DoE
EPA
Co
knifed States
Department
of Energy
Fossil Energy Division
Fossil Fuel Utilization
Washington DC 20540
ANL/CEN/FE-78-13
United States Environmental
Protection Agency
Office of Research and Development
Industrial Environmental Research
Laboratory
Research Triangle Park NC 27711
EPA-600/7-79-157
July 1979
Regeneration of Sulfated
Limestone from FBCs
and Corrosive Effects
of Sulfation Accelerators
in FBCs: Annual Report
Interagency
Energy/Environment
R&D Program Report
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series. These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
5 1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related pollutants and their health and ecological
effects; assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.
-------
EPA-600/7-79-157
DoE-ANL/CEN/FE-78-13
July 1979
Regeneration of Sulfated Limestone
from FBCs and Corrosive Effects
of Sulfation Accelerators in FBCs:
Annual Report
by
G. J. Vogel, I. Johnson, J. F. Lenc, D. S. Moulton,
R. B. Snyder, J. A. Shearer, G. W. Smith, W. M. Swift,
E. B. Smyk, F. G. Teats, C. B. Turner, and A. A. Jonke
Argonne National Laboratory
9700 South Cass Avenue
Argonne, Illinois 60439
EPA No. IAG-D5-E681
DoE No. W-31-109-Eng-38
Program Element No. INE825
Project Officers:
David A. Kirchgessner John F. Geffken
EPA/Industrial Environmental DoE/Fossil Fuel Utilization
Research Laboratory Washington, DC 20540
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY U.S. DEPARTMENT OF ENERGY
Office of Research and Development Fossil Fuel Utilization
Washington, DC 20460 Washington, DC 20540
-------
TABLE OF CONTENTS
Page
ABSTRACT 1
SUMMARY 1
TASK A. REDUCTIVE DECOMPOSITION PROCESS STUDIES 3
1. Experimental Studies . 3
2. Prediction of Process Parameters Based on a
Regeneration Model 5
3. Predicted Effects of Off-Design Conditions on Regenerator
Operating Parameters 19
4. Comparison of New and Earlier Predictions of Effects
of Process Parameters 24
5. Economic Feasibility of Regenerating Sulfated Limestone ... 27
TASK B. ALTERNATIVE REGENERATION PROCESS DEVELOPMENT 28
1. Rotary-Kiln Regeneration 28
TASK D. EFFECTS OF LIMESTONE SULFATION ACCELERATORS ON CORROSION
RATES OF METALS IN AN AFBC 34
1. Atmospheric Combustor Facility for Corrosion Tests 34
2. Corrosion Behavior of Materials in Fluidized-Bed
Environments 42
REFERENCES 62
111
-------
LIST OF FIGURES
No. Title
1. Flow Sheet for Conceptual 635-MW Atmospheric Fluidized-Bed.
Combustor . 6
2. Velocity Ratio vs. Inclination Angle in a Laboratory-Scale
Rotary Kiln . . . .- 39
3, Gas-Sampling Locations and Reactor Temperatures in the
Laboratory-Scale Rotary Kiln Operated at 1040°C and 4 rpm . . . . 31
4. Variation of Percent S02 in Off-Gas with Carrier-Gas
Flow Rate, during Limestone Regeneration in a Small Rotary
Kiln at 1040°C . 32
5. Simplified Piping Schematic of Atmospheric Combustor
Facility ..... ............ 35
6. 152-mm-ID Atmospheric Combustor 36
7. Some Major Components of the New AFBC 37
8. Automatic Control Loops for Atmospheric Combustor Facility .... 39
9. High-Temperature 5-cm-dia Quartz Fluidized-Bed Reactor for
Salt Corrosion Studies ................. 43
10. Average Thickness of Surface Scale and Corrosive Penetration
for Corrosion Coupons Exposed Inside the Bed for 100 h at
1123 K except that the Temperature for Inconel 601 was 1173 K . . 45
11. Average Thickness of Surface Scale and Corrosive Penetration
for Corrosion Coupons Exposed Above the Bed for 100 h at
1123 K except that for InConel 601 the Temperature was 1173 K . . 46
12. SEMs of Type 304 Stainless Steel after a 100-h Exposure at
1123 K 47
13. SEMs of Type 310 Stainless Steel after a 100-h Exposure at
1123 K ................ ......... 48
14. SEMs of Incoloy 800 after a 100-h Exposure at 1123 K ....... 49
15. SEMs of Inconel 600 after a 100-h Exposure at 1123 K ....... 50
16. SEMs of Inconel 601 after a 100-h Exposure at 1123 K 51
17. SEMs of RA333 after a 100-h Exposure at 1123 K . 52
18. X-ray Microprobe Line Analyses for Ni, Fe, and Cr on
Incoloy 800 Specimens Exposed above the Bed for 100 h at
1123 K . . . 53
IV
-------
LIST OF FIGURES (contd)
No. Title
19. SEMs and EDAX Analysis of the Surface of the Scales Developed
on Incoloy 800 Exposed above the Bed for 100 h at 1123 K 54
20. SEMs of Type 304 Stainless Steel after a 100-h Exposure 56
21. SEMs of Type 310 Stainless Steel after a 100-h Exposure 57
22. X-ray Microprobe Line Analyses for Fe, Cr, Ni, 0, and S on
Type 304 Stainless Steel Specimen Exposed Above the Bed
Containing NaCl for 100 h at 923 K . . . 59
23. SEMs and X-ray Images for Fe, Cr, 0, S, and Cl of Type 304
Stainless Steel Specimen Exposed Above the Bed Containing
NaCl for 100 h at 923 K 60
24. SEMs of (a) Sulfide/Mixed-Oxide Interface and
(b) Mixed-Oxide/Iron-Oxide Interface Observed in Type 304
Stainless Steel Specimen Exposed Above the Bed for 100 h
at 823 K 61
-------
LIST OF TABLES
No. Title Page
1. Baseline Conditions for Flow Sheet 7
2. Predicted Effects of Feed Ratio on Regeneration System
Performance and Dimensions 8
3. Predicted Effects of Superficial Gas Velocity on Regenerator
Dimensions . 10
4. Predicted Effects of Solids Residence Time on Regenerator
Dimensions and Performance 11
5. Predicted Effects of Solids Feed Temperature and Gas Inlet
Temperature on Regenerator Dimensions and Performance 1.3
6. Predicted Effects of Regenerator Bed Temperature on Regenerator
Dimensions and Performance 15
7. Predicted Effect of Pressure on Regenerator Dimensions 17
8. Predicted Effect of Changes in Feed-Gas Oxygen Concentration
on Regenerator Dimensions and Performance 18
9. Predicted Effects of Off-Design Conditions on Regenerator
Performance 20
10. Design of Regenerator to Accommodate Turndown 23
11. Regeneration of Sulfated Greer Limestone with Char in a Small
Rotary Kiln 31
12. Regeneration at 1040°C of Spent Limestone Sorbent in a Small
Rotary Kiln 33
13. Nominal Operating Conditions. Atmospheric Fluidized-Bed
Combustor 38
14. Control Loops for Atmospheric Fluidized-Bed Combustor 40
15. Process Levels Necessitating Shutdown 41
16. Compositions of Alloys 43
17. Experimental Conditions 44
VI
-------
REGENERATION OF SULFATED LIMESTONE FROM FBCs AND
CORROSIVE EFFECTS OF SULFATION ACCELERATORS IN FBCs
ANNUAL REPORT
July 1977 - September 1978
by
G. J. Vogel, Irving Johnson, 0. K, Chopra, J. F. Lenc,
D. S. Moulton, J. A. Shearer, R. B. Snyder,
G. W. Smith, W. M. Swift, E. B. Smyk,
F. G. Teats, C. B. Turner, and A. A. Jonke
ABSTRACT
These studies support the national development program
in fluidized-bed combustion. The objectives of this program
are to develop an economically and environmentally accept-
able process for the regeneration of the partly sulfated
product of a fluidized-bed coal combustor, to obtain the
design data needed for the construction of larger regen-
erators, and to determine the possible corrosive effect on
metallic alloys of sulfation acceleration agents added to
the limestone sorbent. A regeneration process model has been
developed; the model has been used to investigate the
effects of the main variables on regenerator size and
performance and to estimate the economic feasibility of
regeneration. The results of an investigation of a regen-
eration process using a rotary kiln in place of a fluidized
bed are reported. An atmospheric pressure FBC was placed
into operation to study the corrosion of metallic alloys by
limestone treated with various sulfation accelerators such
as NaCl or CaCl2. The results of corrosion studies carried
out in a laboratory-scale facility are reported.
SUMMARY
Task A. Reductive Decomposition Process Studies
Experimental. Work continued on characterization of the reductive
decomposition of sulfated limestone in a coal-fired fluidized-bed regenerator.
Three experimental studies were done in the ANL 10.8-cm-ID bench-scale unit:
(1) to ascertain the effect of preheating sorbent feed to the regenerator,
(2) to ascertain the effect of high bed temperature on extent of regeneration,
and (3) to determine if hot spots are present in the regenerator bed near the
coal feed point.
The regeneration model correctly predicted the effect of solids preheat,
although S02 enrichment in the off-gas was not obtained because of the ex-
perimental procedure. A high temperature (1423 K vs. 1373 K) in the regener-
ator resulted in a small increase in the extent of regeneration of CaO.
-------
However, defluidization velocity was about 40% higher than at 1373 K. No hot
spots were detected in the regenerator bed near the coal feed points.
Development of a Regeneration Process Model. On the basis of experimen-
tal results obtained in the ANL 10.8-cm-ID bench-scale fluidized-bed regenera-
tor, a mass- and energy-balance constrained model for the reductive decomposi-
tion of sulfated limestone has been developed. The model allows regenerator
performance and dimensions to be predicted, given specific operating conditions,
Results are presented for the effects of feed ratio (FR), superficial gas vel-
ocity, solids residence time, gas and solids feed temperatures, bed tempera-
ture and pressure, and feed-gas oxygen concentration on regenerator perfor-
mance and dimensions. The effect of off-design conditions on (1) regenerator
operating parameters and (2) methods of designing a regenerator to accommodate
turndown are discussed. Finally, methods of interpreting process development
unit (PDU) experimental data for use in designing larger regenerators are
presented.
Economic Feasibility of Regenerating Sulfated Limestone. An economic
analysis of regenerator and sulfur-recovery systems was done to determine
which factors were important and to what degree. Capital cost, capacity
factor, and sorbent price (including disposal cost) were found to be impor-
tant factors, while fuel price and sulfur credit value were found to be of
lesser importance.
Task B. Alternative Regeneration Process Development
Rotary Kiln Regeneration. Rotary kilns are being evaluated as an alter-
native to fluidized-bed reactors for the regeneration of sulfated limestone.
Studies are being conducted in a laboratory-scale rotary kiln to demonstrate
the process and to obtain data for larger-scale tests. Tests made using sul-
fated Greer limestone mixed with coal char at 1040°C indicated that SC>2 con-
centrations near the equilibrium value of 20% can be obtained. Regenerations
of about 40% of the CaS04 were obtained.
Task D. Effects of Limestone Sulfation Accelerators on Corrosion Rates of
Metals in an AFBC
Atmospheric Combustor Facility for Corrosion Tests. A new PDU-scale
automated atmospheric-pressure fluidized-bed coal combustion facility (AFBC)
was constructed. The facility will be used in an experimental program to
study the corrosive effects on metals of construction of agents (e.g., NaCl,
CaCl2, or Na2CC>3) that accelerate limestone sulfation. The new facility,
including the process-control system for automated operation, is described.
A series of runs is being carried out in the new AFBC (prior to corrosion
experiments) to evaluate the effects on SC>2 retention of adding low concentra-
tions (1.0 mol % or less) of CaCl2 or NaCl to Grove limestone. A total of
31 runs in this series have been conducted. Evaluation of the results awaits
the completion of analyses.
Compositions and geometries of corrosion specimens, as well as locations
of exposure, have been selected. Parameters to be measured during corrosion
tests have been identified.
-------
To obtain baseline corrosion data on a variety of metals, initial corro-
sion combustion experiments are to be conducted with no sulfation accelerator
present in the fluidized bed. Next, 100-h scoping tests will be made to de-
termine the effects of sulfation accelerators on corrosion specimens of various
metals. Following the 100-h tests, those metals and accelerating agents se-
lected for further testing will be subjected to 1000-h tests.
Corrosion Behavior of Materials in Fluidized-Bed Environment. As part of
the study of the effect of NaCl and other salts on the utilization of lime-
stone for S02 absorption in FBCs, a series of corrosion tests were made in a
laboratory-scale fluidized-bed reactor. These tests were done with a gas mix-
ture containing from 200 to 500 ppm S02, 5% 02, and the rest N2. In the
tests, at 850°C and of 100-h duration, samples were exposed within or above the
fluidized bed of sulfated dolomite. Tests were done with salt present in or
absent from the bed.
Metallographic examination of the samples indicated (1) the addition of
NaCl or CaCl2 to the bed increases corrosion; (2) in the presence of salt,
Types 304, 316, and 310 stainless steel perform better than did high-nickel
alloys; (3) the corrosion behavior of stainless steels is relatively insensi-
tive to the NaCl concentration in the bed; (4) the corrosion behavior of
Inconel 600, Inconel 601, and RA333 in the presence of 1 mol % NaCl or
0.1 mol % CaCl2 is comparable to that of the stainless steels; (5) the
internal corrosive attack consists of three distinct zones: internal oxida-
tion, internal sulfidation, and a carburized zone; and (6) specimens exposed
above the bed at 923 and 823 K show extensive corrosive attack.
TASK A. REDUCTIVE DECOMPOSITION PROCESS STUDIES
(E. B. Smyk and R. B. Snyder)
1. Experimental Studies
A fluidized-bed reductive decomposition process for regenerating lime
from partially sulfated limestone, using coal as both a reductant and a heat
source, has been developed in a 10.8-cm-ID regenerator. In earlier work,l>2
experiments were performed to ascertain the effects of bed temperature, solids
residence time (SRT), superficial gas velocity, feed-gas oxygen concentration,
and bed height on process performance (most notably, off-gas S02 concentra-
tion).
In all earlier regeneration experiments at ANL, room-temperature sulfated
sorbent was fed to the regenerator reactor. In an industrial process, the
sorbent would enter the reactor at about 1116 K. The effect of sorbent feed
temperature on CaO regeneration has been tested. Experiments were performed with
600-2000 ym (-10 +30 mesh) and 600-1400 ym (-14 +30 mesh) sulfated Greer lime-
stone from Pope, Evans, and Robbins (PER) with sorbent preheating temperatures
of 1075-1175 K. Preheating the solids feed permitted experiments to be done
in the ANL 10.8-cm-ID reactor with solids residence times (SRTs) as brief as
4.4 min. (The range of SRTs was about 4 to 10 min.) In earlier experimental
work, all regeneration experiments were performed with SRTs greater than
7 min.
-------
In the regeneration experiments with the larger sulfated limestone par-
ticles (600-2000 pm), the extent of regeneration at a high SRT was equivalent
to that projected from previous experiments^ with finer (600-1400 urn) lime-
stone. However, even with feed sorbent preheated to 1150 K at a relatively
low SRT (5.4 min), the extent of regeneration was lower than that projected
for large particles. Because regeneration reactions are relatively rapid, it
is believed that the lower conversion could be due to poorer fluidization of
the larger particles in the rather small reactor.
In the experiments with small sulfated limestone particulate (600-
1400 vim), preheating the solids feed (>1150 K) was beneficial. The extent
of regeneration for SRTs as low as 4.4 min were equivalent to the projections
made from previous experiments performed at higher SRTs (>7 min). Preheating
the solids improved the extent of regeneration at the lower SRTs (<7 min)
and allowed it to reach the predicted values. The SC>2 concentration in the
off-gas generally increased from 6% to 10.4% as the SRT was decreased; this
was expected.
In these experiments, as predicted from the regeneration process model,2
solids preheat did not cause SC>2 enrichment because of the way regeneration
experiments are performed at ANL—C>2 and N2 are mixed to satisfy the oxygen
demand of the reactor while a constant total gas input is maintained. Changes
in gas velocity in these experiments are thereby avoided. Therefore, the SC>2
concentrations obtained in these experiments are equivalent to those expected
in an industrial process in which the solids would be fed hot to the regener-
ator reactor.
The effect of a regeneration temperature in excess of 1373 K was tested
by performing a regeneration and defluidization experiment at 1423 K. This is
the highest temperature at which a regeneration experiment has been performed
at ANL. Regeneration was not significantly greater than that obtained at
1373 K. Moreover, the higher temperature caused a drastic increase (about 40%)
in the defluidization velocity (the minimum gas velocity required to prevent
agglomeration). The net effect of this and the very small increase in the ex-
tent of regeneration is a lower S(>2 concentration in the regeneration reac-
tor off-gas. Also, higher temperatures accelerate the sintering process,
rendering the sorbent less reactive in subsequent sulfation steps. Therefore,
temperatures higher than 1373 K do not seem to offer improved performance in
the regeneration step of an FBC process.
In multicyclic usage of sorbent, loss of sorbent reactivity could be
greater than normal due to localized high-temperature spots near the coal
feed point. The regenerator bed temperature near the coal feed point was
measured with a probe to evaluate the extent of any such hot spots. The
local temperature 2.5 to 9.2 cm from the coal injection point along the jet
path was found to exceed other bed-temperature measurements by not more than
5 degrees K. More detailed results have been previously reported.1
-------
2. Prediction of Process Parameters Based on a Regeneration Model
Experimental development of the fluid-bed, reductive decomposition
regeneration process for partially sulfated limestone, using a PDU regen-
erator, has been successful. The regenerator is 10.8 cm in ID, and the
expanded bed height is 45 cm. The extent of regeneration of the partially
sulfated limestone and the S02 concentration in the off-gas were deter-
mined as a function of various operating conditions. These results were
reported previously.
Here, a process flow sheet is discussed for a conceptual 635-MW FBC
process with sorbent regeneration, using Greer limestone. The performance
of Greer limestone as a function of cycle number was established in a
ten-cycle experiment,3 and the results pertaining to reactivity, calcium
utilization, ash buildup, and elutriation as a function of cycle number
have been incorporated into the ANL-developed regeneration process model.
Results presented here differ somewhat from previous ones since the regen-
eration model has been made rigorous and all material balances have been
closed.
Figure 1 presents a process flow sheet for a 635-MW AFBC burning
Sewickley coal and utilizing Greer limestone as the sorbent, coupled to a
regenerator and sulfur-recovery system. The baseline conditions are pre-
sented in Table 1; these are typical but not necessarily optimum. The sul-
fated limestone is assumed to be introduced into the regenerator at 1116 K
(1550°F), which is the bed temperature in the fluid-bed combustor. The
superficial gas velocity of 1.4 m/s is 12% above the velocity predicted to
be necessary to prevent agglomeration (and consequently defluidization) for
sorbent particles having a mean size of 1500 pm (-1/8 in.) when regenerated
at 1373 K with 2% total reducing gas in the regenerator off-gas. The feed
gas to the regenerator is assumed to be heated to 675 K by recovering waste
heat from the regenerator off-gas.
The coal consumption of the total regenerator system with 1116 K solid
and 675 K gas feed streams was estimated to be 95 Mg/d. The fuel consump-
tion of the total regeneration system is obtained by adding the coal fed to
the regenerator (178 Mg/d) to the coal fed to the sulfur-recovery step
(47 Mg/d) and subtracting the fuel credits for the regenerator off-gas cy-
clone product (51 Mg/d) and the sensible heats of (1) the regenerated sor-
bent (41 Mg/d) and (2) the tail-gas stream from the sulfur-recovery step
(38 Mg/d) that will be recycled to the boiler.
-------
PARTICULATE
REMOVAL
B
0
L
R
TO ATMOSPHERE
672Mg/d (TOTAL)
306 Mg/d SULFATED
•- SORBENT
51 Mg/d UNBURNED CARBON
315 Mg/d ASH
REGENERATED
SORBENT 3332 Mg/d
THERMAL CREDIT-
41 Mg/d COAL
4545 Mg/d (TOTAL)
SULFATED
SORBENT
43180 Mg/d AIR
I
3639 Mg/d
PARTICULATE
REMOVAL
47Mg/d COAL
'920K
ELEMENTAL
SULFUR
92 Mg/d
REGEN. SORBENT
36Mg/d
UNBURNED CARBON
THERMAL CREDIT
(RECYCLED TO
COMBUSTOR)
DRAW OFF
906Mg/d
743 Mg/d SULFATED STONE
163 Mg/d ASH
THERMAL CREDIT-38Mg/d COAL
I— COAL 4779 Mg/d
FRESH SORBENT 907 Mg/d CaO/S = I.I
Fig. 1. Flow Sheet for Conceptual 635-MW Atmospheric
Fluidized-Bed Combustor. Fresh sorbent feed
CaO/S ratio =1.1.
a. Predicted Effect of Feed Ratio on Regenerator Dimensions and
Performance
Once the sulfur content of the coal is given, the S02 retention
necessary to meet EPA emission regulations can be calculated. Based on the
reactivity of a given stone, the CaO/S mole ratio necessary to provide the
needed SC>2 retention can be projected. Tests in which the sorbent was re-
peatedly cycled between the combustor and regenerator have shown that the
sorbent has less reactivity at higher regeneration-cycle numbers. It is thus
possible to predict the CaO/S mole ratio necessary to provide the necessary
SC>2 retention as a function of regeneration-cycle number. Feed ratio (FR) is
defined as the amount of CaO in the fresh feed, divided by the total CaO fed
to the combustor. The process model allows the overall CaO/S ratios needed
for SC>2 retention to be predicted as a function of FR by calculating the
age (regeneration-cycle number) distribution of regenerated sorbent as a func-
tion of FR and adding the resulting reactivity to the reactivity of the fresh
sorbent.
-------
Table 1. Baseline Conditions for Flow Sheet (Fig. 1)
Coal Sorbent
Sewickley coal Greer limestone
28.5 MJ/kg (12250 Btu/lb) 80% CaC03
4.3% S 20% Inert
10.0% Ash
AFBC Boiler
635 MW at 37% conversion efficiency (9200 Btu/kW-h)
Bed Temperature - 1116 K (1550°F)
Pressure - 103 kPa (1 atm)
Bed Area - 767 m2 (8255 ft2)
Combustion Efficiency - 99%
Sulfur Retention - 83%
Fresh/Total Sorbent Feed Ratio (FR)a - 0.2
Regenerator
Bed Temperature - 1373 K (2012°F)
Pressure - 103 kPa (1 atm)
Bed Area - 36.2 m2 (389 ft2)
Bed Height - 0.67 m (2.2 ft)
Gas Velocity - 1.4 m/s (4.5 ft/s)
Solids Residence Time - 7 min
Extent of Regeneration - 65%
Total Regeneration System Fuel Burden - 95 Mg/d (105 T/d)
(about 2.0% of the coal fed to the combustor)
Composition of Regenerator Flue Gas
8.0% S02
2.0% CO
19.4% C02
8.8% H20
61.8% N2
aThe fresh sorbent-feed mole ratio, hereinafter referred
to as FR, is the mole fraction of CaO as fresh feed to
the combustor, compared with the total CaO feed to the
combustor including regenerated sorbent.
Table 2 shows the effects of changes in FR on regeneration system per-
formance and dimensions. At low FR (e.g., 0.10), little fresh sorbent is
needed and little sulfated sorbent is discarded. However, the lower the
FR, the larger the regenerator since more sulfated sorbent is being cir-
culated between the combustor and regenerator. In addition, regenerator
fuel consumption will be higher because of the larger sorbent throughput,
and the S02 concentration in the regenerator off-gas will be lower be-
cause (1) more inerts build up and (2) the sulfated sorbent releases less
sulfur due to the lower reactivity of the stone.
-------
Table 2. Predicted Effects of Feed Ratio on Regeneration System Performance and Dimensions
Combustor
635 - MW AFBC
Greer Limestone
Sewickley Coal
83% S02 Retention
Coal Feed Rate - 4779 Mg/d
Regenerator
Bed Temperature - 1373 K
Gas Inlet Temperature - 672 K
Solids Feed Temperature - 1116 K
Solids Residence Time - 7 min
Extent of Regeneration - 64.6%
Pressure - 103 kPa
Superficial Gas Velocity - 1.37 m/s
FR
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0.90
Fresh
Sorbent
Feed,
Mg/d
643
907
1171
1436
1700
1964
2228
2492
2756
Regenerator
Feed,
Mg/d
5249
3639
2829
2245
1768
1352
977
632
307
Waste,
Mg/d
703
906
1121
1336
1549
1760
1864
2173
2376
Regenerator Coal
Usage, % of
Combustor Usage
3.48
2.97
2.50
2.06
1.66
1.28
0.93
0.60
0.29
Regenerator
Bed Area,
m2
45.7
36.2
29.6
24.1
19.2
14.8
10.7
7.0
3.4
Regenerator
Bed Height,
m
0.76
0.67
0.63
0.62
0.61
0.61
0.60
0.60
0.60
Regenerator
Off-Gas S02
Concentration,
*
7.08
8.03
8.38
8.54
8.60
8.65
8.64
8.60
8.58
oo
-------
Conversely, at higher FR (e.g., 0.40), more fresh sorbent is needed and
more sulfated sorbent is discarded. The regenerator would be smaller and the
fuel consumption would decrease, while the off-gas S02 concentration would
increase. Thus, at low FR, operating cost is low and capital cost is high,
while at high FR, operating cost is high and capital cost is low. Therefore,
the decision as to which FR to incorporate in the design must be made on the
basis of economic factors. This type of analysis is made in a later section.
b. Predicted Effect of Superficial Gas Velocity on Regenerator
Dimensions and Performance
The superficial gas velocity in a fluidized-bed regenerator
can be varied between two limits. Previous work has shown that the bed tends
to defluidize due to particle agglomeration at low gas velocities; at high
gas velocities, there is excessive carryover loss due to particle entrainment.
The range for superficial gas velocity is determined by coal-ash characteris-
tics, as well as by spent sorbent particle size, reducing gas concentration,
and bed temperature. Within the range allowed by these constraints, the
effect of superficial gas velocity on regenerator configuration can be pre-
dicted.
Selection of a solids feed rate and solids residence time
fixes the regenerator fluidized-bed volume. At a selected solids feed tem-
perature, gas inlet temperature, and regenerator bed temperature, the gas
residence time varies directly with solids residence time. It then follows
that at a constant solids residence time, the gas residence time cannot vary
if the off-gas S02 concentration is to remain the same. To utilize a higher
gas velocity and maintain the same gas residence time, the bed would have to
be made deeper and the bed cross-sectional area decreased since bed volume
cannot change (constant solids feed rate and constant solids residence time).
At constant gas and solids residence time, changes in the superficial gas
velocity could only be accomplished by changing the dimensions of the regen-
erator if the same SC>2 concentration is to be maintained in the off-gas.
For a low gas velocity, a regenerator would have a shallow bed and a large
cross-sectional area; for a high gas velocity, a regenerator would have a
deep bed and a small cross-sectional area. This is shown in Table 3. It
should be kept in mind that good fluidization. behavior still demands that a
bed have a height to diameter ratio within a certain range.
A deep bed has a greater pressure drop across it than a shal-
low bed does. This would increase the operating cost and the capital cost
for fans. Since the freeboard height is the main factor in determining the
height of the bed enclosure, increased bed height would not entail a signif-
icant capital-cost penalty for the regenerator. Decreased bed cross-sectional
area could significantly decrease capital cost. The superficial gas velocity
would have to be finally determined on an economic basis within the con-
straints of avoiding defluidization, avoiding excessive particle carryover,
and achieving good fluidization behavior.
-------
Table 3.
10
Predicted Effects of Superficial Gas Velocity
on Regenerator Dimensions
Combustor
635-MW AFBC
Greer Limestone
Sewickley Coal
83% SC>2 retention
FR - 0.20
Coal Feed Rate - 4779 Mg/d
Regenerator
Bed Temperature - 1373 K
Bed Pressure - 103 kPa
Gas Inlet Temperature - 672 K
Solids Feed Temperature - 1116 K
Solids Residence Time - 7 min
Superficial Gas Residence Time - 0.49 s
Extent of Regeneration - 64.6%
Solids Feed Rate - 3639 Mg/d
Coal Feed Rate - 142 Mg/d
Off-Gas S02 Concentrations - 8.03%
Superficial
Gas Velocity,
m/s
0.91
1.37
1.83
2.29
2.74
Expanded
Bed Height,
m
0.45
0.67
0.89
1.12
1.34
Regenerator
Bed Area,
m2
54.2
36.2
26.9
21.5
18.0
Regenerator
Bed Volume
m3
24.1
24.1
24.1
24.1
24.1
c. Predicted Effect of Solids Residence Time on Regenerator
Dimensions and Performance
Previously reported data^ obtained with the 10.8-cm-ID (4.25-in.)
regenerator showed the effect of solids residence time on the SC>2 concentra-
tion in the regenerator off-gas. Sulfur dioxide concentration in the off-gas
decreased with increased solids residence time. This result was obtained
with a bed having a fixed diameter and height. To vary the solids residence
time in the regenerator while operating at a fixed gas velocity, the solids
feed rate as well as the inlet gas 02 concentration had to be changed.
In the current analysis, the inlet gas is assumed to be air
(21% ©2), the reactor configuration is allowed to vary, and the gas velocity
is fixed at 1.37 m/s (4.5 ft/s). At an FR of 0.20 and with an air feed, the
S02 concentration in the regenerator off-gas was found to increase with
solids residence time, as shown in Table 4.
The effects of solids residence time on the requirements for expanded
bed height, bed cross-sectional area, solids feed rate, and regenerator coal
consumption (as a percentage of combustor coal consumption) are shown in
Table 4. In general, increasing the solids residence time increases the
required bed height substantially, changes the required bed cross-sectional
area slightly, decreases the required solids feed rate, and increases the
required regenerator coal usage. Increasing the solids residence time may
also increase sorbent sintering, which might lead to decreased sorbent
reactivity, but this has not been confirmed experimentally.
-------
Table 4. Predicted Effects of Solids Residence Time on
Regenerator Dimensions and Performance
Combustor
635-MW AFBC
Greer Limestone
Sewickley Coal
83% S02 Retention
FR - 0.20
Coal Feed Rate - 4779 Mg/d
Regenerator
Bed Temperature - 1373 K
Superficial Gas Velocity - 1.37 m/s
Gas Inlet Temperature - 672 K
Solids Feed Temperature - 1116 K
Pressure - 103 kPa
Solids
Residence
Time, min
3
5
7
10
15
30
Superficial
Gas Residence
Time , s
0.24
0.36
0.49
0.67
0.98
1.93
Extent of
Regeneration,
%
37.5
53.3
64.6
76.0
86.3
95.5
Regenerator
Solids
Feed Rate ,
Mg/d
4151
3803
3639
3515
3424
3359
Regenerator
Bed Height,
m
0.32
0.50
0.67
0.92
1.35
2.65
Regenerator
Bed Area,
m2
36.5
36.3
36.2
36.1
36.1
36.0
Regenerator
Coal Usage,
% of
Combustor
Coal Usage
2.77
2.90
2.97
3.02
3.06
3.10
Regenerator
Off-Gas S02
Concentration,
mol %
7.02
7.69
8.03
8.30
8.49
8.64
-------
12
d. Predicted Effects of Solids Feed Temperature and Gas Inlet
Temperature on Regenerator Dimensions and Performance
The effects of changes in solids feed temperature and gas inlet
temperature on requirements for regenerator bed height, bed cross-sectional
area, and coal usage, as well as on SC>2 concentration in the off-gas, are
shown in Table 5. In previous analyses, the solids feed temperature to the
regenerator was assumed to be 1116 K (1550°F) (solids are fed directly to the
regenerator from the combustor), and the gas inlet temperature was assumed to
be 672 K (750°F). (The feed gas is preheated by sending it through a heat
exchanger with the regenerator off-gas.) This results in a regenerator with
a cross-sectional area of 36.2 m2 (389 ft2), a bed height of 0.67 m (2.2
ft), an off-gas containing 8.0 mol % S02, and a coal consumption 2.97% that
of the combustor (7-min solids residence time, Table 4).
Table 5 shows that, if both the solids and the gas are fed into the
regenerator at 294 K (70°F), the regenerator must have a cross-sectional area
of 108.5 m2 (1168 ft2) and a bed height of 0.22 m (0.72 ft). The coal
consumption would increase to 9.07% that of the combustor, while the SC>2
concentration in the regenerator off-gas would decrease to 3.24 mol %. Con-
versely, if both the solids and gas are fed into the regenerator at its oper-
ating temperature of 1373 K (2012°F), the regenerator must have a cross-
sectional area of 13.0 m2 (140 ft2) and a bed height of 1.86 m (6.10 ft).
The coal consumption would decrease to 1.04% that of the combustor, while the
S02 concentration in the regenerator off-gas would increase to 16.85 mol %.
When the gas and solids are fed to the regenerator cold rather than
hot, more coal must be fed to supply the sensible heat necessary to raise the
temperature of the reactants. Similarly, additional air must be fed to react
with this additional coal; this dilutes the off-gas and therefore reduces the
SC>2 concentration in the off-gas. The total amount of SC>2 leaving the
regenerator is the same for all cases shown in Table 5.
When gas and solids are fed to the regenerator hot, coal combustion
must supply only the reducing gas plus the heat of reaction for the reductive
decomposition. This decreases coal consumption. Less air is needed to react
with this coal, and the S02 concentration in the off-gas increases because
of less dilution.
The bed volumes are the same for all cases shown in Table 5, and
solids feed rate and the residence time do not change. Since heating of the
inlet gas and solids takes no appreciable time (heating is almost instanta-
neous for the gas and on the order of seconds for the solids), the effective
gas and solids residence times for the reactions are not affected by the feed
temperatures. Therefore, all cases shown in Table 5 are kinetically and
thermodynamically equivalent with respect to reductive decomposition.
At the cold-feed condition in which a large amount of air is re-
quired to react with the extra coal, bed cross-sectional area must be in-
creased to keep the gas velocity constant. Conversely, under the hot-feed
conditions, considerably less air is required to react with the coal; bed
cross-sectional area is much smaller, but bed height must be increased to
maintain the regenerator volume constant. It should be noted that the
-------
13
Table 5. Predicted Effects of Solids Feed Temperature and Gas Inlet
Temperature on Regenerator Dimensions and Performance
Combustor
635 - MW AFBC
Greer Limestone
Sewickley Coal
83% S02 Retention
FR - 0.20
Coal Feed Rate - 4779 Mg/d
Regt -nerati > r
Solids Residence Time - 7 min
Regenerator Temperature - 1373 K
Superficial Gas Velocity - 1.37 m/s
Solids Feed Rate - 3639 Mg/d
Pressure - 103 kPa
Solids Feed
Temperature,
K
294
294
294
294
294
589
589
589
589
589
1116
1116
1116
1116
1116
1373
1373
1373
1373
1373
Gas Inlet
Temperature,
K
294
478
672
922
1373
294
478
672
922
1373
294
478
672
922
1373
294
478
672
922
1373
Regenerator
Bed Height,
m
0.22
0.25
0.28
0.31
0.39
0.28
0.31
0.34
0.39
0.48
0.54
0.60
0.67
0.76
0.93
1.07
1.19
1.32
1.50
1.86
Regenerator
Bed Area,
m2
108.5
97.8
87.1
76.9
62.4
87.1
78.4
70.1
61.9
50.1
44.8
40.2
36.2
31.8
25.9
22.6
20.3
18.3
16.1
13.0
Regenerator
Coal Usage,
% of Combustor
Usage
9.07
8.15
7.28
6.40
5.19
7.25
6.51
5.83
5.12
4.15
3.69
3.31
2.97
2.60
2.11
1.83
1.64
1.47
1.29
1.04
Regenerator
Off-Gas S02
Concentration,
mol %
3.24
3.52
3.90
4.32
5.15
3.90
4.25
4.67
5.18
6.16
6.76
7.38
8.03
8.89
10.42
11.51
12.46
13.44
14.66
16.85
-------
14
regenerator dimensions could be varied if changes of the gas velocity were
allowed. Within previously mentioned constraints (including the constraint
that all regenerator volumes be equal and not vary with gas velocity),
increased gas velocity would require a deeper bed of smaller area, while
decreased gas velocity would require a shallower bed of larger area.
From this analysis, it can be seen that the solids feed tempera-
ture and gas inlet temperature have a large effect on regenerator off-gas SC>2
concentration, regenerator coal consumption, and regenerator bed height and
cross-sectional area. Solids feed temperature has a greater effect on the
above variables than does gas inlet temperature since a greater amount of
sensible heat is required to heat the solids to regenerator conditions than
is required to heat the gas. It might be possible to preheat the feed gas to
a higher temperature than 672 K (750°F) by passing it through tubes immersed
in the combustor. Also, the solids might be preheated to more than 1116 K
(1550°F) by passing them through a carbon burnup cell. It should be noted
that this would result in no net change in coal consumption in the overall
system (combustor-regenerator) since the sensible heat of the reactants would
still have to be supplied. However, there would be an increase in SC>2
concentration in the regenerator off-gas that would result in lower gas flow
rates in the regenerator and sulfur-recovery system. Since the cost of the
sulfur-recovery system is the major capital equipment cost in the overall
regeneration system, an increase in SC>2 concentration in the off-gas should
result in a major cost savings.
e. Predicted Effect of Regenerator Bed Temperature on Regenerator
Dimensions and Performance
The effects of regenerator bed temperature and solids residence
time on extent of regeneration have been documented in a previous report.2
At constant solids residence time, the extent of regeneration increases with
higher bed temperatures; at a constant bed temperature, the extent of regen-
eration increases with longer residence times.
Table 6 illustrates the effect of bed temperature on regenerator
dimensions and performance. The combustor operates at the same conditions
for all cases. In the regenerator, the solids feed and gas inlet tempera-
tures, superficial gas velocity, and solids feed rate do not change; the
overall system FR is 0.20 for all cases; the extent of regeneration is made
the same (64% for all cases) by increasing solids residence time to compen-
sate for decreased operating temperature. In effect, changing the regenera-
tor bed temperature has no effect on the combustor; only the regenerator is
affected.
Solids residence time must be increased from 7 min at 1373 K
(2012°F), to 10.7 min at 1348 K (1967°F), and to 17.2 min at 1323 K (1922°F)
to keep the extent of regeneration constant at 64.6%. Because the solids
feed rate is constant, regenerator bed volume must be increased to increase
the solids residence time. Less coal would be burned to achieve the lower
temperature; therefore, less air would be fed to combust the coal. To main-
tain the superficial gas velocity in the regenerator at 1.37 m/s (4.5 ft/s),
the cross-sectional area of the bed would decrease from 36.2 m2 (389 ft2)
at 1373 K (2012°F), to 33.3 m2 (358 ft2) at 1348 K (1967°F) , and to 30.3 m2
(326 ft2) at 1323 K (1922°F). However, since regenerator bed height would
-------
Table 6. Predicted Effects of Regenerator Bed Temperature on Regenerator
Dimensions and Performance
Combustor
635-MW AFBC
Greer Limestone
Sewickley Coal
83% S02 Retention
FR - 0.20
Coal Feed Rate - 4779 Mg/d
Regenerator
Superficial Gas Velocity - 1.37 m/s
Gas Inlet Temperature - 672 K
Solids Feed Temperature - 1116 K
Solids Feed Rate - 3639 Mg/d
Extent of Regeneration - 64.6%
Pressure - 103 kPa
Bed
Temp,
K
1373
1348
1323
Solids
Residence
Time,
min
7.0
10.7
17.2
Superficial
Gas Residence
T ime , s
0.49
0.81
1.42
Regenerator
Bed Height,
m
0.67
1.11
1.95
Regenerator
Bed Area,
m2
36.2
33.3
30.3
Regenerator
Coal Usage,
% of Combustor
Usage
2.97
2.78
2.60
Regenerator
Off-Gas S02
Concentration,
mol %
8.03
8.46
8.96
-------
16
increase from 0.67 m (2.2 ft) at 1373 K (2012°F), to 1.11 m (3.64 ft) at
1348 K (1967°F), and to 1.95 m (6.40 ft) at 1323 K (1922°F), superficial gas
residence time must also be increased to provide the necessary time for
reaction at the lower temperatures. Coal feed rate would decrease from 2.97%
of that of the combustor at 1373 K (2012°F), to 2.78% at 1348 K (1967°F), and
to 2.60% at 1323 K (1922°F). Because there would be less dilution at the
lower bed temperatures, the S02 concentration in the off-gas would increase
from 8.03% at 1373 K (2012°F), to 8.46% at 1348 K (1967°F), and to 8.96% at
1323 K (1922°F).
Regenerator performance would improve (i.e., there would be less
coal consumption and a higher off-gas SC>2 concentration) at lower tempera-
tures. Bed area would decrease somewhat, but bed height would increase sub-
stantially. This might result in a cost savings for the regenerator itself,
but fan power would have to be increased. With a lower-temperature regen-
erator, the use of lower-cost construction materials and methods might be
possible, sintering and agglomeration might decrease, and there might be less
reactivity loss by the sorbent. Higher off-gas 862 concentrations would
lower the cost of the sulfur-recovery system. It should be noted, however,
that no experiments have been done at temperatures lower than 1323 K (1922°F).
At temperatures below this, the concentration of S02 in the off-gas may be
limited by thermodynamic equilibrium considerations.
f. Predicted Effect of Pressure on Regenerator Dimensions and
Performance
At low pressures (lower than 206 kPa or approximately 2 atm), the
effect of increasing the pressure in the regenerator is the same as the ef-
fect of increasing the superficial gas velocity. That is, a change in design
pressure has no effect on regenerator performance. An increase in pressure
calls for a deeper regenerator bed with a smaller cross-sectional area; the
bed volume remains the same. A decrease in pressure would call for a shal-
lower but larger-bed cross section. This is illustrated in Table 7. These
results cannot be extrapolated to high pressures since thermodynamic equilib-
rium would limit the concentration of SC>2 obtainable.
-------
17
Table 7. Predicted Effect of Pressure on Regenerator Dimensions
Combustor
635-MW AFBC
Greer Limestone
Sewickley Coal
FR - 0.20
Coal Feed Rate - 4779 Mg/d
Regenerator
Bed Temperature - 1373 K
Gas Inlet Temperature - 672 K
Solids Feed Temperature - 1116 K
Superficial Gas Velocity - 1.37 m/s
Solids Residence Time - 7 min
Extent of Regeneration - 64.6%
Solids Feed Jlate - 3§39 Mg/d
Coal Feed Rate - 142 Mg/d
Off-Gas S02 Concentration - 8.03%
Condition
la
Ib
Ic
Id
Pressure,
kPa
52
103
155
206
Regenerator
Bed Area,
m2
72.7
36.2
24.0
18.0
Regenerator
Bed Height,
m
0.33
0.67
1.01
1.34
Regenerator
Bed Volume,
m
24.1
24.1
24.1
24.1
g. Predicted Effect of Changes in Feed-Gas Oxygen Concentration on
Regenerator Dimensions and Performance
In general, air would be the feed gas to a large-scale regenerator.
However, changes in feed-gas oxygen concentration were evaluated as to their
effect on regenerator dimensions and performance as a preliminary step in a
possible economic study of using oxygen-enriched feed gas in a regenerator.
As can be seen in Table 8, oxygen enrichment of the feed gas results in an
increase in off-gas S02 concentration and a decrease in coal consumption.
Coal consumption would decrease because less sensible heat would be needed to
preheat the inert nitrogen. Off-gas S02 concentration would increase be-
cause of less dilution by nitrogen. Bed cross-sectional area would decrease
(at constant superficial gas velocity) because less inert nitrogen would pass
through the bed. Bed height would increase to maintain a constant bed volume
(constant solids residence time) and provide the longer gas residence time
necessary to obtain high off-gas S02 concentrations.
Oxygen enrichment of the feed gas could increase the off-gas S02
concentration considerably. This would bring about a great cost savings in
the sulfur-recovery system (RESOX or other), which is the major capital cost
item in a regeneration system. Therefore, an overall economic evaluation
should be made of using oxygen-enriched feed gas for the regenerator.
-------
Table 8. Predicted Effect of Changes in Feed-Gas Oxygen Concentration on Regenerator
Dimensions and Performance
Combustor
635-MW AFBC
Greer Limestone
Sewickley Coal
FR - 0.20
Coal Feed Rate - 4779 Mg/d
Regenerator
Bed Temperature - 1373 K
Gas Inlet Temperature - 672 K
Solids Feed Temperature - 1116 K
Solids Residence Time - 7 min
Extent of Regeneration - 64.6%
Solids Feed Rate - 3639 Mg/d
Pressure - 103 kPa
Superficial Gas Velocity - 1.37 m/s
Inlet Gas 02
Concentration,
%
15.1
21.0
40.0
60.3
80.1
99.2
Superficial Gas
Residence Time,
s
0.29
0.49
1.12
1.79
2.44
3.08
Regenerator
Bed Area,
m2
60.9
36.2
15.8
9.8
7.2
5.7
Regenerator
Bed Height,
m
0.40
0.67
1.53
2.45
3.35
4.22
Regenerator
Off-Gas S02
Concentration
mol %
5.22
8.03
14.66
19.33
22.53
24.86
Regenerator
Coal Usage,
% of Combustor
Usage
3.66
2.97
2.40
2.23
2.16
2.12
00
-------
19
3. Predicted Effects of Off-Design Conditions on Regenerator Operating
Parameters
In this section, the effects of changes in operating variables on
performance of an operating reactor with fixed bed height and cross-sectional
area are discussed. The effects of changing the solids feed rate, solids
feed temperature, gas inlet temperature, and regenerator bed temperature will
be explored for a regenerator which could be designed to operate with a
635-MW AFBC using Greer limestone and Sewickley coal at an FR of 0.20. For
the baseline conditions (case la, Table 9) of an operating temperature of
1373 K (2012°F), a solids feed temperature of 1116 K (1550°F), a gas inlet
temperature of 672 K (750°F), a solids residence time of 7 min, a solids feed
rate of 3639 Mg/d, and a superficial gas velocity of 1.37 m/s (4.5 ft/s), the
regenerator will (1) have a bed height of 0.67 m (2.2 ft) and an area of
36.2 m2 (389 ft2), (2) require 142 Mg/d of coal, and (3) have an off-gas
S02 concentration of 8.03%.
a. Predicted Effects of a Lower Solids Feed Rate on Regenerator
Operating Parameters
Cases la, Ib, Ic, Id, and le in Table 9 illustrate the effects on
regenerator operating parameters of turning the combustor down (reducing the
electrical and/or steam production of the system). Reducing the load on the
combustor would reduce the solids feed rate to the regenerator. Since the
regenerator volume would be constant (the volume cannot be altered during
operation), the solids residence time would increase, which would mean a
concomitant increase in gas residence time. Since bed height would remain
constant, superficial gas velocity must be lowered. Coal feed rate would
decrease, while off-gas S02 concentration would increase.
Decreasing the combustor output by 50% would decrease the solids
feed rate to the regenerator from 3639 Mg/d to 1713 Mg/d; simultaneously, the
coal feed rate to the regenerator would be reduced from 142 Mg/d to 73 Mg/d
and the SC>2 concentration in the off-gas would increase from 8.03 to 8.48%.
The superficial gas velocity would decrease from 1.37 m/s (4.5 ft/s) to
0.68 m/s (2.23 ft/s). This lower velocity might result in defluidization.
The amount a system can be turned down will be determined by the relation
of operating velocity to defluidization velocity.
-------
Table 9 Predicted Effects of Off-Design Conditions (Solids Feed Rate, Gas Inlet Temperature,
Solids Feed Temperature, and Regenerator Bed Temperature) on Regenerator Performance
Combustor
635-MW AFBC
Greer Limestone
Sewickley Coal
FR - 0.20
Coal Feed Rate - 4779 Mg/d
Condition
la
Ib
Ic
Id
le
Ila
lib
He
Ilia
II Ib
IIIc
Hid
IV a
IVb
Va
Vb
Combustor
Output ,
%
100
90
75
50
33
100
100
100
100
100
' 100
100
100
100
65
41
Solids
Residence
Time,
min
7.0
7.9
9.6
14.9
22.6
7.0
7.0
7.0
7.0
7.0
7.0
7.0
7.0
7.0
10.7
17.2
Superficial
Gas Residence
Time,
s
0.49
0.54
0.65
0.98
1.47
0.96
0.49
0.25
0.68
0.55
0.49
1.44
0.16
1.36
0.81
1.42
Regenerator
Bed Temp. ,
K
1373
1373
1373
1373
1373
1373
1373
1373
1373
1373
1373
1373
1373
1373
1348
1323
Solids
Feed
Temp. ,
K
1116
1116
1116
1116
1116
1373
1116
589
1116
1116
1116
1116
294
1373
1116
1116
Gas Inlet
Temp . ,
K
672
672
672
672
672
672
672
672
1373
922
672
478
294
1373
672
672
Regenerator
Bed Height - 0.67 m
Bed Area - 36.2 m2
Pressure - 103 kPa
Regenerator
Solids
Feed Rate,
Mg/d
3639
3233
2644
1713
1125
3639
3639
3639
3639
3639
3639
3639
3639
3639
2374
1485
Superficial
Gas Velocity,
m/s
1.37
1.23
1.03
0.68
0.45
0.69
1.37
2.68
0.98
1.21
1.37
1.53
4.17
0.49
0.82
0.47
Regenerator
Coal Usage,
% of
Combustor
Usage
2.97
2.99
3.01
3.06
3.09
1.47
2.97
5.83
2.11
2.60
2.97
3.31
9.07
1.04
2.78
2.60
Regenerator
Coal Feed,
Mg/d
142
129
108
73
49
70
142
279
101
124
142
158
433
50
87
51
Regenerator
Off-Gas
S02 Cone . ,
mol %
8.03
8.14
8.28
8.48
8.60
13.44
8.03
4.67
10.42
8.89
8.03
7.38
3.24
16.70
8.46
8.96
KJ
O
-------
21
b. Predicted Effects of Solids Feed Temperature on Regenerator
Operating Parameters
Cases Ila, lib, and lie in Table 9 illustrate the effects of sol
feed temperature on regenerator operating parameters. Case lib is the
at a gas feed temperature of 672 K (750'F) would reduce the amount of sen-
sible heat that must be provided by coal combustion, thereby reducing the
regenerator coal feed rate from 142 Mg/d to 70 Mg/d. Less air Would be
needed to combust this quantity of coal; since the bed cross-sectional area
would be constant, the superficial gas velocity would be decreased from
1.37 m/s (4.5 ft/s) to 0.69 m/s (2.26 ft/s). Because there would be less
dilution, the off-gag S02 concentration would increase from 8.03 to _ij.
-------
22
Case la in Table 9 is considered to be the baseline condition.
Steam and/or electrical production in the plant would be decreased to 65% of
the baseline condition for operation at 1348 K (1967°F) and to 41% of base-
line for operation at 1323 K (2012°F). Regenerator coal consumption would
decrease from 142 Mg/d at 1373 K (2012°F) to 87 Mg/d at 1348 K (1967°F) and
to 51 Mg/d at 1323 K (1922°F). Sulfur dioxide concentration in the off-gas
would increase from 8.03% at 1373 K (2012°F) to 8.46% at 1348 K (1967°F) and
to 8.96% at 1323 K (1922°F), while gas velocity would decrease from 1.37 m/s
(4.5 ft/s) at 1373 K to 0.82 m/s (2.69 ft/s) at 1348 K (1967°F) and to 0.47 m/s
(1.54 ft/s) at 1323 K (1922°F).
e. Design of a Regenerator to Accommodate Turndown
For operation of a regenerator at reduced load (lower solids feed
rate), the superficial gas velocity must be decreased. However, the gas
velocity must always remain above the defluidization velocity (below the
defluidization velocity, the bed material does not fluidize but agglomerates);
this limits the turndown that a system can tolerate. In Table 10, the mini-
mum gas velocity is assumed to be 1.37 m/s (4.5 ft/s). Case I in Table 10
provides for a maximum of 25% turndown—i.e., the system will reach minimum
gas velocity at a system load of 75%. Case II provides for 50% turndown,
while case III provides for 67% turndown. This is accomplished by adjusting
the bed configuration to accommodate a higher superficial gas velocity under
full-load conditions. For no turndown capability (see case Ib in Table 7),
the bed height is 0.67 m (2.2 ft), the bed cross-sectional area is 36.2 m2
(389 ft2), and the superficial gas velocity is 1.37 m/s (4.5 ft/s). The
bed height would be increased from 0.67 m (2.2 ft) to 0.89 m (2.92 ft) to
accommodate 25% turndown, to 1.34 m (4.40 ft) for 50% turndown, and to 2.02 m
(6.65 ft) for 67% turndown. The bed cross-sectional area would decrease from
36.2 m2 (389 ft2) to 27.1 m2 (292 ft2) to accommodate 25% turndown, would
decrease to 17.9 m2 (193 ft2) for 50% turndown, and to 12.0 m2 (129 ft2)
for 67% turndown. The full-load superficial gas velocity would increase from
1.37 m/s (4.5 ft/s) to 1.82 m/s (5.97 ft/s) to accommodate 25% turndown, to
2.73 m/s (8.96 ft/s) for 50% turndown, and to 4.12 m/s (13.5 ft/s) for 67%
turndown. The solids feed rate, the solids residence time, the gas residence
time, the bed volume, the coal consumption, and the S02 concentration in
the off-gas are the same for the full-load part of each case (la, Ila, Ilia).
Excessive particulate carryover as a result of too high a superficial gas
velocity and excessive pressure drop from too deep a bed are the practical
limitations on regenerator turndown capability.
-------
Table 10. Design of Regenerator to Accommodate Turndown
Combustor
635-MW AFBC
Greer Limestone
Sewickley Coal
FR - 0.20
Coal Feed Rate - 4779 Mg/d
Regenerator
Bed Temperature - 1373 K
Gas Inlet Temperature - 672 K
Solids Feed Temperature - 1116 K
Minimum Superficial Gas Velocity - 1.37 m/s
Pressure - 103 kPa
Combustor
Condition
la
Ib
Ic
Ha
lib
He
lid
HIa
IHb
IIIc
Hid
Hie
Output
100
90
75
100
90
75
50
100
90
75
50
33
Solids
Residence
Time,
min
7.0
7.9
9.6
7.0
7.9
9.6
14.9
7.0
7.9
9.6
14.9
22.6
Superficial
Gas Residence
Time ,
s
0.49
0.54
0.65
0.49
0.54
0.65
0.98
0.49
0.54
0.65
0.98
1.47
Regenerator
Solids
Feed Rate,
Mg/d
3639
3233
2644
3639
3233
2644
1713
3639
3233
2644
1713
1125
Regenerator
Bed Height,
m
0.89
0.89
0.89
. 1.34
1.34
1.34
1.34
2.02
2.02
2.02
2.02
2.02
Regenerator
Bed Area,
m2
27.1
27.1
27.1
17.9
17.9
17.9
17.9
12.0
12.0
12.0
12.0
12.0
Superficial
Gas Velocity,
m/s
1.82
1.65
1.37
2.73
2.48
2.06
1.37
4.12
3.74
3.11
2.06
1.37
Regenerator
Coal Feed
Rate,
Mg/d
142
129
108
142
129
108
73
142
129
108
73
49
Regenerator
Coal Usage,
% of
Combustor
Usage
2.97
2.99
3.01
2.97
2.99
3.01
3.06
2.97
2.99
3.01
3.06
3.09
Regenerator
Off-Gas
SO2, Turndown
tool X Desired
8.03
8.14
8.28
8.03
8.14
8.28
25%
50%
8.48 ,
i
8.03
8.14
8.28
8.48
8.60
67?
-------
24
4. Comparison of New and Earlier Predictions of Effects of Process
Parameters
Predictions presented earlier in this report (Sections 2 and 3) differ
in certain instances from predictions made in previous reports. This section
explores the new predictions, compares them with previous predictions in
certain instances and explains the differences.
Since the volume of the regenerator used in these studies is fixed (con-
stant bed area and expanded bed height), the minimum number of variables that
can be simultaneously studied is two. For instance, the effect of solids
residence time may be studied only if (1) superficial gas velocity, (2) re-
generator bed temperature, or (3) feed-gas oxygen concentration is also
varied. The solids residence time in a fixed-volume reactor can only be in-
creased by decreasing the solids feed rate; this simulataneously lowers the
sensible heat requirement, and a lower coal feed rate and less combustion air
are needed. Combustion air can be decreased in two ways: by lowering gas
velocity or by lowering oxygen concentration in the feed gas. Since lowering
of gas velocity might alter fluidization characteristics or cause agglomera-
tion of material by defluidization, feed-gas oxygen concentration was chosen
as the operating variable.
For conditions which call for increasing the amount of combustion air,
doing so by increasing the gas velocity might lead to excessive particle
carryover; instead, oxygen concentration in the feed gas can be varied.
Changing the regenerator bed temperature (the second variable) would
affect gas-solid kinetics and would not be a reasonable approach to estab-
lishing the effect of solids residence time on the extent of regeneration and
off-gas SC>2 concentration.
A model has been developed from the PDU experimental results and has
been used to predict the results of other experiments; agreement has been
good.l In all experimental work, the feed gas oxygen concentration has
been varied. This a useful experimental tool, but for the design of larger
systems, the feed gas should be air, fixing the feed-gas oxygen concentration
at 21%. Results reported previously,2 as well as results presented earlier
in this report (Sections 2 and 3), all assume that the feed gas is air.
a. Relation of Solids Residence Time and Gas Residence Time to
Regenerator Performance
The relationship of solids residence time to superficial gas resi-
dence time at a regenerator temperature of 1373 K (2012°F), a solids feed
temperature of 1116 K (1550°F), and a gas inlet temperature of 672 K (750°F)
for a conceptual 635-MW regenerator may be seen in Table 4. Analogous tables
could be made for different regenerator temperatures and different solids
feed and gas inlet temperatures. Also, relationships between solids residence
time, regenerator temperature, and extent of regeneration have been previ-
ously reported.! jn comparing regenerator performance, it should be kept
in mind that two regenerators that are operated at the same temperature, with
the same solids feed and gas inlet temperatures, and with the same gas and
solids residence time are kinetically equivalent. That is, the extent of re-
generation, SC>2 concentration in the off-gas, and coal/solids feed ratio
-------
25
will be identical. Larger regenerators, such as that assumed in Table 4,
would have slightly lower heat losses and therefore would be more efficient
(have higher off-gas SC>2 concentrations), but this effect would be small in
practical-size units.
b. Relation of Bed Height to Superficial Gas Velocity
This brings us to the relationship of bed height to superficial gas
velocity. Once a gas residence time has been fixed, expanded bed height is
directly proportional to superficial gas velocity. This is shown in Table 3
for a conceptual 635-MW combustor. Since the mass flow rate of gas remains
constant, bed cross-sectional area decreases. The solids residence time is
not changed thereby since the volume of the bed remains constant. Earlier
work^ in which bed height was increased ran into two problems: (1) only
the height of the reducing zone was increased; the height of the oxidizing
zone remained constant, and (2) increasing the height of a very-small-cross-
sectional-area bed decreased the quality of fluidization (already poor) still
further. The first problem has not been solved; it is hoped that the reduc-
ing and oxidizing zones can be made the appropriate sizes by adjusting the
reducing gas concentration in the off-gas. The second problem should be
solved by building larger regenerators for which the bed height/diameter
ratios are not in the "slugging" regime.
Superficial gas velocity must be maintained between the lower ve-
locity limit set by defluidization (which results in agglomeration) and the
upper velocity limit set by particle carryover. In previous studies,2
increases in gas velocity were found to be detrimental to performance. This
was probably due to degradation of already-poor fluidization characteristics
of the experimental PDU. Also, an increase in superficial gas velocity not
accompanied by an increase in bed height decreases gas residence time; this,
in turn, adversely affects bed performance.
c. Relation of Solids Residence Time to Off-Gas S02 Concentration
Sulfur dioxide concentration in the regenerator off-gas is predict-
ed to increase with longer solids residence time (Table 4). In contrast, a
previous report^ implies that shorter solids residence times produce higher
off-gas S02 concentrations. In earlier work, the experimental solids
residence time was decreased by increasing the solids feed rate (at a fixed
regenerator volume). This increased the amount of coal that had to be burned
to provide sensible heat to the solids. Also, the oxygen concentration in
the feed gas was increased so that the superficial gas velocity would not
have to be increased. Since less nitrogen was present to dilute the off-gas,
a higher S02 concentration resulted. If the S02 concentration obtained
in the earlier work should be corrected to a 21% oxygen-feed-gas basis and
the regenerator configuration should be "readjusted" to provide the correct
gas residence time, S02 concentration in the off-gas would increase with
longer solids residence times.
-------
26
d. Relation of Solids Feed Temperature and Gas Feed Temperature to
Off-Gas SC>2 Concentration
Sulfur dioxide concentration in the regenerator off-gas has been
predicted to increase with solids feed temperature and gas inlet tempera-
tures (Table 5). In comparison, previous work1 implies that this effect
is marginally important. If the solids residence time is fixed, the amount
of sensible heat which must be supplied is inversely proportional to the
solids feed temperature and gas feed temperatures. When more sensible heat
was required in the experimental PDU, more coal was fed and a higher concen-
tration of oxygen was used in the feed gas (the superficial gas velocity was
constant). Therefore, the (cold-feed) runs which required the most preheat
were diluted the least. Conversely, the runs which required the least pre-
heat (hot-feed) suffered the most dilution. These effects tended to cancel
each other, and at all feed conditions, SC>2 concentrations in the off-gas
were similar.2 If air is the only feed gas permitted and the regenerator
configuration is adjusted appropriately, SC>2 concentration in the off-gas
would increase with solids feed temperature and gas inlet temperature, as
shown in Table 5.
e. Relation of Regenerator Bed Temperature to Off-Gas S02
Concentration
Previous work2 predicted that off-gas S02 concentration would
decrease with lower regenerator bed temperature. The extent of regeneration
increases with longer residence times at all temperatures of interest. How-
ever, if there is a constant solids feed rate in the regenerator irrespective
of temperature and if the regenerated product has to be returned to the com-
bustor at the same composition, lower-temperature regeneration would result
in a higher off-gas S02 concentration. A longer solids residence time must
be utilized at lower regeneration temperatures in order to keep the same sol-
ids feed rate, composition, reactivity, and extent of regeneration. Since a
lower regeneration temperature requires that less sensible heat be added, less
coal and therefore less air must be fed to the regenerator. Less dilution of
the off-gas results in higher S02 concentrations in the off-gas. This is
illustrated in Table 6. However, bed volume would increase (to obtain a
longer solids residence time); the bed would be deeper and somewhat smaller
in cross-sectional area.
In a system which utilized the same solids residence time at the
temperatures being compared, lower S02 concentrations would be obtained in
the off-gas at lower regeneration temperatures. In addition, higher solids
circulation rates would have to be maintained between the combustor and the
regenerator at the lower regenerator temperature. At the lower temperature,
regenerator cross-sectional area would have to decrease and regenerator bed
height increase.
-------
f. Conclusion
In conclusion, the PDU experimental data are valid and are useful
for design purposes if all conditions are corrected to correspond to a feed
gas containing 21% oxygen. The data relating solids residence time, gas
residence time, solids feed temperature, gas inlet temperature, regenerator
temperature, and superficial gas velocity to regenerator dimensions and per-
formance can be utilized to design regenerators within operating and economic
constraints.
5. Economic Feasibility of Regenerating Sulfated Limestone
An economic evaluation of regeneration of spent sorbent from fluidized-
bed combustion was performed,^ using a Westinghouse cost study along with
data from ANL cyclic combustion and regeneration experiments with Greer
limestone and Sewickley coal. The system assumed was an atmospheric-pressure,
fluidized-bed reductive-decomposition process, followed by a RESOX sulfur-
recovery system with tail-gas incineration.
Capital cost for the system, capacity factor, and sorbent price (includ-
ing sorbent disposal cost) were found to be important factors; fuel price and
sulfur credit value were found to be of lesser importance. At today's
capital cost, a 70% plant capacity factor, $27.56/Mg coal price, and with no
sulfur credit, the sorbent regeneration system would break even at a total
sorbent cost of $l2.68/Mg. Further results were presented for values of all
of the above parameters.
A computer program was developed to incorporate cyclic combustion and
regeneration experimental data for a given coal and a given sorbent in order
to predict the economic viability of regeneration with that particular coal
and sorbent. A second computer program was written to predict the costs of a
regeneration system when system size, various economic factors, coal type,
sorbent type, and preliminary data (or assumptions) on sorbent reactivity are
given. More detailed results have been previously reported.
-------
28
TASK B. ALTERNATIVE REGENERATION PROCESS DEVELOPMENT
(D. S. Moulton)
1. Rotary-Kiln Regeneration
Previous emission-control studies have shown that spent limestone sor-
bents can be regenerated, using reducing agents. Rotary kilns are being
evaluated for use as regeneration reactors. They can be externally fired,
allowing reaction heat to be supplied without diluting the off-gas with
combustion products. An externally fired rotary kiln has a potential for
producing S02 in a concentrated stream. The ease of handling elemental
sulfur, and the utilities' traditional avoidance of chemical processing,
makes elemental sulfur an attractive final form of the sulfur pollutants.
Concentrating the S02 stream simplifies the subsequent reduction to
elemental sulfur and greatly improves the economics.
Studies are being conducted with a small laboratory-scale rotary kiln
in support of further work with a larger industrial kiln. These studies are
(1) to demonstrate rotary-kiln regeneration on a laboratory scale and (2) to
provide information on kiln performance under various operating conditions.
The S02 concentration in the off-gas is measured, and the extent of regen-
eration is determined by sulfur analysis of the solid product.
a. Equipment
The laboratory-scale rotary kiln was constructed from a 26-mm-ID
tube of fused silica. Heat was supplied by a 40-cm annular furnace, and the
tube was rotated by a variable-speed motor via gear reductions and a pulley.
The rotation rate was variable—between about 4 and 80 rpm. The kiln was
fitted with rotary seals at each end. Solids feed rate was controlled by a
rotating bucket driven by a variable-speed motor. The nitrogen carrier gas
could be introduced at either end of the kiln and made to flow either
cocurrently or countercurrently to the solids flow. Countercurrent flow
operation has been difficult to achieve, and most of the data has been
obtained using cocurrent flow.
A 7-mm quartz gas-sampling tube was used to withdraw part of the
off-gas from different points within the kiln. The tube extended into the
kiln from the end, just above the solid reactants. The S02 content was
measured with a Thermo Electron Corporation Model 40 pulsed fluorescence
analyzer. Off-gas for analysis was drawn into the tube and was diluted
with house nitrogen to bring the percent S02 within the range of the
analyzer.
b. Materials and Procedure
Greer limestone, which had been sulfated in a fluidized-bed com-
bustor, was supplied by Pope, Evans, and Robbins. The reductant was char
from Occidental Research Corporation's flash-pyrolysis process. An S02/N2
mixture used for analyzer calibration was obtained from Matheson and certi-
fied as 3.02% S02.
-------
29
The sulfated stone was analyzed for sulfate content by titration
with a solution of barium perchlorate in isopropanol, using thorin and meth-
ylene blue as indicator. The stone was mixed with the char before it was
added to the feeder. The mixing ratio was based on the sulfate analysis of
the stone and the fixed carbon content of the char and was stoichiometric for
the reaction:
2CaSC>4 + C •*• 2CaO + 2SC>2 + CC>2
except where noted otherwise. After passing through the kiln, samples of the
solid product were analyzed for sulfate content by the above method. Some
sulfide is formed by the following reaction:
CaS04 + 2C * CaS + 2CC>2
The sulfide content was obtained from the difference between the sulfate
content before and after oxidation at 850 °C.
Preliminary work indicated that low rotation rates and small incli-
nation angles were required to produce a solids residence time sufficient for
regeneration. The kiln was generally operated at 4-10 rpm and at an inclina-
tion angle of 0.5 to 2.0 deg.
c. Kiln Operation
The laboratory-scale kiln was operated at room temperature to mea-
sure solids residence time under various operating conditions. Whatley-* has
shown that the ratio of the dimensionless velocity number to the slope is
nearly constant over a wide operating range:
const = C
v
where a is the kiln's angle of inclination measured from the horizontal, Cv
is here referred to as the velocity ratio, and 1TV is the velocity number
given by:
1 = —
v u)D
The solids velocity, v, is the kiln length, L, divided by the residence time,
oi is the kiln rotation rate, and D the inside diameter.
Room temperature studies were made to investigate the above rela-
tionships. Residence time was determined from the mass flow rate and the
total amount of material in the kiln. Cy measurements were made at various
angles of inclination and the results, shown in Fig. 2, indicate that Cv is
independent of the angle of inclination only when the angle exceeds about
three degrees.
-------
30
10
cc
o
o
> 2
0
I
I
I
I
0 I 23456
INCLINATION ANGLE,deg
Fig. 2. Velocity Ratio vs. Inclination Angle
in a Laboratory-Scale Rotary Kiln
The residence time, t, is given by
L
t" —
uDC t an a
v
For inclination angles greater than three degrees, Cv can be taken as 3.1;
otherwise, Cv is obtained from Fig. 2. Griswold^ gives a similar equation
which includes the angle of repose of the dry material. Whatley's treatment
was used here because it is based on a dimensionless analysis and may be more
useful for scale-up. Figure 2 was obtained for sulfated limestone having a
45° angle of repose, and Cy values obtained from the graph may have to be
adjusted for material having significantly different angles of repose.
-------
31
d. Results and Discussion
Four runs were made with the SC>2 sampling tube within the hot
zone. The carrier-gas flow was adjusted to obtain the maximum percent S02,
and was usually about 0.2 L/min. The results (Table 11) show that high S02
concentrations occur within the kiln. Gas grab samples taken just downstream
from the kiln for mass spectrometric analysis were in general agreement with
S(>2 levels indicated by the pulsed fluorescence analyzer. Data from the
last three runs (Table 11) indicate that S02 concentrations near the
equilibrium value of about 20%? can be achieved within the kiln.
Table 11. Regeneration of Sulfated Greer Limestone
with Char in a Small Rotary Kiln
Run
780414
780418
780419
Temperature,
°C
1035
1040
1040
Operating
Time ,
min
60
40
120
Residence
Time,
min
19
33
60
Regeneration,
42
20
37
S02
Concentration,
in Off-Gas
lia.b
20b
22b
780425
1040
180
16
23
29b
a
Sulfur dioxide reacted with the stainless steel sampling tube. Quartz
sampling tubes were used in subsequent runs.
Measured with pulsed fluoresence analyzer.
"Measured with mass spectrometer.
SOLIDS AND
CARRIER
GAS INLET
FURNACE
B
Fig. 3. Gas-Sampling Locations and Reactor Temperatures in
the Laboratory-Scale Rotary Kiln Operated at 1040*C
and 4 rpm.
A - 10 cm from end of heated zone, 1040°C
B - 5 cm from end of heated zone, 1005°C
C - Outlet, room temperature
-------
32
In rotary kilns, it is difficult to separate the gas from the solid
at high temperature. Ordinarily, the product gases circulate over some solid
materials which are cooler than the reaction temperature before leaving the
kiln. This could lead to recombination of the SC>2 with the cooler material,
and experiments were made to see if this was a significant limitation.
The kiln was operated at 4 rpm and 1040°C with a solids residence
time of about 20 min. Three different gas-sampling points, shown in Fig. 3,
were used in different runs. The carrier-gas flow rate was varied during
each run, to obtain a plot of percent S(>2 versus flow rate at each sampling
point.
The results are shown in Fig. 4. At low carrier-gas flow rates,
the SC>2 concentration inside the kiln approached 20%, approximately the
equilibrium value. However, by the time the gas reached the kiln outlet,
most of the SC>2 had recombined with the solid. At higher flow rates, there
(VI
O
CO
LJ
O
(E
LU
CL
A Gas Sampled at A, I040°C
o Gas Sampled at B, I005°C
Gas Sampled at C (Outlet)
Room Temperature
4 -
0
0.8 1.2
FLOW RATE, L/min
Fig. 4. Variation of Percent S(>2 in Off-Gas with
Carrier-Gas Flow Rate, during Limestone
Regeneration in a Small Rotary Kiln at
1040°C. Sampling locations are shown in
Fig. 4.
-------
33
was not sufficient time for the gas in the kiln to reach a high S(>2 concen-
tration; however, there was also less S02 recombination during exit. In
the laboratory-scale kiln, recombination with the solid was insignificant
only at the higher carrier gas flow rates where the SC>2 concentration
reached less than one-half the equilibrium value.
The amount of recombination in a larger kiln may differ from the
above results. An increased gas residence time should allow high internal
S02 concentrations at higher gas flow rates. If the solids are heated
rapidly so that the lower-temperature material occupies only a short section
of the kiln or if the incoming solids are saturated with SC>2, the SC>2
stream might exit from the reactor without much recombination occurring.
Otherwise, recombination may limit the maximum S02 concentration in the
off-gas.
High rates of regeneration occur with low percentages of SC>2 in
the off-gas, when the carrier-gas flow rate is high. This effect is shown in
Table 12. At the higher flow rates, recombination was not significant, and
regeneration was more complete in the same residence time. However, under
these conditions, an off-gas stream with a high concentration of S(>2 is not
obtained. Production of a concentrated 862 stream requires that recombina-
tion be minimized, regardless of the type of regeneration reactor used. It
should be an objective in regenerator design to cause the SC>2 stream to
flow past any lower-temperature solids rapidly enough to prevent recombina-
tion or else to separate the gas from the solids at a high temperature.
Table 12. Regeneration at 1040°C of Spent Limestone Sorbent
in a Small Rotary Kiln
Sample
780522
780523
780525
780621
Carrier-Gas Flow,
L/min
0.5
1.0
1.0
1.0
Sulfate
Conversion
to Oxide,
%
33
55
64
19
Sulfate
Conversion
to Sulfide,
%
5
6
18
48a
Excess carbon in the feed—i.e., 1.5 times stoichiometric.
-------
34
TASK D. EFFECTS OF LIMESTONE SULFATION ACCELERATORS ON
CORROSION RATES OF METALS IN AN AFBC
1. Atmospheric Combustor Facility for Corrosion Tests
(J. Lenc, G. Smith, F. G. Teats, and R. Mowry)
This work to study the effects of limestone sulfation accelerators on
corrosion rates of metal components of an AFBC is related to a laboratory-
scale investigation to increase the degree of sulfation of partially sulfated
lime solids within a fluidized-bed combustor by means of additives (i.e.,
sulfation accelerators). Sulfation accelerators such as NaCl, CaCl2, and
Na2C03 added in small amounts to the lime solids increase both the rate
and the extent of sulfation for many limestones.1
Increased S02 capacity of limestone would mean a decrease in the quan-
tity of lime solids required for the combustion process. Such a decrease in
the lime solids requirement would lower the process cost and reduce the
environmental impact of solids waste disposal. However, there is concern
that volatilization of these sulfation accelerators (alkali metal compounds)
might cause unacceptable corrosion of the metal components of the combustion
system. A separate laboratory-scale investigation of the corrosiveness of
salts mixed with sulfated limestones is in progress.1 The following
section of this report (section 2) summarizes corrosion experiments done to
date in laboratory-scale apparatus.
a. Description of Atmospheric Combustor Facility for Corrosion Tests
To measure the corrosion rates of metals of construction in the
presence of sulfation accelerators in a PDU-scale unit, a new automated
atmospheric-pressure fluidized-bed coal combustion facility (AFBC) was
designed and constructed. Major components of this new facility are an air
preheater, an atmospheric combustor, coal and limestone-sorbent hopper-
feeder assemblies, two parallel off-gas particulate-removal systems (each
system consisting of two cyclones and a sintered-metal final filter) and an
off-gas analysis system. Figure 5 is a simplified piping schematic of the
new facility. A schematic diagram of the 152-mm-ID atmospheric combustor to
be used in the new facility is shown in Fig. 6. Figure 7 is a photograph
showing some major components of the new facility.
The AFBC was designed for attended operation one shift per day
and automated operation on the remaining two shifts. Details of the process
control system for automated operation are discussed in the following section.
Anticipated nominal operating conditions for combustion are listed in Table 13.
-------
35
6"DIA ATMOSPHERIC
GAS ANALYSIS
RUPTURE DISK
PRESSURE COMBUSTOR
SECONDARY
CYCLONE
PREHEATER
PRIMARY
CYCLONE
SULFATED
PRODUCT
LIMESTONE
HOPPER
AIR
f J *
("L" VALVE
AIR
COAL HOPPER
ROTARY-VALVE FEEDER
•-AIR
Fig. 5. Simplified Piping Schematic of Atmospheric Combustor Facility
-------
36
127
OA
FREEBOARD
SECTION-
INTERNAL COOLING
COIL, PIPE-1"
SCHED.40.3IOSS
5 REQ'D.
BUBBLE CAP
. DISTRIBUTOR,
165 316 SS -^
I 3 CAPS REQD.
FLUIDIZING-COMBUSTION
AIR, 304 SS
FLANGE, 6"-150 Ib.CS
•-OFF-GASTO CYCLONES
AND FILTERS, 2" DIA
BODY, PIPE- 18"
f'WALL.CS
FLANGE-18" l50lb,CS
^FLANGE- 5" I50lb.CS
•32" SOLIDS OVER-
FLOW, I" DIA
SOLIDS GRAVITY FEED
PORT, I" DIA
FLANGE-5" l50lb,CS
REFRACTORY LINING
(PLIBRICO K-LMIX)
INSULATION LINING
(PLIBRICO VERILITE-R)
SOLIDS FEED LINE,304 SS
Fig. 6. 152-mm-ID Atmospheric Combustor.
Pipe and flange sizes (nominal
sizes) are American standard.
-------
ATMOSPHERIC
COMBUSTER
i,: PRESSURE
CONTROL .IV".*v... « ••
••: VALVE
OFF-GAS J
LINE i;
AIR PREHEATER
^•f FEEDER-HOPPER
;i;3i ASSEMBLIES
HEATED
FLUIDIZING
- AIR LINE
TRANSPORTfe
Fig. 7. Some Major Components of the New AFBC.
ANL Neg. No. 308-78-63A
-------
Table 13. Nominal Operating Conditions.
Fluidized-Bed Combustor.
Atmospheric
Process Variable
Nominal Value
Fluidized-Bed Temperature
Pressure
Superficial Fluidizing Gas Velocity
Goal Feed Rate3
Limestone Feed Rate13
Heat Generation Rate
Fluidized-Bed Height
Fluidizing-Combustion Air Flow Rate
Estimated Coolant Air Required
855°C (1571°F)
152 kPa (1.5 atm)
1 m/s (3.3 ft/s)
895 mg/s (7.1 Ib/h)
428 mg/s (3.4 Ib/h)
26.3 kJ/s (89,800 Btu/h)
813 mm (32 in.)
6800 cm3/s (14.4 scfm)
18900 cm3/s (40 scfm)
Based on eastern bituminous coal, such as Sewickley.
Based on Greer limestone. Ca/S mole ratio «• 3.
b. Automatic Operation of the AFBC Corrosion Test Facility
The fluidized-bed corrosion test facility is designed for fully
automatic unattended operation. The process control system is nearly
complete, and engineering evaluation will be possible after a few minor
alterations. These engineering evaluations will consist of a short series
of continuously attended, 100-h runs to determine the reliability of the
automatic control system. Based on the results of these evaluations, any
necessary changes for automatic unattended operation will be implemented
and tested.
A large portion of the automatic instrumentation provides analyti-
cal data about the process. Upstream from the instruments for analyzing
gaseous constituents in the off-gas, a portion of the gas is passed through
a gas-conditioning system in which the off-gas is dried and the sample-gas
flow is regulated. The dried off-gas is then distributed via a manifold to
various gas analyzers. Infrared analyzers provide continuous analysis for
CO, CH/,., H20, and C02 concentrations. In addition, Q£ is monitored
by a paramagnetic analyzer, total hydrocarbons by a flame-ionization analyzer,
S02 by a pulsed fluorescent analyzer, and NO/NOX by a chemiluminescent
analyzer.
In addition to conditions monitored by the analytical instrumenta-
tion, a number of process conditions are monitored. These process variables
are controlled with a separate control loop for each variable. The six con-
trol loops implemented are shown in Fig. 8 and are described in Table 14.
In addition to the six process-control loops, a relay and timer
sequence is used to route off-gas through one of two parallel particulate-
removal systems (primary and secondary cyclones and a filter) or through both
of them. If an excessive pressure drop occurs across a particulate-removal
system, the off-gas flow is switched to the other system and the first is
taken out of service. If after a short period (about 10 min) with the second
system operating, the pressure drop remains excessive, the first system is
-------
39
returned to use and both systems are used simultaneously. A continued high
pressure drop through both systems for a period of about 10 tnin would initi-
ate automatic shutdown.
V.
NO. 4 LOOP
TRANSPORT AIR
FLUID1ZING AIR
T TEMPERATURE
P PRESSURE
F FLOW
R RECORDER
C CONTROLLER
TC THERMOCOUPLE
S SOLENOID VALVE
Fig. 8. Automatic Control Loops for Atmospheric
Combustor Facility
Experimental operation of the AFBC is monitored and data are re-
corded by an Acurex Autodata 9 data logger with alarm capabilities. The
alarm capabilities can initiate process shutdown in case of a major process
upset. As currently set up, the data logger can check any one of 400 channels
against high or low limits and for a faulty sensor. During continuous alarm
scanning, each channel is tested approximately once every three minutes.
However, faster scanning (every 20 s) is possible with some slight loss of
accuracy. Fast alarm scanning will be used, and experimental data will be
recorded once every 10 min in our experiments.
-------
Table 14. Control Loops for Atmospheric Fluidized-Bed Combustor
Loop
Number
1
2
3
4
5
6
Controlled
Variable
Transport air
flow rate
Fluidizing air
flow rate
Bed temperature
System pressure
S02
Excess air
Set point
2400 cm3/s
(Vj scfm)
7200 cm3/s
(15 scfm)
850°C
153 kPa
('vT.S psig)
750-1500
ppm
17% (3% 02
in off-gas)
Measurement
Orifice AP
Orifice AP
Type K thermocouple
Variable reluctance
transducer
Pulsed fluorescent
S0£ analyzer
Paramagnetic 02
analyzer
Manipulated
Component or Variable
Pneumatic control valve
Pneumatic control valve
Pneumatic control valve
(cooling air flow)
Pneumatic control valve
Sorbent (limestone)
feed rate
Coal feed rate
-------
41
Several unusual process conditions necessitating process shutdown
for safety considerations are given in Table 15. The automatic shutdown is
controlled by a relay and timer sequence which also activates an alarm at the
Laboratory Central Surveillance station. Whenever automatic shutdown is
initiated, power to the feeders and the air preheater is shut off. In
addition, the fluidizing and transport air streams are replaced with house
nitrogen; house nitrogen is also used to purge the filter housings and the
coal hopper. Finally, a bypass valve around the system pressure control
valve is opened to relieve any excessive system pressure.
Table 15. Process Levels Necessitating Shutdown
Limit
Condition
Level
Requiring
Shutdown3
HI CO in flue gas
HI CH4 in flue gas
LO C>2 in flue gas
HI Temperature in bed
HI Temperature in coal hopper
HI Temperature in filter casings
HI Temperature in cyclone receivers
HI Bed AP (e.g., as a result of plugged overflow)
HI AP across gas cleanup system (independent
of the data logger)
HI System pressure
HI Temperature of cooling gas outlet manifold
LO Pressure of house high-pressure air supply
LO Flow rate of fluidizing air
LO Flow rate of transport air
LO System pressure (e.g., as a result of
rupture disc breakage)
LO Pressure of emergency nitrogen supply
HI SO2 in flue gas
HI ©2 in flue gas (air leak into sample system)
>1000°C
>75°C
>100°C
>100°C
>17 kPa (>2.5 psi)
>34 kPa (>5 psi)
>205 kPa (>15 psig)
>3QO°C
<239 kPa (<20 psig)
<2400 cm3/s (<5 scfm)
<900 cm3/s (<2 scfm)
<136 kPa (<5 psig)
<239 kPa (<20 psig)
>1000 ppm
Tentative levels.
data.
To be reevaluated on the basis of actual experimental
After an adjustable delay of 10 to 60 min, steps are taken to con-
serve nitrogen. If the data logger alarm scan indicates that temperatures in
key locations are below a predetermined level, the fluidizing and transport
gas streams are switched back to air; also, the nitrogen purges of the filter
housings and coal hoppers are shut off.
To avoid operating without sufficient nitrogen for emergency shut-
down, a slightly different shutdown process is initiated when the pressure of
house nitrogen is low. In the latter case, the power to the feeders and the
gas preheater is shut off and the bypass valve around the system pressure
-------
42
control valve is opened. This sequence safely shuts down the system in the
event that insufficient nitrogen is available for emergency shutdown. In the
latter case, operation is not allowed to continue because the emergency shut-
down sequence would be inoperative due to the low supply of purge nitrogen.
c. Planned Schedule for Corrosion Experiments in AFBC
Compositions and geometries of corrosion specimens, as well as pref-
erable locations of exposure in the AFBC, have been selected by 0. Chopra of
the Materials Science Division of ANL. Parameters to be measured during cor-
rosion tests have been identified.
Initially, to obtain baseline corrosion data on a variety of metals,
corrosion combustion experiments are to be made with no sulfation accelerator
present in the fluidized bed. Next, 100-h scoping tests will be conducted to
determine the possible corrosive effects of a sulfation accelerator on a vari-
ety of metal corrosion specimens. The accelerator for these 100-h tests will
be selected on the basis of information obtained in laboratory-scale tests.
After the 100-h tests, metals and accelerating agents selected for further
testing will be subjected to 1000-h tests.
d. Preliminary Runs in the AFBC
A series of runs was carried out in the AFBC (prior to corrosion ex-
periments) to evaluate the effects on S02 retention by Grove limestone (1359)
of adding low concentrations (1.0 mol % or less) of CaCl2 or NaCl. In this
series of runs, Sewickley coal (either -6 +100 mesh or -12 +100 mesh) was com-
busted at a bed temperature of 850°C, a pressure of 101.3 kPa (1 atm) , a
fluidizing-gas velocity of 1 m/s, and a fluidized-bed height of 813 mm, with
3% 02 in the dry off-gas. The above variables were maintained at the stated
values in all runs; only the composition of the limestone sorbent and the
Ca/S mole ratio were varied. Grove limestone (-10 +30 mesh), with or without
CaCl2 or NaCl addition, was the sorbent used in this experimental series.
A total of 23 runs were made in which the Grove limestone sorbent
contained nominal concentrations of 0, 0.1, 0.3, or 0.5 mol % CaCl2. Eight
additional runs were completed in which the Grove limestone sorbent contained
nominal concentrations of 0.5 or 1.0 mol % NaCl.
Analytical data for this series of runs are incomplete. Consequent-
ly, the effects of CaCl2 or NaCl on S02 retention by Grove limestone are
yet to be evaluated.
2. Corrosion Behavior of Materials in Fluidized-Bed Environments
(J. Shearer, 0. K. Chopra,* and C. Turner)
As part of the study of the effects of NaCl and other salts on limestone
sulfation, corrosion experiments with several metal alloys were performed to
evaluate the corrosiveness of salts in a simulated flue-gas environment. The
major component of the experimental apparatus was a 5-cm-ID quartz fluidized-
bed vessel (Fig. 9). The bed material consisted of either a batch of fully
sulfated dolomite to which particulate salt was added or a continuous feed of
raw limestone or dolomite impregnated with salt from an aqueous solution.
The compositions of the materials tested are given in Table 16.
*~~. . .
Materials Science Division.
-------
43
BED MATERIAL
AND SALT SUPPLY
HIGH-TEMPERATURE
FLUIDIZED BED
VESSEL
DUST
COLLECTOR
EXHAUST
INLET GAS
SAMPLING
PORT
STAINLESS
LINES
QUARTZ
LINES
SAMPLE COUPONS
ON Pt WIRES
AIR N2 C02 S02
I I THERMOCOUPLE
WELL
QUARTZ FRIT
FLOWMETERS
STAINLESS
LINES
FLUIDIZED BED OF
SULFATED LIMESTONE
CRUSHED QUARTZ
DOWNCOMER TUBE
QUARTZ
LINES
HEAT EXCHANGERS
PREHEATER
Fig. 9. High-Temperature 5-cm-dia Quartz Fluidized-Bed Reactor
for Salt Corrosion Studies
Table 16. Compositions of Alloys (in wt %)
Alloy
Inconel 600
Inconel 601
RA 333
Type 304 SS
Type 316 SS
Type 310 SS
Type 321 SS
Incoloy 800
9Cr-2Mo
Fe
8.0
14.1
18.0
Bal.
Bal.
Bal.
Bal.
Bal.
Bal.
Ni
Bal.
Bal.
Bal.
9.5
12,0
20.5
10.5
32.5
-
Cr
15.
23.
25.
19.
17.
25.
18.
21.
9.
5
0
0
0
0
0
0
0
5
Mo
_
-
3.0
-
2.5
-
-
-
2.1
Mn
0.5
0.5
1.5
2.0
2.0
2.0
2.0
1.5
0.9
Si
0.25
0.25
1.25
0.50
0.50
1.50
1.00
1.00
0.33
C
0.08
0.05
0.05
0.08
0.10
0.25
0.08
0.10
0.09
1
3
0
0
Other
.35 Al
.0 Co,
.40 Ti
.38 Al
Elements
, 0.25 Cu
3.0 W
, 0.38 Ti
-------
44
Corrosion coupons were suspended on platinum wires, both within and
above the fluidized-bed material at 850°C (1123 K) for 100 h. The bed
materials and the fluidizing-gas compositions for the various tests are
listed in Table 17. For batch runs, NaCl was injected as a solid powder
Table 17. Experimental Conditions
Run
Bed Material
Fluidizing-Gas Composition
5% 02, 200 ppm S02, balance N2
5% 02, 200 ppm S02, balance N2
5% 02, 3200 ppm S02, balance 82
5% 02, 3200 ppm S02, balance N2
5% 02 3200 ppm S02, balance N2
1 Fully sulfated dolomite
2 Fully sulfated dolomite,
1.5 g solid NaCl introduced
every 4 h
3 Dolomite treated with NaCl
(3.0 mol % NaCl)
4 Limestone treated with CaCl2
(0.1 mol % CaCl2)
5 Dolomite treated with NaCl
(1.0 mol % NaCl)
into the bed from above (run 2). In runs 3, 4, and 5, salt was introduced
in pretreated dolomite or limestone, prepared by soaking the stone in a salt
solution and drying. Small portions of the treated stone were periodically
introduced into the fluidized bed. The fluidized-bed vessel was equipped
with an overflow tube which maintained a constant bed level. For these runs,
the fluidizing gas contained 3200 ppm S02. After reacting with the bed
material, the fluidizing gas contained 200-500 ppm S02, 5% 02, and the
balance N2 and flowed at approximately 0.9 m/s.
Metallographic examinations (in the Materials Science Division of this
Laboratory) were performed on corrosion specimens, some exposed to beds with
salt and some exposed to beds without salt.
a. Metallographic Examination
The corrosive attack on all specimens was primarily oxidation, with
some sulfidation of the material. In general, the addition of salt to the
fluidized bed increased the corrosion rates. The average thicknesses of the
surface scales and the depths of corrosion-product penetration for the
specimens exposed inside and above fluidized beds are shown in Figs. 10 and
11 respectively. In the absence of salt, all materials developed 2- to
3-ym-thick surface scales when exposed either above or in the fluidized bed
which contained sulfated dolomite. The corrosive attack under the surface
scale in these specimens was minimal. However, under the same environmental
conditions, Inconel 601 specimens suffered considerable internal attack
because the exposure temperature was about 50 K higher, i.e., about 1173 K.
-------
45
Fig. 10. Average Thickness of Surface Scale and Corrosive
Penetration for Corrosion Coupons Exposed Inside
the Bed for 100 h at 1123 K except that the
Temperature for Inconel 601 was 1173 K.
As shown in Figs. 10 and 11, the addition of salt to the fluidized
bed increased the corrosive attack on all materials. The iron-base alloys,
(namely, Types 304, 316, and 310 stainless steel) fared better than the
high-nickel alloys., For these stainless steels, a variation in the amount
of NaCl in the bed had little or no effect on their corrosion behavior. In
the presence of NaCl, the average value of the total corrosive attack, i.e.,
the scale thickness plus the depth of penetration, observed in Types 304,
316, and 310 stainless steel was about 20 urn as compared with about 4 wm in
the absence of salt and about 8 ym in the presence of 0.1 mol % CaCl2.
The corrosion behavior of the nickel-base alloys, i.e. , Inconel 600,
Inconel 601, and RA333, showed a dependence on the amount of salt in the
fluidized bed. The addition of solid NaCl (run 2) increased the corrosive
attack drastically, i.e. , thick surface scales formed and extensive internal
attack occurred. The specimens exposed above the bed generally showed
greater corrosive penetration than those exposed inside the bed. However,
in the presence of 1.0 mol % NaCl or 0.1 mol % CaCl2, the corrosion behavior
for these alloys was comparable to that of the stainless steels.
-------
46
INCOLOY
800
INCONEL
600
RA 333
9 Cr 2 Mo
TOTAL
CORROSION
ALLOY
^
333
o— 0
L
Fig. 11. Average Thickness of Surface Scale and Corrosive
Penetration for Corrosion Coupons Exposed Above
the Bed for 100 h at 1123 K except that for
Inconel 601 the Temperature was 1173 K.
Detailed metallographic examination was made of all the specimens
to determine the distribution of corrosion products. Figures 12 to 17 show
scanning-electron micrographs (SEMs) of the cross sections of specimens of
Types 304 and 310 stainless steel, Incoloy 800, Inconel 600, Inconel 601, and
RA333 exposed under various test conditions. The specimens exposed to the
environment with NaCl or CaCl2 exhibited large cavities along the grain
boundaries. Micrographs of the specimens indicate that the internal corro-
sive attack is caused primarily by the preferential oxidation of the carbide
phases. Carbon from the carbides diffuses into the material and reprecipi-
tates ahead of the oxidation front, as shown in Figs. 14c and 17b. These
specimens showed a depletion of chromium in the surface region. The X-ray
microprobe line analyses for nickel, iron, and chromium on Incoloy 800
specimens, which had been exposed above NaCl-containing and NaCl-free bed
material, are shown in Figs. 18a and 18b, respectively. In the absence of
NaCl, the distribution of nickel, iron, and chromium in the specimen is
relatively uniform although some enrichment of chromium occurs at the surface,
However, in the presence of NaCl, the surface region is depleted of chromium
and iron to a depth of about 60 ym. The specimen matrix in this region
consists primarily of nickel.
-------
47
I
IOU
»w?s<•;;•••:•:• i sr fc •:•:••:•:••>."•.;.•:•:•••:
:M-:r:4r^f<-i:::v,;:^=,:
'"•":- \." '.' •' •• •.*"• ••• '•«-
;•/: ;,•:+* :-'].--'-
11
Fig. 12.
SEMs of Type 304 Stainless Steel after a 100-h Exposure
at 1123 K. (l) Exposed in bed, (II) exposed above bed,
(a) without salt, (b) solid NaCl , (c) 1.0 mol % NaCl,
and (d) 0.1 mol % CaCl2 . ANL Neg. No. 306-78-795
-------
48
10/1
IQ/i
I
II
Fig. 13. SEMs of Type 310 Stainless Steel after a 100-h Exposure
at 1123 K. (I) Exposed in bed, (II) exposed above bed,
(a) without salt, (b) solid NaCl, (c) 1.0 mol % NaCl,
and (d) 0.1 mol % CaCl2. ML Neg. No. 306-78-797
-------
49
15/1
I
II
Fig. 14. SEMs of Incoloy 800 after a 100-h Exposure at 1123 K.
(l) Exposed in bed, (II) exposed above bed, (a) with-
out salt, (b) solid NaCl, (c) 1.0 mol % NaCl, and
(d) 0.1 mol % CaCl2. ML Neg. No. 306-78-798
-------
50
?!«>' ".": '•«'••' XV; '.v>.;-v^v-,><";> ..'jX*^
'•''•»* .•'. v-* !•'•' •'»'* *.'.'.. •'•''••.'•' '' i-
%;•'.:/ •.-. ; ^V^..., :?• •-• , 5/t §
;::; :;k T> -.;.;.,. ;<.'r':' ^ -x.;>: ' '
^•^. • /• <••• • *<$ ^-nfyj-:
^Cf^:;.'v:: •' •*• :%,%'.&:^
iiliffl
.
^^^::.^«-
•v.;-:'''-i ";
I
15/1
II
Fig. 15.
SEMs of Inconel 600 after a 100-h Exposure at 1123 K.
(l) Exposed in Bed, (II) exposed above bed, (a) with-
out salt, (b) solid NaCl, (c) 1.0 mol % NaCl, and (d)
0.1 mol % CaCl2. ANL Neg. No. 306-78-796
-------
51
15
15/Z
1 II
Fig. 16. SEMs of Inconel 601 after a 100-h Exposure at 1123 K.
(I) Exposed in bed, (ll) exposed above bed, (a) with-
out salt, (b) solid NaCl, (c) 1.0 mol % NaCl, and (d)
0.1 mol % CaCl2. ANL Neg. No. 306-78-794
-------
52
a
10/z
?s
\5/J.
15/t
I
II
Fig. 17.
SEMs of RA333 after a 100-h Exposure at 1123 K.
(l) Exposed in bed, (II) exposed above bed, (a)
without salt, (b) solid NaCl, (c) 1.0 mol % NaCl,
and (d) 0.1 mol % CaCl2. ANL Neg. No. 306-78-793
-------
53
\5fJL
Fig. 18. X-ray Microprobe Line Analyses for Ni, Fe,
and Cr on Incoloy 800 Specimens Exposed
above the Bed for 100 h at 1123 K.
(a) Exposure above bed without salt and
(b) above bed with solid NaCl added.
ANL Neg. No. 306-78-787
-------
54
Scanning-electron micrographs and energy-dispersive X-ray (EDAX)
analyses of the surface of the scales formed on Incoloy 800 specimens which
were exposed above the bed (1) free of and (2) containing NaCl, are shown in
Figs. 19a and 19b, respectively. In the absence of NaCl, a continuous oxide
scale formed on the specimen surface. The major elements in the scale are
chromium, manganese, and iron. Minor amounts of calcium, magnesium, silicon,
sulfur, aluminum, and titanium are also observed. Sulfur is probably present
Fig. 19. SEMs and EDAX Analysis of the Surface of the Scales Developed
on Incoloy 800 Exposed above the Bed for 100 h at 1123 K.
(a) Exposure without salt and
(b) with solid NaCl
ANL Neg. No. 306-78-458
-------
55
in calcium sulfate particles that deposit on the scale. When NaCl is present
in the bed, the specimen surface is quite corroded, (Fig. 19b) and the metal
grains underneath the scale are exposed. The EDAX analysis of the surface
shows the presence of nickel, iron, and silicon.
The addition of salt to the fluidized bed also caused internal sul-
fidation of some of the alloys, e.g.. Type 304 stainless steel and RA333
specimens (shown in Fig. 12c and 17c, respectively). The patches of sulfide
were always observed in a region between the oxidation front and areas where
the carbides reprecipitate.
These results indicate that NaCl causes destruction of the normally
protective oxide scales. This process leads to continuous depletion of chro-
mium and iron in the specimen matrix. The absence of a stable and adherent
oxide scale leads to internal oxidation and sulfidation of the alloys. The
sulfides are always observed ahead of the oxidation front. This behavior
shows that diffusion of sulfur in the matrix is faster than diffusion of
oxygen.
For most of the alloys, the internal corrosive attack consists of
three distinct zones. At the zone near the surface, there is internal oxi-
dation. The matrix is depleted in chromium and shows iron-silicon oxides
along the grain boundaries. The second zone consists of patches of chromium
and manganese sulfides. In this region, the partial pressure of oxygen is
low and chromium reacts preferentially with sulfur to form sulfides. The
internal oxidation and sulfidation zones are free of carbide particles. The
carbides in these zones are either oxidized or they dissolve and the carbon
diffuses into the material and reprecipitates as chromium-rich carbides ahead
of the sulfidation zone. The third zone consists of these reprecipitated
carbide particles.
b. Effect of Temperature
To evaluate the effect of temperature on the corrosion behavior of
materials in the presence of salt, corrosion specimens were placed inside and
at different heights above the fluidized bed. The temperature of the speci-
mens varied from 1123 K to 723 K (850 to 450°C). For these tests, the fluid-
ized bed contained 3.0 mol % NaCl. Micrographs of Types 304 and 310 stainless
steel specimens exposed inside and above the bed at different temperatures
are shown in Figs. 20 and 21, respectively.
The corrosion behavior of the specimens that were exposed inside
and above the bed at 1123 K (850°C) is similar to that described in the
previous section. However, the specimens exposed above the bed at lower
temperatures show considerably more corrosive attack. The total depths of
corrosion in Types 304 and 310 stainless steel specimens at 923 K (650°C)
are 80 and 50 um, respectively, and at 823 K (550°C) the values are 600 and
100 pm, respectively. At these temperatures, the surface scales consist of
an outer layer of iron oxide with an inner layer of mixed oxides of iron and
chromium. The alloy matrix shows extensive internal sulfidation and oxida-
tion.
-------
56
a
.••.; •
15/i
Fig. 20. SEMs of Type 304 Stainless Steel after a 100-h
Exposure. (a) Exposed in bed at 1123 K,
(b)-(f) exposed above bed at (b) 1123 K,
(c) 1073 K, (d) 923 K, (e) 823 K, and (f) 723 K.
ANL Neg. No. 306-78-791
-------
57
I5U
t
, m .>V.--Y.:^
' *• - tr ^ '.<•'• 7
f ^^-
Fig. 21. SEMs of Type 310 Stainless Steel after a 100-h
Exposure. (a) Exposed in bed at 1123 K,
(b)-(e) exposed above bed at (b) 1123 K,
(c) 1073 K, (d) 923 K, and (e) 823 K.
ANL Neg. No. 306-78-792
-------
58
The X-ray microprobe line analyses for iron, chromium, nickel, oxy-
gen, and sulfur on Type 304 stainless steel specimens exposed at 650°C above
the bed containing NaCl are shown in Fig. 22. The relative concentrations of
these elements show that the outer surface scale consists mainly of iron and
oxygen. The inner scale contains iron, chromium, and oxygen. The region
below the oxides consists of nickel, iron, and sulfur. The X-ray images for
the above elements and chlorine are shown in Fig. 23 along with a micrograph
of the surface scale on the same specimen. The distribution of the various
elements shows the two oxide layers and the region of internal sulfidation
and oxidation. The X-ray image for chlorine shows that this element is pres-
ent in regions which show a higher concentration of sulfur. This behavior was
observed in all specimens exposed at 923 and 823 K. Chlorine was observed in
the region between the oxide layer and the area of internal sulfidation. The
sulfide/mixed-oxide and the mixed-oxide/iron-oxide interfaces observed in
Type 304 stainless steel specimens exposed at 823 K are shown in Figs. 24a
and 24b, respectively. Cubes of NaCl can be clearly seen in the sulfide
region.
c. Conclusions
1. The addition of NaCl or CaCl2 to the fluidized bed
increases the corrosion rates.
2. In the presence of salt, Types 304, 316, and 310 stain-
less steel perform better than did the high-nickel alloys.
3. The corrosion behavior of stainless steels is relatively
insensitive to the amount of NaCl in the bed.
4. The corrosion behavior of Inconel 600, Inconel 601, and
RA 333 in the presence of 1.0 mol % NaCl or 0.1 mol % CaCl2
is comparable to that of the stainless steels.
5. The internal corrosive attack consists of three distinct
zones: internal oxidation, internal sulfidation, and a
carburized zone where the carbon from the outer zone
reprecipitates as chromium-rich carbides.
6. Specimens exposed above the fluidized bed at 923 and
823 K show extensive corrosive attack. In these
specimens, chlorine was detected at the sulfide-oxide
interface.
-------
59
Fig. 22. X-ray Microprobe Line Analyses for Fe, Cr, Ni, 0,
and S on Type 304 Stainless Steel Specimen Exposed
Above the Bed Containing NaCl for 100 h at 923 K.
ANL Neg. No. 306-78-788
-------
60
Fig. 23. SEMs and X-ray Images for Fe, Cr, 0, S, and Cl of Type 304
Stainless Steel Specimen Exposed Above the Bed Containing
NaCl for 100 h at 923 K. ANL Neg. No. 306-78-789
-------
61
No Cl
Fe,Cr SULFIDE
Fe.Cr OXIDE
Fe,Cr OXIDE
Fe OXIDE
Fig. 24. SEMs of (a) Sulfide/Mixed-Oxide Interface and
(b) Mixed-Oxide/Iron-Oxide Interface Observed
in Type 304 Stainless Steel Specimen Exposed
Above the Bed for 100 h at 823 K.
ANL Neg. No. 306-78-790
-------
62
REFERENCES
1. G. J. Vogel et'al., Supportive Studies in Fluidized-Bed Combustion,
Quarterly Report, July-September 1977, Argonne National Laboratory
Report, ANL/CEN/FE-77-8.
2. J. C. Montagna et al., Fluidized-Bed Regeneration of Sulfated Dolomite
from a Coal-Fired FBC Process by Reductive Decomposition, Argonne
National Laboratory Report, ANL-77-16 (April 1977).
3. G. J. Vogel et al., Supportive Studies in Fluidized-Bed Combustion,
Annual Report, July 1976-June 1977, Argonne National Laboratory
Report, ANL/CEN/FE-77-3.
4. G. J. Vogel et al., Regeneration of Sulfated Limestone from FBCs and
Corrosive Effects of Sulfation Activators in FBCs, Quarterly Report,
October-December 1977, Argonne National Laboratory Report,
ANL/CEN/FE-77-10.
5. M. E. Whatley, A Dimensional Analysis of Solids Transport and Dispersion
in a Rotary Kiln, Oak Ridge National Laboratory Report, ORNL/TM-5898
(July 1977).
6. J. Griswold, Fuels Combustion and Furnaces, McGraw Hill, New York,
p. 421 (1946).
7. W. M. Swift et al., Decomposition of Calcium Sulfate; A Review of the
Literature, Argonne National Laboatory Report, ANL-76-122, pp. 40-42
(December 1976).
-------
63
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/7-79-157
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Regeneration of Sulfated Limestone from FBCs and
Corrosive Effects of Sulfation Accelerators in FBCs:
Annual Report
5. REPORT DATE
July 1979
6. PERFORMING ORGANIZATION CODE
'7. AUTHORISIG. J. Vogel,!. Johnson,J. F. Lenc,D.S. Moulton,
R. B. Snyder,J. A. Shearer, G. W. Smith, W. M. Swift
E. B.Smyk,F.G. Teats ,C. B.Turner, and A. A. Jonke
8. PERFORMING ORGANIZATION REPORT NO.
ANL/CEN/FE-78-13
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Argonne National Laboratory
9700 South Cass Avenue
Argonne, Illinois 60439
10. PROGRAM ELEMENT NO.
INE825
11. CONTRACT/GRANT NO.
IAG-D5-E681 (EPA) and
W-31-109-Eng-38 (DOE)
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
Annual; 7/77 - 9/78
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES (*) Cosponsored by DOE. Project officers: D. A.Kirchgessner
(EPA) and John Geffken (DOE).
is. ABSTRACT The report gives 1977-78 results of studies supporting the national develop-
ment of fluidized-bed combustion (FBC). Program objectives are to develop an eco-
nomically/environmentally acceptable process for regenerating the partly sulfated
product of the FBC of coal, to obtain the design data needed for the construction of
larger regenerators, and to determine the corrosion of metallic alloys by sulfation
acceleration agents added to the limestone sorbent. A regeneration process model
has been developed and used to investigate the effects of the main variables on regen-
erator size and performance and to estimate the economic feasibility of regeneration.
It reports results of an investigation of a regeneration process using a rotary kiln in
place of FBC. An atmospheric pressure FBC was operated to study the corrosion of
metallic alloys by limestone treated with various sulfation accelerators (e.g. , NaCl
and CaCl2). It gives results of laboratory scale corrosion studies.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Pollution Desulfurization
Fluidized Bed Processing
Calcium Chlorides Dolomite
Combustion
Coal
Additives
Sulfur Oxides
Inconel
Automation
Accelerating Agents
Limestone
b.IDENTIFIERS/OPEN ENDED TERMS
Pollution Control
Stationary Sources
Atmospheric Fluidized
Bed Combustor
Incoloy 800
Metal Sulf idation
c. COSATl Held/Group
OTD~
13 B
13H,07A
07B
21B
21D
11G
08G
11F
18. DISTRIBUTION STATEMENT
Release to Public
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
63
20. SECURITY CLASS (Thispage/
Unclassified
22. PRICE
EPA Form Z220-1 (9-73)
------- |