed under the topic of availability since it is assumed they do not
affect the sorbent thermodynamic behavior. While interactions between
the sorbent material and support material may occur, this assumption is
reasonable for the initial screening.
15
-------
TABLE 3. INITIAL LIST OF ALTERNATIVE SORBENTS
EVALUATED (METAL OXIDES)
Li20
Na20
K20
BeO
MgO
CaO
SrO
BaO
La2°3
Ce2°3
Ce°2
Zr°2
V2°4
Mn203
CoO
NiO
CuO
Ag20
ZnO
CdO
^2°3
Sn°2
PbO
Sb2°3
Bi20
uo3
Th02
LiA102
Li2Cr04
Li2Cr2°4
LiFe02
Li2T1°3
LiV03
NaA102
Na2Cr04
Na2Cr2°4
NaFe02
Na_TiO_
2 3
NaVO,,
KA102 SrAl204
K Cr04 SrCr04
v c*-** r\ o^-/"1^- r\
K2Cr2°4 SrCr204
KFe02 SrFe204
K_TiO_ SrTiO-
Z. J j
•j 2. D
BeAl204 BaAl204
BeCrO. BaCrO.
4 4
BeCr204 BaCr^
BeFe00. BaFenO,
24 24
BeTi03 BaTiO
BeV206 BaV206
CaCrO,
CaCr204
CaFe204
CaV206
16
-------
TABLE 4. METAL OXIDES NOT CONSIDERED BECAUSE OF PREVIOUS JUDGMENTS
Sorbent
Rb2°
Cs20
SC2°3
Y2°3
Ti02
Hf02
V2°3
vo2
Nb205
Ta2°5
Cr2°3
Mo°3
W03
Explanation
3
3
2
2
2
2
1
1
2
2
4
3
2
Sorbent Explanation
MnO 1
Mn02 1
Re02 2
FeO 1
Ir02 2
PdO 2
Cu20 1,3
Ga203 2
Si02 4
Ge02 2
As203 2
Double oxide silicates 2
CaTi03 2
1 - Not stable oxide form at combustor conditions.
2 - Insufficient data.
3 - Low melting point.
4 - Other.
17
-------
SECTION 5
PROJECTIONS OF DESULFURIZATION PERFORMANCE
Alternative sorbents having acceptable desulfurization potential
are identified by examining three areas:
• The equilibrium behavior of the desulfurization reaction
MeO + S02 + 1/2 02 ^r=^MeS04
• The melting points characteristic of the sorbent materials
• The metal carbonate stability MeO + CO ^—£ MeCO~ in the
desulfurizer environment.
EQUILIBRIUM DESULFURIZATION PERFORMANCE
tions
Data are available on the equilibrium behavior of the reac-
5-7
MeO + S03, ^ = XSQ P
x1/2x
°2 S°2
SO. + 1/2 0-, K9 = —~ P
z L L Xso3
where X. is the mole fraction of species i and K is the
appropriate equilibrium constant.
The total SO concentration at equilibrium is
K
.
XSO XSO_ + XSO_ P + Yl/2p3/2 '
x 2 3 ^V>
U2
18
-------
The limiting values for K can be determined for each fluidized-hed
combustion variation based upon the pressures, temperatures and oxygen
concentrations given in Table 1; the equilibrium criteria SO concentra-
tion given in the last column of Table 1; and the value of K? obtained
from the literature. Table 5 presents values for the limit curves for
sorbent acceptance.
Curves giving the equilibrium constant for 32 metal oxides are shown
in Figures 1 and 2 (see curves 1 through 32), together with the limit
curves for log K taken from Table 5 (see curves A through D). The
potentially acceptable sorbents and their maximum operable combustor
temperatures have been determined on the basis of Figures 1 and 2.
Results are shown in Table 6. Note that the number of potential
sorbents for the pressurized systems is much lower than that for
the atmospheric pressure boiler. The 29 sorbents included in Table 3
but not listed in Table 6 are entirely below the limit curves and are
thus eliminated from further consideration. In addition, curves are
not shown for Li20, Na20, K20, CaO, SrO, BaO, LiA10_ and Li-TiO.,, all of
which lie far above the limit curves and are,thus, acceptable.
SORBENT MELTING DATA
Many sorbent physical properties are of interest (such as
density, hardness, specific heat, and so forth.). Melting-point data
are critical in identifying the potential operating conditions for
the alternative sorbents. Melting points were collected from standard
l
references and are listed in Table 7 for the oxide, carbonate, sulfate,
o q
and sulfide forms. '
On the basis of these data, the following metal oxides should be
eliminated from further consideration since the melting point of the
oxide or the sulfate is exceeded in the required temperature range
identified in Table 6: K20, Ag20, Bi^, and Zr02. Low melting points
for the components Na2C03> LiC03> Na^O^, PbO, Cr03 and V^ also
limit the applicability of some sorbents.
19
-------
TABLE 5. LIMITING VALUES FOR THE METAL SULFATE DECOMPOSITION
FOR SORBENT ACCEPTANCE
Temp., °C
-10*io Ki <
Atmospheric Pressure
Boiler - 4 v/o On
Pressurized Boiler
2 v/o 0
air)
(low excess
Pressurized Boiler
10 v/o 02 (high
excess air)
Adiabatic Conibustor
16 v/o 00
600
650
760
900
1040
760
900
1040
760
900
1040
760
900
1040
4.159
4.291
4.678
5.186
5.584
3.370
3..814
4.194
.438
,814
.132
• 697
.015
4.343
20
-------
Curve 687466-B
10
- 4
CO
S1 6
I
in
2 5
S1
I
10
i I I
Log^Kj Limit Curves
A Atmospheric Boiler
B Adiabatic Combustor
C Pressurized Boiler Low 02
D Pressurized Boiler - High 0
Dashed Lines Represent
Operating Temperature
Limits from Table 1
600 700 800 900
Temperature, °C
1000
Equilibrium
Constant
Curve Metal Oxide
1100
Temperature Range for Atmospheric Boiler
Temperature Range
for Pressurized Combustors
Figure 1 - Equilibrium constant for MeSO. ^ MeO + SO,.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
NiO
ZnO
CdO
Ce2°3
Ce02
Li2Cr04
LiFe02
LiV03
NaAI02
NaFe02
NaV0
CaALO.
i 4
CaCr04
CaV206
SrAI204
SrCr04
SrTi03
SrV2°6
BaAI204
BaCr04
BaFe204
BaV206
Bi2°3
CoO
ivtgb
PbO
Th02
Zr00
21
-------
Curve 687467-B
crv
O
cfT 6
O
III
r—I
ii
J5 5
30
Log,0 K, Limit Curves
A Atmospheric Boiler
B Adiabatic Combustor
C Pressurized Boiler Low (L
D Pressurized Boiler -High (L
, Dashed Lines Represent
Operating Temperature
Limits from Table 1
Equilibrium
Constant
Curve Metal Oxide
600 700 800 900
Temperature, °C
1000
1100
Temperature Range for Atmospheric Boiler
I- -)
Temperature Range
for Pressurized Combustors
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
NiO
ZnO
CdO
Ce02
Li2Cr04
LiFe02
LiV03
NaAI02
NaFe02
NaV03
K2Cr04
CaAI204
CaCr04
CaV206
SrALO,
i. 4
SrCr04
SrTi03
SrV206
BaAI204
BaCr0
BaTi03
BaV.,0,
L 0
Ag2o
Bi203
CoO
La2°3
IVlgO
PbO
Th02
2
Figure 2- Equilibrium constant for MeS04 2 MeO + SO,
22
-------
TABLE 6. METAL OXIDES ACCEPTABLE BASED ON THERMODYNAMICALLY
PREDICTED DESULFITRIZER SO CONCENTRATION
x
Maximum temperature (°C) at which the alternative sorbent will meet the
equilibrium criteria SO concentration (Table 1)
Atm-Pi
Metal Oxide Boi
Pressurized
essure Boiler^
lera (high and low excess air)
Li20 1040 1040
Na20 1040 1040
K20 1040 1040
CaO 1040 1040
SrO 1040 1040
BaO 1040 1040
LiAK>2 1040 1040
Li2Ti03 1040 1040
CdO 735 NA°
Ce203 800 NA
Ce02 725 NA
Li2CrO, 665 NA
LiFe02 965 1040
LiV03 750 930
NaA102 830 905
NaFeO- 1030 1040
K2CrO, 750 NA
CaAl204 970 1040
SrAl204 1010 1040
SrTi03 925 1025
SrV206 665 NA
BaAl204 1010 1040
BaFe_04 675 NA
BaTiO 1020 1040
BaV.O, 685 NA
A
Ag20d 870 , 990
Bi203 650 NA
La203d 810 NA
MgO 760 NA
PbOd 870 970
ZrO. 780 NA
Adiabatic
Combustorb
1040
1040
1040
1040
1040
1040
1040
1040
NA
NA
NA
NA
1040
900
NA
1040
NA
1040
1040
1010
NA
1040
NA
1040
NA
970
NA
NA
NA
960
NA
Minimum operating temperature of 650°C.
Minimum operating temperature of 900°C.
CNA Not acceptable because maximum temperature limit is less than
minimum operating temperature for the fluidized-bed combustion
variation (Table 1, Column 2).
Llata available to 800°C extrapolated to higher temperatures.
23
-------
TABLE 7. MELTING-POINT DATA
Sorbent
Li20
Na20
K20
CaO
SrO
BaO
CdO
Ce2°3
Ce02
B12°3
La203
MgO
PbO
Zr02
LiA102
Li2Cr04
LiFe02
NaA102
NaFe02
CaAl204
SrAl204
SrTi03
BaAl204
BaFe204
BaTi03
BaV2°6
A12°3
Ti02
Fe203
Oxide Form
1727
920
-v700
2587
2430
1923
Decomposes
1687
2727
Decomposes 187
817
2317
2802
886
2677
1900-2000
1550
—
::
1800
968
1600
—
—
—
~—
2045
1825
2435
1565
690
Melting
Carbonate Form
727
856
—
1339
1500
1740
Decomposes
—
::
Decomposes
Decomposes
Decomposes
Decomposes
—
727
727
727
727
690
856
856
1339
1500
1500
690
1740
1565
1740
690
Point (°C)
I Sulfate Form
—
884
1069
1450
1605
1347
1000
Decomposes
::
Decomposes 405
Decomposes 1150
Decomposes
1170
Decomposes 410
860
860
860
860
690
884
884
1069
1450
1605
1605
690
1347
1347
, 1347
690
Sulfide Form
—
950
—
Decomposes
<2000
1200
1750
<2100
—
2100-2150
1112
—
900-975
900-975
900-975
900-975
690
950
950
Decomposes
2045
1825
690
1200
1200
1200
69Q
( I
24
-------
METAL CARBONATE STABILITY
Figure 3 shows the equilibrium CO partial pressure for the metal
carbonates which remain as potential sorbents. Lines representing the
combustor CO^ exit content are also shown for the three power cycle
concepts (Table 1). If the sorbent CO equilibrium curve (curves 1
through 27) is above the combustor flue gas C09 curve (curves A through
D), the sorbent would probably exist in the combustor largely as the
oxide; if below, largely as the carbonate.
Results are summarized in Table 8. Six of the 27 remaining possible
sorbents may be present as carbonates rather than oxides in the fluid-
bed combustor, at least over some of the temperature range of interest.
While the carbonate form may still function as an acceptable sorbent,
two factors must be checked: the carbonate melting point and the desul-
furization potential of the carbonate.
Li«CO melts at 727°C, so it is unacceptable for the pressurized
concepts and is only acceptable for the atmospheric-pressure boiler over
a very limited temperature range. Thus, Li^O is excluded from further
consideration. Na^CO- melts at 856°C and is then completely unaccept-
able for the pressurized concepts. It will be retained for consideration
with the atmospheric-pressure boiler. NaFeO,, when carbonated becomes
Na0C00 • Fe00» and must face the same limitations as Na-C0_ - exclusion
2 3 2 J £• J
from the pressurized concepts and consideration with the atmospheric-
pressure concept. The other carbonates (CaCO , SrCO , and BaCOg) all have
high melting points. Note that the carbonate form of NaFe02 in the
atmospheric-pressure boiler becomes unstable at 860°C, which is almost
identical with the Na2C03 melting point; thus the sorbent may be
acceptable over the entire temperature range of interest.
25
-------
Curve 687469-B
CM
o
Q_
O
10
9
8
7
6
5
4
3
2
0
-1
-2
-3
-4
-5
-6
-7
-8
-9
-T
19
Combustor C02 Level
A - Atmospheric Press. Boiler
B - Adiabatic Combustor
C - Pressurized Boiler - High C-
D - Pressurized Boiler - Low (L
I
600 700 800 900
Temperature, °C
1000
1100
Curve
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Sorbent
Li20
CaO
SrO
BaO
CdO
Ce203
Ce02
La203
MgO
PbO
LiAI02
Li2Ti03
Li2Cr04
LiFe02
LiV03
NaAI02
NaFeO
K2Cr04
CaAI204
SrAI204
SrTi03
SrV206
BaAI204
BaFe204
BaTi03
BaV206
Figure 3- Metal carbonate stability
26
-------
TABLE 8. METAL CARBONATE STABILITY
Sorbent Atra-Pre
(Curve in Fig. 3) Boil
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
Li-0 C
Na 0 C
ssure Pressurized
er Boiler
C
C
CaO 760°Cb 875°C, 910°C (Low 00)
SrO C
BaO C
C
C
CdO Unknown NAC
Ce000 Unknown NA
2 3
Ce02 Unknown NA
La000 Unknown NA
23 ,
MgO 0
PbO 0
NA
0
LiA102 0 0
Li.TiO_ 0
2 3
Li.CrO, 0
2 4
LiFe02 0
LiVO, 0
NaA102 0
0
NA
0
0
0
NaFe02 860° C C
K.CrO. 0
2 4
CaAl00, 0
2 4
SrAl00. 0
2 4
SrTi03 0
SrV^O., 0
2 6
BaAl204 0
BaFe.O. 0
2 4
BaTi03 0
BaV.O, 0
2 6
NA
0
0
! 0
NA
0
NA
0
NA
Adiabatic
Combustor
C
C
860°C
C
C
NA
NA
NA
NA
NA
0
0
0
NA
0
0
NA
C
NA
0
0
0
NA
0
NA
0
NA
C - Carbonate form stable.
Temperature indicates carbonate decomposition point.
'lIA - not applicable (from Table 6).
0 - Oxide form stable.
27
-------
The equilibrium desulfurization potential of the metal carbonates
will be somewhat reduced compared to the metal oxide form. For the
simple oxides:
C0
SO,,, K =
MeO
= P
g()
MeO
MeC03,
Thus, K =
and
or
P /p
rso3/rso3
where P is the equilibrium S0_ for the metal oxide and P__ is the
""o J OU—
actual partial pressure of CO. in the combust or (Table 1) . K_ is the
metal oxide equilibrium constant in Figures 1 and 2, P is the
SUo
equilibrium S0_ partial pressure for the carbonated sorbent, and
P is the equilibrium C0_ partial pressure for the sorbent in
UU2 *•
Figure 3. Then,
lo%o pso - 10«io pso " losio
- logio Ki -
The effect of carbonate stability on the desulfurization potential may
be predicted from the curves in Figures 1 and 2 (giving log.n P ) and
the ratio PCQ /PCQ from Figure 3 (giving I°g10 PCO
(giving P ) .
.n
1U
) and Table 1
28
-------
For the complex oxide forms only NaFeO- is of interest. For
NaFeO-:
2 NaFe02
2 NaFe02
and
2 NaFe02
Thus,
K /K K
2 I NaS0 + Fe + 00
K2K3 * _ K3 *
and
which is the same form as for the simple metal oxides.
Table 9 gives the quantity log -K-log.-K for the five sorbents of
interest. Comparison with Figure 1 shows that the desulfurization
potential of these metal carbonates is acceptable. Log-^K for NaFe02
does not drop below the limit curves in Figure 1. Since -log.^1^ » 10
for CaO, Na_0, SrO and BaO (not shown in Figures 1 and 2), they are not
greatly effected by carbonation.
29
-------
TABLE 9. METAL CARBONATE DESULFURIZATION POTENTIAL
log.- K - log n K. = login P - login P*
1U 1U 1 ±U bO. 1U bU_
Temperature (°C)
627 727 827 9:
Na2C03
n 1027
Atm-Pressure Boiler
Atm-Pressure Boiler
Pressurized Boiler
Adiabatic Combustor
SrC00
Atm-Pressure Boiler
Pressurized Boiler
Adiabatic Combustor
BaCO.,
Atm-Pressure Boiler
Pressurized Boiler
Adiabatic Combustor
10.2
8.4
7.0
5.8
4.7
1.3
2.4
1.8
3.6
4.7
4.1
6.0
7.1
6.5
0.3
1.4
0.8
2.6
3.7
3.1
4.4
5.5
4.9
— —
0.6
0.0
1.5 0.6
2.6 1.7
2.0 1.1
3.2 2.2
4.3 3.3
3.7 2.7
—
—
—
0
1.0
0.5
1.5
2.6
2.0
Atm-Pressure Boiler
2.1
1.1
0.2
30
-------
SULFITE STABILITY
Sulfite stability consideration does not affect the acceptability of
the sorbents. On the basis of existing data, all of the sulfites appear
unstable at the conditions of interest.
SUMMARY OF DESULFURIZATION POTENTIAL
Table 10 summarizes the screening results for alternative sorbent
desulfurization potential. The maximum operable desulfurizer temperature
is listed for the atmospheric boiler, the pressurized boiler, and the
adiabatic combustor. The sorbent form listed in Table 10, oxide or
carbonate, is the form projected to be stable in the combustor at the
temperature maximum tabulated.
Many of the alternative sorbents in Table 10 are of limited interest
because they have a limited operable temperature, may be very limited
in supply,or are toxic. These sorbents are summarized in Table 11 and are
excluded from further evaluation. They are reserved for further assess-
ment if new information becomes available which indicates merit for any
of them. The remaining 14 potential sorbents listed in Table 11 are the
subject of the evaluation to follow. They are listed as the form stable
in the combustor.
31
-------
TABLE 10. SUMMARY OF ACCEPTABLE SORBENTS BASED ON
DESULFURIZATION PERFORMANCE
Sorbent
Maximum Combustor Operating Temperature , °C
Atm-Pressure
Boiler
Pressurized
Boiler
Adiabatic
Combustor
Na2C03
CaO
CaC03
SrC03
BaCO.,
CdOd
Ce 0 d
Ce02d
La2°3d
MgO
PbO
LiA102
Li.TiO.
2 3
LiFe02
LiV03
NaA102
NaCO, • Fe_0-
J £3
SrAl.O,
2 4
SrTi03
SrV2 0
2. D
BaAl20,
BaFe 0,
BaTi03
BaV206
856a
1040
760b
1040
1040
735C
800C
725C
810°
760C
870C
1040
1040
665C
965C
690a
830C
860C
970C
1010C
925C
665C
1010C
675C
1020C
685C
NAa
1040
910b (low 02)
1040
1040
NAC
NAC
NAC
NAC
NAC
NAa
1040
1040
NAC
1040C
NAa
NAa
NAa
1040
1040
1025C
NAC
1040
NAC
1040
NAC
NAa
1040
NAb
1040
1040
NAC
NAC
NAC
NAC
NAC
NAa
1040
1040
NAC
104 Oc
NAa
NAa»C
NAa
1040
1040
1010C
NAC
1040
NAC
1040
NAC
NA - Not acceptable.
«
Based on melting point.
Based on carbonate stability.
£
Based on desulfurization potential (see Table 6).
d
Insufficient information on carbonate stability.
32
-------
TABLE 11. ASSESSMENT OF ACCEPTABLE SORBENTS BASED ON
DESULFURIZATION PERFORMANCE
Sorbents that Merit Further Consideration:
Na2C°3
CaO/CaCO,
SrC0
NaC0
BaC03
LiAlO,
LiFeO,
BaTiO.
NaAlO,
Li2Ti03
Sorbents That Are Reserved for Further Consideration:
CdO
Ce2°3
Ce°2
La2°3
MgO
PbO
LiVO.
BaV206
a
b
a
b
a
c
a
a
a
a
a
a - Limited temperature range; could not reach
advanced steam conditions; may be given
further consideration at future time.
b - Probable high cost or low availability.
c - Toxic.
33
-------
SECTION 6
PROJECTIONS OF SORBENT REGENERATION POTENTIAL
The regeneration potential of the selected sorbents listed in
Table 11 has been projected. Thermal decomposition, reductive decomposi-
tion, and a pair of two-step processes are evaluated.
THERMAL DECOMPOSITION
Examination of Figures 1 and 2 indicates that none of the
sorbents can utilize thermal decomposition as a means of regeneration
within the permissible temperature range. Since -login P _ is larger
lu
than 1.4 for all of the acceptable sorbents in Table 11, the SO concen-
tration would be too low (less than 4 mole %) to satisfy the regeneration
criteria in Table 2.
REDUCTIVE DECOMPOSITION
The regeneration scheme,
MeS04 •
will occur by using the in situ partial oxidation of coal to supply the
reductants. The use of clean fuels such as natural gas and No. 2 fuel
oil has been excluded. If it is assumed that the sulfidation reaction
is also in equilibrium, the equilibrium S02 concentration may be
projected by examining the reaction equilibrium for the overall reaction
3 MeS04 + MeS i=± 4 MeO + 4 S02
34
-------
For the sorbents in which the carbonate form is stable, the corresponding
reaction is
3 MeSO^ + MeS + ,4 C02 ^=^ 4 MeCO + 4 S02 .
Table 12 and Figures 4 and 5 summarize the equilibrium data and the
minimum temperatures required to produce 4 volume percent S0? (the
minimum acceptance criterion in Table 2) and 10 volume percent SO., both
at 100 kPa and at 1000 kPa pressure. For the sorbents that are stable
carbonates, a 15 percent C0« concentration has been assumed. The
melting-point limit is also listed for each sorbent in Table 12.
Two of the sorbents require a temperature exceeding the melting-point
limit (and also exceeding the maximum allowable regenerator temperature
of 1400°C) in order to achieve the minimum acceptable S0_ concentration
of 4 volume percent (Table 2). These sorbents, Na_0 and BaO, must
be dropped from consideration for reductive decomposition. In addition,
Li-TiCL is not suitable for regeneration by reductive decomposition at
1000 kPa pressure because the required temperatures at that pressure are
above the sorbent melting point. For three of the stable carbonate
sorbents (based on desulfurizer conditions) - CaCO.,, SrCO^, and
BaCOo - the carbonate forms are not stable in 15 percent C02 at a tempera-
ture sufficiently high to achieve 4 percent SO and must be dropped. The
ultimate conclusion from Table 12 is that the barium-based alternative
sorbent (BaO or BaCO ) is not suitable for regeneration by reductive
decomposition. The other sorbents in either the oxide or carbonate
form are suitable for the temperature ranges listed in Table 12.
TWO-STEP REGENERATION SCHEMES
The reaction equilibrium for the first step of the two-step schemes,
f2}
I CO/
MeS04 + 4 \ ^ + MeS + 4
and
35
-------
TABLE 12. REGENERATION BY REDUCTIVE DECOMPOSITION
3 MeSO, + MeS *=±
Cu
Sorbent3 (Fig
4 MeO + 4 SO
rve Melting
. 4) Limit, "C
Na20b 1 884
CaO 2 1450
SrO 3 1605
BaOb 4 1347
LiA102 5 860
LiFe02 6 860
NaA102 7 884
NaFe02 8 884
CaAljO, 9 1450
SrAl204 10 1605
SrTi03 11 1605
BaAl204 12 1347
BaTiO 13 1347
Li2TiO 14 860
3 MeSO + MeS + 4 CO
^=^ 4 MeC03 -\
(Fig
- 4 S02
5)
Na CO 1 856
b
CaC03D 2 1339
SrC03b 3 1500
BaC03b 4 1347
Na2C03 • Fe,,03 8 856
Atmospheric
Pressure
Minimum
Temperature
for °C
4% S02
>1400
920
1180
1400
700
720
590
820
740
760
580
710
720
800
740
NS
NS
NS
690
10% S02
>1400
975
1235
>1400
750
760
620
860
770
800
630
750
760
870b
860
NS
NS
NS
840
Pressurized
Operation
Minimum
Temperature
for °C
4% S02
NAC
1060
1330
>1400
850
820
NA
NA
830
860
700
810
820
1020b
NA
NS
NS
NS
NA
10% S02
NA
1120
1390
>1400
940b
880b
NA
NA
880
900
750
860
870
1150b
NA
NS
NS
NS
NA
Form of sorbent product from regeneration process.
Dropped from consideration for reductive decomposition because minimum
temperature >1400°C.
NA - Not acceptable desulfurizer performance (Table 10).
NS - Carbonate form not stable in 15% C02 to sufficiently high
temperature to achieve 4% SO-.
36
-------
Curve 637468-B
CNl
O
CO
S1
-1
A - 4% S02 at 100 kPa
D -10% SO at 1000
-2 -
Q ._„„
-4 -
-5
600 700 800 900 1000 1100 1200 1300 1400
Temp°C
Figure 4- One-step reductive decomposition equilibrium S02 projection - no sorbent carbonation 3 MeSO. +
MeSZ
Curve Sorbent
1 NaJ)
2
3
4
5
6
7
8
9
10
11
12
13
14
CaO
SrO
BaO
LiAI02
LiFe02
NaAlO
NaFe0
SrTi0
BaTi0
'2
-------
Curve 687*172-6
CM
O
X
g1
600
800
900 1000 1100 1200 1300
Temp, °C
1400
Figure 5- One-step reductive decomposition equilibrium projection - sorbent carbonation (15% COJ
SMeSO. + MeS + 4 C00 Z 4 MeC(L + 4 S00
4 2 3 f.
-------
may be evaluated by considering again the overall reaction
3 MeSO, + MeS « * 4 MeO + 4 SO .
The maximum regenerator temperature for MeS generation with a resulting
A
low S02 emission (XgQ _< 10 in Table 2 is given in Table 13 and is
based on the curves in Figures 3 and 4. Many of the sorbents require
very low regenerator temperatures in order to satisfy this requirement
(500°C). Reaction kinetics may limit the reduction step of the two-step
process in some cases. None of the sorbents is eliminated on the basis
of the sulfide formation step, but the kinetics of this step must be
explored at these low temperatures to see if the reaction rates are acceptable.
The metal sulfide produced in the sulfate reduction step may be
treated by either of two schemes:
MeS + H0 5=? MeO + HS
or
MeS + H0 + C0 « — » MeC0
3
Table 14 and Figure 6 consider the first reacting scheme. Equilib
rium data are shown in Figure 6. Figure 6 also indicates two curves (A
and B) based on mass balances, giving the mole fraction of lUS as a
function of X^ o/3^ n ^or two *nitial m°le fractions of steam of 1.0
and 0.5; the temperature required to obtain greater than 3 mole % H
(the minimum acceptance criterion) and 10 mole % lUS may be obtained
from the figure and is listed in Table 14 for the case where the
reactant gas is 100 percent steam. The maximum possible H2S mole
fraction is also projected in the table. Because four of the
sorbents require a temperature exceeding the melting-point limit and/or
the maximum operable temperature of 1400° C, they are dropped from
consideration for this reaction scheme.
39
-------
TABLE 13. METAL SULFIDE FORMATION
3 MeS04 + MeS *~> 4 MeO + 4 S02
Sorb en t
CRITERION - X < 10~4
S°2
Curve
(Fig. 4)
100 kPa Pressure
Maximum
Temperature, °C
1000 kPa Pressure
Maximum
Temperature, °C
Na20
CaO
SrO
BaOC
LiA102
LiFe02
NaA102
NaFe02
CaAl00.
2 4
SrAl00,
2 4
SrTiOg
BaAl00.
2 4
BaTi03
LiJTiO,
2 3
Na.CO,
2 3
CaC03
SrC03
BaC03
Na2C03 • Fe2C
1
2
3
4
5
6
7
8
9
10
11
12
13
14
(Fig. 5)
1
2
3
4
>3 8
1150
680
890
1070
«600
520
«600
615
540
550
«600
<580
<550
«600
<600
<600
d
770
«600
NA
760
990
1180
«600
590
NA
NA
615
625
«600
580
550
^500
NA
750
910
1000
NA
Belting limit of 884°C.
NA - Not acceptable (see Table 10).
°Melting limit of 1200°C.
See curve in Figure 5 for exceptional behavior.
40
-------
TABLE 14. H2S GENERATION FROM METAL SULFIDE
BY MeS + H20
MeO +
Sorbent
Curve
(Fig. 6)
Temperature,
°C for
XHzS of
>0.03
>0.10
Melting
Limit, °C
Maximum
Temperature, °C
for Maximum H_S
Na20
CaOb
SrOb
BaOb
LiA102
LiFe02
NaA102
NaFe02
CaAl204
SrAl204
SrTi03
BaAl204
BaTi03
Li2Ti°3
1
2
3
4
5
6
7
8
9
10
11
12
13
14
>1400
>1400
>1400
>1400
<1300
>700
<1300
>800
>700
<1300
<1300
<1300
<1300
<1300
>1400
>1400
>1400
>1400
<1300
>980
<1300
>1160
<1300
<1300
<1300
<1300
, <1300
920
1450
1605
1200
860
860
884
884
1450
1605
1605
1200
1200
860
>0.9
0.10
0.40
0.06
0.10
0.60
>0.9
>0.9
>0.9
>0.9
<1150
975
950
950
950
800
<1150
<950
<800
<900
aBasis - reactant is 100% steam.
Drop from consideration for regeneration by this reaction.
41
-------
H2S (Curves A and B Only)
0.2 0.4 0.6 0.8
Curve 687471-B
1.0
Mass Balance
Curves
„ yo -n c
B.X -0.5
Curve
1
2
3
4
5
6
7
8
9
10
11
12
13
14
900 1000 1100 1200 1300
Temperature, °C
Sorbent
CaO
SrO
BaO
LiAI02
LiFe02
NaAI02
NaFe02
CaAI204
SrAI204
SrTi03
BaAI204
Li2Ti03
Figure 6- MeS +H20^ MeO +H2S reaction equilibrium
42
-------
For reaction with steam and C09, equilibrium data are shown in
Figure 7 for a reaction mixture of 50 percent H.O and 50 percent CO.
for the eight sorbents which will form stable carbonates (see Fig-
ure 3). Curves A and B in Figure 7 show H S levies based upon mass
balance. Table 15 projects the temperature required for 3 and 10 percent
H2S at 100 kPa and at 1000 kPa. For this reaction scheme the equilibrium
H.S increases as the temperature decreases, so there is no apparent
maximum H«S limit. Kinetic limitations will control in the actual case.
None of the possible sorbents being considered is excluded for this
reaction scheme.
SUMMARY OF REGENERATION POTENTIAL
Tables 13 and 17 summarize the equilibrium projections for desul-
furization and regeneration. Table 16 considers the atmospheric-pressure
system and Table 17 considers the pressurized system. For those sorbents
that may be stable carbonates under some operating conditions, the behavior
of both the oxide and carbonate forms are indicated in the summary tables
in order to permit identification of transition points from oxide to
carbonate forms.
43
-------
CJ
O
O
X
CM
CD
O
-2
-3
-4
-5
0.1
AH2S(CurvesA&BOnly)
0.2 0.3 0.4 ,0.5
0.6
T
T i r
Yo — vo —
Mass Balance Curves
A: 100kPa Pressure
B: 1000kPa Pressure
1
Curve Sorbenl
1 Na«0
600 700 800 900 1000
Temperature, °C
1100
1200
Figure 7-MeS +H20 +C02 ^ MeC03+H2S reaction equilibrium
2
3
4
7
8
9
10
CaO
SrO
BaO
NaAlO,
NaFeO
CaAI20
44
-------
TABLE 15. H2S GENERATION FROM METAL SULFIDE
BY MeS + H20 + C02<=± Me CO + H2Sb
Sorbent
Curve
(Fig. 7)
100 kPa Pressure
Temperature, °C,
for X^ of
>0.03 1 >0.10
1000 kPa Pressure
Temperature, °C,
f°r *H2S °f
>0.03 | >0.10
Melting
Limit,
°C
Na20
CaO
SrO
BaO
NaA102
NaFe02
CaAl204
SrAl204
1
2
3
4
7
8
9
10
<1300
<500
<720
<780
<670
<950
<630
<730
<1200
«500
<600
<710
<500
<820
<600
<650
<1300
<690
<880
<920
<840
<1300
<700
<850
<1300
<600
<780
<820
<740
<1200
<650
<790
856
1339
1500
1200
856
856
1339
1500
TTo limit on the maximum X_ for this reaction,
HnS
bBasis - reactance is 50% H20 and 50% C02.
45
-------
TABLE 16. THERMODYNAMIC SCREENING OF AFBC SORBENTS
Atmospheric-
Pressure
Sorbenta
Maximum
Desulfurizer
Temperature, °Cb
One-Step Regeneration
Reductive Decomposition
Minimum Temperature
4% S02
10% S02
Two-Step Regeneration
Ist-Step
Maximum
Temperature, °C
2nd-Step with Steam
Temperature
H2S > 3%
Maximum
H2S,%
2nd-Step witl
Temperature
H2S > 3%
i Steam & CC>2
; (°C) for
H2S > 10%
CT>
Na20
Na2C03
CaO
CaC03
SrO
SrC03
BaO
BaC03
LiA102
LiFe02
U2Ti°3
NaA102
Na0CO_ • Al_0,
23 23
NaFe02
Na-CO- • Fe_0,
23 23
CaAl,0,
CaC03 • A1203
SrAl.O,
SrC03 • A1204
SrTi03
BaAl204
BaTiOg
NS
856
1040
760
NS
1040
NS
1040
1040
965
1040
830
NS
NS
860
970
NS
1010
NS
925
1010
1020
NA
740
920
NS
1180
NS
NA
NS
700
720
800
590
NS
820
690
740
NS
760
NS
580
710
720
NA
NA
975
NS
1235
NS
NA
NS
750
760
NA
620
NS
860
840
770
NS
800
NS
630
750
760
NS
665
760
750
NS
910
NS
1000
500
590
500
500
NS
685
<500
615
NS
625
NS
<500
580
550
NA
NS
NA
NS
NA
NS
NA
NS
<1300
>700
<1300
<1300
NS
>800
NS
>700
NS
<1300
NS
<1300
<1300
<1300
NA
NS
NA
NS
NA
NS
NA
NS
>90
10
>90
40
NS
6
NS
10
NS
60
NS
>90
>90
>90
NS
<856
NS
<500
NS
<720
NS
<780
CNSe
CNS
CNS
NS
<670
NS
<856
NS
<630
NS
<730
CNS
CNS
CNS
NS
<856
NS
«500
NS
<600
NS
<710
CNS
CNS
CNS
NS
<500
NS
<856
NS
<600
NS
<650
CNS
CNS
CNS
Represents sorbent forms at each step of the regenerative cycle.
Minimum temperature of 650°C in the desulfurizer, maximum 1040°C.
°NS - Sorbent in column 1 not stable form.
TNA - Not acceptable.
eCNS - Carbonate -for-m of sorbent in column 1 not stable.
-------
TABLE 17. THERMODYNAMIC SCREENING OF PFBC SORBENTS
Pressurized
Sorbenta
Maximum
Desulfurizer
Temperature, °
One-Step Regeneration
Reductive Decomposition
Minimum Temperature
4% S02
10% S02
Two-Step Regeneration
Ist-Step
Maximum
Temperature, °C
2nd-Step with Steam
Temperature
H2S > 3%
Maximum
H2S (%)
2nd-Step with Steam & (X>2
Temperature (°C) for
H2S > 3%
H2S > 10%
CaO
CaCOj
SrO
SrC03
BaO
BaC03
LiA102
LiFe02
Li2Ti03
CaAl-0,
CCO, • Al,0,
J £• J
SrAl204
SrCO, • A100,
3 23
SrT103
RflAl 0 1
^^ 2 A
BaTi03
1040
910e/NAf
NS
1040
NS
1040
1040
1040
1040
1040
NS
1040
NS
1025e/1010f
1040
1040
1060
NS
1330
NS
NA
NS
850
820
NA
830
NS
860
NS
700
810
820
1120
NS
1390
NS
NA
NS
NA
NA
NA
880
NS
900
NS
750
860
870
NSC
750
1100
910
1200
1170
620 .
670
700
690
NS
710
NS
500
650
660
NAd
NS
NA
NS
NA
NS
<1300
>700
<1300
>700
NS
<1300
NS
<1300
<1300
<1300
NA
NS
NA
NS
NA
NS
>90
10
>90
10
NS
60
NS
>90
>90
>90
NS
<690
NS
<880
NS
<920
CNSg
CNS
CNS
NS
<700
NS
<850
CNS
CNS
CNS
NS
<600
NS
<780
NS
<820
CNS
CNS
CNS
NS
<650
NS
<790
CNS
CNS
CNS
Represents sorbent forms at each step of the regenerative cycle.
Minimum temperature of 900°C in the desulfurizer, maximum 1040°C.
CNS - Sorbent in column 1 not stable form.
dNA - Not acceptable.
ePressurized boiler only.
Adlabatic combustor only.
- Carbonate form of Sorbent in column 1 not stable.
-------
SECTION 7
MATERIAL AND ENERGY BALANCE LIMITATIONS
Generalized material and energy balance relationships were developed
for the regeneration schemes to identify any sorbents that might be
limited in performance because of material and energy balance constraints.
Such constraints may arise in practice when the method of supplying the
energy needs of the reaction scheme (e.g., in situ combustion of coal,
preheat of reactant streams) or controlling the reactor temperature
simultaneously dilutes the reactor product to a low level of S02 or H^S.
There are a large number of variables and parameters that affect
the material and energy balances: the sorbent properties and heats
of reaction, the reaction temperature, the extent of reaction, the
extent of competing reactions, the mass ratio of support material to
sorbent material in the sorbent structure, the sorbent circulation rate,
reactant streams preheat temperature, and so on. After the effects of
these variables and parameters were examined in detail by computer
analysis for some of the sorbents,it was found that the regenerator
performance could range from unacceptable to acceptable for a large
number of combinations of these variables. A set of operating condi-
tions and assumed operating parameters and performance variables exist
for each of the sorbents that could lead to optimum regenerator perform-
ance (e.g., maximum SO- or lUS concentration), but a set also exists
that could lead to unacceptable regenerator performance. Material and
energy balances could be used to identify these sets of operating con-
ditions and performance variables; but due to the large number of param-
eters involved,this is more logically conducted in parallel with an
experimental program that measures critical performance factors. The
48
-------
general material and energy balances, in the absence of data, will not
lead to the elimination of any of the alternative sorbents from
consideration.
In terms of projecting the potential of the alternative sorbents,
material balances around the regenerator have been used to project the
maximum possible performance levels. Energy balances have been excluded
from general consideration in keeping with the previous remarks.
REDUCTIVE DECOMPOSITION
For the reductive decomposition regeneration scheme the number of
moles of carbon consumed in the regenerator per mole of SO,, produced is
given by
assuming that all of the coal fed to the regenerator is completely con-
sumed and complete combustion or reaction of the generated CO and H_
occurs. N_. -,/Ncfj is the number of moles of metal sulfide produced per
MeS au2
mole of SO. generated in the regenerator by the competing reaction, a
is the molal ratio of carbon to hydrogen in the coal, and £ is the frac-
tion of stoichiometric air/fuel ratio used in the regenerator for partial
coal combustion. Assuming that N^pc/NsOo = 0, a = 1.1, and that the
minimum value of £ that can be used (due to tar formation and carbon
losses) is 0.5 (50 percent of the stoichiometric air/fuel ratio), then
the minimum value for N /N is 0.81. For comparison, the moles of
c SO-
carbon consumed in the fluid-bed combustor per mole of SO- produced in
the regenerator may range from about 20 to 100 depending mainly upon the
amount of sulfur in the coal. Thus, as a minimum, the regenerator will
consume from about one to four percent of the total power plant coal.
49
-------
The corresponding maximum SCL mole fraction produced in the regen-
erator may be projected from
Nc -i
=-=- (3.762 (1 + l/4a) + 1 + l/2a) - 1.881 (1+4 (NM p/N0rt ) + 1
SO- 2
X
Again, for a = 1.1, NM q/^gO = ^» an^ ^r^SO e
-------
limit the formation of KLS. The system will be either thermodynamically
or kinetically limited, unless for some practical reason temperature con-
trol of the reaction cannot be achieved.
The molar rate of steam required for the reaction is related to the
molar rate of H S production by
v _^_
V C ^1 C
ri«D n«o
for reaction with steam, and
v
"H2S =
for reaction with steam and C02, where NH Q is the number of moles of
H20 fed to the second-step regenerator. The factor R is equal to the
ratio of steam to C02 in the reaction mixture. A value of 6 for
Ng Q/NH S represents about one percent of the power plant fuel energy
for a 4 weight percent sulfur coal. This corresponds to about 17 percent
HnS in the regenerator product gas for the steam reaction and about
9 percent H2S for the steam/CO- reaction.
SUMMARY
Tables 18, 19, and 20 summarize the projections of sorbent performance
taken from both the thermodynamic evaluations and the consideration of
material and energy balances. Table 18 presents the results for the
reductive decomposition scheme, Table 19 for the two-step process at
atmospheric pressure, and Table 20 for the two-step process at 1000 kPa
pressure.
The tables project regenerator temperature ranges, maximum S0? (or
H.S) mole fractions, and minimum carbon (or steam) consumption factors.
For the reductive decomposition scheme, the minimum temperature limit is
based on the equilibrium S02 mole percentage of 4 (the minimum acceptance
.51
-------
TABLE 18. RESULTS OF MATERIAL AND ENERGY BALANCE CONSIDERATIONS
FOR REGENERATION BY REDUCTIVE DECOMPOSITION
N>
Sorbent
(Form stable
in combustor)
AFBC (100 kPa)
Regeneration
Temperature
Range, °Ca
Maximum
so2, %b
Minimum
N /Ncn c
c S02
PFBC (1000 kPa)
Regeneration
Temperature
Range, °Ca
Maximum
S02, Zb
Minimum
Nc/NS02C
Na2C03
CaO
CaC03
SrC03.
LiA102
LiFe02
L12T1°3
NaA102
NaC03 • Fe203
CaAl20,
SrAl204
SrTi03
BaAl204
BaTi03
740- 865
920-1150
920-1150
>1150
700- 860
720- 860
800- 860
590- 884
690- 856
740-1150
760-1150
580-1150
710-1150
720-1150
10
25
25
—
25
25
9
25
10
25
25
25
25
25
1.80
0.81
0.81
—
0.81
0.81
1.98
0.81
1.80
0.81
0.81
0.81
0.81
0.81
—
1060-1150
1060-1150
>1150
850- 860
820- 860
—
—
—
830-1150
860-1150
700-1150
810-1150
820-1150
—
16
16
—
5
6
—
—
—
25
25
25
25
25
—
1.18
1.18
—
3.45
2.90
—
—
—
0.81
0.81
0.81
0.81
0.81
Minimum temperature corresponds to S02 mole fraction of 0.04 based on thermodynamics; maximum
temperature based on sorbent melting or coal-ash melting (assumed to be 1150°C), whichever is lower.
Based on a minimum air/fuel ratio in the regenerator of 50% of stoichiometric or on thermodynamic
limit, whichever is lowest.
Represents the ratio of moles of carbon (in coal to the regenerator) to moles of S02 produced in the
regeneration process based on material balance only. The corresponding value of this ratio in the
combustor may range from 20 to 100.
-------
TABLE 19. RESULTS OF MATERIAL AND ENERGY BALANCE CONSIDERATIONS
FOR REGENERATION BY THE TWO-STEP PROCESS - AFBC
Ui
Sorbent
(Form stable
in combustor)
First-Step
Temperature3
Range, °C
Second Step with Stjeam
Temperature
Range, °Cb
Maximum Minimum
H2S> %C \ NH,0/NH,Sd
Second Step with Steam/C02
Temperature
Range, °Cb
Maximum
H2S, %c
Minimum
NH20/NH7Sd
Na,C03
CaO
CaC03
SrC03
BaC03
LiA102
LiFe02
Li2T103
NaA102
Na2C03 • Fe20,
CaAl204
SrAl204
SrT103
BaAl204
BaT103
400-550
400-680
400-550
400-750
400-770
400-550
400-520
400-500
400-500
400-500
400-540
400-550
400-500
400-550
400-500
300-860
700-860
300-860
300-884
800-884
700-1300
300-1300
300-1300
300-1200
300-1300
99
6
95
39
5
9
60
99
99
99
1.01
16.67
1.05
2.56
20.0
11.11
1.67
1.01
1.01
1.01
300-856
300-500
300-500
300-720
300-780
—
300-670
300-856
300-630
300-730
—
—
—
99
18
18
39
77
—
31
79
57
54
—
~
—
1.01
3.28
3.28
1.78
1.15
—
2.11
1.13
1.38
1.43
—
—
—
BFor first step minimum value of NC/NH2S ls 3.26 for all sorbents based on a minimum air/fuel ratio of 50% of
stoichiometric. Minimum temperature based on estimate of minimum coal partial oxidation temperature; maximum tempera-
ture based on limiting S02 release to 10~2 mole % or sorbent melting.
Minimum temperature based on avoiding steam condensation (300°C) or achieving minimum XH „ of 0.03; maximum
temperature based on sorbent melting poing or achieving XH g of 0.03. 2
c i
Based on thermodynamic equilibrium.
Represents the moles of steam supplied to the reactor per mole of H2S produced. For a 4 wt% sulfur coal a
N« rt/Nu <• ratio of 6 represents about IX of the power plant fuel energy required for the second step.
2 2
-------
TABLE 20. RESULTS OF MATERIAL AND ENERGY BALANCE CONSIDERATIONS
FOR REGENERATION BY THE TWO-STEP PROCESS - PFBC
Sorbent
(Form stable
in combustor)
First-Step
Temperature
Range, °Ca
Second Step with Steam
Temperature
Range, °Cb
Maximum
H2S, %c
Minimum
NH2o/NH2Sd
Second Step with Steam/C02
Temperature
Range, °Cb
Maximum
H2S, %c
Minimum
NH20/NH2Sd
CaO
CaC03
SrO>3
BaCOg
LiA102
LiFe02
L12T103
CaAl20,
SrAl20,
SrTi03
BaAl204
BaTi03
400-760
400-760
400-910
400-1000
400-600
400-590
400-500
400-615
400-625
400-500
400-580
400-550
300-860
700-860
300-860
700-1300
300-1300
300-1300
300-1200
300-1300
99
6
95
9
60
99
99
99
300-690 55
— 300-690 55
300-880 73
300-920 99
1.01
16.67
1.05
11.11 300-700 83
1.67 300-850 82
1.01
1.01
1.01 —
1.41
1.41
1.18
1.01
—
1.10
1.11
—
—
aFor first step minimum value of NC/NH s is 3.26 for all sorbents based on a minimum air/fuel ratio of 50% of
stoichiometric. Minimum temperature based on estimate of minimum coal partial oxidation temperature; maximum
temperature based on limiting S02 release to 10~2 mole % or sorbent melting.
bMinimum temperature based on avoiding steam condensation (300°C) or achieving minimum X^g of 0.03; maximum
temperature based on sorbent melting point or achieving XH s of 0.03.
°Based on thermodynamic equilibrium.
dRepresents the moles of steam supplied to the reactor per mole of H2S produced. For a 4 wt% sulfur coal
a N,
H20/NH2S
ratio of 6 represents about 1% of the power plant fuel energy required for the second step.
-------
criterion), and the maximum temperature limit is based on melting-point
limits or on the ash-softening point (assumed to be 1150°C). For the
two-step regeneration processes the minimum temperature of the first step
(sulfide production) is selected as the minimum to conduct in situ partial
oxidation of coal (/x-400°C). The minimum temperature for the second-
step reaction (300°C) is selected to avoid steam condensation. The maxi-
mum temperature limits are based on achieving the thermodynamic acceptance
criteria or avoiding sorbent melts.
The maximum SCL or E^S mole fractions are based on the lowest of
the thermodynamic, the material, and the energy balance limits. While the
carbon and steam consumptions are large for some of the sorbents, most
of this energy should be recoverable through proper system design. The
limiting factors will be capital investment and system complexity.
Some of the alternative sorbents are of limited potential, although
they do satisfy the acceptance criteria. For example, LiAlO- and LiFeO?
could produce only limited SO- concentrations by reductive decom-
position at 1000 kPa pressure and are applicable over a very limited
temperature range. Li«TiO« also has a limited applicable temperature
range for the atmospheric-pressure reductive decomposition.
The tabulated results provide an indication of the potential per-
formance of the alternative sorbents. The results also identify the tem-
perature ranges at which reaction kinetics should be investigated for the
most promising alternative sorbents.
Some other factors should be considered when the results in Tables 18,
19, and 20 are used. First, the thermodynamics involved with many of
the sorbents are speculative, and the original references should be
reviewed to note their limitations. Secondly, the assessment has been
consistently based on optimistic assumptions in order to project the
maximum possible performance. For example, secondary reactions of
several types could occur which could lead to deactivation of the
sorbent or reduce the level of H0S or S09 concentration in the regen-
£, £•
erator product.
55
-------
SECTION 8
AVAILABILITY
Availability data have been collected from standard references. ~
In general, sufficient commercial production of the simple metal oxides
exists for their availability to be assessed, but the double oxides are
available in such limited production that the availability of the raw
materials required to produce them must be considered along with the
feasibility of industrial expansion for their production. For example,
in 1960 about 3,000 Mg of BaTiO were produced at a cost of about
$16,000/Mg and in 1957 about 12,000 Mg of NaA102 were produced at about
$250/Mg. Neither of these production rates would support a viable
fluidized-bed combustion industry. The availability of sorbent support
materials is also considered.
DATA
The basic raw materials required to produce all of the candidate
sorbents are BaCOg, SrCO , TiO , Fe203, Al-Ck, Li2CO , Na CO , and
CaCO.. The following summary sheets (Tables 21 through 28) of avail-
ability facts relate to these materials. Cost information is also
included in the tables. The information in the tables does not provide
a perspective until a specific model of the sorbent requirements is
developed, but it is interesting to compare the annual production and US
resources of coal to these sorbent materials. In 1973 about 600 x 10 Mg
of coal were produced in the US, 65 percent of which was consumed by
the electric utility industry, 15 percent for coke production, and
15 percent for industrial use. The identified coal resources in the US
9
are about 1,600 x 10 Mg, and the ultimate reserves may be a factor of
two larger. The sorbent resources are insignificant compared to those
of coal (even for bauxite and iron ore), except for sodium carbonate
56
-------
TABLE 21. BaC03 STATISTICS
Applicable Sorbents - BaAl204, BaTi03, BaC03
Commercial Mineral Forms
Barite (BaS04)
Witherite (BaCO ) - almost entirely from England
Mineral Resources (identified)
US - 99 million Mg (US is self-sufficient)
World - 300 million Mg
Mineral Production (1975)
US - 1.1 million Mg Barite
Imports - 0.6 million Mg Barite
Distribution - Mined in eight states, 50% in Nevada; five firms produce
70% (poor distribution)
Commercial Uses - 80% in petroleum industry; remainder glass, paint,
rubber, barium compounds
Price (fob mine, 1975) - $15.00/Mg (Barite)
Compounds
BaC03 - production (1961) - 78,000 Mg; price (1972) - $150/Mg
(chemical grade)
BaTiO., - production (1960) - 3,000 Mg; price (1960) - $16,000/Mg
57
-------
TABLE 22. SrCC^ STATISTICS
Applicable Sorbents - SrC03, SrAI^O^, SrTi03
Commercial Mineral Forms
Celestite (SrSO^)
Strontianite (SrC03)
Mineral Resources (identified)
US - 1 million Mg strontium - not economically recoverable
World - very large, mainly celestite, Mexico and Canada chief
suppliers
Mineral Production (1975)
US - no domestic mining since 1959
World - 36,600 Mg
Imports - 22,000 Mg
Distribution - poor; 6 firms produce compounds in California, Georgia,
New Jersey, Ohio
Commercial Uses - 48% for color T.V. tubes
Price (1975) - $24/Mg of mineral, ^$50/Mg of Sr
Compounds
SrC03 - production (1972) - 15,000 Mg; price (fob, 1970)
- $380/Mg (technical grade), $700/Mg (pure)
SrS04 - price (fob, 1970) - $60/Mg
SrN03 - price (fob, 1970) - $240/Mg
58
-------
TABLE 23. Ti02 STATISTICS
Applicable Sorbents - Li2Ti03, SrTi02, BaTiO,
Commercial Mineral Forms
Rutile (Ti02)
Ilmenite (FeTi03) - 1.8-20% Ti02
Titanium slag
Mineral Resources (identified)
US - 500,000 Mg Rutile - dependent on foreign supply
100 million Mg Ilmenite
World - 13.7 million Mg Rutile
580 million Mg Ilmenite
Mineral Production (1975)
US - 11,000 Mg rutile; 640,000 Mg ilmenite
Imports - 216,000 Mg rutile; 450,000 Mg ilmenite
Distribution - rutile mined in Florida; 76% ilmenite mined in Florida
and New York; six firms produce concentrates in New York,
Florida, Georgia, and New Jersey
Commercial Uses - welding rod, pigment industries
Price (fob, 1975) - $710/Mg rutile, $55/Mg ilmenite
TABLE 24. Fe20 STATISTICS
I
Applicable Sorbents - LiFe02> NaC03 • ^2
Commercial Mineral Forms
Iron ore
Mineral Resources
US - 4 billion Mg recoverable iron
Mineral Production (1975)
US - 82 million Mg
Imports - 47 million Mg
Price (fob, 1975) - $19.00/Mg natural ore
59
-------
TABLE 25. Al STATISTICS
Applicable Sorbents - LiA102, NaA102, CaAL^, SrAI^O^, BaAl204, also
support material
Commercial Mineral Forms
Bauxite (35 - 65% Al^)
Mineral Resources (Bauxite)
US - 40 million Mg (inadequate long term supply)
World - 17 billion Mg
Mineral Production (Bauxite, 1975)
US - about 2 million Mg - dependent on foreign supply
Imports - about 12 million Mg
Distribution - Mined in Arkansas, Georgia, Alabama
Commercial Uses - 91% (1975) used to produce alumina; 88% of alumina
used to produce aluminum; alumina used for adsorbents,
catalysts, ceramics, abrasives, refractories;
0.15 million Mg (1963) of Si02-AL203 cracking
catalyst produced
Price (Bauxite, fob mine, 1975) - $10-20/Mg
AL.O- - production of all types (1975) - 0.2 million Mg; price
(1972) - $120-300/Mg
60
-------
TABLE 26. Li2CC>3 STATISTICS
Applicable Sorbents - LiAL02, LiFe02, Li2Ti03
Commercial Mineral Forms
Brine - as by-product
Spodumene - 1-2% Li.O (LiAlSi-OJ
2. / b
Mineral Resources (reserves)
US - 1.8 million Mg Li20 - self-sufficient
World - 2,9 million Mg L±20
Mineral Production
Not available
Exports (1975) - 800 Mg
Distribution - Mined and milled in N.C.; recovered from brines in Nevada
Commercial Uses - glass, ceramics, grease, air conditioning, storage
batteries, metallurgy
Price (fob, 1975) - $130/Mg spodumene concentrate
Compounds - L^COg
Consumption (1960) - 1,200 Mg; distribution - plants in PA. and
TE.; capacity (1975) - 27,000 Mg/yr; Price (1975) - $l,600/Mg
(chem. grade)
61
-------
TABLE 27. ®*C0 STATISTICS
Applicable Sorbents - Na2C03, NaA102, NaC03 • Fe203
Commercial Mineral Forms
Natural soda ash
Salt cake (Na2SC>A)
Sodium chloride (brine)
Mineral Resources
US - 50 million Mg soda ash (53-63% Na2C03) >400 million Mg salt
cake
Mineral Production (1975)
Natural soda ash - 4 million Mg
Salt cake - 665,000 Mg
Distribution - four natural soda ash producers in Wyoming and
California; three natural salt cake producers in
California, Texas, and Utah
Commercial Uses (Na7CO_) - 47% glass, 25% chemicals; pulp and paper
Price (fob, 1975) - $54-64/Mg natural soda ash
- $45/Mg salt cake
Compounds
Na_CO_ - 9 firms producing from natural soda or brine by Solvay
process
NaAL02 - production (1957) - 11,000 Mg; price (1957) - $200-260/Mg
62
-------
TABLE 28. CaCO STATISTICS
Applicable Sorbents - CaCO^, CaAL20,
Commercial Mineral Forms
Limestone - CaCOo
Dolomite - CaCOg • MgC03
Gypsum - CaSO»
Mineral Resources
Not estimated; projected to be sufficient for thousands of years at
present consumption
Mineral Production (1972)
>500 million Mg Limestone
>60 million Mg dolomite
Distribution - resources in every state; widely distributed by many
companies
Price (fob, 1975) - $2-20/Mg limestone or dolomite
63
-------
and limestone/dolomite. Sodium carbonate represents about three percent
of the resources of coal,and limestone apparently far exceeds coal
resources. The annual production rates of all the sorbent materials
except limestone are very small compared to those of coal; limestone
rates are comparable.
The point of view developed for the availability and suitability of
support materials for the production of alternative sorbents is that:
(1) there are a number of ceramic raw materials occurring in nature
(clays, silicas, feldspar, etc.) that would not be expected to be suit-
able support materials because of their limited availability and poor
attrition resistance; (2) the only support material that may possibly
have sufficient availability and attrition resistance would be a
specially prepared alumina structure.
SORBENT DEMAND MODEL
The demand for a sorbent material may be projected from the
relationship
M=eM C [(1.05)n-l] [f + 0.0476] ,
o o
where M is the annual mass of sorbent required in a year n years after
1990, e is the fraction of the new coal-fired power generation market
captured by fluidized-bed combustion annually, M is the initial mass of
sorbent required for a new fluid-bed combustion plant (Mg per MWe), C
is the US power demand (electric utility coal-fired) in the year 1990,
the year in which fluidized-bed combustion is assumed to become a viable
commercial industry, and f is the fraction of the initial plant sorbent that
must be replaced annually in order to make up for losses due to attrition,
deactivation, elutriation, agglomeration and handling, and so on. The
factor (1.05) accounts for an assumed 5 percent growth in power gener-
ation demand per year. The term M is a function of the plant design,
the sorbent density, and the fraction of support material comprising the
total sorbent structure.
64
-------
In order to project the sorbent demand, or to project acceptable
rates of loss of the alternative sorbents due to all causes, the follow-
ing assumptions are made:
• The sorbent structure consists of 30 weight percent
active sorbent and 70 weight percent alumina support;
• The combustor volume is identical to the volume for
O
the limestone/dolomite combustor at about 1060 dm
3
(30,000 ft ) per 1000 MWe power plant for all of the
sorbents.
• The coal-fired capacity in 1990 is 5 x 105 MWe (C ),
and fluid-bed combustion captures 25 percent of the
new market annually (e).
The results in Table 29 are obtained with these assumptions. The
sorbent loss rate is expressed as a percentage of the bed inventory
lost per hour (a function of f), which leads to a sorbent demand, M (of
the limiting component) in the year 2000 (10 years after the assumed
commercial start) equivalent to the 1975 production level. The selec-
tion of the year 2000 and the production level in 1975 are arbitrary,
but the results do provide a perspective on the tolerable loss rates.
For comparison, the normal attrition rate for limestone and dolomite
in a fluid-bed combustor is on the order of 1 percent per hour, while
total deactivation plus attrition losses are about 10 percent per hour.
The analysis does not provide a clear breakpoint in the loss rate
of sorbent that separates acceptable from unacceptable behavior, but
it is evident that a great increase in demand for any material, such as
is expressed in Table 29, would require a large obligation by industry
to expand and invest capital. Certainly,an alternative sorbent should
not be developed if it cannot support an industry. The sorbents other
than Na2C03, CaO/CaC03, NaA102» NaCC>3 • Fe^, and CaAl^ appear to require
extremely low loss rates when all of the possible sources for sorbent
losses are considered: combustor and regenerator attrition; cyclone
attrition; pneumatic transport attrition; deactivation due to sintering
and poisoning; and handling, maintenance, and preparation losses.
65
-------
TABLE 29. SORBENT AVAILABILITY COMPARISON
a
Sorbent
(Form Stable
in Combustor)
Na.CO.
2 3
CaO/CaCO,,
3
SrC03
BaC03
LiA100
2
LiFeO.
2
Li2Ti°3
NaA100
2
NaC03 • Fe203
CaAljO,
SrAl00.
2 4
SrTi03
BaAl.O,
2 4
BaTi03
Limiting
Component
A100,
2 3
A190,
2 3
SrC03
BaC03
Li0C00
2 3
Li0CO.
2 3
Li2C°3
A100,
2 3
A12°3
A12°3
SrC00
3
SrC03
BaC00
3
BaC03
Percent of the Combustor
Inventory Lost per Hour
(in 2000) to Equal the
1975 US Production Rate
of Limiting Component^3 »c
3.64
2.77
0.011
0.054
0.044
0.113
0.033
2.41
3.07
1.94
0.015
0.010
0.070
0.064
Based on 30 wt% sorbent in an alumina support
Based on the following assumptions: (1) Fluid-bed combustion com-
mercial in 1990; (2) Coal-fired electric utility generation of
2 x 105 MWe in 1990; (3) 5% annual increase in coal-fired power
generation; (4) fluid-bed combustion captures 25% of the new
coal-fired power generation capacity.
t
"See Tables 21-28 for US and world 1975 resource estimates.
66
-------
Even the projections for the five sorbents that have apparently high
availability may be overly optimistic. The limiting component for each
of these sorbents is the alumina support material. Table 29 is based on
all forms of alumina currently produced; but (from Table 25) 88 percent
of this current production is presently utilized in the production of
aluminum. Thus, the increase in alumina production for the catalyst-
type uses assumed here would be in reality much larger, or the tolerable
loss rate should be smaller by a large factor (about 10).
67
-------
SECTION 9
ECONOMICS
Ultimately, the alternative sorbent must provide an economic advan-
tage over the limestone/dolomite-based fluidized-bed combustion system
to be considered for development. The total cost of the sorbent consists
of contributions from the sorbent material (and support) cost and the
capital investment associated with the power plant, both of which are
influenced by the sorbent behavior. Both of these contributions can be
estimated, and the tolerable rate of sorbent losses can be projected.
SORBENT MATERIAL COST
The cost associated with the direct consumption of the sorbent con-
sists of the raw materials cost, the delivery cost of the raw materials
and the cost of preparation of the sorbent particles. The delivery cost
of some of the raw materials could be high because of the poor distribu-
tion of producers in the country. Preparation could also be very expen-
sive. Various preparation techniques could be used. For example, the
raw materials could be transformed first into the desired metal oxide
form, and then the metal oxide could be fixed to the support structure by
slurrying or impregnation, followed by drying and sizing of the sorbent
material. It is assumed that a separate industry would be initiated to
supply the sorbent needs of the utilities rather than on-site preparation
of the sorbent.
The costs of the sorbent raw materials are estimated in Table 30 ;
for an alumina support costing $200/Mg and for the sorbent particles con-
taining 30 weight percent sorbent material. Availability limitations are
neglected, and present day sorbent material costs are assumed; increased
demand would increase these costs. A minor factor has been included
68
-------
TABLE 30. SORBENT RAW MATERIALS COSTS
Sorbent
Na2C°3
CaO/CaCO_
SrC03
BaC03
LiA102
LiFe02
Li2Ti°3
NaA102
Na2C03 - Fe203
CaAl204
SrAl204
SrTiCL
3
BaAl204
BaTi03
Components (Cost
Na2C03 (85)
CaC03 (10)
.. SrC03 (450)
BaC03 (180)
Li2C03 (1,800), Al
Li2C03 (1,800), Fe
Li2C03 (1,800), Ti
Na2C03 (85), A1203
Na2C03 (85), Fe203
CaCO (10), Al 0-
O ^ j
SrC03 (450), A1203
SrC00 (450), Ti00
3 2
BaCO (180), Al 0.
j £• J
BaC03 (180), Ti02
$/Mg)
203 (200)
203 (67)
02 (800)
(200)
(67)
(200)
(200)
(800)
(200)
(800)
Sorbent Raw Materials
Cost ($/Mg)a
166
143
275
194
489
270
678
194
159
181
267
353
206
268
Q
Cost per Mg of sorbent plus support based on 30 wt% sorbent in an
alumina support. The support alumina material is assumed to cost
$200/Mg. Does not include raw material losses during preparation
and handling; 1975 cost basis; assumes delivery cost, is minimal.
69
-------
for delivery but is not expected to be representative of actual delivery
costs except under ideal conditions. Preparation costs may range up
to $2000/Mg, but because there is presently no basis for predicting the
preparation cost, its influence will be observed parametrically.
MAXIMUM ACCEPTABLE TOTAL SORBENT COST
It is assumed that the total cost of power from the fluidized-bed
combustion power plant with an alternative sorbent must not exceed the
cost of power from the limestone/dolomite-based power plant. The
limestone/dolomite-based system could utilize either the once-through
sorbent concept or the reductive decomposition regeneration concept.
Economic comparison will be made for both of these concepts. The maxi-
mum acceptable sorbent cost will be calculated in order to be competa-
tive with limestone/dolomite-based sorbents, and this cost will be
translated into a maximum acceptable sorbent loss rate due to attrition,
deactivation, and so on.
An alternative sorbent may influence the design and cost of many
components in the power plant (e.g., sorbent storage, feeding, com-
bustor design, power generation equipment, particulate control, etc.),
but for this analysis it is assumed that all plant components are
identical in cost when comparing the alternative sorbents with the
limestone/dolomite sorbent, except for the cost of the regeneration
system. The regeneration system consists of the regenerator components,
the sorbent circulation components, the sulfur recovery process, the
C09 recovery process (if applicable), and required auxiliary components.
Spent sorbent processing equipment is not included. Thus, the following
discussion will consider only those components of the total plant.
The maximum performance in terms of S02 or H_S mole fractions from
the regenerator have been projected in Tables 18, 19, and 20. Because
this is the most critical cost-determining factor in the regenerator
system, the capital cost corresponding to this maximum performance
represents the minimum possible investment. The capital investment
for sulfur recovery processes as a function of the fraction S0« or
70
-------
H S has been estimated. Previous cost assessments of regeneration
indicate that the total capital investment for regeneration may be
related directly to the sulfur recovery investment. It is assumed that
the total regeneration system investment is equal to 1.47 times the
sulfur recovery investment for the reductive decomposition regeneration
and 2.0 times the sulfur recovery investment, plus the CCL recovery system
investment for the two-step regeneration processes.
For the comparison it is assumed that the limestone/dolomite-based
reductive decomposition regeneration system investment is $29/kW for
the atmospheric-pressure system and $61/kW for the pressurized system
(based on a 600 MWe power plant). The base limestone/dolomite reductive
decomposition regeneration energy costs are estimated to be 2.55 mills/kWh
for the atmospheric-pressure system and 4.00 mills/kWh for the pressurized
system. The limestone/dolomite-based once-through energy cost is
assumed to be 2.0 mills/kWh for the atmospheric-pressure system (based on
a calcium-(limestone)/sulfur mole ratio of 3) and 1.3 mills/kWh for
the pressurized system (based on a calcium (dolomite)/sulfur mole ratio
of 1.2).
The energy cost associated with the regeneration system consists of
contributions from five factors: capitalization of the investment, oper-
ating and maintenance cost, fuel cost, power plus other auxiliary items,
and the sorbent cost. Based on a 20 percent capitalization plus operat-
ing and maintenance factor, a 70 percent plant capacity factor, a
600 MWe power plant producing 200 Mg of sulfur per day and a coal cost of
$20/Mg, the following energy cost relationship is developed for the reduc-
tive decomposition regeneration:
Minimum Energy Cost (mills/kWh)
= 0.0326 (I) + (N_/N_rt ) 0.1225 + 0.4 + S ,
U d(Jj
where I is the minimum regeneration system capital investment in $/kW,
N /Nso is the minimum coal rate factor in Table 18, and S is the sorbent
cost in mills/kWh (sorbent price times sorbent consumption rate).
71
-------
For the two-step regeneration processes
Minimum Energy Cost (mills/kWh)
= 0.0326 (I) + CN_ n/Nfl ) 0.0174 + 0.8 + S ,
2 ?
where Ng o/^H S *s the steam consumption factor in Tables 19 and 20.
Applying all of the assumptions stated provides a means to estimate
the maximum acceptable sorbent cost (S) for comparison with either the
regenerative base system or the once-through base system. The results
from this comparison are presented in Tables 31, 32, and 33 for the reduc-
tive decomposition and the two-step regeneration processes. Relative con-
clusions may be drawn from the three tables. The results indicate
that there is very little economic incentive to develop alternative
sorbents for pressurized fluidized-bed combustion if the basis of com-
parison is the pressurized calcium-based once-through performance (i.e.,
the maximum permissible alternative sorbent costs in column four of
Tables 31, 32, and 33 are near zero or, in some cases, less than zero.
On the other hand, the economic incentive to develop alternative
sorbents is highest for the pressurized systems if the basis of comparison
is the pressurized calcium-based regeneration performance (i.e., the per-
missible sorbent costs are the greatest in column three of Table 31 and
columns 1 and 2 of Table 33). The economic incentive to develop
alternative sorbents for the atmospheric-pressure system is lower
than for the pressurized system, comparing columns 1 and 3 in Table 31
(i.e., because more expensive sorbent materials may be used in the pres-
surized systems than in the atmospheric-pressure systems, more opportunity
for economic advantage is available in the pressurized regeneration
systems).
72
-------
TABLE 31. MAXIMUM ACCEPTABLE ALTERNATIVE SORBENT COST
FOR REDUCTIVE DECOMPOSITION REGENERATION(S), mills kWha
Sorbent
AFBC
Comparison with
Base ,
Regeneration
Base Once-
through0
PFBC
Comparison with
Base ,
Regeneration
Base Once-
through6
Na2C03
CaO/CaCO.
SrC03
LiAl02
LiFe02
Li0Ti00
2 3
NaA102
NaC03 • Fe203
CaAl00,
2 4
SrAl_0,
2 4
SrTi03
BaAl00.
2 4
BaTi03
0.98
1.43
1.43
1.43
1.43
0.83
1.43
0.98
1.43
1.43
1.43
1.43
1.43
0.43
0.88
0.88
0.88
0.88
0.28
0.88
0.43
0.88
0.88
0.88
0.88
0.88
—
2.71
—
1.68
1.88
-
—
—
2.88
2.88
2.88
2.88
2.88
—
0.01
—
<0
<0
,-,
—
—
0.18
0.18
0.18
0.18
0.18
aThe figures given in the table are the alternative sorbent cost, S
(expressed in mills/kWh) which would be necessary for the alternative
sorbent reductive decomposition regeneration system to have a cost
equal to that of the base limestone/dolomite regeneration and once-
through cases indicated. S could be expressed as a product of the
sorbent cost in $/Mg and the sorbent consumption rate in Mg/kWh.
Base regeneration is atmospheric pressure limestone/dolomite reduc-
tive decomposition.
Base once-through is an atmospheric pressure limestone/dolomite
once-through system.
Base regeneration is a pressurized limestone/dolomite reductive
decomposition.
SBase once-through is a pressurized limestone/dolomite once-through
system.
73
-------
TABLE 32. MAXIMUM ACCEPTABLE ALTERNATIVE SORBENT COST FOR
TWO-STEP ATMOSPHERIC-PRESSURE
REGENERATION(S), mills/kWha
Sorbent
Na2C03
CaO/CaC03
SrC03
BaC03
LiA102
LiFe02
Li2Ti03
NaA102
NaC03 • Fe203
CaAl204
SrAl204
SrTi03
BaAl204
BaTi03
Two-Step Regeneration AFBC
With Steam
Comparison with
Base
Regeneration"3
__
—
—
—
1.31
0.45
1.21
1.12
0.23
0.81
1.17
1.31
1.31
1.31
Base Once-
through0
__
—
—
—
0.76
<0
0.66
0.57
<0
0.26
0.62
0.76
0.76
0.76
With Steam/C02
Comparison with
Base
Regeneration"3
0.85
0.62
0.68
0.75
—
—
—
0.68
0.75
0.75
0.75
—
—
—
Base Once-
through0
0.30
0.07
0.13
0.20
—
—
—
0.13
0.20
0.20
0.20
—
—
—
The figures given in the table are the alternative sorbent cost,
S (expressed in mills/kWh) , regeneration schemes at atmospheric pressure
to have the same energy cost as the base limestone/dolomite regeneration
and once-through cases indicated.
Base regeneration is atmospheric-pressure limestone/dolomite reduc-
tive decomposition.
"Base once-through is a atmospheric-pressure limestone/dolomite once-
through system.
74
-------
TABLE 33. MAXIMUM ACCEPTABLE ALTERNATIVE SORBENT COST FOR TWO-STEP
PRESSURIZED REGENERATION (S), mills kWha
Sorbent
Two-Step Regeneration PFBC
With Steam
Base
Regeneration*"
Base Once-
through0
With Steam/C02
Base ;
Regeneration^ •'
Base Once-
through0
CaO/CaCO,
SrCO,,
2.20
2.20
<0
<0
<0
Da.i^u-
LiA102
LiFe02
Li2Ti°3
NaA102
cal2o4 &2
SrAl204
SrTi03
BaAl204
BaTi03
2.76
1.90
2.73
~™
2.26
2.57
2.76
2.76
2.76
— — f. . j\j
0.06
-------
MAXIMUM ACCEPTABLE SORBENT LOSS RATE
The estimates of maximum acceptable sorbent cost given in Tables 31,
32, and 33 may be translated into maximum acceptable sorbent loss rates
as a function of the sorbent preparation cost. The following relationship
applies:
"~ 0.3 + p /2.8
X =
s
0.0163
CM+C -0.143
M P / pg x 10~3
where X is the maximum acceptable percentage of the combustor inventory
lost per hour; S is the maximum acceptable sorbent cost in Tables 31, 32,
and 33; (L. is the sorbent raw material cost estimated in Table 30; C
is the sorbent preparation cost in $/Mg; p is the sorbent material
S
density; and p is the support material density. It is assumed in this
relationship that the sorbent structure is comprised of 30 weight percent
sorbent and 70 percent support, the particle is 50 percent void, the bed
is 50 percent void, the support material density is about 4 g/cc
(250 lb/ft3), the bed volume is 1060 dm3 (30,000 ft3) per 1000 MWe
plant, the capitalization factor on the original sorbent purchase is 0.1,
and a 0.7 plant capacity factor is in effect.
Evaluation of the cost relationship for the value of S given in
Tables 31, 32, and 33 under the calcium-based reductive decomposition
regeneration base case results in the maximum loss rates reported in
Tables 34, 35, and 36. Loss rates have not been determined on the
basis of comparison with the once-through limestone/dolomite system
because in most cases the resulting loss rates would be inconceivably
small (<0.01 % per hour). The tables state that if (1) the sorbent
performs at its maximum possible level of performance and (2) if the
availability of the sorbent is not limiting so that the present sorbent
costs may be assumed, then the rate of loss in the tables for a given
sorbent preparation cost will result in an energy cost for the power
plant identical with the limestone/dolomite-based power plant.
76
-------
TABLE 34. MAXIMUM SOKBENT LOSS RATES FOR
REDUCTIVE DECOMPOSITION REGENERATION
Sorbent
Na2C03
CaO/CaCO.
SrC03
LiA102
LiFe02
Li2Ti03
NaA102
NaC00 • Fe000
3 23
CaAl204
SrAl204
SrTi03
BaAl20,
BaTi03
System
AFBC
AFBC
PFBC
AFBC
AFBC
PFBC
AFBC
PFBC
AFBC
AFBC
AFBC
AFBC
PFBC
AFBC
PFBC
AFBC
PFBC
AFBC
PFBC
AFBC
PFBC
Maximum Acceptable Percentage
of
Combustor Bed Loss
Per Houra
Sorbent Preparation
0
0.804
1.234
2.339
0.618
0.372
0.437
0.856
1.125
0.148
0.941
0.787
0.941
1.894
0.637
1.283
0.442
0.890
0.807
1.625
0.619
1.247
Cost, $/Mg
500
0.199
0.272
0.515
0.218
0.183
0.215
0.235
0.308
0.085
0.261
0.187
0.248
0.499
0.220
0.443
0.181
0.365
0.233
0.469
0.215
0.434
1000
0.113
0.152
0.287
0.132
0.121
0.142
0.142
0.186
0.057
0.152
0.106
0.142
0.285
0.132
0.266
0.114
0.230
0.135
0.272
0.129
0.259
aBased on sorbent raw materials costs in Table 30 and the maximum
acceptable sorbent costs in Table 31 using the columns reflecting
the base regeneration system comparison.
77
-------
TABLE 35. MAXIMUM SORBENT LOSS RATES FOR TWO-STEP
REGENERATIONS AFBC
Maxim
Coml
Reactant Gas 0
Na9CCL H.0/C09 0.698
<£ j 2. jL
CaO/CaC03 H20/C02 0.535
SrC03 H20/C02 0.293
BaCO, H_0/C00 0.424
3 22
LiA102 H20 0.341
LiFeO- HO 0.269
£» £
Li2Ti03 H20 0.217
NaA102 H20 0.737
H 0/C02 0.447
NaC03 • Fe203 H&0 0.184
H20/C02 0.603
CaAl20^ H20 0.533
H 0/C02 0.494
SrAl204 H20 0.522
H20/C02 0.334
SrTi03 H20 0.404
BaAl20^ H20 0.738
BaTi03 H20 0.567
urn Acceptable Percentage of
bustor Bed Loss Per Houra
Sorbent Preparation
Cost, $/Mg
! 500 1000
0.173 0.098
0.117 0.065
0.104 0.064
0.122 0.072
0.168 0.111
0.073 0.044
0.124 0.086
0.204 0.119
0.124 0.072
0.044 0.024
0.143 0.082
0.140 0.080
0.130 0.075
0.179 0.108
0.116 0.068
0.166 0.104
0.214 0.117
0.197 0.117
Based on sorbent raw material costs in Table 30 and the maximum
acceptable sorbent costs in Table 32 using the columns reflecting
the base regeneration system comparison.
78
-------
TABLE 36. MAXIMUM SORBENT LOSS RATES FOR TWO-STEP
REGENERATIONS PFBC
Maximum Acceptable Percentage of
of Combustor Bed Loss per Houra
^nTnPnt* Rr*f>r*n PT T i~ i on -- . --
Reactant Gas 0
CaO/CaCO. H00/C00 1.900
J z 2
SrC03 H20/C02 - 0.950
BaCO. H00/C0^ 1.299
3 22
LiA102 H20 ' 0.717
LiFe02 H20 1.138
Li2Ti03 H20 0.487
CaAl204 H20 1.487
H 0/C02 1.513
2 A 2
H20/C02 1.025
SrTi03 H20 0.853
BaAl-0, H_0 1.557
BaTi03 H20 1.195
Sorbent Preparation
Cost, $/Mg
500 1000
0.419 0.233
0.336 0.204
0.375 0.220
0.352 0.233
0.311 0.189
0.279 0.192
0.391 0.223
0.398 0.228
0.396 0.238
0.354 0.212
0.349 0.220
0.450 0.261
0.416 0.248
Based on sorbent raw materials costs in Table 30 and the maximum
acceptable sorbent costs in Table 33, using the columns reflecting
the base regeneration system comparison.
79
-------
If the sorbents performed at their maximum level and cost nothing
to prepare, some of them could operate with maximum loss rates similar
to the attrition loss rates of limestones and dolomites (1 to 5 percent
per hour). Total sorbent loss rates for regenerative limestone/dolomite
systems due to attrition, deactivation, and so forth, will be about
10 percent per hour or greater. With expected preparation costs and
system performance, however, it would be expected that the loss rates
must be at least an order of magnitude less than the loss rates of lime-
stone and dolomite due to attrition alone in order to be economically
competitive and even lower to provide sufficient economic incentive to
develop the alternative sorbent. Experimentation will be required to
demonstrate such low loss rates. Again, the pressurized systems
appear to provide the greatest opportunity for an alternative sorbent.
If the comparison with the base once-through system had been made
from Tables 31, 32, and 33, it is doubtful that any alternative sorbent
could compete economically with the limestone/dolomite-based once-through
system unless environmental constraints on spent sorbent place additional
economic requirements on the once-through systems.
The projections presented in the tables not only provide a perspective
or the sorbent loss requirements but also can act as a guide to experi-
mental attrition/deactivation studies and show the relative potential
of the candidate sorbents.
80
-------
SECTION 10
ASSESSMENT OF ALTERNATIVE SORBENTS
Fourteen alternative sorbents have been identified in Table 18 as
possibly being suitable for atmospheric-pressure fluidized-bed combustion:
, CaO/CaC03, SrC03, BaC03> LiA102> 2> 2 ,
NaC03 • Fe203, CaAl^, SrAl^, SrTiOg, BaAl^ and BaTi03. These
sorbents are reported in the form thermodynamically stable under the
combustor conditions. Eleven sorbents are possibly suitable for
pressurized fluidized-bed combustion (Tables 19 and 20): CaO/CaCO ,
Sr2C03, BaC03, LiA102, LiFe02, Li2Ti03, CaAl^, SrAl^, SrTiOg,
BaAl^O, and BaTiO~. Suitable regeneration processes for each of these
sorbents have been identified and applicable temperature ranges and
maximum performance levels have been projected.
Because sorbent losses by attrition and deactivation at a number of
process locations are always to be expected, it appears that the avail-
ability of the active sorbent material or the support material will be
the most important factor in justifying the development of fluidized-
bed combustion with alternative sorbents. None of the alternative sor-
bent materials can compare with limes tone/ dolomite in availability and
wide distribution of producers.
If maximum performance levels could be achieved, it is expected that
sorbent loss rates must be less than 0.1 percent of bed content per hour
in order to compete economically with limestone/dolomite-based fluidized-
bed combustion (reductive decomposition regeneration) . The availability
limitations on the alternative sorbents may be more stringent than the
economic .
Further development is required in attrition behavior, kinetics,
and engineering studies concerning sorbent preparation costs, sorbent
81
-------
availability, the impact of design modifications that reduce attrition
upon fluidized-bed combustion economics, and the environmental impact
of alternative sorbents.
Attrition behavior under simulated process conditions should be
characterized in order to select promising support materials, although
availability may determine the support material and the preparation of
the sorbent may become the issue.
The kinetics of desulfurization and regeneration should be charac-
terized within the applicable range of operating conditions (temperature,
pressure, fuel, etc.) and under conditions which simulate the maximum
performance levels. Unless the high levels of SO- or H^S can be
achieved in the regeneration process, the alternative sorbents will
become economically unfeasible. The degree of sorbent utilization and
regeneration appears to be a relatively unimportant economic factor.
At present, too little is known about the cost of sorbent prepara-
tion. Feasible process technology for sorbent preparation should be pro-
posed and the economic feasibility assessed.
The availability of the sorbent materials and the support materials
may require more modeling in order to assess all of the market factors
properly. The area of industrial expansion into alternative sorbent prep-
aration and distribution must also be explored.
Alternative sorbents may require alternative fluidized-bed combustion
design concepts to satisfy their limitations and promote their advantages.
The impact of alternative designs that would reduce sorbent attrition/
deactivation should be evaluated.
Finally, the environmental impact of alternative sorbents (e.g.,
trace metals released to the environment during combustion/regeneration,
or upon disposal of deactivated sorbent) may lead to elimination of
certain candidate materials. The complex sociological interactions of
these materials must be explored before development is committed.
82
-------
9. High Temperature Properties and Decomposition of Inorganic Salts,
Part 2. Carbonates, U.S. Dept. of Commerce, Nat. Bureau of Standards,
Nov. 1969.
10. Mineral Facts and Problems, Bureau of Mines, Dept. of the Interior,
1970.
11. United States Mineral Resources, Geological Survey Professional
Paper 820, U.S. Govt. Printing Office, Wash., DC 1973.
12. Commodity Data Summaries, 1976, Bureau of Mines, U.S. Dept. of the
Interior.
13. Minerals Yearbook, 1972 and 1973, Vol I, Bureau of Mines, Wash.,
D.C., 1974 and 1975.
14. Encyclopedia of Chemical Technology, Second Edition, Kirk-Othmer,
New York; Interscience Publishers, 1963.
15. O'Donnell, J. J., A. G. Sliger, "Availability of Limestones and
Dolomites," Presented at the Second International Lime/Limestone
Wet Scrubbing Symp., New Orleans, Nov. 1971.
16. Archer, D. H., et al., Evaluation of the Fluidized-Bed Combustion
Process, Vols. I-III, Report to EPA, Westinghouse Research and
Development Center, Pittsburgh, PA, November 1971, Contract 70-9,
NTIS Numbers PB 211-494, 212-960, 213-152.
17. Keairns, D. L., et al., Pressurized Fluidized-Bed Combustion Process
Development & Evaluation, Vols. I and II, Pressurized Fluidized-
Bed Boiler Development Plant Design, Vol. Ill, Report to EPA,
Westinghouse Research and Development Center, Pittsburgh, PA,
December 1973, EPA-650/2-73-048 a, b and c, NTIS Numbers 231-162,
231-163, 232-433.
18. Keairns, D. L., et al., Pressurized Fluidized-Bed Coal Combustion
Development, Report to EPA, Westinghouse Research and Development
Center, Pittsburgh, PA, September 1975, EPA-650/2-75-027C, NTIS
Number PB 246-116.
83a
-------
SECTION 11
REFERENCES
1. "Fundamental Study of Sulfur Fixation by Lime and Magnesia,"
submitted by Battelle Memorial Institute to Public Health Service,
HEW, Contract No. PH86-66-108, June 30, 1966 (NTIS No. PB 176843).
2. "A Survey of Metal Oxides as Sorbents for Oxides of Sulfur," sub-
mitted by AVCO Corporation, Contract No. 86-67-51, Peb.1969 (NTIS
No. PB 105190).
3. Borgwardt, R. H., D. C. Drehmel, T. A. Kittleman, D. R. Mayfield, and
J. S. Bowen, "Alkaline Additives for Sulfur Dioxide Control," Annual
Report, EPA, March 29, 1971.
4. Brancaccio, J., C. V. Flach, "Use of Dry Fuel Additives to Reduce
SO Emissions from a Full Size Industrial Coal Fired Boiler," Pre-
sented 65th Annual Meeting of the Air Pollution Control Association,
Miami Beach, Florida, June 1972.
5. "Applicability of Metal Oxides to the Development of New Processes
for Removing S09 from Flue Gases," Final Report, Tracer, Contract
PH 86-68-68, July 31, 1969.
6. "Evaluation of Reactive Solids for SO- Removal During Fluidized-
Bed Coal Combustion," Catalytica Associates, Inc., EPRI TPS75-603,
Oct. 1975.
7. "Identification of Regenerable Metal Oxide S02 Sorbents for Fluidized-
Bed Coal Combustion," Radian Corporation, EPA-650/2-75-065, July 1975.
8. Handbook of Chemistry and Physics, 52nd ed., The Chemical Rubber
Co., Cleveland, Ohio, 1971.
83
-------
APPENDIX A
PROGRAM PLAN
Preliminary screening of alternative sorbents for fluidized-bed com-
bustion has identified 14 candidates. Rather than attempt to evaluate
all of these sorbents within the limited resources of the program, three
of the most promising materials will be evaluated experimentally and
economically.
The following tasks will be performed with the selected sorbents:
1. Evaluation of availability, cost, and properties of selected
sorbents and support materials through contacts with industrial
suppliers.
2. Evaluation of alternative sorbent commercial preparation
methods and preparation costs.
3. Preparation of the selected sorbents, using commercial prepara-
tion techniques, and evaluation of the prepared sorbent physical
and chemical properties.
4. Collection of data on desulfurization kinetics and regeneration
kinetics under high-performance conditions. Testing of sorbent
attrition at process conditions.
5. Assessment of plant performance and cost with the selected
sorbents.
6. Proposal of development plan for the best alternative sorbent
if it is commercially advantageous compared to the naturally
occurring calcium-based sorbent materials.
SELECTION OF SUPPORT MATERIAL
On the basis of availability considerations the support material
will be alumina. The prepared alumina properties to be utilized must
84
-------
provide a tradeoff between low cost, high attrition resistance, high
sorbent kinetics promoted by large pore diameters and the amount of
active sorbent that can be loaded. Alumina suppliers will be contacted.
SELECTION OF SORBENT MATERIALS
The 14 sorbents from which three materials have been selected are:
Na2C03, CaO, SrC03> BaC03, LiA102, LiFe02> Li2Ti03> NaA102, NaC03 - Fe^,
CaAl204, SrAl204, SrTi03> BaAl^ and BaTiOg. The following table
(Table Al) summarizes the technical potential, the cost and availability
potential and the selection logic for these materials. The evaluation
of the three selected alternative sorbents BaCO CaAl 0, and BaTiO ,
will provide information about other sorbent materials by induction.
SORBENT TESTING
Three critical test areas will be considered: attrition testing,
desulfurization kinetics, and regeneration kinetics. The following general
description of these tests provides an indication of their scope.
Attrition Tests
• Basic parameters. Temperature, pressure, velocity, gaseous
environment, number of regeneration cycles, sorbent material
structure, and sorbent preparation technique.
•• Description. Samples will be fluidized under desulfurizer con-
ditions and rates of fines generation will be measured. Both
fresh and cycled samples will be tested.
• Criteria. Observed loss rates must be less than the economic
limits previously determined for each sorbent.
Desulfurization Tests
• Basic parameters. Temperature, pressure, gaseous environment,
number of cycles, sorbent composition, support structure, and
particle size.
85
-------
TABLE Al. ALTERNATIVE SORBENT SELECTION
Sorbent
Desulfurizer
Potential
Atmospheric
Na-CO, Limited
temp, range
CaO Good
Pressurized
Reductive Decomposition
Potential
Atmospheric
No Limited
so,
Good Good
Pressurized
Z-Step Regeneration
Potential
Atmospheric
No Good
Limited Limited
Pressurized
Availability
Potential
Cost
Potential
Selection Logic
No Good — Drop - Limited applicability;
highly corrosive nature
Good Good Good Drop - Good potential, but
CO
SrCO,
Good
Good
No
S0
No
H2S
Good
Good
Poor
BaC03
L1A102
LiFeO,
L
Li2T103
NaAlO-
NaCO, • Fe,0,
3 23
CaAljO^
SrAl20A
SrTi03
BaAl.O.
Good
Good
Good
Good
Limited
temp, range
Limited
temp, range
Good
Good
Good
Good
Good
Good
Good
Good
No
No
Good
Good
Good
Good
No
Good
Good
Limited
Good
Limited
SO,
Good
Good
Good
Good
No
Limited
so2
Limited
SO,
No
No
No
Good
Good
Good
Good
Good
Good
Limited
H2S
Good
Good
Good
Good
Good
Good
Good
Good
Good
Limited
HZS
Good
No
No
Good
Good
Good
Good
—
—
—
—
Good
Good
Good
Poor
Poor
—
similarity with
existing data from
Argonne
Drop - Poor availability
Select - Reasonable potential
Poor Drop - Poor cost potential
Poor Drop - Poor cost potential due
to limited performance
Poor Drop - Poor cost potential
Drop - Limited applicability;
highly corrosive nature
— Drop - Limited applicability;
highly corrosive nature
Good Select - Reasonable potential
Drop - Poor availability
— Drop - Poor availability
Drop - Reasonable potential,
but similarity with
BaTiO,
Good
Good
Good
Good
Good
Good
Select - Reasonable potential
-------
• Description. Sorbent desulfurization kinetics will be determined
on TGA for fresh and cycled material, up to 20 cycles (if fresh
material meets criteria). High performance conditions will be
utilized (low SO^ approaching the required level).
• Criteria. Kinetics should be within the same order of magnitude
as the kinetics of limestone/dolomite sorbents. Deactivation on
cycling must be limited to levels less than permissible sorbent
attrition loss rates.
Regeneration Tests
• Basic parameters. Reaction scheme (reductive decomposition, two-
step process with steam or steam and CCO, temperature, pressure,
gaseous environment, number of cycles, sorbent composition, sup-
port structure, and particle size.
• Description. Sorbent regeneration kinetics will be determined
on TGA for fresh and cycled material (if fresh material meets
criteria). High performance conditions will be utilized (high
S02 or H2S).
• Criteria. Kinetics should be within the same order of magnitude
as the kinetics of limestone/dolomite sorbents. No criterion
has been developed for the extent of sorbent regeneration (change
in utilization).
Specific sorbent test conditions for BaCCL, CaAl20,, and BaTiCL are
listed in Tables A2, A3, and A4. The program testing logic will be a set
i
of specific tests arranged in series with a decision to halt testing or
continue testing following each test step. This will minimize the number
of tests required and will force an assessment of each test result to be
performed.
This program will require between 1 and 1.5 years to complete, based
on the present resources available.
87
-------
TABLE A2. BaCO. TEST CONDITIONS
Particle size - <2000 pm; fines less than 100 ym eliminated
Desulfurization (TGA)
• 100 kPa pressure - 700-1000°C (scan range to find kinetic
limitations), 15% C02» 1000 ppm SO (check carbonate stability)
• 1000 kPa pressure - 900-1000°C (scan range), 15% 00 ,
1000 ppm S02 (check carbonate stability)
Regeneration (TGA) - Two-step process
• First Step
100 kPa pressure - 400-1000°C (scan range), 15% CO , 10%
H£ + CO
1000 kPa pressure - 400-1200°C (scan range), 15% C0_, 10%
H2 + CO
• Second Step
100 kPa pressure - 400°C, 25% H2S - 650°C, 8% H2S; remainder
H20 and C02 (equal parts)
1000 kPa pressure - 400°C, 25% H2S - 800°C, 8% H2S
Attrition (2 in. unit)
• Desulfurization conditions; 2 m/s velocity maximum losses
<0.05% of bed per hour
88
-------
TABLE A3. CaAl^O, TEST CONDITIONS
Particle size - <2000 urn; fines less than 100 ym eliminated
Desulfurization (TGA)
• 100 kPa pressure - 700-970°C (scan range), 15% C0_, 1000 ppm
so2 2
• 1000 kPa pressure - 900-1000°C (scan range)
Regeneration (TGA) - Reductive decomposition
• 100 kPa pressure - 800°C, 4% SO - 1000°C, 10% SO ; 15%
C02, 5% H2 + CO ' Z 2
• 1000 kPa pressure - 900°C, 5% SO - 1000°C, 10% SO ; 15%
C02, 5% H2 + CO Z
Regeneration* (TGA) - Two-step process
• First Step - 1000 kPa pressure, 400-690°C (scan), 15% C02, 10%
H2 + CO
• Second Step - 1000 kPa pressure, 300-650°C, 25% - 8% H2S, equal
parts steam and CO.
Attrition (2 in. unit)
Desulfurization conditions; 2 m/s velocity maximum losses <1% of
bed per hour
*Second regeneration scheme is examined only if the reductive decomposi-
tion scheme does not satisfy acceptance criteria.
89
-------
TABLE A4. BaTi03 TEST CONDITIONS
Particle size - <2000 pm fines less than 1000 ym eliminated
Desulfurization (TGA)
• 100 kPa pressure - 700-1000°C (scan range), 15% CO ,
1000 ppm SO-
• 1000 kPa pressure - 900-1000°C (scan range)
Regeneration (TGA) - Reductive decomposition
• 100 kPa pressure - 800° C, 4% SO - 1000°C, 10% SO.; 15%
C02, 5% H2 + CO Z Z
• 1000 kPa pressure - 900°C, 5% S02 - 1000°C, 10% S02
Regeneration* (TGA) - Two-step process
• First Step - 1000 kPa pressure; 400-660°C (scan range), 15% CO-,
10% H + CO
• Second Step - 1000 kPa pressure; 300-1000°C, 25% - 10% H2S,
remainder steam
Attrition (2 in. unit)
Desulfurization conditions; 2 m/s velocity maximum losses <0.06%
of bed per hour.
*The second regeneration scheme is examined only if the reductive decom-
position scheme does not satisfy acceptance criteria.
90
-------
APPENDIX B
ENGINEERING ASSESSMENT OF INTIMATE COAL/SORBENT MIXTURES
FOR S02 CONTROL IN FLUIDIZED-BED
COMBUSTION APPLICATIONS
PURPOSE AND SCOPE OF STUDY
This study has been carried out to investigate the technical and
environmental feasibility and economic potential of "intimate coal/
sorbent mixtures" when used in a fluidized-bed combustion system for
power generation. Various classes of intimate coal/sorbent mixtures
were first qualitatively screened for feasibility on the basis of their
probable performance assessment. Intimate coal/sorbent mixtures selected
as potentially feasible in the initial screening were then subjected to
an engineering assessment of technical and environmental performance.
Areas such as SO and NO control, trace metal and particulate control,
j£ X
solid waste and plant efficiency, and design factors for the fluidized-
Bl
bed combustor were considered. Because no actual performance or kinetic
data exist for the intimate coal/sorbent mixtures, only potential per-
formance could be addressed and problem areas identified.
Economic potential was examined by using optimistic performance
assumptions for the intimate coal/sorbent mixture. Process alternatives
for the preparation of the mixtures were identified and cost projections
for the preparation systems were generated.
BASIS FOR THE STUDY
The study specifically considers coal-fired fluidized-bed combus-
tion power plants (atmospheric-pressure boilers or pressurized boilers)
with once-through sorbent sulfur-removal systems. Limestones or dolo-
mites, which are presently utilized as sulfur sorbents in these systems,
91
-------
are considered for the sorbent materials in the intimate coal/sorbent
mixtures. Performance and economics of the intimate coal/sorbent mixture
system are compared to the conventional once-through system in order to
evaluate potential.
Projections for the sorbent utilization performance in a conven-
tional fluidized-bed combustor with once-through sorbent are summarized
in Table Bl. Cost and environmental control projections for con-
ventional fluid-bed combustion systems have been presented in the
•DO
literature. For this study the potential range of sorbent utiliza-
tions expected in a conventional fluid-bed combustor is taken to be 50
to 90 percent for dolomites and 35 to 60 percent for limestones.
INITIAL SCREENING OF CONCEPTS
An "intimate coal/sorbent mixture" may be defined as a mixture of
coal and sorbent in which the coal and sorbent are in intimate contact
during the combustion and desulfurization occurring in the fluid-bed
combustor. In the conventional once-through system, individual
particles (-3 mesh) of coal combust within a fluidized-bed of sorbent
particles (-3 mesh). Desulfurization does not appear to be kinetically
controlled but is hindered by diffusion and convection of SO- from
the burning coal particle to the sorbent particle and diffusion within
the pores of the partially reacted sorbent particle. In an intimate
coal/sorbent mixture these resistances might be substantially reduced,
resulting in a higher utilization of the sorbent material and/or lower
S0_ emissions from the combustor.
Alternative Intimate Coal/Sorbent Mixtures
A number of general classes of intimate coal/sorbent mixtures can
be conceived:
(1) Individual fine particles
(2) Consolidated particles
(3) Impregnated particles
(4) Continuous phase sorbent material or combustible material.
92
-------
TABLE Bl. PROJECTIONS OF THE CALCIUM-TO-SULFUR FEED RATIOS IN FLUIDIZED-
BED BOILERS WITH CONVENTIONAL ONCE-THROUGH SORBENTS3
Sorbent
Limestone
Uncalcined
Calcined
Dolomite
Half-calcined
Calcined
Pressure, 100 kPa
816°C 954°C
— —
2.8/1 >5/l
— —
2.2/1 80% sulfur removal efficiency
from reference B2.
-------
The first class, "individual fine particles," could consist of a
fluidized-bed of very fine sorbent particles (^40 ytn) in which the coal
(-3 mesh) is combusted, or very fine coal particles combusted in a bed
of sorbent (-3 mesh), or a bed of very fine sorbent and coal particles
undergoing combustion and desulfurization, possibly in the presence of
large particles. While this class might result in high combustion and/or
desulfurization rates and may be potentially attractive in one of its
forms, the design and operation of the fluid-bed combustor might require
major modification in order to develop an economical system. In any
case, this first class does not completely comply with the definition
of an intimate coal/sorbent mixture in that a definite gas-phase
resistance between the sorbent and coal particles will exist. This
class has not been considered further in this study.
"Consolidated particles" may consist of fine sorbent particles and
fine coal particles uniformly mixed and bound together into pellets, balls,
or granules of about -3 mesh. Consolidated particles of this type have
been produced in the laboratory. ' The fine sorbent and coal particles
might also be arranged into pellets, balls, or granules having the sorbent
on the outside and the coal centralized, or in the opposite arrangement
with the sorbent in the center. Similar particles have been produced
812
in the iron ore industry. Alternating layers of fine sorbent
particles could be bound to the surface of an individual or small group
of coal particles to form an intimate mixture.
The third general class of mixtures, "impregnated particles," could
consist of coal particles (-3 mesh) impregnated with sorbent (i.e.,
absorbed into the pores of the coal) or sorbent particles (-3 mesh)
impregnated with coal. This is a technique applied frequently with
catalyst production. Intimate contact between sorbent and coal would
result.
The last class of mixtures, a "continuous sorbent phase" or a "con-
tinuous combustible phase," might consist of a slurry of coal in a
94
-------
TJ C
Ca(OH)2 solution, or a slurry of sorbent particles in a coal/oil
suspension. Under the high-temperature conditions of the cotnbustor, the
initially continuous phase would not remain so. The Ca(OH)2 solution
would be immediately transformed into a very fine precipitate of CaO
which would react quickly with S0_ to form a fine calcium sulfate
precipitate. The coal/oil suspension would crack thermally and combust,
as in the fluidized-bed combustion of liquid fuels. This class would
ultimately behave similarly to the first class of individual fine particles.
From the four classes of intimate coal/sorbent mixtures, only the
consolidated particles class was considered further. The ability of
coal to become impregnated with sorbent or sorbent to become impregnated
with coal to the extent required (about 10 to 30 weight percent sorbent
in coal) appears very doubtful. The slurry of coal in a Ca(OH)_ solution
would result in the entrainment of a large amount of extremely fine (^5 ym
or less) particles from the bed. Also involved would be a great quantity
water and an associated reduction in plant thermal efficiency due to
the evaporation of this water. Only the consolidated particle concept
appears to be feasible and to merit further assessment.
Alternative Coal/Sorbent Consolidated Particle Characteristics
A coal/sorbent consolidated particle may exhibit many character-
istics which will greatly influence the performance of the fluidized-
bed combustor and will determine the type of equipment required to pre-
pare the particles. Table B2 summarizes the consolidated particle
characteristics, their ranges of values, and the influences which they
have on the plant performance and design.
If the relationships between the consolidated particle charac-
teristics and the factors they influence were known, the optimum
characteristics and the preparation equipment that would provide them
could be selected. Without this information a qualitative judgement
is required to assess these factors.
95
-------
Alternative Coal/Sorbent Consolidated Particle Preparation System
A general schematic process flow diagram for the consolidated
particle preparation system is shown in Figure Bl. The preparation
system will in general consist of handling equipment for coal and
sorbent, crushing and pulverizing equipment for coal and sorbent, mixing
equipment for the coal and sorbent, size enlargement equipment to form
the consolidated particles, and equipment to feed the consolidated
particles to the fluidized-bed combustor.
A number of options exist for the separate stages of the prep-
aration system, many of which are related to the consolidated particle
characteristics discussed in Table B2. Some of these options are
addressed in Table B3. In addition to these options, if coal is pul-
verized to a very fine size, a mechanical cleaning process might be
utilized to remove considerable ash, trace elements, and pyritic sulfur
R6 B7
following the pulverization. ' An absorbent for trace element con-
trol might also be incorporated into the consolidated particles. These
additional options have not been considered further in this study.
PROJECTIONS OF ENVIRONMENTAL AND TECHNICAL PERFORMANCE
Factors that are of concern to the environmental and technical
performance of the fluidized-bed combustor are technical factors
such as the combustion efficiency, heat transfer rates, bed mixing, or
bed temperature uniformity; and environmental factors such as SO
X
control, NO control, trace element release, particulate control, solid
X
waste, sorbent utilization, and so on. Without carrying out detailed
modeling or experimental investigation of the behavior of coal/sorbent
consolidated particles, it is impossible to evaluate these factors.
It is possible to speculate, however, about the possible behavior of
the consolidated particles in a fluidized-bed combustion system.
In order for a fluidized-bed combustion power plant to be economical,
the fluidization velocity in the combustor must be in the range of
96
-------
1
Raw
Coal
Dwg. 6379/02
To
Fluidized Bed
Coal
Handling
Coal Size
Reduction
Binder
1 !
i
Coal&
Sorbent
Mixing
Bin
i
der
1
Coal&
Sorbent
Size
Enlargement
Combi
i
jstor
i
Consolidated
Particle
Feeding
Sorbent
Handling
Sorbent
Size
Reduction
1
Raw
Sorbent
Figure Bl. Coal/sorbent consolidated particle preparation system
-------
TABLE B2. COAL/SORBENT CONSOLIDATED PARTICLE CHARACTERISTICS
CHARACTERISTIC
RANGE
INFLUENCES
Sorbent Type
Sorbent Particle
Size
Limestones or dolomites
250 to 2 ym; broad or
narrow distribution
oo
Coal Particle Size
Ratio of Sorbent
to Coal
500 to 2 urn; broad or
narrow distribution
5 to 50 wt% sorbent
Consolidated Particle
Size
6000 to 100 ym; narrow
distribution
Cost of sorbent, wt% sorbent in consol-
idated particle, desulfurization reaction
rate
Desulfurization reaction rate, utilization
of sorbent, rate of elutriation from corn-
bus tor, type of pulverization equipment and
cost and auxiliary power, partlculate con-
trol equipment, combustor particle size
distribution, residence time of sorbent
particle in bed
Combustion rate, degree of homogeneity of
consolidated particle, pulverization
equipment and cost, and auxiliary power,
combustor particle size distribution
Determined by type of sorbent, sulfur con-
tent of the coal, the utilization of
sorbent, sulfur removal efficiency of com-
bustor; influences cost and power for con-
solidated particle preparation, desulfur-
ization reaction rate
Combustion rate, particle internal temper-
ature, bed fluidization velocity, heat
transfer coefficient, size distribution in
bed, bed residence time of sorbent, bed
mixing uniformity, type of preparation
equipment
-------
TABLE B2 (Continued)
CHARACTERISTIC
RANGE
INFLUENCES
Consolidated
Particle Shape
Consolidated Particle
Compaction
Distribution of Coal
and Sorbent in the
Consolidated Particle
Binder Type and
Amount311"515
Cylindrical, spherical,
irregular
Loose to compact
Sorbent on surface, coal
on surface, uniformly
mixed, in multiple layers
Pitch, asphaltum, oil,
lime sodium silicate,
cellulose, bentonite
(clay), water; 2 to
50 wt% binder or pres-
sure only (as in
briquetting)B14
Type of size enlargement equipment, fluid-
ization velocity, heat transfer, bed mix-
ing, general fluidization properties
Reaction rates, particle strength and
density, bed-size distribution, prepara-
tion equipment
Combustion and desulfurization rates,
utilization of sorbent, residence time of
sorbent in bed, strength of particle, type
of size enlargement equipment
Strength of particle, attrition rate and
elutriation rate, bed particle-size distri-
bution, emissions from combustor, operating
cost of plant
-------
TABLE B3. CONSOLIDATED PARTICLE PREPARATION SYSTEM OPTIONS
Coal and Sorbent Handling - Conventional equipment
Coal and Sorbent Size Reduction -
Type of equipment: crushers and grinding mills (slow speed,
medium speed, high speed and fluid-jet pulverizer catagories);
equipment type depends on size reduction required.B8-B11
Degree of crushing size-reduction compared to grinding size-
reduction: an optimum exists in terms of power and cost for the
extent of crushing followed by grinding to produce a given particle
size.
Coal and Sorbent Mixing -
Type of equipment: tumblers, stationary shell, shell and internal
device rotate, impact mixers
Continuous or batch^-'--'-
Coal and Sorbent Size Enlargement -
Type of equipment: compaction (as pellet mill or briquetting
equipment, tumbling (as disk or drum), briquetting followed by
curshing, pressing followed by granulation: depends upon
characteristics of the product requiredBll-B14
Consolidated Particle Feeding -
The strength of the particles and their flowability may affect
the equipment design
100
-------
about 0.6 to 3.0 m/sec. In any case, the fluidization velocity will
certainly be greater than the terminal velocities of the individual
coal and sorbent particles composing the consolidated particles.
Because of this practical constraint, it appears that the con-
solidated particle attrition behavior (before, during, and following
combustion) is the single most critical consolidated particle property
relating to the performance of the fluidized-bed combustor. The type
of binder material may also then be a critical factor. The consolidated
particle attrition behavior for a given fluidization velocity will
determine three important characteristics of the fluidized bed — the
particle size distribution in the bed, the residence times of coal and
sorbent particles in the bed, and the nature of fluidization in the bed.
These characteristics will in turn determine the load on the particulate
control equipment and the fines recycle rate, the heat transfer rates
and temperature gradient in the combustor, the extent of SO control
j£
and the utilization of the sorbent, the particulate emissions from the
plant, and the nature of the solid waste. They may also relate to
trace element release and NO control.
x
Table B4 summarizes the speculative performance of coal/sorbent
consolidated particles. The overall performance to be expected is very
uncertain.
Some evidence does exist, though limited, that the basic sulfur
removal process will function in the consolidated particles (i.e., the
ability of high-calcium content lignite to burn with reduced sulfur
B4
•n-l f
release, and the ability of laboratory-fabricated consolidated
particles to desulfurize in a conventional pulverized-coal boiler).
ECONOMIC PROJECTION
The economics of the coal/sorbent consolidated particle concept are
estimated on the basis of a set of very optimistic performance assumptions
in order to identify the maximum economic potential for the concept. The
major assumption applied is that 100 percent utilization of the sorbent
101
-------
particles (limestone or dolomite) can be realized, and all other performance
aspects are acceptable. It is also assumed that the cost of coal and
sorbent handling equipment, coal/sorbent consolidated particle feeding
equipment, and all other plant equipment other than the consolidated par-
ticle preparation system are identical in cost to the conventional
fluidized-bed combustion plant (i.e., the combustor, solids feeding
system, and particulate removal equipment are identical in cost to the
conventional plant). Thus, only the cost of the consolidated particle
preparation system is considered.
A 600 MWe power plant is selected for the estimates, and a plant
heat rate of 9.5 MJ/kWh is assumed (excluding power requirements for
the consolidated particle preparation).
Economic Incentive
The potential economic advantage or disadvantage of the consolidated
particle concept is demonstrated by comparing the operating-cost savings
obtained by the assumed reduction in the rate of sorbent consumption (due
to assumed 100 percent utilization) with the increased operating cost asso-
ciated with capitalization on the preparation system capital investment.
Comparison may also be made with the cost of auxiliary power (or loss
of plant efficiency) needed for consolidated particle preparation.
A number of parameters are involved: the weight fraction of sulfur
in the coal (and the binder material), the heating value of the coal
(and the binder material), the type of sorbent (limestone or dolomite),
the utilization of the sorbent in the conventional fluid-bed combustor,
and the cost of the sorbent.
The required sulfur removal efficiency, e , for a given coal
s
(assuming no binder sulfur or binder sulfur or binder heating value) is
1 - e r (1)
S to 2 x 106
s
102
-------
where S is the environmental standard for sulfur emission from the plant,
(kgS02/GJ), H is the coal-heating value of the coal (MJ/kg),and co is
S
the weight fraction of sulfur in the coal. The Federal EPA emission
standard, S, is currently 0.516 kgSO /GJ for large, conventional coal-fired
boilers, H may range from 23.2 MJ/Kg up to 35.0 MJ/kg for eastern coals,
and cog may range from 0.01 up to 0.06. Figure B2 shows the required
sulfur removal efficiency as a function of the parameters in Equation (1).
For a once-through sorbent, the consumption rate of sorbent is given
by
M = e/U , (2)
c>
where M is the sorbent feed rate in moles calcium per mole of sulfur in
the coal, and U is the average utilization (fraction of the calcium
reacted) of the sorbent in the combustor. For the coal/sorbent con-
solidated particle it is assumed that IJ = 1.0 and thus
M = e . (3)
S
Figure B3 relates the maximum difference between the sorbent feed rates
in the conventional fluidized-bed combustion and consolidated particle
cases as a function of the conventional sorbent utilization and coal
sulfur content.
The operating cost savings due to the assumed reduced sorbent feed
rate with the consolidated particle concept is expressed in Figure B4 in
mills/kWh. The figure assumes a dolomite sorbent with a molecular weight
of 184, a sorbent purchase and disposal cost of $ll/Mg and a coal-heating .
value of 23.2 MJ/kg . For a limestone sorbent having a molecular
weight of 100, the cost savings in Figure B4 would be divided by 1.84.
A coal-heating value of 35 MJ/kg for a coal sulfur content to would be
S
equivalent to the cost savings for a 23.2 MJ/kg heating value coal
having a sulfur content of to /1.5. The cost savings is directly
0
proportional to the purchase and disposal cost of the sorbent material
(the disposal cost is assumed identical for the conventional and consoli-
dated particle cases).
103
-------
TABLE B4. SPECULATIVE PERFORMANCE OF COAL/SORBENT
CONSOLIDATED PARTICLES
Consolidated Particle Attrition -
• Attrition can occur because of mechanical forces (feeding, bed-
mixing distributor air jets) or because of decrepitation during
combustion of the particles.
• Some softening of the coal during heat-up may initially
strengthen the consolidated particles.
• During combustion ash and sorbent particles will probably
continuously break off as combustion proceeds inward. Ash
from the binder material may also break off. Complete
reduction of the consolidated particle to fine ash (coal
and binder) and individual sorbent particles is expected,
with all of this material being elutriated from the combustor
unless a binder is identified that would retain the
structural integrity of the consolidated particle even
following combustion.
Combustion -
• The consolidated particles' may be more porous than single
coal particles and may combust more rapidly and more
uniformly.
to • If mechanical attrition of the consolidated particles occurs
to anv extent, coal particle elutriation (carbon carry-over)
could result in lower combustion efficiency.
• The consolidated particle temperature will be greater than
the bulk bed temperature (possibly by several hundred degrees
centigrade) during combustion.
General Fluidization Behavior -
• The average particle size in the combustor will probably have
a terminal velocity less than the superficial fluidization
velocity, resulting in a high cyclone-fines recirculation rate
requirement.
• The bed will probably be a fast or elutriated fluidized-bed
with heat transfer and bed mixing much different from a
bubbling fluidized bed.
104
-------
TABLE B4 (Continued)
Binder Behavior -
• Ideally, a binder would provide resistance to mechanical
attrition and decrepitation due to combustion, resulting
in a combusted particle that maintains physical strength
and contains the coal ash and utilized sorbent in a single
mass. It appears that none of the binders listed in
Table B2 will display this property.
• Water as a binder would quickly evaporate and lead to break-up
of the consolidated particle unless coal softening provides
sufficient liquid bridges to maintain the particle strength.
• Other proposed binders will probably combust or decompose
during the combustion of the consolidated particle.
S02 Absorption and Sorbent Utilization -
• The sorbent particles should be of a size (probably <40 urn)
that would be expected to react quickly 0\/2 min) with S02 to
a very high level of utilization.
• The sorbent residence time in the combustor depends upon
the consolidated particle combustion rate, attrition rate,
and the rate of elutriation of the sorbent particles from
the bed.
• Desulfurization will occur while the sorbent particles are
situated within the consolidated particle and also following
attrition when the sorbent particle is exposed to the bulk
combustor gas.
• The sorbent particles with the consolidated particle should
be exposed to higher S02 concentrations than is the sorbent
in a conventional fluid-bed combustor due to decreased
dilution. Higher reaction rates should result.
• On the other hand, high C02 concentrations in the consolidated
particle may limit calcination of the sorbent and reduce
sulfur reactivity.
• Because of the possible high consolidation particle temperature
during combustion, the equilibrium S02 driving force will be
reduced relative to its value at the bulk bed temperature.
Also, the sorbent particles may quickly sinter, leading to
deactivation and limited utilization.
105
-------
TABLE B4 (Continued)
NO Control -
formation appears to be more sensitive to the combustion
conditions (excess air, bed temperature) and possibly the
sulfur level in the gas than it is to the distribution of
coal and sorbent. While the mechanism of NOX formation is
not well understood, it is not felt that the coal/sorbent
consolidated particle concept would significantly influence
NO emission from the combustor.
x
Trace Element Release -
• Because the sorbent particles are exposed to a higher temper-
ature than in a conventional fluid-bed combustor, the release
of trace elements from the sorbent may increase.
• The binder material may also release trace elements.
• The trace elements released from the coal, sorbent, and
binder may have an opportunity to react with an absorbent
incorporated into the consolidated particle. Coal ash
also may act as a trace element getter. A binder such as
bentonite (contains kaolinite and montmorillonite) may
also function as a getter.
• Washing the pulverized coal during consolidation particle
preparation may eliminate some trace element content.
Particulates -
• The use of consolidated coal/sorbent particles will probably
place a greater demand on the particulate control equipment
and may result in high particulate emissions from the plant.
Turbine protection from the erosion standpoint will be more
difficult in the case of the pressurized system.
• On the other hand, identification of an indestructible binder
could improve the particulate control situation.
Solid Waste -
• Coal-ash quantity will probably be identical to a conventional
fluidized-bed combustion power plant.
106
-------
TABLE B4 (Continued)
• Spent sorbent waste quantity may be reduced if an improvement
in sorbent utilization can be realized, but the waste fines
may be more difficult to handle than the coarse waste from
a conventional fluidized-bed combustion power plant.
• Additional waste from the binder material will also exist.
• Leaching properties of the solid waste may also be modified
as a result of the binder.
107
-------
Curve 684848-A
I
0.9
0.8
& °-7
c
.2
[o
S 0.6
0.5
0.4
0.3
0.2
0.1
0
/ Standard = 0.516 kg S02/GJ
Coal Heating Value,
MJ/kg
23.3
35.,0
0.01 0.02 0.03 0.04 0.05
Weight Fraction Sulfur in Coal
0.06
Figure B2- Combustor sulfur removal efficiency required
to achieve U.S. EPA emission standard
108
-------
Cirve 684847-A
4.0 -
CD CD
>
CD
£5
CD -f
B J5
t5 Q-
00 -a
.23 -2
to 45
CK: S
T3 O
CD «/>
CD C
U_ 5
CD C.
O CD
CD t^
*t *^i
5 S
E <->
3.0 -
1.0 -
X
-------
Curve 684852-A
to
O S:
| £
JB O
CD j_,
fT\ QJ
1/5 "CD
o o_
O _
mt_t CD
C ~~
CD
o
00
CD CD
5t= ^
S= re
Q O
6 c
=) CD
s-e
X °
ra CO
3.0
2.0
1.0
0
0.02
W$=0.01
I
0.06
I
Sorbent Material - Dolomite
Sorbent Utilization =100% in Sorbent/Fuel Pellet
Sorbent Cost = $ll/Mg
Coal Heating Value =23.2 MJ/kg
Plant Heat Rate = 9.5/VU/kWh
For Limestone Sorbent Material Divide Cost
Difference by 1.84
For Coal Heating Value of 35 MJ/kg
(in the Curves)Substitute 1.5 W for W
J
Cost Difference Is Directly Pro-_
portionaltothe Sorbent
Material Cost
Ws E Weight
Fraction Sulfur
jncoal
I
0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8
Sorbent Utilization in Conventional Sorbent Case
Figure B4- Maximum sorberrt cost savings
-------
Maximum Acceptable Costs for Consolidated Particle Preparation
The power plant total energy cost resulting from the consolidated
particle preparation system consists of contributions from fixed charges
on the capital investment, operating and maintenance costs, auxiliary
power costs, and the cost of auxiliary fuel and chemicals (binder, etc.).
Neglecting all items except the capital charges permits the calculation
of a maximum break-even capital investment for the preparation system
equipment, where fixed charges for the capital investment equal the
maximum savings shown in Figure B4 for the consolidated particle system.
The maximum break-even capital investment is shown in Figure B5 for a
dolomite sorbent costing $ll/Mg, a coal-heating value of 23.2 MJ/kg, a
capacity factor of 70 percent and a capitalization of 15 percent/yr. The
maximum break-even investment is directly proportional to the sorbent
purchase and disposal cost.
In a similar manner a maximum break-even auxiliary power requirement
for the preparation system may be determined (assuming no capital charges,
operating and maintenance cost, or auxiliary fuel and chemicals). In
this case the auxiliary power is treated as a reduction in the power
plant heat rate. This calculation is represented in Figure B6 as a per-
centage of the power plant output for a dolomite sorbent, a sorbent cost
of $ll/Mg, and a total power plant energy cost without consolidated
particle preparation of 19.0 mills/kWh (based on a pressurized combustion
B2
system with a plant heat rate of 9.5 MJ/kWh.
The actual break-even capital investment and break-even power require-
ment will be less than the maximum break-even values given in Figures B5
and B6 if all factors are simultaneously considered. In some cases
binder material may be expensive and maintenance costs very high.
Consolidated Particle Preparation System Cost Estimate
For the purposes of providing a cost estimate for the consolidated
particle preparation system, the following basis has been selected:
• 600 MWe plant
• Coal rate of 227 Mg/hr
111
-------
Curve 684851-A
90
70
* 60
50
0.05
0.04
S 40
20
10
I I I I
Sorbent Material - Dolomite
Sorbent Utilization =100% in Sorbent/
Fuel Pellet
Sorbent Cost =$ll/Mg
[Coal Heating Value =Z3.2MJ/kg
No Auxiliary Power or Fuel, No
O&MCost, No Chemicals Assumed
Capacity Factor of 70 %
LCapitalizationat 15%
For Limestone Sorbent Material
Divide Break-Even Cost by 1.84
For Coal Heating Value of
35 MJ/kg
Substitute 1.5 W for
,0.03
0.02
W =0.01
W in the Curves
Break-even Cost Is
.Directly Proportional
,to the Sorbent
Material Cost
W E Weight
o
Fraction -
Sulfur in
Coal
l
1
1
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Sorbent Utilization in Conventional Sorbent Case
Figure B5- Maximum break-even capital investment
112
-------
10
Curve 684853-A
CD
CD
03
CD
GO
O
03
a.
CD
i_
Q_
o>
o
t
CO
O.
"8
•»-»
CO
o
CO
8
7
•1 5
S3.
CL.
i_
CD
1
"c
CD
0
10.06 i'ii
Sorbent Material - Dolomite
Sorbent Utilization =100%Insorbent/ _
Fuel Pellet
Sorbent Cost =$ll/Mg
Coal Heating Value =Z3.2MJ/kg
No Capita I Investment, Aux. Fuel,
O&MCost, or Chemicals Assumed
Plant Energy Cost without SorbenT
preparation =19.0Mills/kWh
For Limestone Sorbent Material-
Divide by 1.84 (approximate)
W I Weight Fraction SuIfur_
s in coal
0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9
Sorbent Utilization in Conventional Sorbent Case
Figure B6- Break-even power requirement
113
-------
• Sorbent rate of 27 Mg/hr
• Coal grinding to 99 percent - 200 U. S. Sieve Size
• Sorbent grinding to -25 ym
• Consolidated particle size of -3000 ym
• Consolidated particle shape spherical (balling apparatus)
• Binder-bentonite
• Uniform distribution of coal and sorbent in consolidated
particle.
These characteristics are expected to result in the lowest-cost prep-
aration system of the alternatives considered. Systems requiring higher
compaction, greater degree of pulverization, or more complex distribution
of sorbent and coal will be more expensive and will consume more power.
A flow diagram of the selected preparation system is shown in Fig-
ure B7. Raw coal and raw sorbent are fed to separate trains of pulveriza-
tion equipment. The raw coal and raw sorbent are each first crushed to
about -300 ym. They are then fed to grinding equipment for further
size reduction and classification to the desired size. The pulverized
coal and sorbent are then mixed with the addition of binder material and
conveyed to the size enlargement equipment. Balling drums are used to
generate -3000 ym spherical consolidated particles, which are conveyed
to the combustor feeding system for introduction to the fluid-bed
comb ustor.
The major pieces of equipment are described in Table B5. Coal and
sorbent crushing and consolidated particle feeding equipment are excluded
from the cost estimate since they are assumed to be identical with
analogous equipment in the conventional fluid-bed combustion plant.
Multiple units are required because of limitations on capacity and are
probably preferable from the standpoint of reliability and control (the
power plant will consist of multiple fluidized-bed combustors; probably
B2
four in a 600 MWe plant) .
Because the pulverization step is probably the most expensive, the
most power-consuming, and the highest-maintenance operation in the
114
-------
Dwg. I682B27
Raw
Coal Coal Co;
rrnchar * Oili * rUOUci
QJ i/rusner Q\ i
Over
Reti
Compressor
Air ^^=4
Raw lf
Sorbent Sorbent ,- ,, Mi"
in Crusher m "(internal •
(1) crusher (i) classifier)
Coal/ Sorbent
Consolidated Particles
To *
Combustor
Feeding System
" Pulve
(
31
nzer
Classifier
•size (internal)
jrn
A Exhaust
Dust
Collector
, Product
Collector
_!'
F»
fc Dust /^y~
Collector ^
Fdll
Fan Air
Product u .
Collector m
\
thaust
n
i
Feeder
1 * '
Mi)
i
^
-------
TABLE B5. DESCRIPTION OF SELECTED PREPARATION EQUIPMENT
Equipment
Type
Includes
Number
Required
Specification
Coal Pulverizer
Ball mill
Sorbent Pulverizer Fluid energy-
type such as
Mikro-Atomizer
Coal and Sorbent
Mixer
Continuous
Muller
Fans, classifier,
product collector,
dust collector
Compressor, product
collector, dust
collector
8 227 Mg/hr
coal ground to 99% -200
mesh
20 27 Mg/hr
sorbent ground to -25um
diameter
8 254 Mg/hr
sorbent and coal uniformly
mixed with binder material;
binder ^25 Mg/hr
Size Enlargement Balling drum
Vibrating screen,
recycle conveyor
281 Mg/hr
-3000 vim consolidated coal/sorbent
particles
-------
preparation system, the degree of size reduction must be carefully
selected. Figure B8 shows the power required for coal pulverization for
two general classes of size-reduction equipment: mechanical mill-type
grinders and fluid energy-type grinders. Data have been collected
from the literature and show that the power requirement rises steeply
with decreased product diameter and that the fluid-energy mill uses
less power for a given size reduction when the product size drops
8—11
below about 40 urn. The same behavior is seen for sorbent pulveriza-
tion, shown in Figure B9.
Cost projections for pulverization systems based on projections
from a few cost data points are presented in Figure BIO. The cost in
$/kW rises steeply as the degree of pulverization increases. The
mechanical mill-type pulverizers and fluid-energy mills are about on the
same line for sorbent product particle sizes greater than around 50 urn,
but the fluid-energy mills cost more for the coal pulverization because
of the large number required. A trade-off exists between power required
and capital investment when selecting from the two general classes of
grinders. It appears that pulverizing coal to less than approximately
50 ym diameter would not be economically feasible. Sorbent pulverization
to -25 ym does appear feasible using fluid-energy mills, although a
large number may be required (about 20 for a 27 Mg/hr rate) .
A cost breakdown for the consolidated particle preparation systems
is shown in Table B6. Power requirements are also indicated. A total
capital investment of $42.7/kW is projected with a power requirement of
12.6 MWe (2.1 percent of plant power).
Table B7 compares the projected cost investment and power require-
ments for limestone and for dolomite sorbents with the maximum break-even
capital investment and power requirement presented in Figures B5 and B6.
117
-------
Curve 684849-A
oo
CD
T3
O>
CD
a.
10
Fluid
Energy
• Mechanical Mill Type Grinders
o Fluid Energy Pulverizers
250 Mm
177 um
140 Mm
105 um
90 Mm
80 Mm
50 Mm
25 Mm
10 Mm
= 70% - 200 mesh
=75% - 200 mesh
=85% - 200 mesh
=90% - 200 mesh
=97% - 200 mesh
=99% - 200 mesh
=98% -325 mesh
= 5Mm average
=2Mm average
J I
Mechanical
Transition
For Fluid
Energy
Rate = 227 Mg/hr
Feed Size From Crusher =
- 3000 Mm
Grindability =55 Low Moisture
Content(~3%)
J I
-1
-10
Maximum Particle Size, Mm
-100
Figure B8- Coal pulverization power requirement
-------
Curve 684850-A
•g 100
i i
•25 Mm =5 Mm average
•15 urn =3.5 Mm average
• Mechanical Mill-Type Grinders
o Fluid Energy Pulverizers
I I I
Fluid
Energy
Mechanical
Mill
Transition
For
Fluid
Energy
o>
a.
Feed Size =- 3000 Mm
Low Moisture Content
Rate = 9-45 Mg/hr
i i i
-10
Maximum Particle Size, Mm
-100
Figure B9- Sorbent pulverization power requirement
-------
to
o
Curve 684854-A
i i i
Coal
Pulverization
O>
.—
CO
o
r
£
Sorbent
Pulverization
-100
-10
I I I
I I I
1 I
Coal Rate = 227 Mg/hr
Sorbent Rate =45 Mg/hr
Mechanical Mill-Type Grinder
— Fluid Energy Pulverizers
J I
V I
10 100
Equipment Cost, $/kW(1976)
I I I
Figure BIO- Cost of size reduction system for coal and sorbent
-------
TABLE B6. COST ESTIMATE FOR CONSOLIDATED PARTICLE PREPARATION
Installed Equipment Cost
Coal-grinding system
Sorbent-grinding system
Coal and sorbent mixing
Size enlargement system
Bins, feeders, elevators, etc.
Total Direct Cost
Indirect Cost
Contingency and Fee (18%)
Total Cost
Basis -
• 600 MWe plant
• 227 Mg/hr coal
• 27 Mg/hr sorbent
• 1976 cost basis
$/kW
11.0
10.0
3.2
3.8
1.0
29.0
7.2
6.5
42.7
Power (MWe)
8.5
2.1
1.0
0.5
0.5
12.6
121
-------
TABLE B7. COMPARISON BETWEEN BREAK-EVEN FACTORS AND ESTIMATES*
Weight Percent
Sulfur in Coal
Limestone Sorbent
Maximum Break-Even Cost
Cost ($/kW) Estimate ($/kW)
Dolomite Sorbent
Maximum Break-Even
Cost ($/kW)
Cost
Estimate ($/kW)
6 20-60 -N 6.5-55 ~N
5 16-53 / 5.5-45 /
NJ
IV)
4
3
2
1
6
5
4
3
2
1
13-40 V 38-43 4.0-35 V 40-48
9-27 1 3.0-25 f
5-16 I 1.5-14 I
1.5-5 J 0.5-4 J
Maximum break-even Power Maximum break-even Power
power (%) estimate (%) power (%) estimate (%)
2.3-6.5
2.0-5.6
1.5-4.5
1.1-3.0
0.6-2.0
0.2-0.5
0.8-6.2
0.7-5.2
1.9-2.1 0.5-4.2
0.4-3.0
0.2-1.8
0.1-0.5
2.1-2.5
aBasis - see Figures B5 and B6.
100% utilization of the consolidated particle sorbent
Limestone/dolomite cost - $ll/Mg ($10/ton) (purchase and disposal)
Limestone utilization ranges from 0.35 to 0.60 in conventional fluid-bed combustion
Dolomite utilization ranges from 0.50 to 0.90 in conventional fluid-bed combustion
The left-hand figure in the break-even columns for cost and power refers to maximum
utilization of the sorbent, the right-hand figure refers to minimum utilization.
The left-hand figures in the cost and power estimate columns refer to 1% sulfur coal
and the right-hand columns refer to 6% coal.
-------
It is evident from Table B7 that for a sorbent cost of $ll/Mg
and 100 percent utilization of the consolidated particle sorbent:
• If maximum utilization of limestone or dolomite sorbents can be
obtained in conventional fluidized-bed combustors (60 and 90 per-
cent, respectively), either the capital investment or the auxiliary
power requirement alone makes the consolidated particle concept
appear economically not competitive.
• If only minimum utilization of limestone or dolomite sorbents
can be obtained in conventional fluidized-bed combustors (35 and
50 percent, respectively), either the capital cost or the auxiliary
power alone still makes the consolidated particle concept appear
economically not competitive for coal-sulfur contents of less
than 4 weight percent.
• If all cost factors are considered (capital charges, auxiliary
power, operating and maintenance, auxiliary fuel and chemicals,
coal losses, etc.) simultaneously, the consolidated particle
concept will not be economically competitive with a sorbent cost
of $ll/Mg, at any coal sulfur content.
The only circumstances under which the consolidated sorbent concept
could be economically competitive are:
• If all the optimistic assumptions applied could be realized -
100 percent utilization of the sorbent in the consolidated
particle, no additional capital investment associated with the
fluid-bed combustor, particulate control equipment, feeding
equipment, spent sorbent processing, etc., - and, simultaneously,
• If the sorbent material were very expensive, (>$20/Mg), the
available coal contained more than 5 weight percent sulfur and
the conventional sorbent utilization performance was very poor.
123
-------
ASSESSMENT
The only technically feasible intimate coal/sorbent mixture
that could be identified for the current fluidized-bed
combustion design concept is the consolidated coal/sorbent
particle concept.
Attrition of the consolidated particle is the most critical
factor influencing the performance'and feasibility of the con-
cept. Modifications to the combustor design would probably be
required in order to apply the consolidated particle concept.
The performance (technical and environmental) cannot be esti-
mated without initiating a test program. The overall technical
and environmental performance of the consolidated particle con-
cept could conceivably be worse than or better than the conven-
tional fluid-bed combustor, but it is highly unlikely that any
significant improvement in performance is to be realized.
Except under very extreme conditions, the consolidated particle
concept will not be economically competitive with conventional
fluid-bed combustion concepts.
Washing the pulverized coal during consolidated particle
preparation could reduce trace elements, ash, sulfur, and the
sorbent requirement. The economics of this option have not
been investigated.
The most attractive consolidated coal/sorbent particle from
the standpoint of technical and environmental impact would
utilize a binder that would maintain the coal-ash and sorbent
particles in discrete, consolidated particles following com-
bustion. A binder that will effect this behavior has not been
identified.
124
-------
REFERENCES
Bl. Henschel, D. B., "The U.S. Environmental Protection Agency Program
for Environmental Characterization of Fluidized-Bed Combustion
Systems," Fourth International Conference on Fluidized-Bed Combus-
tion, ERDA, McLean, Virginia, December 1975.
B2. Fluidized Bed Combustion Process Evaluation, submitted by Westing-
house Research Laboratories to EPA, Report No. EPA-650/2-75-027-C,
Sept. 1975 (NTIS No. PB 246116).
B3. Chauhan, S. P., H. F. Feldman, E. P. Stambaugh, J. H. Oxley,
"(Hydro) Gasification of Battelle Treated Coal," paper presented
at 170th National Meeting ACS, Chicago, IL, August 1975.
B4. Buttermore, W. H., Sulfurtain Process: "Trapping Out Sulfur
Dioxide While Burning Pulverized Coal," Coal Age, 80_(2) , 148 (1975).
B5. Eastman, D., "Production of Sulfur-Free Carbon Monoxide Containing
Gas," U.S. Patent 829,844, 1960.
B6. Gorin, E., H. E. Lebowitz, "The Removal of Sulfur and Mineral Water
From Coal," in Coal Processing Technology, AIChE, 1974.
B7. Schultz, H., Hattman, E. A., Booher, W. B., "The Fate of Some Trace
Elements During Coal Pretreatment and Combustion," Trace Elements
on Fuel, S. P. Babu, Ed., Advances in Chem. Series 141, p. 139, 1975.
B8. Gryling, G. R., Ed., Combustion Engineering, Combustion Engineering,
Inc., New York, 1966.
B9. "Sulfur Oxide Removal From Power Plant Stack Gas - Sorption by
Limestone or Lime Dry Process," TVA, 1968, NTIS No. PB178972.
BIO. Cremer, H. W. and T. Davies, Eds., Chemical Engineering
Practice, Volume III, Solid Systems, New York: Academic Press,
Inc., 1957.
Bll. Perry, J. H., Ed., Perry's Chemical Engineers' Handbook,
5th Edition, New York: McGraw-Hill Book Co., 1968.
125
-------
B12. Knepper, W. A., Ed., Agglomeration, New York: Intersciences
Publishers, 1968.
B13. Leaver, R. H., "Evaluating Industrial Pelleting," Chemical
Engineering, 155, Jan. 1976.
B14. Lowry, H. H., Ed., Chemistry of Coal Utilization, Supplementary
Volume, New York: John Wiley & Sons, Inc., 1963.
B15. Moore, J. E., W. B. Pietsch, "Briquetting and Compacting of Lime
and Lime-Bearing Materials," International Briquetting & Agglomera-
tion Proceedings, 38, August 1967.
B16. A Development Program on Pressurized, Fluidized-Bed Coal Combustion,
Quarterly Report, submitted by Argonne National Laboratories,
Oct. 1, -Dec. 31, 1975, to ERDA, ERDA 14-32-0001-1780,
NTIS ANL/ES-CEN-1014.
126
-------
TECHNICAL REPORT DATA
(Please read Instructions on tile reverse before completing)
1. REPORT NO.
EPA-600/7-78-005
2.
3. RECIPIENT'S ACCESSION- NO.
TITLE AND SUBTITLE Alternatives to Calcium-based SO2
Sorbents for Fluidized-bed Combustion: Conceptual
Evaluation
5. REPORT DATE
January 1978
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Richard A. Newby and Dale L. Keairns
8. PERFORMING ORGANIZATION REPORT NO.
3. PERFORMING ORGANIZATION NAME AND ADDRESS
Westinghouse Research and Development Center
1310 Beulah Road
Pittsburgh, Pennsylvania 15235
10. PROGRAM ELEMENT NO.
EHE623A
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 PERIOD COVERED
Final; 12/75-12/76
14. SPONSORING AGENCY CODE
EPA/600/13
IB. SUPPLEMENTARY NOTES jERL-RTP project officer is D. Bruce Henschel, Mail Drop 61,
919/541-2825. EPA-650/2-75-027c is a previous report related to this work.
16. ABSTRACT
repOrt gjves results of a. conceptual engineering evaluation to screen
supported metal oxides as alternatives to natural calcium-based sorbents (limestones
and dolomites) for SO2 control in atmospheric and pressurized fluidized-bed combus-
tion (FBC) processes. Alternative sorbents were evaluated, using three acceptance
criteria: SO2 removal capability in the combustor, predicted by thermodynamics;
SO2 concentrations achievable in the regenerator off-gas , according to thermodyna-
mics; and SO2 concentrations off the regenerator, achievable based on material and
energy balances . The evaluation identified 14 potentially acceptable sorbents for
atmospheric FBC, and 11 for pressurized FBC. Cost estimates were prepared to
project the maximum acceptable loss rates for the alternative sorbents due to attri-
tion and/or deactivation. Loss rates must be less than 0.1% of bed inventory per hour
in order to compete economically with natural calcium-based sorbents , even if maxi-
mum thermodynamic performance were obtained. U.S. resources of some minerals
may be of extreme importance for many of the alternative metal oxide sorbents
considered.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Fluidized-bed Pro-
cessing
Combustion
Sulfur Dioxide
Sorption
Oxides
Limestone
Dolomite (Min-
eral)
Air Pollution Control
Stationary Sources
SO2 Sorbents
Metal Oxides
13B
13H,07A
21B
07B
07D
08G
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report}
Unclassified
21. NO. OF PAGES
143
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
127
------- |