United States      Industrial Environmental Research  EPA-600/7-80-013
Environmental Protection  Laboratory          January 1980
Agency        Research Triangle Park NC 27711
Miniplant and Bench
Studies of Pressurized
Fluidized-bed  Coal
Combustion: Final 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:

     1. Environmental Health Effects Research

     2. Environmental Protection Technology

     3. Ecological Research

    4. Environmental Monitoring

    5. Socioeconomic Environmental Studies

    6. Scientific and Technical Assessment Reports (STAR)

    7. Interagency Energy-Environment Research and Development

    8. "Special" Reports

    9. Miscellaneous Reports

This report has been assigned to the INTERAGENCY ENERGY-ENVIRONMENT
RESEARCH AND DEVELOPMENT series. Reports in this series result from the
effort funded under the 17-agency Federal Energy/Environment Research and
Development Program. These studies relate to EPA's mission to protect the public
health and welfare from adverse effects of pollutants associated with energy sys-
tems. The goal of the Program is to assure the rapid development of domestic
energy supplies in an environmentally-compatible manner by providing the nec-
essary environmental data and control technology. Investigations include analy-
ses of the transport of energy-related  pollutants and their health and ecological
effects;  assessments of, and development of, control technologies for energy
systems; and integrated assessments of a wide range of energy-related environ-
mental issues.
                        EPA REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved
for publication. Approval does not signify that the contents necessarily reflect
the views and policies of the Government, nor does mention of trade names or
commercial products constitute endorsement or recommendation for use.

This document is available to the public through the National Technical Informa-
tion Service, Springfield, Virginia 22161.

-------
                                       EPA-600/7-80-013

                                              January 1980
    Miniplant and Bench Studies
of Pressurized Fluidized-bed Coal
      Combustion:  Final Report
                         by

         R. C. Hoke, E. S. Matulevicius, M. Ernst, J. L. Goodwin,
      A. R. Garabrant, I. B. Radovsky, A. S. Lescarret, R. R. Bertrand,
        L. A. Ruth, V. J. Siminski, M. S. Nutkis, M. D. Loughnane,
            H. R. Silakowski, M. W. Gregory, and A. Ichel

               Exxon Research and Engineering Co.
                       P. 0. Box 8
                  Linden, New Jersey 07036
                  Contract No. 68-02-1312
                 Program Element No. INE825
              EPA Project Officer: D. Bruce Henschel

            Industrial Environmental Research Laboratory
          Office of Environmental Engineering and Technology
                Research Triangle Park, NC 27711
                      Prepared for

            U.S. ENVIRONMENTAL PROTECTION AGENCY
               Office of Research and Development
                   Washington, DC 20460

-------
                                  ABSTRACT


     The pressurized fluidized bed combustion (PFBC)  of coal  and  regeneration
of spent sorbent were studied in the continuous  480  Ib coal/hr  (220  kg/hr)
"miniplant" and the 28 Ib coal/hr (13 kg/hr)  bench PFBC units.  The  ability
of the PFBC system to reduce SOg emissions by 90& or more  was demonstrated
using both dolomite and limestone sorbents.  The dynamic response of SO?
emissions to sudden changes in the sulfur content of the coal,  and in the
dolomite to coal feed ratio, was also measured.   NOX control  tests carried
out in the bench scale unit indicated that either two stage  combustion or
ammonia injection could reduce NOx emissions  below the already  low levels
normally measured in PFBC.

     Particulate emissions were studied using a  number of  particulate control
devices.  A three stage cyclone system was shown to  be unexpectedly  efficient.
However, it could not consistently reduce particulate emissions to the level
required by the current New Source Performance Standard for  utility  boiler.
A high temperature/pressure ceramic fiber filter was successfully tested,
giving collection efficiencies high enough to meet current emission  standards.
Further testing of a granular bed filter was  unsuccessful  due to  poor removal
efficiencies and various other operating problems.   Conventional  low tempera-
ture/low pressure electrostatic precipitator  and bag house systems were also
tested.  Both were shown to be capable of reducing particulate  emissions,
following the three stages of cyclones, to levels meeting  the current New
Source Performance Standard.

     Regeneration of spent sorbent was studied in the miniplant and  bench
units.  In the miniplant, a series of extended tests were  completed  which con-
firmed the operability of the system and the  large reduction in fresh lime-
stone resulting from the use of regeneration. An approximate four fold reduc-
tion in fresh limestone feed requirements results from the use  of regeneration.
S02 retentions were in excess of 90% at all times.   No loss  in  sorbent
activity due to regeneration was observed. Bench scale tests indicated that
coal  could probably be used as the fuel for regeneration in  place of natural
gas used in the miniplant program.

     A series of sampling campaigns was completed which generated samples used
by another contractor for Level I and Level II comprehensive analyses of
emissions from PFBC.

     This report 1s submitted in fulfillment  of  Contract Number 68-02-1312 by
Exxon Research and Engineering Company under  sponsorship of  the Environmental
Protection Agency.  Work was completed in August 1979.
                                     111

-------
                              TABLE OF CONTENTS

                                                                      Page
Abstract                                                              ill-
List of Figures                                                        v
List of Tables                                                         ix
Acknowledgements                                                      xii
Sections
     I    Summary                                                      1
    II    Introduction                                                 13
   III    Miniplant Combustion Studies                                 17
          Equipment, Materials, Procedures                              17
          Experimental Results and Discussion                          47
    IV    Particulate Measurement and Control                           86
          Particulate Measurement                                      87
          Cyclone Studies                                              95
          Ceramic Fiber Filter Evaluation                              105
          Granular Bed Filtration Studies                              123
          Conventional Particulate Control                             147
     V    Regeneration Studies                                        148
          Equipment and Procedures                                    148
          Experimental Results and Discussion                         151
    VI    Comprehensive Analysis  of Emissions                         168
          Level  I and II Comprehensive Analysis  Tests                  168
          Presence of Mg3(CaS0)   in 3rd Cyclone Flyash                169
   VII    Bench Unit Studies                                          171
          Combustion Studies                                          171
          Bench Regeneration  Studies                                  206
  VIII    References                                                 221
    IX    List  of Publications                                        223
     X    Appendices                                                 225
                                    1v

-------
                                LIST OF FIGURES
                                                                        Page
 II-l      Pressurized Fluidlzed Bed  Coal  Combustion System               14
 II-2      Exxon Fluidized Bed Combustion Miniplant                       16
III-l      Exxon Fluidized Bed Combustion Miniplant                       18
II1-2      Coal  and Limestone Feed System                                 19
III-3      Cross Section of Coal/Sorbent Feed Line                        21
III-4      Combustor Vessel                                               23
III-5      Miniplant Combustor Cooling Coll  1A After 1200 Hours           24
          Operation (Close Up)
III-6      Miniplant Combustor Cooling Coil  1A After 1200 Hours           25
          Operation
III-7      Cyclone Design Dimensions                                       27
III-8      Variable Pressure Reducing Nozzle                              30
II1-9      Controlled Condensation Coil Attached  to a S03-S02             32
          Impinger Train
111-10    Miniplant Flow Schematic                                       33
III-ll    Balston Filter Particulate Sampling System                     34
111-12    Original HTHP Particulate  Sampling System                      36
111-13    HTHP Particulate Sampling  System as Modified                   37
111-14    Southern Research Institute 5-Cyclone  Train                    38
III-l5    University of Washington 7-Stage Impactor                      39
111-16    Schematic of Alkali Probe  Train                                41
111-17    Coal  Particle Size Distribution                                43
111-18    Dolomite Size Distribution                                     46
111-19    Selected Combustor Temperatures During Crashdown and Restart   49
111-20    Miniplant Flow Schematic Hot Corrosion/Erosion                 50
          Test Configuration

-------
                         LIST OF FIGURES (Continued)

                                                                        Page
111-21    6E Turbine Blade Specimens                                     51
111-22    Westinghouse Boiler Tube Probes                                53
111-23    Natural Gas Injection - Temperature Profiles                   54
111-24    SOo Retention Adjusted to 2 s.  Gas Residence Time (tg).       57
          Dolomite Runs Having Actual tg Between 1.5 and 2.5 s.
111-25    S02 Retention Adjusted to 3 s.  Gas Residence Time (tg).       58
          Dolomite Runs Having Actual tg Between 2.5 and 3.5 s.
111-26    First Order Rate Constant Vs. Calcium Utilization              60
          Miniplant Dolomite Runs (Approximate Only)
II1-27    Measured and Calculated S02 Retention Vs. Ca/S for             61
          Miniplant Dolomite Runs
111-28    Calculated S02 Retention Vs. Ca/S Miniplant Dolomite Runs      62
111-29    SOo Retention Vs. Ca/S Ratio for Limestone No. 1359 at         63
          Calcining Conditions
111-30    Change in Coal  Sulfur/S02 Emissions (Run 99)                   67
111-31    Instantaneous SO? Response (Run 99) (Champion to               68
          Illinois No.  6 Coal)
111-32    Instantaneous S02 Response (Run 99) (Illinois No. 6 to         69
          Champion Coal)
111-33    Run 100 S02 Emissions (10 Minute Averages)                     70
111-34    Run 100 Instantaneous S02 Response (Champion Coal)             71
111-35    Run 100 Instantaneous S02 Response (Champion Coal)             72
II1-36    Correlation of NOX Emissions                                   74
111-37    Schematic of  the Alkali  Probe Train                            76
111-38    Particle Concentration in Cyclone Gas Outlets                  79
111-39    Primary Cyclone Dipleg Size Distribution                       81
111-40    Magnetization Curve for the Miniplant Flyash Sample at 25°C    82

                                     v1

-------
                         LIST OF FIGURES  (Continued)
                                                                        Page
111-41     Effect of Temperature on the Magnetic  Induction  Force           83
          of the Miniplant Flyash Samples
 IV-1      Comparison of Particle Size Distribution Via  Bahco Vs.          88
          Coulter Counter of Secondary Cyclone Captured Material
 IV-2      Comparison of Particle Size Distribution Via  Bahco Vs.          89
          Coulter Counter of Tertiary Cyclone Captured  Material
 IV-3      Balston Filter/Coulter Counter Vs.  APT Impactor                 92
          Particle Size Distribution
 IV-4      Effect of Temperature on IKOR Monitor Reading                  94
 IV-5      Tertiary Cyclone Collection Efficiency                         99
 IV-6      Basic Cyclone Design                                          102
 IV-7      Filter Housing Pressure Vessel Cross Section                   106
 IV-8      Acurex Test Filter Installation Schematic                     108
 IV-9      Acurex HTHP Ceramic Bag Filter Site                           109
 IV-10    Acurex HTHP Ceramic Bag Filter Pressure Drop and Flow         111
 IV-11     Acurex High-Temperature Ceramic Bag Gas Inlet Schematic       114
 IV-12    Acurex HTHP Bag Filter Outlet Loading Vs. Face Velocity       117
          (Averaged Over First 6 Hours of Exposure)
 IV-13    Acurex Ceramic Bag Filter - Bag No. 5 Particulate             118
          Penetration History
 IV-14    Ceramic Filter Bag No. 3                                      120
 IV-15    Ceramic Filter Bag No. 3 Closeup of Vacuumed Strip            121
 IV-16    Ceramic Filter Bag No. 9 After Run 96                         122
 IV-17    Modified Granular Bed  Filter  Element  (Without Shroud)         125
 IV-18    Modified Filter Bed                                           126
 IV-19    Granular Bed  Filter Schematic                                 127
 IV-20    Filter Bed with Internal Baffle                               129
                                     vii

-------
                         LIST OF FIGURES (Continued)

                                                                       Page
  IV-21      Plexiglas Model GBF                                         130
  IV-22      Increase in Outlet Particulate Concentration with Time      135
  IV-23      Exxon Mark IV Granular Bed Filter (Single Bed)              141
  IV-24      Granular Bed Filter Installation Schematic                  142
  IV-25      Granular Bed Filter Flow and Pressure Drop Vs. Time         146
  V-l       Mlniplant Sol Ids Transfer System (Schematic)                150
  V-2       S02 Retention Adjusted to 2 s Residence Time Vs.            156
            Ca/S Ratio
  V-3       Combustor Bed Age and S02 Emissions Vs. Time (Run 102)      158
  V-4       Combustor Bed Age and S02 Emissions Vs. Time (Run 103)      159
  V-5       Combustor Bed Age and S02 Emissions Vs. Time (Run 105)      160
  V-6       Regenerator and Combustor S02 Emissions Run 105             165
VII-1      Schematic of Batch Combustor Unit                           172
VII-2      Bench Combustor Showing Locations for Ammonia Injection     186
VII-3      Emission Indices for Ammonia Injection Program Showing      189
           95% Confidence Intervals
VI1-4      Ammonia Injection Program:  Effect of Excess A1r Level      195
           on NOX Emissions
VI1-5      Response Time                                               196
VII-6      Emission Indices for Simulated  Flue Gas Redrculation       199
            (SFGR) Program Showing 95% Confidence Intervals
VII-7      Comparison  of the Simulated Flue Gas Rec1rculat1on          202
           (SFGR) Program with  the Ammonia  Injection Program
VII-8      Reduction 1n  N0« Emissions by Using Two Control              208
           Techniques  Simultaneously
VII-9      Bench Regeneration Unit                                     211
VII-10     Fuel  Injection Mode                                          213

-------
                                LIST OF TABLES
                                                                        Page
III-l       Coal  Composition                                             44
II1-2       Sorbent Composition                                          45
II1-3       Run Summary 90% S02 Retention Study                           56
111-4       S02 Dynamic Response Summary                                 65
III-5       Emission of Sodium, Potassium, Chlorine,  Vanadium            77
           in Flue Gas
III-6       Particulate Emission Particle Size Distribution              78
III-7       Particulate Concentration and Size Ranges                    78
           Representing a Number of Runs
 IV-1       Particulate Concentration and Size Miniplant                 95
           Cyclone System
 IV-2       Tertiary Cyclone Total Efficiency Summary                    97
 IV-3       Tertiary Cyclone Fractional  Efficiency Summary               98
 IV-4       Third Cyclone Operating Conditions                           100
 IV-5       Tertiary Cyclone Design Dimensions                           101
 IV-6       Summary of Cyclone Test Program Results                      104
 IV-7       Acurex HTHP Bag Filtration Summary                           112
 IV-8       Acurex HTHP Bag Filter Inlet Particulate Loadings            113
 IV-9       Acurex HTHP Bag Filter Collection Efficiency                 116
 IV-10     Granular Bed Filter Run Summary                              132
 IV-11     Particle Size Distributions                                  137
 IV-12     Granular Bed Filter Cleaning Program                         143
 IV-13     Comparison of Particulate Size Distribution of Material       143
           Before and After Filter Test Element
 IV-14     Run 115 Filter Test Summary                                  145
                                       ix

-------
                          LIST  OF TABLES  (Continued)

                                                                       Page
   V-l      Mini pi ant  Regenerator Run Summary                            153
   V-2      Minlplant  Fluidlzed  Bed Coal Combustion Run Summary          154
   V-3      Sorbent  Elutr1at1on  Losses                                   162
   V-4      Composition of  Regenerator Off Gas                           163
   V-5      Minlplant  Regenerator Mass Balances                          167
VII-1      Location of Gas Injection Points Bench Unit NO               174
           Control  Studies
VII-2      Designed Experiment  for Study of Staged Combustion           175
           in Bench Unit
VI1-3      Effects of Staged Combustion on NO  Levels at                178
           8 Atmospheres
VI1-4      Effects of Staged Combustion on NOX Emissions at             179
           5 Atmospheres
VII-5      Effect of the Bench  Unit's Operating Parameters              180
           upon SOp Emissions at 5 Atmospheres
VII-6      CO Emissions for Staged and Unstaged Combustion at           182
           5 Atmospheres
VU-7      Ammonia Injection Run Conditions                             185
VII-8     Ammonia Injection Location                                   187
VI1-9      Results of Program to Reduce NOX by Ammonia Injection        191
VII-10    Results of Ammonia Injection Runs 5 and 7                     192
VII-11     Comparison of the Results  of This Work and an Earlier        193
          Work (21) on Ammonia  Injection  to Reduce NOX
VI1-12    Response Time for Ammonia  Injection Run 5                     197
VI1-13    Simulated Flue Gas Recirculation Program - Summary           200
          of Results
VII-14    Significance of Changes  1n Emissions - Simulated Flue        201
          Gas Recirculation Program

-------
                         LIST OF TABLES (Continued)
                                                                       Page
VI1-15    Comparison of Ammonia Injection and Simulated Flue           203
          Gas Recirculation Results

VII-16    Variation in Ca/S Mole Ratio                                 205

VII-17    Combined NO -Control  Techniques Program.  Changes in         207
          NOX Levels x

VII-18    S09 and CO Emissions for Combined NO -Control Techniques     209
          RuR 2                               x

VI1-19    Bench Regenerator Run Summaries                              216

VI1-20    Coal Fueled Bench Regenerator Run Summaries                  219
                                     xi

-------
                              ACKNOWLEDGEMENTS


     The authors wish to express their appreciation  to  the many  individuals
who played major roles in the conduct of this  program at Exxon Research and
Engineering Company.  We wish to acknowledge the efforts of  the  operating
and mechanical crews, Tony AHobelli, Tom Barresi, Jeff Bond, Jim  Bond, Jack
Fowlks, Ted Gaydcs, Phi Hi pa Givens, Bob Groth,  Ed Hellwege, Frank Huber,
Pete Madorma, Tom Morrison, Sal  Pampinto, Bob  Robinson, Warren Spend, Ted
Sutowski, Luther Tucker, and George Walsh.  We also  wish to  thank  Bill Dravis,
a designer of the Research Technology Services Division and  our  math clerks
Beth Fell and Sue Gregory.  A special acknowledgement goes to Nancy Malinowsky
who typed this report.

     We also wish to acknowledge the cooperation we  received from  the other
EPA contractors who worked with  us on the mini pi ant  on  a number  of joint
programs.  This includes Chris Chaney, Mike Shackelton, Steve Schliesser and
Clyde Stanley of Acurex Corporation, Rick Parker of  Air Pollution  Technology,
Bob Hall, Willy Piispanen and Paul Fennelly of GCA Company,  Ken  Murphy and
Clem Thoennes of General Electric Company, Joe McCain of Southern  Research
Institute and many of their co-workers.

     The personnel of the Industrial Environmental Research  Laboratory of
the EPA have been most helpful and deserve special thanks.   We wish to express
our gratitude for the help of Bruce Henschel,  the EPA project officer, Pic
Turner and Bob Hangebrauck.
                                      xii

-------
                                  SECTION  I

                                   SUMMARY


     The pressurized fluidized bed combustion  of coal  (PFBC) and  regeneration
of sulfated S02 sorbent were studied in  the  continuous  miniplant  unit and the
semi-continuous bench unit.   In the miniplant  combustion  program, a series of
runs were made to verify that PFBC could attain  the 90% S02  retention level
using dolomite or limestone  sorbents.   The dynamic  response  in  502 emissions
to a sudden change in the sulfur content of the  coal,  and in the  dolomite to
coal feed ratio, was also measured.  Emission  of NOX»  other  gases and parti-
culates was measured.  Combustion efficiency results were updated.

     The measurement and control of particulate  emissions was studied.   Various
particulate measurement methods were employed  and cross checked.  Particulate
emission control using cyclones, and using cyclones in series with a high
temperature/pressure ceramic fiber filter and  with  a granular bed filter was
studied.  Particulate removal from the flue  gas  at  low temperature/pressure was
also studied, using a conventional electrostatic precipitator and bag house.

     Regeneration of sulfated limestone was  studied in a  series of extended
tests.  A series of sampling campaigns was also  completed in which samples
were taken which were used to conduct a comprehensive  analysis  of emissions
with and without regeneration.

     Combustion and regeneration studies were  also  carried out  in the bench
unit.  NOX control methods were evaluated in a series  of  combustion tests.
The use of coal as the regenerator fuel  was  also studied.

COMBUSTION STUDIES

     The miniplant combustor consists of a refractory  lined  vessel 10 m  (33
ft) high with an inside diameter of 32 cm (12.5  in).  A number  of vertical
water-cooled tubes are mounted in the combustor  to  remove the heat of com-
bustion.

     Premixed coal and sorbent are injected into the combustor  a  single  point
28 cm (11 in) above the fluidized bed support  grid. The  combustor normally
operates at pressures up to  950 kPa (9 atm), at  temperatures up to the ash
agglomeration temperature of the coal  (usually less than  960°C),  at super-
ficial velocities of up to 2 m/s (7 ft/sec)  and  with expanded beds of up to
3.3 m (12 ft).  The coal feed rate is normally less than  160 kg/hr (350
Ib/hr).  Flue gas leaving the combustor passes through three cyclones in
series to remove most of the particulate matter.  Particulates  captured  in the
first cyclone are recycled to the combustor to improve combustion efficiency.
Particulates captured in the second and third  stage cyclones are  rejected
through lock hoppers.  Spent sorbent is also rejected  from the  combustor
through a lock hopper system to maintain a constant bed level in  the combustor.

-------
      Runs  were made  with:   an  Eastern  bituminous Pittsburgh seam coal
 (Champion) containing about 1.5%  sulfur with a particle size distribution of
 300 to 2400 microns; an  Illinois  No. 6 coal containing about 3.5% sulfur
 screened to a  particle size range of 400 to 3400 microns; and an Ohio coal
 (Valley Camp)  containing 2.5%  sulfur screened to a size range of 300 to 3400
 microns.  A Virginia limestone (Grove  No. 1359) and an Ohio dolomite (Pfizer
 No.  1337)  were used.   Both  were screened to a size range of 700 to 2400
 microns.

      Operational performance of the combustor was very good.  At the comple-
 tion  of the  program  in August  1979, over 3700 hours of coal combustion time
 was  accumulated.  Twelve runs  of  100 to 250 hours duration were included in
 the  total.   The ability  of  the mini plant to conduct short daily runs was also
 demonstrated.   In one  period,  fourteen 8 to 10 hour runs were completed in
 fifteen  working days.  Multiple purpose runs were also conducted.  In some
 cases,  as  many as three  EPA contractors, in addition to Exxon, were present at
 the  same time, conducting sampling and equipment evaluation tests during
 extended combustion  runs.   During this reporting period, a total of 1117 hours
 of test  time was dedicated  to a materials evaluation program sponsored by the
 Department of  Energy (DOE), under a cooperative agreement between DOE and EPA.
 Gas turbine and boiler tube materials were evaluated under realistic PFBC
 operating  conditions.  Evaluation of test results will be made by General
 Electric and Westinghouse,  the DOE contractors involved in the materials test
 program.

     S02 retention studies  for EPA were conducted to confirm the sorbent
 requirements,  indicated in earlier tests, needed to assure 90% or higher S02
 retention.   The work was  done in support of the most recent New Source Perfor-
 mance Standards for large coal fired boilers.  A number of runs were made with
 Ohio and Illinois coals,  dolomite and limestone sorbents at various calcium to
 sulfur  (Ca/S) molar ratios.  It was found that 90% and higher S02 retention
 levels could be reached with either sorbent.  In fact, retention levels as high
 as 99*%  were measured.  As expected, dolomite was shown to be more reactive.
 A Ca/S  ratio of 1.5 will  assure 90% S02 retention at a gas superficial  residence
 time of  2 s while limestone use will require a Ca/S ratio of between 3.5 and 4.0,

     The recent S02 retention results obtained with dolomite sorbent were
 analyzed using the simple first order kinetic expression developed previously
 (2).  The recent data did not show the expected effect of gas residence time
on S02 retention.  However, the lack of a gas  residence time effect may have
been due to the fairly narrow range of residence time  variations.  It may also
indicate that the reaction rates may be affected in larger FBC units such as
the miniplant, by other operating parameters such as  solids recycle and
rejection.   These factors may make it difficult to  obtain satisfactory results
 using simple rate expressions, the rate expression  must be refined to account
 for these factors.  However, this rate expression can  be used as an approx-
 imate guide until a better understanding of the FBC system 1s developed.

     The dynamic response of the S02 emissions  to sudden changes in the sulfur
 content  of the coal, and  in the Ca/S ratio,  was also  studied.  A new auxiliary
 coal/sorbent feed system  was built and installed to permit rapid switching

-------
from the current feed system to the new one to cause the rapid change  in  coals
and coal/sorfaent ratios.   The time to switch from one system to the  other was
about 40 s.  The work was done in cooperation with General  Electric, another
EPA contractor, who is developing a dynamic SOg response model for assessing
automatic control options.  It was found that the S02 emissions were stabilized
in 7 to 8 minutes following a step change in the sulfur content of the coal
from 1.9 to 3.4%.  However, when the Ca/S ratio was changed suddenly from 1.4
to 0.4, the time to stabilize was 110 min.  The change back to the higher Ca/S
ratio required 300 min to reach a stable emission level.  The reason for  the
difference in the two response times (110 and 300 min) is not apparent.
However, it is clear that the response to a change in the Ca/S ratio has  a  time
constant much greater than that for a change in the coal sulfur content.  These
results will be used by General Electric to develop the emission response model
model.  The model will be described in a separate report by General  Electric.

     NOX emissions measured recently follow the same trend line with excess air
developed in earlier studies.  At an excess level of 20%, NOX emissions
averaged 0.2 +0.1 Ib/MBTU.  Even at excess levels in the range of 60  to  100%,
emissions are no greater than 0.4 Ib/MBTU.  Therefore, NOX emissions are  still
well within the New Source Performance Standard even with the recent reduction
in the NSPS to 0.6 Ib/MBTU.
         emissions were measured using the controlled condensation method in
place of the older Method 8 method.  Method 8 results averaged 12 +.12 ppm,
controlled condensation results averaged 6 +_ 9 ppm.  In both cases~the range
of values was 0 to 30 ppm.  Although the controlled condensation method gave
an average value one half that of the Method 8 average, the uncertainty ranges
were so wide that the difference may not be significant.  No positive con-
clusions could be drawn from the results regarding the cause or the factors
affecting the degree of $03 formation.  The reduced sulfur compounds, HgS, COS
and CS2, were found to be less than the detectability level of 1 ppm in the
flue gas.  Methane averaged 7 + 5 ppm, ethane 4 + 4 ppm, C3 through C^ hydro-
carbons were generally less than 1 ppm, the detectability limit.  Sodium and
potassium present in the flue gas were measured by quenching and filtering a
flue gas sample, then extracting the collected particulates and sample system
with hot water.  Sodium concentration in the extract was equivalent to 2 to
3 wppm in the flue gas.  Potassium was found to be between 0.3 and 0.5 wppm
in the flue gas.  No vanadium was detected,  Chloride levels of about 50 wppm
in the flue gas were also measured.

     Particulate emissions in the flue gas, after passage through three con-
ventional cyclones in series, generally ranged from 0.03 to 0.15 g/Nm^, com-
pared to the NSPS of about 0.035 g/Nm3 (13 ng/J) for particulates.  The par-
ticulates generally had a median particle size of 1 to 2 ym, with 80 to 90%
smaller than 5 ym.

     Carbon combustion efficiency measurements averaged 99.3% for the runs
made during this period, about 0.8% higher than expected based on the correla-
tion published in the previous report (1).

-------
 PARTICULATE  MEASUREMENT AND CONTROL

      An  important  technical issue to be resolved before pressurized fluidized
 bed  combustion  can be applied commercially is the degree of particulate
 removal  needed  to  protect the gas turbine.  A related issue is which tech-
 nology to use to achieve the needed degree of the particulate control.  In
 addition to  meeting  the particulate removal requirements set by the gas tur-
 bine, the environmental New Source Performance Standards published by the
 Environmental Protection Agency must also be met.  Although the EPA is not
 directly concerned with studying particulate control to protect gas turbines,
 it 1s responsible  for the evaluation of such particulate control devices to
 the  extent that these devices could determine the level, size and composition
 of particulate matter emitted from a PFBC system.  Therefore a particulate
 control  program was  begun in which pre-turbine devices such as granular bed
 filters, a ceramic fiber filter and high efficiency cyclones were evaluated.
 In addition, post-turbine devices were also tested, based on the realization
 that  the degree of pre-turbine cleanup may not be sufficient to meet the more
 stringent New Source Performance Standards for particulates.  In this case, a
 trailer mounted electrostatic precipitator (ESP) and a trailer mounted bag
 house were connected to the mini plant flue gas system and tested under typical
 low  pressure, low  temperature conditions.

      In addition to evaluating the particulate control  devices, It was also
 necessary to develop and improve particulate measurement systems.  Such sys-
 tems were used to determine particulate concentrations in the flue gas before
 and after the control devices and to measure particulate size distributions.
 For most of the current program, particulate concentrations were measured by
 passing isokinetlc samples of the flue gas through high temperature total  fil-
 ters.  Particle size distribution was measured using a Coulter Counter on a
 sample of particulate captured on the filter.  A question arose as to whether
 the size distribution measured by the Coulter Counter was the same as that
 occurring in the flue gas.  The degree of particulate agglomeration on the
 filter and redlspersion during the Coulter Counter measurement were unknown,
 and could affect the measured particle size.    The primary concern was that
 redlspersion in the Coulter Counter resulted in a particle size distribution
 significantly finer than that actually existing in the flue gas.  To answer
 the question, a series of samples were obtained by Southern Research Institute
and Air Pollution Technology, Inc. using a high temperature, pressure cascade
 impactor developed by Air Pollution Technology.  This is a different parti-
culate sampling system which would not cause possible agglomeration and redis-
persion effects.  The impactor results indicated a larger concentration of fine
particles and a mass median size of about 0.8  m in the flue gas following the
third cyclone, compared to 1.6  m measured by the Coulter Counter.  Electron
micrographs of the material  captured on the impactor stages also showed very
 little sign of agglomeration.  These findings indicated that the results of
the filter/Coulter Counter method differed somewhat from cascade impactor
results, but to the degree expected, based on measurements made in other par-
ticulate systems.  The Coulter Counter is definitely not biased toward the
 finer particles by breakdown of agglomerates in the Coulter Counter during
sample preparation.  A cross  check was also made between the Coulter Counter
 and a Bahco inertia! system.   Both devices gave very similar particle size
 distributions.  An attempt was also made to use a miniature five stage cyclone

-------
train to obtain participate concentration  and  size  distributions.  Satisfactory
particle size distributions could not be obtained with  this  system due  to
overlapping size differentiation on the cyclone stages.  A continuous particle
monitor was also tested but did not perform satisfactorily.   This device
measured the rate of transfer of electrical  charge  from particles in the duct
to a stationary rod as the charged particles impinged on the rod.  The  system
was very sensitive to temperature changes  and  was not reliable.   It was con-
cluded that either the filter/Coulter Counter  system or a cascade impactor
system was satisfactory.  The filter/Coulter Counter system  is easier to use
and probably gives a better measure of particulate  concentration, but gives a
coarser size distribution.

     Particulate control studies included  the  testing of the three-cyclone
system installed on the mi nip!ant in addition  to a  number of other systems.
The second cyclone was found to be 90 to  95% efficient, reducing  the concen-
tration of particulates entering the cyclone with a mass median diameter of
20 to 25 um to 0.4 to 1.2 g/Nm3 in the cyclone outlet.   The  outlet particulate
stream had a median size of 3 to 5 ym,  The third cyclone was found to  be
about 90% efficient, reducing the concentration of  particulates entering it
with a mass median diameter of 3 to 5 ym  to 0.03 to 0.15 g/Nm3 in the outlet.
Particulate in the flue gas leaving the third  cyclone had a  mass  median par-
ticle size of 1 to 3 ym.  The efficiency  of third stage cyclone was indepen-
dently verified by a team from Southern Research Institute and Air Pollution
Technology, Inc., using the same high temperature/pressure cascade Impactor
system mentioned previously.  The impactor results  verified  those measured
using total filters and Coulter Counter particle size measurements.  Both
methods indicated the cyclone cut diameter (particle size captured with 50%
collection efficiency) to be about 0.7 to  0.9  ym.  The  measured efficiency
is significantly greater than that expected from theoretical efficiency models.
The reason for the difference is not known.

     Although the cyclone system, at times, did just meet the EPA New Source
Performance Standard for particulates (0.035 g/Nm3), it did  not do so con-
sistently.  Further studies were carried  out with cyclones of differing designs
to see if the performance could be optimized and higher efficiencies obtained.
Two new cyclones were built similar to those currently  being tested at  the
National Coal Board PFBC facility in Leatherhead, U.K.   They were tested on the
miniplant flue gas in place of the existing conventional third stage  cyclone.
Contrary to expectations, the new cyclones were found  to be  no more  efficient
than the existing third stage cyclone.  Also,  no significant effects  of pres-
sure, inlet velocity, temperature or coal  type on cyclone performance  were
found.  An increase in  inlet  particulate concentration  did  increase  overall
efficiency significantly.  Comparisons will be made with the performance  of
the similar cyclones at the National Coal  Board PFBC unit as the  data  become
available.

     Tests were also conducted on the performance of a  high  temperature/pressure
ceramic fiber filter provided by Acurex Corporation. A single cylindrical
filter element consisting of  a mat of Saffil alumina,  was mounted in  a  heated
pressure vessel and used to filter a slipstream of flue gas  extracted  between
the second and third cyclones.  The filter was cleaned  periodically by a  com-
bination of reverse flow of compressed air and short,  high  pressure pulses.

-------
 Pressure drops were continuously recorded and  were  found  to  be  0.2 to 2 kPa
 after cleaning, rising to a maximum of 14 kPa  before  cleaning.   Filtration
 efficiencies were high, ranging from 96 to 99.5%.   Outlet particulate loadings
 averaged 0.013 g/Nm3, well  under the EPA emission standard for  particulates
 and significantly lower than that obtained with  the three stage cyclone system.
 An extended test was attempted  but was terminated after 19 hours due to failure
 of the filter.  The failure was caused by a combination of corrosion of the
 very fine filter support screen and excessively  high  reverse pressure drops
 during bag cleaning.  Corrosion of the support screen was probably accelerated
 by a failure in the temperature control  system during this test.  The support
 screen used in these tests  was  the only type available at the time and it was
 recognized that it would have a short lifetime at test conditions.  The tests
 did indicate the ceramic filter was operable under  PFBC conditions.  High
 collection efficiency at high face velocities  was measured.   In general, the
 test filter exhibited performance similar to that of  a conventional bag
 filter.   However,  additional work is  needed to develop a  better mechanical
 support  system to  prevent bag failure.

      A granular bed filter  was  also evaluated  as a  high temperature, pressure
 particulate control  device.  This is  a system  in which particulates are
 removed  by passage through  a bed  of granular material.  A  number of small  beds
 operated in parallel  are used to  reduce  the  pressure  drop.  The beds are
 periodically cleaned by the  reverse flow of  clean gas.  The  "blow back" gas
 fluidizes  the granular  filter media at a  velocity sufficient to blow the
 collected  dust off,  but low  enough to  prevent  blowing  the  filter media itself
 out of the filter  vessel.  The  removed dust  settles outside the filter and is
 collected  at the bottom of a containment  vessel.  Previous studies with this
 system were beset  with  operating  problems, primarily caused by plugging of the
 filter inlet sections with particulates,  poor  bed cleaning and loss of filter
 media  during cleaning.   Filtration  efficiency was low and furthermore, decreased
 with  time.   Modifications were made to the system in an attempt to improve
 performance.   This  included testing a  number of inlet screen sizes, using
 baffles  to  prevent  the  loss of filter media and testing filter media  of various
 sizes  and  densities, also to prevent loss during the cleaning step.  However,
 the  test program was again unsuccessful due  to poor filtration efficiency,
 loss of  filter media and other serious operating problems which persisted  to
 the end of  the program.  The lowest outlet particulate concentrations  measured
 were 0.07 to  0.11 g/Nm3, but the concentrations increased with time during  the
 tests  by as much as a factor of three.  At no time did the filter performance
 equal  that  required by the particulate emission standards.  Since the  filter
 was installed after the second stage cyclone, the outlet  loadings could  be
 compared to  those from the third stage cyclone.  The comparison  indicated  that
 the cyclone was more efficient and moreover, the  efficiency was  maintained.
 The poor filter performance was  believed to be  due to  poor cleaning.   Filter
 beds examined after a test program were generally found to contain  high concen-
 trations of dust intimately mixed with the filter media.   Tests  made with
 transparent models at ambient conditions showed that the  dust was adhesive  and
 only partly removed during blow  back.  Some of  the dust adhered  to  the filter
media and was mixed into the filter bed by the  motion  of  the  fluidized filter
media.  Attempts to improve bed  cleaning by modifying  the  blow back conditions
were partly successful.  However, outlet concentrations were  still  high and

-------
increased with time.  Attempts  to prevent filter media  Toss were also unsuc-
cessful .  Screens were found to plug,  even when screen  opening sizes as large
as 10 mesh were used.  Larger filter media and very  dense  filter media were
tested but did not prevent the  loss.  Baffles were also unsuccessful.  Trans-
parent model tests were made and indicated that the  loss was  probably caused
by a surge in the blow back gas flow caused by the sudden  opening of the blow
back valve.  It was also suggested that the bed height  was  too low* not
allowing enough space for disengagement of the fluidized filter media.  In
addition to these problems, the operation of the filter caused periodic upsets
in the miniplant pressure control system, which in turn caused coal feeding
problems.  Pressure drops across the filter were also high, probably due to
poor bed cleaning.

     A final attempt was made to determine if an operable  granular  bed  system
could be developed.  A small single bed filter was installed  and  tested on  a
flue gas slipstream after the second stage cyclone.   The blow back  system was
modified to bring in some of the reverse flow gas  at the  interface  between
the filter media and the dust layer.  This blow back gas was  directed
horizontally across the top of the bed and was  intended to shear  off the dust
layer without disturbing the filter media enough to  cause  its loss.  A  single
test was made with the modified filter but was  unsuccessful.   Pressure  drops
were very high after the first  few minutes of operation.   Increasing the
severity of the cleanup step again resulted in  loss  of  filter media.
Unfortunately, the testing of this filter was terminated  due  to lack of time
before all operability questions could be answered.   An unresolved  issue was
whether the loss of filter media could have been  prevented by modifying the
blow back procedure, possibly by introducing the blow back gas more gradually.
The filter face velocity was also very high in  the  test and the effect  of
lowering the velocity on filter performance was also unresolved.

     A conventional low temperature, low pressure  electrostatic precipitator
(ESP) and bag house were also tested on the flue gas from  the miniplant in  a
series of long term tests.  The purpose of the tests was to determine if an
ESP or bag house could be used  after expansion of  the flue gas through  the  gas
turbine, to meet particulate emission  standards.   This  assumed that cyclones
may be sufficient to protect the gas turbine from  excessive wear  but would  not
be sufficient to meet environmental  standards.  The  tests  were performed using
mobile, trailer mounted systems operated by Acurex Corporation for  the  EPA.
Both systems appeared to be applicable in this service. The  ESP  overall
efficiency was 87%, corresponding to an emission level  of  about 0.02 g/Nm3.
The bag house overall filtration efficiency was 99.3%,  corresponding to an
emission level of about 0.001 g/Nm3.

REGENERATION STUDIES

     A series of extended runs  was made in which sulfated  limestone was con-
tinuously regenerated and recirculated to the combustor.   The primary objective
of the runs was to determine the reduction in fresh  limestone requirement
resulting from regeneration of the sul fated limestone.

-------
      The regenerator  consists  of a  refractory lined vessel with an inside
 diameter of 22 cm  (8.5  1n)  and an overall  height of 6.7 m (22 ft).  Gaseous
 fuel  is  burned in  a plenum  below the  fluidized bed to achieve the reaction
 temperature.   Additional  fuel  is injected  directly into the fluidized bed just
 above the fluidizing  grid to create a reducing zone in which the CaS04 reduc-
 tion  reaction  occurs.   Supplementary  air is injected directly into the bed
 at a  higher elevation to  create an  oxidizing zone.  The oxidizing environment
 at the top of  the  bed assures  high  selectivity to CaO, the desired product of
 the regeneration reaction,  by  minimizing the formation of CaS, an undesired
 byproduct.

      Three extended runs  totalling  370 hours were completed 1n the test series.
 Operations were generally good.  Fresh limestone, at a Ca/S ratio ranging
 from  0.7  to 1.5 was fed to  the combustor.  SOg emissions from the combustor
 were  low,  the  SO?  retention exceeded  90% at all times and was normally above
 97%.   In  a once through system, a Ca/S ratio of 3 to 4 would have been
 necessary  to achieve the  same  level of S02 retention.  The recirculating
 regenerated sorbent rates were much higher than the fresh limestone rates,
 giving total Ca/S  ratios  entering the combustor usually between 6 and 12 (one
 was 54).   The  regenerated sorbent appeared to have the same activity of fresh
 limestone  with no  sign of diminished activity due to regeneration.  There was
 also  no  evidence of loss  in activity during the runs.  Loss of sorbent by
 attrition  and  entrainment from the combustor and regenerator was somewhat
 greater  than the corresponding loss from a once through operation.  In fact,
 the fresh  limestone Ca/S  ratios fed to the combustor were usually those
 required  to maintain constant sorbent levels in the combustor and regenerator.
 It  had been planned to reduce fresh limestone feed rates to the point where
 the S02 retention  in the combustor was about 90%.   This could not be done,
 since  sorbent  feed would have had to be reduced to a point where it would have
 been  inadequate to make up for attrition/entrainment losses, and bed depth
would  have  dropped.  In a once through operation,  the entrainment losses are
about 1% of the bed/hr.   In the regeneration runs, the entrainment losses
averaged about 1.8%.  S02 concentration in the regenerator off gas was low,
ranging from 0.2 to 0.5% equivalent to 6 to 16% of the calculated equilibrium
concentration.   These levels are lower than what might reasonably be expected
in a commercial system.  As  pointed out in previous reports (1,2), due to the
size of the miniplant regenerator,  the S02 level  in the miniplant regenerator
off gas is determined by energy and mass balance requirements rather than
chemical  equilibria.  It had been  planned to use a higher sulfur coal  in these
tests  to  maximize  the  S02 concentrations.  Unfortunately, a  prepared high
sulfur coal could  not be obtained  and a  lower sulfur coal  was  used,  contri-
buting to the low  S02  concentrations.

     However, the  runs did establish the minimum degree to which the  sorbent
requirement would  be decreased  by regeneration.  This  is  a  factor  of  3  to 4
The runs  also indicated  that regenerated  sorbent activity was  high and
activity  loss due  to regeneration was  low.
                                     8

-------
COMPREHENSIVE ANALYSIS  OF EMISSIONS

     The comprehensive  analysis  of emissions  program  initiated  previously was
concluded.  In the earlier tests,  a  Level  I sampling  and  analytical program
was conducted in cooperation with  Battelle Columbus Laboratory, the EPA con-
tractor responsible for the program  at that time.   In the current series of
tests, a set of samples was taken  during  a run  with Illinois  No. 6 coal and
analyzed for inorganic  elements  by spark  source mass  spectroscopy (SSMS).
The results were forwarded to Battelle to be  included with the  Level  I results
obtained by them in the earlier  tests  which used Champion (Eastern) coal.
Also, in the current series of tests,  sampling  campaigns  were conducted in
cooperation with GCA Corp., the  current EPA contractor.   In this case, Exxon
assisted 1n sampling but essentially all  analytical work  will be done, inter-
preted and reported by  GCA.  In  the  recent tests,  a Level I sampling  campaign
was conducted during an extended run in which the regenerator was in  operation.
A Level II sampling campaign was also  conducted with  the  combustor operating
in a once- through fashion.
     A brief investigation into the formation of the double salt MgsCaCSO^H  or
MgS04 *n tne PFBC solid waste products was conducted.  These materials,  if
present, could cause environmental  problems upon disposal  since they are water
soluble.  No MgS04 was found in any of the samples.  MgsCa(S04)4 was found in
solid samples collected in the third stage cyclone and the gas turbine test
section downstream of the cyclone.   However, S02 concentrations in the run from
which samples were taken was considerably higher than normal.  Also, the resi-
dence time in the mi nip! ant ducting is probably greater than that which  will
occur in a commercial plant and the average temperature somewhat lower.   All  of
these factors promoted the formation of the double salt.  Additional tests
should be conducted in newer PFBC units to determine the extent of the double
salt formation under more realistic conditions.

BENCH UNIT STUDIES

     Programs were carried out in both the bench combustor and regenerator sec-
tions.  The combustor had been modified to permit continuous solids feeding and
removal.  The combustor was then used to evaluate three NOx control methods:
two stage combustion, NHs injection and simulated flue gas recirculation.  The
regenerator was used in a series of tests using sul fated sorbent produced in
the mini pi ant, studying the use of natural gas and coal to fuel the regenera-
tion section.

     The bench combustion unit consists of a refractory lined combustor vessel
which normally operates at temperatures of 840 to 950°C, pressures of five to
eight atm, superficial velocities of 1 to 2 m/s and coal feed rates of 1 to 12
kg/hr.  The inside diameter is 11.4 cm, the interior height is about 4.9 m.
Three sets of vertically mounted water cooled coils are located inside the com-
bustor to remove heat of combustion.  Coal is fed by a pneumatic injection
system similar to that used on the miniplant.  Sorbent is fed separately by
means of a transfer line lock hopper consisting of two cycling valves mounted
in the sorbent feed line.  Provisions were also added to inject supplementary
air into the combustor at various locations to study two stage combustion, and
to inject NHa and N2 to study the effects of NHa injection and simulated flue
gas recirculation on NOX emissions.

-------
      Two sets  of runs  were made studying the effect of two stage combustion.
 The first was  made  at  8 atm  pressure.  The runs were subject to frequent
 temperature excursions and bed agglomeration caused by low superficial veloc-
 ities 1n the bed below the point where the supplementary air was Injected.
 As  a result, the test  series was not completed.  The pressure was reduced to
 5 atm to Increase the  velocity and the second set of tests was run.  The
 location of the  supplementary air Injector was varied during the Initial tests
 at  8 atm.   Injection of the  air Into the bed, 15 cm above the grid, rather
 than above the bed, was necessary to promote CO burnout.  Despite the problems,
 some data  were obtained during the 8 atm tests which Indicated that NOX emis-
 sions could be reduced about 30% by using two stage combustion.  The test
 series at  5 atm  pressure was completed with no serious operating problems.
 The supplementary air  Injection point was raised to 33 cm above the grid to
 Increase the residence time  1n the lower reducing zone thereby promoting NO
 reduction  reactions.   However, the runs at 5 atm resulted 1n only a 10 to 20%
 reduction  1n NOX  emissions.  The Increase from 10 to 20% reduction occurred as
 the amount of  primary  combustion air was decreased from 90 to 75% of the
 stolchlometrlc amount  at a fixed level of total combustion air (primary plus
 supplementary).   The reason  for the apparent effect of pressure on the effect-
 iveness  of two stage combustion Is not known.  NOX emissions were also shown
 to  Increase with  overall  excess air and temperature as had been noted in
 previous studies  (1,2).

      Staged combustion increased S02 emission levels about 20%.  The effect
 was  believed to  be caused by low oxygen concentrations in the reducing zone,
 depressing the calcium sulfation reaction which requires the presence of
 oxygen.  CO emissions  were increased up to 20% by staging, but only at low
 excess air levels.  The increased S02 and CO emissions could be offset by
 increasing the gas residence time in the oxidizing zone slightly.  A more
 serious  concern  is the effect of two stage combustion on boiler tube materials.
 This  potential  problem must be addressed 1f two stage combustion is to be
 studied  further.

      NH3 injection was also evaluated as a means of reducing NOX emissions
 from  PFBC.   The use of NHo to reduce NOX emissions was developed by Exxon
 Research and Engineering Company and successfully applied to conventional
oil,  gas and coal fired furnaces.   The current program was designed to test
the effectiveness of NHo  injection under PFBC conditions.   A series of tests
was  conducted by Injecting NH3 at  NH3/NO ratios ranging from 1  to 8 at three
locations,  below the grid Into the combustion air, into the combustor at the
bed overflow port (168 cm above the grid) and into the combustor freeboard
region,  290  cm above the grid.  Combustor pressure was approximately 7 atm
during the  tests.  The effect on NOX emissions was found to be a strong
 function of location and the NHo/NOx ratio.  Injection below the grid into the
 combustion air resulted in a 50% Increase in NOX emissions.  Injection at the
 bed  overflow port had  no  effect on emissions.  Injection in the freeboard zone
 decreased  NOX  emissions 30 to 50%.  Increasing the NH3/NOX ratio Increased NOX
 emissions  at the  below-grid location, but decreased emissions at the freeboard
 zone location.  The results are probably related to local  oxygen concentra-
 tions and  temperatures.  Injecting below the grid exposed  NH3 to high 02 con-
 centrations  and high temperatures  near the coal  feed point.  These conditions
                                      10

-------
promoted NH3 oxidation rather than  the  NO/NH3  reaction.   In the freeboard
zone, the lower temperatures  (~700°C) and  Og concentrations favored NO reduc-
tion.  At the intermediate location,  the temperature, and possibly 02 concen-
tration conditions, promoted  each of  the competing reactions equally, resulting
in no significant effect on NOX emissions. The  optimum temperature for the
NH3/NO reaction found in this study (~700°C) is  lower than expected based on
previous studies and further  work may be needed  to determine if it is the true
optimum.  If such a temperature is  required, NHs could be injected ahead of the
gas turbine.  The desired temperature would occur within  the turbine itself,
where the NOX removal reactions would take place.

     The effect of flue gas recirculation  on NOX emissions was simulated by
adding nitrogen to the combustion air inlet stream.  The  ratio of N? to air
was set at 0.10 vs 0.2.  Excess air and temperature  were  also varied.  The use
of N2 addition was found to cause an  apparent  decrease in NOX emissions.
However, the addition of N2 always  caused  a decrease in the excess air level
since more coal had to be fed to heat up the cold N? added to the inlet air.
When the results were plotted against excess air and compared to results
obtained without N2 addition, no significant  effect  of  N£ addition was seen.
Therefore, it  is concluded that flue gas recirculation  would  have  no  effect
on controlling NOX emissions.  S0£  and  CO  emissions  were  also  increased
significantly, probably because of lowered oxygen partial  pressure and gas
residence time.

     A series  of tests was conducted  to determine if a  combination of two
stage combustion and NH3 injection  would result in  further  reductions  in  NOX
emissions.  The runs were made by adding the  control techniques  one  at a  time
and then in combination to a  run made under conditions  typical of  those  used
in the NOX control program.  It was found  that either  two stage  combustion  or
ammonia injection decreased NOX emissions  25  to 30%.  However,  the combination
of the two techniques resulted in a reduction  of only  26%.   Therefore, the
combination did not offer any advantages over  the use  of  either  method alone.
It should be mentioned that only one run was made and  that  additional  testing
might be needed to determine  if a positive interaction  exists.

     Regeneration studies were carried  out in  the modified  bench regenerator.
The unit now is equipped for  continuous feeding of  sulfated  sorbent  and  with-
drawal of regenerated sorbent.  The regenerator vessel  is refractory  lined  to
an inside diameter of 9.5 cm  and has  an interior height of  4.6 m.  The air
and fuel addition systems are similar to those used  in  the  miniplant  regen-
erator.  A below-grid grid burner provides most of  the  heat  required  by  the
system.  Additional fuel is added above the  grid to  create  a  reducing  zone
where CaS04 reduction occurs.  Supplementary  air is  added higher in  the  bed
to create the  oxidizing zone  needed to  oxidize the  undesirable  CaS byproduct
and complete combustion of the fuel.   The  test program  was  divided into  two
segments, natural gas fueled  regeneration  and  coal  fueled regeneration.   Most
of the work concentrated on developing  satisfactory  operating  techniques.   The
work done with natural gas-fueled regeneration included a study  of the most
effective way  to introduce fuel to  the  regenerator and  a  limited study of
operating variables and their effect  on sorbent regeneration.  Studies made
                                      11

-------
 with coal-fueled  regeneration were  preliminary  1n nature and were Intended to
 determine the feasibility  of operating  a regeneration unit with coal.  The
 Incentive to  use  coal  Instead of  natural gas of course, 1s to reduce operating
 costs and  minimize  dependence on  natural gas supplies.

      Attempts  were  made to add all  the  natural  gas fuel directly to the bed,
 using the  below-grld burner only  to preheat the bed to temperatures high
 enough to  Initiate  methane combustion.  This was done after 1t was found that
 about 30%  of  the  natural gas fed  to the burner was used to overcome heat
 losses from the burner plenum.  The attempts to add all the fuel directly to
 the  bed were  not  completely successful  due to the occurrence of frequent high
 temperature excursions.  The temperature excursions were believed to be caused
 by unstable combustion conditions in the regenerator.  Operations with natural
 gas  addition  to the burner as well as to the bed were more successful.  A
 series of  six  runs was made.  However,  because of the limited amount of data
 obtained it was not possible to draw conclusions concerning the relationship
 between operating variables and the extent of regeneration.  The S02 concen-
 tration 1n the off gas varied from 0.2  to 0.6%, equivalent to 15 to 30% of
 the  calculated equilibrium concentrations.  These low SO? concentrations
 occurred since the concentrations were not limited by kinetics or thermo-
 dynamics,  but rather by heat and material  balance requirements.

      A series of runs was made with coal added to the bed instead of natural
 gas.   The  results of the few runs were  promising.  The only serious operating
 problem was maintaining good temperature control.  Some of the runs were
 subject to high and erratic bed temperature.  Modifications, such as the use
 of nitrogen instead of air as the coal transport medium, improved temperature
 control significantly.   However, control problems persisted until  the end of
 the  program, possibly due to poor solids and gas mixing caused by the small
 regenerator diameter.  Solids regeneration levels (the percentage of the sul-
 fated  sorbent converted to CaO in the regenerator) were around 40%, somewhat
 lower than the levels of 40 to 80% measured with natural  gas fueled runs.
 S02 concentrations in the off gas were about 17% of the calculated equilibrium
 concentrations, roughly comparable to the  levels measured with natural  gas
 fuel.  More work is needed before coal-fueled regeneration can be completely
evaluated.
                                       12

-------
                                 SECTION  II

                                INTRODUCTION


     The pressurized fluidlzed bed  combustion  of coal  is  a  new  combustion
technique which can reduce the emission of $03 and  NOX from the burning of
sulfur-containing coals to levels meeting EPA  emission standards.   This is
done by using a suitable S02 sorbent such as limestone or dolomite  as  the
fluidlzed bed material.  In addition to emissions control,  this technique
has other potential advantages over conventional  coal  combustion systems
which could result in a more efficient and less costly method of electric
power generation.  By immersing steam generating surfaces in the fluidized
bed, the bed temperature can be maintained at  low and  uniform temperatures  in
the vicinity of 800 to 950°C.  The  lower  temperatures  allow the use of lower
grade coals (since these temperatures are lower than ash  slagging tempera-
tures), and also decrease NOX emissions.   Operation at elevated pressures,  in
the range of 600 to 1000 kPa, offers further advantages.   The hot flue gas
from a pressurized system can be expanded through a gas turbine, thereby
increasing the power generating efficiency.  Operation at the higher pressure
also results in a further decrease  in NOX emissions.

     In the fluidized bed boiler, limestone or dolomite is  calcined and
reacts with SOg and oxygen in the flue gas to  form  CaS04  as shown  in reaction
v' / •

                         CaO + S02  + 1/202  +  CaS04                      (1)


     Fresh limestone or dolomite sorbent feed  rates to the boiler  can be
reduced by regeneration of the sulfated  sorbent to  CaO and recycle  of the
regenerated sorbent back to the combustor.  One regeneration system, studied
by  Exxon Research and Engineering Company in the past, is the so-called one
step regeneration  process in which sulfated sorbent is reduced  to  CaO in a
separate vessel at a temperature of about 1100°C according to equation (2).
The goal is to produce SOg in the regenerator off gas  at a sufficiently high
concentration to be recovered in a by-product sulfur plant.

                               CO                 C02
                       CaS04 + H2  +  CaO + S02 + H20                     (2)


     A diagram of  the  pressurized fluidized bed combustion and  regeneration
process is shown in Figure II-l.

     Exxon Research and Engineering Company, under contract to  the EPA, has
built two  pressurized  fluidized  bed combustion units to  study  the combustion
and regeneration processes.  The smaller of the  two units, the  bench scale
unit, was  built under  contract CPA 70-19 and was described  in  previous
reports  (1,2,3,4).  Those reports also described regeneration  and combustion
studies carried on in  the bench  unit.  This report describes work done in the
                                      13

-------
                               FIGURE ll-l
            PRESSURIZED FLUIDIZED BED COAL COMBUSTION SYSTEM
                                      AIR COMPRESSOR
          GAS TURBINE
     CONDENSER
   COAL AND
SORBENT MAKEUP
                         HIGH EFFICIENCY
                         SEPARATOR
                                             SOLIDS
                                            TRANSFER
                                            SYSTEM
                                                                r\TO SULFUR
                                                                  RECOVERY
                                                               SEPARATOR
                                                             I
                                                          DISCARD
                              COMBUSTOR
Tl
FUEL
                                                        REGENERATOR

-------
bench combustor evaluating various  NOx  emission control methods.  Regenera-
tion studies were conducted in the  bench regenerator evaluating fuel injection
methods and the use of coal as the  regenerator fuel.   These results are also
included in this report.

     The larger unit, called the mini pi ant,  was designed  under EPA  Contract
CPA 70-19 and built under Contract  68-02-0617.  Figure II-2 shows a photo-
graph of the miniplant.  The shakedown  and operation of the unit was funded
under Contract 68-02-1312.  Previous reports (1,2,3,5) described design,
shakedown and operation of the unit.  This report  includes additional results
from the operation of the combustion and regeneration  sections of the mini-
plant.  The combustion program included tests to verify and document further,
the ability of PFBC to reduce S02 emissions  by 90% or  more.   The dynamic
response of S02 emissions to a rapid change  in coal sulfur content  and dolo-
mite to coal feed ratio was also measured.  This was done to  provide the  basis
for the evaluation of an S02 emission control concept  by  General Electric
Company under a separate EPA contract.   Additional data on the emission of
NOX, other gases and particulates were  also obtained.   Particulate  control
studies evaluating high efficiency  cyclones, a high temperature/pressure
ceramic fiber filter, a granular bed filter and a  low  temperature/pressure
electrostatic precipitator and bag  house were also completed. This work  was
done in cooperation with a number of other EPA contractors including Acurex
Corporation, Southern Research Institute and Air  Pollution Technology,  Inc.

     A series of regeneration tests was also conducted.  Three extended runs
were completed in which activity maintenance of  the regenerated  sorbent,  SOg
concentrations in the regenerator off gas and S02  retention  in  the  combustor
were measured.

     A series of sampling campaigns was conducted in  cooperation with  another
EPA contractor, GCA/Technology Division, to provide samples  for  Level  I and
Level II comprehensive analysis test programs.   Samples were  taken  during
operation of the combustor alone and during regeneration  tests  in which the
combustor and regenerator were both in operation.

     The period of performance discussed in this  report is August  12,  1977
to August 7, 1979.
                                      15

-------
                 FIGURE 11-2
EXXON FLUIDIZED BED COMBUSTION MINIPLANT
                    16


-------
                                 SECTION  III

                        MINIPLANT COMBUSTION STUDIES


     Combustion studies  have been carried out  in  the  EPA/Exxon pressurized
fluidized bed combustor  referred  to  as  the miniplant.   The miniplant is shown
schematically in Figure  III-l.  This unit has  provisions  for continuous
addition of coal and sorbent and  continuous withdrawal  of sulfated sorbent.
It normally operates at  pressures of about 9 atm, temperatures of 840 to
950°C,  with superficial  velocities of about 1  to  2.5  m/s, feeding 100 to
160 kg/hr of coal  and 15 to 20  Nm3/min  of combustion  air. As of August 1979,
the combustor has  been operated for  a total of approximately 3700 hours in a
series  of individual runs up to 250  hours duration.   This section of the
report  describes the combustor  equipment, operating procedures, combustor
performance and combustion results.   A discussion of  the  regeneration work is
given in Section V.

EQUIPMENT, MATERIALS, PROCEDURES

     This section will focus on the  major system  components which include:
(1) solids feeding system, (2)  combustor  with  internal  subcomponents,  (3)
flue gas system, (4) temperature  and pressure  control,  (5)  flue gas  sampling
and analytical system, (6) process monitoring  and data  generation system,  (7)
combustor safety and alarm system, (8)  coal and sorbent properties,  and  (9)
operating procedures.  A detailed description  of each of these  systems can be
found in earlier reports (1.2)  and only a brief discussion  will be  included
here.  Changes in the equipment made since the last report  are  also  included.

Soljds  Feeding System

     Figure III-2 displays a schematic of the  miniplant coal and  sorbent
feeding system.  Crushed and sized coal and limestone or dolomite are  held
in separate storage bins (20 tonnes  for coal  and 3 tonnes for sorbent) under
atmospheric conditions.   On demand,  the solids from the bins are  proportioned
to a specific coal/sorbent ratio.  Inverters  control  the motor  speeds  of
separate coal and sorbent screw feeders and volumetrically  control  the coal/
sorbent ratio.  A blending screw  transports the mixture into a  solids  feed
vessel.  The coal/sorbent mixture is held in  this vessel  until  refill  of  the
injector vessel is required.

     The solids feeding  system provides for continuous  solids delivery  (coal
and sorbent) from the injector vessel to the  pressurized combustor,  while
allowing intermittent refilling of the injector vessel  (193 kg  operating
capacity).  Load cells located under the injector vessel  monitor  the solids
feed rate and actuate control signals for the refill  cycle.  Prior  to  initia-
tion of a refilling operation, the injector vessel,  feed vessel,  and the  pair
of solids storage bins remain isolated from each other.  When the  load cell
under the injector vessel detects a  solids loading of less  than 102  kg,  91
kg of solids are automatically transferred pneumatically from the  feed  vessel
                                     17

-------
                                             FIGURE 111-1

                             EXXON FLUIDIZED BED COMBUSTION MINIPLANT
                                                                    COOLING
                                                                     WATER
                                                                                   TO  SCRUBBER
CO
                                                          ORIFICE
                                                         r	
      THIRD STAGE
       SEPARATOR
                                                                    AIR (PRESSURE CONTROL)
                        CYCLONE
                       SEPARATOR
                                                                   TOOLING
                     COOLING WATER
                        OUT   IN
                                                                                   TO
                                                                              SCRUBBER
                                                                       CYCLONE
                                                                       SEPARATOR
                                                                           SOLIDS
                                                                           DISCHARGE
                               SOLIDS
                               REJECT
                              VESSELS
    FEED
   WATER
 RESERVOIR

 COAL AND
 LIMESTONE
 FEED SUPPLY
AUXILIARY
AIR
COMPRES
       r
  LIQUID FUEL STORAGE
    NATURAL
      GAS
  COMPRESSOR

MAIN AIR
COMPRESSOR

-------
LIMESTONE
   BIN
             COAL
              BIN
                                   FIGURE 111-2
                           COAL & LIMESTONE FEED SYSTEM
                                               CONTROLLERS
                                   AP SIGNAL -eee*>
 [FEEDER I    |  FEEDER  |
       BLENDER
J
        VENT

 04  _n|


3j~
      FEED
      VESSEL

h*18"-»-
    	S-	&3
     HIGH PRESSURE AIR
                                   D-D
                                                 TC
                                              I  I I  I
                                                 AP   TEMP
                            t
                                 i	>^-
                               INJECTOR
                                VESSEL
                                                                     MAI
                                                                    I  II II U U I
                                                                    COMBUSTOR
                   	A
                                                                 1/2"  S.S. PIPE
                                                                   (LOOPED)

-------
 to the  pressurized injector vessel without interrupting feed to the combustor.
 Refilling  is usually completed in about 5 minutes.  After refilling, the feed
 vessel  is  again isolated from the injector vessel, vented, and filled with
 solids  from the storage bins.  The feed vessel is then isolated and repres-
 surized to await another cycle.

     Solids in the injector vessel are continuously aerated by the pressurized
 air stream, and the vessel is automatically controlled at a pressure level
 slightly above that in the combustor.  Originally the port used to sense com-
 bustor pressure was near the top of the combustor where plugging with bed
 material would not occur.  Maintenance of the proper pressure differential
 between the combustor and the injector vessel during minor bed upsets was dif-
 ficult.  These upsets sometimes resulted in hot solids back flow into the
 injector vessel.  For this reason, the combustor pressure tap was relocated
 to a port very near the coal  injection nozzle.  A large purge flow rate in the
 pressure sensing line and a high coefficient of damping 1n the measurement
 system were used to control plugging and oscillations.  This system has worked
 without incident for over one year.

     The coal  and sorbent mixture 1s discharged from the injector vessel
 through a 1.3 cm diameter orifice and pneumatically conveyed by a stream of
 dried transport air through a 1 .6 cm ID stainless steel pipe.  Originally
 the pipe was S-shaped.  To increase the pressure differential between the
 combustor and the injector a  longer conduit with one 180° and four 90°  bends
was installed.  The conduit is 17 m long with 30 cm radius bends.  The  pres-
 sure differential  between the injector vessel and the combustor was increased
 from about 20 kPa to 70 kPa.   The erosion of the coal  transport conduit at
 the location of the four bends consistently occurred after about 200 hours of
operation when the conduit was made of 0.37 cm thick, type 304 or 316 stain-
 less steel heavy wall pipe.  Heat treated, type 410 stainless steel pipe
 bends although much harder than 304 or 316 stainless steel, did not produce
any better results.  Carbonized steel pipe failed after only 60 hours of
operation.  Limited data were obtained with type 304 stainless steel  pipe
 that had been alonized.  After approximately 150 hours of operation, the
alonized conduits were still  functioning.

     A comparison of the locations where the holes were eroded in the conduit
 bends of various materials showed that in all cases the failure occurred when
 the angle of solids impingement upon the wall was between 15 and 20 degrees.
A segment from a type 304 stainless steel  bend showing the hole location
after 200 hours of operation  is shown in Figure III-3.

     An auxiliary solids feed system was constructed to allow changing  from
 one coal/sorbent feed to another instantaneously.  The capacity of this new
 vessel  is equivalent to about 8 hours of operating inventory.  It 1s situated
 on load cells to monitor feed rate.  There is no provision for on-line  refil-
 ling.  The coal/sorbent blend exits the auxiliary vessel  through a 1.3  cm
 orifice and is pneumatically  conveyed by a stream of dried transport air
 through a 1.5 cm ID polyurethane tube.   This material  was chosen to minimize
 line erosion, because of the  excellent resilience of polyurethane.  This  tube
 is shielded in the event of rupture.   The tube is 12.5 m long and winds once
 around the 2.4 m diameter vessel  before it reaches the combustor.

                                     20

-------
               FIGURE 111-3
CROSS  SECTION OF COAL/SORBENT FEED LINE
                 I

-------
      Final  entry of  solids  Into the combustor 1s through a 1.3 cm ID nozzle
 located 28  cm above  the fluldlzlng grid and horizontally extending about
 2.5 cm beyond the reactor wall.  Both the primary and the auxiliary feed
 systems have  their own Identical entry ports located 180" apart on the com-
 bustor.  The  tip of  the probes Include ten 0.79 mm diameter holes which sur-
 round  the solids  feed opening.  They are used to continuously Inject an
 annular stream of sonic-velocity air to assist penetration of the solids
 feed Into the fluldlzed bed and to protect the feed nozzles from blockage
 with bed  solids.   Only one  probe Is used at any one time.  The probe not 1n
 use 1s  continuously  purged  to prevent blockage of the openings with sol Ids.
 The flow  of solids Into the combustor 1s controlled to maintain constant
 temperature 1n the combustor.

 Combustor

     The  combustor consists of a 61 cm ID steel  shell, refractory lined to an
 Inside  diameter of 33 cm.   The 9.75 m high unit 1s fabricated 1n flanged sec-
 tions to  allow Insertion and removal of the cooling colls.  Various ports are
 strategically  located to allow for material  entry and discharge.  Numerous
 taps are  also  provided for monitoring both pressure and temperature.  A
 schematic of the combustor  1s shown 1n Figure III-4.

     Heat removal from the combustor 1s provided by cooling colls located in
 discrete vertical zones above the grid.  Each coll  has a total  surface area
 of 0.55 m* and consists of vertically-oriented loops constructed of 1/2-Inch
 Schedule 40 316 stainless steel  pipe.  The number of colls normally varies
 from one  to four depending on the combustor operating conditions and the
 amount of cooling required.  A high pressure pump is used to pump the cooling
water through  a closed-loop arrangement consisting of a dem1neral1zed feed
water reservoir, cooling coils,  and a heat exchanger.  The flow rate and exit
 temperature from each coil  can be separately controlled and monitored.  Baf-
 fles were installed on the cooling colls  to prevent excessive  erosion by the
 bed solids (2).

     Only one  failure of the baffled cooling colls has been recorded after
 1200 hours of  operation.  A hole (Figure II1-5)  developed on the lower por-
 tion of the coil closest to the bed support grid, facing the center of the
 combustor, approximately 45° of arc along a bend.  The lowest portion of the
 coll was  only  23 cm above the fluldlzlng grid.  The section of coil was
 analyzed  and the failure was attributed to erosion,  Very little oxidation
 or sulfidatlon was present.  The lack of sulfldation 1s not surprising, since
 the coil water outlet temperature 1s controlled  at 140°C.  At  this Internal
 temperature, little sulfidatlon  1s expected.  Most of the 180° bends 1n the
 lower section  of the coll  also exhibited  a decided sharpening  of the leading
 edge.   This is also attributed solely to  erosion.  The entire  coil  1s shown
 in Figure III-6.  A similar coll  located  further from the fluidlzing grid was
 exposed over the same time period with no significant erosion.   The erosion
 problem was apparently solved by welding  small  rectangular steel "guards"
 horizontally at the bottom of the colls.   These  should prevent erosion but
 have been used for only 500 hours of operation to date.
                                     22

-------
   FIGURE 111-4
COMBUSTOR VESSEL
       23

-------
              FIGURE 111-5

MINIPLANT COMBUSTOR COOLING COIL 1A
 AFTER 1200 HOURS OPERATION (CLOSE UP)
                  24

-------
                 FIGURE 111-6

MINIPLANT COMBUSTOR COOLING COIL 1A AFTER
            1200 HOURS OPERATION
                     25

-------
      The  combustion  air to the unit 1s provided by a main air compressor
 having  a  capacity of 40 Nm3/min at 1030 kPa (1400 SCFM at 150 pslg).   Preheat
 of the  combustor during startup 1s made possible by a natural gas burner which
 1s housed 1n  the bottom plenum section of the combustor.  Once the fluldlzed
 bed temperature reaches approximately 430°C, a liquid fuel system 1s  used to
 heat the  bed  further to the coal Ignition temperature.  Both systems  are more
 completely described In a previous report (1).

      With continuous addition of sol Ids to the combustor, there must  be pro-
 vision  for continuous removal to control  combustor expanded bed height.  The
 expanded  bed  height can be controlled at any level above 2.3 m by the con-
 tinuous withdrawal  of bed solids through a port located 2.3 m above the
 fluldlzlng grid.  Originally, solids  flowed by gravity through a refractory
 lined pipe Into a "pulse pot" from where they were pneumatically transported
 by controlled nitrogen pulses to a pressurized lock hopper.  This system did
 not give  precise solids flow control  and was replaced with a dual valve sys-
 tem.  Solids now flow by gravity to a refractory lined sliding gate valve.
 This valve and a ball valve behind 1t alternately open and close an adjustable
 number of  times per hour.  Each cycle rejects 4 kg of bed material, the amount
 of material that fills the volume between the two valves.  From the ball  valve.
 the sol Ids empty Into a lock hopper which 1s manually emptied at regular
 Intervals.

     The combustor  fluldlzlng grid consists of 332 3/32 Inch holes and four
 Independent cooling water loops consisting of five channels each.  It has
 performed since run 50, without Incident.  A more complete description Is con-
 tained 1n a previous report (1).

 Flue Gas System

     Combustion gases exit the combustor and Immediately pass through two
cyclones 1n series.  From the second cyclone, the gas passes through  a long
 duct to a large pressure vessel.  This vessel was designed to contain a Ducon
 granular bed filter or other high efficiency hot gas cleanup device.   After
completion of the granular bed filter tests (run 66), the filter was  replaced
with a third cyclone Installed within the vessel.  When the third cyclone Is
used, no flow enters the vessel; 1t 1s merely a pressure containment  device.
After passing through a third cyclone (or the granular bed filter 1n  the
earlier tests), the flue gas 1s sampled,  expanded and sent to a wet scrubber
for final  cleanup.   The dimensions of all  three cyclones are shown 1n Figure
III-7.  The dimensions of the second  stage cyclone have been modified since
It was originally built.  Originally, the barrel diameter was 31.8 cm and the
inlet section sized to give an Inlet  velocity of about 10 m/s.  After run 18.3,
the inlet dimensions were changed to  Increase the inlet velocity to about
30-35 m/s.  The outlet diameter was decreased.  The final change was  made
after run 47 to the dimensions in Figure  III-7.  These changes decreased the
 barrel diameter but maintained the inlet section dimensions to keep the Inlet
velocity range fixed.

     The  granular bed filter, used 1n tests prior to run 67, 1s described in
 Section IV.
                                      26

-------
                                  FIGURE  111-7
                        CYCLONE DESIGN  DIMENSIONS
                  w h—
          INLET
        SECTION
                                           T
                                            ^o
                                           I
                                                              \ i
                                                              i,
                                                  DIMENSIONS (CM)
SYMBOL
  Db
 W
 h:
      DESCRIPTION
Inlet Type
Barrel Diameter
Inlet Width
Inlet Height
Outlet  Pipe  Insertion
Barrel Height
Cone Height
Outlet  Pipe  Diameter
Outlet/Inlet Ratio
Diameter of  Cone at Bottom
Function
CYCLONE
NO. 1
Tangential
32.4
7.62
25.4
20.3
43.2
45.7
14.6
0.87
12.7
Recycle
CYCLONE
NO. 2
Tangential
17.8
4.45
10.2
13.3
35.6
35.6
8.89
1.37
9.0
Cleanup
CYCLONE
NO. 3
Tangential
15.2
3.81
7.62
11.4
40.6
20.3
6.2
1.04
9.0
Cleanup
                                      27

-------
     The primary Intent of the first cyclone 1s to redrculate larger unburned
 carbon particles back to the combustor to Improve combustion efficiency.   The
 particles fall down the dlpleg and are relnjected with a nitrogen pulse about
 66 cm above the fluldlzlng grid.

     Particulates collected 1n the second stage cyclone are removed through
 a lock hopper.  The cyclone, which was originally cast with Grefco LUecast
 75-28 refractory, was recast after run 47 with Resco RS-17-E, and Is still  1n
 service after 2-1/2 years.  None of the severe pitting that was present with
 the earlier refractory was found.

     From the second cyclone, the flue gas enters the large pressure vessel
which contains the third cyclone.  This cyclone 1s of conventional  design  and
constructed of 316 stainless steel.  Particulates captured by the third
cyclone are removed through a lock hopper.

     Following the third cyclone, the flue gas may either be expanded through
a pressure control  nozzle or a turbine test section supplied by General
Electric under contract with the Department of Energy (DOE Fireside Corrosion
Contract No. EX-76-C-01-2452).  The test section contains 24 stationary
blades of various materials that simulate both the Impulse and reaction vanes
of a gas turbine.  The turbine section was tested for 1100 hours of operation
with three stage cyclone partlculate cleanup as part of the DOE Fireside
Corrosion program.   A more complete description 1s given 1n a report to DOE
 (6).

     In order to simulate gas turbine operation, care must be taken to prevent
the gas temperature from dropping below 843°C (1550°F).   Below this tempera-
ture, gas phase alkali will condense on duct walls or on partlculate matter.
This would unrealIstlcally reduce the vapor phase alkali concentration to
which the turbine blades are exposed.  For this purpose, four natural  gas
Injection ports were located between the second and third cyclones.  The
locations of the four ports were staggered to maintain gas temperatures
between 857°C and 900°C.  The total  amount of natural  gas (typically 0.6 to
0.9 Nm3/m1n) 1s controlled to maintain a set third cyclone Inlet temperature.
The flow distribution between the four probes 1s manually controlled.   The
Injection probe consists of a closed end 3/8 Inch alonlzed tube projecting
to the duct centerllne with a 0.10 cm diameter hole facing downstream.  The
probes are purged with nitrogen when not 1n use.

     The natural  gas Injection system 1s frequently used to hasten  thermal
equilibrium In the  mlnlplant off gas system.  Heating the refractory to
steady state would  otherwise require 8 to 12 hours.  With natural  gas  Injec-
tion, this 1s reduced to two hours.   The effects of natural  gas Injection
will  be reported In a following section.

Temperature Control

     The rate of solids feed to the  combustor 1s automatically controlled  1n
order to maintain a specific operating temperature within the combustor.  This
1s accomplished, through a series of controls which adjust the pressure


                                     28

-------
differential  between the combustor  and  the  Injector vessel, thereby varying
the coal  feed rate 1n a way to  maintain the proper temperature,  This system
1s described  in detailed In a  previous  report  (1).

Pressure Control
     The FBC mlnlplant combustor has  the  capability of operating at pressure
levels of up to 10.5 atmospheres (140 pslg).   Normally,  pressure control Is
achieved by restricting the discharge flow of  all  the gas from the combustor,
so as to achieve an Increase 1n back  pressure.  This 1s  done by use of a
silicon carbide converging nozzle Inserted 1n  the  discharge line.  Adjustment
of the combustor pressure 1s accomplished by metering high pressure air Into
the discharge line just upstream of the flow nozzle.  A  2 Inch-ball valve
equipped with a pneumatic positioner  and  actuator  regulates the amount of
air added 1n response to a signal from the pressure controller.  A more com-
plete description 1s contained 1n a previous report  (1).

     During the time that the combustor gas passed through the General Electric
turbine test section, this turbine section acted as the  back pressure regulator
for most of the gas.  The converging  nozzle was  disconnected during these
tests.  However, to maintain proper pressure control, a  small slip stream  of
gas (-10%) was diverted prior to the  test section  and expanded through a smaller
variable nozzle.  Makeup air was added just upstream of  this nozzle through
the same valves and controllers that  controlled  pressure 1n the  normal opera-
ting mode.  This variable nozzle (Figure III-8)  was adjusted to  maintain 5 to
9 Nm3/min makeup air flow.  The nozzle was constructed of 316 stainless  steel
and performed satisfactorily for over 700 hours  of testing.

Sampling and Analytical Systems

Gas Sampling System--
     Flue gas is sampled at a point about 20 m downstream of  the third  stage
cyclone.  The system Is designed to produce a  solids-free,  dry  stream of flue
gas at approximately ambient temperature and  atmospheric pressure whose  com-
position, except for moisture, is essentially unaltered  from  that of  the
original flue gas.  The conditioned gas 1s analyzed  in  a series  of continuous,
on-Hne analyzers for S02, CO, C0£, Og and NOX concentration.   The flue  gas
sampling and continuous analytical system was  described  in  a  previous  report
(2).  Another sample of flue gas can be extracted  which  has  been filtered,
cooled and depressured, but not dried.  This  system  1s  used to  obtain  batch
samples of flue gas for analysis of SOa and SOa by wet  chemistry methods.
The method to determine $03 concentration was  modified  since  the publication
of the previous report.

     Previous measurement of $63 from the FBC mlnlplant consisted of  flowing
the flue gas through an 80% Isopropyl alcohol  (IPA)  Impinger.   Because  of
potential problems with this method, such as possible contamination of IPA
lots with small quantities of  peroxides which oxidizes  SOg to  SOa, collection
of S03 1s now accomplished by  using a controlled condensation coll.  The coil
1s maintained at 60eC, which is well above the water dew point and low enough
to remove most of the 503 from  the gas phase.


                                      29

-------
            FIGURE 111-8
VARIABLE PRESSURE REDUCING  NOZZLE
                30

-------
Figure III-9 shows the sampling train as  it is  now  used.   Participate
material  is removed at 9 atm by a  Balston filter heated to 288°C, well above
the $03 dew point under FBC conditions.   The cleaned gas  is delivered  to the
gas sampling train through a heated (288°C) stainless  steel line  followed  by
heated teflon tubing.  Collection  of S03  as H2S04 is accomplished by using a
Goksoyr-Ross controlled condensation coil.  The 80% IPA impinger  is used as
a backup for 863 collection with IPA known to have  a negligible peroxide
contamination.

     H2S04 collected in the condensation  coil and the IPA impinger  is  titrated
with barium perchlorate reagent and thorin indicator.   Comparable results
were obtained with an acid/base titration using bromphenol  indicator.  Two
H202 impingers are used to trap S02 which is also determined using  barium
perchlorate/thorin titration.

Particulate Sampling Systems

     The miniplant has three particulate  sampling systems.  These systems  were
installed or modified as the need for more and different  particulate sampling
information arose.  The oldest system referred to as the  "Balston Filter  Par-
ticulate Sampling System" is described in a previous report (1).   The  Balston
filter elements consist of borosilicate glass fibers bonded with  an epoxy
resin.  A newer system is capable of sampling at high temperature and  high
pressure with an in-situ size distribution measuring device.  This  system is
referred to as the high temperature high  pressure (HTHP)  particulate sampling
system.  It was constructed during this reporting period  to sample  the flue
gas with a 5-stage cyclone train developed by the Southern Research Institute
or a University of Washington 7-stage impactor.  Both the Balston filter and
the HTHP sampling systems sample particulates in the gas  leaving the tertiary
cyclone.  The third system samples the particulates in the gas entering the
tertiary cyclone.  This system, referred to as the  Upstream Balston Filter
Particulate Sampling System, was only installed after run  94.  It was  used to
determine cyclone capture efficiency.  The locations of all of these systems
is shown in  Figure 111-10.

      The old  Balston  Filter  Particulate Sampling System has been modified
since  its description  in a  previous report  (1).  Originally the system con-
tained three  18 cm Balston  total filters  in  series  followed by coolers, a
water  knockout, flow  control valve and rotameter.   The secondary and tertiary
Balston filters were  removed from the system.   During the  sampling it was
found  that  their weight gain contributed  only a  negligible amount to the
total  loading and what was  captured  had  the  appearance of  wet corrosion pro-
ducts. A wet test meter was also added  after  the rotameter.  This  allowed
for a  more  accurate  measurement of sample  flow  than was previously possible
with  a rotameter and a timer.  A schematic  of  the  system  as modified at the
end of this  reporting  period is shown in  Figure  III-ll.   This system has
provided a  large  data  base  of  particulate  measurements.   It has been used with
probe sizes  between  1.09  and 0.46 cm and  isokinetic flow  rates between 20 and
200 Ndm3/min, all with good  results,
                                      31

-------
 SLIP
    FROM
PFBC EXHAUST
           FIGURE 111-9

CONTROLLED CONDENSATION COIL
    ATTACHED TO A SO3-SO2
         IMPINGER TRAIN
to
ro
                                                   S
                                      TEFLON
                                      TUBING
                BALSTON
                 FILTER
                       STAINLESS STEEL MANIFOLD
                      HEATED SECTIONS
                     TEMPERATURE 288°C
                                                            CONTROLLED
                                                           CONDENSATION
                                                                COIL
                                                                  .
                                                                  1
                                                                  !
                                                            00
                   FRITTED
WET TEST METER
                                                            0
                                                                    0
        0
        0
                                                                                 "\
                                                               H^    H2O2   BLANK     IPA

                                                                  SO3-SO2 IMPINGER TRAIN
                                                                                        ICE
                                                                                       BATH

-------
                                          FIGURE 111-10


                                  MINIPLANT FLOW SCHEMATIC
OJ
CO
.^-~— >.
/^N

c

M
B
U
S
T
0
R












^i



<

N/G,
Kl //-»
k ) 2ND IN/ii
\X rvrinMF

1ST
Cyclone


sS


x^
1
1








N/G 	 '
UPSTRI
BALST
EAM
ON -

1
















G.E.
TURBINE TEST
HTHP
PARTICULATE
SAMPLING



J
FILTER




PAKIICULAIh T"
SAMPLING ..L
SYSTEM N/G






SYSTEM









\ /
V



(
SECTION
_ - -i
r >






1 ' 1 '

ALKALI '
\ — SAMPLING
SYSTEM


BALSTON
FILTER
PARTICULATE "~





SAMPLING
SYSTEM
. _ GAS SAMPLING
»-*-~ SYSTEM
x-v
"?_
t~*—>3 	
MAKE UP AIR
OFF GAS
COOLER




N/G i
ACUREX ESP
OR BAGHOUSE
   N/G = NATURAL GAS INJECTORS
   3RD
CYCLONE
                                              t

-------
     FLUE
      GAS
 1/2"
PROBE
CO
                                          FIGURE 111-11

                            BALSTON FILTER PARTICULATE SAMPLING SYSTEM
                                   TC
                           TEMP.
                           VALVE
BALSTON
 Fll TFR
 NLTtK
                        STEAM HEAT
                        EXCHANGER
                                                                                     WATER
                                                                                  CONDENSER
                                                              FLOW
                                                              MEASUREMENT  FLOW CONTROL
                                                              ROTAMETER    /    VALVE
                                                                 WATER '
                                                               KNOCKOUT
                                                       WET TEST METER (1  SCF/Rev)

-------
     The HTHP sampling system is  designed  to obtain  isokinetic,  isothermal
samples of participate laden  flue gas.   The system consists of a long 0.95 cm
(3/8 inch) diameter tube projecting  into the flue gas  stream along the duct
centerline.  Once outside the flue gas  duct, the tube  is  heated with electric
heaters to maintain a set gas temperature. A  high temperature (Kamyr) ball
valve provides system isolation during  combustor operations.  Following the
valve, the gas enters a pressure  furnace which contains the sampling device.
This furnace is large enough  for  most in-stack sampling devices.   It is
electrically heated to a controlled  temperature (usually  1600°F) and pres-
surized with helium to prevent sample leakage  or high  temperature  corrosion
of the steel.  After the furnace, the gas  passes through  a backup  Balston
filter and some gas coolers.   Then the  gas passes through a water  knockout;
another filter to remove the  condensed  acid mist; and  the flow control sys-
tem.  Initially, the flow control system was automated to maintain a set  flow
rate over a span of sampling  device  pressure drops.  This automatic system
was prone to plugging and corrosion  failure.   The automatic flow controllers
were replaced with a manual  valve.  This valve has performed satisfactorily
for many samples.  Following  pressure let  down, the  gas passes through a  gas
saturator and a wet test meter that  measures total flow.  Figure 111-12  is a
schematic of the original system.  Figure  111-13  is  a  schematic  of the system
as modified at the end of the reporting period.

     One additional modification  that was  made to the  HTHP sampling system,
that could not easily be shown in the figure,  is  the provision to  bypass  the
furnace and the small (7 cm)  backup  Balston filter and install an  18 cm
Balston total filter in their place.  In this  way, the probe and flow  control
system can be used for sampling at 500°C with  one 18 cm Balston  filter as well
as at 850°C with the other devices.   The advantage of  the Balston  filter  is
rapid and reliable particulate concentration measurements.

     The in-situ size distribution measuring devices used in  the HTHP  parti-
culate sampling system were a Southern Research Institute 5-cyclone train and
a University of Washington 7-stage impactor.   The  5-cyclone  train  (Figure
111-14) was housed in the pressure furnace in  the vertical  position.   Two
cyclones were upside down during sampling.  A  complete description of  the
cyclones is given  in a separate report by SoRI which also details  cold flow
calibrations  (7).  The 7-stage impactor was also installed  in  a  vertical
position, with a precutter cyclone before the  impactor.   The  cyclone was  there
to prevent overloading of the first stage due  to large concentrations  of
large particles.   The impactor is described in a technical  bulletin  (8),  a
picture is shown in Figure 111-15.

     The third particulate sampling system was installed  after run 94  to
measure particulate loadings entering the tertiary cyclone.   This  system was
built mainly  to verify the high collection efficiency of the  tertiary  cyclone.
This efficiency was previously obtained by completing a mass  balance  with the
outlet filter sample and the captured lock hopper material.   This  sampling
system consists of a 6 mm probe projecting to  the duct centerline.  Filter
isolation  is  provided by two bellows seal  valves (650°C temperature limit).
Immediately after  the valves is the 18 cm Balston total  filter.   Following the
filter a large cooling coil  provides natural  convective cooling.  This is
followed by a water knockout, manual flow control  valve,  gas  saturator and

                                     35

-------
                                        FIGURE 111-12

                         ORIGINAL HTHP PARTICULATE  SAMPLING SYSTEM
                                                     PROBE   I
                                                   NITROGEN X
                                                     PURGE	T_
                                        HEATER
                                        KAA)
         FLOW
      CONTROLLER
               SAMPLE
              NITROGEN
                PURGE
                                                         *
CO
o>
                     BACK PRESSURE
                     REGULATOR
 0!
-tk-L
                         t
                                             BALSTON
                                              FILTER
      WATER
    KNOCKOUT
 BACK PRESSURE
'REGULATOR
                                         TEMP.
                                         VALVE
                                    WATER
             GAS
           SATURATOR
                                        SAMPLING
                                         DEVICE
                                         HEATER
                                       VENT
                         PRESSURE
                         VESSEL
                             WET TEST
                               METER
                                   HEATER
                                           VESSEL
                                      PRESSURIZING
                                            AIR
                                                          FLUE
                                                          GAS

-------
                                            FIGURE 111-13
                            HTHP  PARTICIPATE SAMPLING SYSTEM AS MODIFIED
      BALSTON
CO
--4
     HIGH  TEMP
       FILTER
   FLOW CONTROL
       VALVE
                         WATER
                       KNOCKOUT
                     GAS
                  SATURATOR
BALSTON








V
OFF GAS
COOLER
1
VATER


1-ILItK
(BACK
UP)

1

                                        VENT
 SAMPLING
   DEVICE
                                                  HEATER
                                    WET TEST METER
    PROBE
NITROGEN I
    PURGE X
                                                                                 HEATER
              SAMPLE
              NITROGEN7HIGH   rAAAn
              .  PMRnp  4TEMP   r  v v v I
                        VALVE
                                                                           HEATER
                                                             PRESSURE
                                                             VESSEL
                                                                       —H!	
            VESSEL
         PRESSURIZING
            Helium
                                 3/8"
                                PROBE
                                         FLUE
                                         GAS

-------
                FIGURE  111-14
SOUTHERN RESEARCH INSTITUTE 5-CYCLONE TRAIN
                     38

-------
                FIGURE 111-15
UNIVERSITY OF WASHINGTON 7-STAGE IMPACTOR

                     39

-------
 a wet test meter.   The flow 1s  set  to  1sok1net1c conditions by timing the gas
 flow through the wet test  meter.  This  system has verified efficiencies
 obtained earlier by the mass  balance technique.  Low flow rates are required
 due to the limited  natural  convectlve  cooling.

 Alkali Probe Train

      An alkali  probe train was  designed and constructed by Exxon Research to
 enable acquisition  of a hot pressurized flue gas sample before the turbine
 test section (Figure 111-10).   The  probe was designed to measure vapor phase
 Na  and K concentrations of the  flue gas at 840°C and 9 atm pressure.  Figure
 II1-16 1s a  schematic of the  alkali probe train.  The temperatures shown are
 those for test  4 of run 78; these are  representative of normal operating
 conditions.

      The  hot pressurized flue gas enters the probe at system temperature and
 pressure  and 1s  preflltered through three layers of astroquartz.  Na and K
 vapors  1n the flue  gas  are  then condensed on the walls of an air cooled quartz
 tube,  which  lowers  the  flue gas temperature from approximately 840°C to
 200°C.  Alkali metals which condense on partlculates In the flue gas are col-
 lected  In a  Balston filter  after exiting the quartz tube.  Flow 1s manually
 controlled by a needle  valve and measured by a wet test meter.

 Process Monitoring  and  Data Generation  System

     Data  characterizing the mlnlplant  operating conditions are recorded on
 5 multipoint recorders.  In addition, at one minute Intervals, the same out-
 put  1s  recorded by  a data logger system consisting of a D1g1trend 210 data
 logger  with  printer and a  Kennedy 1701  magnetic tape recorder.  Approximately
 100  pieces of data  are  logged with  one-half Involving temperature measurement
while  the  rest deal  with pressure, material  flow rate or flue gas composition,
 The  system 1s described 1n a previous report (1).

     Signals  from the data logger are scanned every minute and are stored on
magnetic  tape.  The magnetic tape, containing about 6000 Items of data per
 hour of run  time, 1s fed to a computer  which converts the logger output to
 flow rates,  pressures, etc. with the proper  dimensions.   The  data are then
averaged and  standard deviations are calculated  over preselected time Inter-
vals (usually 10 to  30 m1n.).   other quantities  are also calculated.   This
 Includes average bed temperature, based on four  thermocouple  readings covering
the 15-114 cm Interval above the flu1d1z1ng  grid,  superficial  gas velocity,
excess air, as well  as the Important gas concentrations.

Combustor Safety and Alarm System

     A  process alarm system was  designed to  warn of Impending  operational
problems.  Two general alarm categories exist.   The  first, dealing  with  less
critical situations, alerts the  operator of  the  problem  so that  appropriate
corrective action can be taken.   The second  class of more critical  alarms
results In the Immediate or time delayed shutdown of the complete system or
specific subsystems.  An alarm condition is  brought  to the attention  of  the


                                    40

-------
                                        FIGURE 111-16
                              SCHEMATIC OF ALKALI PROBE TRAIN*
        25°C
VENT
100 kPa
                  NEEDLE VALVE
        WET TEST
         METER
FLOW RATE
0.06
(2 SCFM)

  SECOND
  BUBBLER
                                  710kPa T>70°C T >210°C
                                         (160°F) (410°F)
                                           FIRST
                                          BUBBLER
AIRCOOLED
S.S. PROBE
                                                                        =-£
         T > 877°C
          (1610°F)
         ^—FLUE
           GAS
                                                                  /
                                                                 / QUARTZ
                                                                 *-  TUBE
                                                             BALSTON
                                                              FILTER
                                                         KNOCK OUT
       FRONT
       FILTER
* TYPICAL TEMPERATURES AND CONDITIONS DURING TEST 4 OF RUN 78

-------
 operators by a flashing light above the control  panel  accompanied  by a  high
 pitch sound.  The sensitivity of the Individual  alarms 1s controlled by pot-
 entiometers located beneath the control panel.  The system 1s  described 1n a
 previous report (1).

 Coal and Sorbent Properties

 Coal —
      Coals used 1n the mlnlplant variables study were  a  high volatile bitum-
 inous coal from the Consolidation Coal  Company's "Champion" preparation  plant
 1n Pennsylvania, an Illinois No. 6 seam coal  obtained  from Carter  Oil Company's
 Monterey No. 1  mine, and an Ohio coal  obtained from the  Valley Camp  mine.

      The Champion coal  was  partly classified  to  remove fines smaller than 50
 U.S. Mesh.  The Illinois coal  was screened to 6  x 40 U.S.  Mesh to  prevent
 plugging of the primary injector feed  vessel. Partlculate size distribution
 and composition data for the coals are shown  in  Figure 111-17 and  Table  III-l
 respectively.   The Valley Camp coal  was only  used on a very limited  basis
 during a coal  strike.   This coal  was purchased from Curtiss-Wright and no size
 analysis was obtained.   The coal  was prescreened 6  x 50  U.S. Mesh.

 Sorbent

 Sorbent—
      Grove limestone (BCR No.  1359)  and Pfizer dolomite  (BCR No. 1337) were
 the primary sorbents used in the  mlniplant variables study.  The composition
 of these stones  1s  given in Table III-2.   Most of the  runs were made with the
 stone screened  to  8 x 25 U.S.  Mesh to give the distribution of dolomite shown
 1n Figure 111-18.   The  limestone  was only  used on a  limited basis, no size
 distribution was obtained.

 Operating Procedures

      Prior  to initiating a  run, a detailed checkout  procedure Is followed to
 Insure  that  the system  is ready for operation.   This includes various equip-
ment checks, alarm  system checks, calibration of flue gas analyzers, activa-
tion of  process monitoring  and control   systems, and the turning on of all
cooling water systems.   All   runs are begun with an initial bed  of sorbent In
the combustor.  This consists of either a charge of sulfated limestone or
dolomite, or the bed material from the  previous run.  Fresh, uncaldned  lime-
stone has been used, but is  not preferred.

     The first operation of  start-up involves heating the gently fluidlzed
sorbent bed by burning natural gas in the burner  plenum followed  by injection
of kerosene into the bed.  Once the bed temperature is  ~650°C,  coal injection
is begun.  As soon as coal ignition is  confirmed  by a sharp temperature
increase, the other fuels are shut off.  Combustor temperature,  gas flow rate
and pressure are than rapidly Increased to their  operating values.   A more
complete procedure 1s contained In a previous  annual report (1).
                                      42

-------
                                                  FIGURE 111-17
                                       COAL  PARTICLE SIZE  DISTRIBUTION
CO
                                              PARTICLE SIZE, \im

-------
TABLE III-1.  COAL COMPOSITION
Run Number
60-65
66-70
71
72
73-75
76 and 77
78.1
78.2-78.10
79
80
81-88
89-96
97,98,99.1,
99.2,99.5-99.7
99.3 and 99.4
100
101 and 102
103 and 104
105
106
107-115
Coal Type
Champion
Illinois
Illinois
Ohio
Ohio
Ohio
Illinois
Illinois
Illinois
Illinois
Illinois
Champion
Champion
Illinois
Champion
Champion
Champion
Champion
Illinois
Champion
Moisture
2.19
1
1
1
1
1
3
12
12
12
12
2
3
13
3
2
2
2
11
2
.96
.60
.76
.82
.8
.16
.82
.5
.86
.9
.3
.12
.0
.12
.28
.94
.52
.36
.40
Ash
12.50
9.68
9.47
8.76
7.27
6.77
11.57
8.88
7.98
8.54
8.12
9.82
12.20
9.18
8.91
6.64
8.61
6.97
7.89
8.34
Ultimate Anlysls
Total
Carbon H S N Cl
71.35
67.35
68.73
73.41
74.33
75.13
65.58
57.76
60.63
67.67
58.81
71.91
68.04
60.34
70.44
74.35
72.43
79.41
67.39
--
4.74
5.40
5.41
5.23
5.35
5.52
5.63
5.02
5.48
5.81
4.78
5.24
4.76
4.70
4.88
5.23
5.20
5.87
5.10
--
1.40
4.00
4.00
2.84
2.63
2.47
4.17
3.52
3.33
3.53
3.37
1.65
1.92
3.4
1.87
1.38
1.61
1.64
3.52
1.84
1
1
1
1
1
1
2
2
1
1
1
1
1
1
1
1
1
1
1

.54 0.07
.24 —
.24 -
.35 -
.29 --
.16 -
.41 -
.26 -
.17 --
.29 --
.09 -
.43 -
.42 -
.02 -
.48 -
.58 —
.49 -
.47 -
.33 --
—
6.31
10.37
9.58
6.70
7.93
7.18
7.55
11.0
10.11
11.99
10.94
7.70
8.63
9.65
9.41
8.58
7.81
2.04
4.43
—
Heating
Value
Ib/BTU
12,514
12,575
12,645
13,171
13,351
13,711
12,042
10,988
11,043
11,528
10,892
12,809
12,581
10,912
12,984
13,728
13,151
13,612
11,969
12,874

-------
TABLE III-2.  SORBENT COMPOSITION
Quarry
Grove
Pfizer
Sorbent Type
Limestone
Dolomite
Weight Percent
CaO
94.64
55.78
MgO
1.01
37.04
S102
0.04
0.59
A1203
0.46
0.45
Fe203
0.117
0.23

-------
                                          FIGURE 111-18
CO
o

8
                                  DOLOMITE SIZE DISTRIBUTION

                                            Run 100

                             T—I—I	T
—•  »o  oo  NJ   a-

88888
o
g
o
                                                                                                 >

                                                                                                 Z
                                       PARTICLE  SIZE,

-------
EXPERIMENTAL RESULTS AND DISCUSSION

Combustor Operations

     During the month of August 1979,  experimental  miniplant studies  under  the
present contract were completed with run 115.   The  combustor was  shut down
after completing over 3700 hours of coal combustion operation.   Twelve runs
were made of over 100 hours duration;  five of these were over 200 hours dura-
tion.  The longest continuous run of the combustor  was 250 hours  long.

     The miniplant also made many short runs in rapid succession.  During the
ceramic bag filtration studies, fourteen 8-10 hour  runs were completed in
fifteen working days.  This was possible because of the rapid startup proce-
dure developed for the miniplant.  Coal was typically burned 1  to 1-1/2 hours
after the start of the first shift.

     The ability to turn the unit around rapidly was also demonstrated.  For
example, during the month of June 1979, the combustor was operated for nearly
400 hours in 3 run segments for two different test  series.  These runs were
coordinated with two other EPA contracts (Acurex and GCA Corp.).   Several runs
were also made in a "piggyback" fashion with up to  three other EPA contractors
participating.

     Although the operation of the miniplant became rather routine, some
operating problems still occurred.  The impact of these problems was usually
minimized by employing a number of operating procedures which were developed
over a period of time.

     Initially coal feeding was one of  the more serious operating problems.
After a number of equipment modifications described in an earlier report (3),
the system  performed satisfactorily.  Some problems still persisted and were
dealt with  by modified operating procedures.  The erosion of the coal feed  line
described in an earlier section of this report was such a problem.  The 316
stainless steel line eroded through after about 200 hours operation at a bend
in the line.  The impact of this problem was minimized by replacing all bends
in the coal feed system as soon as one  bend eroded.  This required only 10-15
minutes, during which time the combustor was fed liquid fuel.  Before long
runs, entirely new  bends were  installed, further reducing run interruptions.

     Coal feed plugging was only a minor problem during this period.  Plugging
of the coal feeder  would occur occasionally if coal was fed with a combination
of high moisture and high  fines content.   Feeding of coal with a moisture con-
tent as  high as 112 was possible if fines  smaller than 420 ym (40 U.S. Mesh)
were removed.  Drier coals with moisture contents of 1 to 3% were fed without
incident when  fines  less  than  300 ym  (50 U.S. Mesh) were removed.  The drier
coals were  easier to process and generated  less waste  as fines.

     The availability of  properly  processed coal was much more a problem than
coal feed  plugs.  No supplier  of dry  Illinois  No.  6  coal was  found for the
quantities  needed.   The  high moisture coal  was  used  because  of this.   Unpre-
pared  Illinois  coal  was  dried  by spreading it  in a  parking  lot and raking  it.


                                       47

-------
 Coal prepared "In this manner contained 11% moisture.  The supplier who  did this
 work elected to discontinue after processing only our most urgent supply needs.
 Regeneration runs were made with low sulfur Champion coal  because no  supplier
 of prepared Illinois could be found.

      Problems associated with the back flow of hot bed solids  through the coal
 transfer line into the coal  injector vessel  became severe during  the  testing
 of the granular bed filter.  During the filter blow back  cycle, a high  pressure
 pulse would upset the pressure differential  between the combustor and the
 injector vessel.  Hot solids would back flow to the injector vessel and  ignite
 the coal  in the vessel.   Several  runs were terminated for this reason.   The
 problem,  however, disappeared when testing of the granular bed filter stopped.

      Corrosion of the wet scrubber used to ensure compliance with EPA emission
 standards was a continual  problem.  It required frequent  repairs  during  turn-
 around times.  The severity of the problem was  lessened with the  injection of
 NH3 into  the scrubber to neutralize the acids  formed.

      Operations  during winter months  were  difficult.   Because of  the start/stop
 nature of many of the mini pi ant  runs,  frequent  draining and  blowing of water
 systems was  necessary.   Even with  these precautions, freeze  ups still  occurred.
 Productivity during the  winter was much lower for  these reasons.

      Small  problems with the main  fluidizing air compressor  frequently required
 attention during a  run.   One failure during a run  gave a chance to observe
 the effect of bed  slumping and hot start.  When the compressor failed, it
 stopped all  fluidizing air to  the  bed.   The temperature profiles during this
 period  are shown in Figure  111-19.  The  unit was restarted within 1/2  hour
 after  repairs  were  made  to the compressor.  The restart was completely suc-
 cessful with  no  bed agglomeration.

 Hot  Corrosion/Erosion Testing of Materials--
     During this reporting period, a total  of 1117 hours of combustor  operation
were devoted  to  "Hot  Corrosion/Erosion Testing of Materials for Application to
Advanced  Power Conversion Systems  Using Coal-Derived Fuels."  This work  was
 sponsored  by  DOE, under  a cooperative agreement with EPA,  to test the  effect
of exposing gas  turbine  blade and  boiler tube materials to real  PFB conditions.
 The gas turbine  specimen test  section was designed and built by General
Electric.  The air  cooled boiler tube probes were designed and built by
Westinghouse.  The  evaluation of results from these two systems is the res-
ponsibility of General Electric and Westinghouse.  The combustor flue  gas
system was modified as shown 1n Figure 111-20 to permit good pressure  control
in the  presence of  the turbine test section.  Most of the  flue gas was expanded
through the turbine test section; 10% was diverted for pressure control.

     A picture of the blade specimens after 1000 hours of  exposure is  shown in
Figure  111-21 .  As can be seen 1n the photograph, little visible  erosion is
present.  Further metallographic examination is being done by General  Electric.
Results will  be reported by them 1n the future.
                                      48

-------
                                  FIGURE  111-19
      SELECTED COMBUSTOR TEMPERATURES  DURING CRASHDOWN AND RESTART
                RESTART
• Primary Cyclone  Discharge
A 3rd Cyclone Inlet
• Port #9 (in-bed)
Numbers refer to height above
       fluidizing grid, (cm)
1415
1400
1345
                                                                        1330
                                                                                         o
                                                                                        o
                                                    450
                                                   - 350
                                  CLOCK TIME

-------
                                              FIGURE 111-20

                                       MINIPLANT FLOW SCHEMATIC
                              HOT CORROSION/EROSION TEST CONFIGURATION
en
o
p.


c
0
M
B
S
T
O
R









1st





\ /
\
Cyclone


x
S





/










/
N/G-<


i
2ND
' CYCLONE

























h







VG-^


1

ALKALI
HTHP PR°BE
PARTICULATE TURBINE TEST
PROBE SECTION



- OFF GAS
T COOLER
1
N/G


I 1 I » t_ -J 
-------
           FIGURE 111-21
      GE TURBINE BLADE SPECIMENS
  L
1000  HOURS
         S T  CASCADE
I	SECOND  CASCADE
    FOURTH  CASCADE
                         1 C"
               51

-------
      Twenty-eight of the air cooled boiler tube probes designed  by  Westinghouse
 were exposed for up to 1117 hours.  These were controlled  at  specified tem-
 peratures, and located both in and above the fluidized bed.   Each probe con-
 tained two test materials.  A picture of two probes  is shown  in  Figure 111-22.
 Results of the metallographic examination of the specimens will  be  reported
 separately by Westinghouse in the future.

      Exxon will report run conditions including particulate loadings and
 other exposure conditions in a report to DOE later in  1979.

 Effect  of Natural  Gas  Injection--
      Natural  gas  injection is used in the miniplant  to maintain  temperatures
 above 840°C (1550°F) for  the DOE  corrosion tests  as  well as to shorten the
 time necessary for thermal  equilibrium in downstream components.  Without
 natural  gas injection, the temperatures  at the  tertiary cyclone would only
 reach their steady operating values  after about 12 hours of operation.  With
 natural  gas injection  this is reduced to 1-1/2  to 2  hours.  For these reasons,
 natural  gas injection  is  used routinely  during  miniplant operations when
 tests of components  outside the combustor are run.

      The effects of  natural  gas injection on flue gas  composition and parti-
 culate  characteristics  have been  investigated.  One  concern is that instan-
 taneous  combustion of  the gas  would  cause hot spots, which could increase
 formation of  NOX,  or alter  particulate characteristics.  Temperature measure-
 ments have shown that  the gas  does not begin to burn until  it is several  cen-
 timeters  downstream  of  the  injection  point.  Combustion then occurs  uniformly
 over  the next 2 meters  before  it  is  complete as indicated by a temperature
 decline.   The temperature profile  for the  injection of natural gas can be
 seen  in  Figure  111-23.  During this  study  the amount of natural  gas  normally
 distributed among 4  probes  was injected  in only the first probe.  During
 normal operations, the  temperature rise  is no more than 20°C,  however, the
 shape of the  profile for  each  probe  is similar.

      The  only effect of natural gas injection on flue gas composition that has
 been measured besides a 0.5 to 1.0% reduction in flue gas oxygen is  a drastic
 reduction  in  CO.  During  run 81, CO emissions fell from 525 ppm  1-1/2 hours
 after coal  injection was  begun to 30 ppm after natural  gas  injection into  the
 flue gas  lines was started.  Reduction in CO emissions  was  also  observed  in
 runs 79 and 80 with  natural gas injection.  CO emissions observed in the  past
 for miniplant runs without natural gas injection were in the range of 100  to
 200 ppm.   For the runs with natural gas injection, the  residence time of  the
 flue gas  in the piping at the temperature range of 815  to 940°C  was  about
 twice as long as for runs without  natural gas injection. Thus,  natural gas
 injection could provide increased  opportunity for burnout of CO.

     The effect of natural gas injection on flue gas  particulates  is also  of
concern, especially  if the gas is  responsible for the high  collection
efficiencies measured with the tertiary cyclone.  Studies in the  miniplant
 have shown that cyclone collection efficiencies  are not significantly affected
by the injection of natural gas.   The slight effect that is measured is usually
explained by the effect of temperature on the cyclone inlet velocity.
                                      52

-------
                 FIGURE 111-22
      WESTINGHOUSE BOILER TUBE PROBES
            After 1117 Hours Exposure
                 (850°C In-Bed)
                                                          Top
 ALLOY  --
Haynes  188
Hasten oy X
                                                         Bottom
                                                         Top
ALLOY —
 Hasten oy X
Haynes  188
                                                          Bottom
                       53

-------
                                  FIGURE 111-23

                  NATURAL GAS INJECTION - TEMPERATURE PROFILES
                           T
T
    900
    850-
U
o
                                                  10
                    DISTANCE DOWNSTREAM OF 2ND CYCLONE (M)
                                                                     76.7
                                                                   METHANE
                                                                    FLOW
                                                                 (Ndm3/MIN)
                     15

-------
$02 Emission Control

     As a result In a change 1n the  New Source  Performance Standards for coal
fired utility boilers, the allowable S02 emission  level was decreased from
1.2 Ib/MBTU to a sliding scale which, for higher sulfur coals, requires 90%
SOz retention.  A series of runs were made to demonstrate that PFBC could
attain the 90% SOg retention level using dolomite  or limestone sorbents.  Ohio
and Illinois coals, containing 2.5 and 4% sulfur respectively were tested.
Ca/S molar ratio was  the primary variable in the study.  With dolomite sorbent,
the Ca/S ratio was varied from 1.5 to 2.1. With limestone, it was varied from
3.7 to 7.6,  Superficial gas phase residence time  in these tests was varied
slightly from 1.6 to  2.2 s.  Other conditions are  summarized in Table III-3,
which also tabulates  the S02 retention results.  As seen, 90% SO? retention
can easily be reached with either sorbent. Retention  levels as high as 99+%
were measured.  As expected, dolomite is more active.  A Ca/S ratio of 1.5
will assure 90% S02 retention at a gas residence time  of about 2s while lime-
stone use will require a Ca/S ratio  of between  3.5 and 4.0.

     In subsequent runs, additional  data were obtained in the high S02 reten-
tion area.  All SO? retention data were then plotted against Ca/S ratio for
both dolomite and limestone sorbents and analyzed.  In the plots, the Ca/S
ratio set on the coal/sorbent blender was used  as  the  correlating parameter.
In earlier runs, a Ca/S ratio calculated by mass balance was used because of
blender speed control problems.  Since the problems  were solved by installing
an electronic speed control system on the blender, it  was decided to use  the
"set" Ca/S ratio since it appeared to be more  reproducible,  possibly because
it did not require reliance on a series of chemical  analyses of the solid and
gaseous products.

Results with Dolomite Sorbent—
     Figure 111-24 gives S02 retention results  using Pfizer  dolomite sorbent,
Champion,  Illinois and Ohio coals,  at gas residence times  from 1.5 to  2.5s.
Retention  results shown in the figure were corrected to  a  residence time  of  2s
using the  simple first order rate expression  described in  the  previous  report
(2).  Figure 111-25 gives results of runs made with a  residence  time of 2,5
to 3.5s, corrected to a residence time of 3s.   Since a significant amount of
data scatter occurs on each plot, values from the  smooth curves  were used to
construct  a curve relating the first order rate constant for the  S02  sorbent
reaction,  k, to the sorbent utilization.  A set of values  for  k  was  calculated
from the first order rate equation by using a range of retentions at constant
residence  times.  The sorbent utilization was obtained from the  smooth curves
by using the expression for a sulfur balance:

                    % Ca Utilization * S02 retention  (%)
                                                ratio
      If  the  S0£ reaction rate is indeed first order, then the values from the
 curves in  Figures  111-24 and 111-25 would overlay on a k vs. utilization curve,
 all  other  variables  being equal.  However, when this was done, two k vs. uti-
 lization curves were obtained, one for each residence time.  The reason for
 this was that  the  3s residence time data did not show a significant improve-
 ment in  S02  retention over the 2s data, as would be predicted by the first
                                       55

-------
                             TABLE  I I 1-3.  RUN SUMMARY 90$ S02 RETENTION STUDY



     Nominal Operating Conditions  Run  68  Run 69  Run 70  Run 71  Run 72  Run 73  Run 74.1   Run 74.2  Run 75

     Run Length (hrs)              10       11      11      13      13      11.5       4.5      5.75     9
     Pressure (kPa)                930     930     930     930     930     930       930      930      930
     Temperature (°C)              946     947     936     933     943     939       933      937      936
     Superficial Velocity (m/s)      1.7     1.8      1.9     1.7     1.8     1.8       1.7      1.7      1.7
     Expanded Bed Height (m)         2.8     3.0      3.0     3.8     3.7     3.8       3.7      3.7      3.8
     Gas Residence Time (s)           1.6     1.7      1.6     2.2     2.1     2.1       2.2      2.2      2.2
     Excess A1r («)                17       14      23      13      11      10        16.5     18.9     17
     Coal  Feed Rate (kg/hr)         127     132     127     122     116     117       113      112      115
     Ca/S Molar Feed Ratio           2.0     1.5      1.5     1.5     1.8     2.1       7.6      4.8      3.6
     Sorbent                         PD      PD       PD       PD      PD      PD        6L       GL       GL
     Coal                             IL      IL       IL       IL      OH      OH        OH       OH       OH
<*    Emissions
     S02 (ppm)                      50     197      94       60       60       4         1       30      170
     NO' (ppm)                      79      56      81       40       30       50        80       60       50
     02  (%)                         3.2     2.6     4.0     2.4     2.1     2.0       3.3      3.7      3.4
     C0? (%)                        16      16      17       17       22       18        19.5     19       15.2
     CO2 (ppm)*                      ............        ......
     S02 Retention                  98.5    94.2    96.9    98,1     99.0     99.7      99.9     98.2     90.9
     Notes:  PD  *  Pfizer 1337 Dolomite
             GL  =  Grove 1359 Limestone
             IL  *  Illinois Coal  - 4% S
             OH  *  Ohio Coal  - 2.5% S
             *  CO Analyzer Malfunctioned

-------
                              FIGURE 111-24

          SO2 RETENTION ADJUSTED TO 2 S. GAS RESIDENCE TIME (tg).
         DOLOMITE RUNS HAVING  ACTUAL tg BETWEEN 1.5 AND 2.5 S.
  100
   90-
   80-
   70-
z
o
*•  60
 CM
   50-
   40-
   30
     0
                            Ca/S MOLAR RATIO
                               57

-------
   10
   90
   80
Z
o
   70
 CN
   60
                              FIGURE 111-25


          SO2 RETENTION ADJUSTED TO 3 S. GAS RESIDENCE TIME (tg).

          DOLOMITE RUNS HAVING ACTUAL tg  BETWEEN 2.5 AND 3.5  S.
                                              T
                                                     T
   50
   40
   30
                        t
0
                           Ca/S MOLAR RATIO
                                 58

-------
order expression.  A plot of the  mean  k vs.  utilization  curve  derived  from the
two retention curves is presented in  Figure  111-26.   This  curve  is a compromise
between the upper and lower limits formed by the  2s  and  3s residence time data.

     At present, there is no explanation for the  lack of uniformity in the two
groups of data, but one obvious observation  would be that  the  reaction does not
strictly comply with first order  kinetics.  This  may be  a  problem inherent in
large scale FBC operations where  additional  operating parameters, such as
solids recycle and rejection, compound the difficulty of obtaining representa-
tive samples necessary to characterize the system.

     To study the effect of residence time on S02 retention, the k vs.  utiliza-
tion curve in Figure II1-26 was  used to construct retention curves  for four
residence times.  2s and 3s retention curves were calculated and plotted with
the miniplant data curve in Figure 111-27.  As seen, the two calculated surves
straddle the data curve reasonably well, indicating that effects of gas resi-
dence time on S02 retention become more difficult to discern as residence time
Increases.  This can be seen more clearly in Figure 111-28, where retention
curves for all  four residence times were drawn.  Here it becomes obvious that
below 2s, residence time has a significant effect on SOg retention,  while
above 3s this effect would necessarily become less substantial, particularly
at Ca/S ratios  greater than 1.

     Additional  work  is obviously needed  before  the effect of gas phase resi-
dence time as well  as  other  kinetic parameters can  be accurately predicted.
Figure  111-28 can  be  used  as  an  approximate  guide until a better understanding
is developed.

Results with  Limestone Sorbent--
      Figure  111-29 gives  S02  retention  results obtained with  limestone sorbent
at temperatures high  enough  to calcine  the  limestone  substantially.  As seen
in the  previous figures with  dolomite,  the  data  scatter is  significant.  With
limestone,  the  degree of  calcination  and porosity of the  calcined stone are
other  parameters in addition to  the kinetic  effects discussed above.  Again,
more work  is  needed to improve understanding of  the desulfurization system.

      It should  also be mentioned that limestone  not only  is less active than
dolomite  under  normal  PFB combustor operating conditions, but  becomes com-
pletely inactive at temperatures at or below 760°C  (2).   This must be con-
sidered if limestone is  to be used and 1f low temperatures  are  expected, for
example,  to  aid in decreasing the steam output from the combustor.

S02  Dynamic  Response

      The  new S02 emission standards call for 90% retention of the S02  from new
coal  fired utility boilers.   This retention must be averaged  over 30  days.   It
 is expected that utilities would operate at slightly higher retentions  to avoid
exceeding  the new limit  and avoid costly violations.  The incentive therefore
exists  to  develop a system to control S02 retention close to  the desired  level
without large safety margins. With  this goal  in mind,  General  Electric has  set
out, under contract to the EPA,  to develop  a control  system that would adjust
 sorbent feed rates to assure a specified S02 retention  despite variations in

                                        59

-------
                             FIGURE 111-26
Z
o
u
as
Q
Of

O
          FIRST ORDER RATE CONSTANT VS.  CALCIUM UTILIZATION

                       MINIPLANT DOLOMITE RUNS

                          (APPROXIMATE ONLY)
     3.
     3.
     2.5
     2.0
     1.5-
     1.0
      .5
             T	1	1	Tl	1	1	1	1	r
                 _L
_L
J.
J_
       0    10    20   30    40   50    60    70


                      CALCIUM UTILIZATION (%)
_L
                                                       J.
                           80   90   100
                             60

-------
                               FIGURE 111-27

           MEASURED AND CALCULATED SC>2 RETENTION VS. Ca/S FOR
                        MINIPLANT DOLOMITE RUNS
   100
    90
    80
g
t—
LLJ
    70
 cs
    60
    50
    40
    30
               CALC.  FOR tg = 3 S.
                                   MINIPLANT DATA
                                   FOR tg = 2 to 3 S.
                                            CALC. FOR tg = 2 S.
               0.5
       1.5          2

Ca/S MOLAR RATIO
2.5
                                   61

-------
           FIGURE 111-28
CALCULATED SO2 RETENTION VS. Ca/S
    MINIPLANT DOLOMITE RUNS
     1.0         1.5        2.0
       Ca/S MOLAR RATIO
2.5
3.0
              62

-------
  100
   90-
   80-
 , 70
Z
o
                               FIGURE 111-29

                    SO2 RETENTION VS. Ca/S RATIO FOR
              LIMESTONE NO. 1359 AT CALCINING CONDITIONS
tu
Of
 CM
8
60-
   50
   40
     0
                          345
                          Ca/S MOLAR RATIO
8
                                    63

-------
 the  sulfur  content of the coal, and In the reactivity of the sorbent.  Tests
 were carried  out on  the miniplant  in cooperation with GE, to determine the
 emissions response to  a step change in either coal sulfur or dolomite feed
 rate,  with  other variables held constant.

 Experimental  Equipment and Procedures--
     At  GE's  request,  the nvfniplant was run with no primary cyclone recycle
 during the  S0£ response tests.  The combustor was allowed to reach a steady
 state  for 4 to 5 hours before a change of condition was made.  During this
 period,  the old, continuous, primary feed system was used.  Once steady state
 emissions were achieved, the change was made to the auxiliary feed system
 containing  a  different coal/sorbent blend.  The auxiliary feeder was used for
 8  hours, the  approximate capacity  of the vessel.  After 8 hours on the
 auxiliary feeder, the  change was made back to the primary feeder and the
 original coal/sorbent  blend.  The  combustor was then run for 18 to 20 hours
 with the primary feeder.  The 18 hours were required to assure two changes of
 the  bed  material Inventory.

     The procedure for changing from one feeder to another consisted of:

     1.  Calibrating the analyzers 30 minutes before the change.
     2.  Closing the valve at the  bottom of the active feeder.
     3.  Allowing the  coal transport Hne to clear of sol Ids (~20 sec.)
     4.  Changing over AP taps, transport air and electronic controls
         to the other  vessel.
     5.  Opening the valve at the  bottom of the other vessel.
     6.  Confirming coal feed by non-zero SOg emissions and combustor
         temperature increase.

 The  time required for  the change of feed vessels was 40 to 80 sec, depending
 on crew  experience and operating conditions.

 Results and Discussions —
     Shakedown—Shakedown of the auxiliary feed system went smoothly during
 runs 97 and 98.  Several  modifications  made after run 97 allowed  the time
 required for the change to be reduced  from 4  minutes to 40 sec during run 98.
 The resultant combustor temperature drop was  reduced from  260 to  70°C.   Table
 III-4 lists  the results of step changes in the coal  source during shakedown
as well as  1n the actual  runs.   The time listed 1s  the time required from the
confirmation of coal  feed by non-zero  SOg  emissions  to the first  time the S02
emission reached the  equilibrium value.   A large, short S02  "spike"  was
 Ignored in  this summary since it was due to the rapid  combustor temperature
drop and subsequent coal  feed  rate  increase caused  by  the  action  of  the  coal
feed controllers.   These  temperature fluctuations may  be unique to the mini-
plant and may not appear  in  a  commercial  installation.

     Change  in Coal  Sulfur—The first  actual  response test conducted on  the
minlplant (run 99)  was a  step change in coal  sulfur  content at a  constant
dolomite mass feed rate.   The minlplant was started  with Champion coal and
dolomite (1.9% sulfur, Ca/S  = 1.40). After 4 hours,  the combustor was switched
to the auxiliary feeder which  contained Illinois No.  6 coal  and dolomite (3.4%
sulfur, Ca/S » 0.76).  Then, after  8 hours, the combustor  was switched back to

                                      64

-------
                              TABLE III-4.
S02 DYNAMIC RESPONSE SUMMARY
in


Run #
97(D
970)
98(D
98(D
98^
98^
99
99
100(3>
100(3>
Type
of
Change
% S
%-S
% S
% S
% S
% S
% S
% S
ca/s
Ca/S

/0\
From(2)
Champion
Illinois
Champion
Illinois
Champion
Illinois
Champion
Illinois
Ca/S = 1 .43
Ca/S =0.38


To
Illinois
Champion
Illinois
Champion
Illinois
Champion
Illinois
Champion
Ca/S =0.38
Ca/S = 1 .43
S02
Before
(ppm)
90
675
225
675
225
675
225
675
375
500

S02 at Steady
State (ppm)
675
225
675
225
675
225
675
225
500
300
Time from
Feed Change to
Steady State (min)
6-3/4
7
6
4
5
7-1/2
7-1/2
8-1/4
110
200
     (1)  Shakedown Runs
     (2)  Champion Coal       - 1.92% Sulfur (Ca/S = 1.40 during tests where % S changed)
          Illinois No. 6 coal  - 3.4% Sulfur (Ca/S = 0.76 during tests where % S changed)
     (3)  Champion Coal Used
          Pfizer dolomite used in all tests

-------
 the Champion coal, dolomite blend for  20 hours.   The  S02 emission, averaged
 over 10 minute intervals,  over the entire run  is  shown  in  Figure  111-30.
 Figures 111-31 and 111-32  show the continuous  S0£ analyzer output immediately
 before and after the two changes  in coal  sulfur.   The response times of 7.5
 and 8.2 minutes were determined from these figures.   During the change to the
 higher sulfur coal,  the dolomite  mass  feed rate actually increased 19% although
 it was intended to hold it constant.  The effect  of a 19%  increase in dolomite
 mass  feed rate was superimposed on the 104% increase  in coal sulfur mass feed
 rate  occurring because  of  the  change.   The change  back to the lower sulfur coal
 also  had a  19% decrease in dolomite mass  feed  rate.   The unintended change in
 the dolomite  mass  feed  rate was  caused by slight  changes in the sulfur content
 and heating value  of the coal  actually used, versus prior analyses of other
 batches  used  to set  conditions.

      During the combustion of  Champion coal, S02  emissions were close to the
 predicted level  of 170  ppm.  This  prediction comes from Figure 111-25 of the
 previous  section.  The  predicted  SQ2 emission  for  the Illinois No. 6 coal  is
 1080  ppm, much higher than the  725  ppm measured.

      Change in Sorbent  Feed Rate—During  run 100,  the S02 emission response to
 a  setp change  in Ca/S from 1.43 to  0.38 using  Champion coal and dolomite was
 investigated.   The combustor was  started  with  the  higher Ca/S which was then
 rapidly changed to the  lower Ca/S  by switching to  the auxiliary feed system.
 The first attempt  to switch  to  the  auxiliary feed  vessel failed due to a plug
 in  the transport line.   The  plug  was cleared while the combustor was maintained
 on  liquid fuel.   The low Ca/S coal  was  fed  from the auxiliary vessel for 20
 minutes to  determine the likelihood of another plug.  The combustor was switched
 back  to the high Ca/S blend  until  the  SOg  emission stabilized at the preswitch
 level  (about 2  hours).   The  combustor  was  again switched to the low Ca/S blend
 in  the auxiliary feeder.   After 8  hours of operation on the low Ca/S blend,
 the change  was  made  back to  the higher  Ca/S blend.  During this final run seg-
 ment, coal  was  fed for  19  hours to  assure  that steady state had been attained.
 From  Figure 111-33 the  response time for the step  from Ca/S = 1.43 to 0.38 is
 shown to  be 110  minutes.   The reverse  step required 300 minutes to achieve
 steady state.   It  is unexplained why the original   steady emission, at startup
 (about 375  ppm), was not the same as the  final  segment emission (300 ppm).  At
 no time during  the run was  the average  SOg emission as close to the predicted
emission as expected.   The  predicted emission for Ca/S = 1.40 Is  170 ppm.   The
actual average emission during the  19 hour final  portion of the run was 300
 ppm.  During the low Ca/S  portion (mid  portion) of the run, the actual  average
emission was only 500 ppm  compared  to 750 ppm predicted.  This  may be due  to
 the fact  that the low Ca/S  operating region has not been well  defined experi-
mentally.  The time required during the second change  (to the  higher Ca/S)  to
reach the "startup emission" of 375 ppm was approximately 120 minutes.   This
suggests that there is a reversibility  in the response  times  for these
changes.  The SOg emission  just prior to and just after  the coal  sorbent blend
changes can be  seen in  Figures  111-34 and 111-35.   The immediate  effect of
both changes ts  not noticeable  since the response times  are fairly long.

 Conclusions—
     For the typical  90% S02 retention  levels expected to be used in most  pres-
surized fluidlzed bed combustors, the response to  100% increase in coal  sulfur

                                       66

-------
                                                    FIGURE 111-30
CT»
              Q_
              O.
              z
              O
               CN
1100

1000

 900

 800

 700

 600

 500

 400

 300

 200

 100
                                   CHANGE  IN COAL SULFUR/SO2 EMISSIONS  (RUN 99)
PREDICTED
EMISSION
170  PPM
                        COAL FEED
                          UPSETS
                       CHAMPION
                          COAL
                       Ca/S=1.40
               I
                                     PREDICTED EMISSION
                                          1080 PPM
                                                                                         i   r
                                                             8.2 MINUTE
                                                              RESPONSE
7.5 MINUTE
 RESPONSE
             ILLINOIS NO. 6 COAL
                 Ca/S = 0.76    I
               i  i   i  I  i   i  i	i
                                                         I
                    CHAMPION COAL Ca/S= 1.40
                     i  i  i   I  i  i   i  i   I  i   i
                     0
                           10          15          20
                              TIME INTO RUN  (HRS)
                                                         25
                                                    30

-------
                                                    FIGURE 111-31
en
CD
                      +22.5
                      -22.5
                                        INSTANTANEOUS SO2 RESPONSE (RUN 99)

                                        (CHAMPION TO  ILLINOIS NO. 6 COAL)
                                                           AFTER CHANGE
                                                                    ILLINOIS NO. 6 COAL


                                                                    Ca/S = 0.76
                                                           COAL SULFUR CHANGE



                                                           BEFORE CHANGE






                                                                    CHAMPION COAL


                                                                    Ca/S = 1.40
                                        I
                                 300
600
900
1200     1500
1800     2100    2400
                                                 SO2 EMISSIONS (PPM)

-------
                                                    FIGURE 111-32
IO
                       +22.
                       +15.0-
LU
O
z

I
u


o
u.
    -15.0
-22.5
    0
                                       INSTANTANEOUS SO2 RESPONSE  (RUN 99)
                                        (ILLINOIS NO. 6 TO CHAMPION COAL)
                                                               AFTER CHANGE
                                                                       CHAMPION COAL
                                                        8.2 MINUTES
                                                                       Ca/S = 1.40
                                                               PRIOR TO CHANGE
                                                                       ILLINOIS NO. 6 COAL

                                                                       Ca/S = 0.76
600
                                                  900      1200     1500

                                                  S02  EMISSIONS (PPM)
                                                       1800
2100
2400

-------
                                    FIGURE 111-33
                               RUN TOO SO2 EMISSIONS
                                (10 MINUTE AVERAGES)
    600
                     i  i  i   i  |   I  i  i   i  |   i
                     PREDICTED EMISSION 750 PPM
                                                          i  i
    500-
I   400
Q_
to
z

§   300
—
LLJ
200
     100
         ATTEMPTED
          CHANGE
          Ca/S=1.43

         — PREDICTED
            EMISSION
            170 PPM
            GAS SAMPLING
         SYSTEM INOPERATIVE
             Ca/S =1.43

PREDICTED EMISSION 170 PPM~
                                I
                               10           15

                                TIME INTO  RUN (MRS)
            20
                                                               25
30

-------
                                   FIGURE 111-34
    +22.5
    +15.0-
o
z
<

u


o
Qi
   -15.0
   -22.5
                      RUN 100  INSTANTANEOUS SO2 RESPONSE

                                (CHAMPION COAL)
                      600
                                             AFTER CHANGE
                                                Ca/S = 0.38
                 CHANGE IN Ca/S
                                             PRIOR TO CHANGE
                  Ca/S= 1.43
 900     1200    1500


SO2 EMISSIONS (PPM)
1800     2100     2400

-------
                                  FIGURE 111-35
    +22.5
    +15.0-
     +7.5-
LLJ
0
z
u
o
exi
Ll_

UJ

5
     -7.5-
    -15.0-
                      RUN 100 INSTANTANEOUS  SO2 RESPONSE

                                (CHAMPION COAL)
                                              CHANGE  IN Ca/S
                                                PRIOR TO  CHANGE
                       600
                               900     1200     1500


                               SO2 EMISSIONS (ppm)
1800     2100     2400

-------
was fairly rapid (~8 minutes).   The  response to  a  73% change  in  dolomite  feed
rate required 2 to 5 hours  to  reach  a  stable emission.   This  would  indicate
that 1t may not be possible to  prevent sudden changes in SO?  emissions which
respond rapidly to variations  in coal  sulfur content with changes in  sorbent
feed rate which require much longer  response times.

     The fact that most emissions were far from  the  predicted emission  levels,
may suggest that there is an initial short response  time coupled with a very
much longer response time.   Chemical analyses of the sorbent, bed extract
material, and cyclone dumps have been sent to GE.   They are incorporating these
data into their S02 emission response model and  will issue a  separate report
with their conclusions.

NQX Emissions

     NOX emissions measured in all minlplant runs are plotted in Figure 111-36
as a function of percent excess air.  Data follow the same trend Tine shown
in previous reports  (1).  NOX emissions are well below the recently modified
New Source Performance Standard of  0.6 Ib/MBTU for coal fired utility boilers.
Most PFBC design studies have indicated the excess air level  will most  likely
be around 20%.  At this excess air  level, NOX emissions are expected to be
0.2 4^0.1 Ib/MBTU.   Emissions are expected to be below 0.4 Ib/MBTU, even  at
excess  air levels in the range of 60 to 100%.

     An evaluation of methods which could decrease NOX emissions below those
in Figure 111-36 was carried out in the bench PFB combustion unit.  The
results of this study are given in  Section VII.

Other Gaseous Emissions

     503 emissions in the combustor flue gas were measured for most runs.  For
runs up to run  75, the EPA Method 8 method was used.   For subsequent runs, the
controlled condensation method described in Section  III was used.  Results
for all measurements is given in Appendix Table G.   Two measurements out of a
total of 40 appeared to be in error and were discarded.  These were readings
of 213  pprn from run  71 (Method 8),  and 73 from run  100  (controlled condensa-
tion).  Excluding these two measurements, 16 Method  8 measurements averaged
12 + 12 ppm, 22 controlled condensation measurements averaged 6 +_ 9 ppm.   All
meaFurements averaged 9 £  11 ppm.   In both cases, the  range of measurements
was 0 to 30 ppm.  This is  the same  range reported previously when only the
Method  8 technique was used (1).  Aside from the fact  that the controlled
condensation method  gave an average reading exactly  one  half of the Method 8
average, no positive conclusions could be drawn from the results.  Even  the
difference between the average measurements  is not  statistically significant
due to  the large  degree of uncertainty.   In  both cases,  the degree of uncer-
tainty  was approximately +, 100% of  the average values.   Large variations were
also measured within a single run.  Therefore, the  variation appears to  be
random  and no conclusions  could  be  drawn regarding  the  cause of SOs  formation
or  the  factors  affecting the degree of $03  formation.

      CO emissions were again in  the range  of 50 to  200 ppm as reported pre-
viously (1).  A few  measurements were 1n the 300 to  500 ppm  range, but could
have  been  caused  by  analyzer problems.

                                        73

-------
          FIGURE 111-36




CORRELATION OF NO  EMISSIONS

-------
     Reduced sulfur compounds,  H2S,  COS  and  C$2  1n  the  flue  gas were measured
by gas chromatography (GC)  a number  of times and were consistently  less than
the detectability limit of  1 ppm.  Hydrocarbons  were also measured  by GC.
Methane averaged 7 +_ 5 ppmt ethane 4 + 4 ppm, 03 through Cc  hydrocarbons were
generally below the detectabillty  limit  of 1 ppm.  GC results  are given in
Appendix Table F.

     Emission of sodium, potassium,  chlorine and vanadium  in the flue gas was
also analyzed using the system  described previously.   In this  system, a sample
of high temperature, pressure flue gas is extracted from the ducting following
the third stage cyclone, partly filtered, cooled, filtered  again, cooled  fur-
ther and bubbled through an impinger train  (see Figure 111-37).   Figure  111-37
also shows the different locations where the four elements  are condensed,
deposited or absorbed and subsequently analyzed.  The  filters are  extracted
with boiling water and the extract analyzed.  Unused filters are  also  extracted
to give a blank for the filter material  itself.  The quartz tube  is washed  and
the wash solution analyzed.  Table III-5 gives the distribution of  the  four
elements and particulates found in the six locations in the train.   Sodium
totalled 2 to 3 wppm in the inlet flue gas, potassium 0.3  to 0.5  wppm.
Generally, over 90% of these elements were  found on the front and final  fil-
ters (sampling locations 1  and 3 in Figure  111-37).  The  front filter was not
retaining all the particulates impinging on it, as seen by the large fraction
(60 to 70%) of particulates captured by the final filter  in samples 78 and
79-1.  The front filter media was removed for sample 79-2.  However, the
fraction of total alkali captured on the final  filter was  greater than the
fraction of particulates captured on the final  filter.  Therefore,  sodium and
potassium compounds were condensing in the  quartz tube on the surface of
particulates which were removed by the final filter.  Very little alkali
compounds were condensing and collecting on the  surface of the quartz tube
itself.

     The concentration of chlorides measured in  the flue gas was about 50
wppm.  This represents about 60% of the total Cl entering with the coal  and
dolomite feeds.  As seen in Table III-5 chlorides were not detected on the
particulates  but only  in the knockout condensate.  Therefore, chlorides were
probably all  present as HC1 at the  low  temperature conditions in the sampling
probe.  Sodium and  potassium were then  probably present as  the sulfates in the
sampling system, although  they probably occurred as chlorides in the high
temperature  flue gas prior  to sampling.

      No vanadium was detected in  any of  the samples.

Particulate  Emissions

      Particulate  emissions  from the miniplant depend heavily  upon  the parti-
culate cleanup  system  used.  The  emission when  the flue gas  is cleaned by  a
granular  bed  filter is very different from  that when it is  cleaned with cyc-
lones.  During  this reporting  period  almost all  runs were made with 3 stages
of cyclones  in  series.  For this  reason,  this section  will  deal mostly with
emissions measured in  the  flue gas  cleaned  in  this manner.   Emissions using
other  control  technology are contained  in Section  IV.


                                      75

-------
                                       FIGURE 111-37

                            SCHEMATIC OF THE ALKALI PROBE TRAIN
CT»
           VENT
           1  ATM
                     T
   WET TEST
    METER
           2ND
         IMPINGER


1
1

I
ER
NEEDLE
VALVE
1

1--,
1 1
3\ -r
T\ 1
1 1ST '
IMPINGER
(T)
>240°C
//
¥ n
* 1
, \
\ BALST
\ FILT
KNOCK O
CON DEN SA
©<~:
            o
ANALYZED SAMPLE LOCATIONS
                                                                        AIR COOLED
                                                                         S.S.  PROBE
                                                                                      843°C
                                                                                      ATM
                                                              FILTER    QUARTZ
                                                                        TUBE

-------
          TABLE II1-5.  EMISSION OF SODIUM, POTASSIUM, CHLORINE, VANADIUM IN FLUE GAS
Run/Sample No.
Element
Concentration (wppm)

Distribution (wt. %)
Location
1  - Front Filter
2 - Quartz Tube
3 - Final Filter
4 - Condensate
5 - 1st Impinger
6 - 2nd Impinger
         78/4
   Na    K   V
  2.06  0.54 0
  7
< 1
 89
  3
< 1
< 1
37
 3
59
 1
 0
 0
79/1
irt.
• ••
39
0
61
0
0
0
Na
1.84
1
2
96
1
0
0
K
0.28
5
14
80
1
0
0
V Cl
0 47.4
- 0
- 0
- 0
- 100
- 0
- 0
Part.
™ ™
31
0
69
0
0
0
Na
3.23
< 1
2
95
2
0
0
79/2
K
0.38
6
8
83
3
0
0
y___ci 	
0 53.8
- 0
- 0
- 0
- 100
- 0
- 0
Part
~ *
3
0
97
0
0
0

-------
      Particulate emissions during this  reporting  period  generally  ranged  from
 0.03 to 0.15 g/Nm3 (0.013 to 0.065 gr/SCF)  with 3 stages of cyclone cleanup.
 These gas particulate concentrations  were measured with  Balston  total  filters
 in either the Balston filter sampling system or the HTHP sampling  system.
 The size distribution of particulates in  the filter cake was obtained  with the
 Coulter Counter utilizing techniques  described 1n Appendix  D.  The size dis-
 tributions of most samples are recorded 1n  Appendix M.   Table  III-6 gives par-
 ticle  size distributions  in  the  flue  gas  leaving  the third  cyclone for two
 runs which span the expected  range.

          TABLE  1II-6.  PARTICULATE EMISSION  PARTICLE SIZE DISTRIBUTION
                                 Particle Size (urn)
  Run      5%       10%       25%        50%       75%       90$       95%
  No. Less Than Less Than  Less Than  Less Than Less Than Less Than Less Than

  78

  80
0.58
0.56
0.66
0.66
0.98
0.85
1.9
1.2
4.6
1.8
8.5
3.4
11.0
5.4
     The particulate loading of the PFBC effluent at various locations within
the off gas system during a typical run (run 80) is shown in Figure 111-38.
Run 80 is typical of good long term operations.  The concentrations are broken
down into five principal size ranges for material leaving each cyclone in the
gas stream.  These loadings were obtained from mass balance calculations around
the cyclones.  Size analysis was performed on a Balston filter catch, as well
as second and third cyclone dump material  taken at a similar time as the Bal-
ston filter flue gas particulate samples were being collected.  These size
analyses together with the mass collected by each cyclone and on the Balston
filter were used to calculate the loadings and size distributions leaving the
primary and secondary cyclones.  The only assumption made was that the par-
ticles did not change size significantly during their pass through the cyclone.
Total particulate concentrations and median particulate sizes in the flue gas
at the same three points, and median particulate sizes in material captured in
the secondary and tertiary cyclones, are given in Table II1-7.  These data
cover the range of values measured in a large number of individual samples from
many runs.  A comparison of the data in Figure 111-38 and the range of values
in Table III-7 shows that run 80 was a typical  run.

               TABLE III-7.  PARTICULATE CONCENTRATION AND SIZE
                             RANGES REPRESENTING A NUMBER OF RUNS
    Cyclone

  Recycle
   (Primary)

  Second
  Third
Part. Cone,
   (g/m3)
  Passing

    8-12
  0.4-1.2

 0.03-0.15
Mass Median Size (pm)
CapturedPassing

               20-25
  20-25

   3-5
3-5

1-3
          Cyclone Efficiency
                  (X)
90-95

85-94
                                      78

-------
PARTICULATE LOADING (gr/SCF)
o — •
o o — ' o
o o • • •
— — — 0 0



n -o
-— ^^ — ~
r^ O >
vy ^
m -y ^

rn

to
ECONDAR
CYCLON
OUTLET
*



o Q ~*
5 P 3
p"* /"^ H^
™ z 3
rn "^


1 1 1 1 1 1 1 1 } 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1






1 i r\ ^r\t im /*"*"* i*1 o' ^
1 IU £U|irn \£.i »Q /Q)











20-40 jim (0%)
>40|im (0%)


(0-5 um (92 0%}

— | 5-1 Opm (7.3%)

10-20|im (0%)
20-40 nm (0%)
>40pm (0%)
i t i I 1 1 1 1 i i t i i i i 1 1 i i i i i 1 1 1 1 i i t i i i 1 1 1 1

-o
70
n
m
O
O
y^
O
rn
z
"^
IS
zz

•
n
O
z
m
s
oo
0
c
— <
r™
m
—I

0 0 — -
* • • \J
o — o x














-n
O
c
70
m
T
&











       oNiavoi

-------
      The primary cyclone  was  Intentionally designed with a low efficiency.
 Its primary function Is to  return  the  larger  particles  back to the bed.  Dur-
 ing normal  operations, the  material  in the cyclone dlpleg has the size dis-
 tribution shown  1n  Figure II1-39.  At  times this dlpleg has become plugged
 and operations have continued as In  run 81.   During this type of operation,
 there is a  large Increase 1n  the material captured 1n the second cyclone, a
 slight Increase  1n  the third  cyclone capture  and a large reduction in the bed
 overflow solids  rate.  This change in  the mass  balance envelope reflects the
 greater amount of material  entrained from the combustor and recycle cyclone
 and captured  in  the last  two cyclones.  At the  exit of the third cyclone, the
 only significant difference from normal operations when the primary cyclone
 dlpleg  plus is a 3-5 percentage point  increase  in particles over 10 ym.  The
 particulate concentration and average  particle  size are not changed signi-
 ficantly.  A  further discussion of run 81 is  contained in another report (6).

      The  secondary  cyclone  is in reality the  first cleanup cyclone.  As shown
 in  Table  III-7 this  cyclone captures 90 to 95% of the particulates in the flue
 gas  exiting the  recycle cyclone.   The  average particle size of the particulates
 captured  by and  passing through the  cyclone is 20-25 urn and 3-5 um respectively.

      The  third cyclone in the train  ~ the second cleanup cyclone --  captures
 approximately 90% of the  particulates  which exit the second cyclone.   The par-
 ticulates exiting this cyclone with  the flue  gas have a mean diameter of 1  to
 3 ym.   Of the particulates  leaving the cyclone with the flue gas, normally
 80  to 90% are smaller than  5 ym.

      During the  course of particulate  sampling, many samples of particulates
 were  found to have magnetic particles  in them.  These samples required special
 techniques for Coulter Counter size analysis, these techniques are described
 in Appendix D.   The  possibility of enhanced particulate collection by magnetic
 interactions was  investigated.  Figure 111-40 is a magnetization curve for
 particles captured  by the third cyclone during runs 99 and 100.   Figure 111-41
 shows the effect of temperature on magnetic induction.  Clearly this  effect
 1s not  useful in the temperature range of interest (800-900°C).   Furthermore
 based on  the type of response curves  seen in the two figures, it is evident
 that the magnetization is  primarily due to the presence of ferrites in the
ash.

Combustion Efficiency

     Carbon combustion efficiencies for runs  made since the  last report were
consistently above 99%.   Only  three runs out  of a total  of 39 were below 99%.
The correlation published  in the previous  report (1)  which relates combus-
tion efficiency to the average bed  temperature,  excess air level  and  gas
residence time, was used to  predict combustion efficiencies  for  these  runs.
The calculated efficiencies  were consistently lower than the  measured  effi-
ciencies.  The average calculated  efficiency  was 98.5% compared  to 99.3%
measured, an average miss  of 0.8 combustion efficiency units.  The standard
error for the correlation  was  previously reported to  be  0.6  units  but  without
the bias shown 1n the recent runs.   If  a consistent measurement  error  occurred
                                     80

-------
                                                          FIGURE 111-39
                                      PRIMARY CYCLONE  DIPLEG SIZE DISTRIBUTION (RUN 85)
oo
            z
            <
            ffi
            Z
            u_
                     2000    1000  600 400
200
100  60   40      20


  PARTICLE SIZE (jim)
1086

-------
CO
ro
                                                        FIGURE 111-40


                              MAGNETIZATION CURVE FOR THE MINIPLANT FLYASH SAMPLE AT  25°C
                        100
                                                 I
                                                   I
                    oo
                    LLJ
                    to
                    I/O
o
 *.
5
 •*
z

B    50
u

Q
z

u
I—
UJ
Z
o
                                                                        0*
                                                50                     100

                                                MAGNETIC FIELD, H, OERSTEDS
                                                                        T50

-------
                            FIGURE 111-41

         EFFECT OF TEMPERATURE ON THE MAGNETIC INDUCTION
               FORCE OF THE MINIPLANT  FLYA5H  SAMPLES
0.03
                    200
300      400

  TEMP. °C —
500
600
700
                             83

-------
in the recent runs,  the error was  apparently less  than 1%.   Therefore,  a
conservative estimate of the average  combustion efficiency  from recent  runs,
obtained from the correlation, would  be 98.5%.   This  is within  the  targeted
range for a commercial PFBC facility.
                                      85

-------
                                 SECTION IV

                      PARTICULATE MEASUREMENT AND CONTROL


     An important technical  issue to  be resolved before  pressurized fluidized
bed combustion can be applied  commercially  is the degree of particulate
removal needed to protect the  gas turbine.  A related issue is which tech-
nology to use to achieve the needed degree  of the particulate control.  In
addition to meeting the particulate removal requirements set by the gas tur-
bine, the environmental New  Source  Performance  Standards published by the
Environmental Protection Agency must  also be met.

     The degree of particulate removal  needed to protect the gas  turbine
from erosion, hot corrosion  and deposition  has  not  been  defined as yet.
Studies, including the 1000  hour turbine cascade exposure  test recently com-
pleted in the mi nip!ant for  the Department  of Energy, are  now underway to
define the needed degree of  particulate control.  Although the EPA is not
directly concerned with studying particulate control  to  protect gas turbines,
it is responsible for the evaluation  of such particulate control  devices to
the extent that these devices  could determine the  level, size and composition
of particulate matter emitted  from  a  PFBC  system.   Therefore a particulate
control program was begun in which  pre-turbine  devices  such as granular  bed
filters, a ceramic fiber filter and high  efficiency cyclones were evaluated.
In addition, post-turbine devices were also tested, based  on the  realization
that the degree of pre-turbine cleanup may not  be  sufficient to meet  the more
stringent New Source Performance Standards  for  particulates.   In  this case, a
trailer mounted electrostatic  precipitator (ESP) and a  trailer mounted  bag
house were connected to the miniplant flue gas  system and  tested  under  typical
low pressure, low temperature  conditions.   A number of these  tests were  con-
ducted in cooperation with other EPA  contractors.   A ceramic  fiber  filter,
built  and tested by Acurex Company, was provided by Acurex and  tested jointly
with them.   The mobile ESP and bag  house were  also tested  in  cooperation with
Acurex.  These devices are operated by Acurex,  for the EPA,  on  a  number  of
industrial gas sources.

     The cyclone tests were carried out primarily by Exxon.   However, Southern
Research Institute and Air Pollution Technology, both EPA contractors,  con-
ducted a series of  particulate sampling tests  to confirm cyclone performance
results obtained by  Exxon using a different particulate sampling system.

      In addition to  evaluating the particulate control devices, it was  also
necessary to develop and Improve particulate measurement systems.  Such sys-
tems were used  to determine particulate concentrations  in the flue gas  before
and after the  control  devices  and to measure particulate size distributions.

      The results of the  particulate measurement and control programs are
reported in  this section.
                                     86

-------
 PARTICULATE MEASUREMENT

      Particulate measurements  In  the  high  pressure, high temperature flue gas
 are difficult to make.   A lot  of  data have been obtained with 18 cm Balston
 high temperature total  filters.   These filters were used on all three sampling
 systems described earlier.   The particles  captured on the filters were sized
 with a Coulter Counter  Model TA11 as  described in Appendix D.  However, there
 is some question as  to  whether the particulate size distribution measured by
 the Coulter Counter  is  the  same as that actually occurring in the flue gas
 before sampling.  For this  reason sampling was attempted with the Southern
 Research Institute (SoRI) 5-cyclone train  and the University of Washington
 (UoW)  7-stage impactor  using the  HTHP sampling system for an 1n-s1tu size
 distribution  measurement.   Verification of results was also made by other EPA
 contractors (Southern Research Institute and Air Pollution Technology) with
 Impactors  designed for  the  high temperature, high pressure conditions.  In
 order  to understand  better  the instantaneous particulate emission, an on-line
 particle concentration  monitor was used.   This device was developed by IKOR,
 Inc. of Gloucester,  Massachusetts.  It is  based on a charge transfer principle.

 Balston Filter Measurements

     Balston  filter  samples have  been used to build a large data base of cyc-
 lone,  granular bed filter and  ceramic barrier filter performance.   This total
 filter  gives  a good  measurement of the particulate concentration provided good
 sampling procedure is followed.   Obtaining the size distribution of this
 material can  be  a  problem.  The Coulter Counter method used to size the parti-
 culates  requires  that a small  amount of solids be dispersed in an  agitated
 electrolyte (Isoton  II).  The  particles are then drawn through a precision
 orifice.  A small  current continuously passing through the orifice is Inter-
 rupted  each time a particle passes through.  The degree of Interruption Is
 proportional  to  the  particle volume.   An accumulator records  the number of
 these interruption pulses whose magntiude falls 1n each of the 16  accumulator
 channels.   The particle diameter reported Is  for a sphere of  the same volume
 as the  particle.

     One question  that arises is  whether the  particulates are agglomerated in
 the flue gas  or on the filter and to  what degree they are redlspersed in the
 agitated electrolyte.  Depending  on  the relative extent of agglomeration and
 redispersion, the measured particle size distribution  could be either coarser
 or finer than that occurring 1n the flue gas.   Also  sufficient particulate
must be captured to  form a filter cake.  If there 1s  no cake  formed,  filter
material may contaminate the sample.   Filter material  consists  of  fine,  long
 cylindrical particles with a volume mean diameter of about  9  ym.

     A cyclone or other  cleanup device sees the aerodynamic equivalent  dia-
meter.  This  1s different from  the volumetric  diameter measured  by  the  Coulter
 Counter.  In order to determine the validity of Coulter  Counter measurements
 for approximating aerodynamic size, some material  was  sent  to  General  Electric
 for size analysis with a Bahco  particle size analyzer.   A comparison  between
 the Coulter Counter and  the  Bahco  results  for  both second and  third stage  cyc-
 lone capture material is shown  1n  Figures  IV-1  and IV-2.  The  distributions


                                     87

-------
                                                    FIGURE IV-1
oo
00
            LLI
            N
            Q
            LU

            <
            O
            I—
            z
            _i
            O
                    COMPARISON OF PARTICLE  SIZE DISTRIBUTION VIA BAHCO VS. COULTER  COUNTER
                               OF SECONDARY CYCLONE CAPTURED MATERIAL (RUN #78)
                                                     EXXON (COULTER COUNTER AND SONIC
                                                                       SIFTER)

-------
00

10
                                            FIGURE  IV-2



             COMPARISON OF PARTICLE SIZE DISTRIBUTION VIA BAHCO VS. COULTER COUNTER

                          OF TERTIARY CYCLONE CAPTURED MATERIAL (RUN #78)
              N

              co


              O
              LLJ
co


Z

<
I
I—

oe.
UJ


s
Qi

O
              O
100





 90





 80





 70





 60





 50





 40





 30





 20





 10





  0
                                            n
                      GE-BAHCO
                                EXXON (COULTER COUNTER)
                                                         Q
                                                          \
                                                    i  i  i
.81.0
                                 2.0   3.0 4|  5 |  6  7 8  910

                                           4.2 5.4


                                          PARTICLE SIZE (jim)
                                                     20
                                                       30  40

-------
are remarkably similar for  two  such different techniques.  The second cyclone
material  was preclassified  with a  sonic  sifter  for  the size distribution above
45 ym and combined with the Coulter Counter  size  distribution to give the
Exxon result.  Balston filter captured  particulates cannot be analyzed with
the Bahco, because there is an  insufficient  quantity of material available for
this instrument.

High Temperature High Pressure
Particulate Sampling	

     The question as to the difference  between  the  size  distribution measured
by the Coulter Counter and  the  true size of  particulates  in  the  flue gas  before
sampling was also addressed.  It was  answered as  part of a joint program  with
Southern Research Institute (SoRI) and  Air Pollution Technology, Inc.  (APT),
both EPA contractors.  The  joint program was intended to resolve the question
of the validity of the filter/Coulter Counter method and thereby confirm  cyc-
lone efficiencies based on  this method.

     The minimum particle size detectable by the Coulter Counter in this
application was estimated to be approximately 0.6 \m.  Thus, the possibility
existed  that the fraction efficiency results might have been biased by using
the mass balance technique with Incomplete size distributions.   Further,  the
ultrasonic deagglomeration of particles prior to performing  the Coulter anal-
yses may have resulted in aerodynamically large agglomerates being measured
by the Coulter Counter as individual  primary particles.  The magnitude of the
possible errors in the fractional efficiency curves resulting from these
biases are difficult  to assess.

     Cascade  impactors have  been used for a number of years  for determinations
of control device  fractional efficiencies over the size range from approx-
imately  0.3 ym  to  10  ym.  Recent work at APT (9),  has shown that cascade
impactor performance  at temperatures and pressures like those in the miniplant
flue gas lines  can be predicted to good accuracy by current theories.  A
pair of  cascade impactors which were designed  by APT, to operate at high pres-
sures and  temperatures was made available by the EPA for a series of indepen-
dent tests  of the miniplant  tertiary cyclone outlet size distribution and
fractional  collection efficiency.

      Inconel  shim  substrates were used  for  particle collection surfaces for
each stage.   Ceramic  fiber  backup filters were used to collect those particles
which were not  removed by  the  impaction stages.  Qualitative verification of
the  performance of the Impactors was obtained  by Coulter Counter analyses
 (where applicable) and by  electron microscopy  of the various stage catches
from  typical  runs.   Previous experience by  Exxon and the samples obtained dur-
ing  this joint  program by  Southern Research and APT showed that the particles,
at the sampling conditions,  were  highly adhesive.   This  permitted valid
impactor results  to  be obtained  even though bare metal substrates were used.
 (Ceramic fiber  substrates  were on  hand  for  use had this  not been the case.)
                                       90

-------
      Figure IV-3 shows comparative results  of  size  distributions of the par-
 ticles In the tertiary cyclone exhaust  stream  obtained by Coulter analyses of
 filter catches from mlnlplant run  105 by Exxon and  those obtained by means of
 cascade impactors during the same  run by SoRI  and APT.

      As seen, the cascade impactor results  Indicate a larger concentration of
 fine particles and a mass median particle size of about 0.8 micron, where the
 Coulter Counter mass median  particle size averaged  1.6 microns.

      Electron micrographs of material captured on the various impactor stages
 Indicated that, with the exception of the first stages, the participate matter
 collected in the Impactor stages was fine,  non-agglomerated participates.
 Electron microscopy also revealed  that  much of the  aerodynamically large*par-
 ticulate matter on the first impactor stage at the  cyclone Inlet was agglo-
 merates of smaller particles.

      These findings indicated that the  results of the Balston filter/Coulter
 Counter method used by Exxon differed somewhat from cascade impactor results
 but to the degree expected,  based  on measurements made in other particulate
 systems.  They also indicated that the  Coulter Counter results were not being
 biased toward the finer particles  by breakdown of agglomerates in the aqueous
 dispersing medium used in the Coulter Counter, since the Coulter Counter gave
 a  coarser distribution than  the impactor.

      Prior to the joint impactor sampling program with SoRI and APT, samples
 were taken by Exxon with a University of Washington (UoW)  Mark III cascade
 Impactor.   The 7-stage impactor was used to obtain background experience with
 impactors  at high temperature and  high  pressure.   The impactor was used with
 the HTHP particulate sampling system.   It was heated to  750°C and pressurized
 to  900 kPa.   Several  runs were made using the Impactor which was designed for
 the dilute sampling  conditions.  These runs showed that  the particulate was
 very adhesive and would  stick  to bare metal  substrates.  Also 1t was shown
 that the  pre-cut  cyclone  used ahead of the Impactor to prevent overloading
 of  the first stages  was  not  needed.  It captured  very little material  and
 after  most runs the  first two stages of the impactor were  empty.

      Particle size distributions obtained with the UoW Impactor were Indeter-
minate.  The fmpactor was erroneously run at very high flow rates.   This  tended
 to  increase  jet velocities to the point where much of the  material  that should
 have remained  on  a given stage may have been "washed" off  onto a following
 stage.  This may  be  the reason that the Balston backup filter captured a  sub-
 stantial portion  of  the particulate.  The stage cut diameters were calculated
with classical formulas, and the size distribution obtained.   The size dis-
 tribution was much finer than that of a  Balston filter (with  Coulter Counter
 size analysis) taken at the same time.   This 1s further  evidence that  the  flow
rates were much too high.

Mini Cyclone Train Sampling--
     The SoRI  5-cyclone train was  used  to try to  obtain  gram  size quantities
of particulates which pass through  the  tertiary cyclone.   This  sampling device
was also intended to give an in-situ distribution  of the particulates  once it


                                     91

-------
                        FIGURE IV -3
LU
N
oo

LLJ

U
      BALSTON FILTER/COULTER COUNTER VS. APT IMPACTOR
                 PARTICLE SIZE  DISTRIBUTION
        FLUE GAS FOLLOWING THIRD CYCLONE (RUN  105)
   10.0


    I*
    6.0-
    5.0-

    4.0

    3.0


    2.0
l.(
    0.3
    0.2
0.1
 2%
                                     CASCADE IMPACTOR
                                     BALSTON FILTER/
                                     COULTER COUNTER
                 1
                  1
1
1
1
i
i
i
I
I
            5   10    20   30  40 50 60  70  80    90  95

              PERCENTAGE SMALLER THAN PARTICLE SIZE
                                                      98%
                               92

-------
 was calibrated.   The  cyclones were used with the HTHP sampling system.  They
 were heated  to  between  720 and 870°C and pressurized to 900 kPa.  The problems
 associated with  the 5-cyclone train were mostly due to the very dilute nature
 and small size of the particulates in the flue gas.  Long sample times were
 required and  small amounts of particles had to be removed from the relatively
 large catch cups.  The  system was very prone to leaks and contamination by
 oxidized metal and the  anti-seizing compound used on the nuts and bolts that
 hold it together.  The  total loading of particulates measured was typically
 one-half the  loadings obtained with concurrent Balston filter sampling.  Also,
 the size differentiation of the cyclone train was much too broad at the samp-'
 ling conditions  to describe the particle size distribution.  The cyclone cap-
 ture  efficiencies probably due to re-entrainment of the dust from the catch-
 pots  on each  cyclone.   Therefore, with the cyclone stage cut diameters a func-
 tion  of sampling duration, calibration was not possible.  For these reasons
 the SoRI 5-cyclone train was not deemed suitable for use in the downstream
 location to measure particulate loadings or size distributions.  Modifications
 may have improved its performance under these conditions, but this was out of
 the scope of  this program.

 Continuous Particulate  Monitor

      A continuous, on-line particulate monitoring system was evaluated during
 run 80.  The monitoring system (manufactured by IKOR, Inc., Gloucester,
 Massachusetts) makes  use of the surface charge that particulates accumulate
 during their  flow through the combustion unit.

      The IKOR probe measures the total  electrical  charge of impacting solids.
 The probe consisted of  a 0.63 cm diameter solid rod of type 316 stainless
 steel that was 35.5 cm  long.  The last 8.9 cm of the probe was immersed in
 the 10 cm diameter exhaust duct of the miniplant at a location 2.1  meters
 downstream of the GE  turbine test cascade.  The probe was mounted on the
 duct's center line and  normal  to the gas flow direction.   A constant purge of
 dry N£ was maintained around the probe to prevent condensation on the probe
 insulator.

      The output of the  probe was correlated with gas temperature measured near
 the probe location and with average particulate loadings measured by Balston
 filter samples.  The  data plotted in Figure IV-4 compares the IKOR signal
strength with gas temperature  during run 80 at relatively constant dust load-
 ing.  The particulate loading data obtained using the Balston filter during
run 80 are summarized in Appendix M-6.  The particulate data show that the
average mass  mean diameter of the particles captured by the Balston technique
following the third cyclone was  1.2 ym and the average loading was  0.051 g/Nm^
The loading varied between 0.034 and 0.067 g/Nm3 during the IKOR testing
period.  The  samples contained 6% of particles larger than 5 ym.

      The data in Figure IV-4 show that when the particulate loading downstream
of  the third  cyclone averaged  0.051  g/Nm3, the IKOR probe signal  strength
varied almost linearly with gas  temperature.   Natural  gas was injected into
 the combustor exhaust stream upstream of the  third cyclone.  The gas tempera-
 ture  at the probe location was varied by adjusting the amount of natural  gas
 flow.

                                     93

-------
                      FIGURE IV-4




EFFECT OF TEMPERATURE ON IKOR MONITOR READING (RUN 80)


0
^—
II
>
SENSITI'
©
LU
w
— J
D
u_
tfS



IUU
90


80
70
60
50
*J\J
40
30


20

10
0
65
O -, o-
Z Z °co
0^1-7
< ^ co \
O t °. d
— 1 ^^ ^^
i j
• v\
" XCONSTANT
"•/J LOADING
* IBUT TEMP.
•"/• JCHANGEVI;
f ICH4 INJEC-
./ [TION
/ !
• J '
• / !
COMBUSTOR ON LIQUID FUEL X1 \
1 \
• i i * i i !
0 700 750 800 850 S




-
—
-

<\



-

-

>0(
             GAS TEMPERATURE @ PROBE (°C)

-------
      The  unexpected  sensitivity of the IKOR signal to probe temperature made
 comparison  of signal  amplitude with gravimetric sampling very difficult.
 Because of  this  and  the  unknown effect of thermionic emission from the probe
 or  pressure  vessel walls,  further testing was stopped.  IKOR agreed to investi-
 gate  the  problems and possibly develop a temperature compensator for the unit.

 CYCLONE STUDIES

      The  miniplant was initially designed with only two stages of cyclone
 cleanup.  The  gas was sampled and expanded after the secondary cyclone.  The
 particulate  loading of the gas was much too high to pass through any gas tur-
 bine.  A  granular bed filter was then added to the flue gas stream.  The pro-
 blems of  this  device  are contained in this and a previous report (1).  A con-
 ventional cyclone was installed upstream of the granular filter to lower the
 dust  loading and possibly improve efficiency.  This cyclone performed very
 well  and  lead  to a halt in granular bed filter testing.  All subsequent hot
 gas cleanup was accomplished with three stages of cyclones.  However, further
 tests of  hot gas cleanup systems were conducted on a slipstream of flue gas
 withdrawn after the second stage cyclone.  Low temperature systems were also
 tested.   These tests are discussed in subsequent sections.

 Cyclone Efficiency Testing,

     Particulates in the flue gas  leaving the miniplant recycle cyclone pass
 through the two additional  cyclones which remove ~99% of the particulate
matter.  Mass balance calculations based on particulate material  captured in
 the second and third cyclones and  in the flue gas  leaving the  third cyclone
 are used to determine the particulate size and concentration at the exit of
 all  three miniplant cyclones.

     These measurements and calculations show that material  captured by the
 second cyclone has a mass median particle size of  20 to 25 urn.   The overall
 cyclone efficiency is about 95%.  Material  captured by the third cyclone has
a mass median size of 3 to 5 ym.  The overall  efficiency of this cyclone is
about 90%.  Particulate concentration in the flue  gas leaving  the third cyc-
 lone is generally 0.03 to 0.15 g/Nm3.   The mass median size of the particulate
ranges from 1 to 3 microns as determined by Coulter Counter.  These results
are  summarized in Table IV-1  .  Detailed particulate loading and size data
for  each run are shown in Appendices M-3 through M-6.

                  TABLE IV-1.  PARTICULATE CONCENTRATION AND
                         SIZE MINIPLANT CYCLONE SYSTEM


                  Part. Cone.       Mass Median Size
                      (g/Nm3)            (Microns)               Cyclone
      Cyclone       Passing       Captured     Passing     Efficiency (%)

      Recycle         8-12            --        20-25

      Second        0.4-1.2         20-25        3-5            90-95

      Third        0.03-0.15         3-5         1-3            85-94

                                      95

-------
     The particulate emissions  after the  tertiary  cyclone  have been as  low as
0.03 g/Nm3.  This  is slightly below the New Source Performance Standard  (0.03
Ibs/MBTU) which corresponds to  -0.035 g/Nra^.  The  performance was  not con-
sistent however, and usually exceeded the emission standard.  This performance
is not good enough to use cyclones  alone  to meet the  emission standards.

     The third cyclone fractional  efficiency is  of greatest  interest, since
if a three stage cyclone system is  used  in a commercial  PFBC system, the  third
stage must be very efficient to prevent  damage to  the gas  turbine  behind  it.
The collection efficiency of the tertiary cyclone  has been measured over  some
20 runs with run durations between  8 and  250 hours.  Throughout  this period
the total collection efficiency averaged  90% with  a standard deviation  of
only +_ 3%.  This summary is shown in Table IV-2.  As  can be  seen from the
table, these measurements were made with  3 different  coals and 2 different
sorbents with a variety of Ca/S ratios.

     Both the total efficiency and  the fractional  (grade)  collection efficiency
was much higher than expected.  This cyclone, which was  designed with classical
handbook formulas, averaged a cut diameter (50% efficiency), over  this  period,
of 0.88 vim ^0.02.  Table IV-3 shows the  grade efficiencies  measured during
runs 68 through 100.  These efficiencies  were obtained from  outlet samples
and captured samples sized with a sonic  sifter and a  Coulter Counter.   The
inlet concentrations were calculated by  size differentiated  mass balance under
the assumption  that no size change of particles occurred through the cyclone.
During runs 99 and 100 this assumption was checked with  filter samples  taken
upstream of the cyclone.  Although cyclone performance may not be  sufficient
to meet emission standards, cyclones performing as well  as those used  on the
miniplant may be sufficient to protect the gas turbine from serious damage  by
erosion.

     The main purpose of the particle size distribution study  (described ear-
lier) by the Southern Research Institute and A1r Pollution Technology was to
measure cyclone fractional collection efficiency.  Figure IV-5 shows the frac-
tional efficiency of the third cyclone during run 105 as calculated from; (1)
the SoRI/APT cascade impactor data,  (2)  the Exxon total  filter/mass balance
technique with  Coulter Counter size analysis, (3) the Lelth and l_1cht (10)
cyclone  fractional efficiency model.  The cyclone operating conditions  are
shown 1n Table  IV-4.  The  Impactor and the  total filter/Coulter Counter effi-
ciencies agree  fairly closely except in the small particulate size range.
The cyclone cut diameter  (50% efficiency)  1n  both cases was approximately 0.7
microns.   Therefore, the cyclone efficiencies calculated  from cascade Impactor
data substantially confirms  the efficiency  obtained  from  total filter/Coulter
Counter  data.   However,  the  predicted fractional  efficiency curve is signifi-
cantly  lower than  the measured results.   The  Lelth and Licht model, shown in
Figure  IV-5, comes  closer  to predicting miniplant  third cyclone performance
than other available models  tested.  The  reason for  the lack of agreement
between  the measured and  predicted  results  is not  understood at the present
time.   It  may  be  due to  the  fact that many  of the models  are semi-empirical
and  based  on data  obtained with  particulate generated by  other  systems at lower
temperatures and  pressures.
                                      96

-------
           TABLE IV-2.  TERTIARY  CYCLONE TOTAL EFFICIENCY SUMMARY
Run #
Total Eff.
67.1
67.2
67.3
68
71
72
73
74.1
74.2
75
78.2
78.4
78.10
79
80
81
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
Average Of
All Runs
+ Is
88.4
83.4
88.7
97.0
82.9
91.3
94.3
94.6
94.5
89.1
89.2
90.0
90.1
93.9
84.6
91.2
92.4
95.9
94.7
92.0
94.2
92.6
88.6
90.9
91.1
88.6
89.8
89.9
90.4
91.1
90.7
86.9
89.6
88.4
90.6 +3.3


Coal  Type
                            Illinois
                            Illinois
                            Illinois
                            Illinois
                            Illinois
                            Ohio
                            Ohio
                            Ohio
                            Ohio
                            Ohio
                            Illinois
                            Illinois
                            Illinois
                            Illinois
                            Illinois
                            Illinois
                            Illinois
                            Illinois
                            Illinois
                            Illinois
                            Illinois
                            Illinois/Ohio
                            Ohio/Champion
                            Champion
                            Champion
                            Champion
                            Champion
                            Champion
                            Champion
                            Champion
                            Champion
                            Champion
                            Champion
                            Champion
Sorbent Type     Ca/S Ratio
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Limestone
                                       Limestone
                                       Limestone
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomi te
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                       Dolomite
                                     1.25
                                     1.25
                                     1.25
                                     2.00
                                     1.52
                                     1.79
                                       14
                                       60
                                       80
                                       65
                                       45
                                     1.45
                                     1.45
                                     1.45
                                     1.41
                                     1.56
                                     1.45
                                     1.45
                                     1.45
                                     1.45
                                     1.45
                                     1.45
                                  1.40/1.87
                                     1.87
                                     1.50
                                     1.50
                                     1.50
                                     1.50
                                     1.50
                                     1.50
                                     1.40
                                     1.40
                                     1.40
                                     0.38
                                   97

-------
            TABLE IV-3.   TERTIARY  CYCLONE FRACTIONAL EFFICIENCY SUMMARY



              Particle Size Captured with Stated Efficiency (urn)

   Run No.           50%           75%          85%          90%          95%
    68              N/M          N/M          1.15         1.60          2.15
    71              1.10         1.80         2.30         2.60          3.20
    72.1            N/M          0.95         1.55         2.10          3.15
    73.1            0.70         1.00         1.15         1.20          1.50
    74.1            N/M          0.85         1.15         1.55          2.10
    74.2            N/M          N/M          0.80         1.55          2.50
    75              0.80         1.45         1.85         2.10          2.60
    78.2            0.85         1.20         1.50         1.70          2.50
    78.4            1.05         1.40         1.70         1.85          2.45
    78.10           0.75         1.20         1.55         2.00          N/M
    79              1.15         1.45         1.65         1.80          2.00
    80              1.05         1.65         1.95         2.30          4.50
    81              N/M          N/M          1.35         1.60          2.20
    86              N/M          N/M          1.55         2.20          3.20
    87              N/M          1.85         2.20         2.35          2.70
    96              N/M          N/M          1.45         2.00          3.00
    99.7            N/M          N/M          0.90         2.00          4.50
    99.7*           N/M          1.00         2.50         4.00          6.35
   100.3            0.70         1.25         1.60         2.00          2.30
   100.3*           0.65         1.25         1.60         1.90          2.35
           Is   0-88 ^ 0.02   1.31 + 0.3   1.57 +. 0.4   2.02 + 0.6   2.91 +.  1.13
 *  Efficiency calculated based on upstream Balston  total  filter  sample.

N/M  =  Beyond range of Coulter Counter.
                                       98

-------
    100




     90




     80




     70




     60




     50




     40
                                           FIGURE IV-5


                            TERTIARY CYCLONE COLLECTION EFFICIENCY

                                            (RUN 105)
u

LLJ
z
o
O    30
U



     20
     10-
      0

      15
                     BALSTON FILTER/COULTER COUNTER

                     CASCADE IMPACTOR

                     LEITH  AND  LIGHT MODEL (THEORETICAL)
                J	I	I	I    I
_l_J	I	I	I    I
               10  9 8   7   6
  .9 .8  .7  .6   .5    .4
                                        PARTICLE SIZE, jim

-------
           TABLE  IV-4.   THIRD  CYCLONE OPERATING CONDITIONS  (RUN 105)


                    Pressure                 700 kPa
                    Temperature              635°C
                    Flow rate                14.6 Nnr/min
                    Inlet Velocity           36 m/sec
                    Pressure  drop            4 kPa

     Knowlton and Bachovln (11)  found that   pressures  up to 5.6 M Pa had little
effect on cyclone performance.  Their work was based on much higher dust load-
ings than the current study.   Perhaps other  effects, peculiar to the PFBC sys-
tem such as the nature  of the  particulates,  or the  operating temperature and
pressure are responsible for  the high efficiency.   Additional work will be
needed to explain the observed results.

     The conclusions reached  by this study were basically  to confirm the high
third cyclone efficiencies measured with the Balston filter/Coulter Counter
technique.  The Balston filter/Coulter  Counter technique was not being  biased
toward finer size distributions by particle  deagglomeration in  the aqueous
dispersing medium used in Coulter Counter measurements.   In fact, the Coulter
Counter gave a coarser size distribution than  the  impactor, which is as expected
based on other measurements made in other particulate  systems.

     Further work is required to optimize cyclone  performance under PFB con-
ditions.  The emissions of particulate  with  3  stage cyclone cleanup have, at
times, met the new source performance  standard of  12.9 ng/J (0.03 Ibs/MBTU),
set by the EPA.  However, the performance was  not  consistent enough to  meet
the standard over the long term.  The  improvement  in  performance  required is
small and may be met by cyclone optimization.

Cyclone Optimization and Variables Study

     A cyclone variables and optimization study was undertaken  during  the last
7 mini plant runs.  During these runs (nos.  109 through 115) three different
cyclones were exposed to various PFB operating conditions  in the  tertiary
position.  Each cyclone was located inside the vessel  that once contained the
granular bed filter elements.  In this way, the  plunger type valves that were
used to provide blowback shutoff of the granular  filter elements  could  be
used to shut off the cyclones.  All three cyclones could be tested, one at a
time, during every run.  Each cyclone was equipped with an orifice  in  the inlet
line.   In  this way, valve leakage could be detected and flow to the cyclones
verified.  A AP cell was used to measure the pressure drop between  the  pressure
vessel  inlet and outlet.  This was the same location used in previous measure-
ments with only one cyclone.

     The dimensions of  the cyclones are shown in  Table IV-5 and Figure  IV-6.
The  "A" cyclone was the  original miniplant tertiary cyclone which had  almost
2000 hours of combustor  operating time at the start of the test program.   The
"B"  and  "C" cyclones were newly constructed, scaled down versions of cyclones
currently  being tested  by American Electric Power and Stal-Laval  in England
at  the  National Coal Board facility in Leatherhead.

                                      100

-------
             TABLE IV-5.   TERTIARY CYCLONE  DESIGN  DIMENSIONS

                                  (cm)
ymbol
--
%
W
hi
ho
"B
he
Do
0/1
Dd
Dp
hp
•essur
Description
Inlet Type
Barrel Diameter
Inlet Width
Inlet Height
Outlet Pipe Insertion
Barrel Height
Cone Height
Outlet Pipe Diameter
Outlet/Inlet Ratio
Diameter of Cone at
Dust Pot.
Dust Pot Diameter
Dust Pot Height
e Drop @ 30.5 m/sec
Cyclone
A
Tangential
15.2
3.81
7.62
11.4
40.6
20.3
6.2
1.04
None
None
None
6.6
Cyclone
B
Scroll
17.8
3.66
8.18
7.62
22.9
44.5
5.25
0.76
1.94
12.4
25.4
11.0
Cyclone
C
Scroll
17.8
3.66
8.18
7.62
22.9
44.5
3.51
0.31
1.94
12.4
25.4
28.6
Inlet Velocity  8409C
  930 kPa (calculated)
                                 101

-------
     FIGURE IV-6
BASIC CYCLONE  DESIGN
          102

-------
      The parameters  that  were  changed  in  this study included;  cyclone geometry,
 inlet velocity, inlet temperature,  inlet  particulate loading and combustor
 coal  feed.  Cyclone  geometry was  studied  by using more than one cyclone at
 each  condition.  Inlet velocities were  lowered during one test by lowering
 the combustor fluidization air by 26%.  The lower combustor superficial
 velocity also lowered the cyclone inlet temperature.  Cyclone inlet tempera-
 ture  was also varied  by not injecting,  or injecting very little natural gas,
 This  in  turn  varied the inlet  velocity.   Inlet particulate loading was varied
 by  deactivating the secondary  cyclone for one series of tests.  The effect
 of  combustor  coal  feed  on cyclone performance was determined by varying the
 combustor  feed  coal from  the Champion coal used for most of the test to
 Illinois  No.  6  coal.   During this test the dolomite sorbent feed rate was
 approximately constant.

      The  cyclone  performance tests were carried out in one day (6-10 hour)
 runs.  The combustor  was  allowed to equilibrate for 1-1/2 to 2 hours before
 any testing was done.   After steady state was attained, Balston total  filter
 samples were  taken simultaneously at the cyclone inlet and outlet.  At the
 conclusion of the  sampling a different cyclone was put on line.  The next
 cyclone  reached steady  state during the turn around of the filter (one-half
 hour) and  the sampling  was started all over again.  The cyclone particulate
 catch was also  sampled  before and after every gas particulate sample.

     The total  collection efficiencies are shown in Table IV-6.  For the
 baseline case the  total efficiency of the "A" cyclone was 84 +_ 2%.  The B and
 C cyclones averaged 83  +_ 4% and 72 + 6% total  collection efficiency respec-
 tively under  baseline conditions.  The standard deviations  are such that
 most of the non-baseline data are within the data scatter of the baseline
 data.  This obscured many of the effects intended to  be studied.

     The cyclone grade efficiencies  were also  calculated for each of the
 tests.  These were obtained from the inlet and outlet  filter sample concen-
 tration and size distributions.  These grade efficiencies contained as  much
data scatter as the total  efficiencies.   For the baseline case,  the cyclone
cut diameters averaged 1.5,  1.1, and 1.7 ym for the A,  B and C cyclones
respectively.

     Further data concerning  particle size distribution  in  the flue gas
prior to, and after,  the cyclones are presented  in Appendices  M-5 and M-6.

     The pressure drop measured across  all  cyclone systems  was  fairly high
 (60-90 kPa).   However, if the  pressure drop  due  to the  orifices  and  the rest
of the flow distribution system are  subtracted,  the pressure drops  of the
cyclones were close to the calculated values.   The pressure  drop  obtained  for
the "A", "B"  and "C"  cyclones was 6, 12  and  40 kPa respectively.   These com-
 pare favorably with calculated  pressure  drops  of 6, 12 and  29  kPa.

     The conclusion reached from this  study  was  that no  significant  effects of
 pressure, inlet velocity,  temperature,  and combustor coal feed were  found.  An
 increase in Inlet particulate  loading did  increase overall efficiency signi-
ficantly.  The slight decrease  in the  cut  diameter it caused was  within  the
data scatter of the measurements.  None  of the cyclones worked as well  as
expected.

                                     103

-------
                       TABLE IV-6.  SUMMARY OF CYCLONE TEST PROGRAM RESULTS
                                                                                      Particle Size
                                                                                      Collected with
Run
No.
109
m
113
no
112
113
114
105
109
111
113
no
no
112
114
109
m
115
112
Cyclone
AM \
A ri
i 1 i
A
A
A
A
A/o ^
A(3)
B(D
B/f {
gU )
B
B
B
B
C(D
C\ f
/I \
C(D
C
Inlet
Velocity
(m/sec)
42
42
42
39
25
42
42
36
41
41
41
34
38
25
41
41
41
41
25
Inlet
Temp.
CC)
840
840
840
760
700
840
840
635
840
840
840
650
760
700
840
840
840
840
700
Coar2^
Type
CH
CH
CH
CH
CH
IL
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
CH
Inlet
Loading
(q/m3)
1.87
1.61
0.85
1.35
1.54
1.73
4.83
1.62
1.23
1.13
1.79
1.36
1.24
1.25
5.06
1.19
0.91
0.54
1.14
Inlet
Loading Total Collecti
(g/m3) Efficiency (J
0.268
0.143
0.163
0.160
0.152
0.269
0.304
0.176
0.255
0.179
0.249
0.240
0.138
0.302
0.393
0.387
0.191
0.171
0.236
85
85
81
88
90
85
94
89
79
84
86
82
89
76
92
68
79
68
79
Stated Efficiency
on (u)
;) 50%
NA
NA
1.5
1.0
1.2
1.4
0.7
1.2
1.0
1.2
1.0
1.4
1.2
1.3
1.1
2.1
1.3
1.7
1.2
75%
2.0
NA
1.8
1.5
1.4
1.7
1.3
1.6
1.6
1.4
1.4
1.8
1.4
1.6
1.4
2.8
1.8
2.1
1.3
902
4.0
NA
2.1
2.8
1.7
2.2
1.9
2.1
2.9
2.0
1.8
2.4
1.8
2.2
2.0
3.8
2.7
3.4
1.8
(1)  Baseline expected commercial/design operating conditions
(2)  CH = Champion Coal (Ca/S = 1.25)
     IL = Illinois No. 6 Coal (Ca/S = 0.76)
(3)  Pressure Reduced to 700 kPa, All others at 900 kPa

NA - Data Not Available

-------
     The "A" cyclone did not perform at the 907» collection efficiency  measured
over the previous 2000 hours.  This was probably due to changes  in  geometry
due to erosion, repair and thermal  stress.  The "B" and "C" cyclones also
did not collect particulates with the expected efficiency.  Ironically,  this
may be due to a lack of exposure.  Any rough welds or edges had  not had  suf-
ficient exposure to the erosive gas to smooth them to their steady  operating
cond-itions.  As can be seen in Table IV-2 the total  collection efficiency of
the "A" cyclone was also lower and  slightly more eratic during Its  first 100
hours of exposure (run 66-68} than  later on.

     Comparisons of "B" and "C" cyclone performance with the Van Tongeren
Model 850 cyclones used by AEP and  Stal-Laval in England are underway.  The
two cyclones used 1n England are geometrically similar to the "B" and  "C"
cyclones.  The results from these tests will be followed to develop more
reliable scale up procedures for cyclones in PFBC service.

CERAMIC FIBER FILTER EVALUATION

     As mentioned previously, cyclones do not appear to be capable  of  consis-
tently meeting the New Source Performance Standard for particulates.   However.
cyclones may be adequate to protect gas turbines, based on early results from
extended materials tests.  Two devices, with the potential for meeting the
environmental emission standard as  well as the turbine requirement, were
tested on a small flue gas slipstream from the miniplant.  The first of these
was a ceramic fiber filter developed and supplied by Acurex Corporation under
an EPA contract.

     Previous development work on the ceramic fiber filter concept  is
reported elsewhere (28).  The testing of the fiber filter on the miniplant
will  also be described in a separate report by Acurex.

Test Description

     The test device used in this study was provided by Acurex Corporation.
It consisted of a single filter element contained in a heated pressure vessel
with all necessary cycle controls.   The filter consisted of a loosely  packed
mat of Saffil alumina generally about 1 cm thick sandwiched between an open
weave fine gage 304 stainless steel support screen.  The alumina mat,  support
screen sandwich was formed into a cylindrical filter element which  was slipped
over a heavy gage support cage and  clamped at the top and bottom.

     The objective of this study was to demonstrate the feasibility of ceramic
filtration under actual PFBC conditions.  Data were also obtained on  the effects
of cleaning cycle, face velocity, temperature and bag age on filtration
efficiency.

     A cross section of the filter housing is shown in Figure IV-7.  Hot,
dusty inlet gas enters the unit from the side, below the test filter.   This
gas impacts against a plate on the dust hopper.  Heavy particles may  remain
in the hopper while others travel upwards to the filter element. The  filter
element was 10  cm in diameter by 45.7 cm long.  A heater element surrounded
                                     105

-------
                 FIGURE IV-7


FILTER HOUSING PRESSURE VESSEL CROSS SECTION
                            r\  GAS
                            V OUTLET
    REVERSE FLOW

       INLET
    GAS INLET [)
                                   TEST FILTER
                                    HEATER ELEMENT
                                   DUST HOPPER
                      106

-------
 the  test filter and was used to maintain gas temperature in the test filter
 zone.  After removal of particles by the test filter, hot gases exit the
 chamber through a pipe in the top of the vessel.

     Figure IV-8 illustrates the installation of the test filter on a flue
 gas  slipstream downstream of the miniplant second stage cyclone.  The flue gas
 flow rate through the filter was 1.0 to 2.3 Nm3/min.

     The electronic time sequencer controlled the operation of the cleaning
 cycle.  Cleaning cycle parameters were adjustable but the basic cleaning
 sequence was as follows:

     t  Start on a timed interval by closing a solenoid valve downstream
        taking the filter off-line

     t  Start a gentle reverse flow of unheated gas

     •  Release one or more cleaning pulses, (amplitude, duration  and
        pulse interval  are all  adjustable)
     •  Wait several seconds for dust removed during pulsing to fall  into
        the dust hopper

     t  Stop reverse flow

     •  Open the downstream solenoid valve, returning the filter to service


     Figure IV-9 is a photograph of the filter unit  installed on the miniplant.
The slipstream for the bag filter leaves the main flow duct through a 1  inch
 pipe.  Two high-temperature valves, a 1-inch Mohawk  ceramic gate valve and a
1-inch Kamyr ball  valve, were used to isolate the filter from the  PFBC.  Just
before the filter vessel, a gas bypass line allowed  extra gas to be withdrawn
from the PFBC to maintain temperature in the inlet line.  This bypass line was
also used to preheat the inlet line prior to the start of filtration.

     The filtered gas leaving the top of the filter  pressure vessel  cooled
down to 440°C before it entered the Balston total  filter shown in  Figure IV-8.
The weight gain of this filter was used to  determine the outlet particulate
concentration.  The gas was further cooled  and the water removed in a knockout
vessel  before it was measured through a flow orifice and expanded  through  a
ball valve.  Pressure drop across the ceramic filter was continuously measured
and recorded.  Inlet particulate concentration was measured by extracting  a
sample and passing it through a Balston total  filter.

     The filter was evaluated during runs 82 through 96.  Runs 82  to 85  were
devoted to system shakedown.  Typically, a  run lasted for one working day. At
the end of that time the Acurex technician  changed the test filter and a new
test was begun the following day.

     Several problems with valves occurred  during the shakedown runs  82  through
85.  The Kamyr valve failed during run 83.   It was removed  and not replaced.
The Mohawk valve bonnet leaked during run 84.   That  gasket  was replaced  with a
copper gasket and  the valve performed satisfactorily until  the alumina gate
cracked during run 93.   The solenoid valve  that shuts off the filtered gas

                                     107

-------
                                            FIGURE  IV-8
                            ACUREX  TEST FILTER INSTALLATION  SCHEMATIC
                 GAS BYPASS
o
00
                     KAMYR
                        WATER
                        KNOCK
                        OUT
                  SAMPLE PROBE
                                                Reverse
1
1
^

1



Flow
Valve
1
PULSE '
WAIVE
                                      BALSTON
                                      TOTAL
                                      FILTER
COOLER
FLOW ORIFICE
                       910 kPa
                       CLEAN AIR
                       1300 kPa
                       CLEAN AIR
           ELECTRONIC
              TIME
           SEQUENCER
                                       %* 1/2" KAMYR
                                               i
                                              .j
  1" BALL
(FLOW RATE
CONTROL)
                          1" SOLENOID
                         OPERATED BALL
                                                                                •>TO  SCRUBBER
                                    -> TO BALSTON FILTER SAMPLING SYSTEM
                        1/2"  KAMYR
                  1" MOHAWK
                       OFF GAS FROM SECOND CYCLONE

-------
              FIGURE IV-9
ACUREX HTHP CERAMIC BAG FILTER SITE
                                            FILTER SUPPORT CAGE

-------
flow during cleaning  failed  in  run  85.  The  teflon  seat  of  this valve had
become damaged by hot gas.   The valve was  replaced  with  a solenoid operated
ball valve which functioned  well  for the remainder  of  the test.  Otherwise,
shakedown ran smoothly and was  mainly used to  optimize the  cleaning  phase of
the filtration cycle.

Test Results

     Pressure drop across the filter  bags  varied  as a  function  of  time  in a
manner typical of fabric filters.  Figure  IV-10 illustrates typical  pressure
drop and flow recordings.  Baseline pressure drop was  defined as the pressure
drop of a bag at the start of the filtration cycle.  Graphically,  this  point
corresponds to the first point in each  cycle in Figure IV-10.   Many  filters
were operated for several hours at baseline  pressure drops  close to  that  of
a new bag.  The baseline pressure drop  was always between  0.1 and  5.0 kPa,
usually between 0.2 and 2.0 kPa.  As shown in Table IV-7,  the baseline
pressure drop did increase with time in a  number of runs,  but this is typical
of  fabric filters during initial operation.   Pressure drops before cleaning
were never allowed to exceed 14  kPa to reduce the chance of bag failure.
High pressure drops  generally caused the  inner filter support screen to bulge
into the cage on which the filter was fastened.  Baseline  pressure drops  were
slightly higher when  filtering Champion coal than when filtering Illinois
No. 6 coal under similar conditions.  Outlet particulate loadings were
slightly lower with  Champion coal than with Illinois  No. 6 coal.  The overall
influence  of  coal type was small, and could not  be quantified from the
relatively few tests  completed at the miniplant.   Filter bags were used for
from 4.5 to  19 hours.  A summary of test  conditions,  pressure drops, and out-
let particulate  loadings is  shown  in Table  IV-7.

      Filtration  efficiencies for the Acurex ceramic bag filter  were generally
over  90%,  ranging from 96 to 99.5%.  An exact filtration efficiency was dif-
ficult  to  determine  because  of problems in measuring  the filter inlet parti-
culate  concentration.  Filter  inlet particulate  concentration was  measured or
calculated by three  methods:   (1) Balston total  filter catch on an extracted
sample, (2)  mass  balance around  the third miniplant cyclone, (3) mass balance
around  the ceramic bag  filter.   The results obtained  by these three techniques
were  not consistent  as  shown in  Table  IV-8.

      The test filter inlet  line  (Figure IV-11) was  1-inch  schedule 80 pipe
taking  a  sample  from a  4-inch  schedule 5  pipe.   Isokinetic flow would have
been  1.15  Nm3/min.   This flow  alone, would  have  cooled  to  450°C before reach-
ing the filter  vessel.   A bypass flow  of  1.4-2.3 Nm3/min in addition to the
filtered  gas, was drawn  through  the line  to help maintain  temperature near
800°C.   The  flow into the filter inlet line was  therefore  200 to 300 percent
isokinetic.   The Balston total  filter  inlet sample (1/4-inch probe) was incor-
rectly  operated  isokinetically with respect  to  the filtered gas only, neglect-
 ing the bypass  gas which was present at that  point.   Therefore, the  Balston
 total  filter samples were taken  at only 30  to  50%  isokinetic rates.  For this
 particle size range  and  loading, isokinetic  flow appears to be important, and
 this  inlet loading,  measured with the  Balston filter, may  be treated as  a
 lower limit of particle concentration.


                                       110

-------
                              FIGURE  IV-10
        ACUREX HTHP CERAMIC BAG FILTER PRESSURE DROP AND FLOW
O 70
 CN
I

 E 60
O  50
m  40
0£


   30
at  20
LLJ


E  10



    0
                             CLEANING

                              ^PULSES
                                                              OBASELINE
                                  5 MIN
                            TIME
?1.5
0
"-0.5
cd
LLJ
1—
_l
0
s 	 ,
-






CLEANINGr1
CYC LE ^








                           TIME
                                      -5 MIN
                                  111

-------
TABLE IV-7.   ACUREX HTHP BAG FILTRATION SUMMARY

































(1)
(2)
(3)
(4)


(5)

Run No
83
D4
85
86
87.1
87.2
87.3
88.1
80.2
88.3
89.1
89.2
89.3
89.4
90
91.1
91.2
91.3
92.1
92.2
93.1
93.2
94
95
96.1
96.2
96.3
96.4
96.5
96.6
96.7

Coa
Outlet
1 Load
Flow Face Average
Rate Velocity Tempera tui
Ba
-------
     TABLE IV-8.  ACUREX HTHP BAG FILTER INLET PARTICULATE  LOADINGS



Run No.
86
87.1
87.2
87.3
88.1
88.2
88.3
89.1
89.2
89.3
89.4
90
91.1
91.2
91.3
92.1
92.2
93.1
93.2
94
95
96.1
96.2
96.3
96.4
96.5
96.6
96.7
Inlet
Sample by
Balston Total
Filter at Inlet
0.40
0.31
0.30
0.32
0.30
0.60
0.60
0.48
0.48
0.37
0.37
0.47
0.40
0.40
0.57
N/A
N/A
1.22
1.22
0.53
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Particulate Loading (g/m )
Calculated from Calculated from
Mass Balance Mass Balance
Around 3rd Cyclone Around Filter
0.79 N/A
1.09 N/A
1.09 N/A
1.09 N/A
0.77 N/A
0.77 N/A
0.77 N/A
1.08 N/A
1.08 N/A
1.08 N/A
1.08 N/A
1.35 0.48
1.06 0.48
1.06 0.48
1.06 0.48
0.92 0.84
0.92 0.84
1.16 0.84
1.16 0.84
1.16 0.73
0.96 0.73
1.06 0.73
0.94 0.73
0.94 0.73
1.65 0.73
1.65 0.73
1.65 0.73
1.65 0.73
N/A = Not Available
                                  113

-------
                                         FIGURE IV-11

               ACUREX HIGH-TEMPERATURE CERAMIC BAG FILTER GAS INLET SCHEMATIC
   ACUREX
  CERAMIC
    BAG
    FILTER
                                      BYPASS GAS TO  SCRUBBER
                                          1.4-1 .9  Nm3/mm
                               ACUREX BAG FILTER
                               INLET PROBE (1" SCH. 80)
                               1 .9-3.4 Nm3/min
                               (200-300% ISOKINETIC)
KAMYR
VALVE
MOHAWK
 VALVE
1 .0 to 2.0  Nm /min
                    BALSTON FILTER
                     SAMPLE FLOW
                  0.034-0.056 Nm3/min
                                       FLUE GAS FROM
                                       2ND CYCLONE
                   VENT
                                            BALSTON
                                            FILTER
                                       X
                                        4 INCH SCHEDULE
                                             5 PIPE
                                WET TEST
                                 METER
                                                        •LUE GAS
                                                      16.6 Nm3/min
                                                   TO 3RD CYCLONE

-------
     A mass balance around the filter test vessel was attempted to resolve the
Inlet loading Issue.  Weight of the filter bags was  not determined before
exposure, so a tare of 0,2 kg was assumed  by weighing other unexposed bags.
These inlet loadings, intermediate to the  other two  results, can still be con-
sidered low because of the multitude of places where particulates could have
been lost during cleaning and dismantling  operations.  However, from these
three calculation methods a reasonable estimate can  be made of the actual
inlet particulate concentration.

     Size distribution of the inlet particulate matter was not obtained since
the samples were not taken under Isokinetic conditions.

     The bag filter outlet particulate concentration was determined by passing
the entire filtered gas flow through a Balston total filter.  The total par-
ticulate concentration was obtained by weighing the  total filter before and
after exposure.   Overall  ceramic bag filter efficiencies are shown 1n Table
IV-9.  These were calculated using the three methods of determining inlet
concentration discussed previously.  Despite some uncertainty  in  the  inlet
particulate concentrations, the collection efficiencies  calculated  by the
three methods were generally in good agreement.

     A size distribution of the outlet particulates  could  not  be  obtained.
The amount of particulates on the Balston filter  was so  low  that  Insufficient
material was available for Coulter Counter analysis.  The  filters were washed
off with Isoton II in an attempt to remove particulates  without mechanical
brushing.  This method caused enough Balston filter  material to be washed  Into
solution to obscure completely the flyash particulates.  A clean  Balston
filter, not exposed to any flyash but also washed with Isoton  gave  a  sample
which had a size distribution similar to that obtained from  a  used  filter.

     During the  tests at the miniplant, eight single and one double thickness
bags were exposed to PFBC conditions, as shown in Table  IV-7.  Most bags were
exposed for 6 hours or more.  Averaging the face  velocity and  exit particulate
concentration over the first 6 hours of new bag exposure and plotting  outlet
loading as a function of face velocity provided the  data shown 1n Figure IV-12.
Bag number 4 results were not recorded on  Figure  IV-12  because of problems with
the outlet filter.  Bag number 7 was a double thickness  bag  ("2 cm).   Normally
filter media thickness was about 1 centimeter.  Bag  7  was  physically  less dis-
torted and its pressure drop was less than bag 8  (a  1-cm thick bag which was
run at similar conditions), and its filtration efficiency was much lower.  The
reason for the lower AP and lower efficiency for  the double-thickness  bag is
not certain (it probably had a leak).  As  seen  in Figure IV-12, the parti-
culate penetration of a bag seems to increase with face  velocity.  The trend
lines shown on the curve were selected by drawing a  line through  the  Illinois
No. 6 results, which fell on a straight line, and then drawing a  suitable
parallel line for the Champion coal.  However, the data may  not be  as  precise
as implied by this curve.  For example, one objective  of the off-line cleaning
used in this study is to offset the increased penetration with velocity  that
is normally seen  in filter tests and suggested  by Figure IV-12.   More testing
1s still required to determine better the effect  of  face velocity and off-line
cleaning on penetration.


                                     115

-------
CT>
           Run  No.
                        TABLE IY-9.  ACUREX HTHP BAG FILTER COLLECTION EFFICIENCY
   Outlet
Particulate
  Loading
   (g/m3)
86
87.1
87.2
87.3
88.1
88.2
88.3
89.1
89.2
89.3
89.4
90
91.1
91.2
91.3
92.1
92.2
93.1
93.2
94
95
96.1
96.2
96.3
96.4
96.5
96.6(D
96.7(1)
0.0093
0.001
0.009
0.003
0.021
0.014
0.011
0.023
0.023
0.023
0.023
0.009
0.007
0.003
0.004
0.006
0.005
0.007
0.005
0.066
0.012
0.016
0.007
0.007
0.016
0.030
0.187
0.244
Collection Efficiency (%) Based
Upon Alternative Inlpt nptprmi nations
Sampled by
Balston Total
Filter at Inlet
97.6
95.2
97.1
99.1
93.0
97.7
98.1
95.3
95.3
93.9
93.9
98.1
98.3
99.1
99.4
__
..
99.4
99.6
87.6
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
Calculated from
Mass Balance
Around 3rd Cyclone
98.8
98.7
99.1
99.8
97.3
98.2
98.5
97.9
97.9
97.9
97.9
99.3
99.4
99.7
99.7
99.4
99.5
99.4
99.6
94.3
98.8
98.6
99.4
99.4
98.6
97.4
83.5
78.4
Calculated
from
Mass Balance
Around Fil
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
N/A
98.1
98.6
99.3
99.2
99.3
99.4
99.2
99.5
90.9
98.4
97.8
99.1
99.1
97.9
95.9
74.4
66.6
ter




























Average

  98.2
  97.0
                                                                                                98.
                                                                                                99.
                                                                                                95.
                                                                                                98,
                                                                                                98.
                                                                                                96.
                                                                                                96.
                                                                                                95.
                                                                                                95.
                                                                                                98.
                                                                           98.8
                                                                           99.4
                                                                           99.4
                                                                                                99.
                                                                                                99.
                                                                                                99.
                                                                                                99.
                                                                                                90.
                                                                                                98.
                                                                                                98.
                                                                                                99.
                                                                                                99.
                                                                                                98.
                                                                                                96.
                                                                                                78.
                                                                                                72.5
          (1)  Bag Failed
              N/A   =   Not Available

-------
                    FIGURE IV-12

  ACUREX HTHP BAG FILTER OUTLET LOADING VS. FACE
VELOCITY (AVERAGED OVER FIRST 6 HOURS OF EXPOSURE)

               FACE VELOCITY, (FPM)
0.06
CO
E
Z 0.05
CD
Z
< 0.04
O
LU
=1 0.03
u
Qi
°-
5 0.02
i—
O
0.01

5 10 lo zo
i i 1*1
7*
• CHAMPION COAL
O ILLINOIS NO. 6 COAL
8 BAG NUMBER
* DOUBLE FILTER THICKNESS BAG
•
•




3A^
9 3^^^
**~'&T^
^6
i i I i i i
123 456
0.030
0.025
u.
0)
o
•
o
ro
o
LOADING,
0.015 uj
D
y
fe
0.010 £
LU
5
O
0.005

               FACE VELOCITY,  m/min
                         117

-------
                          FIGURE IV-13



            ACUREX CERAMIC  BAG FILTER—BAG NO. 5

               PARTICIPATE PENETRATION HISTORY
    0.015
    0.010
z
o
u


o
u
LJJ
13

y  0.005

c*.

o_
O
                \
\RUN 90

      \RUN 91.1
         x


                        l
                          iii
                     l
       5             10


       BAG AGE (HOURS)
                                                    15
                            118

-------
      Outlet loading plotted in Figure IV-12 was  averaged  over  the  first  6
 hours of operation.  Outlet loading actually decreased as a  function  of  time
 in the same fashion that a conventional  filter media test would  show  in  similar
 tests under ambient conditions.  This decrease as  a  function of  time  is  shown
 as Figure IV-13.  This bag was exposed to Champion coal at 775°C for  a total
 of 13 hours.  Along with the decrease in filter  particle  outlet  loading  there
 was an increase in baseline pressure drop from 0.1 to 3.0 kPa  as expected as
 the filter cake formed and the efficiency improved.

      The effect of bag cleaning conditions  was not studied.  After adjusting
 cleaning conditions during the first two runs, conditions were fixed  for the
 rest  of the runs at 6  pulses of 1300 kPa air for about 0.75  s  each with a 3 s
 interval  between pulses.

      Examination of the filters after a  test showed  that  the dust cake was
 deposited mostly on the surface of the filter media  and the  dust cake could be
 removed easily.  Figure IV-14 shows bag  number 3 immediately after removal
 from  the filtration vessel.  Figure IV-15 is a close up of the same bag after
 a  strip was vacummed clean.  This  strip  had  the  appearance of a  virtually new
 bag indicating very little dust penetration  into the media.

      At the conclusion of the series of  short runs,  run 96,  a long continuous
 test  of the bag filter was attempted at  conditions deemed  optimum for extensive
 testing.   Filtration commenced smoothly,  however, the  baseline pressure drop
 across  the bag continued  to increase during  the  first  6 hours of filtration.
 A  possible  reason  for  this  increase  became clear at  the conclusion  of the run
 when  it was  discovered  that  the pressure  regulator used to set  the  pressure  of
 the reverse  flush  air was  set  only  slightly  higher than filter  vessel  system
 pressure.   It  is possible  that a slightly higher combustor pressure could have
 reduced reverse  flush air  flow to a  level too small to clean  the  filter
 effectively.   Since  the reverse flush air flow rate was not measured,  this
 hypothesis cannot  be confirmed.

     After 9 hours of filtration during run  96, a pressure drop of  over  12.5
 kPa across the  filter was  thought to be excessive for continued bag life.  The
 vessel pressure was  reduced to 300  kPa and a full high pressure  (1300 kPa)
 pulse blow back was  initiated  into the lowered system pressure.  This  blow
 back reduced the pressure drop almost back to the clean baseline  condition
 (0.5 kPa).  However, outlet particulate concentration increased over the  next
 10 hours until   it was almost identical to the inlet concentration.   The  run
was terminated  at that point.  As expected, the bag had failed  (see Figure
 IV-16).  The bag failure probably began with the  high pulse pressure blow back
against the low system pressure, and was made worse by subsequent blow backs.
The final cause of the failure appeared to be high  temperature  corrosion  of
the thin 304 stainless  steel filter support screen.   The blown  out  appearance
was probably caused by the high pressure pulse into the lowered system pressure
environment after the corrosion weakened support  screen failed.

     It is recognized that corrosion of a metal support screen  is a  potential
problem in long term applications.   The metal support screens used  in  these
tests  were used only for ease of construction.  In  addition,  the  early failure


                                      119

-------
       FIGURE IV-14
CERAMIC FILTER BAG NO. 3
           120

-------
  Vacuumed Strip
          FIGURE IV-15



CERAMIC FILTER BAG NO. 3 CLOSEUP

       OF VACUUMED STRIP
             ,1
 I •   ' I ••
         *'
   v,
 r,,Jh,v'
                  Ere
                  •H
                    \ I
                        121

-------
              FIGURE IV-16
CERAMIC FILTER BAG NO.  9  AFTER  RUN 96

                   122

-------
 of the  support  screen experienced in run 96 was probably caused by a malfunc-
 tion  in the  temperature control system which caused the filter element to be
 overheated.   Flexible woven ceramic fabric screens are available to perform
 this  function for applications requiring greater corrosion resistance.

 Summary and  Conclusions

      During  the miniplant tests, baseline pressure drops of under 2 kPa were
 maintained for over 6 hours average duration at face velocities of up to 6.0
 m/min with removal efficiencies of 95 to 99%.

     The average particulate concentration at the bag filter outlet was measured
 to be 0.013  +_ 0.005 g/Nm3.  This is less than one-half the EPA New Source Emis-
 sion Standard for particulates (0.03 Ibs/MBTU).  The particulate concentration
 was as  low as one order of magnitude below the emission standard.  In compar-
 ison to 3 stage cyclone cleanup, the average improvement was approximately a
 factor of 7.  The particulate concentration with the ceramic filter easily
 meets both the EPA and most gas turbine standards,

     These tests proved the ceramic filter was cleanable while subjected to
 flyash generated under PFBC conditions.  High collection efficiency at high
 face velocity was also shown.  In general, the test filter exhibited perfor-
 mance similar to that which would have been expected from a filter unit opera-
 ting under more common conditions.

     The failure of the filter element in the final test was a result of an
unusual  combination of events and is not an indication of an inherent filter
 problem.  Further investigation of this potentially promising technique 1s
felt  to be warranted.

 GRANULAR BED FILTRATION STUDIES

     Granular bed filters were also evaluated as high efficiency particulate
control  devices.  Originally, the objectives  of the test program were to measure
 particulate removal  efficiency, operational  stability and long term life of the
 filter hardware.  However, operating difficulties, primarily caused by plugging
of the filter inlet sections by the particulates,  poor bed cleaning and loss
of filter media during cleaning caused a change in the program direction.
The program was then directed toward modifying the granular bed filter system
to overcome the operating problems.  The previous  report describes the filter
systems studied, the problems encountered and modifications made to solve the
 problems (1).  These modifications allowed a  24 hour run to be completed.
 However, filtration efficiency decreased sharply during the test.  It was
 found that the filter beds had not been adequately cleaned and gas was blowing
 through the filter beds in "rat holes."  Additional modifications were sub-
 sequently made and the modified systems tested in  an attempt to improve perfor-
mance.  These latter tests are described in  this report.

Equipment

     The filter configuration used at the start of the test program discussed
 in this report was described in detail  in the previous report (1).  This system

                                     123

-------
was installed on the miniplant following the second cyclone,  and  was  capable
of accepting the total  flue gas flow.   It consisted of two filter elements,
each composed of five filter beds, as  shown in Figure IV-17.   A schematic  of
one of the filter beds  is shown in Figure IV-18.   The filter  media contained
in each bed consisted of granular alumina or quartz.  This design had no
screens across the dirty gas inlet opening at the top.  This  was  done to pre-
vent inlet screen plugging.  The freeboard height was also increased  to 18 cm
to prevent loss of the filter media during the reverse (upflow) cleaning step
with clean compressed air.  In the cleaning step, the beds are fluidized by
reverse-flow air, the fine particulates retained  by the bed from  the  previous
filtration step are entrained from the filter beds and blown  out  of the inlet
opening at the top of the beds in a reverse direction.  Figure IV-19  is a
schematic which illustrates the operation of a filter element during  the
filtration and blow back (cleaning) steps.  The fluidizing grids  at the bottom
of each bed were also redesigned to give better gas distribution  during blow
back.  Each grid contained 56, 0.5 cm diameter holes.

     Each filter element was installed in a shroud as shown in Figure IV-19.
Flyash blown from an element during blow back is  retained within the  shroud,
falls to the bottom and is collected in a lock hopper.  Two filter elements
were installed in parallel after the second stage cyclone in a single large
refractory lined pressure vessel.  The inside dimensions of the vessel are
approximately 2.4 m diameter by 3.4 m inside height.  The vessel  is connected
to the second stage cyclone by 12 inch pipe refractory lined and sleeved  with
stainless steel to an inside diameter of 10 cm.

     Dirty gas entering the filter vessel is piped to each shroud, passing
through orifices which measure flow rate to each  filter element.   Clean gas
leaving each shroud fills the interior of the pressure vessel before  leaving
through a single outlet line.  Blow back air enters each filter element
through flanges at the top of the pressure vessel and flows in a reverse
direction through each filter element.  A blow back nozzle and seal plate  are
dropped down to engage the top of the filter element during blow  back and
direct the blow back gas in the proper direction.  This is shown  in Figure
IV-19.  The blow back gas leaving a filter element flows in a reverse direc-
tion through the clean gas inlet system into the  other filter element which
is in a filtration step.  Each element is blown back separately.

     A natural gas burner was installed to preheat the interior of the pres-
sure vessel to a temperature above the dew point of the combustor flue gas
before starting a filtration test.  The burner fires into the vessel  through a
side port for an 8 to 12 hour period prior to the start of a run.  The filter
vessel is at atmospheric pressure during this period.

     Natural gas injection into the flue gas piping between the second stage
cyclone and the filter vessel was used to keep the flue gas temperature above
840°C.  Gas was injected at four  points, 2.3 m, 6.7 m, 9.8 m and 14.6 m down-
stream of the second stage cyclone.  Approximately 0.03 Nm3/m1n of natural gas
was  injected through each of the  four inlets.

     All of the above systems are described in more detail in  the  previous
report  (1).

                                      124

-------
             FIGURE IV-17
MODIFIED GRANULAR  BED FILTER ELEMENT
          (WITHOUT  SHROUD)
                125

-------
    FIGURE IV-18
MODIFIED FILTER BED
tMspmw
DIRTY
C, A ci ••—•
INLET

1

— 	
______ 	 _

1





                    ._ ____.
        CLEAN GAS
          OUTLET
                                     l  1 .3 cm

                                       t
                                       21.9cm
FLUIDIZING GRID
 "(56-^.36 cm
   i did. holes)
       -7.3 cm
       21 .9cm
           126

-------
                         FIGURE IV-19
                GRANULAR BED FILTER SCHEMATIC
        FILTRATION CYCLE
BLOWBACK CYCLE
FILTER MEDIA
SHROUD
             Ly/y>
                         CLEAN
                         GAS EXIT
                      DIRTY GAS
                       FIVE BEDS

!f
1
7
—
8*
M
[be
u
•-
C
-1
11
C
»
i
:
)
i
j


i
j
OB
M
aa
J
ml
FLY ASH,
F-
'*^ DIRTY GAS
FLUIDIZED
FILTER MEDIA
LOCK
HOPPER

LOCK
HOPPER
                             127

-------
System Modifications--
     In order to improve the operation  of the  filter a  few modifications were
made to the filter system.   The  modifications  were  used at various times dur-
ing the test program.  None of them were permanent  or  needed  for all tests.
The section dealing with test results will  specify  which modifications were
used for the various runs.   This section will  describe  the modifications.

     Blow back gas ejector--A blow back gas ejector was installed  to replace
the blow back gas inlet nozzle and seal  plate  assembly.  It was described in
the previous report (1).  Briefly, the  ejector used a  small amount of motor
air during blow back, but pulled in additional  clean gas  by the action of the
ejector to supplement the blow back gas.  There were no moving parts and no
sealing surfaces.  The ejector was tested during two runs and removed.

     Inlet screens--In the above sections describing the  design of the filter
it was mentioned that the inlet screens were removed to prevent plugging and
the bed freeboards increased to prevent loss of filter media  during  blow  back.
However, loss of filter media still occurred despite the  increased bed  free-
board and attempts to minimize blow back gas velocity.  Therefore, for most
runs, screens were strapped over the inlet slot.  Three screen mesh  sizes were
used, 10 x 10, 20 x 20 and 50 x 50.

     Baffles—Another modification was  made to prevent loss of  filter media,
in this case, by the installation of baffles in the filter  beds  near the  gas
inlet.  Figure IV-20 is a sketch of the baffle arrangement.   The  baffles were
tested in one run.

     Filter media —Initial tests were carried out using quartz  granules with
a  particle size range of 250 to 600 ym.  In subsequent tests, 850 to 1400 urn
alumina particles were used.  A very dense iron oxide   (speculite) was also
tested in two runs  in an attempt to minimize filter media losses.  The  specu-
lite particle size range was 400 to 2000 ym.

     Outlet piping--Initially. the clean gas from the   filter  elements  filled
the interior of the pressure vessel and left through a single vessel outlet
line.  However, heat losses  from the vessel were high  and the temperature
drop between the filter inlet and outlet lines was too great.  The filter
element outlets were then piped directly to the vessel outlet and the new
piping and the filter shrouds were insulated to reduce heat  loss.

Ambient Temperature Models —
     Because of various operating  problems, especially loss  of filter media
and poor  bed cleaning,  transparent plastic models of the filter beds were
built  and  tested.   All  the models  operated at  ambient  temperature and close
to atmospheric  pressure.  The first model was  simply a plastic pipe, 2.4 cm
in diameter.  The  pipe  was  filled  with  granular filter media, dust was added
manually  and  the  bed  blown  back with compressed air.

     A two-dimensional  plexiglas model  of  a granular bed filter was also
built  and  tested.   A  schematic  of  the model is  shown in Figure IV-21.  The
system was  designed to  operate  in  both  the  filtration  mode and blow back
mode  although  it was  only used  to  study blow  back  operations.  The model

                                     128

-------
                    FIGURE IV-20




              FILTER BED WITH INTERNAL BAFFLE
I / / / / /// / 77
BAFFLES
vr//// ///n
///////
1.3cm-*| [««-
j

i

WA
V //////.
/ /




i
r
u
i™
}
/ /
/ \


1
"^
^
r
i-
=
j
s
*
•>
:,
n



•

=
D
S
•

                             6.9 cm
                    129

-------
                                             FIGURE IV-21

                                         PLEXIGLAS MODEL GBF
CA>
O
                                               FILTRATION
                                                 VENT
                                 BLOW BACK
                                    VENT
                                              FILTER
3-WAY
VALVE
       IROTAMETER

          SOLENOID
           VALVE
                                                                   /^PRESSURE GAUGE
                                                                   T




.ATE
TOR









i^^^mmmmm


\

w,

1
jr
*
M 1
FILTRATION
FLOW SHUT-OFF
VALVE


1

















—
W

—
m










PL!
—
M

—
VA


:X







GLA.

-^~~





             BACK PRESSURE
            FLOW CONTROL
                VALVE
IROTAMETER
                                                                                     BLOWBACK
                                                                                       AIR
                                                                           PRESSURE
                                                                          REGULATOR

-------
 consists  of a  two  dimensional slice of  two filter beds each with a filtra-
 tion area of 0.01  m2.   Each  filter bed  is 23 cm high with two 1.3 cm inlet
 slots.  The filter media  is  supported by a 50 x 50 mesh screen supported by
 a  fluidizing grid.  Dirty gas enters the filter through two inlets, each feed-
 ing  one side of the  filter beds.  Filtered gas exits the filter through a
 5  cm central core  and is  passed through a backup filter to allow an estimation
 of collection  efficiency  to  be made.  Blow back air enters through the central
 core and  is  distributed to the filter beds.  After passing through the filter
 beds, the blow back  air is then vented  to the atmosphere.

     Another simple  plastic model was built.  This was 7.6 by 7.6 cm in cross
 section and  7.6  cm high.   The chief feature of this unit is the design of the
 blow back system.  Blow back air can be added through the bed support grid as
 in the  other units,  but it can also be added through a ring sparger, located
 at the  top of  the  bed along the wall.  The blow back air from the sparger is
 directed  horizontally and  was intended  to shear the dust layer from the top
 of the  bed without disturbing the filter bed itself to a large degree.

     A  second  high temperature/high pressure filter was built and tested.
 This was  a much  smaller unit operated on a slipstream of flue gas withdrawn
 after the  second cyclone.  This unit and results of the tests are described
 in a later section.

 Experimental Results

     Six  runs  were made on the larger filter system during the reporting
 period.   A summary of run  conditions is given 1n Table IV-10.  The changes 1n
 filter  configuration described above were made during the test program.  In
 addition,  blow back conditions were varied to examine the effect of blow back
 gas velocity,  duration and frequency on cleaning of the filter beds.  At the
 conclusion of  the  six runs, an attempt was made to use the granular bed filter
 in a 100  hour  test as part of the DOE sponsored program on hot corrosion and
 erosion of gas turbine materials.

     The experimental program was not  successful.   Three major problem  areas
developed which were not satisfactorily resolved:

     •  low particulate outlet concentrations  could not be maintained
        for more than a  few hours,
     •  loss of filter media  could not be  prevented,
     •  serious operating problems persisted  to  the end of the program.

 Outlet  Particulate Concentration and Size Distribution--
     Outlet  particulate concentrations as low as 0.07 to  0.11  g/Nm3 were
 measured  in  several runs.  However, in all  cases,  the concentrations Increased
 during  the runs as much as a factor of three In  8  to 10 hours.  Results from
 two  typical  runs are given 1n Figure IV-22.   It  is believed  that this  poor
 removal  efficiency and the increase in particulate concentration with  time
 were due  to  poor dust removal from the beds  during blow back.   This was suppor-
 ted  by  the observation that the filter beds,  after completion of a run, were
 generally  found to contain a very high concentration of dust intimately mixed

                                     131

-------
                              TABLE IV-10.   GRANULAR BED FILTER  RUN  SUMMARY
GO
ro
Run No.

Filter Description

Number of Parallel Elements
Number of Beds/Element
Retaining Screen
Miter Media
Bed Depth (cm)
Filter Media Part. Size (ym)
Other Information
Operating Conditions

Preheat Temperature (°C)
Filter Inlet Temperature (°C)
Filter Outlet Temperature (°C)
Filter Inlet Pressure (kPaa)
Filter Face Velocity (m/s)
Filter AP After Blow Back (kPa)
Filter AP Before Blow Back (kPa)
Run Length (hrs)

Blow Back Conditions

Superficial  Velocity (m/s)
Duration (s)
Interval  Between Blow Back (min)

Partlculate 'Emissions

Outlet Partlculate Concentration (g/Nm )
                                                                60
                                                                 2
                                                                 5
                                                              50  Mesh
                                                              Quartz
                                                               6.4
                                                              250-600
                                    61
                                     2
                                     5
                                  50 Mesh
                                 Alumina
                                    6.4
                                 850-1400
                                                      Ejector  on  One  Element     Ejector on One Element
  Run Unsuccessful
Because of Poor Flow
Distribution and High
   AP Measurement
       760
       840
       690
       900
0.26 and 0.47
  9.6 and 30({
  24 and 450)
       5.5
                                                                                         0.61
                                                                                           8
                                                                                         10-15
                                                                                          1.1
                                          (1)
                                         0)
         (1)   Values  for  the  two  ftlter elements are different due to unequal flow split.

-------
                        TABLE IV-10 (CONT'D).  GRANULAR BED FILTER RUN SUMMARY
CO
CO
Run No.
Filter Description
Number of Parallel Elements
Number of Beds/Element
Retaining Screen
Filter Media
Bed Depth (cm)
Filter Media Part. Size (urn)
Other Information
Operating Conditions
Preheat Temperature (°C)
Filter Inlet Temperature (°C)
Filter Outlet Temperature (°C)
Filter Inlet Pressure (kPaa)
Filter Face Velocity (m/s)
Filter AP After Blow Back (kPa)
Filter AP Before Blow Back (kPa)
Run Length (hrs)
Blow Back Conditions
Superficial Velocity (m/s)
Duration (s)
Interval Between Blow Back (min)
Particulate Emissions
Outlet Particulate Concentration (g/Nm )
                                                     62.1
                                                       2
                                                       5
                                                    10 Mesh
                                                    Alumina
                                                      6.4
                                                   850-1400
                                                      0.30
                                                        8
                                                       15
                                                      0.069
  62.2
    2
    5
 10 Mesh
 Alumina
   6.4
850-1400
760
830
700
900
0.41
14 to 20
35
2
__
830
700
900
0.41
20 to 24
35
2
  0.30
    4
   10
  62.3
    2
    5
 10 Mesh
 Alumina
   6.4
850-1400
                830
                700
                900
               0.41
                24
                35
                 2
  0.30
    2
   10
               0.48
   63
    2
    5
  None
Speculite
   1.9
425-2000
                   760
                   900
                   800
                   900
                  0.47
                22 to 33
                   41
                   12
  0.46
    8
    6
             0.11,0.16,0.27

-------
                         TABLE IV-10 (CONT'D).  GRANULAR BED FILTER RUN SUMMARY
CO
Run No.
Filter Description
Number of Parallel Elements
Number of Beds/Element
Retaining Screen
Filter Media
Bed Depth (m)
Filter Media Part. Size  (ym)
Other Information
Operating Conditions
Preheat Temperature (°C)
Filter Inlet Temperature (°C)
Filter Outlet Temperature (°C)
Filter Inlet Pressure (kPaa)
Filter Face Velocity (m/s)
Filter AP After Blow Back (kPa)
Filter AP Before Blow Back (kPa)
Run Length (hrs)
Blow Back Conditions
Superficial  Velocity (m/s)
Duration (s)
Interval  Between Blow Back (min)
Partlculate Emissions
Outlet Particulate Concentration (g/Nm )
                                                          64
                                                           2
                                                           5
                                                         None
                                                       Speculite
                                                          6.4
                                                       425-2000
     760
     950
     840
     900
    0.36
10 and 14
22 and 26
      7
                                   65
                                                                                          (2)
                                                               ,
                                                              U
                                                         0.82
                                                        4 to 5
                                                        3 to 5
                                                    0.64,0.66,0.62,0.27
                                    2
                                    3
                          20 Mesh (One Element)
                                 Alumina
                                   6.4
                                850-1400
                   Baffles On One Element (No Screen)
   760
   850
   800
   860
  0.51
 7 to 34
34 to 62
   8.5
                                  0.61
                                   4
                                   4
                                0.11,0.14
      (1)   Values for the two filter elements are different  due to unequal  flow split.
      (2)   A third cyclone was installed  between  the second  cyclone and  the filter  vessel during  this
           run,  thus reducing the inlet particle  concentration  to the filter

-------
                              FIGURE IV-22


       INCREASE IN OUTLET PARTICULATE CONCENTRATION WITH TIME
w
 i_
&
z
O  0.12
U

O   0.08
u
3
ID
y
t   0.04
ID
O
       0
       0
             a MINIPLANT RUN 63

             O MINIPLANT RUN 65
                                                        D
                                               O
     EPA EMISSION  STANDARD

   j	1	i     '      '
    4          6

TIME INTO RUN (MRS)
                                                   8
                                                               0.3
                                                               0.2
                                           O
                                           c
                                           70
                                           ^H
                                           n
                                           c
                                           n
                                           O
                                           Z
                                           n
                                           m
                                           Z
                                      0.1   5
                                                                   I
                                                                     CO
10
                                135

-------
with the filter media.   Dust concentrations  of 10 to 3Q% were  typically mea-
sured.  Tests made in the transparent models showed that the dust  was  only
partly removed during blow back, some of it  adhered to the filter  media and
was mixed into the bed during blow back by the motion of the fluidized filter
media granules.  The fine particles present  in the PFBC flue gas are apparently
very adhesive and are not removed easily during blow back. This mixing pro-
cess probably explains the observed uniform  distribution of the fine parti-
culates throughout the filter beds after extended runs.  Once  the  fine parti -
culates were mixed into the filter media, they could be carried out through the
bed support grid during the filtration step  because of locally high gas veloc-
ities at the grid holes.  As seen in Figure  IV-22, at no time  were outlet con-
centrations measured which meet the current  emission standard  of 0.03  Ib/MBTU.

     Attempts were made to improve bed cleaning by blowing back more  frequently
for longer periods and at higher velocities.  The blow back velocity was
limited by loss of filter media (with alumina or quartz media) to  0.3  to  0.6
m/s.  The duration of the blow back was usually 4 to 6 s, although durations  as
short as 2 s and as long as 8 s were tested.  The frequency of blow back  was
reduced from once every 15 min. to once every 4 min.  As a result  of  these
changes, the degree of filter cleaning appeared to improve, but the best  run
made, run 65, still showed an increase in outlet particulate  concentration  with
time.

     Another factor which could be responsible for the retention of particulates
in the filter beds was the recycling of particulates from bed  to bed  during the
blow back step.  The shrouds which surround  each element may impede the settling
into the lock hoppers of the particulates blown from the elements during  blow
back.  That  is, if a new filtration step began before the particulates fell
into the hoppers, the particulates could get carried back into the filter ele-
ments.  This could result in high internal recirculation rates between elements,
causing the  particulates to  build up in the beds and thereby causing an
increase in  the outlet particulate concentration.

      In discussions with representatives of Ducon Company, the original  sup-
plier of the filters, Ducon  claimed that the filtration velocities used in
these tests  were excessively high and could have promoted dust penetration
(12).  Ducon suggested a velocity of about 0.2 m/s.

      Regardless of the reasons, the outlet concentrations measured in  these
tests were consistently  higher than those measured when a third stage cyclone
was  used instead of the  filter.  Therefore, the filter showed no efficiency
advantage over  the cyclone.  It should also be mentioned that the best run in
the  series of six shown  in Table IV-10, run 65, was made with a cyclone
installed between the second stage cyclone and the filter.  The cyclone was
not  designed for the flow rate used in run 65 and was  probably not too effi-
cient, but 1t still probably contributed to the improved performance observed
in  that  run.  The cyclone was later removed and replaced with a more efficient
cyclone which was used for the tests described in the  earlier section on
cyclone  performance.
                                       136

-------
      Filtration  efficiencies were not measured during the test program since
 Inlet partlculate  loadings could not be measured.

      Typical  particle size distributions of the dust entering and leaving the
 granular  bed  filter are given in Table IV-11.  Additional particle size data
 after the  filter for some of the filter runs are presented 1n Appendix M-6.

                   TABLE IV-11.  PARTICLE SIZE DISTRIBUTIONS

                             Percent Less Than Particle Size (ym)
                        5%     10%    25%    50%    75%     90%    95%

       Filter Inlet     1.5    2.3    3.5    6.9    13      32     45

       Filter Outlet    1.1    1.4    2.0    3.0     5.1    12     15


      The outlet size distribution is relatively coarse when  compared  to typical
distributions measured from the third stage cyclone.  The cyclone consistently
produced participates with a median size of 1 to 2 ym compared to 3 ym shown
in Table IV-11.  The coarser particulate distribution 1n the filter outlet was
also  probably caused by dust penetration resulting from poor bed  cleaning.

Loss  of Filter Media--
      The loss of filter media during blow back continued throughout the program
and was never satisfactorily prevented.  Screens were used in most runs to pre-
vent  loss, but the screens consistely plugged with flue gas  partlculate even
though screen opening sizes as large as 10 mesh were used.  The material
plugging the screens was a loose agglomerate of dry, fine particles which
could be removed by brushing.  A photograph of a plugged screen was given in
the previous report (1).  Small  openings above the screens and below  the top
flanges of each filter bed were used to allow entry of dirty gas  and  exit of
blow  back gas.  This arrangement minimized but did not prevent loss of filter
media.  Alumina, 850 to 1400 ym, was used in all  runs after  60 except 63 and
64 in which 425 to 2000 ym speculite was used.  The screens  were  removed in
runs  63 and 64 since it was believed that the very dense and coarse speculite
would not be lost from the beds.  However, loss of the filter media stm
occurred.   In addition, 1n run 64, a temporary upset caused  an Increase in the
CO concentration in the flue gas.  This apparently reduced the speculite (iron
oxide) to Fe° which was then oxidized after the CO concentration  decreased.
This  is a highly exothermic reaction.  As a result, the filter was severely
damaged.  Two filter beds could not be repaired.   Fine Iron  oxide dust was
also  found 4n the filter outlet piping sections.   The use of speculite was
discontinued.

      Testing of the two dimensional  transparent model  then began  1n an attempt
to determine the cause and possible cure for the filter media loss problem
A series of tests were conducted in which various internal baffles and inlet
gas slot sizes were used.  The most significant finding from these tests  was
that  most of the filter media loss occurred at the beginning of the blow back
The sudden opening of the blow back valve caused  a surge 1n  gas flow  which blew
filter media from the vessel.  The baffles did not prevent loss caused by the


                                     137

-------
flow surge although some designs  did  decrease  losses caused  by high blow back
velocities following the initial  surge.   Baffles were  installed  in one filter
element and tested in run 65.   Screens were removed  from  this filter element
and retained on the other element.  Both elements  showed  loss of filter media,
the baffled element without screens showed  the greater loss, 32% of the start-
ing media bed compared to 18%  from  the second  element.  The  use  of unscreened
baffled filters was then discontinued.

     The plexiglas model tests would  indicate  that slower acting blow back
valves or surge capacity between  the  valve  and the filter would  decrease the
loss of filter media caused by the  flow  surge.  Ducon  representatives also
claimed that filter bed freeboards  50% larger  than those  used in these tests
are needed to prevent loss during blow back (12).

Other Operating Problems--
     In addition to the problems  described  above,  other operating problems
occurred which persisted until the  end of the  program  despite attempts to
improve operations.  One problem  was  an  increase  in  the combustor pressure
which occurred during filter blow back.   At times, the increase  was  large
enough to cause the combustor  pressure to exceed  the pressure in the coal
injection vessel.  When this happened, the  hot solids  in  the combustor would
back up into the coal injector vessel, causing a  fire  in  the vessel.  A
nitrogen purge system on the coal injector  was used  to extinguish the fires,
but these incidents always caused the termination of the  run and removal of
some charred coal from the Injector vessel.

     Temperature drops across  the filter was another problem.  Although  this
did not affect performance of the filter, it did cause problems  for  the  gas
turbine materials test.  Flue gas temperatures were  required to  stay above
840°C at all times to prevent condensation  of alkali sulfates,  the material
which could cause hot corrosion of the  turbine parts.   Injection of  cold blow
back air and heat losses from the filter elements to the  interior of the pres-
sure vessel resulted in temperature drops in excess  of those required.   Injec-
tion of natural gas into the filter outlets, insulation of the  filter  shrouds
and piping of the clean flue gas  from the filter outlet directly to  the  pressure
vessel outlet decreased the temperature  drop to an acceptable  level.

     An attempt was also made to replace the blow back nozzle  and seal  plate
with an ejector.  This would decrease the amount of cold  blow back  air needed
and also replace the seal  plate which occasionally did not work satisfactorily.
An ejector was installed on one filter element and tested in runs 60 and 61.
The second element was equipped with the nozzle and  seal  plate arrangement.
In the tests with the ejector, it was found that the pressure  drop  through  the
ejector during the filtration step was  high, giving  rise  to unequal  flow rates
between the two filter elements.   The element with the ejector  was  not cleaned
as well during blow back and the outlet loading from the  test  (run  61)  was
very high, 1.0 g/Nm3.  The ejector was  removed and not tested  further.

     Poor bed cleaning also led to high pressure drops across  the filter.   At
the beginning of a run, the pressure drops  were usually around 14 to 20 kPa
after bed blow back, increasing to about 28 to 35 kPa before blow back.   The
                                      138

-------
 pressure  drops usually increased as the run progressed, indicating poor bed
 cleaning.   Pressun drops of 28 to 35 kPa after blow back and 40 to 60 kPa
 before  blow back were measured after 8 to 15 hours of filter operation.

     Another intrinsic problem with the filter was the fact that essentially
 all problems were non-corrective and as such, required termination of the run,
 a long  wait to cool down the filter, followed by a long period to remove,
 clean and replace the filter elements.

     The final  run made with the filter was an attempt to run a 100 hour gas
 turbine exposure test (run 66).  In this case, three filter elements were used
 in parallel, each with three filter beds.  The filters were charged with
 alumina filter media (850 to 2000 ym).  The inlet slots were covered with 20
mesh screens except for a 3 mm split above the screens to allow gas flow if  the
 screens plugged.   Filter blow back occurred every 4 min. at 0.3 m/s for 6 s.
These were judged to be the best conditions to assure a successful  test. Over
the first four hours of operation, a gradual increase in the filter AP was
 noted which then increased sharply to 207 kPa.  The blow back velocity and
duration were increased to improve bed cleaning.  The AP dropped to 70 kPa,  but
a significant amount of filter media was lost.  As a result of the unstable
 pressure drops and frequent high velocity blow backs, the combustor pressure
control was upset and hot solids back flowed into the coal  injector causing  a
fire.  The run was terminated after 17 hours operation.  After the run, it was
 found that the high pressure drops were caused by partly blocked filter support
 screens.  At this point, the granular bed filter program involving the use of
the multi-bed filter elements contained in the large pressure vessel was ter-
minated.

     During the test program, a series of tests were also made with the trans-
 parent models in an attempt to determine the cause of the poor bed cleaning.
 Some of the test results were described in a previous section of this report.
 They showed how a portion of the dust particles adhered to the filter media
 and were mixed into the filter bed by the motion of the fluidized filter media
 granules during blow back.  In order to prevent this, a modified blow back
 scheme  was  tested.  In this case, most of the blow back air entered the bed
 through a ring sparger located at the interface between the filter media and
 the dust layer.  Blow back air was directed horizontally across the interface.
 This resulted in a shearing action which blew off the dust layer without
 disturbing  the filter media.  Blow back upwards through the filter media was
 minimized as was the mixing effect caused by it.  This technique looked pro-
 mising  enough to be tested under PFBC conditions.

     In order to test this as well as other concepts which could improve the
 filter  performance, a program was planned to test a modified filter on a flue
 gas slip  stream.  The slip stream tests were to be made with a small, single
 bed filter  incorporating a number of design modifications.  The design of  the
 new unit  would build on the lessons learned in the previous test program and
 would  attempt simply to show if the granular bed filter concept was feasible
 for pressurized fluidized bed combustion applications.
                                      139

-------
Granular Bed Filtration Slip Stream Test

     The single element granular bed filter device was  tested on a minlplant
slip stream during run 115.  The filter element  was  a further modification
of the "modified Exxon filter element"  described 1n  Figure  IV-18.  The  loca-
tion and Instrumentation of the test vessel which contained the filter  element,
are almost Identical  to those of the Acurex high temperature ceramic  filter
vessel.  This was done to allow a fair  comparison of the  two high efficiency
hot gas cleanup devices.

     The filter element tested was named  Exxon Mark  IV  to distinguish it  from
earlier designs.  The filter element body was the same  one  used in full scale
tests.  Three things  were changed:

     1.  The fluldizing grid was replaced with a support  grid with 1  cm
         diameter holes, in order to reduce AP and reduce the velocity
         of the grid  jets during below-bed blow  back.

     2.  A fluidizing coil was installed  just above  the support grid.
     3.  An above-bed blow back coil was  installed.

These three modifications were intended to lower the baseline pressure drop,
reduce bed loss during blow back and improve the effectiveness of  the blow
back.  Both coils have 0.84 mm holes every 1.6 cm along the coll.   In the
below-bed coil, the holes are oriented  approximately 45°  from the  vertical
to provide upward fluidizing jets.  In  the above-bed coil,  the  holes  are  hori-
zontal  to blow the dust cake off the surface of  the  filter  media by a shearing
action without disturbing the filter media granules  to  a  great extent.  In
this way the superficial blow back velocity also increases  just above the bed
to blow more particulate and less filter  media out of the bed.  Air could
still be added below  the support grid to  aid in  fluidization of the bed.  A
schematic of the element 1s shown in Figure IV-23.

     The test vessel  which contained the  Exxon Mark  IV  element, was installed
on an isokinetlc slip stream taken from a point  between the second and third
cyclones.  A schematic of the installation is shown  in  Figure  IV-24.   The
valves and electronic timers are the same ones used  for the Acurex ceramic
filter tests.  The high pressure (1300  kPa) air  supply  is used on  the above-
bed coil for a pulsed blow back.  The lower pressure air  supply  (1000 kPa)
1s used 1n the in-bed coll and for the  below-bed blow back  air.   Particulate
samples were taken before and after the ganular  bed  filter  using  the  Balston
filter method.

Results  and  Dlscussions--
      The Exxon  Mark  IV  filter was  run  for 4  hours during run 115.  Before
 filtration  was  begun, the  inlet  lines  were preheated with  flue gas to a  tem-
 perature above  400°C  to prevent  condensation 1n  the line.   The filter could
 not be preheated.  The  filter  element  was charged to a 7 cm bed depth with
 -7 kg of -8+14  U.S. mesh alumina.   The blow back cycle was  program 1  in  Table
 IV-12.  The filtration  period  was  5 m1n  with a  20 sec cleaning cycle.  After
 the first 45 m1n it  became apparent that the pressure drop could not be  con-
 trolled with this  blow back program.   At this time,  the  cleaning cycle was


                                     140

-------
                              FIGURE IV-23

                  EXXON MARK IV GRANULAR BED FILTER
                              (SINGLE BED)
DIRTY
GAS
INLET
  IN-BED
BLOW BACK-
   COIL
               CLEAN GAS OUTLET
                                  t
                                                            ABOVE-BED
                                                            BLOW BACK
                                                               COIL
SUPPORT
 GRID
                             BELOW-BED
                             BLOW BACK
                                141

-------
                                  FIGURE IV-24
                   GRANULAR BED  FILTER INSTALLATION SCHEMATIC
FROM 2ND_
CYCLONE
     TO
3RD CYCLONE
                       -^-PREHEAT VENT
                                 •—t><-j	-J<	
                                           1300 kPa AIR

                                        950 kPa AIR
-•<	
                                                                             BLOW BACK
                                                                              SHUTOFF
                                                                               VALVE
                                                  WATER
                                                KNOCKOUT

-------
           TABLE IV-12.  GRANULAR BED FILTER CLEANING PROGRAM
          Program 1

Event                                                        Time (m:s)
  0       Forward Flow Off, Reverse Flow On                     0:00
  1       Above Bed Pulse (1.5 sec)                             0:02
  2       Above Bed Pulse (1.5 sec)                             0:07
  3       Above Bed Pulse (1.5 sec)                             0:14
  4       Reverse Flow Off, Forward Flow On                     0:19
  5       Reset To Start Cleaning Cycle                         5:19
          Program 2

Event                                                        Time (m:s)
  0       Forward Flow Off, Reverse Flow On                     0:00
  1       Above Bed Pulse (1.75 sec)                            0:01
  2       Above Bed Pulse (1.75 sec)                            0:05
  3       Above Bed Pulse (1.75 sec)                            0:10
  4       Above Bed Pulse (1.75 sec)                            0:14
  5       Above Bed Pulse (1.75 sec)                            0:17
  6       Forward Flow On, Reverse Flow Off (Filtration)         0:19
  7       Reset To Start Cleaning Cycle                         5:19
        TABLE  IV-13.  COMPARISON OF PARTICULATE SIZE DISTRIBUTION
            OF MATERIAL BEFORE AND AFTER FILTER TEST ELEMENT
     Sample

 Before Filter

 After Filter

       Test 1

       Test 2


Parti cl
e SI
Vol . % Finer
5id
1.4
1.4
1.4
10%
1.7
1.7
1.7
25%
2.4
2.2
2.4
bu%
3.1
2.8
3.0
ze (ym)
Than Size
75%
4.2
3.8
4.0


90%
5.6
5.1
5.4


95%
6.4
6.4
6.2
                                  143

-------
changed to program 2 1n Table IV-12.   After  3  hours,  the  filter was  isolated
from the miniplant and the Balston total  filter which follows  the  granular
bed filter test vessel was replaced.   This Balston sampling  filter was  com-
pletely destroyed by the high temperature of the gas  exiting the filter
vessel.  A large amount of particulate matter was found in the off gas
piping before and after the Balston total filter.  This was  evidence of a
large amount of particulate penetration.   A second Balston filter  was also
destroyed.  After the removal of the second filter, the Balston filter  vessel
failed and the test was terminated.  The size distribution of  the  particulates
removed from the filter support screen is shown in Table  IV-13.  The average
particle size of 3 microns is almost identical  to the average  size of
material before the filter.  Therefore it appears that very  little filtra-
tion took place.

     The filtration velocity that was used during these tests  was  higher than
any tested before on the miniplant.  This happened because the AP  cell  which
measures the pressure drop across the measuring orifice was  miscalibrated.
It was calibrated for 0-30 kPa rather than 0-30 in We. This error resulted
in actual filtration face velocities much higher than expected.

     During the second Balston total  filter  sampling  test, the pressure drop
across the filter was constant throughout the filtration  cycle.  This indi-
cated a complete loss of filter media.  This, along with  the failure of the
sampling filter vessel, caused the termination of the test.

     After the run, the plenum of the test vessel was opened and approximately
7  kg of material was  collected.  This material was alumina bed material, which
had been blown out of the filter bed, containing less than 10% flyash.

     A summary of granular bed filter test conditions and results for Run 115
is shown in Table IV-14.  Even with all of the things that went wrong during
the test, some conclusions are possible.  With a single element and controlled
blow backs, the maintenance of a proper AP was very difficult.  The pressure
drop continued to rise even during periods when  some bed material  was lost.
The pattern of pressure buildup with time can  be seen in Figure IV-25.   Per-
haps lower face velocities would have kept the pressure drop from rising as
quickly.  However, the fact remains there was  a  continued failure to clean
the filter properly and check the increased pressure drop even with signifi-
cant loss of filter media.  The operating problems seen in this run, poor
filter cleaning and loss of filter media, are  the same problems that termi*
nated the earlier program.

     Unfortunately, the testing of the slip stream filter was  terminated
before all operability questions could be answered.  An unresolved issue was
whether the loss of filter media could have been prevented by  modifying the
blow back procedure,  possibly by introducing blow back air more gradually
instead of in short pulses.  The effect of lowering the filter face velocity
on filtration performance was also unresolved.
                                      144

-------
                TABLE  IV-14.   RUN 115  FILTER TEST SUMMARY
           Run Length  (total)
           Number of Beds
           Design
           Filter Medium
           Bed Depth (cm)
           Filter Media Particle Size (ym)
           Vessel Preheat
                         4 Hours
                         1
                         Exxon Mark IV
                         Alumina
                         7
                         1410-2380
                         Inlet Pipe Only
                              Run Breakdown
Segment
Run Length (hrs)

Operating Conditions
Filter Inlet Temperature (°C)
Filter Outlet Temperature (°C)
Pressure (kPa)
Baseline AP
   Clean Bed (kPa)
AP Before Blow Back (kPa)
Superficial Velocity (m/min)
Duration:  Filtration (min)
Blow Back Conditions
Superficial Velocity
Duration:  Blow Back
Blow Back Program
(s)
                    0.75
                  400
                  350
                  900

                   28
                   35
                   38
                    5
 NA
20
 1
                2.25
                NA
              630
              900

               38
               43
               52
                5
 NA
20
 2
              NA
              630
              900

               23
               23
               45
                5
NA
 20
  2
NA  =  Not Available
                                    145

-------
                        FIGURE IV-25
     GRANULAR BED FILTER FLOW AND PRESSURE DROP VS. TIME
   3.7
   3.1
°°E  2.5|-
o
II I
     g
     «O
      v-i
     0
               5 MIN
                                FLOW
                            TIME-
                                                    60
                                                    40
                                                        70




                                                        -o
                                                     20
                            146

-------
 CONVENTIONAL  PARTICULATE CONTROL

     A  series of long term (4 to 10 day) continuous tests were carried out in
 which a conventional low temperature, low pressure electrostatic precipitator
 (ESP) and bag house were tested on the flue gas from the miniplant.   These
 tests were conducted during minlplant runs 103 through 108, with the three
 stages of high temperature/pressure cyclones in operation; the low tempera-
 ture/pressure devices thus represented a sourth stage of particle cleanup.
 The purpose of the tests was to determine if an ESP or bag house could be
 used, after expansion of the flue gas through the gas turbine, to meet partl-
 culate emission standards.   The possibility being considered was that cyclones
may be sufficient to protect the gas turbine from excessive wear but may not
 be sufficient to meet environmental standards.  The tests were performed
 using mobile, trailer mounted control devices operated by Acurex Corporation
 for the EPA.   These units received a stream of expanded, diluted miniplant
 flue gas during 3 runs each for the ESP (runs 103, 104,  105) and the bag
 house (runs 106, 107, 108).  Due to the system configuration,  the flue gas
was diluted by 50 to 70% with air used for pressure control.  A sketch of
the test configuration is shown in Figure 111-10.

Results
     Preliminary results with the EPA mobile ESP indicate  the  applicability
of conventional electrostatic precipitation for  control  of cyclone-cleaned
PFBC particulate emissions.  Overall  results from 17  days  of operation  (runs
103, 104 and 105) reflect 87% efficiency, corresponding  to an  emission  level
of 8.6 ng/J (0.02 g/Nm3).

     Preliminary results with the EPA mobile bag house also indicated the
applicability of conventional fabric  filtration  for control of cyclone-cleaned
PFBC particulate emissions.  Overall  results from 15  days  of operation  (runs
106, 107 and 108) reflect 99.3% efficiency, corresponding  to an emission level
of 0.46 ng/J (0.001 g/Nm3).

     A more detailed and comprehensive accounting of  these tests will be
contained in a report to be published in late 1979 by Acurex.
                                    147

-------
                                  SECTION  V

                            REGENERATION STUDIES


     In 1976, the combined operation  of the  mini pi ant  combustor and regenerator
sections was demonstrated.  That demonstration  run,  reported  in the previous
annual  report, illustrated the operability of the system  and  demonstrated that
sorbent regeneration does reduce the  amount  of  makeup  sorbent required.  The
next step in the regeneration study,  and the primary objective of  the combined
combustor-regenerator test program conducted in 1979,  was to  quantify the
effect of certain key variables on the performance  of  the combined combustor/
regenerator system.   The variables studied were makeup Ca/S  ratio  and sorbent
recirculation rate.

EQUIPMENT AND PROCEDURES

Equipment

     The equipment and materials were described in  detail in  the  previous
two annual reports (1,2).  No major changes  have  occurred since,  and only
brief summaries are included in this  report.

Air System--
     The two separate air systems are burner air  and supplementary air.  All
air is supplied by the main air compressor.  Automatic control systems,  con-
sisting of control valves, flow measuring  orifices, and electronic control-
lers, are used to regulate air flow.   Burner air  is supplied to  the burner,
located beneath the fluldizing grid, in sufficient  quantity to completely
burn the fuel (natural gas),  Supplementary  air is  added about halfway  up  the
bed in order to create an oxidizing zone in  the upper  portion of the bed.

Fuel System—
     The two fuel systems are burner fuel  and  supplementary fuel.  Automatic
control systems, similar  to those used for air flows, are used to regulate
the flow of natural gas.  Burner fuel is supplied to the burner where it is
burned with an approximately stoichiometric amount of air.  Supplementary
fuel is added directly to the regenerator column just above the fluidizing
grid in order to produce  reducing gases (CO, H2).

Off Gas Handling--
     Hot  pressurized  gases leaving the regenerator are cooled in a single
pass double  pipe heat exchanger and expanded to nearly atmospheric pressure
across  a  control valve.   Dust  is  removed from the gas upstream of the cooler
by a cyclone  and upstream of  the  pressure control valve  by a  stainless steel
knockout  vessel.

      Off  gases  from  the  regenerator are sent to a Research-Cottrell scrubber
for cleanup before  venting.   Ammonia  is injected to neutralize the scrubber
water.
                                      148

-------
 Gas Sampling System--
      A slipstream of the off gas  is  taken  downstream of the pressure reducing
 valve.   The gas is filtered  (Balston Model  33  filter) and dried  (Perma-Pure
 Model  PD-1000-24S self-regenerative  membrane-type dryer) before  entering the
 gas analyzers.

 Fluidizing Grid--
      The fluidizing  grid has  88 holes of 3.6 mm  (9/64 in) diameter for passage
 of the  fluidizing gases  (from the  burner located beneath grid) and 14 water
 cooling channels  of  4.8  mm (3/16  in)  diameter.   The cooling water flow is
 controlled independently through  six  separate cooling zones.  The 14 channels
 were separated  into  groups of 3,  2,  2, 2,  2, and 3, each group or zone having
 its own water supply.

 Burner--
      No changes were made to  the  regenerator burner since the last report.
 This  unit  is identical to that used  in the miniplant combustor and is described
 in a  previous annual report  (2).

 Sorbent Transfer  System--
      The solids transfer system used to circulate solids between the combustor
 and regenerator is shown schematically in  Figure V-l.   Pressure in the regen-
 erator  is  maintained slightly higher than  that in the combustor.  Solids in
 the regenerator-to-combustor  transfer line move into the combustor when a
 pulse of nitrogen  is applied  to the lower  end of the transfer line.  The flow
 rate of solids  is  controlled  by adjusting  the frequency, duration, and inten-
 sity of the pulse.   Two  slide valves are used in the combustor-to-regenerator
 transfer line in  order to prevent back flow of gas from the  regenerator up  the
 line.   These automatic valves trap solids  in the piping between them.  Solids
 are discharged  into  the  regenerator when the bottom valve is opened.  The two
 solids'  take-off  plugs shown  in Figure V-l  are inserted into the ports during
 startup  to prevent solids from entering the lines.  Plugging can occur if the
 solids  become wet due to water condensation during startup.   The manual  slide
 valve in the regenerator-to-combustor line is also closed during startup and
 during  upsets.

     The components of the sorbent transfer system (valves,  expansion joints,
 etc.) are described in detail in a previous annual  report (2).   The transfer*
 lines themselves were fabricated from 6 inch schedule  40 carbon steel  pipe
and refractory lined to an inside  diameter  of 7.6 cm.   The sloping portions
of  the lines were sleeved with 2-1/2  inch Schedule 10  316 stainless  steel
 pipe, which has  an inside diameter of 6.7 cm.

Operating Procedures

Startup of Transfer System--
     The solids  transfer  system is not operated  during  startup  to  prevent
moisture from entering the transfer lines.   Plugs  are  inserted  in  the  solids
takeoff ports so that solids  cannot spill  into  the lines  from the  combustor
and regenerator  vessels.   Nitrogen is  pulsed  into  the transfer  line  pulse pots


                                     149

-------
                    FIGURE V-l
   MINIPLANT  SOLIDS TRANSFER SYSTEM (SCHEMATIC)
REGENERATOR
COMBUSTOR
                SOLIDS TAKE
                 OFF PLUG
                                 AUTO
                                 SLIDE
                                VALVE
                     NITROGEN
                       PULSE
                        150

-------
 1n order to prevent solids  from backing  up  Into  the  lines.  The slide valves
 are also cycled occasionally 1n order  to dislodge any sol Ids that may have
 passed by the takeoff plugs.  Pressure 1n the  regenerator 1s set slightly
 above that 1n the combustor (usually 2.5-10 kPa) and the beds are heated to
 near operating temperatures.  Transfer of sol Ids Is started by pulling the
 plugs out of the takeoff ports.  The transfer  rate 1s controlled by setting
 the cycle time of the slide valves  1n  the combustor-to-regenerator line.

 Starting Reducing Conditions--
      The regenerator 1s  operated  under net  oxidizing conditions during heatup.
 Under these conditions,  no  regeneration  of  sulfated sorbent occurs.  When the*
 combustor and regenerator bed  temperatures  are uniform and close to the
 desired operating temperatures, the switchover to reducing conditions 1n the
 regenerator 1s made.  This  1s  accomplished  by  increasing flow of supplementary
 air to the required  value and  then  increasing  the flow of supplementary fuel
 Supplementary air flow 1s always  increased  before supplementary fuel  so that*
 air is not added to  a column already filled with a reducing gas.  Temperature
 is continuously monitored and  flow  rates  of burner air, burner fuel,  supple-
 mentary air,  and supplementary fuel are  all  adjusted to yield the desired bed
 temperature.   Oxygen and  CO concentrations 1n  the off gas are also monitored
 and the supplementary air flow is corrected to produce low concentrations of
 CO (under 5000 ppm).   The flow rate of supplementary air is the minimum value
 which will  just produce  CO  1n  the oxidizing zone.

 Shutdown--
      It is important during  shutdown, to empty the transfer lines of  sol Ids;
 otherwise plugging of the lines may occur when the unit is  restarted.  Hence
 the first step 1s to  shut the  plugs 1n the sol  Ids takeoff ports, thereby    '
 preventing solids from entering the lines.  Cycling of the  slide valves  1n
 the combustor-to-regenerator Hne and pulsing of nitrogen in the regenerator-
 to-combustor  line is  continued until the mlniplant is shut  down.  This  assures
 that the  lines  are emptied.

      Normal shutdown  can  be accomplished  several  ways.   One method is to  press
 the  "emergency  stop"  button, which shuts  all flows  of air and  fuel, fills the
 column with nitrogen  and slowly depressurizes the column.   This  method  is
 preferable  if  the final composition of solids is  important,  as  nitrogen  plus
 the  rapid  temperature drop "freezes" the  composition  of the  bed.

     Another method of shutdown is to turn off  supplementary fuel  and air
 flows, in  that order.  Burner fuel is  shut next.   Column  pressure  and burner
 air  flow must be brought down together  so as not  to create  high  superficial
 velocities, which would blow solids out of the  column.

 EXPERIMENTAL RESULTS AND DISCUSSION

 Shakedown

     After some cold flow tests, a hot  test, run  101, was made to  insure  that
 the solids transfer lines and control systems were operating smoothly and  to
determine the operating pressure for the  test program (7 atm).  Shakedown went


                                    151

-------
very smoothly with the  combined  units operating as well as they had during the
first 100 hour demonstration  run In  1976.

Operating Performance

     During the regenerator test series  (runs  102, 103 and 105) over 400 hours
of operation were logged with solids circulating  between  the  regenerator and
combustor, 370 of these hours were with  sorbent regeneration.   Including the
100 hour demonstration  run in 1976,  the  combined  combustor-regenerator system
was operated for more than 500 hours.  Overall, operations of the  two units
was smooth.  Combustor  problems  were limited to infrequent interruptions in
coal feed due to coal flow problems  which  were quickly remedied.   Problems in
regenerator operation,  which  were basically limited  to two areas,  required
maintenance during and  between runs  but  did not result in forced shutdown.
The two problem areas were off gas handling and solids transfer,

     Regenerator off gas passes  through  a  cyclone before  being cooled and
pressure reduced.  The cyclone was operational during run 102, but during  the
second day of run 103,  the downcomer became plugged  with  fines, probably a
result of wall buildup from the previous run coupled with moist particulates
from the startup of 103.  A large, semi-batch  granular bed  sand filter was
put on stream to replace the cyclone rather than  shutting down to  clear  the
cyclone.  The high dust loading 1n the regenerator off gas  prior to the
installation of the sand filter and during the filter blow  back intervals
caused appreciable erosion of the pressure control  valve  stem and  housing.

     The major problem with the solids transfer  system was  a blockage in  the
combustor to regenerator line that prevented sol Ids movement at the start  of
run 103.  It was found that a piece of refractory wall  in the 45°  angle  sec-
tion just above the  pulse  pot had fallen into the opening.   The section  was
removed, the refractory recast, and the run resumed.

     The solids transfer system also proved to be a limiting factor on the
solids circulation rate.   As solids are transferred from the combustor to
regenerator, they are  trapped and held between the two automatic knife valves.
While the solids are trapped between cycles,  they cool.   If the transfer leg
temperature drops too  low, water vapor may condense making the sol Ids mushy
and preventing  solids  flow.  The lowest solids circulation rate which seemed
safe for the existing  system was 10-12 kg/hr.

     During  the regeneration test series, additional testing was done by
outside  contractors.   Acurex  Company tested a mobile electrostatic precipita-
tor using  a  slip  stream  of combustor flue gas during runs 103 and 105.  GCA
Technology  Company performed  Level  I comprehensive analysis  tests on both
the combustor  and  regenerator during run 105.  These test programs are
discussed  elsewhere  in this  report.

 Run Conditions  and Summaries

      Run summaries giving operating conditions for  both  the  combustor and
 regenerator are presented in Tables V-l and V-2.  During the 3 runs, six
 steady  state conditions  were reached.   Temperatures, pressures, superficial

                                     152

-------
in
CO
                        TABLE V-l.   MINIPLANT  REGENERATOR RUN SUMMARY

                               Regenerator  Operating Conditions

Run No.                                 102      103.0      103.1      103.2     103.3        105

Run Length, (reducing)  hrs             97.5        54         48         20        56.5       80.7
Pressure, kPa                         710        710        710        710       710        705
Air Flow Rate, m3/min                   2.86       3.08       2.72       2.52      2.31       2.3
Fuel Flow Rate, m3/m1n                   0.26       0.30       0.27       0.23      0.20       0.23
Average Bed Temp - Reducing Zone     1025        1004        1003       1008       974       1010
              °C - Oxidizing Zone    1053        1024        1021       1025       985       1031
Superficial Gas Velocity, m/s
                 - Reducing Zone        0.64       0.78       0.67       0.50      0.58       0.60
                 - Oxidizing Zone       0.91        1.08       0.98       0.87      0.78       0.77
Expanded Bed Height, m                   1.65       1.37       1.38       1.38      1.37       1.85
Air to Fuel Ratio, 4>
                 - Reducing Zone        1.36       1.31       1.37       1,35      1.32       1.27
                 - Oxidizing Zone       0.92       0,95       0.94       0.91      0.96       1.02
Solids Recirculation Rate, kg/hr       88.5        22.7       22.7       12.2      12.2       15.4
Average SOg Emissions,  %                0.32       0.40       0.46       0.25      0.16       0.51
Average CO  Emissions,  ppm           1610         934        484        231       996       1438
Average C0£ Emissions,  %                9.8        9.44       9.42       8.38      8.69      10.9
Average 02  Emissions,  %                0.17       0.21       0.31       0.42      0.18       0.26

-------
TABLE V-2.  MINIPLANT FLUIDIZED BED COAL COMBUSTION RUN SUMMARY
Combustor
Operating Conditions
Run Length, hrs.
Pressure, kPa
Air Flow Rate, m3/min
Temperature Gradient, °C/m
Average Bed Temperature, °C
Superficial Velocity, m/sec
Settled Bed Height, m
Initial
Final
Expanded Bed Height, m
Coal Feed Rate, kg/hr
Ca/S Molar Feed Ratio-Set*
Ca/S Molar Ratio-Equivalent**
Excess Air, %
Sorbent
Coal
102
3/13-17/79
107
700
12.4
-15,2
908
1.43

2.06
0.96
2.17
69.5
1.5
54
20.3
GL
CH
103.0
3/29-4/1/79
81
700
13.3
-12.2
913
1.52

2.11

2.65
76.6
1.35
12
16.8
GL
CH
103.1
1/4-3/79
49
700
13.4
-14.0
901
1.52

__
._
2.83
77.7
0.68
11
17.7
GL
CH
103.2
4/3-4/79
22
700
14.1
-12.3
901
1.59

__
__
1.97
79.4
0.68
6.1
13.0
GL
CH
103.3
4/4-6/79
59
700
14.4
-17.7
902
1.62

-_
1.42
2.24
79.9
0.93
6.4
18.9
GL
CH
105
4/30-5/5/79
99
700
14.0
-55.9
894
1.46

2.23
1.71
3.10
77
1 .29
8.3
45.8
GL
CH
Flue Gas Emissions - Combustor
S02, ppm
NOX, ppm
CO,  ppm
C02, %
021  *
Results - Combustor
    Retention, %
Ca Sulfation, %
Lb SO?/M BTU
Lb NOx/M BTU

GL  =  Grove Limestone
CH  =  Champion Coal
                  27
                 126
                  90
                  14.0
                  4.0
                  97.1
                  5.1
                  0.06
                  0.19
             28
             77
             71
             13.7
              3.3
41
56
62
14.1
3.4
22
65
64
14.1
2.6
77
55
56
13.7
3.6
97.5
18.2
0.06
0.12
96.4
23.7
0.09
0.09
98.0
22.2
0.05
0.10
93.0
28.0
0.17
0.09
                                                                                              70
                                                                                              59
                                                                                              77
                                                                                              12.4
                                                                                               4.1
                                                            93.6
                                                            23.4
                                                             0.15
                                                             0.09
              **
Fresh sorbent feed only.
Includes recirculating sorbent from regenerator,  as  well  as  fresh
sorbent feed; assumes 100% regeneration of the recirculated  sorbent,

-------
 gas  velocities,  bed  heights, coal feed rates, sorbent type, coal type, and
 reducing atmosphere  composition were held fairly constant throughout the runs.
 Solids  redrculation rate and Ca/S ratio fed to the combustor were the vari-
 ables studied for their effects on combustor and regenerator $03 emissions,
 sulfur  retention, and makeup sorbent rate needed to meet EPA emission stand-
 ards.

 Results

     In all cases, sulfur retention exceeded 90% with makeup Ca/S ratios rang-
 ing  between 0.68 and 1.5, compared to a Ca/S ratio of 3 to 4 which is required
 to achieve the same results for once through operation with limestone.  The
 significant reduction in fresh (makeup) sorbent requirements is due to the
 regenerated sorbent which circulates from the regenerator to the combustor.

    The
 Ca/s
 The
     r  /c    -  0.32 x SCA x (SFR x SCR + 1.79 RSR x (1  - SSL)  x (1  + SCR}}
     Ca/SEQ  -                           CS x SFR	

     where SCA = % CaC03 in sorbent,  for limestone ~100%
           SFR = coal + fresh sorbent feed rate, kg/hr
           SCR = sorbent to coal  ratio, kg/kg
           RSR = recirculating sorbent rate, kg/hr
           SSL - sorbent sulfation level  of regenerated  sorbent, wt.  fraction
            CS = coal sulfur content, %

 Comparison of the fresh sorbent Ca/S  ratio with the Ca/S equivalent  shows  that
 the rate of circulating regenerated sorbent is  much larger than the  fresh  sor-
 bent rate and therefore, probably plays a larger role  in controlling  S0£ emis-
 sions.

 Combustor SOg Retention—
     As  mentioned previously, S0£ retention exceeded 90% for the runs  made,
covering a range of Ca/S ratios.   An  effective  method  to evaluate sorbent
regeneration is to compare SOg retention  as  a function of CB/SEQ for  the
regeneration runs to the data obtained  for once through  operation.   Figure
V-2 shows such a comparison.  All  SOg retentions were  adjusted  to a  2  second
 gas residence time using the first order  rate expression.

     Figure V-2 shows that the data points generated by  the regeneration runs
fall  around the curve for once through  operation with  limestone,   The  values
for Ca/Sfcq are based on an assumption of  100% sorbent  regeneration due to
difficulty in obtaining regenerated bed samples.  If the samples  were  not
 fully regenerated, i.e., the assumption was  invalid, the data points would
 shift to the left toward a lower  CB/SEQ.   For example, one regenerator bed
 sample was obtained during run 105 when a pressure upset caused some regen-
 erator bed to be blown out of the column  into the cyclone.   The sample was
analyzed and found to contain 14.6% CaS04 by weight.  The equivalent  Ca/S
molar ratio dropped from 8.3 to  6.0 when  the sorbent sulfation  level was
 taken into account in the Ca/Sgq  expression.

                                   155

-------
                          FIGURE V-2

  SO2 RETENTION ADJUSTED TO 2 s RESIDENCE TIME VS Ca/S RATIO
(N
z
o
I—
Z
UJ
LU
oe.
 CN
    100
     90
     80
     70
     60
      50
      40
                  'LIMESTONE
           DOLOMITE
                      ONCE-THROUGH OPERATION
                    • WITH LIMESTONE
                    • REGENERATION RUNS
       0   1
                   3456789
                       Ca/S MOLAR RATIO*
10  11
        * FRESH SORBENT ONLY FOR ONCE-THROUGH RUNS
         FRESH PLUS RECIRCULATED SORBENT FOR REGENERATION
         RUNS
                          156

-------
      The scatter in the data is  not  appreciable and  from this treatment of
 the data it can be concluded that  the  regenerated sorbent behaves as if it
 were fresh sorbent.

 Sorbent Activity Loss-
      Previous  work done in  TGAs  and  batch units indicated a decline in sorbent
 activity as the sorbent was  cycled between sulfations and regenerations (13),
 In a continuous unit such as  the miniplant, the sorbent does cycle a number of
 times but the  cycling is not  comparable with batch cycling experiments.  Sorbent
 lost by attrition must be replaced with makeup (fresh) sorbent.  The fresh
 sorbent feed serves to maintain  the  age distribution of the sorbent in the
 system, and therefore the activity of  the bed, at a  steady state value.  From
 the outset of  a run,  the average age of the sorbent  bed changes exponentially
 and the steady state bed age  distribution is approached asymptotically.  The
 rate at which  the bed steady  state is  reached is determined by the system time
 constant,  t, where t  = bed  inventory/feed rate.  (Based upon the range of
 values  for t in these miniplant  runs,  the exponential expression would suggest
 that the combustor bed should approach within 90% of its steady state level in
 about 90 hours  of operation.)  Therefore, unless the sorbent suffers rapid
 and severe deactlvation, sorbent deactivation would not be obvious in a con-
 tinuous system  once a steady  state bed age distribution has been reached.

      Figures V-3,  v-4 and V-5 illustrate the change in average bed age with
 time as well as  the combustor S02 emissions for the regeneration test runs.
 When the regenerator  was brought into reducing conditions and regenerated
 sorbent began  to  enter the combustor at high rates, combustor SOg emissions
 dropped markedly.   As  seen in the figures, the combustor S02 emissions leveled
 out at  a  steady  state value a long time before the bed reached its steady
 state age  distribution.  If the  sorbent experienced deactlvation, this would
 have been  detectable  as  an increase  in combustor SOp emissions during the
 initial  line-out  period.

      Based  on the  results exhibited  in Figures V-3 through V-5,  it can be
 concluded  that during  the residence  time of a  sorbent particle,  minimal  or  no
 sorbent deactlvation occurs  or if deactlvation does occur, it does not greatly
 alter the overall activity of the bed.

      The results of the regeneration  test  series  which  are reported  here con-
 tradict the  findings of the  regenerator demonstration run reported previously
 (1).  In the earlier run, after about 50 hours  combustor S02  emissions  began
 to  increase  rapidly for a period of 50  hours,  then  leveled out at 550 ppm.
 The  increase in emissions was assumed to have  been  caused by  a gradual  decline
 in  the  activity of the regenerated  sorbent.  This  hypothesis  1s  not  supported
 by the most  recent miniplant regeneration  studies where there was  no  sign of
 sorbent deactlvation.  It 1s also not consistent  with the analysis of the
 recent  runs described above.  This  analysis  points  out  that 1f deactlvation
 occurs, its effect would only be  seen during the  initial  part of  the  run and
 not as  far  into the run as 50 hours.  The  increase  in combustor S02 emissions
 during  the  previous run may  have  been a result of operational difficulties
rather than sorbent deactlvation.  If sorbent regeneration or solids  transfer
was  inhibited for any reason, such  as a bed agglomerate,  sulfur retention in
 the combustor would be adversely  affected.

                                     157

-------
en
TO
          I
          Q_
          Z
          O
          on
          to
cs
          Qi
          O

          fc
          Z)
          CO
                                              FIGURE V-3


                      COMBUSTOR BED AGE AND SCX EMISSIONS VS. TIME (RUN 102)
                0   '     30
                  START

              REGENERATION
                                      SO  EMISSIONS
                      60
90      120      150

 HOURS INTO RUN
180
210
                                                                                     100
                                                                          80
                                                                                     60
                                                                         40
                                                                                     20
                                                                                          o

                                                                                          >
                                                                                          O
                                                                                          >
                                                                                          O
                                                                                          m
                                                                                          O
                                                                                          m
                                                                                          CO
                                                                                          C
                                                                                          ;H

                                                                                          O

                                                                                          z

-------
                                  FIGURE V-4


           COMBUSTOR BED AGE AND SO2 EMISSIONS VS. TIME (RUN 103)
oo

Z

o

OO
10



UJ
O

CO

Z)
CD

5

O
u
        REGENERATION
                              90      120     150


                               HOURS INTO RUN
                                                                             03

                                                                             rn

                                                                             O
o
m


^
ff-


-T3

?0

O


o




O
 O
 m
                                                                             TO

                                                                             55

                                                                             c
                                                                             O
                                                                             z

-------
en
o
          O
          CQ


          O
                                        FIGURE V-5

                  COMBUSTOR BED AGE AND SO2  EMISSIONS VS.  TIME (RUN 105)
OO
z
o
8   100
           CM
    SO2  EMISSIONS
                                                                                  100
                                                                                  80
                                                                                  60
                                          40
                                                                                  20
                                                                                       oo
                                                                                       m
                                                                                       D
                                                                                       O
                                               -o
                                               -o
                                               •yo
                                               o

                                               3
                                               5
                                               S
                                               oo
                                                                             O
                                                                             m
                                                                                       oo
                                                                                       C

                                                                                       O
                                                                                       z
                0  I     30
                 START
             REGENERATION
90      120      150


 HOURS  INTO RUN
                                                      180
210

-------
 Sorbent  Elutriation  Losses--
      During  the  test series, the minimum Ca/S ratio fed to the combustor was
 set  by sorbent attrition  rather than  by sorbent desulfurization activity.  To
 maintain  bed levels  in the combustor  and regenerator, bed lost through attri-
 tion was  replaced  by the  incoming makeup sorbent.  Sorbent losses represent
 elutriation  of the fines  formed by attrition.

      Since sorbent elutriation was the factor limiting the Ca/S ratio, it is
 important to compare elutriation losses for the once through process and the
 regenerative process to see the effect regeneration has on attrition and
 subsequent elutriation.   Sorbent elutriation losses for the once through
 process  (miniplant combustor only) for both Grove limestone and Pfizer dolomite
 were presented in  a  previous report (2).  Sorbent losses for the regeneration
 test series  are  presented in Table V-3.  Losses for the once through process
 using  both uncalcined and precalcined limestone are included for comparison.
 The  sorbent  loss calculations are based on the calcium contents of the
 materials entering and leaving the system and are essentially calcium losses.

      Combustor sorbent losses expressed as the Ca/S ratio needed to offset the
 loss,  and as  percent of fresh sorbent, are extremely high.  However, combustor
 elutriation  losses expressed as a percent of bed inventory for the regenera-
 tion runs do  not differ greatly from the values for once through operation.
 With the  combined  combustor-regenerator unit, the regenerator off gas provides
 an additional source of sorbent elutriation,  Regenerator sorbent losses
 expressed as  a percent of bed inventory tend to be higher than percent of bed
 inventory combustor  losses.  A possible explanation for higher regenerator
 elutriation  losses is that the regenerated sorbent (CaO) is less attrition
 resistant than sulfated sorbent (CaSCty).  The sorbent is also subject to
 thermal shock as it  enters the regenerator.  This may also contribute to
 higher attrition rates.

     Combined losses from both the combustor and regenerator reflect the fact
 that the miniplant,  under the conditions of the regeneration test series, was
 operating in the range of the minimum Ca/S ratio needed to maintain bed levels.
 In a number of the regenerative cases, elutriation losses exceeded the feed
 rate of fresh sorbent.  Regeneration therefore can reduce the Ca/S ratio
 needed to meet EPA emission standards but  not beyond  the point were elutria-
 tion losses are larger than the makeup sorbent ratio.   To specify the minimum
 Ca/S ratio necessary to provide 90% or less S02 removal  in the miniplant, a
 high sulfur coal  would have to  be burnt so  that bed activity, rather than
 attrition, would be the limiting factor at  those lower percentage S02
removals.

S02 Content in Regenerator Off  Gas--
     Miniplant regenerator S02  emissions are not limited by thermodynamics.
 For the operating conditions  used  in the test series,  thermodynamics  predicts
 an equilibrium S02 concentration of -"3%.   In all  runs,  S0£ emissions  were
much lower than the equilibrium S02  concentration.  Average SO^  emissions
 ranged from 0.16 to 0.51%, equivalent  to 6  to 16%  of  the equilibrium  concen-
 tration (see Table V-4),
                                     161

-------
                                 TABLE V-3.  SORBENT ELUTRIATION LOSSES
ro
Sorbent
Run No.
UncaUlnedW
GL
Precalcined ' '
GL
102
GL
103.0
GL
103.1
GL
103.2
GL
103.3
GL
105
Superficial
Gas
Velocity
m/s
i 1.4-2.2
) 1.6-2.5
1.43
1.52
1.52
1.59
1.62
1.46
Combustor
Feed Rate
Makeup
Ca/S
1.5-2.8
2.5-4.0
1.5
1.35
0.68
0.68
0.93
1.29
Ca/S Loss
Equivalent
0.2
0.2
2.4
0.80
0.58
0.85
0.45
0.55
Fresh
Feed
12
8
160
58
86
126
48
43
Losses
Vol. %
Bed/Hr
1.1
0.8
3.2
1.3
0.9
2.1
1.0
0.9

Total
Feed
12
8
6
11
9
23
11
11
Reg.
Losses
Vol . %
Bed/Hr
—
1.6
6.2
1.6
4.3
1.6
0.7
Combined Losses
Ca/S Loss
Equivalent
0.2
0.2
2.8
1.5
0.78
1.4
0.63
0.67
Fresh
Feed
12
8
185
113
114
200
68
51
Vol. %
Bed/Hr
1.1
0.8
2.8
2.1
1.1
2.6
1.1
0.8
    GL   =   Grove  Limestone
    (1)   Once-through operation.

-------
(71
CO
                            TABLE V-4.  COMPOSITION OF REGENERATOR OFF GAS
                                                    Equilibrium
                                 Avg. S02           S02 Cone. @              S02 Emissions
               Run No.        Emissions (%)       Conditions (%)       (% of Equilibrium Cone.)
102
103.0
103.1
103.2
103.3
105
105 Peak
0.32
0.40
0.46
0.25
0.16
0.51
1.0
3.4
3.0
3.0
3.1
2.6
3.2
3.5
9
13
15
8
6
16
29

-------
     The low S02 concentrations are  caused  by limitations  imposed  by  the  size
of the minlplant regenerator and combustor.   S02  concentration  in  the regen-
erator off gas 1s determined by mass and  energy balance  constraints rather
than chemical equilibria or reaction rates.   Heat losses from the  regenerator
are high and require the addition of more hot gas than  is  needed to satisfy
the requirements of the regeneration chemical  reactions.  Also, the amount of
CaS04 (sulfated limestone) fed to the regenerator 1s set by the sulfur  content
of the coal and the size of the combustor.   The regenerator is  actually over-
sized for the CaS04 rates possible in the recent  test series.   Reducing the
size of the regenerator or increasing the size of the combustor is not  prac-
tical and coals with a higher sulfur content were not available for this
program.

     In addition to these constraints, the superficial  gas velocity 1n  the
regenerator also had to be high enough to exceed  the minimum fluidization
velocity and promote vigorous sol Ids mixing.  Operating pressure  had  to be
lowered from 900 kPa to 700 kPa in order to achieve the proper  superficial
velocity at the desired operating temperature without diluting  the off gas
even further.

     Therefore, the SOg levels measured 1n this test program are  set  by the
limitations  imposed by the minlplant system and do not represent  typical
results from a larger facility which would not have the same limitations.
The  larger  facility would  be designed to allow the S02 levels to approach
those  predicted at chemical equilibrium.

     The  last portion of run 105 was a test to determine the maximum $03  con-
centration  attainable in the regenerator off gas  and to compare this  with
thermodynamic predictions.  The test was carried  out by establishing oxidizing
conditions  1n the regenerator to stop further sorbent regeneration and allow
the  sulfation level of the  circulating sorbent to Increase.  It was  Intended
to do  this  for a period of  time until the combustor $03 emissions increased
to a fairly high level.   This would  indicate the sorbent was highly  sulfated.
When regeneration conditions were re-established, the initial S02 concentra-
tion from the regenerator  would approach an  equilibrium concentration.

      After 6 hours  of oxidizing conditions,  SO?  emissions  from the combustor
had  not increased,  Indicating  that  before oxidizing conditions had been
established, the sorbent  had  been regenerated  to a  high degree and was still
very active.  Due  to  time limitations, reducing  conditions had to be established
 1n the regenerator  following  the  6  hour  oxidizing period  but before  combus-
tor  emissions  had  shown  any increase.  Regenerator  and  combustor  S02 emissions
 following the start of  reducing conditions  are plotted  as  a function of  time
 in Figure V-6.   The cyclic nature of the regenerator S02  emissions 1s  due to
 sorbent transfer  from the combustor which  cools  the regenerator bed  and
 decreases the regeneration reaction.  The  Initial  increase in  combustor  S02
 emissions was probably caused by regenerator gas passing  into  the combustor
 since the regenerator operates at a slightly higher pressure to facilitate
 solids transfer.   Regenerator S02 emissions seemed  to  peak and level out at
 1.0%.  At the existing conditions,  thermodynamics predicts an  equilibrium  con-
 centration of 3.5%.  Therefore the  peak  emission was only 29%  of  the equilib-
 rium value (see Table V-4).

                                      164

-------
en
en
       z
        cs
       ex.
       o
       LLJ

       z
       LLJ

       O
       LLJ
       Qi
                                                 FIGURE V-6

                                REGENERATOR AND COMBUSTOR SO2 EMISSIONS

                                                  RUN 105
             0
             OXIDIZING
           60                     120

MINUTES INTO REDUCING CONDITIONS
                                                               O
                                                               O

                                                               OO
                                                               C
                                                               CO

                                                               o
                                                               •70


                                                               O
                                                               co
                                                               °2

                                                               O
                                                               Z
                                                               CO

-------
     Thermodynamics  was  obviously  not  the  limiting  factor  in determining regen-
erator off gas SOg concentrations.   The  Champion  coal used  in the test series
had a low sulfur content (< 2%)  which  led  to  a  low  sulfur  load on the unit,
thereby limiting regenerator SOg emissions due  to mass  and  heat  balance con-
siderations as discussed previously.

Mass Balances--
     Table V-5 summarizes steady state sulfur mass  balances for  the regenera-
tor runs.  Also included are total  mass  balances  for  each  run.   Total mass
balance closure was  99+% in all  cases.  Sulfur  balances  for three of the runs
(102, 103.1, and 105)  were very  good,  just slightly over 100%.   Sulfur recovery
for run 103.0 was 120%,  indicating that  the initial sulfur  inventory (from the
initial bed charge of sul fated sorbent)  was being depleted  by regeneration and
was affecting the mass balance.

     The last two segments of run  103  (103.2  and  103.3)  had low  sulfur recov-
eries, 74% and 48%, respectively.   Since the  total  mass balances were 100%,
the cause of the deficient sulfur  balances is not obvious,  A possible explana-
tion is that the regeneration process  was  not keeping up with the  sulfation
process, increasing the sulfur inventory of the bed.   Combustor  bed  probe
samples exhibit a slight increase  in sulfur level toward the  end of  the  run,
supporting this possibility.  Another  explanation would be a  sudden  change  in
sulfur load  (coal sulfur content)  which  would not be  accounted  for in  the
sulfur balance.  During run 103, an average coal  sulfur content of 1.61% was
used.  In addition, two coal samples were taken during run 103.3 to  determine
the consistency of the coal sulfur content.  The samples were taken  15  hours
apart; the first had a sulfur content of 1.84%, the second dropped to  1.37%.
If this drastic change in coal sulfur content was not just transient,  the  sul-
fur mass balance would be in error.

Conclusions

      Continuous operation of the miniplant provided a realistic way of measur-
 ing the  potential benefits  of a regenerative system compared to a once
through  system.   Regeneration can  reduce  the makeup sorbent requirement needed
to meet  EPA  emission  standards  by  a factor of  3  to 4 but not below the makeup
sorbent  ratio  needed  to  compensate for  attrition losses and thus maintain bed
inventory.   Sorbent activity  loss,  if it  occurs, was not apparent from the
performance  of the  combined combustor-regenerator  system on the basis of
retention  or combustor  SOg  emissions.   Results of  this  study may have been
affected  by  the fact  that  all work fell in the high  SOg retention range.
      If a higher sulfur  coal  (~4%)  would  have  been available for the test
 series, perhaps more conclusive  results would  have been obtained.  High sulfur
 coal would have allowed  lower makeup  Ca/S ratios  to  be tested without attri-
 tion being the limiting  factor.   The  increased sulfur load would have also
 improved the SOg concentration in the regenerator off gas, an important factor
 in downstream sulfur recovery.
                                      166

-------
                             TABLE V-5.  MINIPLANT REGENERATOR  MASS  BALANCES
en
--J
    Run No.
 102
103.0
103.1
103.2
103,3
  105
Sulfur Mass Balance
Sulfur Input
Coal
TOTAL
Sulfur Output
Regenerator Off Gas
Combustor Off Gas
Combustor Overhead Solids
2° Cyclone
3° Cyclone
Regenerator Overhead Solids
% Sulfur Recovery
% of Sulfur Entering
100
100

69.4
2.9

22.8
2.9
3.8
101.7
100
100

81 .9
2.4

19.5
3.2
13.8
120.2
100
100

82.3
3.2

13.6
2.4
4.0
105.9
100
100

40.7
1.6

19.6
3.1
9.4
74.3
100
100

24.1
7

10.9
2.3
3.9
47.7
100
100

78.4
6.3

13.5
1.6
2.4
101.4
     Total Mass Balance
99.88%
99.86%
99.89%
 99.79%
 99.87%
100.6%

-------
                                SECTION VI

                     COMPREHENSIVE ANALYSIS OF EMISSIONS
LEVEL I AND II  COMPREHENSIVE  ANALYSIS  TESTS

     The U.S.  Environmental Protection Agency  has  developed  a  phased approach
to assess the  environmental impact  of  solid, liquid  and  gaseous emissions  from
a process (26).  This  approach  includes sampling and analysis  for  a wide range
of organic and inorganic species.   The first phase of the  assessment (Level  I)
is intended to provide preliminary  environmental assessment  data and identify
principal problem areas.  The objective of the second phase  (Level  II)  is  to
obtain more detailed and accurate  data than  is available from  a Level  I  study.
Level II studies are intended to identify and  quantify specific compounds
whose presence could be inferred from  the results  of Level I tests.  Level  II
studies may also be initiated on the basis of  results of the biotests
employed as part of Level I  (27).   Exxon has  participated in Level  I and
Level II testing programs using the miniplant  facility.   This  work was  done
in cooperation with two other EPA Contracts, Battelle Columbus Laboratories
and GCA/Technology Division.

Level I Emission Measurements

     Level I comprehensive emission measurements  from the PFBC miniplant unit
without the regenerator unit operating were conducted in cooperation with
Battelle Columbus Laboratories during run series  50 in early 1977 (14).  These
results will be reported in detail  in a separate  report prepared by Battelle
and Exxon.  In this test series, an eastern (Champion) coal  was burned with
Pfizer dolomite sorbent.  In a subsequent test (run 69), made  with Illinois
coal and Pfizer dolomite, samples of solid materials were obtained and sent
to Battelle for analysis.  The samples were analyzed for  inorganic elements
using spark source mass spectroscopy  (SSMS), the same method used on the
solids samples from run series 50.  The purpose of this supplementary test
program was to provide a data base for Illinois coal use, to add to that
already established for Champion coal.  The data will be  incorporated by
Battelle in the interpretation of the  Level I results.  The results of the
analyses are reported  in Appendix I.

     A second  series of  Level  I tests  was conducted  during  run 105 on the
miniplant  with the  regenerator also in operation.   Champion coal  and Grove
 limestone  were used in  run 105 which  was  conducted  in cooperation with GCA/
 Technology Division.   The principal effect of the GCA program was the collec-
 tion of  samples from  the combustor  flue gas and the  regenerator flue gas
 using  the  Source  Assessment  Sampling  System (SASS).   Two  SASS tests were  con-
 ducted on  both the  combustor flue  gas  after expansion,  diluted with air used
 for  pressure  control,  and on the undiluted regenerator  off  gas.   Two SASS
 train  tests were  also  conducted on  a  filtered, undiluted  slip stream of the
 combustor  flue gas.   Other gas  samples were taken by GCA  for  analysis of  $03,
 HC1,  nitrogen  species  and Ci to Cj  hydrocarbons.  Gas  samples were also
 collected  by  Exxon  personnel for analysis of  the  volatile sulfur  compounds.

                                    168

-------
 In  support  of  the  comprehensive analysis,  solids  samples and scrubber slurry
 samples  were also  collected  by Exxon.   The details of the sampling analysis
 effort were presented  In a test plan  prepared  by  GCA prior to sampling (15).
 GCA is responsible for analysis of the  samples, and will report the results of
 the Level I tests  from run 105.

 Level  II  Emission  Measurements

     An  evaluation of  the survey data of run 50 series by Battelle Columbus
 Laboratory  (14)  indicated the need for  a Level II characterization of the
 emissions from the miniplant unit.  A series of Level II tests was conducted
 in  cooperation with GCA during run 107  on  the miniplant without the regenera-
 tor unit  operating. Operating conditions  were similar to those used in the
 run 50 series.   Champion coal was burned with Pfizer dolomite sorbent.

     The  gas and solid sampling techniques were the same as in run 105.  A
 total of  5  SASS  train  tests were completed by GCA.  Three SASS train tests
 were completed on  the  combustor flue gas after expansion, diluted with pres-
 sure control air.   Particulates from the first SASS test on the diluted
 combustor flue gas  were collected at about 100°C instead of 200°C, the normal
 sampling  temperature,  to check for organic deposition on the particulates at
 lower temperature.   Two other SASS train tests were completed on a filtered
 undiluted slip stream  of the combustor  flue gas for analysis of gaseous
 emissions.  Other  gas  samples were taken by GCA and Exxon personnel for
 analysis of $03, C]  to  Cj hydrocarbons and other sulfur compounds.  Parallel
 solid samples were  also collected by Exxon personnel.  Details of the sampling
 analysis effort are  included in a test plan prepared by GCA prior to sampling
 (16).  The  Level  II analytical  results of run 107 will  be reported by GCA.

 PRESENCE  OF Mg3(CaS04)4 IN 3RD CYCLONE  FLYASH

     As was reported by Argonne (17), the formation of Mg3Ca(S04)4 or MgSO/i
 should be avoided  when dolomite is used in PFBC if the sulfated solid waste*
 is  to be disposed  of as landfill.  In the presence of water, Mg3Ca(S04)4 1s
 unstable, decomposing  to MgS04 which is soluble in water, and CaS04 which is
 relatively  insoluble in water.   The chemistry related to the formation of the
 binary salt Mg3Ca(S04)4 is not  known to date.  However, it had been reported
 (17,18,19,20)  that Mg3Ca(S04)4  can form during the sulfation of dolomite under
 conditions  similar to  those in  PFBC.  Therefore, a short study was undertaken
 to determine if the binary salt is formed in  the waste solids produced in the
miniplant.  Samples from run 67 of the bed overflow solids,  partlculate
captured in the second and third  stage cyclones and solids found in the tur-
bine test section downstream of the third cyclone were analyzed by X-ray dif-
fraction.   In  run 67, the flue  gas contained  about 700 ppm S02 and 3  to 5%
02.   At the flue  gas pressure,  9  atm, MgS04 was thermodynamically possible at
temperatures below about 840°C.   No evidence  of the binary salt or MgS04  was
found in the second cyclone  particulates even though  the  temperature  was
between 770 and 800°C.   However,  third cyclone  material and  turbine test
section deposits  did contain  Mg3Ca(S04)4 but  no MgS04.  The  temperature in
the  third cyclone was 700 to  820°C,  the  turbine test  section  660  to 820°C.
The  absence  of Mg3Ca(S04)4  in the  second cyclone material  but  its  presence


                                    169

-------
in the third cyclone and turbine test section material  could  be  due  to  a
residence time effect.   The residence time  in the  piping between the two
cyclones is greater than the residence time between the combustor and the
second cyclone.  The temperature is  also lower in  the third  cyclone  and tur-
bine section thus favoring the formation to a greater extent.

     The extent of magnesium conversion in  the third cyclone and turbine  sec-
tion solids was estimated by a sulfate balance to  be from 30 to  100%.  How-
ever, the 700 ppm $03 level during run 67 was higher than the S02 levels  which
will be experienced under the New Source Performance Standard (1.2 Ib SOg/lO^
BTU or less, depending upon the percentage  reduction required in a particular
case).  Furthermore, the residence time of the flyash in the piping  from a
PFBC combustor to an efficient third stage  particulate removal device should
be shorter than in the miniplant and the gas temperatures higher.  Therefore,
formation of soluble Mg303(804)4 will be less under higher S02 retention
conditions, shorter residence times and higher temperatures between the com-
bustor and the particulate removal devices.  Additional tests should be con-
ducted to determine the extent of the binary salt formation under more
realistic conditions.
                                     170

-------
                                 SECTION VII

                             BENCH UNIT STUDIES


     Programs  were  carried out  1n both the bench combustor and regenerator
sections.   The combustor  had  been modified, as described in the previous
report (1), to permit  continuous solids feeding and removal.  After an initial
check out  period, the  combustor was used to evaluate three NOX control methods,
two stage  combustion,  NH3 Injection and simulated flue gas redrculation.  The
regenerator was used  1n a series of tests using sulfated sorbent produced in
the miniplant, studying the use of natural gas and coal to fuel the regenera-
tion section.   The  following  sections  describe the results of the  bench unit
combustion and regeneration tests.

COMBUSTION STUDIES

     The bench combustion unit  consists of a  refractory lined combustor vessel
which normally operates  at temperatures of 840 to 950°C, pressures of five to
eight atm, superficial velocities  of  1 to 2 m/s and  coal feed rates of 1  to
12 kg/hr.

Equipment

     The unit was described in  detail  in  previous  reports  (1,2,3). Additional
modifications were made to permit  the NOX  control  studies  to be carried  out.
A  brief description of the combustor  section  and  the recent  modifications are
described  below.

Combustor  Vessel —
     The combustor (shown 1n Figure VII-1)  consists of four sections  of 25 cm
diameter standard wall carbon  steel  pipe,  lined  with Grefco 75-28 refractory,
to an inside diameter of 11.4  cm.   The height of the vessel  above the fluid-
Izing grid is about 4.9 m.  Below the grid is a  61  cm burner section  lined
with Grefco Bubblite  refractory.   The preheat burner employs a mixture of
natural gas and air to heat  (and fluidize)  the bed up to temperatures suf-
ficiently  high to  ignite the coal.

     Three sets of vertically  mounted 316 SS water cooled coils are located
inside the combustor  to  assist in temperature control.  These coils remove
50-60% of  the  heat of combustion.

     Two flanged Inlet lines are welded into the final section above the grid
at angles  of  60° to the  horizontal.  These permit the charging of fresh sor-
bent  and the  return of the solids from the primary cyclone.

      The  second and third sections of the combustor have solids draw off lines
Inclined at 60° to the horizontal; these are used to control bed  height and
volume.   Solids drawn off through them are discharged to lock  hoppers.  The
lower  draw off line  is located 1.09 m above the grid and provides a  bed volume
of 0.0111  m3.   It  was used for all tests described in this report.   The  higher
line is located 1.85  m above the grid and provides a bed volume of 0.0189 m3.

                                      171

-------
                                   FIGURE Vll-l
                       SCHEMATIC OF BATCH COMBUSTION UNIT
                  COMBUSTOR
                     SHELL
        SORBENT
          FEED
        HOPPER  X/
                   Refractory
                     Lining
 SUPPLEMENTARY AIR1
TRANSPORT
   AIR
V                                                HOPPER
                                                       COOLING WATER OUT
                                                          OFF GAS PRECOOLER
COOLING WATER IN
          OFF GAS COOLER
   COOLING WATER OUT
                                                                    TO  SCRUBBER
                                                                    TO ANALYTICAL
                                                                        TRAIN
                                                         BACK PRESSURE
                                                SOLIDS    REGULATOR
                                                OVERFLOW
                          BURNER

-------
      The use of these draw off facilities  and  variations  In operating para-
meters permit operating over a  wide  range of  gas  residence times.

Ancillary Systems--
      Coal is fed to the combustor by a  means of  a  pneumatic transport coal
injector capable of feeding up  to 14 kg/hr  for  periods  of  up to  8  hours
before refilling.  Sorbent is charged through a separate system.  The system
consists of two cycling valves  which trap a small amount of sorbent  between
them and discharge the contained sorbent into the combustor intermittently.
Flue gas leaving the combustor  passes through two cyclones, a  cooler and
filter before expansion across  a control valve.

Combustor System Modifications—
      The combustor was modified to permit  the  study of the effects  of  staged
combustion, ammonia injection and simulated flue  gas recirculation on  NOX
emission  levels  (see Figure VII-1).  Modifications  were made  to introduce
supplementary combustion air (for two stage combustion) and ammonia directly
into the  fluidized  bed, or above the bed, during combustor operation.   Nitrogen
(for simulated flue gas recirculation) and ammonia could also be injected into
the inlet air, underneath the grid  plate.

       Probes  used  to  inject supplementary air or ammonia  into the combustor
were constructed of 3/8 inch 316  SS  tubing sealed off at one end.   The probes
extended  horizontally  across the  diameter of the combustion zone to the far
wall,  and contain  three or  four  holes drilled horizontally.  The holes were
sized  and located  so  that  high velocity gas  streams could  be obtained and
directed  to  impinge upon the combustor walls.  This provided adequate mixing
of the incoming  gases  with  the contents of the fluidized  bed.   Injection
locations were varied, depending  upon the  test conditions; they are summarized
in Table  VII-1.

       The reasons  for  the  various probe locations are discussed in  detail in
those  parts  of  the text relating  to the specific program.
       The NH3 addition  system  consisted of a cylinder of NHs maintained at
 about 40°C in a  heated  water bath.   NH3 gas was  transferred through heated
 lines to prevent condensation  and metered through a rotameter.   N2 and H? can
 also be metered  and mixed  with
       Other than for these modifications,  the  bench  unit  combustor remains
 essentially as described  in the  previous report (1).

 Two Stage Combustion

      The program to study the effects of staged combustion upon NOX  emission
 was developed as a 2 X 3 X 2 factorial  design  and included the effects of
 overall excess air, primary air to fuel ratio  (before injection of secondary
 air), and bed temperature (see Table VII-2).

      Initially, it was intended to conduct the designed study at a combustor
 pressure of 8 atmospheres, but operational difficulties prevented this.  At 8
 atmospheres, superficial gas velocities 1n the primary combustion (reducing)

                                      173

-------
                                TABLE VII-1.  LOCATION OF GAS INJECTION
                                 POINTS BENCH UNIT NOX CONTROL STUDIES
          Program
-•   Staged Combustion
    Ammonia  Injection
    Simulated  Flue  Gas
     Recirculation
Injector
   No.
    1
    2
    3

    1
    2
    3
                                                                 Injector Location
Distance Above Grid
        200
         15
         33
Distance Relative to Top of Expanded
                  (cm)

                 78 above
                140 below
                 89 below
Bed
        168                               46 above
      Ammonia mixed with combustion air -  no injectors  used
        290                              168 above

          N  added to combustion air - no  injectors  used

-------
                   TABLE VI1-2.  DESIGNED EXPERIMENT FOR
                 STUDY OF STAGED COMBUSTION IN BENCH UNIT
                                        Bench Unit Combustor Run Number
Primary
Average
Overall
Air (% of Stoich) •>
Bed Temp. (°C) ->
Percent
Excess Air
Unstaged Staged Combustion
Combustion 90% 75%
840 930 840 930 840 930

15
21.1    20.1
21.2    20.2
21.3    20.3
30
17A1    22.1
17A2    22.2
17A3    22.3

-------
 zone were too low for good  fluidization.   This  resulted  In repeated episodes
 of bed agglomeration.  As a consequence,  all  of the  runs  in the designed
 experiment were made at 5 atmospheres.  Champion coal  (-16 mesh, 2.19% sulfur)
 was used in all  runs, and the Ca/S  ratio  was  set at  3.0.

      Originally, it was planned  to  use  the entire  bed  as  the reducing zone
 and to inject the supplementary  air above the bed  to oxidize the CO formed.
 In the Initial  runs at 8 atm,  the supplementary air  probe was located 200 cm
 above the grid,  or about 45 cm above the  top  of the  bed.  Pfizer dolomite was
 selected as the sorbent. This scheme was not practical.  The dolomite
 attrited rapidly and was elutriated from  the  bed early in the run.  This
 resulted in smaller (and shorter) beds, and greater distances between the top
 of the bed and  the point of air  injection.  As  a consequence, the off gas
 temperature at  the point of air  injection was on the order of 600-650°C rather
 than at the bed  temperature of 870-930°C.   This  resulted  in Incomplete burnout
 of CO.  The ignition point  of  CO in air at  atmospheric pressure is about 610°C.
 Occasionally,  there were brief temperature  rises above the air injection point
 indicating that  combustion  of  CO was  taking place  sporadically.  In light of
 the above, 1t was decided to place  the  probe  in  the bed at 15 cm above the
 grid and to use  Grove limestone  (-8+25 mesh)  as the sorbent to reduce attri-
 tion losses.

      These changes provided  adequate  CO burnout, but resulted 1n a limited and
 unstable reducing zone.   During  several  runs  at 8 atm, fusion of the combustor
 bed occurred  as  a consequence  of erratic  coal feed rates and  poor mixing in
 the bed,  particularly at  low primary air  flow rates.   Partial  disassembly of
 the combustor revealed  that  the  refractory lining immeidately above the grid
 had been  eroded.   The diameter was  14 cm  to 15 cm rather than the original  dia-
 meter  of 11.5 cm.   This  resulted in low gas velocities 1n this  area and
 inadequate mixing in  the  bed.  As a consequence, the low velocities and poor
 solids  mixing were probably  responsible for fusion of the bed.   Another pos-
 sibility  was  that CaS was formed and, upon occasion, was  carried into  the
 oxidizing  zone.   The  oxidation of CaS to CaS04 is highly  exothermic, and  would
 result  in  excessive  bed  temperatures and fusion  of the bed material.

       The  refractory  lining was repaired and the diameter of  the combustor
 restored  to the original  11.5 cm.  At this point, the combustor  pressure  was
 reduced  from 8 atmospheres to 5 atmospheres for  subsequent runs,   These
 changes resulted  1n higher gas velocities  and improved mixing of the bed
 solids.  The supplementary air probe was relocated  to a height of 33 cm above
 the  grid in order  to provide a larger reducing zone than  was  obtainable when
 the  probe was located 15 cm  above the grid.

     The results of the study were  evaluated by  Analysis  of Variance (ANOVA)
 techniques for both main effects  and interactions.   If an effect could  not  be
 demonstrated at the 90% confidence  level (or higher), it  was assumed not  to
 exist.

     Emission levels were characterized  as "Emission  Indices" 1n  terms  of
 pounds of the components in  the flue gas per MBTU of fuel  supplied.  This
approach is preferable to using the  component  concentrations  in  the flue  gas,
as the latter is affected by excess  air  levels.


                                    176

-------
     Detailed results  of the  test  program are given  1n Appendix P-2.

Effect of Operating Parameters  Upon  NOx  Emlssions--
     Data obtained at  8 atm_pressure"—The results of the present study were
compared with those from earlier work to measure NOx emissions during unstaged
combustion.  The earlier results were reported  1n a  previous  report  (2).  The
results of the earlier work,  at excess air  levels of 0% to  40%, were combined
with results of four new unstaged  runs  (runs  12.1, 13.1, 14.1, 17.1) made at
8 atm.  A "t Test" with both  paired  and  unpaired comparisons  was used to deter-
mine the effect of staging.  The  staged  tests  (runs  10.1, 11.1, 11.2, 12.2,
13.2, 14.2, 17.2) were carried  out at conditions similar to those  in Table
VII-2.  The test indicated that staged  combustion reduces NOX emissions about
30% below those resulting from  unstaged  combustion  (see Table VI1-3).  The
"t Test" Indicated that the confidence  level  was greater than 90%  when the
unpaired comparison was made  and  over  98% when  the  paired comparison was made.

     Data obtained at 5 atm pressure—The data  obtained at  5  atmospheres —
through the completion of the test matrix in Table  VII-2  — are  summarized  in
Table VI1-4.  Staged combustion reduced the emission Index  (Ibs  N02/MBTU)  from
0.272 (Ibs/MBTU)  (without staging) to 0.245 (Ibs/MBTU) with the  primary  air
at 90% of the stoichlometric amount required to burn the  coal.  This is  a
reduction of only 10%.  A reduction of 20% to 0.220 (Ibs/MBTU) was achieved
by reducing the primary air to 75% of stoichlometric.  The effect  was  observed
at the 90% confidence level.   These reductions are lower than those observed
at 8 atm pressure.

      The  staged  and  unstaged data obtained in this series  were also analyzed
to determine  the  effects of variables other than staging  on NOX  emissions.
For  all  runs, staged  and unstaged, increasing the overall  excess air levels
from 15%  to  30%  increased  the NOX emissions from 0.222 (Ibs/MBTU)  to 0.270
 (Ibs/MBTU),  an  increase of 22%,   Increasing the combustor temperature from
843°C  to  927°C  increased emission Indices from 0,218 (lb/MBTU) to 0.273  (lb/
MBTU),  a  25%  Increase with increasing temperature.

      These results  are  generally  consistent with previous  findings from the
bench  and  minlplant units  (1,2,3).

      No  statistically significant Interactions were found.

 Effects  of Operating  Parameters on  S02  Emissions—
      The S02 emissions observed during  the  staged combustion matrix at 5 atm,
 are  shown 1n Table VII-5.  Staged combustion appears to Increase  the $63 emis-
 sion levels.   The emission index  increased  from 1.66 for all  unstaged runs to
 1.90 (lb/MBTU)  for staged  runs, an  Increase of 16%.  Effects of temperature
 and  excess air on S02 emission levels were  also found.

      High combustion temperatures,  regardless  of staging or  excess  air levels,
 decreased S02 emission levels.   Increasing  the combustor's operating tempera-
 ture from 840°C to 930°C reduced  S02 emission  by 21%.
                                      177

-------
            TABLE VI1-3.  EFFECTS OF STAGED COMBUSTION
                  ON N0  LEVELS AT 8 ATMOSPHERES
                                                         Overall
                           NQX (Ib/M BTU)	     Excess Air (EA)





7
s2
Run No.
12.2
13.2
14.2
17.2
—
«•
Staged
0.14
0.19
0.19
0.21
0.182
0.030
Unstaged(l)
0.192
0.235
0.256
0.317
0.250
0.052
A
0.052
0.045
0.066
0.107
0.067
0.028
(%)
7.8
15.7
19.6
30.6
--
— ^
(1)  Calculated from regression equation  tNOx (lb/M BTU) -
     0.00548 (% EA) + 0.149], for all unstaged combustor runs.
v~
-£•  =  TfTJfr  =  1.374 or 37% increase in NOX levels with unstaged
X~       '                 combustion over the NOX levels with
                          staged combustion.

      r
1  -  37-  »  1  -  0.728  »  0.272 or 27% reduction in NOX levels
      X~c                              when using staged combustion.
                                178

-------
              TABLE  VI1-4.   EFFECTS  OF STAGED COMBUSTION  ON  N0y  EMISSIONS  AT 5  ATMOSPHERES
                                                  Emission  Indices  for NOX  (lb N02/M BTU)
10
Unstaged
Primary Air (% of Stolen) -»- Combustion
Average Bed Temp. (°C) -»• 840 930
Overall Percent Excess Air
15 0.20 0.32
30 0.26 0.31
Sub Column Means = 0.230 0.315
Column Means = 0.272
Group Means
Group 1 (840°C) = 0.218
Group 2 (930°C) = 0.273
Grand Mean = 0.246
Staged Combustion
90%
840 930 840
0.22 0.23 0.15
0.23 0.30 0.25
0.225 0.265 0.200
0.245
Significant
Effect
Temperature
Staging
Excess Air
75%
930
0.21
0.27
0.240
0.220
Effects
Conf, Limits
95%
90%
95%
Row
Means
0.222
0.270

-------
                    TABLE VII-5.  EFFECT OF THE BENCH UNIT'S OPERATING PARAMETERS
                                 UPON S02 EMISSIONS AT 5 ATMOSPHERES
00
o
                                                  Emission Indices  for  SOz  (Ibs/M  BTU)
Primary
Average
Overall


Air (% of Stoich) -»•
Bed Temp. (°C) -»•
Percent Excess Air
15
30
Unstaged
Combustion
840
2.25
1.48
930
2.02
0.890
Staged Combustion
90%
840
2.23
2.16
930
1.79
1.63
75%
840
1.56
2.53
930
1.78
1.56
Row
Means
1.94
1.71
       Sub Column Means =
       Column Means =
1.86    1.45
    1.66
2.20    1.71
    1.95
2.04    1.67
    1.86
       Group Means
       Group 1 (840°C)  =  2.04
       Group 2 (930°C)  =  1.61
       Grand Mean       =  1.82
                	Significant Effects
                	Effect	
                Temperature
                Excess Air-Temperature  Interaction
                Excess Air-Staging  Interaction
                                     Conf. Limits
                                         95
                                         90
                                         95

-------
      Unstaged combustion  at  high  excess air levels  (30%) produced the lowest
S02 emission levels.   The  mean  emission level under  these conditions was 1.18
(lb/MBTU) (the average of  1.48  and 0.890), while the mean emission level for
all other twelve conditions was 1.97  (lb/MBTU), a 40% difference 1n emission
levels.  Combustion under  the condition of high excess air levels (30%) and
high temperature (930°C),  regardless  of staging, materially decreased S02
levels.  The mean SO? emission  level  under these conditions was 1.36 (lb/MBTU),
while the mean for all other  eleven conditions was 1.82  (Ib/MBTU) or a reduc-
tion in emission levels of 25%.
      The effects of the combustor  operating  parameters upon SOg emission
levels may be interpreted In light  of the  following  equations.

                             CaC03  - — »  CaO + C02                        (1)

                                                                          (2)

      The detrimental effect of staging is probably  due to  the need to have a
sufficient oxygen level  to promote  equation 2.   Since oxygen is depleted in
the reducing zone, equation 2 is hindered  there.  Equation  2 will occur in the
oxidizing zone, but because the gas residence time in the oxidizing zone is
less in the staged configuration, equation 2 is further hindered,

      The effects of the higher combustor  operating  temperatures upon SO? emis-
sion levels stems from the fact that the sorbent 1s  more completely calcined
(Eq. 1) at930°C than 1t is at840°C.  The higher the degree of calcination,
the greater the degree of porosity and the more reactive the sorbent.

      The apparent effect of excess air 1n the unstaged runs is  not under-
stood.  Although the increase in excess air could promote  equation 2, it is
believed that at excess levels of 15% or so, oxygen  1s  present in sufficient
excess  to have no further effect.

      Overall, the SO?. emissions measured  in this test series  were  fairly  high,
compared to results  obtained 1n the m1n1plant.  The difference 1s  due to a
much lower gas phase residence time  in the bench studies (-0.5 s)  compared  to
normal  mlniplant operations  (2 to  3  s).

Effects of Operating Parameters Upon CO Emissions--
      Staged combustion Increases CO emission moderately,  and  then only at  low
levels  of primary air (see Table VII-6).   The mean value of the CO emission
Index of 0.292 (Ibs/MBTU) for a primary air rate of 90% of stolchiometrlc
1s  not  significantly different from  the mean value of 0.277 (Ibs/MBTU)  for
unstaged combustion.  However, the mean value of the CO emission indices of
0.335  (Ibs/MBTU) for a  primary air rate of 75%  of stoichlometric Is  20% higher,
which  is significantly  different from  the  value of unstaged combustion.

       CO emission  indices are markedly Increased by low combustion temperatures
an  effect seen 1n  previous  studies  (2).   Decreasing the combustor 's  operating
 temperature  from 930 °c  to 840 °C  Increased  the mean CO emission indices from
                                      181

-------
             TABLE VI1-6.  CO EMISSIONS FOR STAGED AND UNSTAGED COMBUSTION AT 5 ATMOSPHERES
                                                   Emission Indices for CO (Ibs/M BTU)
oo
Unstaged
mary Air (% of Stoich) ->• Combustion
rage Bed Temp. (°C) -> 843 927
Percent Excess Air
15 0.35 0.25
30 0.33 0.18
Sub Column Means = 0.340 0.215
Column Means = 0.277
Group Means
Group 1 (843°C) = 0.355
Group 2 (927°C) = 0.248
Grand Mean = 0.302

90%
843 927
0.29 0.22
0.38 0.28
0.335 0.250
0.292


Temperature
Staging
75%
843 927
0.34 0.29
0.44 0.27
0.390 0.280
0.335
Sianificant Effect
Effect

Temperature, Excess Air Interaction
Staging, Excess
Air Interaction
Row
Means
0.290
0.313



Conf. Limit
99%
95%
95%
95%

-------
0.248 (Ibs/MBTU)  to  0.355  (Ibs/MBTU), an increase of 43%.  Over the range
of 15 to 30% excess  air, the  level of excess air has no significant effect
upon the CO emission indices.  This  finding was unexpected.

      High excess air levels  (30%) and  low combustion temperatures (840=0)
Increase CO emission levels.   At  these  operating conditions the mean CO emis-
sion index was 0.38  (Ib/MBTU)  while  the mean value  for other operation con-
ditions was 0.30 (Ib/MBTU), or an increase of  27%.  This may be the result
of the combined effect of  higher  levels of CO  formation at the lower tempera-
ture and the introduction  of  large amounts of  cold  secondary air  into the bed,
which has the effect of causing  local cooling, thereby preventing CO burnout.

      The combination of high excess air  levels (30%) and  low primary air rates
(75%) also increases CO emission levels.   This effect may  be due  to the  higher
levels of CO formed in the bed at low primary  air  rates  and the introduction
of large volumes of cold secondary air, which  would retard CO burnout.

Conclusions--
      Two stage combustion was shown to reduce NOX emissions by about 20%.
However, S02 and CO emissions were both increased  about  20% by staging.
Increasing the gas phase residence time in  the oxidizing section  of the  com-
bustor slightly should offset the increased $03 and CO emissions.  Two  stage
combustion could also create a boiler tube materials problem  if cooling coils
were subjected to alternate low and high oxygen concentrations.   This  was not
addressed  in  this study but must be at some point, if two stage combustion is
to  be considered further.

NH3  Injection

       Exxon  Research  and  Engineering Company  has developed a process  based on
the  selective, homogeneous,  gas  phase  reduction of NO by NH3.   The amount of
NH3  needed is  comparable  to  the  amount of NO  reduced.  Temperature has an
effect on the effectiveness  of the  reaction.  When the temperature is too low,
NHa  and  NO tend  to  remain unreacted, when the temperature is too  high, NHs
tends to form additional  NO.  Thus, 1t 1s possible to achieve an  efficient
reduction of NO  with little  NH3  remaining, but only within a narrow temperature
range.  The NOX  destruction  and  production reactions are:

                          4NO +  4NH3 + 02 + 4N2 +  6H20                       (1)

                              4NH3 + 502 •»> 4NO + 6H20                         (2)

 Reaction (1) dominates  at temperatures around 950°C whereas reaction (2)
 dominates above 1100°C.

       R. K. Lyon (21) proposed  a free  radical chain mechanism for the reaction
 of NH3, 02 and NO to reduce  NOX:

                              NH2 +  NO  *  N2  +  H +  OH                          (1)

                                NH2  + NO  * N2  + H20                          (2)
                                     183

-------
                                 H + 02 + OH + 0

                               0 + NH3 + OH + NH2

                              OH + NH3 -»• H20 + NH2

                               H + NH3 -> H2 + NH2

 The  addition of a third component such as hydrogen, carbon monoxide or various
 hydrocarbons which form reactive radical intermediates, reduces the optimal
 temperature for the reaction (2).  Hydrogen is preferred since it is itself
 not  an air pollutant.  Although hydrogen reduces the optimal  temperature for
 the  reaction, it has the disadvantage of decreasing the selectivity with which
 NH3  reduces the NO.  Hence, if too much hydrogen is added, NH3 may react to
 form NO rather than to reduce NO.  The preferred levels of the gases based  on
 tests at low pressure are:

                            NH3     NH3/NO = 0.5-3.0

                            H2      H2/NH3 < 3

                            02      0.1-20.0 volume %

      At H2/NH3 ratios around 2/1, the NOX reduction reaction can proceed at
 700°C.  By selecting the proper H2/NH3 ratio, the reaction can be accomplished
 at any temperature between 700°C and 950°C (22).

      Laboratory data has shown that reductions  in NOX emissions  up to 70%  are
 possible.  The process has been commercially demonstrated  in  gas  and oil-fired
 steam boilers and process furnaces.  Until  this  work, no tests have been made
 on a pressurized fluldized bed coal  combustion system.

 Experimental  Conditions—
      A series of eight runs covering 27 conditions were made to  test the effect
 of NHs injection on NOX emissions.   The NH3/NOX  and H2/NH3 ratios were varied
 and 3 injection levels were tested.   Table VII-7 lists the runs which were
made.  Figure VII-2 shows the location of the 3  injection  levels  and Table
 VII-8 indicates the port (and height above the grid)  which was used at each
 level.

      The injection level, or probe  location,  was  set prior to the start of  a
 run.   In each experimental  day,  a  series  of  steady state periods  was  obtained.
 Each series began with a steady  state period at  baseline conditions  (no  NHi
 injection)  and usually ended at  baseline  conditions to see if the emissions
 levels returned to the same base  level.

      Constant operating conditions were  as  follows:

        Champion Coal  (-16 mesh,  2.19%  S)
        Grove Limestone (-8 + 25 mesh)
        Ca/S  mole ratio:   3.0
                                    184

-------
        TABLE  VII-7.  AMMONIA INJECTION RUN CONDITIONS
                                NH3  Injection
Run Number

    1.1
    1.2
    1.4

    3.1
    3.2

    4.1
    4.2
    4.3

    5.1
    5.2
    5.3
    5.4
    5.5
    5.5
    5.7

    6.1
    6.2
    6.3

    7.1
    7.2
    7.3
    7.4

    8.1
    8.2
    8.3
    8.4
    8.5
Locat1on

     LI



     LI


     LI



     L2
                        (1)
                 Approx.
                  Temp.
816
816
816
954
                                                 Conditions
      L3
      L3
      LI
 704
 704
 816
           NWNOx
 No Injection
0.88      0
0.88      2.00

 No Injection
2.27      1.48

 No Injection
2.00      1.00
 No Injection

 No Injection
2.18      2.94
1.27      2.50
 No Injection
3.41      0
8.05      0
 No Injection

 No Injection
 0.82     0
 No Injection

 No Injection
 1.36       0
 2.25       0
 No  Injection

 No Injection
 1.24       0
 2.29       0
 1.39       1.46
  No Injection
 (1)  See Figure VII-2 and Table VII-8 for description of
      each location.
                              185

-------
                        FIGURE VII-2

           BENCH COMBUSTOR SHOWING  LOCATIONS
                  FOR AMMONIA INJECTION
             COMBUSTOR
                SHELL
REFRACTORY
LINING
            AMMONIA
                 •>
                (L3)
            AMMONIA
                (LI)
                COILS
AMMONIA (L2)
                                   1st
   SOLIDS
 OVERFLOW
                              GRID
                    MAIN AIR
                           186

-------
           TABLE  VII-8.   AMMONIA  INJECTION  LOCATION
  Probe                                             Height Above
Location          	Description	            Grid  (cm)

   L3             Inject ammonia near top                290
                  of column

   Ll             Inject ammonia above bed               168
                  (at overflow port)

   12             Mix ammonia with main air          Below Grid
                  stream
                              187

-------
         Pressure:  657 to 758 kPaa
         Bed Temperature:  900°C
         Excess Air:  15%

 Results and Discussions--
       The operating conditions for each of the  runs  is  summarized in Appendix
 Table P-3.  In all, 27 steady state periods were  run which  included 13 periods
 at baseline conditions (with no ammonia injection).  Figure VII-3 shows the
 emission indices for NOX, S0£ and CO,  respectively.  These  bar graphs show
 the 95% confidence interval  for each steady state level.  Standard deviations
 for NOX, S02 and CO emission index, used to draw  the confidence interval (or
 2 standard deviation bars),  were calculated from  the 13 base level steady
 states.

       NOx Em is sums--In the  5 and 7 run series  a  statistically significant
 change in the NOX emissions  index occurred.   Within  each of the run series 1
 3,4, 6, and 8 the confidence intervals overlap,  suggesting that a real change
 in the level  may not have occurred.  Significant  changes in the NOX emissions
 from  the baseline level  occurred  in  runs  5.5, 5.6, 7.2, and 7.3.

       Table  VII-9 summarizes  the  results.  The NOX emission index is  a strong
 function of  the  NH3 injection  location  and NH3/NOX ratio.   With injection  into
 the main air  stream (run  5)  the NOX  level increased  50%.  With injection above
 the bed  (runs  1, 3, 4,  and 8)  the  NOX level remained unchanged.  With  the
 injection  near the top  (runs  6 and  7),  the  NOX level decreased 30 to  50%.
 Table VII-10  shows that  with  injection  into the main air (run 5)  the  NOX'
 level  increased  in proportion  to  the amount of NH3 injected (NH3/NOX  ratio)
 and with injection near  the  top  (run 7)  the NOX level decreased in proportion
 to the amount  of NH3  injected.

       Because  changing the injection location changes the  injection temperature
 the effect of  injection location on the NOX emissions may  be simply a  tempera- '
 ture  effect.   If  this is  true, then it would appear that in  this  combustion
 system,  the NH3-NO  reaction is faster at 700°C,  the NH3-02  reaction is  faster
 at  950°C, and  at  820°C the rates of the two reactions are  equal.   However
 oxygen partial pressures and mixing may be different, as well  as  temperature
 at  each  injection  location.  High oxygen partial  pressures,  which  probably   *
 exist  immediately above the grid, favor NOX production.  This may  explain NO
 production at  injection location 12.  Slugging may cause a  high oxygen  levelX
 above  the bed  (LI) also.

      Table VII-11 compares the results  of this  work to  the  laboratory data
 obtained in another study (21).  The optimal  temperati-e for the NOX reduc-
 tion reaction without hydrogen addition  appears  to have  shifted down from
 950°C in the earlier laboratory study to 700°C in  this work  with the bench
 combustor.  This  shift is possibly due,  in part, to the  presence of CO or
 unburned hydrocarbons in the  flue gas.   CO and hydrocarbons  may act in place
 of  hydrogen as a  source of radicals for  the  NH3-NO reaction.

      Effect on SQj) and CO Emissionsj--Examination  of  the bar graphs for SO?
and CO emission indices (Figure VII-3) reveals that no significant change
occurred due to ammonia injection  location or  NH^/NO  ratio.
                                                0   X

                                     188

-------
                            FIGURE VII-3
        EMISSION INDICES FOR AMMONIA INJECTION PROGRAM
               SHOWING 95% CONFIDENCE  INTERVALS
      0.'
X
LLJ
II
00 \
oo .of
o
z
      0.3
0.2
      0.1
                    NO  EMISSIONS
                       x
X
LLJ
o
Z
 LO \
 LO -Q
 CN
 O
 «, ,
 X
 LLJ
 Q


 ll
 o
 u
           .11.21.4   3.13.2   4.1 4^4.3    5.1 5.25.3 5.4 5^5j5 5J
                  AMMONIA INJECTION RUN  NUMBER
               INJECT ABOVE BED 	*h— INJECT INTO    *\
                                             MAIN AIR
                         INJECTION LOCATION
                                189

-------
                   FIGURE VII-3 (CONT'D)


   EMISSION  INDICES FOR AMMONIA INJECTION  PROGRAM
           SHOWING 95% CONFIDENCE INTERVALS
Q
z
£.  D
O £

Q
Z



Of
00
 CN
Q
i
00
O
u
0.3



0.2


0.1
0.8

0.7

0.6

0.5

0.4

0.3

0.2

0.1
  0
                     NO  EMISSIONS
                        x
        QLJ	1	1	1	1	1	1	1	1	L
                      SO2 EMISSIONS
                      CO EMISSIONS
8
1
           6.1  62 6.3   7.1 72 7.3 7A   8.1 82 8.3 8.4 8.5
           AMMONIA INJECTION RUN NUMBER
              INJECT 	>f<	  INJECT
            NEAR TOP                 ABOVE BED
                    INJECTION LOCATION
                            190

-------
TABLE VII-9.  RESULTS OF PROGRAM TO
  REDUCE NOX BY AMMONIA INJECTION
 Temperature
Signi ficant
Run
Series No.
1
3
4
5
6
7
8
Injection
Level
LI
LI
LI
L2
L3
L3
LI
at Injection
Point (°C)
816
816
816
954
704
704
816
Change in NOX
Emission Index?
No
No
No
Yes
Yes
Yes
No
Increased/Decreased
NOX Levels
—
—
—
Increased
Decreased
Decreased
_ __
Maximum
Change (%)
—
—
—
50
30
50
__

-------
            TABLE VI1-10.  RESULTS OF AMMONIA INJECTION RUNS 5 AND 7
         NH3 Injection
Run        Conditions
No.    NH3/NOX    H2/NH3

          Base (No NH3
           Injection)
5.2
5.3
5.5
5.6
2.18
1.27
3.41
8.05
2.94
2.50
0.0
0.0
                        (Ib NQ2/MBTU)
Base
Level

 0.21
                              Level with
                            NH3 Injection
                            NOX Emissions
                                     Change
                                   From Base
                                   Level (%)
          Significant
             Change
0.24
0.23
0.27
0.32
+14
+10
+29
+52
No
No
Yes
Yes
7.2

7.3
          Base (No NH3
           Injection)
1.36
2.25
0.0
0.0
                    0.18
              0.12

              0.09
-33

-50
Yes
Yes
                                    192

-------
TABLE VII-11 .   COMPARISON OF THE RESULTS  OF THIS  WORK AND  AN
     EARLIER WORK (21) ON AMMONIA INJECTION TO REDUCE NOU
                  Temperature (°C)
Added
NH3
"-
Yes
Yes
Yes
Yes
Yes
Yes
H2
No
No
No
No
Yes
Yes
Earlier
Work (21) This Work (PFBCC -
(Lab Data) Bench Combustor)
1093
954
816
949 704
871
704
Change in NOX
( Increase/Decrease)
Increase i
Increase *
No Change
Decrease 4-
Decrease 4-
Decrease 4-
                             193

-------
       Effect of Excess Air Level on N0y-F1gure VII-4 shows  that  the  excess  air
 level  varied only slightly from a mean of 17.2%.  The two  steady  states  for
 which  the NOX emissions Increased significantly (run 5}  and  the two steady
 states  for which the NOX emissions decreased significantly (run 7) are shown
 1n addition to the 13 base level steady state periods.  Figure VII-4  shows
 that the change in NOX emissions due to NH3 Injection 1s not due  to a change
 In the  excess air levels.  The uncontrolled NOX emissions  1n this study  are
 generally lower than In earlier bench combustor studies  (2). This 1s pre-
 sently  unexplained.

      Response Time—Figure VII-5 shows graphically  how  response times were
 determined for run 5.  The response time 1s the time required for the system
 to reach a new steady state NOX level  once a change  1n the ammonia Injection
 conditions (NHs/NOx and Hg/NHs) is made.  Table VII-12 shows  the response
 times for run 5.  The system response is very rapid, averaging 10 minutes.

 Ammonia Material Balance—
     NHs material  balances were attempted by analyzing all  solid  and  gaseous
 product streams.  The balances were very low and variable,  ranging from  14 to
 69%.  Analysis of sol Ids collected by the cyclones and by  the final filter
 Indicated that ammonium salts were forming and precipitating on the collected
 sol Ids.  The compounds were most likely NfyHSOs and/or NfyHSO^  The  mass of
 ammonia which would have to have been lost to account for  the low mass
 balances was very little, ranging from 4 to 130 gm.   Therefore, the poor
 balances could have been caused by the formation and deposition of ammonium
 salts,  possibly 1n the flue gas ducting.  Other sampling and analytical
 problems may have also occurred which contributed to the low balances.

 Conclusions—
      NHj injection at temperatures around 700°C in  PFBC flue gas can reduce
 NOX emissions 30 to BQ%.  The temperature level 1s lower than expected and
 further work may be needed to determine 1f it is the true  optimum.  If such
 a temperature is required, NH3 could be Injected ahead of  the gas turbine.
 The desired temperature would occur within the turbine Itself, where  the NO
 removal reactions  would take place.

 Simulated Flue Gas Recirculation

      In the simulated flue gas redrculatlon (SFGR)  program, nitrogen was
mixed with the main air stream to simulate flue gas  reclrculatlon.  The
 variables studied  were:

      Recirculation Ratio, R * N2/A1r  * 0.10, 0.20
      Average bed  temperature = 840, 900°C
      Excess A1r = 15, 30%

 Constant operating conditions were:

      Pressure = 808 kPaa
      Ca/S mole ratio =3.0
      Champion coal  (-16 mesh, 2.19 wt.  % S)
      Grove limestone (-8  + 25 mesh)


                                   194

-------
                        FIGURE VII-4
                AMMONIA INJECTION PROGRAM:
        EFFECT OF EXCESS AIR LEVEL ON NOx EMISSIONS
U.H
0.3
S
Y~
co
0 n o
Z O'2
_Q
X
O
Z
0.1

0
1 1 1
A
A

• •
••
•»o»-
4h V ^k
v^
•*A\
A \
O
MEAN OF THE BASE LEVEL
NH3 DATA:
(EXCESS AIR: 17.2%)(NOX: 0.17 Ib/MBTU)
i i i
AMMONIA
INJECTION
                   10         20

                        EXCESS AIR(%)
                             30
40
BASE LEVEL (No NH3 INJECTION)

WITH NH3 INJECTION (RUN 7)
WITH NH3 INJECTION (RUN 5)
                              195

-------
a.
a.
LU
o
z
                          FIGURE VII-5
                         RESPONSE TIME
                    CHANGE IN
                OPERATING VARIABLE
       STEADY-STATE NO. 1
STEADY-STATE  NO. 2
                            RESPONSE
                              TIME
                           TIME (MIN)
                               196

-------
TABLE VII-12.   RESPONSE  TIME  FOR AMMONIA  INJECTION RUN 5
Run
No.
5.1
5.2
5.3
5.4
5.5
5.6
5.7
Operating Variables
Subject to Change
NH3/NOX
H2/NH3
Response
Time
(min)
Base Level (No Injection)
2.18
1.27
Base
3.41
8.05
Base
2.94
2.50
Level (No Injection)
0.0
0.0
Level (No Injection)
10
10
20
7.5
5
7.5
NOX Level
(ppm)
145
210
200
160
210
250
170
                                Avg. = 10
                           197

-------
 The  N£  flow system used to cool down the combustor during an emergency shut-
 down was used to inject N£ to simulate flue gas recirculation.  The No is
 mixed with the main air stream far upstream of the injection point and there-
 fore, perfect mixing may be assumed.

 Results and Discussion--
      The run conditions for the nine steady state periods are summarized in
 Appendix Table P-4.  In each run, a steady state period was maintained without
 nitrogen addition and then in a second period, nitrogen was added  without
 adding additional air.  However, additional  coal  had to be burned  to  heat the
 nitrogen which was injected cold.  Therefore, the excess air levels reported
 for  the second period are all lower than for the first.  Figure VII-6 shows
 the  bar graphs for each emission index (E.I.) for N02, S02, and CO, and the
 95%  confidence interval for the emission levels.  The error bars shown on
 the  graphs were drawn using the standard deviations  derived from a  statistical
 analysis of the ammonia injection program data.

     Table VII-13 summarizes the results of the SFGR program,  Table VII-14
 indicates the significance of changes in the emissions for each run.   Signi-
 ficant reductions in the NOX emissions, observed in  runs 2 and 4,  are cor-
 related with changes in the excess air level  which accompanied N2  addition
 during these tests.  Increases in both the S02 emission index  and  CO emission
 index are correlated to changes in the excess air level and the gas residence
 time.  These changes in the NOX, SC^ and CO emissions are discussed below.

     NOy Emissi'ons--Figure VII-7 compares the NOX emissions vs.  excess air
 results for the SFGR program with the baseline NOX emissions from  the ammonia
 injection program (no NH3 injection).  Table  VII-15  shows that the  mean of the
 SFGR data (all data) nearly coincides with the mean  of the ammonia  injection
 data (base levels only).  Also, because NOX  emissions decrease with decreasing
 excess air levels,  it is believed that the lower  NOX emissions which  resulted
 from N2 addition in these tests are really the result of lowering  the excess
air level.   If the  test had been conducted in a way  which would  have  prevented
an accompanying decrease in excess air with  N^ addition (e.g., by  increasing
the air rate), it is expected that no change  in NOX  emissions  would have been
observed.   Therefore,  flue gas recirculation  should  not have an  effect on
reducing NOX  emissions.

     SO? and  CO Emissions—As seen in Table  VII-13,  both S02 and CO emissions
 increase as a  result of N2 addition.   Although the increase in SO?  emissions
could not be  shown  unequivocally to be statistically significant (Table VII-14),
it was consistent.

     Adding nitrogen to  the main air  has  two  effects:

     1.   The  partial  pressure of oxygen is  lowered.

     2.   The  gas  velocity is  increased, which lowers  the residence
         time  of the combined gas  stream.
                                    198

-------
                          FIGURE VII-6

EMISSION  INDICES FOR SIMULATED FLUE GAS RECIRCULATION (SFGR)
        PROGRAM SHOWING 95% CONFIDENCE INTERVALS
X
LLJ
Q
Z

Z S
O 2
      x
     O
     Z
     Z =
     LLJ

      CN
     Q
     Z
           o..
           0.3
        ca
      CO
      00
           0.1
            2.4
            2.0
            0.8
            0.7
       0.6
       0.5
       0.4
   o  0.3
       0.2
      8
                        NO  EMISSIONS
                           X
                          _|	I	|	 |	i
                    SO2  EMISSIONS
             m

                           P
        i
                    CO  EMISSIONS
^
^i

                                       g
                 VA
i
                    i
               1A.1  1A.3   2.1     3.1      4.1
                  1A.2        2.2      3.2       4.2

                        SFGR RUN NUMBER
  Note: The  standard deviation values used to draw  the error bars are  from
        a  statistical analysis of the  Ammonia Injection  Program data.

                             199

-------
TABLE VII-I3.  SIMULATED FLUE GAS RECIRCULATION
          PROGRAM - SUMMARY OF RESULTS
SFGR
Run
No.
1A.1
1A.2
1A.3

2.1
2.2

3.1
3.2

4.1
4.2

Independent Variables
R
(%) Bed Temp. (°C)
874
11.2 886
S82
Average = 881
874
23.4 887
Average = 881
906
11.2 897
Average = 901
819
11.3 817
Average =818
Excess
Air (%)
18.6
12.9
17.2

18.4
7.4

26.5
25.0

16.7
13.0

Dependent Variables
Emission
NOx
0.19
0.16
0.19

0.22
0.07

0.20
0.17

0.29
0.13
Avg =
0.18
Index (1
S02
0.89
1.17
0.72

0.81
1.09

1.01
1.07

1.18
1.72
Avg =
1.07
b/MBTU)
CO
0.14
0.34
0.33

0.17
0.27

0.14
0.45

0.29
0.56
Avg =
0.30
                      200

-------
TABLE VI1-14.  SIGNIFICANCE OF CHANGES IN EMISSIONS -
      SIMULATED FLUE GAS RECIRCULATION PROGRAM
                   Emission Index
(Ib/MBTU)
Run No.
-v
lA.ll
1A.3J
1A.2
2.1
2.2
3.1
3.2
4.1
4.2


1A.2
2.1
2.2
3.1
3.2
4.1
4.2

lA.fl
1A.3J
1A.2
2.1
2.2
3.1
3.2
4.1
4.2

Base
Level

0.190

0,216

0.203

0.293


0.804

0.813

1.014

1.177


0.237
0.170

0.135

0.292


Level Change
With N? From Base
Addition Level (%)
NOX Emissions

0.164

0.068

0.173

0.129
S02 Emissions

1.168

1.094

1.066

1.722
CO Emissions
0.341

0.274

0.446

0.562
201


-14

-67

-15

-56


+45

+35

+5

+46

+44

+61

+230

+93

Significant
Change?


Maybe

Yes

Maybe

Yes


Maybe

Maybe

No

Maybe

No

No

Yes

Yes


-------
                               FIGURE VII-7

       COMPARISON OF THE SIMULATED FLUE GAS RECIRCULATION (SFGR)
             PROGRAM WITH THE AMMONIA INJECTION PROGRAM
        i—
        CQ
       o
       z
           0.4
           0.3
       O  0.2
           0.1 -
                                    O Mean of all of the
                                    SFGR data: (Excess Air:
                                    17%)(NO : 0.18
                                    lb/MBTU)X
                                    I
                             I
SFGR PROGRAM
SFGR PROGRAM
AMMONIA
INJECTION  PROGRAM
      10         20         30

            EXCESS AIR (%)

(BASE LEVEL)
(WITH N2 ADDITION)
(BASE LEVEL)
                                                         40
                                 202

-------
     TABLE  VI 1-1.5.   COMPARISON OF AMMONIA  INJECTION  AND
         SIMULATED  FLUE  GAS  RECIRCULATION RESULTS
                                      Mean Value
                        Excess Air (%)       NOX (lb NOT/MBTU
Ammonia Injection            17 9                  n 17
(Base Levels Only)           "'*

Simulated Flue Gas
 Recirculation               17.3                  0.18
(All Data)
                              203

-------
     The first effect, the lowering of the oxygen partial  pressure, may reduce
both the rate of reaction of SOg to form CaS04 and the rate  of  reaction of CO
to form C02-  As was mentioned earlier, some excess oxygen is required for
the reaction of S02 with CaO according to the reaction:

                           CaO + S02 + 1/202 -»• CaS04

Therefore, reducing the oxygen partial pressure to low levels,  may decrease
the S02 retention and increase the S02 emissions.  CO reacts with 02  in the
gas phase according to the reaction:

                               CO + 1/202 -*• C02

Similarly, decreasing the oxygen partial pressure substantially will  decrease
the rate of CO combustion and increase CO emissions.

     The second effect, the lowering of the gas residence  time, means that
there is less time for both of the above reactions to occur. Hence,  adding
Ng to the gas stream means that the rate of each reaction  and the time
available for each reaction to occur are less.  Therefore, both the S02 and
CO emissions increase with N2 addition as shown by these runs.

      For each steady state period, the sorbent feed rate  was adjusted to
maintain the Ca/S mole ratio at a constant value of 3.0.  However, Table
VII-16 shows that the Ca/S ratio varied about the average  of 3.0, ranging
from 2.65 to 3.36.  In two runs, Ca/S increased when N2  was  injected  and in
two runs Ca/S decreased when N2 was injected.  This variation was not system-
atic; Ca/S did not change in proportion to the amount of N2  injected  into the
main air.  Therefore, any systematic changes in S02 emissions would not be
expected to be caused by changes in the Ca/S ratio.

     Conclusions—-N? addition to the main air stream has no  direct effect on
NOX emissions.  The observed decrease in the NOX emissions in this program
can be explained by changes 1n the excess air level; No  addition lowered the
percent excess air which in turn lowered the NOX emissions.  If the excess
air level had been held constant, no change in NOX emissions would have been
expected.

      N2 addition results in increased S02 and CO emissions  probably  because
of lowered oxygen partial pressure and the gas residence time.

Combined Techniques

      A series of tests was conducted to determine if a  combination of NOX
control techniques could result in further reductions in NOX emissions.
Since only staged combustion and ammonia injection caused  a  reduction in NOX
emissions, only the combination of these two methods was studied.

      Two runs were completed and are discussed below.  In run  1, two steady
states were run at base level conditions, two with NH3 Injection, and then
one with staged combustion.  No sulfur retention data were obtained because
the S02 analyzer was inoperable during this run.  In run 2 one  steady state

                                      204

-------
TABLE VII-16.  VARIATION  IN  Ca/S MOLE  RATIO
Run No.
1A.1
1A.2
1A.3
2.1
2.2
3.1
3.2
4.1
4.2
Without
No Addition
(Base Level)
2.97
3.10
3.36
3.05
2.55
With
N2 Addition
3.10
2.85
2.87
2.87
Change in
Ca/S (%)
+2.1
-15.2
-5.9
+8.3
             Average =2.98
                      205

-------
was  run  at  base level conditions, one with staged combustion, one with both
NH3  injection and staged combustion and then one more with staged combustion
only.  The  results are summarized in Appendix Table P-5.  The constant opera-
ting conditions for both runs were:

                              Pressure     505 kPa
                          Temperature     900°C
                                 Ca/S     3.0
                           Excess Air     25%
                                 Coal     Champion, -16 mesh, 2.19% S
                            Limestone     Grove, -8 + 25 mesh
      Ammonia injection probe location     290 cm above the grid
      Supplementary air probe location     33 cm above the grid
            Primary Air as % of Stoich     75% (2-stage combustion)
                               NH3/NOX     1.5 (ammonia injection)

The values of the staged combustion and the NHs injection variables are those
found to be promising in the previous testing.

      Table VII-17 shows the changes in NOX levels in each run as changes
were made from base level  to various NOX control  and combination of NOX con-
trol techniques.

      Table VII-17 shows that in this work, either ammonia injection or staged
combustion or a combination of the two techniques tend to lower NOX emissions
by 15 to 50% below the base level.  The reduction in NOX emissions by staged
combustion 1s reasonably consistent with the earlier study at 505 kPa in
which reductions of 10 to 20% were observed.   The reduction in NOX emissions
by ammonia Injection at 505 kPa with an NH3/NOX ratio of 0.89 to 1.43 is
consistent with the earlier ammonia injection program at 808 kPa in which  a
reduction of 33% was observed with an NH3/NOX ratio of 1.36.  Figure VII-8
shows that the reduction 1n NOx emissions from the base level when both
techniques were used simultaneously was not greater than the reduction when
either technique was used alone.  However, since only one data point was
obtained when using the two techniques simultaneously, 1t cannot be stated
with any certainty whether an interaction does or does not exist.  Perhaps,
if a larger NH3/NOX ratio had been used, a greater reduction in NOX may have
been achieved.

      S02 and  CO emission  data  were  obtained  for  run  2.   Table  VII-18 shows
that the  S02 emissions  Increased from 0.87 to  1.03  Ib/MBTU  going  from base
level conditions (run  2.1)  to staged combustion (run  2.2) and the  CO  emis-
sions were unchanged at  0.26  Ib/MBTU.   However, over  the  entire run  both the
S02 and  CO emissions increased; runs 2.2 and  2.4  were both  with staged  com-
bustion  only,  but  the  S02  emissions  Increased  from  1.03  to  1.22 Ib/MBTU and
the CO emissions Increased  from 0.26 to 0.35  Ib/MBTU.  These  upward  drifts  in
the emission levels are  unexplained.

BENCH REGENERATION STUDIES

      The bench  unit regenerator,  because  of  its  relatively small  size, pro-
vides a  means  to experiment with operating conditions  and methods  quickly

                                    206

-------
                   TABLE  VII-17.
COMBINED NOX-CONTROL TECHNIQUES PROGRAM.
 CHANGES IN NOX LEVELS.
Run
No.
1.1
1.2
1.3
1.4
1.5

2.1
2.2
2.3
2.4

Excess
Air («)
25.1
22.6
25.2
23.3
22.0

31.3
24.5
25.3
24.2

Average Bed
Temp. (°C)
891
896
894
900
901
Avg. - 896
876
883
887
883
Avg. = 882
Control
NFh Injection
( Yes/No J(NH3/NOX
No
Yes 0.89
No
Yes 1 .43
No
No
No
Yes 1 .61
No
Technique
Staged
Combustion
) (Yes/No)
No
No
No
No
Yes
No
Yes
Yes
Yes
NOX
Mb/MBTU)
0.21
0.15
0.15
0.09
0.10
0.17
0.14
0.13
0.15
                                                                                        Change  in  NOX
                                                                                          from  Base
                                                                                        Level  (1)  (%)
                                                                                            -15


                                                                                            -49

                                                                                            -43
                                                                                            -21

                                                                                            -26

                                                                                            -15
(1)   Base level  NOX emission = 0.18 Ib/MBTU
     (Average of Runs 1.1, 1.3, 2.1)

-------
Z
g
fc
^>
CO
5
o
(J
Q
LLJ
O
YES
    NO
                           FIGURE VII-8

             REDUCTION  IN  NOX EMISSIONS BY USING
            TWO CONTROL TECHNIQUES SIMULTANEOUSLY
         0.21 )
    —    0.15>
         0.17/
                 AVG. =0.13
                        NOX
                   (Ib NO2/MBTU)
AVG. =0.18
(BASE  LEVEL)
                  NO
                            0.13
O.is)
0.09)
                                                         AVG. =0.12
                                               YES
                               NH3 INJECTION
                              208

-------
              TABLE  VII-18.  S02 AND CO EMISSIONS  FOR
               COMBINED  NO  -CONTROL TECHNIQUES RUN 2
Run                                        Emission  Index  (Ib/MBTU)
No.            Control  Technique           S02CO
2.1         Base level                      0.87                 0.26
2.2         Staged-combustion              1.03                 0.26
2.3         Staged-combustion              1.15                 0.32
            with ammonia injection
2.4         Staged-combustion              1.22                 0.35
                                 209

-------
 and  Inexpensively.   Promising results obtained in the bench regenerator can
 then be  scaled up to miniplant regenerator operation,

      The test program in the bench regenerator was divided into two segments,
 natural  gas fueled regeneration and coal fueled regeneration.  Most of the
 work concentrated on developing satisfactory operating techniques.  The work
 done with natural gas-fueled regeneration included a study of the most
 effective way to introduce fuel to the regenerator and a limited study of
 operating variables and their effect on sorbent regeneration.  Studies made
 with coal-fueled regeneration were preliminary 1n nature and were intended
 to determine the feasibility of operating a regeneration unit with coal.
 The  incentive to use coal instead of natural gas is to reduce operating costs
 and  to reduce reliance upon gaseous (or liquid) fuels.

      The results of the bench unit regenerator test program will  be discussed
 in the following sections.

 Equipment

      The batch regenerator unit was modified to permit semi-continuous opera-
 tion and renamed the bench regenerator unit.  The modified facilities were
 discussed in the previous annual  reports (1,2) and will  not be described in
 detail.

      The regenerator vessel  is 4.57 m high and.1s constructed of 12 inch car-
 bon  steel pipe and lined with Grefco 75-28 refractory to an inside diameter
 of 9.52 cm.  The burner plenum chamber below the grid 1s 0.69 m high and is
 lined with Grefco Bubblelite refractory.

     Sulfated sorbent is intermittently charged to the regenerator from a
 lock hopper via a line equipped with two cycling valves to meter in the sor-
 bent  (see Figure VII-9).  The sorbent is Introduced near the bottom of the
 fluid bed (9.5 cm above the grid)  to insure complete mixing.  Sul fated sor-
 bent from miniplant operations is  used.  Regenerated sorbent exits the bench
 regenerator through a bed overflow line which leads to an overflow lock
 hopper.  The below-bed burner is  used to heat up the bed and provides some
of the energy required by the regeneration reactions.  A reducing zone is
established immediately above the  grid by adding supplementary fuel  directly
to the bed.  This creates the CO  and H2 rich zone needed to carry out the
regeneration reactions.  Supplementary air is then added further up the bed
This creates a mildly oxidizing zone at the top of the bed which oxidizes
any undesirable CaS formed in the  reducing zone.   All remaining fuel  is also
burned 1n the oxidizing zone, to maximize fuel  u^e efficiency.

Natural  Gas-Fueled Regeneration

Operating Procedures--
     The Initial  charge of bed material  is heated to the desired temperature
under oxidizing conditions by operation of the below-bed burner.  Sometimes,
small amounts of supplementary fuel  are added directly to the bed to Increase
bed temperature,  but not enough is added to produce reducing conditions 1n
the bed.   Supplementary fuel  can  only be added when bed  temperature is above
650°C, the temperature at which natural  gas will  ignite easily.

                                   210

-------
                              FIGURE VII-9
                        BENCH REGENERATION UNIT
SORBENT
 FEED
HOPPER
      AUTOMATIC
        VALVES
      FLUIDIZING GRIDH-
                            SUPPLEMENTARY
                            AIR
                                    COOLING
                                    WATER IN
   f
               BURNER
          COOLING WATER OUT
                OFF  GAS PRECOOLER

           COOLING  WATER  IN
                                                        COOLING
                                                       WATER OUT
                                                     -OFF GAS COOLER
         BACK  PRESSURE   Tr.
           REGULATOR     IU
                      SCRUBBER
 OFF
 GAS
FILTER
                                                                  TO
                                                             ANALYTICAL
                                                                TRAIN
SORBENT
OVERFLOW
HOPPER

-------
     When the bed temperature 1s uniform and close to the  desired  run  tempera-
ture, the switchover to reducing conditions is made.   This 1s  accomplished
by Increasing flow of supplementary air to the required value  and  then
Increasing the flow of supplementary fuel.  Supplementary  air  flow is  always
Increased before supplementary fuel to avoid adding air to a column already
filled with a reducing gas.  Temperature is continuously monitored and flow
rates of burner air, burner fuel, supplementary air,  and supplementary fuel
are all adjusted to yield the desired bed temperature.  Oxygen and CO  con-
centrations in the off gas are also monitored and the supplementary air flow
is corrected to produce low concentrations of CO (under 5000 ppm).  The flow
rate of supplementary air 1s the minimum value which  will  just produce CO 1n
the oxidizing zone.

Shakedown-
     Shakedown of the bench regenerator was completed 1n two runs.  During the
first shakedown run, all  systems except the sorbent feeder operated success-
fully.  The opening of the feed line in the regenerator was found  to be par-
tially blocked with fused bed and refractory, which had been left  from an
earlier batch run.  The next run proved to be more successful.  The sorbent
feed system functioned properly and the high temperature necessary for regen-
eration was achieved using the air preheater.

     The main operational difficulty during shakedown runs was maintaining a
satisfactory bed height.   Although the solids overflow port is 1.0m above
the grid, the highest expanded bed height obtained was under 0.5 m. It is
believed that the bed slugged badly and the solids were moving, piston-Uke,
up the regenerator and out the sorbent overflow due to high superficial  gas
velocities.  This was a continuing problem during the test program. The
effect was partly offset by minimizing the gas velocity.

Effect of Fuel Injection Mode-
     One objective of the work done with natural gas-fueled regeneration was
to determine the best way of introducing the natural  gas.   The energy
required to maintain the temperature of the unit can  be supplied in two ways.
The burner below the fluidizing grid can be operated  at oxidizing  conditions
and supplementary fuel added to the bed to create the reducing zone (Mode 1),
or all fuel can be added directly to the bed where it would partially  burn In
the fluidizlng air (Mode 2) (see Figure VII-10).  In  both  modes of opera-
tions, the reducing gas 1s produced by the Incomplete combustion of methane
which 1s added directly to the bed through the supplementary fuel  probe.
Adding all the fuel to the bed eliminates the heat losses  associated with
operation of the burner, thereby decreasing the fuel  requirements  and  pos-
sibly increasing the S02 content in the off gas.  One drawback to  adding all
the fuel directly to the bed is that the entering cold fuel must heat  up
to Ignition temperature before it burns.  The heat up requires a certain
residence time in the bed, causing the methane to move further up  the  bed
before it Ignites.  As the level of combustion varies, the location of the
reducing zone also varies which might have a detrimental effect on tempera-
ture control and on regeneration.
                                    212

-------
                            FIGURE VII-10

                        FUEL INJECTION MODE
      BED	
FLUIDIZING
   GRID
    SUPPLEMENTARY
         AIR
    SUPPLEMENTARY
         FUEL
    BURNER	(j-*
            AIR —
FUEL
SUPPLEMENTARY
     AIR

FUEL
             MODE NO.  1
                  MODE NO.  2
                                  213

-------
      Burner  section heat 1ogses--In determining the best mode of Introducing
 fuel  Into  the regenerator, it was observed that a significantly smaller
 amount of  fuel was required to maintain bed temperature when all  the fuel
 was added  directly to the bed as opposed to operation with the burner (e.g.,
 0.079 compared to 0.045 NnvVmin).  To account for this discrepancy in energy
 requirements, the heat losses associated with burner operation were deter-
 mined.   Losses Include radiation and convection of heat from the burner
 plenum,  and  heat loss to the cooling water in both the fluidizlng grid and
 the burner head.  These heat losses were calculated and are summarized below:

      Radiation and convection from plenum:

       Q1  =  AkAT  =  0.63 m2 x 21.8 JM'W1 (533-300) K

                    =  3200 Js"1

      Flu1d1zat1on grid cooling water:

       Q2  =  mCpAT  =  4 x 3.17 x 10"2kg sec"1 (4.187 x 103Jkg~1K~1)(306-294)K
                     =  6364 Js"1

      Burner cooling water:

       Q3  =  mCpAT  =  3.17 x lO'^gs"1 (4,187 x  103Jkg~V1)(328-294)K
                     =  4508 Js"1

     Total  heat loss   =   Q1  + Q2 + Q3  -  14072 Js"1

     where  Q  =  heat loss
            A  =  area
            k  =  combined heat transfer coefficient for  radiation  and
                  convection
           AT  s  temperature differential
            m  =  mass flow  rate
           Cp  =  heat capacity

     The heating value of methane is  3.73 x  10 Jm" .  Therefore,  the  heat
loss associated with  burner  operation 1s equivalent to the  heating  value of  a
flow rate of 3.78 x  lO-^nrs-"1  of methane.  Thus,  the difference 1n  fuel
requirements between  operation with and without the burner  can  be  largely
explained by the heat losses  associated with burner operation.  Since about
0.079 Nm3/m1n of methane  are required to operate  the regenerator with the
burner,  a significant portion (about  30 percent)  of the fuel  requirement Is
used to  overcome heat losses.

Results  and Discussion—
     Mode No. 2 operations—Six runs  were made 1n  the bench regenerator to
evaluate operation in which  all  the fuel (methane)  is added directly  into the
bed.  In five of the  runs, supplementary fuel  was  injected  at the wall of the
reactor.  In the sixth run (R-8), a new supplementary fuel  probe designed to
give better fuel  distribution across  the bed was  used.  Operation with all of
the fuel  added to the bed  generally resulted in an  uniform  temperature pro-


                                     214

-------
file throughout the bed  but  controlling the temperature to the desired level
proved difficult.   During  the  series of six runs, only one steady state regen-
eration condition  was  reached.  The steady state was a 15 minute segment of
run R-6.  Operating conditions  for run R-6 are presented in Table VII-19.

     Many of the regenerator runs in which all the fuel was added directly
Into the bed were ended  by high temperature shutdowns.  Controlling the tem-
perature proved difficult  when mode No. 2 was used.  The factors which might
cause a temperature runaway  are:

     1.  While in oxidizing  conditions during heat up:

         a.  an increase in  fuel  flow, thereby adding additional energy input
         b.  a decrease in air flow thereby increasing the temperature of
             the preheated incoming gas

     2.  While in reducing conditions:
         a.  a decrease in fuel flow  causing  oxidizing conditions and
             oxidation of CaS  (an exothermic  reaction)
         b.  an increase in air flow  causing  oxidizing conditions

     3.  In either reducing or oxidizing  conditions:

         a.  a decrease or stoppage in  solids feeding
         b.  poor mixing, including fuel  bypassing,  air  bypassing,  and
             poor solids circulation, causing hot spots  and  agglomeration

Possible reasons for the problems of 1n-bed  fuel  injection  into the regenerator
were examined and it is believed that the cause is related  to mixing.  In
order  for  fuel addition into the bed  to  work, the fuel must  be heated above
Its ignition temperature shortly after injection into  the bed.  If  the  cold
fuel entering the regenerator through the probe does not mix with  the hot bed
solids  and is not brought up to Ignition temperature,  the "flame" may be
blown  out.  It may be that the large, sudden, changes  in temperature were
caused  by  ignition or extinction of the methane flame.   These changes could
have been  brought about by shifts 1n the pattern of solids  mixing within the
fluidized  bed.

     The hypothesis that poor solids  and gas  mixing results  in "flame blow
outs"  is supported by experimental  evidence.   During many of the runs,  it was
observed that the Og concentration in the off gas Increased  suddenly, although
no change  in operating conditions (fuel  or air flow rates)  had occurred.  The
sudden increase in the 02 concentration in the off gas would be characteristic
of the  effects of a flame blow out.  The increase of 02 concentration, since
the air was  no longer used  1n combustion, led to oxidation of the CaS present.
an exothermic reaction.  The exothermic reaction resulted 1n high bed tempera-
tures  and  the subsequent  high  temperature shutdowns.

      It 1s possible that  the problem of a flame out could have been lessened,
or even eliminated, by  heating to higher temperatures the air entering the
regenerator through the fluidizlng grid.  This would have reduced the heat
transfer requirement  produced  by solids backmixing.  Unfortunately, with the
equipment  at hand,  it was possible to heat the incoming air only to  150-
200°C.

                                      215

-------
(M
                             TABLE  VI1-19.   BENCH REGENERATOR RUN SUMMARIES


                Avg.                   Sup.   Gas     /?\   Solid                                Extent
                 Bed         Expanded   Gas    Res.  ^  '   Res.   S02; Mole %                  of Solids
Run
Number
R-eO)

R-9

R-10

R-ll.l

R-11.2

R-12

R-13

R-14

Temp.
K
1370

1340

1350

1350

1350

1345

1395

1360

(1) Mode No.
(9\ * = <;tr
P
Bed
atm Height m
6.9

7.5

7.6

7.6

7.6

7.6

7.6

7.5

0.9

0.6

0.7

0.6

0.6

0.8

0.6

0.6

2 Operation (al
lirMnmetr-ir air
Vel.
m/s
0.7

1.1

1.0

0.9

0.9

0.9

1.2

1.1

1 fuel
tn fm
Time
sec.
1.34

0.56

0.70

0.66

0.70

0.89

0.50

0.55

added
al vati
Equiv Time
Ratio min.
1.43

1.42

1.64

1.85

1.83

1.50

1.57

1.45

to bed,
irt v ^US
29

48

35

43

29

51

22

37

burner
Average
Peak
(Range) Equil
1.17
(0.24-1.74)
0.62
(0.18-0.99)
0.58
(0.18-1.17)
0.59
(0.09-1.32)
0.24
(0.06-0.92)
0.27
(0.03-0.66)
0.53
(0.1-1.35)
0.43
(0.15-0.84)
not used).
0

0

0

0

0

0

0

0


% S02
. % S02
.25

.25

.25

.29

.20

.15

.15

.15


Regeneration
%
NA

NA

44

71

71

48

83

79


             NA =  Data not available.

-------
     Also, 1n a larger bed  with  improved  solids mixing, the difficulty in
maintaining the flame should  be  lessened  and  in-bed  injection of the fuel
should be feasible.

     Mode No. 1 operations—The  alternative to in-bed fuel injection, as
described previously, is operation of the burner  supplemented by in-bed fuel
injection.  The disadvantage  of  this  method of operation  is the increased
heat losses which accompany burner operation; increased heat losses require
additional air and fuel inputs thereby diluting the  off gas SOg concentra-
tion if thermodynamics is not limiting.  An advantage of  this method is that
the burner head serves as a flame holder, an  element that was evidently
lacking in the previous method described. The combustion gases rising from
the burner also provide the heat necessary to ignite the  incoming  supple-
mentary fuel, eliminating the problem of flame blow  outs.

     Six  successful runs were made with the  burner in operation while supple-
mentary fuel is added above the fluidlzing grid  to produce a  reducing zone.
The results of these six runs, which represent  the variable  study, are pre-
sented in the next section.

     Results of process variable study--0perat1on with  both  burner and supple-
mentary fuel resulted  in a steadier and more easily controlled  bed tempera-
ture.  The important operating conditions for runs R-9  through  R-14 are  sum-
marized in Table VII-19.   It was intended to determine  the effect of operating
variables on SOg concentration in the off gas and extent of sorbent regenera-
tion.  However, because of the data scatter and the limited  amount of data
obtained, it was not  possible to draw conclusions about the  relationships
between operating variables and sorbent regeneration.

     S02  concentration  in  the regenerator off gas can be limited by thermo-
dynamics, kinetics or  the  heat and mass balance.  The SOg concentration  from
the  bench unit, as in  the  case of the miniplant, is limited  by the heat  and
mass  balance.  The operating  variables in Table VII-19 (temperature, pressure,
gas  residence  time,  solids residence time, and equivalence ratios) are either
thermodynamic  or kinetic variables used to define the system.  Since the
system  is heat and mass  balance limited,  no  conclusive relationships between
the  variables  and the S02  concentration  in the off  gas were expected.  Indeed,
from  the  data  in Table VII-19,  no  such relationships were apparent, confirm-
ing  that  the  S02 levels  were  not  kinetically or thermodynamics11y controlled.
A further consequence of the  heat  and mass balance  limitation is very low SOg
concentrations in  the regenerator  off  gas.   As seen, the concentrations ranged
from 0.2  to  0.6%.  The approach to thermodynamic  equilibrium of the system can
be expressed  as  the  ratio  of  peak  S02  concentration to the S02 concentration
at equilibrium for the prevailing  temperature and pressure.  From Table
VII-19,  it  is  seen that this  ratio was  below 0.3  (or 3Q% of equilibrium) for
all  runs.

Coal  Fueled  Regeneration

      The objective of the  test  series  using  coal  in the  bench regenerator was
 to evaluate  the  feasibility  of  coal-fueled regeneration  in a PFB.  Factors
which needed to  be evaluated  were:   uniformity of temperature in  the fluidized

                                   217

-------
bed, temperature control, tendency for bed agglomeration,  S02  levels  pro-
duced, and degree of reduction of CaSCty to CaO.   As can be seen,  these  are
primarily operability questions.

Operating Procedures--
     Coal is fed to the regenerator by using the same coal  feeding  and  con-
trol system used for the bench combustor.  The system was  described in  the
previous reports (1,2).  A coal probe, similar to the design used for the
bench combustor, is directed downward at about a 45° angle and delivers coal
at the inner regenerator wall about 8 cm above the fluidizing  grid.  Coal  is
fed to the regenerator after the bed temperature is heated above  650°C  with
the natural  gas burner.

Experimental Results and Discussions--
     Nine experimental runs (Runs R-15 through R-23) were  made with coal-
fueled regeneration.  Four runs, R-19 to R-22, were shut down  soon  after coal
injection was started because of high and erratic bed temperatures  which could
have resulted in serious bed agglomeration.  The other five runs  were more
successful; about ten hours of operation on coal were accumulated.  Operating
conditions for the steady state periods are reported in Table  VII-20.  The
major operational problems during the runs were plugging of the off gas filter
maintaining proper superficial gas velocities, and bed temperature  control.    '

     It was necessary to end three of the runs,  R-15, 16 and 18,  because the
sintered metal off gas filter became plugged, resulting in loss of  column
pressure control.  Filter plugging might have been caused  by an increased
solids loading in the regenerator off gas due to the carryover of unburnt
coal and ash particles.  Another factor which might have contributed  to the
plugging of the filter was increased bed solids  carryover  due  to  excessive
superficial  gas velocities.

     There is a problem in maintaining proper superficial  gas  velocities when
fueling the regenerator with coal.  When operating the regenerator  with coal,
combustion air is introduced to the regenerator from two sources:  primary
combustion air enters through the grid, and coal transport air enters through
the coal probe, which is positioned about 8 cm above the fluidizing grid and
aimed downward.  Transport air flow is approximately 0.26  m^/min, which is
almost half of the total air flow.  Exactly how the air entering  via  the coal
probe contributes to bed fluidization is unknown.  Therefore,  to  insure good
bed fluidization, a minimum superficial gas velocity of 0.61 m/s  was  main-
tained just above the fluidizing grid by adjusting the primary air  flow.
Operating pressure was varied during the coal run series to adjust  the  super-
ficial gas velocity to a level which would insure good bed fluidization and
avoid slugging of the bed.

     The ability to control bed temperature is crucial to  the  operation of a
coal-fueled regenerator.  Temperature runaways to 1100°C can  result in  exten-
sive bed agglomeration because of the presence of coal and ash in the bed.
During the coal runs, a few temperature runaways occurred, resulting  in a
high temperature shutdown of the unit and a limited amount of  bed fusion.
                                    218

-------
no
H-•
10
                       TABLE VII-20.  COAL FUELED BENCH REGENERATOR RUN SUMMARIES


                Avg.                  Sup.  Gas     m  Solid                                Extent
                Bed         Expanded  Gas   Res.   4>v'  Res,   SOe; Mole %   Peak % S02     ofSollds
          Run   Temp.   P      Bed    Vel.  Time  Equiv  Time     Average    TZ7T\—j <-A   Regeneration
        Number    K    atm  Height m  m/s   sec.  Ratio  min.     (Range)    tqui .  » au2        %


         R-17   1422   4.6     2.1    1.6   1.32  1.52     62       1.32         0.16           40
                                                                  (0-3.5)

         R-18   1422   4.7     0.7    1.5   0.44  1.54     16       2.1          0.17           38
                                                                 (0.1-3.7)

         R-23   1338   6.1     0.8    1.5   0.51  1.56     31       0.9           —            40
            A     A1r Required for Complete Combustion
            <(>	Mr Supplied	

            Assumes  100%  Combustion Efficiency

            Sorbent:   Sulfated Dolomite

            Coal:      Champion

-------
     When the unit was operated in oxidizing conditions, temperature control
was not a problem.  However, temperature control was difficult under reducing
conditions.  To maintain reducing conditions in the unit, the coal  feed  rate
must be such that the air-to-fuel ratio is substoichiometric.  With the  large
amount of transport and primary air required, the coal  feed rate needed  to
produce reducing conditions exceeded that needed to satisfy the energy
requirements of the unit.  That is, operating temperatures were too high and
resulted in high temperature shutdowns.  To reduce the  coal  feed rate, com-
bustion air flow rates also had to be reduced.

     To achieve this reduction in air rate, the coal feeding system was  mod-
ified to use nitrogen, instead of air, as the transport medium, making the
primary air the only source of combustion air.   Only one run, R-23, was  made
with the modified coal transport system.  During the run bed temperatures
were more easily controlled than in previous runs.  However, after  25 minutes
of steady state regeneration, bed temperatures  again began to rise.  At  this
point, the run was ended to avoid agglomeration of the  bed.   The tendency for
the bed temperature to rise after a given period of steady state regeneration
has been noted not only in the bench regenerator studies but also in the
batch studies carried out in the batch and miniplant regenerator units.   The
most probable reason for the temperature increase is that the bench unit
operates as a "batch regenerator" during startup because the unit was charged
with sulfated sorbent prior to the run.  Initially, the energy released  by
the combustion of the coal is absorbed in the endothermic regeneration of
this large mass of startup bed material.  When  the entire bed inventory  is
regenerated, the endothermic regeneration reaction takes place at a lower
rate, since only the incoming sorbent is being  regenerated.   Unless the
coal feed rate is reduced to compensate for the lower heat requirement,  a
temperature increase could result after the initial charge is regenerated.
Changes in the off gas composition which occurred concurrently with the
temperature rise support this explanation.  The S02 level dropped even
though the regenerator remained in reducing conditions.

     The results of the few runs in the bech regenerator are promising.   No
problems were encountered which would rule out  the use  of coal.  Maintaining
good bed fluidization while still maintaining a constant bed height was  a
problem as it was with natural gas fueled regeneration.  The partial bed
agglomerations that did occur were probably a result of poor fluidization.
The small diameter (9.5 cm) of the regenerator  makes good gas and solids
mixing and fluidization difficult.  Temperature control was a problem at
times but again, the cause could have been poor gas and solids mixing in the
small regenerator vessel.  Regeneration levels  for the  coal  runs (-40$)  were
not as high as the natural gas runs.  A possible explanation is that the
concentration of CO and H2 in the reducing zone was lower than in the pre-
vious runs.  Obviously, more work must be done  in the area of coal-fueled PFB
sorbent regeneration on both small and larger scales to make a complete
evaluation.  Such questions as where and how to add supplementary air and
coal to maintain proper fluidization, mixing and temperature control must be
answered.  Modifications to the regenerator vessel, such as a tapered cross
section in the lower part of the vessel, should also be considered.
                                    220

-------
                                SECTION  VIII

                                 REFERENCES
1   Hoke, R. C., et al., "Miniplant Studies  of Pressurized  FTuidized-Bed
    Coal Combustion:  Third Annual  Report,"  EPA-600/7-78-069,  April  1978.

2.  Hoke, R. C., et al., "Studies of the Pressurized Fluidized-Bed  Coal
    Combustion Process," EPA-600/7-77-107, September 1977.

3.  Hoke, R. C., et al., "Studies of the Pressurized Fluidized-Bed  Coal
    Combustion Process," EPA-600/7-76-011, September 1976.

4.  Hoke, R. C., et al., "A Regenerative Limestone Process  for Fluidized-Bed
    Coal Combustion and Desulfurization," EPA-650/2-74-001, January 1974.

5.  Skopp, A., et al., "Studies of the Fluidized Lime-Bed Coal Combustion
    Desulfurization System," December 31, 1971.

6.  Nutkis, M. S., et al.,  "Hot Corrosion/Erosion Testing of Materials for
    Application  to Advanced Power Conversion Systems Using Coal-Derived Fuels
    Task II -  Fluidized Bed Combustion," to be published.

7  Smith,  W.  B. and  Wilson, R. R.,  "Development and Laboratory Evaluation
    of a Five  Stage Cyclone System," EPA-600/7-78-008, January 1978.

 8.  "Pilat  (University of  Washington) Mark  III Source Test Cascade  Impactor,"
     Pollution  Control Systems  Corporation,  Bulletin 76-3A.

 9.   Parker, R.,  et  al.,  "High  Temperature and  Pressure Particle Collection
    Mechanisms," Second  Symposium  on the  Transfer and Utilization  of
     Particulate Control  Technology,  Denver, CO,  July 23-27,  1979.

10   Koch, W.  H.  and Licht, W., "New Design  Approach Boosts Cyclone
     Efficiency," Chem Eng. No. 7,  1977.  p  80 ff.

11.   Knowlton,  T. M. and Bachovchin, D.  B.,  "The Effect of  Pressure and
     Solids Loading on Cyclone  Performance," Presented at AIChE 70th Annual
     Meeting, New York,  N.Y.,  November  1977.

12.  Zenz, F. A., Private Communication with R. C.  Hoke.

13   Jonke, A., et al.,  "Supportive Studies  in Fluidized  Bed  Combustion,"
  '  Argonne National  Laboratory, EPA-600/7-77-138, December  1977.

14   Murthy, K.  S., Howes, J.  E., Nack, H.  and Hoke, R. C., "Emissions from
     Pressurized Fluidized-Bed Combustion Processes," Environ. Sci. and  Tech.,
     13(2), 197-204, February 1979.
                                     221

-------
15.  6CA Corporation, GCA/Technology Division, "Test Plan for Level  I
     Characterization of Emissions from Exxon PFBC Miniplant Unit with
     Regeneration," 1979.

16.  GCA Corporation, GCA/Technology Division, "Test Plan for Level  II
     Characterization of the Emissions from the Exxon Fluidized-Bed  Combustion
     Miniplant," 1979.

17.  Snyder, R., et al., "Supportive Studies in Fluidized-Bed Combustion,"
     Argonne National Laboratory, Quarterly Report, ANL/CEN/FE-77-8,
     July-September 1977, p. 31-32.

18.  Hubble, B. R., et al., Argonne National Laboratory, "Chemical,  Structural,
     and Morphological Studies of Dolomite in Sulfation and Regeneration
     Reactions," paper presented at the Fourth International Conference on
     Fluidized-Bed Combustion held at the Mitre Corporation, McLean, VA,
     December 9-11, 1975.

19.  Hubble, B. R., et al., "A Development Program on Pressurized-Bed
     Combustion," Argonne National Laboratory, Annual Report, ANL/ES-CEN-1011
     July 1, 1974-June 30, 1975, p. 84-86.

20.  Cunningham, P., et  al., "A Development Program on Pressurized Fluidized
     Bed Combustion," Argonne National  Laboratory, Annual Report,
     EPA-600/7-76-019, July 1975-June 1976, p. 134.

21.  Lyon, R. K., International  Journal of Chemical Kinetics, 8, p.  315-18
     1976.

22.  U.S. Patent 3900554, "Method for the Reduction of the Concentration of
     NO in Combustion Effluents Using Ammonia."

23.  Pigford, R. L.^and  Sliger, G., "Rate of Diffusion-Controlled Reaction
     Between a Gas and a Porous Solid Sphere," Ind. Eng. Chem. Process.
     Des. Develop., 12(1), 1973, pp.  85-99.

24.  Szekely, J., Evans, J. W. and Sohn, H. Y., Gas-Solid Reactions.
     Academic Press, New York, 1976,  pp. 125-168.

25.  Borgwardt, R. H. and Harvey, R.  D., "Properties of Carbonate Rocks
     Related to S02 Reactivity," Environ. Sci. and Tech., 6/4), 1972,
     pp. 350-360.

26.  Lentzen, D. E., et  al., "IERL-RTP Procedures  Manual Level 1 Environmental
     Assessment (Second  Edition)," EPA-600/7-78-201, October, 1978.

27.  Duke, K. M., et al., "IERL-RTP Procedures Manual:  Level 1 Environmental
     Assessment Biological  Tests for  Pilot Studies," EPA-600/7-77-043,
     April  1, 1977.

28.  Shackleton, M. A.,  "Extended Tests on Saffil  Alumina Filter Media."
     EPA-600/7-79-112, May 1979.


                                     222

-------
                               SECTION IX

                           LIST OF PUBLICATIONS


1.  Hoke, R. C., Gregory,  M.  W.,  "Evaluation of a Granular Bed Filter for
    Partlculate Control  1n Fluidlzed Bed Combustion," Proceedings of the
    EPA/ERDA Symposium on  High Temperature/Pressure  Particulate Control,
    Washington, DC, September 20-21, 1977.

2.  Nutkis, M. S., Hoke, R. C., Gregory, M. W,, Bertrand, R. R.,
    "Evaluation of a Granular Bed Filter for Particulate Control in Fluidized
    Bed Combustion," Proceedings  of  the  Fifth  International Conference on
    Fluidized Bed Combustion, Vol.  Ill  p 504,  Washington, DC,  December
    12-14, 1977.

3.  Ruth, L. A., Hoke, R.  C., Nutkis, M. S., Bertrand,  R. R.,  "Pressurized
    Fluidized Bed Coal Combustion and Sorbent  Regeneration,"  Proceedings
    of the Fifth International  Conference  on Fluidized  Bed  Combustion,
    Vol. Ill p 756, Washington,  DC,  December 12-14,  1977.

4.  Ernst, M., Loughnane, M. D.,  Bertrand, R.  R., "Instrumental  Methods  for
    Process Definition in a Pressurized Fluidized Bed Coal  Combustion  Pilot
    Plant," AIChE 85th National  Meeting, Philadelphia,  PA,  June 4-8,  1978.

5  Ruth, L. A., "Regenerate Sorbents  for Fluidized Bed Combustion,"
    Final Report prepared under NSF RANN Grant AER75-16194, June 1978.

6  Hodges, J.  L.,  Hoke,  R. C., Bertrand,  R.  R., "Prediction of Temperature
    Profiles  in  Fluid Bed  Boilers," Journal of Heat Transfer Vol. 100, No.  3,
    p 508-513,  August 1978.

7.  Ruth,  L.  A.,  "Regeneration of the Sulfur Acceptor in Fluidized Bed
    Combustion,"  Fluldization, Proceedings of the Second Engineering
    Foundation  Conference,  p 303-313, 1978.

8  Ruth,  L.  A.,  Varga, G.  M., "Developing Regenerable S02 Sorbents for
     Fluidized Bed  Coal  Combustion Using Thermogravimetrlc Analysis,"
     Thermochimica  Acta, 25,  p 241-55, 1978.

 9.  Murthy, K.  S., Howes, J. E.,  Nack, H., Hoke, R.  C., "Emissions from
     Pressurized Fluidized-Bed Combustion  Processes," Envir. Sci and Tech.
     Vol. 13,  No. 2, p 197-204, February.1979.

10   Hoke, R.  C., Ruth, L. A., Ernst, M.,  "Control of Emissions  from the
     Pressurized Fluidized Bed Combustion  of Coal,"  AIChE 86th National
     Meeting, Houston, TX, April  1-5, 1979.
11   Siminski, V. J., Ernst, M., "Operating Experiences  at Exxon  Research and
                  '"  "-^	 ~      its in a Pressurized Fluid1z<
                                            Meeting, Houston, TX, April
3 III! I I O "^ I ) •• **••  *-i ii%* v |  ' ' • |   wpwi %» v i ny  fc./\ f
Engineering with  Materials and  Components  in a  Pressurized  Fluidized Bed
Combustor (PFBC)," AIChE 86th National  Me<
     1-5, 1979.

                                      223

-------
12.  Hoke, R. C., Ruth, L. A., "Control  of Emissions  from  the  Pressurized
     Fluidized Bed Combustion of Coal,"  American  Flame Research  Committee
     Boston, MA, April 30, 1979.

13.  Hoke, R. C., Ruth, L. A., Ernst, M., Nutkis, M.  S., Garabrant, A. R.,
     Goodwin, J. L., Radovsky, I.E., "Miniplant  PFBC Studies,"  Pressurized
     Fluidized Bed Combustion Technology Exchange Workshop,  Secaucus, NJ
     June 6, 1979.

14.  Ruth, L. A., Varga, G. M., "New Regenerate  Sorbents  for  Fluidized Bed
     Coal Combustion," Environmental  Science and  Technology.  (13) p 715-720
     June 1979.

15.  Ernst, M., Hoke, R. C., Siminski, V. J., McCain, J. D.,  Parker, R.,
     Drehmel, D. C., "Evaluation of a Cyclonic Type Dust Collector for High
     Temperature, High Pressure Particulate  Control," Second Symposium on the
     Transfer and Utilization of Particulate Control  Technology, Denver, CO
     July 23-27, 1979.

16.  Ernst, M., Shackelton, M. A.,  Drehmel,  0. C., "Ceramic  Filter Tests at
     the EPA/Exxon PFBC Miniplant," Second Symposium  on the  Transfer and
     Utilization of Particulate Control  Technology, Denver,  CO,  July 23-27.
     1979.

17.  Jahnig,  C. E., Shaw,  H., Hoke, R. C., "Continuous  Sorbent Regeneration
     in Pressurized Fluidized Bed Combustion," Proceedings of the 14th
     Intersociety Energy Conversion Engineering Conference, Vol. I, p 933,
     Boston,  MA, August 5-10, 1979.
                             OTHER PRESENTATIONS


Bertrand, R. R., "Pressurized Fluidized  Bed  Combustion,"  Fluidized Combustion
of Coal  Symposium, MIT Industrial  Liaison  Program,  January 18, 1979.
                                     224

-------
                                   SECTION X
                                  APPENDICES

                                                                         Page
A       Effect of Particle Size on Conversion Rate of Limestone           227
        and Dolomite:  Thermal Gravimetric Experiments
B       Testing of Various Stones for Ability to Absorb S02               230
C       Data Management Section                                           232
D       Participate Size Distribution and Concentration                   237
        Measurement Procedures
E       Analytical Techniques                                             241
F       Comparison of 862 Measurements by UV, GC, Wet  Chemistry           243
G       Results of Flue Gas $03 Analysis                                  246
H       Regenerator GC Analyses                                           247
I       Mass Spectrographic Analyses of Solid Samples  of  Run  69           248
j       Miniplant Component Mass  Balances                                 251
«       Miniplant Run Objectives                                          253
L       Mini plant Fluidized Bed  Coal Combustion  Run  Summary               254
M       Particle Size Distribution                                        271
 M_l    Spent  Pfizer  1337  Dolomite  Sorbent                                271
 M-2     First  Cyclone Dip! eg                                              272
 M_3    Second Cyclone  Capture                                           273
 M-4     Tertiary Cyclone  Capture                                         276
 M-5     and Grain  Loadings of Flue Gas  Participates  Before                278
         Tertiary Cyclone
 M-6     and Grain  Loadings of Flue Gas  Parti culates  After                 280
         Tertiary Cyclone
 N        Miniplant Solids  Analysis                                         286
                                       225

-------
                            APPENDICES (Continued)

                                                                         Page
0       Mint pi ant Sample Shipments                                        301
P       Bench Combustor Run Summary                                       305
 P-l     Initial Checkout                                                  3Q5
 P-2     Two Stage Combustion - NOX Control                                308
 P-3     NH3 Injection - NOX Control                                       312
 P-4     Simulated Flue Gas Recirculation - NOX Control                    31g
 P-5     Combined NOX Control Methods                                      318
                                      226

-------
                                APPENDIX A

               EFFECT OF PARTICLE SIZE ON CONVERSION RATE OF
          LIMESTONE AND  DOLOMITE:  THERMAL GRAVIMETRIC EXPERIMENTS


     The sulfatlon of limestone and dolomite has been described by models
which are based on the idea  that the conversion rate is governed by diffusion.
Several models, called "grain" models, assume that individual limestone par-
ticles are composed of many  sub-particles, or grains (23,24).  These grains
are imagined to be non-porous.  When limestone reacts with SOg, S02 diffuses
through the pores separating the grains  (intergranular pores) and reacts on
the grain surfaces.  The speed of the  reaction is determined by the local S02
concentration inside the pores and  by  the chemical reaction rate at the sur-
face of the grain.  Alternatively,  and this  is probably the case, when SO?
reacts with limestone or dolomite,  the speed of the  reaction is determined
by the local S02 concentration  inside  the  pore and by  the thickness of the
solid reaction product layer which  forms on  the surface of each grain.

     The grain model implies the  primary importance  of physical properties
such as porosity,  pore size distribution,  surface  area, and  grain  size,  in
determining the conversion rate  of a  limestone or  dolomite particle.   Experi-
mental studies have confirmed that these properties  are  important.

     The effect of particle size on the  conversion rate  of limestone  depends
on which step  in  the reaction scheme is  rate controlling.  If diffusion
through  intergranular pores is controlling,  then  a concentration  gradient of
SO? could  exist from the surface of the  particle  to  its  center.   Grains  near
the surface would  sulfate faster because of the higher S02 concentration and
a  shell  of reaction  product (CaSO^ would form on the surface of the particle.
In this  situation, particle size has the greatest effect on conversion rate
and the  time required to achieve a given fractional  conversion of the par-
ticle  is  proportional to 1/RS where R  is the particle radius.  At the other
extreme  is the situation where S02 is spread uniformly through the pores of
the  particle and  all  of the particle's  interior reacts at the same rate.  In
this  case, particle  size will have no effect on conversion rate.

      In  order  to  explore the effect of  particle size on conversion rate,
limestone  and  dolomite  particles of different sizes were sulfated in a TGA
at conditions  of  900°C, 0.25% S02, 5% 02, and balance N2.  Four particle
size  ranges were  investigated:  -8 +  25 mesh, -16 + 40 mesh, -100 + 200 mesh,
and  -325 mesh.  These size  ranges correspond approximately to average par-
ticle diameters of about 1500, 700, 100 and 40 ym, respectively.  Table A-l
aives data on  initial  conversion rates, the time to reach 15 percent conver-
sion, and the  percentage conversion after 75 and 150 minutes for each of the
 four average particle sizes.

      A number  of significant observations can be made from the data of Table
A-1.
                                     227

-------
              TABLE  A-l.   EFFECT  OF  PARTICLE SIZE ON THE SULFATION OF DOLOMITE AND LIMESTONE
                                            Dolomite  (Pfizer)
ro
ro
oo
     Average Particle
Initial  Rate,
Time for 15%
     (1)   Average  of 3  runs.

     (2)   Average  of 2  runs.
% Conversion
Diameter, um
1500(1)
700^
100
40<2>

Average Particle
Diameter, ym
1500
700
100<2>
4°(2)
mg S03/(mg CaO)(min.)
0.24
0.044
0.100
0.087
Limestone
Initial Rate,
mq S0.i/(mq CaO)(m1n.)
0.010
0.017
0.056
0.055
Conversion, min.
32
4.0
1.5
2.0
(Grove)
Time for 15%
Conversion, min.
150
150
3.5
3.3
?5 min
28
40
39
75

%
75 min
6.4
5.9
59
79
150 min.
32
44
45
90

Conversion
150 m1n.
7.1
6.7
70
89

-------
     1.   Particle size has a large effect on the percentage  conversion
         or utilization of both limestone and dolomite.   For example,
         the  percentage conversion of limestone increases by about  a
         factor of ten as particle size is reduced from  1500 to 100 urn.

     2.   Coarse dolomite achieves much higher utilizations than coarse
         limestone.  However, utilizations for fine (100 and 40 ym)
         limestone and dolomite are comparable.

     3.   The  initial rates for dolomite are higher than  for limestone
         for  each of the particle sizes studied.

     4.   Initial rates, time for 15% conversion, and percentage
         conversion after 75 and 150 minutes are all more sensitive
         to changes in  particle size for  the larger size particles.

     Each of the above  observations can  be  interpreted  in terms of the grain
model.  Observation  (1),  the effect of particle  size on  utilization, implies
that a concentration  gradient  of SO? through the  particle does  exist and
that diffusion of S02  through  a  sulfate  shell  surrounding the  particle does
influence the conversion  rate.   Observation (2)  suggests  that  dolomite has
larger intergranular  pores  than  limestone.   Indeed, pore  diameters for a
limestone and dolomite, different  from the  stones used  in this study, were
measured by Borgwardt and Harvey (25).   Their  limestone had pores  about 1 ym
in diameter whereas their dolomite had 0.5-20  ym pores.  Because of  the wider
intergranular pores,  large dolomite particles  achieve higher utilizations
than large limestone particles.   However, as particle size  is  reduced, so is
the  influence of Intergranular diffusion.  Thus, for  small  particles, lime-
stone and dolomite give comparable utilization.

     Observation (3), that the initial  rate for dolomite is higher than  for
limestone  is consistent with the hypothesis that dolomite has  larger pores.
Also, dolomite may have smaller grains than limestone.   Observation  (4),  that
the  effects  of particle size become less as size 1s reduced, again implies
that intergranular diffusion becomes less important as  the particle  size
becomes  smaller.

      There are  practical implications for fluidized bed coal combustors  from
the  above  results.  Using smaller sorbent particles, particularly when the
sorbent  used is limestone, can result in a significant reduction in  the
auantity of  sorbent required.  Although, the use of smaller particles will
Result  1n  higher entrapment rates, oversized cyclones can be used to return
sorbent  fines  to the bed.  Lower superficial gas velocities may also be used
to reduce entrapment.  Whether or not  the use of fine particles is  practical
would depend on the trade-off of reduced sorbent and sorbent  disposal costs
vs.  Increased equipment  costs  (I.e.. the cyclone recycle)  and  potentially
 lower equipment capacity.
                                     229

-------
                                APPENDIX B

                          TESTING OF VARIOUS STONES
                          FOR ABILITY TO ABSORB S02


     Several  limestones and  dolomites were screened 1n the thermogravlmetrlc
analyzer (TGA) for their ability to absorb S02.  Two gas compositions were
used, one to promote calcination and the other to  suppress it.  Table B-l
gives the reaction conditions and the fraction of  each stone sulfated after
60 minutes.  The "standard"  sorbents, Grove  limestone and Pfizer  dolomite
are shown for comparison,

     The data of Table B-l shows that there  1s considerable variability In
reactivity among the stones, even when  testing was done  under  calcining con-
ditions (condition "A" in Table B-l).   The activity of  Plum Run dolomite is
fairly high, even under conditions  which  suppress  calcination  of  CaCOs
(condition "B" 1n Table B-l) because  the  MgCOj  in  the dolomite always cal-
cines, Introducing porosity Into the  stone.   The  substantial MgO  content of
the low grade limestone 1s probably responsible  for  its  good activity under
condition B.  The high grade Mulzer limestone,  however,  shows  fair activity
even under "non-calcining" conditions,  even  though 1t  contains little MgO.
This result was  probably  caused by partial  calcination  of the  CaCOs in  the
stone.  The other sorbents showed relatively low activity even under "cal-
cining" conditions.   The  reasons are not known but could be due to other
structural differences  1n the sorbents.

      It  should  be noted  that all stones of Table  B-l  were tested as powders.
The fractional  sulfations observed would have been quite different, probably
much lower,  had the  stone bed tested 1n the form  of 2-3 mm granules, commonly
used 1n  fluldized bed combustors.
                                     230

-------
                    TABLE B-l.  REACTIVITY OF VARIOUS STONES UNDERGOING SULFATION
                                                                 Fraction Sulfated 1n 60 Minutes
ro
to
Sample (Powder) Test
(
Dolomite, Plum Run^ '
Limestone, Mulzer, High Grade' '
Limestone, Consolidation Coal,
Low GradeU/
Limestone, Ste. Genevieve, High Purity^ '
Limestone, Dolomitic^ '
Limestone, Grove, BCR No. 1359
Dolomite. Pfizer, BCR No. 1337
Test Conditions
A - 870-900'C; 0.25% SO?. 5% Op. bal N£
Sample heated to 870 C in N2 prior to sulfation
B - 870°C; 0.25% SO?, 5% 0?. bal CQ2
Sample heated to 870 C in Cp2 prior to sulfation
Conditions "A" Test Conditions "B"
"Calcining") ("Non-Calcining")
0.51 0.29, 0.58(2)
0.36 0.11, 0.13^
0.45 0.28, 0.37^2)
0.17
0.062
0.30
0.39



to suppress calcination
     (1)   Samples supplied by American Electric Power Service Corp.
     (2)   Replicate run
     (3)   Samples supplied by Western Materials Company

-------
                                APPENDIX C

                           DATA  MANAGEMENT SYSTEM


     The data generation and  management  system  for  the minlplant  combustor
was thoroughly described 1n the  previous annual  report  (1),   The  entire sys-
tem has been augmented to Include  the collection, computation,  and  summariza-
tion of minlplant regenerator data.   Description of the major additions
follow.

HOKE SYSTEM

     Figures C-l and C-2 are examples of the tabular form of regenerator data
printout received from a run.  This  printout 1s then used 1n conjunction with
the combustor printout and the continuous recorder  charts to determine the
steady  state period.

INPUT/OUTPUT PROGRAMS

     Regenerator segments of the computer printout generated by the I/O
program are  shown 1n  Figures C-3 and C-4.  The first figure  is an example of
the  experimental data  inputs used 1n the program and Figure  C-4  is an  example
of the  calculated outputs.  Addition regenerator data, such  as off gas  emis-
sions  and  solids chemical analyses and  particle size distributions, are
 Inputed in the  appropriate sections.  The regenerator effluent streams  are
 Included 1n  the overall  mass accountability section,
                                      232

-------
                                                  FIGURE  C-l
                                   ANALYSIS PROGRAM
4/30/79
                                                          REGENERATOR SECTION
ro
CO
CO
TEMPERATURESIDfcG
1
2
3
4
5
6
7
8
9
10
11






TIME
PORT
PORT
PORT
PORT
PORT
PORT
PORT
PORT
PORT
PORT
RUN
5.12
5.13
5.13
5.13
5.12

NO.
NO.
NO.
NO.
NO.
NO.
NO.
NO.
NO.
NO.

18
19
20
21
22
F)

2
3
A
5
7
6
9
10
11
12
1
29
29
29
29
29


(5IN)
( 11IN)
( 17IN)
I23IN)
( 35IN)
(41IN)
(48IN)
(56IN)
(64IN)
(74IN)
23456789 10 11
1820. 1837. 1852. 1859. 1874. 1889. 1885. 1887. 1890. 1893.
1819. 1833. 1849. 1856. 1871. 1886. 1881. 1884. 1887. 1889.
1818. 1832. 1848. 1855. 1869. 1883. 1878. 1880. 1883. 1886.
1815. 1833. 1851. 1857. 1871. 1884. 1879. 1880. 1883. 1888.
1814. 1834. 1852. 1858. 1871. 1885. 1880. 1880. 1884. 1887.

-------
                                                  FIGURE C-2
                         MINIPLANT ANALYSIS PROGRAM
                                                                     A/30/79
                                                           REGENERATOR  SECTION
oo
      12     TIKE

      REDUCTION ZONE
 13     SUPERFICIAL  VELOCITY(FT/SEC )
 14     AVERAGE  BED  TEMPERATURE(DEC  F)
 15     AIR-FUEL  RATIO(PHI)

 OXIDIZING  ZONE

 16     SUPERFICIAL  VELOCITY(FT/SEC )
 17     AVERAGE  BEL)  TEMPERATURE ( DEG  F)
 18     AIR-FUEL  RATIO(PHI)

 19     EXPANDED  BED HEIGHT(FT)
20     OFFGAS CO LEVEL(PPM)
21     OFFGAS S02 LEVEL(VPCT)
             RUN
                12
 13
15
16
17
18
19
20
21
             5.12
             5.13
             5.13
             5.13
             5.12
              18 29
              19 29
              20 29
              21 29
              22 29
2.14 1842.98     1.27
2.14 1840.33     1.26
2.14 1838.95     1.27
2.15 1839.76     1.27
2.15 1840.09     1.27
       2.75 1886.46    1.01
       2.75 1883.37    1.01
       2.75 1880.16    1.01
       2.74 1881.18    1.02
       2.75 1881.35    1.02
                       6.20 1274.00    0.55
                       5.26 1347.48    0.45
                       7.10 1249.00    0.52
                       7,56 1385.56    0.34
                       7.03 1424.60    0.52

-------
                                                     FIGURE C-3
ro
CO
en
          lFMPFBATUHf(Ci!.LSIUS)
PORT NO. 2 ( 5 IN)
FORT NO. 3 ( 11 IN)
PORT NO. 4 ( 17 IN)
FORT NO. 5 ( 23 IN)
PORT NO. 7 ( 35 IN)
PORT NO. 3 ( 41 IN)
PORT NO, 9 ( 43 IN)
FOiiT NO, 13 ( 56 IN)
FORT NO. 11 ( o4 IN)
FORT MO. 12 ( 74 IN)
fORT NO. 13 ( 83 IN)
PORT NO. 1-1 ( 92 IN)
PORT NO. If (11 PI IN)
PORT NO. 13 (158 IN)
FORT NO. 21 (156 IN)
FORT NO. 34 (194 IN)
FORT NO. 26 (218 IN)
CYCLONE DISCHARGE
COOLER OFFGAS
PRiSSURSS(KPA)
NOMINAL OPERATING
R-C PRESSURE DROP
GRID PRESSURE DROP
BSD PRESSURE DROFS
994.
1334*
1012.
1016.
1025.
1032.
1029.
1230.
1031.
1332.
13c6.
1013.
871.
775-
935.
886.
881.
1443.
420.

735.961
5.961
2.194

                                  SETTLFD BED HFIGHT(M)

                                     INITIAL
                                     FINAL
                                                                                    HalSHI  ABOVE  :
                                     SUFP AIR PROBE
                                     SUFF FU&L F*OfcE
                                  BED BULK DENSirT(GR/Cl3

                                     INITIAL
                                     FINAL
                                                                                                                  1.6?
                                                                                                                  3.74
                                                                                                                  2.13
                                                                                                                ******
                 FORTS  29  TO  371  (5  IN  TO 41  IN)     6.573
                 FORTS  29  TO  34  (5  IN  TO 224  IN)   12.413
          FLOW HATES(M3/MIN)

              BURNER  AIR
              BURNER  FUEL
              SUPPLEMtNTART  AIR
              SUFPLEMENTARlf  FUEL
 1.85
 3.18
 2.46
 2.2G
         C-R TRANSFER RATF(KG/HE)
15.4
         AUXILLIAfilf N? FI3»'{13/1IN)
 .?. 1

-------
                                                     FIGURE  C-4
                                                 REGENERATOR SECTION
         SUPERFICIAL moCITT(M/SKC)

           F.EDUCIN3 ZON2
           OXIDIZING ZONE
                                         3.64
                                         0.73
                  SULFATIOM LEVEL DF SORPSNTUPCT)

                     TO REGENERATOR
                     FROM REGENERATOR
   23.4?
********
                 EFT TFMPFPATURFS(CKLSIUS)
            REDUCING ZONE - PORTS 2 TO !>
            OXIL'IZINO ZONE-  POSTS 7 TO 12
                                      10B6.79
                                      1339.77
                  R£SEH«RAriO,M Ll'JZL(FCI)                           39.76
                  COMBUSIOR FEED SULFUR IN R&3KN OfFJAS(rfPCT)       73.CG
ro
CO
at
SETTLED BED HEI3HT(M)
EXFfcKt'ED BED HF.IGHT(M)
GAS RFSIDFNCE TIME(SEO)
SOLIDS RESIDENCE TIME(HRS)
1 .85
2.34
3.23
         ENERGK IMFUT(MBTU/HR)
                                         3.51
         AIR TO FUSL RflriO(PHI)

         HECUCINC  ZONF
         OXIDIZING  ZONE
                                         1.28
                                         1.33

-------
                                APPENDIX D

                       PARTICLE SIZE DISTRIBUTION AND
                    CONCENTRATION MEASUREMENT PROCEDURES


     The measurement of PFB  generated participates is 1n Its Infancy.
Sophisticated high temperature high pressure 1n-s1tu size distribution measur-
ing devices are being developed and tested to understand better the partl-
culate emissions.   While these systems are being developed, simpler systems
are being used to  generate data which will give some Indication of the par-
tlculate emissions and particle size distribution.  In the roiniplant this is
done by sampling for flue gas  particulates with Balston total filters.  Some
stream concentrations which  cannot  be sampled directly are calculated based
on mass balances,   Size analysis  of all  fine particulates  (-45 ym) is done
with the Coulter Counter Model TA11.  The  procedure developed for the mini-
plant will be described in  this section.  This section assumes a working
knowledge of particulate sampling and Coulter Counter operations.  It is meant
to supplement, rather than  replace  good  sampling  practice  and/or the Coulter
Counter operations manual.

Ralston Filter Sampling

     The total filters used on the  mlniplant are Type 30/25 Balston  filters.
They use a 2.5 cm diameter by 18 cm long Grade  BH filter tube.   This filter
Is 99.95%  efficient  for 0.6 ym particles.  The  filter cartridge  is  stored in
Us plastic  wrapping in a dry box to make sure  1t is dry.   Shortly before
sampling,  the  cartridge 1s unwrapped and weighed on an analytical  balance.
The weight  is  recorded, the filter support screen is Inserted, and the filter
assembled.   When  everything else 1s ready, the filter is installed for out-
side/In  flow.   The  sampling is started  soon after filter installation to pre-
vent unknown quantitites of gas and particulate from leaking through the
valve  onto  the filter.  The combustor side of the valve 1s continually purged
with air  or  nitrogen to prevent plugging of the probe.  If sampling must be
delayed  after the filter is installed,  the filter 1s pressurized and purged
to minimize leakage further.

      To  start the sampling, the  purge flows are shut off, and the hot isola-
tion  valve opened.   The flow  is  set on  the flow control system to the iso-
kinetic sampling  rate.  Temperatures, flow, and flow rate are recorded every
15-20  minutes. Most sampling after the tertiary cyclone requires 2 hours to
build  up sufficient filter  cake.   Samples taken before the cyclone only
require 1  hour.  After sampling, the high temperature isolation valve is  shut
and both purges are again  turned on.  The filter 1s allowed to cool and  is
 removed from the  system and taken  into  the lab.
 Note:  Coulter Counter is a registered trademark of Coulter Electronics Inc.
        Isoton and Accuvette are registered trademarks of Coulter
        Diagnostics Inc.


                                       237

-------
      In  the  lab, a one square foot piece of aluminum foil  1s  tared  and  the
 filter Is opened over the foil.  The filter cartridge 1s  carefully  removed
 and any  loose particles are dumped onto the foil.  The Inner  support  screen
 1s removed from the filter cartridge and the cartridge 1s  wrapped with  the
 foil.  The cartridge 1n foil Is weighed on the analytical  balance and the  tare
 weight subtracted to obtain a partlculate weight.  The concentration  1s
 obtained by dividing this by the total  gas flow and multiplying  by  a  conver-
 sion  factor.  The cartridge with foil wrapping 1s then stored  in a  dry  box.
 Unless the filter appeared wet, the filter 1s rewelghed on a  random quality
 control check basis only,

 Fine  Particle Sizing by Coulter Counter

     Most fine particle sizing 1n the miniplant 1s done with  the Coulter
 Counter Model TA11  with either 30 ym and/or 100 ym aperture probes.   Material
with a fair fraction of particles larger than 45 ym 1s prescreened  with a
sonic sifter through the appropriate screens.  The two distributions  are then
combined.

Equipment--
     The Coulter Counter used during these studies is  a Model  TA11.   Modifica-
tions have been made to decrease outside disturbances  and  increase  precision
The instrument rests on a large well  grounded metal  plate.  The  sample  stand'
next to the Instrument is completely enclosed in a Faraday cage  constructed
of a lucite box covered with wire mesh.  The sample stand  and  the Faraday
cage also rest on the grounded plate.  The sample vacuum pump  is outside
the cage, separately isolated and grounded.  The electric  stirrer motor sup-
plied with the sample stand was replaced with an pneumatic  stirrer.   This
equipment configuration has given very  repeatable performance  and has correla-
ted well  with other units in other labs.

     The electrolyte use in this study  is Isoton II.   The  Isoton (1%  NaCL  In
distilled water) 1s continuously filtered through 0.45 and  0.2 ym millipore
filters so that it  is ultra clean.  This has achieved  12 second  background
counts (30 ym aperture) as low as 10.

     The sifter used is an ATM sonic sifter with an assortment of sieves
from 5600 to 45 ym  (3-1/2 to 325 U.S. Mesh).  Various  screens  are used to
bracket the expected size distribution.

Procedure—
     A small  piece  of the Balston filter cake is carefully  removed  to avoid
contamination by the filter substrate.   The filter cake or  the minus  45 ym
(-325 U.S. Mesh) material  from the sonic sifter is used as  the sample.  The
sample (~ 0.1 g) is placed in a small plastic vial  of  known volume  called  an
Accuvette.  A cationic (Type 3A) dispersant is added  (~10  drops) until the
solid is  wetted. Ultra clean Isoton II is  added to  fill the Accuvette half
way.  The mixture is placed in an ultrasonic bath for  5-15 minutes.  While
the Accuvette is 1n the ultrasonic bath, the Coulter aperture and sample
beaker (500 ml)  are rinsed several  times with ultra  clean  Isoton II.  Several
rinses are required to reduce the background (no sample) count to less than
100 in 12 second accumulation with the  30 ym aperture  (less than 30 with the

                                    238

-------
100 ym aperture).   After  a  low background count 1s achieved, the Isoton II is
retained in the sample beaker and  several drops of sample are placed in the
beaker.  Sample is added  while stirring  until the proper sample concentration
1s reached (-2,5 to 55»).  Coulter  accumulation and operation from this point
are the same as in the operations  manual.

Magnetic Particulates—
     The above procedure  works well  for  most ordinary  PFB particulates.
Occasionally particulate  samples  are found  that are  fairly magnetic.  The
test for magnetism 1s carried  out with a simple pocket magnet.  When magnet-
ism is evident, further standardization  of  technique and care must  be exercised
to obtain a true distribution.   Failure  to  dilute the  sample properly may
lead to magnetic particle agglomeration  and settling.  The  idea is  simply  to
deaggl omerate  and dilute the sample enough  that agglomeration due  to magnet-
ism is negligible.

     For this  procedure, a working sample of 0.100  g is  used.   This sample
1s wetted with 25 drops of dispersant (Type 3A)  in  a 500 ml  beaker.  This
sample is diluted with 300 ml  of ultra clean Isoton II.   The beaker with its
contents is placed in the ultrasonic bath for 5-15  minutes.  Once the
sample is well dispersed, a 25 ml aliquot is taken  with  vigorous  mixing of
the  sample.   This aliquot 1s added to the 440 ml  of ultra clean Isoton II  in
the  500 ml  beaker on  the sample stand.  This is sufficient for a good accum-
ulation with  the  Coulter Counter.  The approximate concentration of the
sample is  18  micrograms  sample/ml.

      This  procedure  has  been found to give repeatable size distribution.  It
 Is  not as  sensitive  to slight delays in sampling as the other less rigid
 procedure.   Successive Coulter Counter accumulations  do not tend to finer
 size distribution as  happened with  the  other procedure.

      Balance Calculations,
      Some flue gas streams such  as  the  second cyclone  inlet cannot be sampled
 for particulates directly with a filter.   A calculation and sampling procedure
 has been developed to obtain particle size and  concentration  information  by
 mass balances.  These calculations  have been used  to determine  second cyclone
 Inlet and outlet concentration as well  as  second and third cyclone efficien-
 cies.

      Usually in the operation of the miniplant, Balston  filter  samples  of the
 flue gas after the third cyclone are taken at  planned  intervals.   Lock  hopper
 dumps of the cyclones during or  immediately  after  the  filter  sampling period
 are also sampled.  The Balston filter catch,  second and  third cyclone dump
 samples are analyzed for particle size distributions.   The amount of material
 1n the lock hopper is averaged over 3-5 of the bihourly dumps.  In this way,
 lock hopper hang  up and small inconsistencies are removed.  The various steps
 of the differential size distribution which is obtained from the size  analysis
 are multiplied by the average dump weight and divided by the gas flow rate.
 The various steps of the Balston filter differential  size distribution are
                                      239

-------
multiplied by the participate loading  measured.  All size distribution steps
are on the same weight/flow basis  and  a  size  differentiated mass balance may
be completed starting with the third cyclone.  Once concentrations in all
streams are known, cyclone efficiencies  may be calculated directly.  The only
assumptions in this calculation procedure  are that loadings are fairly con-
stant and that there is no change  in particle size through the flue gas
system.
                                      240

-------
                                APPENDIX E

                           ANALYTICAL TECHNIQUES
analysis of SoTids

     Sol Ids from combustion and regeneration runs were analyzed for SO^2,
CO-?-2, Ca+2, Mg+2,  Na+,  carbon and total sulfur.  The analytical techniques
that were used are  described  below.
     SO,
     CO
        -2
        -2
                           -   The  sample was treated with acidic
Ca"
Mg
      Total
      Sulfur
      Total
      Sulfur
        -  after Aug. 1978
       Total
       Carbon
                                    BaCl2 solution.
                                    was weighed.
             •  before Dec. 1978  -
                                              The BaS04 precipitate
                           -  HC1  was  added to  an  acidified  sample.
                              The solution was  stripped  with N£  and
                              the gas  passed through drlerlte,  CuS04
                              and ascarite.  CO?'2 was determined  from
                              the weight gain of the ascarite.
             -  after  Dec. 1978
       -  before Aug. 1978  -
        -  before  Aug. 1978  -
The sample was digested by heating
vigorously in a medium of perchloric
add/nitric acid.  The determination of
Ca, Mg and Na was made by atomic
absorption.

The sample was fused with NagCOo at
950°C, then dissolved with HC1.  The
determination of Ca and Mg was made by
atomic absorption.

 (Dletert  Sulfur Method)  - The sample was
 combusted in  an oxygen atmosphere at
 1250°C.   The  S02-S03  products in the
 effluent  gas  were  analyzed by an automatic
 Leco  tltrator.

 The sample was added to a ^2®$  catalyst
 and combusted in an oxygen atmosphere  in
 an Induction furnace at 1650°C.  The
 formed 1s selectively measured  by an
 Infra-red detector.  A Leco  IR-23 sulfur
 analyzer was used (ASTM D-1552).

 (Carbon on Catalyst Method)  - The sample
 was combusted 1n an oxygen atmosphere at
 1200°C.  The COg evolved was determined
 from the weight gain of ascarite.
                                      241

-------
     Total  -  after Aug. 1978
     Carbon
Analysis of Flue Gas by
Wet Chemical Analysis

     S03    -  before July 1978
     S03    -  after July 1978
     SO,
The sample was combusted in an industion
furnace at 1650*C followed by removal  of
sulfur, conversion of any CO to C02,
trapping of H£0 and trapping of the C02
on molecular sieves.  In the next cycle,
the molecular sieves are heated to 300°C
where the C02 1s expelled and measured by
thermal conductivity detection.  A Leco
WR-12 carbon analyzer was used.
The amount absorbed by an 80% Isopropanol
solution was determined titrimetrlcally
using 0.01N barium perchlorate as the
titrant and thorin as the indicator.

SOs was collected as H2S04 by using a
Goksoyr-Ross controlled condensation coil
(maintained at 60°C above the water dew
point).  The amount of SO? is determined
by $04 titratlon with O.OlN barium per-
chlorate as the titrant and thorin as
the indicator.

The amount absorbed by a 3% hydrogen
peroxide solution was determined titri-
metr.ically using 0.01N barium perchlorate
as the titrant and thorin as the Indica-
tor.
                                     242

-------
                                APPENDIX F
                       COMPARISON OF SOe MEASUREMENTS
                          BY  UV, GC, WET CHEMISTRY


     S0£ concentration 1n  the mini plant flue gas was continuously analyzed  by
a UV Instrument,  Periodic measurements of SOe concentration were also made
by analysis of grab samples by the  standard wet chemistry method  (barium
perchl orate to thorln end  point) and by GC.  The results from the UV  analysis
were consistently lower than the wet chemistry and  GC  results.  The UV results
averaged 14% lower than wet chemistry  and 29% lower than GC results.  Assuming
the UV measurements were truly low  by  14%, the Impact  on the measured $03
retention results would only be two percentage points  at the 85%  S02  retention
level and one percentage point at the  95% S02 retention level.  This  degree
of bias, if it did occur,  would have a negligible  effect on the estimate  of
dolomite requirements needed to retain 85 to  95% of the S02.

     The difference between UV and  GC  results was  not  considered  Important,
since the wet chemistry method is  generally  regarded as the  standard  method
for SOa measurement.  Therefore, it would  be concluded that  the GC results
are 15% higher than the wet chemistry results.   A  comparison  of UV, wet
chemistry and GC results  1s given in Table  F-l .
                                      243

-------
        TABLE  F-l.   COMPARISON OF S02 MEASUREMENTS
                 BY  UV, GC, WET CHEMISTRY
UV
72
62
89
131
23
106
90
85
269
279
49
602
201
140
140
82
54
3
5
7
360
150
108
319
220
310
59
322
228
320
10
14
132
28
412
23
98
210
we
57
79
-
136
24
-
-
104
251
315
.
_
-
191
.
_
-
14
.
-
437
195
180
323
256
335
116
324
208
305
8
3
88
146
406
64
327
210
GC
—
.
118
_
.
127
137
.
_
_
307
973
257
_
188
103
55
_
3
7
_
_
-
_
-
_
-
-
-
-
_
-
—
_
_
.
_
_
                                            % Difference
 Run               S02  (POT)
 No,
  69
              62      79      -            -27
                                                     -33
  70          131     136      -             -4
                                            -4
                                                     -20
                                                     -52
  71           85     104      -            -22
                                            +7
                                           -13
                                                    -527
                                                     -62
                                                     -28
  72          140     191      -            -36
                                                     -34
                                                     -26
                                                      -2
  73            3      14                  -367
                                                     +40
                                                       0
  78          360     437      -            -21
                                           -30
                                           -67

 79          319     323      -             -1
                                           -16

 80          310     335      -             -8
                                          -100

 81          322     324      -             -1

 99          228     208      -             +9
100          320     305      -             +5
102           10       8                   +20
                                           +78
103          132      88      -            +33
                                          -421
                                            +1
                                          -178
                                          -234
                                            0
                          244

-------
   TABLE  F-l  (CONT'D).  COMPARISON OF S02 MEASUREMENTS
               BY UV, GC, WET CHEMISTRY
                                           % Difference
Run
No.
104

105

UV_
134
23
44
S02 (DPm)
WC__ GC
158
36
17
UV-WC
UV
-18
-56
+61
UV-SC
u\r
— _
—
--
Averages                                14+49(1)   29+20(2)
 (1)  Three Points Excluded  *  -367, -421, -234%

 (2)  One Point Excluded  -  -527%
                            245

-------
                          APPENDIX G



               RESULTS OF FLUE GAS S03 ANALYSES
Run
No.
nj^^ ,
62

65

67
v •
69
w*'
70
/ w
71
/ 1

72
73
I **
74
75
/ *»
76
« »



$03 Analytical Concentration
Method (DOTI)
Method 8

Method 8

Method 8

Method 8

Method 8

Method 8


Method 8
Method 8
Method 8
Method 8
Controlled Cond.




1.8
0.9
34.2
0
19.6
26.5
3.6
3.6
24.4
10.6
10.0
30.0
213
2.5
2.5
0
23
0.1




Run
No.
78



79

80

81
99
100
102

103





104

105
503 Analytical
Method
Controlled Cond.



Controlled Cond.

Controlled Cond.

Controlled Cond.
Controlled Cond.
Controlled Cond.
Controlled Cond.

Controlled Cond.





Controlled Cond.

Controlled Cond.
Concentration
(ppm)
0.6
9.9
12.8
0,4
1.1
11.1
3.2
0.3
6.8
28.1
73
0.7
0
0
0
21.3
1.7
6.7
29.8
0.5
0
0
Averages 0)



All Points
Method 8 Results -
Controlled Cond. -
8.6 + 11
12 Tl2
6+9
ppm
ppm
ppm






Two Samples Excluded - Run 71  (213 ppm), Run 100 (73 ppm)
                                 246

-------
                            APPENDIX H



                      REGENERATOR GC ANALYSES
Run/
Sample
No.

102 #1
#2
#3
#4
103 #1
#2
#3
#4
#5
#6
#7
#8
#9
105 #1
#2
#3
#4
#5
#6
#7

H2S
(ppm)

15
9
< 1
75
<0.1
230
12
2
2
< 1
< 0.1
< 0.1
< 0.1
< 0.1
115*
77
180
—
100
93
Dr
CS2
(ppm)

< 1
< 1
< 1

-------
                      APPENDIX I
MASS SPECTROGRAPHIC ANALYSES OF SOLID SAMPLES OF RUN 69



El ement
Li
Be
B
F
Na
Mg
Al
Si
P
S
Cl
K
Ca
Sc
T1
V
Cr
Mn
Fe
Co
Ni
Cu
Zn
Ga
Ge
As
Se
Br
Rb
Sr
Illinois
Coal
Cone.
ug/g
50
3
400
< 2
5000
-1%
-10%
~8%
300
-7%
2000
~1%
3000
< 20
210
130
130
200
-7%
30
200
70
30
30
30
10
< 5
0.3
35
70
Filter
Part
Cone.
ug/g
30
3
125
50
3500
~12%
-2%
~3%
120
~2%
40
-1%
~12%
< 20
3000
200
80
40
6000
5
400
35
10
8
15
3
< 1
0.5
70
2000
Pfizer
Dolomite
Cone.
ug/g
20
0.01
5
34
350
~12%
500
2000
0.2
200
120
300
-12%
< 2
15
2
12
40
180
3
3
4
10
0.2
0.1
8
< QJ
5
2
150
Initial
Bed
Cone.
ug/g
10
0.5
100
17
1000
-12%
-1%
-2%
120
-2%
40
1000
-12%
< 2
300
30
40
40
6000
2
100
20
20
2
15
8
< 1
2
10
150
Final
Bed
Cone.
ug/g
30
0.5
100
30
3500
~12%
-1%
-2%
120
~Z%
40
2000
-12%
< 2
400
30
40
40
6000
1
100
20
10
2
15
2
< 1
2
10
150
Bed
Overflow
Cone.
ug/g
10
0.8
100
17
1000
-12%
500C
-2%
120
-2%
80
500
-20%
<2
300
40
120
120
6000
2
100
35
4
3
40
8
<0.03
2
10
300

2° Cyclone
Cone.
ug/g
10
2
100
17
1000
-12%
-2%
-2%
120
-2%
160
3000
-12%
< 2
1000
100
40
40
6000
3
100
35
10
3
10
8
< 0.03
2
34
300

Bed Probe
Cone.
ug/g
10
0.8
100
17
350
-16%
-1%
-1%
60
-2%
80
2000
-20%
< 2
1000
40
20
40
6000
2
100
35
10
3
60
24
< 0.05
2
7
300

-------
ISi
•>
IO
Element
   Sr
   Y
   Zr
   Nb
   Mo
   Ru
   Rh
   Pd
   Ag
   Cd
   In
   Sn
   Sb
   Te
   I
   Cs
   Ba
   La
   Ce
   Pr
   Nd
   Sm
   Eu
   Gd
   Tb
   Dy
   Ho
   Er
   Tm
   Yb
                                           APPENDIX  I  (CONT'D)

                         MASS SPECTROGRAPHIC ANALYSES  OF SOLID  SAMPLES OF RUN 69
Illinois
Coal
Cone.
yg/g
70
14
20
< 7
12
< 1
< 0.4
< 1
< 0.3
2
< 0.2
1
4
< 0.5
2
2
100
12
20
4
12
1
0.5
< 2
0.4
3
0.3
1
1
1
Filter
Part
Cone.
yg/g
2000
30
140
14
100
< 0.2
< 0.1
< 0.4
< 0.2
< 0.5
< 0.2
0.8
0.4
< 0.1
< 0.07
10
360
20
60
4
25
< 2
1
3
0.5
5
1
1
< 0.2
< 1
Pfizer
Dolomite
Cone.
yg/g
150
0.4
0.1
0.1
2
< 0.2
< 0.04
< 0.4
< 0.1
< 0.2
< 0.2
0.8
0.2
< 0.05
< 0.07
0.1
4
0.4
1
0.05
0,3
< 0.2
< 0.03
< 0.2
< 0.05
< 0.2
< 0.05
< 0.2
< 0.06
< 0.05
Initial
Bed
Cone.
ug/g
150
4
8
1
8
< 0.2
< 0.1
< 0.4
< 0.04
< 0.2
< 0.3
0.8
0.4
< 0.1
< 0.07
2
60
6
8
0.3
0.1
< 1
< 0.2
< 2
< 0.5
< 0.5
< 0.1
< 0.3
< 0.06
<0.2
Final
Bed
Cone .
yg/g
150
4
40
4
8
< 0.3
< 1
< 0.4
< 0.04
< 0.2
< 0.2
2
0.4
< 0.1
< 0.07
3
60
6
8
1
1
< 1
0.3
< 2
0.5
1
0.1
0.4
0.1
0.2
Bed
Overflow
Cone.
yg/g
300
5
14
4
8
< 0.5
< 1
< 1
< 0.1
< 0.3
< 0.5
2
1
< 0.1
< 0.07
3
60
12
28
1
10
< 4
< 0.3
< 3
< 0.05
< 2
< 0.3
< 0.4
< 0.1
< 0.3

2° Cyclone
Cone.
yg/g
300
5
40
4
8
< 0.5
< 0.4
< 1
< 0.2
< 0.7
< 1
2
1
< 0.1
< 0.2
10
140
12
28
3
15
< 3
< 0.5
< 2
< 0.6
< 3
< 0.3
< 0.3
< 0.1
< 0.3
                                                                                                   Bed  Probe
                                                                                                     Cone.
                                                                                                     yg/g
                                                                                                     300
                                                                                                       5
                                                                                                      14
                                                                                                       4
                                                                                                      16
                                                                                                     < 0,
                                                                                                     < 0,
                                                                                                         1
                                                                                                         07
< 1
< 0,
< 0,
< 0,
  8
  1
< 0,
< 0,
  1
140
  7
 14
  1
 10
< 3
< 0.5
< 2
< 0.2
< 1
                                                                                                        ,2
                                                                                                        ,4
                                                                                                        ,1
                                                                                                     < 0.3

-------
ro
en
o
                                            APPENDIX  I  (CONT'D)


                         MASS  SPECTROGRAPHIC  ANALYSES  OF SOLID  SAMPLES OF RUN 69



El ement
Lu
Hf
Ta
W
Re
Os
Ir
Pt
Au
Hg
Tl
Pb
B1
Th
U
Illinois
Coal
Cone.
yg/g
0.4
< 5
<2
< 2
< 1
< 2
< 1
< 2
< 1
< 10
< 2
< 2
< 1
< 2
< 1
Filter
Part
Cone.
yg/g
< 0.2
< 3
< 0.2
0.3
< 0.2
< 0.2
< 0.1
< 0.2
< 0.07
< 2
14
32
0.2
20
12
Pfizer
Dolomite
Cone.
yg/g
< 0.03
< 0.6
< 0.2
<0.1
< 0.06
< 0.2
< 0.1
< 0.2
< 0.07
<1
< 0.05
0.1
<0.03
0.2
0.2
Initial
Bed
Cone.
yg/g
< 0.06
< 0.6
< 0.2
< 0.2
< 0.2
< 0.2
< 0.1
< 0.2
< 0.07
< 1
<0.05
0.1
< 0.03
1
1
Final
Bed
Cone.
yg/g
0.1
< 0.6
< 0.6
< 0.2
< 0.06
< 0.2
< 0.1
<0.2
< 0.07
< 1
< 0.05
0.1
< 0.03
4
4
Bed
Overf 1 ow
tone.
yg/g
< 0.2
< 0.6
< 0.2
< 0.2
< 0.2
< 0.2
< 0.1
< 0.2
< 0.07
< 1
< 0.3
0.1
< 0.03
4
4

2° Cyclone
Cone.
yg/g
< 0.2
< 0.6
< 0.6
< 0.2
< 0.2
< 0.2
< 0.1
< 0.2
< 0.07
< 1
1
14
< 0.2
5
4

Bed Probe
Cone.
yg/g
< 0.2
< 0.6
< 0.6
<0.2
< 0.2
< 0.2
< 0.1
< 0.2
< 0.07
< 3
< 0.3
0.1
< 0.1
4
4

-------
                                               APPENDIX J
ro
Run No.

 61
 62.1
 62.3
 63
 67.1
 67.2
 67.3
 67.4
 68
 69
 70
 71
 72.1
 73.1
 74.1(1
 74.2(1
 76
77
99.2
99.3
99.4
99.5
                   Total  Mass
MINI PLANT COMPONENT MASS BALANCES

         Weight Percent

S        C Total        Ca       Reactive 02
                                    92.12
                                    91.93
                                    93.39
                                    97,77
                                    91.64
                                    96.05
                                    81.02
                                    91.19
                                   102.83
                                   100.27
                                   111.31
                                   102.69
                                   125.74
                                   105.14
                                   118.07
                                   117.83
                                   100.49
                                   102.17
                                   84.43
                                   80.10
                                   87.93
                                   86.29
100.34
100.97
102.05
101.15
99.52
99.62
99.80
99.71
100.65
99.33
99.39
99.83
99.80
99.38
97.68
98.89
99.59
99.91
91.75
93.17
94.80
96.08
77.07
33.83
293.19
227.28
82.94
102.83
94.56
90.37
155.56
111.35
102.50
123.92
78.03
125.89
66.39
68.20
97.54
128.51
83.26
67.00
62,82
87.18
87.89
88.90
90.96
98.13
109.82
110.60
86.39
98.27
105.17
104.94
126.16
117.37
157.59
122.13
136.44
136.45
119.23
115.26
88.55
92.57
110.97
S3. 77
313.29
173.99
__
__
96.63
95.83
115.42
104.81
123.27
89.09
88.03
122.35
67.59
88.50
68.13
125.22
133.60
126.76
75.73
63.61
65.93
93.43
  Mg
                                                                                    191 .19
                                                                                     84.62
 88.03
 90.05
113.82
100.42
131.83
 81.45
 83.65
104.39
 59.95
 75.93
                                                                                    81.24
                                                                                    99.75
                                                                                    84.92
                                                                                    75.12
                                                                                    60.65
                                                                                    86.48
  Solids
Inorganics
  127
  205
  264
  189
  101
  101
  106
  103
  146
   98,
  100,
  110,
   69,
   94,
   70,
   90.
   83.
.87
.09
.21
.20
.50
.98
.55
,27
.42
,93
,43
83
56
95
61
59
08
             111.77

-------
ro
in
ro
      Run No.
Total  Mass
100.2
100.4
102(2)
103.0(2)
103.1(2)
103.2(2)
103.3(2)
105.2(2)
Average
+ Is
93.88
96.58
99.88
99.86
99.89
99.79
99.87
100.60
98.79+
2.48
                                           APPENDIX  J  (CONT'D)

                                    MINI PLANT COMPONENT  MASS  BALANCES
Weight Percent

C Total       Ca
61.44
62.25
101.70
120.16
105.91
74.32
47.65
101.16
101.16+
51 .32"
81.14
108.30
93.92
97.96
100.51
102.34
101.32
92.30
105.85+
17.33
110.71
44.75
174.92
118.27
118.94
198.85
70.44
65.22
111.90+
53.65"
Reactive 02

    82.69
    82.41
    80.44
    75.91
    79.88
    77.99
    81.43
    79.99
                                                    93.37+
                                                    13.18"
      Pfizer Dolomite Used for All Runs Except as Noted.

      (1)  Grove Limestone

      (2)  Grove Limestone - Combined Combustor-Regenerator Operation
                                                                                     93.35
                                                                                     35.59
                                       91.11 +
                                       31.1
  Solids
Inorganics
   130.00
    90.83
    76.62
   104.35
    81.55
    78.41

   114.11 +
    46.15"

-------
                                APPENDIX K
                         MINI PLANT RUN OBJECTIVES

    Run N0t            _ Major Objective
60                     GBF - ejector evaluation (ejector replaced plunger
                                                 type blow back nozzle)
61                     GBF - ejector evaluation
6& vary filter         GBF - coarse alumina  filter media/large  Inlet
 \nedia and                 retaining screens
 Jblow back            GBF .  Specul1te filter media
6^ conditions
04                     GBF -  specul He/natural  gas  Injection used
6c                     GBF - Internal  baffles vs.  screens In filter bed
                              media retention
g6                     DOE Fireside Corrosion/Erosion Test - 1st attempt
67                     DOE Fireside Corrosion/Erosion Tests • 100 Hr
                              Shakedown
 g3_75                   High SOg retention
 7g                      High S02 retention using fine particle dolomite
 11                      Testing of  unit modifications for DOE materials
                          corrosion  test program -  bed sol Ids overflow  system
 7R a!                   DOE Fireside Corrosion/Erosion  Tests, 78  (250  hours),
 '                        79 (100  hours),  80  (215  hours),  81  (170  hours)
 92-96                   Test Acurex ceramic  bag filter
 gy_98                   Shakedown of auxiliary coal  feed  system
 gg                      S02 response to step change 1n  coal  sulfur content
 IQQ                     S02 response to step change In  Ca/S ratio
 IQ-J                     Shakedown of regenerator
 102-103                 Sorbent regeneration
 104                     Acurex - electrostatic predpltator
  1Q5                     Sorbent regeneration; comprehensive analysis
  106                     DOE Fireside Corrosion/Erosion Tests (265 hours)
  107-108                Acurex mobile bag house
  10g_il4                 3° cyclone efficiency tests
  •J15                     Slipstream GBF  - Exxon Mark IV
                                       253

-------
                    APPENDIX L.  MINIPLANT  FLUIDIZED  BED  COAL  COMBUSTION RUN SUMMARY
ro
in

Operating Conditions
Run Length, hrs.
Pressure, kPa
A1r Flow Rate, m3/m1n
Temperature Gradient, °C/m
Avg. Bed Temperature, °C
Superficial Velocity, m/sec
Settled Bed Height, m
Initial
Final
Expanded Bed Height, m
Coal Feed Rate, kg/hr . .
Ca/S Molar Feed Ratio-Set UJ.
Ca/S Molar Feed Rat1o-Calc. ^'
Excess Air, %
Sorbent
Coal
Flue Gas Emissions
S0?, ppm
NO', ppm
CO, ppm
C0?, %
027 %
Results
SOg Retention, %
Ca Sulfatlon, %
Lb SOp/M BTU
Lb NO^/M BTU
PD • Pfizer Dolomite
CH • Champion Coal
GL = Grove Limestone
* * No analyzer
61
8/24/77
5.0
795
17.6
4.1
940
1.77

1.4
1.5
3.6
103
0.75
--
36.05
PD
CH

—
115
156
11.6
5.6

__
52
--
0.18
(1) Ca/S
62.1
9/1/77
2.67
795
17.6
8.1
929
1.77

1.8
—
3.7
109
1.15
--
37.7
PD
CH

—
160
*
11.9
5.8

—
29
--
0.24
determined based
feed system.
(2) Ca/S calculated based
bed material .
62.3
9/1/77
5.0
795
17.6
6.5
932
1 .77

--
1.8
4.2
103
0.00
--
38.3
PD
CH

126
89
*
11.7
5.3

87
35
0.28
0.14
upon settings
upon level of
63
9/15/77
10.5
790
17.6
8.2
931
1 .80

1.8
1.7
4.1
102
0.00
--
28.9
PD
CH

632
93
162
12.2
6.3

37
38
1.41
0.15
on coal and
sulfatlon of
64
9/29/77
10.5
850
15.6
5.5
939
1 .48

1.7
2.0
3.6
95.6
0.00
—
14.6
PD
CH

593
87
*
14.7
2.7

43
33
1.25
0.13
sorbent.
spent

-------
                APPENDIX  L  (COHT'D).  MINIPLANT FLUIDIZED BED COAL COMBUSTION  RUN  SUMMARY
rsi
01
01
  Operating Conditions

Run Length, hrs.
Pressure, kPa
Air Flow Rate, m3/rnin
Temperature Gradient, °C/m
Avg. Bed Temperature, °C
Superficial Velocity, m/sec
Settled Bed Height, m
  Initial
  Final
Expanded Bed Height, m
Coal Feed Rate, kg/hr
Ca/S Molar Feed Ratio-Set
Ca/S Molar Feed Ratio-Calc.
Excess Air, %
Sorbent
Coal

Flue Gas Emissions
S0?, ppm
NO*, ppm
CO,  ppm
CO,, %
02;  %

Results
    Retention, %
Ca Sulfation, %
Lb SO?/M BTU
Lb MT/M BTU
     A
                                    65
                                 10/13/77
                                    7
                                 910-960
                                17.4-20.2
                                    6.52
                                  937
                                1.45-1.66
                                    2.0
                                    3.4
                                   95
                                    0.00

                                   41.7
                                    PD
                                    CH
80-380
13-98
20-50
 1-13
  4.6
 41
                 66
              11/29/77
                17
               930
                18
                36
               918
                 1.6
                                                                      67.1
                                                                  12/12-13/77
                67.2
             12/14/77
                                                        .6
                                                        .4
                 2.0
                 3.2
               130
                 1.25

                10
                 PD
                 ILL
                                                     370
                                                      20
                                                     205
                                                      12
                                                       1.9
                                                      89

                                                       0.68
                                                       0.03
 28
930
 19.8
 40.1
915
  1.7
  3.3
118
  1.25
  1.8
 19
  PD
  ILL
                                                                736
                                                                100
                                                                300
                                                                 14
                                                                  3.0
                                                                 74
                                                                 41
                                                                  1.60
                                                                  0.16
 25
930
 19.8
 30.1
915
  1.7
  3.4
130
  1.25
  1.2
 14
  PD
  ILL
              725
               65
              160
               15
                2.5
               77
               66
                1.43
                0.09
   67.3
 12/15/77

   10
  930
18.7-19.3
   31.5
 875-915
    1.6
    3.2
  125
    1.25
    1.5
   14
    PD
    ILL
                718
                 65
                312
                 12
                  2.5
                 77
                 50
                  1.41
                  0.09
     PD  =   Pfizer  Dolomite
     ILL  =   Illinois Coal No.
     CH  =   Champion Coal

-------
APPENDIX I (CONT'D).  M1NIPLANT FLUIDIZED BED  COAL  COMBUSTION RUN SUMMARY
            °C/m
            °C
    Operating Conditions

 Run Length, hrs.
 Pressure, kPa
 Air Flow Rate,
 Temperature Gradient,
 Avg. Bed Temperature,
 Superficial Velocity, m/sec
 Settled Bed Height, m
   Initial
   Final
 Expanded Bed Height, m
 Coal Feed Rate, kg/hr
 Ca/S Molar Feed Ratio-Set
 Ca/S Molar Feed Ratio-Calc.
 Excess  Air, %
 Sorbent
 Coal

 Flue Gas  Emissions

 S02,  ppm
 NO  ,  ppm
 CO,   ppm
 COp,  %
  fm   M

 Results
 S02 Retention, %
 Ca Sulfation, %
Lb SOp/M BTU
Lb NO^/M BTU

 ILL  »  Illinois Coal No. 6
 PD   -  Pfizer Dolomite
 *  CO Analyzer Malfunctioned
  67.4
12/16/77

 37
930
 19.3
  38
875
  1.6
                        1.8
                        3.0
                      115
                        1.25
                        1.9
                       31
                        PD
                        ILL
                      739
                      105
                      200
                       12
                       74
                       38
                       1.60
                       0.61
                                         68
                                       3/8/78
 12
930
 18.4
 16
946
  1.7
                   1.8
                   2.8
                 127
                   2.0
                   1.6
                 17
                   PD
                   ILL
                 50
                 79
                  *

                 16
                  3.2
                 98
                 60
                  0.09
                  0.11
   69
3/16/77

  11
 930
  19.0
  17.1
 947
   1.8
                 1.8
                 3.0
               132
                 1.5
                 1.5
                14
                 PD
                 ILL
              197
               56
                *
               16
                2.6
               94
               64
                0.37
                0.07
   70
3/22/78

  11.75
 930
  20.6
  28
 936
   1.9
                  1.8
                  3.0
                127
                  1.5
                  1.4
                 23
                  PD
                  ILL
                 94
                 81
                 *
                 17
                 4.0
                96
                70
                 0.20
                 0.12
  71
4/6/78

 13
930
 19.0
 35
933
  1.72
                  2.3
                  3.S
                122
                  1.52
                  1.7
                 12.6
                  PD
                  ILL
                 61
                 38
                 *
                 16.9
                  2.4
                 98
                 56
                  0.12
                  0.05

-------
ro
in
           APPENDIX  L  (CONT'D).

   Operating Conditions	

Run Length, hrs.
Pressure, kPa
Air Flow Rate, m3/m1n
Temperature Gradient,  °C/m
Avg. Bed Temperature,  °C
Superficial Velocity,  m/sec
Settled Bed Height,  rr,
  Initial
  Final
Expanded Bed Height, m
Coal Feed Rate, kg/hr
Ca/S Molar Feed Ratio-Set
Ca/S Molar Feed Rat1o-Calc.
Excess Air, %
Sorbent
Coal
Flue Gas Emissions
     S0
     N0
     ppm
  x. Ppm
CO,  ppm
c> !
     Jesuits

     S02 Retention, %
     Ca Sulfation, %
     Lb SOo/M BTU
     Lb NCT/M BTU
          A
                                     MINIPLANT FLUIDIZED BED COAL COMBUSTION RUN SUMMARY

                                     -«!;lo           72-2           73-1            73-2
                                     4/13/78         4/13/78        41978        4/19/78
                                        9.3
                                       915
                                       20.4
                                       11
                                       943
                                        1.84
                                         3
                                       116
 ,7
1.79
1.8
9.37
PD
OH
                                   60
                                   28
                                    *
                                   22.4
                                    2.1
                                   96
                                   55
                                    0.13
                                    0.04
4/13/78
   2.66
 694
  17.0
  12
 940
   2.02
   2.2
   3.9
 112
   1.79
   1.3
   1.02
   PD
   OH
                23
                50
                *
                18.8
                0.2
                99
                76
                0.04
                0.01
4/19/78
   8.5
 930
  19.6
  20
 939
   1.78
   3
 117
   2
   1
 ,8
  14
  5
9.12
PD
OH
                  4.0
                 51.0
                  *
                 17.9
                  2.0
                 99
                 67
                  0.01
                  0.07
                                             3
                                           930
                                            19.6
                                            13
                                           946
                                             1.79
                                                                                 1.7
                                                                                 2.7
                                                                               119
                                                                                 2.14
                                                                                 1.8
                                                                                 5.43
                                                                                 PD
                                                                                 OH
                  6.0
                 32.0
                  *
                 17.5
                  1.2
                 99
                 55
                  0.01
                  0.05
  74.1
4/28/78
   4.5
 930
  19.2
  40
 933
   1.68

   2.2

   3*7
 113
   7.6
   6.8
  16.5
   GL
   OH
                              1.0
                             79.6
                              *
                             19.5
                              3.3
                             99
                             14
                              0.002
                              0.13
     OH  «   Ohio,  Valley  Camp
     PD  =   Pfizer Dolomite
     GL  *   Grove  Limestone
                                  *   CO Analyzer Malfunctioned

-------
                APPENDIX  L  (CONT'D).  MINIPLANT FLUIDIZED BED COAL COMBUSTION RUN SUMMARY
ro
VI
co
    Operating Conditions

 Run Length, hrs.
 Pressure, kPa
 Air Flow Rate, m^/min
 Temperature Gradient, °C/m
 Avg. Bed Temperature, °C
 Superficial Velocity, m/sec
 Settled Bed Height, m
   Initial
   Final
 Expanded Bed Height, m
 Coal Feed Rate, kg/hr
 Ca/S Molar Feed Ratio-Set
 Ca/S Molar Feed Ratio-Calc.
 Excess  Air, %
 Sorbent
 Coal
 Flue Gas  Emissions
 S02,  ppm
 NO  ,  ppm
 CO,   ppm
 C02,  %
0      v
 2
 Results
 S02 Retention, %
 Ca Sulfation, %
Lb SO./M BTU
Lb N03/M BTU
                                        74.2
                                      4/28/78

                                         5.75
                                       930
                                        19.5
                                        27
                                       937
                                         1.71
                                         2.2
                                          .8
                                          .2
112
  4,
  8,
 18.9
  GL
  OH
                                        31.4
                                        62.7
                                        *
                                        19.0
                                        3.7
                                       98
                                       11
                                        0.07
                                        0.10
   75
5/12/78
   9.0
 922
  19.5
  30
 936
   1.71

   2.2
   2.5
   3.8
 115
   3.65
   4.1
  17.0
   GL
   OH
               173
                50
                 *
                15.2
                 3.4
                90
                22
                 0.34
                 0.07
                                76
                               6/1/78
                 77
               6/8/78
 12.3
922
 15

936
  1.32

  1.5
  2.3
  2.0
 87.3
  1,0
  1.2
 13.3
  PD
  OH
                422
                 43
                 *
                 16,0
                 2.8
                74
                60
                 0.91
                 0.07
  7.5
912
 17.5
 41
943
  1.53

  2.2
  2.0
  2.3
 92.0
  1.25
  1.2
 27.3
  PD
  OH
              446
               47
                *
               14.3
                5.1
               70
               60
                1.05
                0.08
    78.1
6/19-20/78

    34
   925
    21.2
    60
   895
     1.81

     2.0

     3.8
   101
     1.25

    34J
     PD
     ILL
              286
               45
                 *
               12.5
                5.6
               88

                0.81
                0.03
    * CO Analyzer Malfunctioned
    GL  «  Grove Limestone
    PD  •  Pfizer Dolomite
                                        OH  -  Ohio,  Valley  Camp
                                        ILL •  Illinois  Coal  No. 6
                                        *   CO Analyzer Malfunctioned

-------
                APPENDIX L (CONT'D).  MINIPLANT FLUIDIZED BED COAL COMBUSTION RUN SUMMARY
ro
tn

-------
               APPENDIX L (CONT'D).   MINIPLANT FLUIDIZED  BED  COAL  COMBUSTION RUN SUMMARY
ro
   Operating Conditions
 Run Length, hrs.
 Pressure,  kPa
 Air Flow Rate, m^/min
 Temperature Gradient,  °C/m
 Avg.  Bed Temperature,  °C
 Superficial Velocity,  m/sec
 Settled Bed Height, m
   Initial
   Final
 Expanded Bed Height, m
 Coal  Feed  Rate, kg/hr
 Ca/S Molar Feed Ratio-Set
 Ca/S Molar Feed Ratio-Calc.
 Excess Air, %
 Sorbent
Coal
Flue Gas Emissions
 SO    ppm
 NO  t  ppm
 CO,   ppm
 m    t
 n 2'  I
 Wo    «*

 Results
 S02 Retention, %
 Ca Sulfatlon, %
Lb SOp/M BTU
Lb N(T/M BTU
     /\
78.7
6/26/78
24
925
21.2
35
936
1.77
M •»
_.
3.3
141
1.45
1.7
15
PD
ILL
78.8
6/27/78
24
925
21.2
33
937
1.88
..
w —
2,9
146
1.45
__
11.6
PD
ILL
78.9
6/28/78
24
925
21.2
33
938
1.77
— —
_.
2.9
145
1.45
__
15
PD
ILL
78.10
6/29/78
24
925
21.2
26
895
1.81
— _
2.0
3.8
145
1.45
--
34.21
PD
ILL
79
7/31-8/4/78
100
925
21.4
_-
936
1.88
2.1
1.8
3.1
132
1.45
—
24.2
PD
ILL
                                      113
                                       28
                                        *
                                       14
                                        3.2
                                       95
                                       55
                                       0.27
                                       0.05
                                                 363
                                                  41
                                                   *
                                                  15.6
                                                   3.2
                                                  0.72
                                                  0.06
260
 36
   *
 16
  2.8
 90

  0.62
  0.06
286
 15
   *
 12.5
  5.5
 87

  0.81
  0.03
293
 56
   *
 14.6
  4.2
 88

  0.70
  0.10
   *  CO Analyzer Malfunctioned
   ILL  -  Illinois Coal No. 6
   PD   -  Pfizer Dolomite

-------
                APPENDIX L (CONT'D).  MINIPLANT FLUIDIZED BED COAL COMBUSTION RUN SUMMARY
ro

Operating Conditions
Run Length, hrs.
Pressure, kPa
Air Flow Rate, m^/min
Temperature Gradient, °C/m
Avg. Bed Temperature, °C
Superficial Velocity, m/sec
Settled Bed Height, m
Initial
Final
Expanded Bed Height, m
Coal Feed Rate, kg/hr
Ca/S Molar Feed Ratio-Set
Ca/S Molar Feed Ratio-Calc.
Excess Air, %
Sorbent
Coal
Flue Gas Emissions
S02, ppm
NOX, ppm
CO, ppm
rn #
UU« , h
02f %
Results
S02 Retention, %
Ca Sulfation, %
Lb SOp/M BTU
Lb NO^/M BTU
80.1
9/13-16/78
72.3
922
21.4
—
929
1 .77
2*
.1
--
3.9
126
1.41
--
27.2
PD
ILL

217
78
*
13.0
5.0

93
__
0.40
0.10
80.2
9/16-22/78
142.6
922
21.5
--
910
1.76

--
1.3
2.1
130
1.41
1.6
17.8
PD
ILL

152
121
*
12.4
3,5

95
59
0.27
0.16
81
10/9-16/78
171
912
21.4
—
933
1 .80

2.1
1.1
1.9
113
1 .56
1.3
28.5
PD
ILL

296
86
*
13.2
5.1

90
69
0.61
0.13
83
11/28/78
9.9
900
20.6
__
920
1.83

__
1.8
3.0
119
1.45
__
37
PD
ILL

— —
120
__
11.5
6


•» «•
__
0.22
84
11/29/78
6.8
900
20.5

921
1.7

— —
2.8
4.6
128
1 .45
• tm
23
PD
ILL

115
92

14.1
4.2

95
y >j
0.28
0.16
     *  CO Analyzer Malfunctioned
     PD   *  Pfizer Dolomite
     ILL  •  Illinois  Coal  No.  6

-------
ro
at
PO
                APPENDIX L (CONT'D).  MINI PLANT FLUIDIZED BED COAL COMBUSTION RUN SUMMARY

Operating Conditions
Run Length, hrs.
Pressure, kPa
Air Flow Rate, m-Ymin
Temperature Gradient, °C/m
Avg. Bed Temperature, °C
Superficial Velocity, m/sec
Settled Bed Height, m
Initial
Final
Expanded Bed Height, m
Coal Feed Rate, kg/hr
Ca/S Molar Feed Ratio-Set
Ca/S Molar Feed Ratio-Calc.
Excess Air, %
Sorbent
Coal
Flue Gas Emissions
S02, ppm
NO , ppm
CO , ppm
C0?, %
02, «
Results
85
11/30/78
6
900
21 .0
--
918
1.7

--
2.8
4.6
132
1.45
--
21
PD
ILL

146
100
__
--
4.1

86
12/1/78
9
900
20.8
--
921
1.7

--
2.9
4.7
125
1.45
--
24
PD
ILL

50
100
--
13
4.3

87
12/4/78
11.3
900
20.8
--
918
1 .7

--
2.9
4.7
122
1.45
— -
18
PD
ILL

150
105
--
16
3.7

88
12/5/78
9.5
900
20.6
~
918
1.7

--
2.0
3.3
115
1.45
--
28
PD
ILL/V

145
105
__
11.4
5.1

89
12/6/78
10.7
900
19.2
-_
917
1.7

--
2.2
3.6
101
1.4
— —
23
PD
V/CH

155
150
—
14
4.8

    S02  Retention,  %
    Ca Sulfatlon, %
    Lb SO?/P BTU
    Lb NO:/M BTU
                            94
991
93
    PD
    ILL
Pfizer Dolomite
Illinois Coal  No. 6
93
85
0.35
0.17
ILL/V
V/CH
0.12 0.38
0.18 0.19
« Illinois/Valley Camp
« Valley Camp/Champion
0.39
0.20

0.37
0.26


-------
                APPENDIX L (CONT'D).  MINIPLANT FLUIDIZED BED COAL COMBUSTION RUN SUMMARY

-------
            APPENDIX L (CONT'D).  MINIPLANT FLUIDIZED BED COAL COMBUSTION RUN SUMMARY

Operating Conditions
Run Length, hrs.
Pressure, kPa
Air Flow Rate, nr/mln
Temperature Gradient, °C/m
Avg. Bed Temperature, °C
Superficial Velocity, m/sec
Settled Bed Height, m
Initial
Final
Expanded Bed Height, m
Coal Feed Rate, kg/hr
Ca/S Molar Feed Ratio-Set
Ca/S Molar Feed Rat1o-Calc.
Excess Air, %
5» Sorbent
* Coal
Flue Gas Emissions
SO ppm
NO , ppm
CO , ppm
C02, %
02 , *
95
12/15/78
5.5
900
18.1
__
914
1.6

--
1.8
3.0
98
1.5
__
21
PD
CH

290
102
90
8.9
4.2
96
12/18/78
21.0
900
17.9

927
1.6

_.
2.3
3.8
103
1.5
»_
36
PD
CH

250
100
40
14
6.2
97
1/23/79
5.5
912
17.7
64
902
1.53

2.1
1.8
3.0
116
*
_-
38.4
PD
*CH/ILL

**
124
**
11.7
6.3
98
1/24/79
6.25
922
18.1
64
904
1.55

—
--
3.1
109
*
--
37.3
PD
*CH/ILL

**
111
**
7.8
6.2
99.1
1/25/79
4.0
912
17.4
74
895
1.41

—
--
3.0
90.3
1 .40
—
34.9
PD
CH

203
120
435
9.7
6.0
ResuUs

S0£  Retention,  %
Ca Sulfation, %
Lb SO?/M BTU
Lb NCT/M BTU
                            73

                             0.69
                             0.17
78

 0.55
 0.16
      0.18
0.17
PD
CH
*
Pfizer Dolomite
Champion Coal
Switching between Champion
(Ca/S - 1.4) and Illinois (Ca/S
 "CH/ILL
83

 0.51
 0.21
                                                        **
                                         0.76) coals
Champion and Illinois Coals used - Champion
Analyses used when necessary.
S02 Response Test - No Steady  Emissions

-------
ro
o»
en
           APPENDIX  L  (CONT'D).

   Operating Conditions

Run Length, hrs.
Pressure, kPa
Air Flow Rate, m3/niin
Temperature Gradient, °C/m
Avg. Bed Temperature, °C
Superficial Velocity, m/sec
Settled Bed Height, m
  Initial
  Final
Expanded Bed Height, m
Coal Feed Rate, kg/hr
Ca/S Molar Feed Ratio-Set
Ca/S Molar Feed Rat1o-Calc.
Excess Air, %
Sorbent
Coal
Flue Gas Emissions
S02, ppm
NO  • ppm
CO;  ppm
CO,, %
02,  *
Results

S02 Retention, %
Ca  Sulfation,  %
Lb SO?/M BTU
Lb N(T/M BTU
     A
                                     MINIPLANT FLUIDIZED BED  COAL  COMBUSTION  RUN  SUMMARY
                                       99.2           99,3           99.4           99.5
                                     1/25/79
                                        3.0
                                      912
                                       16.9
                                       62
                                      911
                                        1.38
 3,0
93.0
 1 .40
 1.2
33.9
 PD
 CH
                                       194
                                       125
                                        61
                                        10.0
                                         5.9
                                        85
                                        70
                                         0.46
                                         0.21
             1/25/79

                3.67
              912
               16
               56
              910
                1.37
  ,9
  .2
 3.3
99.4
 0.76
 0.95
30.1
 PD
 ILL
               740
               110
                34
                 9.39
                 5.6
                70
                73
                 1.87
                 0.20
1/25/79
   4.0
 912
  16.9
  58
 911
   1.37
   3.7
  99.4
   0.76
   0.93
  30.0
   PD
   ILL
              721
              113
               15
               11.3
                5.4
               70
               75
                1.82
                0.21
1/25-26/79

     7.0
   912
    16.8
    63
   913
     1.37
     3.9
    82.8
     1.40
     1.1
    33.1
     PD
     CH
                 171
                 121
                  11
                  10,
                   5.8
                  85
                  75
                   0.45
                   0.23
  99.6
1/26/79

   6.0
 912
  16.5
  62
 913
   1.35
   3.9
  81.8
   1.40

  32.2
   PD
   CH
                   210
                   118
                     6
                    11.2
                     5.7
                    82

                     0.55
                     0.22
     PD    «   Pfizer  Dolomite
     CH   --   Champion Coal
     ILL   =   Illinois Coal No. 6

-------
                 APPENDIX L (CONT'D).   MINIPLANT  FLUIDIZED BED COAL COMBUSTION RUN SUMMARY
ro
                             °C/m
                             °C
   Operating Conditions

Run  Length, hrs.
Pressure,  kPa
Air  Flow Rate, m^/min
Temperature Gradient,
Avg.  Bed Temperature,
Superficial Velocity, m/sec
Settled Bed Height, m
   Initial
   Final
Expanded Bed Height, m
Coal  Feed  Rate, kg/hr
Ca/S Molar Feed Ratio-Set
Ca/S Molar Feed Ratio-Calc.
Excess Air, %
Sorbent
Coal
Flue Gas Emissions
      S0
      N0
     ppm
  x, PPm
CO,  ppm

C°2« 5
      Results
      S02 Retention, %
      Ca Sulfation, %
      Lb  SOo/M  BTU
      Lb  NO^/M  BTU
  99.7
1/26/79

   6.83
 912
  16.5
  61
 913
   1.34
   2.4
   4.0
  81.0
   1.40
   1.0
  30.6
   PD
   CH
                                  183
                                  120
                                   11
                                   12.24
                                    5.42
                                  84
                                  84
                                   0.48
                                   0.23
 100.1
1/30/79

   8,0
 912
  17.6
  61
 914
   1.45

   2.1
  85.2
   1.43
   0,90
  41.9
   PD
   CH
                360
                135
                 83
                 14
                  6.3
                 67
                 75
                  0.93
                  0.25
 100.2
1/30/79

   4.0
 912
  17.6
  61
 914
   1 .45
   3.5
  80.9
   0.38
   0.68
  38.2
   PD
   CH
                508
                100
                 40
                 10
                  5.4
                 51
                 76
                  1.39
                  0.20
  100.3
1/30-31/79

    2.0
  912
   17.6
   65
  916
    1,41
    3.5
   85.8
    0.38

   35.3
    PD
    CH
               488
               129
                61
                15.4
                 6.1
                56

                 1.25
                 0.24
                                                                                               100.4
                                                                                              1/31/79

                                                                                                 7.0
                                                                                               912
                                                                                                17.6
                                                                                                57
                                                                                               913
                                                                                                 1.43
                                                                                                 3.6
                                                                                                83.9
                                                                                                 1.43
                                                                                                 0,89
                                                                                                16.1
                                                                                                 PD
                                                                                                 CH
                 364
                 135
                  76
                  12.2
                   3.3
                  66
                  75
                   0.96
                   0.25
      PD   »  Pfizer Dolomite
      CH   =  Champion Coal

-------
               APPENDIX L  (CONT'D).  MINIPLANT FLUIDIZED BED COAL COMBUSTION  RUN SUMMARY
ro
   Operating Conditions
Run Length, hrs.
Pressure, kPa
Air Flow Rate, m3/m1n
Temperature Gradient, °C/m
Avg. Bed Temperature, °C
Superficial Velocity, m/sec
Settled Bed Height, m
  Initial
  Final
Expanded Bed Height, m
Coal Feed Rate, kg/hr
Ca/S Molar Feed Ratio-Set
Ca/S Molar Feed Rat1o-Calc.
Excess A1r, %
Sorbent
Coal
Flue Gas Emissions
     SO   ppm
     NO ,  ppm
     CO,  ppm
     C0?,  %
     02f  %
     Results
     S0? Retention,  %
     Ca Sulfatlon, %
     Lb SOo/M BTU
     Lb NO^/M BTU
          A
                                       100.5
                                      1/31/79

                                         6.0
                                       912
                                        17.6
                                        53
                                       912
                                         1.43
 3.8
83.6
 1.43

37.5
 PD
 CH
                                  274
                                  129
                                   86
                                   11.5
                                    5.4
                                   74

                                    0.72
                                    0.24
              100.6
             1/31/79
                4.5
              912
               17.5
               54
              913
                1.43
 2.3
 3.8
83.7
 1.43
 0.96
36.9
 PD
 CH
              311
              135
               71
               10.8
                6.2
               71
               74
                0.81
                0.25
    T02*
3/13-17/79

   107
   700
    12.5
   -15
   908
     1.43

     2.1
     1.0
     2.2
    69.5
     1.5
    19.
    20.3
     GL
     CH
               27
              126
               90
               14.1
                4.03
               97
                5
                0.06
                0.19
   103.0*
3/29-4/1/79

    81
   700
    13.4
   -12
   913
     1.52

     2.1

     2.7
    76.6
     1.35
     5.4
    16.9
     GL
     CH
                   28
                   77
                   71
                   13.8
                    3.32
                   97
                   18
                    0.06
                    0.12
                                            103.1*
                                           4/1-3/79

                                             49
                                            700
                                             13.5
                                            -14
                                            901
                                              1.52
 2.8
77.7
 0.68
 4.1
17.7
 GL
 CH
                   41
                   56
                   62
                   14.2
                    3.5
                   96
                   23
                    0.09
                    0.09
     PD   «   Pfizer Dolomite
     GL   =   Grove Limestone
     CH   •   Champion Coal
                                        * Combined Combustor-Regenerator Runs;
                                          Combustor Data Only

-------
                APPENDIX L (CONT'D).  MINIPLANT FLUIDIZED BED  COAL  COMBUSTION  RUN SUMMARY
r\>
en
CD
    Operating  Conditions
 Run Lenght,  hrs.
 Pressure,  kPa
 Air Flow Rate,  m3/min
 Temperature  Gradient,  °C/min
 Avg.  Bed Temperature,  °C
 Superficial  Velocity,  m/sec
 Settled  Bed  Height, m
   Initial
   Final
 Expanded Bed  Height, m
 Coal  Feed  Rate, kg/hr
 Ca/S  Molar Feed Ratio-Set
 Ca/S  Molar Feed Ratio-Calc.
 Excess Air, %
 Sorbent
 Coal
 FlMe  Gas Emi s sions

S02,  ppm
NO^.  ppm
CO;   ppm
     Results
     S02 Retention, 5
     Ca Sulfation,  %
     Lb SO-/M BTU
     Lb NO^/M BTU
     GL  a  Grove  Limestone
     PD  «  Pfizer Dolomite
     CH  «  Champion  Coal
                                       103.2*
                                      4/3-4/79
                                        22
                                       700
                                        14,2
                                       -12
                                       901
                                         1.59
  2.0
 79.4
  0.68
  4.4
 13.1
  GL
  CH
                                       22
                                       65
                                       64
                                       14.2
                                        2.7
98
22
 0.05
 0.10
             103.3*
            4/4-6/79

              59
             700
             14.4
             -18
             902
               1.62
  1.4
  2.2
 79.9
  0.93
  3.3
 19.0
  GL
  CH
             77
             55
             56
             13.8
              3.7
93
28
 0.17
 0.09
                    104
                4/17-20/79
      83.5
     850
      11.5
     -42
     899
       1.62

       2.1
       1.8
      3.0
      93
       1.49
       1.4
     42.1
      PD
      CH
                71(5-160)
                  81
                 104
                  13.7
                   3.8
   93(99-84)
     66
0.16(0.01-0.37)
      0.13
         Combined  Combustor-Regenerator  Runs;
         Combustor Data  Only
    105*
4/30-5/5/79

    99
   700
    14.0
   -56
   894
     1.46
                                                                                           2.2
                                                                                           1.7
                                                                                           3
                                                                                          77
       1
                                                                                             29
                                                                                             0
    45.8
     GL
     CH
                         70
                         59
                         77
                         12.5
                          4.1
   93
   23
    0.15
    0.09

-------
ro
o>
vo
           APPENDIX L (CONT'D).   MINIPLANT FLUIDIZED BED COAL COMBUSTION RUN SUMMARY
                                  106.1           106.2          106.3           107            108
   Operating Conditions       5/29-6/10/79   5/29-6/10/79   5/29-6/10/79    6/18-23/79     6/26-29/79

Run Length, hrs.                  43              5            217             108             71
Pressure, kPa                    902            902            912             912            912
Air Flow Rate, m3/m1n             20.6           19.7           20.2            ie.5           16.4
Temperature Gradient, °C/m       -64            -69            -"             -71           -TOO
Avg. Bed Temperature, °C         914            904            908             872            877
Superficial Velocity, m/sec        1.71            1.62           1.64            1.31           1.31
Settled Bed Height, m
  Initial                          2.0            -             --              2.1            2.0
  Final                            -             --             J-?             LI            0.96
Expanded Bed Height, m             2.4            3.3            *•'             1.7            1.6
Coal Feed Rate, kg/hr            139            145            '**              84.4           86.3
Ca/S Molar  Feed Ratio-Set          L35           1.19           1-25            1.25           1.25
Ca/S Molar  Feed Ratfo-Calc.        —             ~             "              ] «4            1 -2
Excess Air, %                     15.6            5.8            9.0            31.4           23.0
Sorbent                            PD             PD             «>              PD             PD
Coal                               ILL            ILL            ILL             CH             CH

Flue Gas Emissions

S02, ppm                          21.4           37.8           91.4            41.5           27.8
NO  , ppm                          47.5           42.4           64.1            26.4           44.6
CO,  ppm                         332            507            540              64.7          130.8
C0?, %                            14.8           16.4           13.4            12.8           14.2
02.  %                             3.1            1.1            2.1             5.1            4.0
Results
S02 Retention, %                  93             95             99              99             98
Ca Sulfatlon, %                    _.             __             __             72             79
Lb S02/M BTU                       0.02           0.10           0.11            0.04           0.0
Lb NO^/M BTU                       0.06           Q.06           0.08            0.13           0.08
     PD   «  Pffzer Dolomite
     CH   -  Champion Coal
     ILL  •  Zllfnofs Coal  No.  6

-------
                  APPENDIX  L  (CONT'D).   MINIPLANT FLUIDIZED BED COAL COMBUSTION RUN SUMMARY
^J
o

Operating Conditions
Run Length, hrs.
Pressure, kPa
A1r Flow Rate, m3/m1n
Temperature. Gradient, °C/m
Average Bed Temperature, °C
Superficial Velocity, m/sec
Settled Bed Height, m
Initial
Final
Expanded Bed Height, m
Coal Feed Rate, kg/hr
Ca/S Molar Feed Ratio-Set
Ca/S Molar Feed Rat1o-Ca1c.
Excess Air, %
Sorbent
Coal
Flue Gas Emissions
S0?. ppm
N(£, ppm
CO, ppm
CO-. %
n *- #
* Near end of run Ca/S « 0,
109
7/25/79
9.5
915
17.6
—
870
1.65

—
..
2.5
87
1.25
—
19.5
PD
CH

70
140
175
15
3.5
sorbent
110
111
7/26/79 7/27/79
6.5
890
17.6
--
870
1.65

—
.-
2.8
98
1.25
--
39
PD
CH

50
110
175
15.5
6.0
hopper
** Auxiliary feed system used to switch coal
N.A. • Data not available.
PD " Pfizer Dolomite
CH • Champion Coal
ILL » Illinois Coal No. 6








8.0
900
17.6
—
870
1.65

--
-.
2.9
98
1.25
--
16
PD
CH

160
90-175
200
17.5
3.0
empty
feed.




112
7/30/79
6.5
900
13.1
--
870
1.22

—
--
3.4
79
1.25*
—
19.5
PD
CH

60(500)*
175
200
17.5
3.5






113
7/31/79
8.2
915
17.6
--
870
1.65

--
--
3.4
93
1.25(0.76)**
—
23
PD
CH/ULL)**

120(600)**
145
200
16.5
4.0






114
8/1/79
4.6
915
17.6
--
870
1.65

—
--
3.4
95
1.25
--
21.5
PD
CH

no
90
N.A.
16
3.8






115
8/7/79
6.7
915
17.6
--
870
1.65

--
--
3.5
99
1.25
--
25
PD
CH

50
N.A.
200
15
4.0







-------
                     APPENDIX M-l .   PARTICLE  SIZE  DISTRIBUTION  SPENT  PFIZER  1337
                                   DOLOMITE  SORBENT (EXCEPT AS NOTED)
ro
Particle Size

Run No.
61
62
63
64
65
65
66
67
68
69
70
71
72
73
74(1)
75(1)
76
78

80
81
81
81
99







Material
Initial Bed
Final Bed
Initial Bed
Final Bed
Initial Bed
Final Bed
Final Bed
Final Bed
Final Bed
Final Bed
Final Bed
Final Bed
Final Bed
Final Bed
Final Bed
Final Bed
Final Bed
Bed Overflow
Dump No. 72
Bed Overflow
Initial Bed
Final Bed
Bed Overflow
Bed Probe No. 1
No. 3
No. 5
No. 11
No. 12
No. H
No. 32
5%
Less Than
___
230
800
550
500
420
240
...
300
300
175
96
100
115
600
420
150
105

130
760
620
125
820
600
510
-__
640
700
—
10%
Less Than
• • —
290
920
650
740
560
330
445
460
450
210
150
125
160
700
520
200
135

150
940
760
425
935
890
600
910
750
820
—
25%
Less Than
175
510
1080
820
900
770
700
740
650
850
300
290
185
280
810
780
350
210

205
1100
940
740
1200
970
830
1200
1025
1100
1100
50%
Less Than
1004
880
1360
1075
1200
1075
1325
1755
870
1200
480
640
360
620
1020
1000
690
430

530
1350
1300
1075
1550
1250
1100
1550
1325
1400
1400
(Vim)
75%
Less Than
1520
1230
1800
1400
1500
1450
1875
2150
1150
1630
980
1150
980
1100
1350
1250
1170
1000

1450
1600
1700
1450
1900
1600
1400
1850
1775
1750
1700

90%
Less Than
2000
1700
2040
1850
1900
1900
2200
2330
1560
2000
1380
1600
1400
1400
1700
1600
1500
2000

2000
1900
1950
1850
—
1950
1750
--
2000
--
2000

95%
Less Than
2200
1900
2200
2050
2100
2150
2300
2500
1800
3000
1700
1850
1700
1600
1900
1800
1780
5600

2500
2500
2350
2250
--
--
1950
—
--
--
—
    (1)  Grove Limestone Sorbent

-------
ro
^j
ro
                    APPENDIX  M-2.   PARTICLE SIZE DISTRIBUTION PRIMARY CYCLONE DIPLE6




                                                   Particle Size (ym)
Run
No.
61
62
65
67
81
85
88
5%
Less Than
.._
—
—
350
—
—
___
10%
Less Than
180
—
680
33
60
77
25%
Less Than
—
900
no
1375
54
80
97
50%
Less Than
140
1300
272
1950
105
no
140
75%
Less Than
185
1850
880
2175
170
180
190
90%
Less Than
890
2100
1400
2435
345
500
290
95%
Less Than
1250
2300
1750
2600
—
—
610

-------
                         APPENDIX M-3.  PARTICLE SIZE DISTRIBUTION SECONDARY CYCLONE CAPTURE
CJ



Run/Sample No.
61
61
61
62

63

64


65

67








68.1
68.2

69
69.1
69.2
70.1
70.2
71
72.1
73.1



No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.



6
7
3
9
2
3
6
2
6
7
12
20
25
30
33
44
49
50
6
8.1
8.2
7
5
10
6
10
4
8.1
6
5%
Less Than
4.4
5.0
3.4
4.0
3.7
3.5
4.0
3.5
4.0
4.0
5.0
4.4
5.0
4.9
5.0
4.7
4.5
4.0
4.0
4.8
2.1
8.0
10.0
10.0
2.0
8.6
7.0
8.2
7.6
5.0
3.8
4.8
10%
Less Than
6.4
6.8
4.5
5.6
5.0
5.4
6.0
5.0
5.0
5.6
6.5
5.8
7.1
7.3
7.0
6.9
6.4
5.8
6.1
7.2
2.6
13.0
13.9
14.0
3.0
10.8
10.0
11.3
10.0
6.6
5.0
6.4
Particle Size
25% 50%
Less Than
11.4
13.0
6.9
10.5
9.4
10.0
11.0
9.8
9.8
10.0
11.5
10.0
12.5
13.0
12.0
11.5
10.5
10.5
11.3
13.0
4.3
26.5
35
32
6.5
19
16
30
17.5
10.5
7.6
12.5
Less Than
24
26
11
20
17
19
23
18
16
18
20
17
22
23
20
20
19
18
19
23
9.2
48
64
60
11
50
28
53
29
16
13
25
(ym)
75%
Less Than
42
48
18
50
36
38
49
30
30
35
35
31
39
38
35
34
31
34
33
39
16
84
100
100
17
69
50
59
50
25
25
43
90%
Less Than
60
70
25
70
73
62
120
50
58
56
54
50
57
59
49
50
47
51
47
53
24
—
112
no
24
119
90
86
160
35
40
64
95%
Less Than
90
90
31
105
113
105
230
60
80
86
70
67
100
80
60
60
55
69
63
70
32
--
120
115
30
230
200
130
215
50
54
100

-------
                    APPENDIX M-3 (CONT'D).  PARTICLE SIZE DISTRIBUTION SECONDARY CYCLONE CAPTURE
r\j
Run/Sample No.
74.1 No.
74.2 No.
75
76
78


79
80
81


87
96
99






TOO






102
103

104

No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
4
10
6
6
52
85
130
18
64
15
74
86
5
4
13
14
15
21
22
23
36
4
5
6
12
13
14
31
30
33
99
6
30
5%
Less Than
14
8.0
4.3
7.0
4.4
6.0
4.4
7.4
5.9
5.8
5.4
6.4
52
4.4
10.0
13.5
11.0
9.0
10.5
14.0
12.5
10.5
10.5
9.0
9.0
7.4
9.0
8.2
5.2
4.3
4.2
4.4
3.8
10%
Less Than
20
14.0
5.8
9.6
6.0
8.2
6.0
10.0
8.1
8.0
8.0
8.4
70
6.4
15.0
20
15.0
12.4
14.0
19
18
14.0
15.0
13.0
12.5
10.5
13.0
11.5
7.2
5.6
5.8
5.6
5.2
Particle Size (ym)
25% 50% 75%
Less Than Less Than Less Than
72
38
10
15
9.6
13
9.6
15
13
14.5
15
15.5
115
11.5
29
38
27
23.5
26.5
35
38
28
28.5
26
24.5
19.5
27
22.5
10.3
9.4
9.8
9.8
9.0
430
18.5
27
16
23
15.5
28
24
33
36
35
165
21
61
88
56
51
55
68
98
58
68
60
56
40
64
50
32
15
18
16
16
850
35
70
32
43
32
55
47
92
100
86
240
47
135
170
135
130
125
150
215
115
175
155
150
92
155
125
96
26
54
33
32
90%
Less Than
—
150
127
63
74
68
92
92
—
--
155
370
100
270
330
275
275
300
300
370
212
360
310
310
190
300
270
220
50
160
72
66
95%
Less Than
—
--
--
90
100
94
120
120
--
--
—
460
150
385
450
390
410
370
410
470
300
480
425
425
300
420
385
320
94
260
135
105

-------
                   APPENDIX M-3 (CONT'D).  PARTICLE SIZE DISTRIBUTION SECONDARY CYCLONE  CAPTURE
        Run/Sample No.
        105  No. 28
             No. 44
        106  No.  2
             No. 11
             No. 23
             No. 57
        107  No. 15
5%
Less Than
3.7
4.2
5.0
7.0
4.2
4.8
4.6
10%
Less Than
5.0
5.6
6.6
14.0
6.0
6.4
6.0
Particle Size
25% 50%
Less Than Less Than
8.5
10.0
10.5
34
11.5
11.5
10.8
14
17
18
84
23
23
21
(ym)
75%
Less Than
27
32
38
165
52
52
48
90%
Less Than
64
82
83
310
105
105
100
95%
Less Than
125
150
140
430
160
160
150
IM
->J
cn

-------
APPENDIX M-4.  PARTICLE SIZE DISTRIBUTION TERTIARY CYCLONE CAPTURE



Run/Sample No.
67







68.1
68.2
71
72.1
73.1
74.1
74.2
5

76
78



79



80
81

86

No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.

7
12
20
25
30
33
43
48
6
8
4
8
6
4
10
6.1
6.2
6
31
52
85
120
1
18
42
50
64
15
74
5


Less
1
1
1
2
1

1
1
1
1
1
1
1
1
1
1
1
1
1
1
2
1
1
1
2
2
1
1
1
1
1
5%
Than
.30
.30
.80
.30
.55
--
.60
.60
.50
.50
.35
.15
.30
.20
.50
.25
.85
.60
.40
.80
.10
.65
.75
.60
.40
.20
.35
.20
.80
.60
.80
10%
Less
1
1
2
1
2
1
2
2
1
2
1
1
1
1
1
1
2
2
1
2
2
2
2
2
2
2
1
1
2
2
2
Than
.60
.75
.30
.66
.00
.15
.00
.10
.90
.00
.75
.40
.60
.50
.90
.60
.10
.00
.75
.30
.50
.20
.15
.10
.90
.90
.70
.40
.20
.10
.30
Particle Size (pm)
25% 50% 75%
Less Than
2.3
2.6
4.3
2.4
3.4
1.6
3.5
3.8
3.0
3.4
2.8
2.1
2.3
2.1
3.0
2.3
2.8
3.0
2.5
3.5
3.4
3.1
2.7
2.9
4.2
4.3
2.5
1.9
3.1
3.1
3.2
Less Than
4.0
4.7
7.4
4.2
6.4
2.5
6.6
7.2
5.2
6.1
4.9
3.5
4.0
3.4
5.0
4.2
5.3
5.0
3.5
5.1
4.8
4.5
3.4
4.4
5.9
6.1
3.9
3.3
4.8
4.7
4.7
Less Than
7.0
8.4
12.3
7.3
11.0
4.6
10.5
11.5
8.8
10.5
8.2
6.0
7.6
5.8
8.8
8.4
--
10.0
5.2
7.6
6.6
6.6
4.5
6.4
8.9
8.8
6.0
6.2
7.8
7.1
7.2
90%
Less Than
12.0
12.5
17.5
12.0
18
8.0
13.5
16
17
16
15.0
9.5
30
8.5
23
32
--
44
8.4
10.0
8.8
10.0
8.0
9.6
12.5
11.5
8.6
10.0
12.0
10.0
11.0
95%
Less Than
18
16
25
20
25
11.0
16
19
__
30
44
16
100
10.5
100
__
--
__
11.0
12.5
10.0
13.0
11.8
10.2
16
12.0
9.8
12.7
14.3
12.7
14.0

-------
                      APPENDIX M-4  (CONT'D).   PARTICLE  SIZE  DISTRIBUTION TERTIARY CYCLONE CAPTURE
f\>
Particle Size (ym)

Run/Sample No.
87 No. 5

96 No . 4
No. 10
99 No. 13
No. 15

No. 21
No. 23
No. 34
No. 36
100 No. 2
No. 4

No. 6
No. 10
No. 12
No. 14
No. 31
102 No. 30
103 No. 33
No. 52
No. 75
No. 99
104 No. 6
No. 18
No. 30
No. 42
105 No. 28
No. 44
108 No. 5
5%
Less Than
1.75
2.10
1.45
1.40
1.42
1.30
1.42
1.30
1.45
1.13
1.42
1.10
1.13
1.30
1.26
1.20
1.40
1.30
1.15
1.05
1.10
1.10
1.35
1.20
1.43
1.60
1.45
1.50
1.59
1.45
1.59
10%
Less Than
2.30
2.70
1.80
1.80
1.75
1.59
1.70
1.59
1.70
1.35
1.65
1.40
1.35
1.70
1.45
1.45
1.70
1.50
1.40
1.36
1.41
1.35
1.70
1.40
1.70
1.95
1.78
1.80
1.90
1.80
1.90
25%
Less Than
3.4
3.7
2.6
2.6
2.4
2.2
2.4
2.2
2.2
1.8
2.2
1.9
1.8
2.2
2.0
1.9
2.2
2.1
1.9
2.1
2.1
1.9
2.5
1.9
2.3
2.6
2.5
2.6
2.6
2.5
2.5
50%
Less Than
5.1
5.3
3.8
3.9
3.3
3.1
3.4
3.1
3.3
2.6
2.9
2.7
2.6
3.1
2.8
2.7
3.0
2.8
2.6
3.4
3.3
2.8
3.6
2.8
3.3
3.9
3.7
3.9
3.4
3.4
3.3
75%
Less Than
7.0
8.0
6.3
6.3
5.2
5.4
5.8
4.9
5.3
3.8
4.2
3.9
3.8
5.0
4.0
3.9
4.4
3.9
3.8
5.5
5.0
4.5
5.4
4.4
5.0
6.3
5.8
6.6
4.8
4.9
4.6
90%
Less Than
11.0
12.0
11.0
10.0
8.4
8.8
9.8
7.6
8.8
5.8
6.3
6.0
6.3
8.0
6.8
6.0
7.0
6.3
5.9
8.5
8.0
6.9
8.0
6.9
7.3
9.8
8.8
9.8
7.1
7.0
7.1
95%
Less Than
13.0
15.0
15.0
12.0
10.0
10.0
12.0
9.6
11.0
7.6
8.4
7.8
9.2
10.0
9.4
7.8
9.2
8.0
7.4
10.0
10.9
9.2
10.0
9.1
8.9
12.0
10.0
12.0
8.8
8.8
9.4

-------
                                 APPENDIX M-5.  PARTICLE SIZE DISTRIBUTION AND GRAIN

                               LOADING - FLUE GAS PARTICULATES BEFORE TERTIARY CYCLONE
ro
~j
oo
/o \Grain Loading
Run/Sample No/ ; gr/SCF
96 BF-2
99 BF-4
99 BF-6
100 BF-1
100 BF-4
100 BF-6
100 BF-8
106 BF-2
106 BF-4
106 BF-1 4
106 BF-18
106 BF-21
106 BF-25
106 BF-35
106 Avg. (1)
106 Avg. (2)
107 BF-5
107 BF-7
107 BF-11
108 BF-2
108 BF-6
109 BF-2
109 BF-4
109 BF-6
110 BF-2
110 BF-5
110 BF-8
0.463
0.335
0.364
0.492
0.476
0.635
0.346
3.451
1.771
0.563
0.718
0.623
0.739
0.212
1.09
0.583
0.402
0.597
0.49
0.336
0.728
0.819
0.540
0.522
0.598
0.546
0.593
5%
Less Than
1.26
1.16
1.15
1.35
1.00
1.00
1.20
1.50
1.65
1.70
1.75
1.70
1.80
1.50
1.70
1.79
1.40
1.35
1.33
1.45
1.50
0.84
1.15
1.35
1.35
1.60
1.50
10%
Less Than
1.59
1.40
1.45
1.75
1.26
1.30
1.45
1.90
2.1
2.2
2.2
2.2
2.4
1.80
2.1
2.2
1.50
1.45
1.45
1.70
1.79
1.08
1.42
1.70
1.60
1.90
1.78
Particle Size (ym)
25% 50% 75%
Less Than Less Than Less Than
2.2
1.9
2.1
2.6
1.7
1.8
2.0
3.0
3.4
3.5
3.4
3.3
3.6
2.6
3.3
3.4
1.9
1.8
1.8
2.5
2.6
1.5
1.9
2.3
2.3
2.6
2.7
3.1
2.9
3.0
3.6
2.4
2.4
2.8
5.0
5.6
6.0
5.6
5.6
6.0
4.2
5.7
5.8
2.8
2.7
?.8
3.8
3.5
2.2
2.6
3.0
3.2
3.3
3.3
4.5
4.8
5.2
5.5
3.9
3.5
4.0
8.6
8.4
10.0
9.1
9.2
9.4
7.0
9.1
9.3
5.8
4.6
6.0
7.0
5.5
3.5
3.7
4.2
4.4
4.4
4.6
90%
Less Than
6.35
9.0
8.6
8.0
7.6
5.7
6.0
14.0
12.5
14.0
12.7
13.5
13.0
10.5
13.1
13.1
12.0
8.4
11.0
11.4
9.0
5.8
5.0
5.8
6.0
5.6
6.0
95%
Less Than
8.4
11.0
11.0
10.0
10.0
8.0
8.0
18
14.8
17.3
15.0
16
16
14.0
16.2
16.2
16
11.8
14.3
14.0
11.6
7.2
6.0
7.0
7.1
6.4
7.1
      (1)  All
      (2)  Excluding 2, 4

      (3)  BF indicates Balston filter samples taken upstream of the tertiary cyclone.

-------
                       APPENDIX M-5 (CONT'D).   PARTICLE SIZE  DISTRIBUTION  AND  GRAIN
                         LOADING - FLUE GAS PARTICIPATES BEFORE  TERTIARY CYCLONE
 Run/Sample No.
                (2)
                                         Particle Size (ym)
Grain Loading     5%       102       25%       5Q%       75%        90%
    gr/SCF    Less Than Less Than Less Than Less Than  Less Than  Less  Than Less Than







ro
^J
10


111
111
112
112
112
113
113
113
114
114
115
BF-2
BF-5
BF-2
BF-5
BF-8
BF-4
BF-7
BF-9m
BF-2
BF-4U>
BF-2
0.399
0.494
0.675
0.545
0.500
0.784
0.373
0.760
2.12
2.22
0.237
1.1
i;2
1.6
1.4
1.2
1.1
1.4
1.4
1.3
1.4
1.4
1
1
1
1
1
1
1
1
1
1
1
.37
.5
.9
.7
.4
.4
.7
.7
.7
.8
.75
2.0
2.1
2.6
2.4
2.0
1.9
2.2
2.4
2.7
2.7
2.4
3.0
2.9
3.3
3.0
2.8
2.7
3.2
3.0
4.4
4.0
3.0
4.5
4.2
4.3
4.1
3.9
3.7
4.8
4.2
6.0
5.8
4.2
6.3
5.8
5.8
5.6
5.8
5.5
6.0
6.0
8.0
7.4
5.6
7.6
6.8
6.9
6.6
6.9
7.0
7.1
7.6
9.1
8.2
6.5
(1)  Secondary Cyclone Disabled
(2)  BF Indicates Balston filter samples taken upstream of the tertiary cyclone

-------
                  APPENDIX  M-6.   PARTICLE  SIZE  DISTRIBUTION AND GRAIN LOADING
                         FLUE  GAS PARTICULATES  AFTER  TERTIARY CYCLONE


                                                          Particle Size  (ym)
               (3)Grain  Loading      5%       10%        25%      50%       75%       90%       95%
  Run/Sample No.      gr/SCF    Less Than  Less  Than Less  Than Less Than  Less Than Less Than Less Than
61 (1
62.1
62.3
64


65

67.1
rsj
CO
0 67,2

67.3
68.2

69.1

70.1
70.2
70.2
71
72.1
73
1}
(1)
(1)
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.
No.

No.
No.
No.
No.



1
2
3
1
2
• 1
2
1
2
2
1
2
1
2

1
1
1
1



(1)
(1)
(1)
(3)
(3)







(2)
(2)
(2)





0.456
0.026
0.206
0.280
0.290
0.271
0.048
0.061
0.057
0.053
0.081
0.058
0.066
0.033
--
0.416
0.350
__
--
--
0.054
0.040
0.023
2.3
2.2
1.2
1.0
1.1
1.05
1.2
1.0
• V
•1 •
-_
_-
--
__
--
1.7
1.4
1.1
1.2
1.0
—
—
~
3.2
2.6
1.4
1.2
1.3
1.3
1.4
1.2
^^
^—
__
__
—
__
—
2.2
1.8
1.4
1.7
1.1
—
--
--
7.5
5.3
2.1
1.8
1.9
1.8
2.2
1.9
— —
1.0
1.1
1 .0
—
1.1
1.6
3.6
2.9
2.2
2.8
1.6
1.3
1.1
—
16
16
3.1
3.1
2.7
2.6
3.3
2.8
1.3
2.5
1.7
1.8
1.4
1.7
2.1
6.5
5.0
3.8
5.6
2.7
2,1
1,7
1.2
40
30
7.0
5.5
3.9
3.7
6.3
4.6
3.5
9.4
2.9
8.0
4.7
2.9
4.8
9.5
8.5
7.8
9.6
5.3
4.5
2.6
3.2
45
17
10.5
6.3
6.0
16
10.0
12.5
16.7
10.5
14.0
10.7
11.0
20
13.0
13.0
11.0
12.6
8.6
12.5
4.5
13.0
—
23
17
11
9.7
24
16
17
20
17
20
15
20
40
16
18
13
14
11
19
7
18
(1)  Runs prior to Installation of 3rd cyclone.        (2)  Third cyclone bypassed.
(3)  Runs prior to run 80, the sample no. refers to the order in which Balston filter samples
     were taken with  the Old Balston  filter sampling system.

-------
                             APPENDIX  M-6  (CONT'D).   PARTICLE  SIZE  DISTRIBUTION AND GRAIN
                                LOADING -  FLUE GAS  PARTICULATES  AFTER  TERTIARY CYCLONE
                                                                   Particle  Size  (ym)
                         Grain Loading     5%       10%       25%        50%        75%        90%       95%
r\>
oo
Run/Sampl
74.1
74.2
75 No.
No.
78 No.
No.
No.
No.
No.
No.
No.
No.
No. 1
No. 1
e No/1'


1
2
2
3
4
5
6
7
8
9
0
1
Avg of Test 2-11
79 No.
No.
No.
on DO i V ^/
oU DD~ 1
80 BA-2
80 BA-3
80 BB-4
80 BA-5
80 BB-6
80 BA-7
1
2
3







gr/SCF
0.024
0.024
0.024
0.031
0.033
0.034
0.028
0.039
--
C.042
0.056
0.045
0.036
0.054
0.041
0.012
0.0086
0.034
0.0143
0.0227
0.0257
0.021
0.023
0.021
0.028
Less Than
__
1.1
_..
—
0.48
0.6
0.62
0.46
0.48
0.48
0.79
0.74
0.6
0.79
0.58
0.54
0.47
0.47
0.52
0.55
0.54
0.53
0.56
0.46
0.58
Less Than
__
1.35
_ _
—
0.56
0.75
0.78
0.54
0.55
0.58
1.00
0.97
0.76
1.10
0.66
0.68
0.53
0.56
0.60
0.61
0.61
0.59
0.64
0.53
0.71
Less Than
1.15
2.0
1.10
1.15
0.74
1.10
1.10
0.74
0.79
0.74
1.49
1.47
1.30
1.75
0.98
1.00
0.69
0.75
0.69
0.76
0.78
0.70
0.91
0.70
1.05
Less Than
1.70
3.8
1.95
1.70
0.96
1.95
3.1
1.10
2.1
1.00
2.5
2.5
3.0
2.5
1.90
1.75
0.98
1.10
0.92
1.06
1.11
0.93
1.30
0.98
1.50
Less Than
2.5
8.0
5.5
9.3
1.4
4.8
6.8
2.8
5.9
1.5
5.5
5.7
5.7
3.7
4.6
3.7
2.4
1.8
1.2
1.9
2.0
1.3
2.0
1.41
2.4
Less Than
8.0
15.0
12.5
14.5
5.0
8.5
10.8
6.1
12.0
3.1
9.2
9.6
8.8
5.7
8.5
6.3
8.8
2.2
1.8
6.0
4.9
1.8
3.4
2.2
4.5
Less Than
13.5
20.0
16.0
18.5
8.7
9.8
12.5
8.4
14.0
5.04
11.5
11.5
11.0
8.0
11.0
8.8
13.5
8.0
2.2
9.0
9.0
2.5
4.9
5.0
6.3
      (1)  ~Runs~pr1or to  run 80  the  sample  number  refers  to  the order  1n which  the Balston filters
          were  taken with  the old Balston  filter  sampling system.
      (2)  Runs  after run 80 BA  refers to Balston  filter  samples taken with  the HTHP sampling system;
          BB refers to samples  taken with  the old Balston filter sampling system.

-------
                       APPENDIX  M-6  (CONT'D).   PARTICLE  SIZE  DISTRIBUTION  AND  GRAIN

                          LOADING -  FLUE GAS  PARTICULATES AFTER  TERTIARY CYCLONE
                                                                 Particle  Size  (ym)

                         Grain Loading     5%       10%       25%       50%       75%       90%       95%
ro
oo
ro
Run/Sample No.
80 BB-8
80 BA-9
80 BB-10
81 BA-4
81 BA-10
81 BB-11
81 BA-13
81 BA-16
81 BB-18
81 BB-20
83 BB-1
84 BB-2
86 BA-1
87 BA-1

89 BB-6

91 BB-3
96 BB-3
99 BB-1
99 BB-2
99 BB-3
99 BB-5
L; gr/SCF
0.0145
0.024
0.023
0.063
0.031
0.037
0.037
0.039
0.021
0.018
0.027
0.050
0.028
0.028
(2)
0.050
(2)
0.0459
0.049
0.036
__
0.094
0.055
Less Than
0.47
0.96
0.47
0.56
0.56
0.63
0.56
0.58
0.52
0.58
1.26
1.40
0.88
0.94
1.15
0.60
1.10
1 .20
1.00
1.00
0.74
0.63
0.96
Less Than
0.54
1.18
0.56
0.60
0.60
0.80
0.62
0.70
0.59
0.63
1.70
1.75
1.10
1.15
1.30
0.68
1.20
1.38
1.26
1.25
0.98
0.83
1.20
Less Than
0.73
1.45
0.76
0.79
0.76
1.20
0.77
1.00
0.72
0.80
2.20
2.20
1.50
1.50
1.70
0.92
1.46
1.78
1.65
1.63
1.40
1.20
1.60
Less Than
1.10
1.95
1.13
1.2
1.1
1.7
1.2
1.4
1.6
1.2
2.8
2.8
2.0
2.2
3.2
1.2
2.7
2.3
2.2
2.1
1.9
1.6
2.1
Less Than
1.6
2.8
1.6
2.8
1.7
3.4
3.2
3.1
5.8
2.9
3.5
3.5
2.9
4.6
8.6
2.0
6.8
3.1
3.7
2.5
2.5
2.4
2.8
Less Than
2.5
4.7
2.4
7.5
6.3
6.0
9.0
9.4
11.5
5.8
4.3
4.0
6.3
10.0
19.0
7.1
12.7
7.4
8.0
3.0
3.5
3.9
4.0
Less Than
4.0
6.0
5.0
11.5
13.0
8.0
14.0
16.0
16.0
8.0
5.0
4.9
10.08
13.0
28.5
11.5
20.2
12.7
11.0
3.5
5.0
5.4
7.0
       (1)   Runs  prior  to  run  80 the sample number refers to the order in which the Balston filters were

            taken with  the old Balston  filter sampling system.
       (2)   100 jjm aperture used  on  Coulter Counter:   all  others used  30 \im aperture.

-------
                       APPENDIX M-6  (CONT'D).   PARTICLE  SIZE  DISTRIBUTION  AND  GRAIN

                          LOADING -  FLUE GAS PARTICULATES  AFTER  TERTIARY  CYCLONE
r\j
oo
/, x Grain Loading 5%
Run/Sample No/' gr/SCF Less Than
100 BB-2
100 BB-3
100 BB-5
100 BB-7
102 BB-5
103 BB-4
103 BB-7
103 BB-11
103 BB-15
104 BB-1
104 BB-2
104 BB-3
104 BB-4
104 BB-5
105 BB-1
105 BB-2
105 BB-3
105 BB-5
106 BA-1
106 BA-3
106 BA-5
106 BA-6
106 BA-7
106 BA-8
106 BA-9
106 BA-1 3
0.059
0.060
0.048
0.049
0.025
0.044
0.075
0.029
0.099
0.057
0.099
0.117
0.084
0.140
0.077
0.074
0.072
0.106
0.083
0.029
0.049
0.037
0.460
0.048
0.094
0.80
1.10
0.63
1.15
0.84
0.63
0.75
1.00
0.60
1.00
1.20
0.79
1.15
1.00
1.00
0.64
1.05
0.88
1.26
0.79
0.74
0.64
0.79
1.10
1.70
0.79
0.79
10%
Less Than
1.35
0.79
1.40
1.10
0.75
0.96
1.26
0.72
1.26
1.40
1.10
1.35
1.17
1.26
0.80
1.30
1.17
1.49
1.05
0.93
0.84
1.05
1.40
2.20
1.15
1.10
Particle Size (ym)
25% 50% 75%
Less Than Less Than Less Than
1.75
1.15
1.85
1.43
1.05
1.3
1.6
1.0
1.6
1.8
1.4
1.8
1.4
1.7
1.1
1.7
1.5
1.9
1.4
1.2
1.1
1.4
2.2
?.1
1.6
1.7
2.2
1.5
2.4
1.8
1.4
1.7
2.1
1.6
2.1
2.2
1.9
2.4
1.9
2.2
1.5
2.2
2.0
2.4
2.1
1.7
1.7
2.0
3.5
4.9
3.1
2.8
2.7
2.1
3.1
2.3
2.1
2.4
2.6
3.2
2.65
2.9
2.5
2.9
2.4
2.8
2.0
2.6
2.5
3.0
3.3
2.4
2.8
2.8
5.6
6.9
4.5
4.3
90% 95%
Less Than Less Than
3.1
3.0
3.9
2.7
5.0
3.5
3.0
5.6
3.1
3.5
3.0
3.5
2.8
3.1
2.5
3.1
3.0
3.5
5.6
4.6
5.0
4.0
7.8
9.2
6.3
6.3
5.0
6.35
4.6
3.0
9.2
5.5
3.5
8.6
3.8
3.9
3.3
3.9
3.1
3.7
3.1
3.6
3.2
3.8
7.4
7.0
8.0
5.0
9.8
10.0
8.0
9.2
      (1)  Runs prior to run 80 the sample number refers to the order in which the Balston filters were
           taken with the old Balston filter sampling systems.

-------
                     APPENDIX M-6  (CONT'D).  PARTICLE SIZE DISTRIBUTION AND GRAIN
                        LOADING -  FLUE GAS PARTICULATES AFTER TERTIARY CYCLONE
ro
oo
Run/Sample No.

106 BA-17
106 BA-20
106 BA-24
106 BA-28
106 BA-32
106 BA-34
Avg. (2)
Avg. (3)

107 BA-4
107 BA-6
107 BA-10
1C8 BA-1
108 BA-5

109 BA-1
109 BA-3
109 BA-5

110 BA-1
110 BA-4
110 BA-7
                    (1)
                  Grain Loading
                                          5%
                       10%
   Particle Size (ym)
25%       50%       75%
90%
95%
     111
     111
    BA-1
    BA-4
gr/SCF    Less Than Less Than Less Than Less Than Less  Than  Less  Than  Less Than

0.055
0.063
0.054
0.047
0.043
0.043
0.056
0.050

0.099
0.117
0.121

0.118
0.129

0.118
0.112
0.170

0.105
0.061
0.070

0.084
0.079
0.56
0.64
0.56
0.58
0.59
0.60
0.75
0.65
0.74
1.05
1.00
0.79
1.05
0.93
0.80
1.00
0.64
0.56
0.82
0.69
0.58
0.62
0.79
0.65
0.65
0.68
0.70
0.95
0.80
0.98
1.35
1.30
1.05
1.30
1.26
1.21
1.30
0.77
0.60
1.20
0.84
0.68
0.8
1.2
0.9
0.8
0.9
1.0
1.3
1.1
1.4
1.9
1.7
1.4
1.7
1.7
1.7
1.7
1.0
0.8
1.7
1.2
0.8
1.2
1.7
1.3
1.2
1.3
1.4
2.0
1.5
1.9
2.5
2.2
1.9
2.2
2.3
2.1
2.1
1.4
1.3
2.4
1.6
1.3
2.4
2.7
2.7
1.9
2.1
2.1
3.1
2.5
2.5
3.2
2.9
2.5
2.6
2.9
2.6
2.6
2.0
2.8
3.2
2.4
2.0
4.9
4.9
5.5
4.6
4.6
3.8
5.3
4.7
3.1
4.8
3.8
3.4
3.0
3.9
3.2
3.2
3.2
5.7
4.0
3.6
3.3
8.0
7.2
8.0
8.0
7.0
6.3
7.6
7.2
3.6
8.0
5.0
4.0
3.4
5.3
3.7
3.85
4.3
7.6
5.1
4.8
4.7
      (1)  Runs prior to run 80 the sample number refers to the order fn which the Balston filters were
          taken with the old Balston filter sampling system.
      (2)  Excluding 7
      (3)  Excluding 7, 8, 9, 13

-------
                 APPENDIX M-6 (CONT'D).   PARTICLE SIZE DISTRIBUTION AND GRAIN
                    LOADING - FLUE GAS PARTICULATES AFTER TERTIARY CYCLONE
                                                           Particle  Size  (ym)
               /,x  Grain  Loading     5%       10%       25%       50%       75%        90%       95%
 Run/Sample  No.   '      gr/SCF    Less Than  Less  Than  Less  Than  Less  Than  Less Than  Less Than Less Than





ro
00
en


112 BA-1
112 BA-4
112 BA-7
113 BA-3
113 BA-6
113 BA-8


114 BA-1
114 BA-3
0.067
0.132
0.203
0.109
0.072
0.118


0.133
0.172
0.60
0.72
0.56
0.62
0.68
0.79


0.68
0.78
0.68
0.90
0.64
0.73
0.78
0.98


0.82
0.98
0.9
1.3
0.8
1.0
1.1
1.3


1.2
1.4
1,4
1.7
1.3
1.4
1.4
1.7


1.7
1.9
2.4
2.3
2.4
2.6
2.0
2.2


2.6
2.9
4.9
3.4
4.2
5.2
3.3
3.9


4.6
4.9
6.4
4.3
5.4
6.6
4.8
5.2


6.5
6.4
115 BB-4              0.075        0.74      0.92      1.3       1.8       2.6       4.3       5.6
(1)   Runs prior to run 80 the sample number refers to the order In which the Balston filters  were
     taken with the old Balston filter sampling system.

-------
                                APPENDIX  N.  MINIPLANT SOLIDS ANALYSIS
         Run No,

          61
           Source
CO
at
          62
 Initial Bed
 Final  Bed
 Second Cyclone No,
                No,
 GBP Filter 2
     Filter 3
     Filter 2
     Filter 3
                                       1
                                       2
                                 Dump No. 3
                                 Dump No. 3
                                 Dump No. 6
                                 Dump No. 6
                    GBF Flange Fines
                    Participates - GBF Outlet
                    Bed Probe Sample No. 1
                                     No. 2
                                     No. 3
                    Dip!eg - 1st Cyclone
Initial Bed
Final Bed
Second Cyclone No,
               No,
               No,
               No.
GBF Filter 2
    Filter
    Filter 2
    Filter 3
    Filter 2
                                       1
                                       2
                                       6
                                       7
                                Dump No. 1
                                Dump No. 1
                                Dump No. 2
                                Dump No. 2
                                Dump No. 7
                       Filter 3 Dump No. 7
                   GBF Flange Fines
                   Particulates - GBF Outlets
                   Bed Probe Sample No. 2
                                    No. 5
                                    No. 6
                   01 pi eg  - 1st Cyclone
 32.0
 38.0
 16,4
 13.7
  8.42
  8.26
 10.4
 24.4
 12.8
 14.5
  7.70
  6.82
 12.1

40.2
37.3
37.6
12.7
Weight Percent
S
7.21
10.50
4.24
3.34
4.44
3.83
4.65
4.51
4.26
2.41
7.04
7.23
8.68
5.45
10.0
10.51
3.24
3.06
2.83
3.09
5.67
5.20
4.70
4.20
4.51
5.33
5.60
5.98
9,10
11.0
11,0
5.67
S04
22.52
33.28
13.48
10.84
13.20
10.39
13.56
12.47
10.42
15.85
21.83
24.93
28.70
20.38
33.37
32.15
9.03
9,20
8.08
7.17
15.57
14.24
12.49
12.38
12.57
11.54
17.02
18.20
27.96
34.66
32.39
16.62
C03
24.06
0.62
11.27
4.05
3.73
3.94
1.42
0.84
4.14
1.17
27.60
0.54
0.71
2.52
19.91
0.53
4.78
2.17
0.32
0.28
2.89
4.20
4.33
3.93
0.68
0.50
1.53
0.01
1.71
0.65
1.78
1.44
Total C
5.01
0.28
5.62
3.05
1.59
2.14
1.44
1.70
1.63
1.19
5.96
0.35
0.37
1.51
5.46
0.42
3.50
1.87
2,21
3.16
0.98
1.49
1.72
1.86
1.52
0.85
1.75
2.78
0.37
0.58
0.32
0.61
                                                  1
                                                  4
                                                  7,
                                                  5,
                                                  5,
                                                  5,
                                                  5,
                                                  4.
                                                  4,
   ,83
   ,92
    19
    91
    44
    53
    39
    73
    28
                                                                              3.27
                                                                              5.
                                                                              2,
                                                   .27
                                                   .36
                                                  4.10
                                                  7.37
 0,43
 0.88
 4.33
 4.18
 2.26
 1.93
 4.62
11.6
 5.06
 5.05
 1.62
 1.32
 4.54

 1.56
 1.08
 1.02
 4,91

-------
       Run No.

        63
00
        64
         APPENDIX N (CONT'D),

	Source	

Initial  Bed
Final Bed
Second Cyclone No. 3
               No. 6
               No. 9
GBF Filter 2 Dump No.  3
    Filter 3 Dump No.  3
    Filter 2 Dump No.  7
    Filter 3 Dump No.  7
    Filter 2 Dump No.  10
    Filter 3 Dump No.  10
GBF Flange Fines Filter 2
                  Particulates
             No.
             No.
             No.
                 Filter 3
                 1
                 2
                 3
Bed Probe Sample No. 2
                 No. 3
                 No. 4
                 No. 5
                 No. 6
Dip!eg - 1st Cyclone

Initial Bed
Final Bed
Second Cyclone No. 2
               No. 3
               No. 6
             Dump No,
                                        3
                               Dump No. 3
                               Dump No. 6
                               Dump No. 6
                      Filter 2 Dump No. 8
                      Filter 3 Dump No. 8
GBF Filter 2
    Filter 3
    Filter 2
    Filter 3
                               MINIPLANT SOLIDS ANALYSIS
                                                       Weight  Percent
                              Ca

                              37.7
                              36.4
                                                 7,
                                                 7,
                                                 4,
                                                 3,
                                                 2,
                                                 2,
                                                 2
                                                 2
                                                 1
                                 62
                                 15
                                 93
                                 34
                                 81
                                 15
                                 ,01
                                 ,32
                                 ,62
 4.47
 4.55
 4.74
 4.14
 5.02
78.0
35.9
39.8
36.1
35.6
 9.42

32.6
30.0
 4.93
 6.13
 3.46
 0.72
 2.16
 0.89
 0.92
 1.96
 1.80
s
8.55
10.32
2.31
2.64

1.49
1.55
2.14
1.15
1.21
1.08
4.93
3.97
-
8.00
9.83
10.10
9.39
10.41
3.13
9.36
11.6
1.55
0.80
1.00
0.34
1.52
0.59
0.45
0.48
1.17
S04
27.36
33.96
6.13
4.25
4.18
6.47
4.14
3.98
2.82
3.30
2.87
11.12
7.66
-
29.07
30.97
32.54
33.63
34.23
8.02
28.90
34.16
5.27
4.77
3.24
0.61
4.28
1.00
5.03
4.71
1.39
C03
11.88
0.50
0.51
0.63
0.16
0.36
0.30
0.06
0.05
0.03
0.24
0.29
0.53
-
6.37
0.43
0.42
0.46
0.79
1.10
3.35
0.78
0.18
0.25
0.11
0.05
0.12
0.27
0.02
0.22
0.81
Total C
2.52
0.23
4.51
4.38
5.00
0.81
1.15
0.80
0.49
0.24
0.59
1.76
1.22
0.12
1.34
0.12
0.30
0.50
0.33
0.57
2.14
0.35
7.30
15.51
4.92
<0.05
1.14
0.18
<0.05
0.17
0.42
0.73
1.56

0.60
0.54
0.14
0.23
0.17
0.06
0.34
0.25
0.16
0.34
0.47

-------
                              APPENDIX N (CONT'D).  MINIPLANT SOLIDS ANALYSIS
                                                                            Weight Percent
ro
oo
CO
Run No.
64
(Cont'd,







65










67


67.1







Source
GBF Flange Fines No. 2
No. 3
Participates No. 1
No. 2
No. 3
Bed Probe Sample No. 4
No. 7
No. 9
Dipleg - 1st Cyclone
Initial Bed
Final Bed
Second Cyclone No. 2
No. 6
GBF Filter 2 Dump No. 3
Filter 3 Dump No. 3
Filter 2 Dump No. 7
Filter 3 Dump No. 7
Bed Probe Sample No. 3
No. 7
Dipleg . 1st Cyclone
Initial Bed
Final Bed
Dipleg - 1st Cyclone
Second Cyclone Nos. 5-9
Nos. 10-14
Bed Overflow Nos. 5-9
Nos. 10-14
Third Cyclone Nos. 5-9
Nos. 10-14
Bed Probe Sample
Bed Probe Sample
Ca
4.56
5.52
5.60
4.92
4.50
43.4
41.1
32.8
5.53
37.8
37.4
5.10
2.80
1.52
8.25
1.06
1.78
39.1
30.5
6.40
22.3
18.9
15.1
7.3
7.1
20.8
21.6
6.7
6.8
19.6
22.1
S
2.17
1.85
3.82
4.52
5.81
8.85
8.78
10.09
1.59
10.6
12.72
2.62
1.31
0.74
0.44
0.78
1.55
11.37
12.78
2.99
6.87
8.17
7.07
5,39
5.03
7.16
6.62
6.05
6.90
10.53
7.01
S04
3.23
8.19
9.28
15.32
24.88
28.56
31.24
33.56
4.83
28.32
35.95
6.26
3.27
1.64
1.15
1.78
3.01
33.42
35.51
6.38
14.54
24.82
17.95
14.01
20.84
21.08
14.13
19.85
20.16
15.72
24.71
CDs
0.03
0.02
_
—
—
0.45
0.27
0.52
0.08
7.95
0.11
0.52
0.10
0.22
0.10
0.01
0.07
0.53
0.62
0.01
.
12.36
9.38
0.45
0.51
17.19
15.77
0.03
0.01
17.80
20.27
Total C
0.46
1 .09
-
1.61
1.63
0.44
0.34
O.05
0.23
2.21
0.29
6.21
5.05
0.12
0.21
0.17
0.62
0.25
0.12
0.29
4.43
2.15
1.94
4.69
4.20
2.63
2.67
1.26
1.23
3.18
4.50
Mg
0.40
0.28
0.23
0.23
0.21
0.56
0.65
0.68
0.36
0.75
0.59
0.24
0.16
0.19
0.35
0.05
0.12
0.62
0.70
0.27
11.7
10.7
8.0
3.1
3.4
11.2
12.0
3.9
4.1
10.9
12.2

-------
                            APPENDIX N (CONT'D).  MINIPLANT SOLIDS ANALYSIS
         Run No.

           67.1
         (Cont'd.)
           67.2
           Source
ro
oo
           67.3
           67.4
Turbine Section No. 1
(Pressure Reduc. Section)
Turbine Section No. 4
(Turbine Housing After
 Distribution Plate)

Second Cyclone Nos. 18-22
               Nos. 23-27
Bed Overflow Nos. 18-22
             Nos. 23-27
Third Cyclone Nos,
              Nos
Bed Probe Sample
Bed Probe Sample
                                       18-22
                                       23-26
Second Cyclone Nos. 28-32
Bed Overflow Nos. 28-32
Third Cyclone Nos. 28-32
Bed Probe Sample

Second Cyclone Nos. 42-46
               Nos. 47-51
Bed Overflow Nos. 42-46
             Nos. 47-51
Third Cyclone Nos. 41-45
              Nos. 46-50
Bed Probe Sample
Bed Probe Sample
Bed Probe Sample
Bed Probe Sample
 8.4
 8.4
20.6
20.1
 6.9
 6.8
16.3
20.9

 7.3
23.1
 6.1
23.3

 4.3
 6.9
22.1
22.1
 6.7
 6.9
23.5
20.4
22.1
23.1
s
4.04
9.18
5.26
5.44
10.19
10.19
8.02
6.19
8.07
8.73
4.99
8.16
8.02
8.74
5.17
5.78
7.45
6.63
9.30
8.78
6.75
11.84
7.99
6.38
S04
11.46
29.55
16.37
17.68
34.36
34.61
16.20
18.64
29.55
23.20
15.34
28.83
21.53
29.96
15.66
16.19
19.20
23.81
28.78
28.29
24.80
21.58
22.59
23.22
C03
0.58
0.39
0.84
0.49
12.80
10.00
0.13
0.08
15.74
20.62
0.51
17.86
0.21
16.44
0.49
0.48
24.12
21.95
0.09
0.04
21.53
22.43
21.38
18.87
Total C
0.33
0.18
3.43
3.66
1.78
2.47
2.26
1.16
4.39
3.73
4.01
4.44
0.40
3.50
4.78
4.53
4.53
5.07
1.06
1.15
4.91
2.46
4.31
5.26
 4.3
 4.2
12.1
10.9
 3.5
 3.4
11.9
12.1

 3.6
13.7
 3.6
13.5

 2.1
 3.4
12.8
12.4
 5.0
 4.9
13.4
11
12,
                                                                                                    ,9
                                                                                                    .7
                                                                                                 13.4

-------
                            APPENDIX N  (CONT'D).  MINIPLANT SOLIDS ANALYSIS
                                                                          Weight  Percent
        Run  No.

          68
          68.1
          68.2
ro
         69
           Source
 Initial Bed
 Final Bed

 Second Cyclone No.  6
                No.  7
 Third Cyclone No.  6
               No.  7
 Bed Probe Sample No. 6
                  No. 7
                  No. 8

 Second  Cyclone No.  8.1
                No.  8.2
                No.  9
                No.  10
 Third Cyclone  No. 8.1
               No. 8.2
               No. 9
               No. 10
 Bed Probe Sample No.   9
                 No. 10
                 No. 11
                 No. 12

Initial  Bed
Final  Bed
  Ca
 ^•^•^•••^H

 22.6
 21.6
 15.8
 14.9
 11
 11
   .0
   .1
23.8
24.2
23.7

17.4
18.7
18.5
17.6
10.8
                                                11
                                                12
                                                13.1
                                                25
                                                22
                                                21,
23.7

27.2
24.1
S04
30.50
37.12
23.19
21.70
24.01
25.90
38.80
33.26
32.98
25.68
24.96
24.92
25.54
24.42
24.95
25.00
21.20
35.36
33.84
36.05
32.73
29.53
38.16
C03
17.87
0.60
4.49
5.32
0.72
1.10
0.93
0.71
1.40
4.93
5.21
5.04
5.13
1.28
1.57
0.97
1.32
0.88
0.69
0.58
0.72
6.42
0.71
Total C
3.44
0.18
2.29
2.78
4.14
4.63
0.12
0.31
1.04
2.01
2.40
1.90
2.53
4.62
5.78
3.40
3.16
0.21
0.32
0.62
0.44
1.34
0.13
Mg
13.8
14.0
10.0
9.3
5.9
6.0
14.6
15.2
15.6
11.4
11.7
12.1
11.1
5.79
5.85
5.68
7.3
16.6
15,2
15.1
16.1
12.9
11.9

-------
                   APPENDIX N (CONT'D).  MINIPLANT SOLIDS ANALYSIS
Run No,
 69.1
 69.2
 70

 70.1
Source
Ca
                                                                 Weight  Percent
                                                                     Me
Second Cyclone No. 5
No. 6
No. 7
Bed Overflow No. 3
Bed Probe Sample No. 1
No. 5
No. 6
No. 7
No. 8
Balston Filter No. 1
Second Cyclone No. 9
No. 10.1
No. 10.2
No. 11
Bed Overflow No. 5
No. 6.1
No. 6.2
No. 7
Bed Probe Sample No. 10
No. 11
No. 12
Balston Filter No. 2
Initial Bed
Final Bed
Second Cyclone No. 6
No. 7
Bed Overflow No. 1
No. 2
Balston Filter No. 1
Bed Probe Sample No. 1
No. 8
No. 9
No. 10
15.6
16.5
15.0
18.0
22.9
23.5
22.4
15.1
19.3
14.1
13.4
14.0
13.7
10.5
23.2
23.0
22.1
20.0
18.7
21 .0
23.4
1 .99
24.7
24.6
16.2
17.7
21.9
26.5
13.8
25.5
32.8
24.4
26.2
27.02
26.40
28.78
35.76
26.29
40.38
40.44
38.36
31.77
44.74
27.61
9.43 27.94
9.70 26.79
24.92
34.29
12.19 36.88
10.94 33.05
32.42
40.20
12.92 40.28
39.02
24.48
31.12
33.68
27.72
29.78
36.19
29.04
30.56
37.45
35.25
37.93
31.39
	 
-------
                            APPENDIX N tCONT'D).  MINIPLANT SOLIDS ANALYSIS
          Run No,
ro
<&
ro
          71
           Source
          72
 Initial  Bed
 Final  Bed
 Second Cyclone  No.
 Bed Overflow  No.
 Third Cyclone No.
 Bed Probe Sample No.
Initial Bed
Final  Bed
Ca
70.2 Second Cyclone No. 8
No. 9
No. 10.1
No. 10.2
No. 11
Bed Overflow No. 3
No. 4
No. 5.1
No. 5.2
No. 6
Balston Filter No. 2
Bed Probe Sample No. 11
No. 12
No. 13
No. 14
No. 15
^•^^•••M
15.3
11.7
15.1
16.2
16.8
16.1
19.9
18.0
21.7
24.6
11.9
24.4
22.7
24.0
23.9
20.5


). 4
9
, 4
No.
No.
No.
No.







1
5
6
18


25.3
22.4
14.9
22.0
10.6
24.5
25.1
23.0
22.3
21.1
18.9
Weight Percent
S04
28.28
28.10
28.22
28.63
27.57
32.06
31.71
35.75
35.84
33.43
27.52
33.44
34.69
29.39
35.30
39.57
27.70
31.95
27.60
32.94
23.58
33.74
31.85
33.44
34.51
39.55
33.96
C03
2.99
2.59
2.40
2.63
2.70
6.11
5.74
4.82
6.12
4.86
_
1.05
1.12
1.30
1.41
0.56
2.21
3.22
0.91
9.35
1.32
5.93
3.86
2.22
3.16
3.56
0.46
Total C
1.13
1.14
1 .00
0.98
1.04
1.40
1.25
0.98
1.28
0.97
0.42
0.31
0.27
0.30
0.33
0.13
0.53
0.69
0.96
2.00
1.09
1.36
0.88
0.57
0.81
0.95
0.15
Mg
9.92
7.85
10.5
9.59
9.40
9.09
10.8
12.3
12.7
13.5
6.43
13.7
13.4
13.3
13.2
12.2
14.5
13.0
7.74
12.1
4.71
13.0
13.6
12.1
12.9
13.8
12.8

-------
                            APPENDIX N (CONT'D).   MINIPLANT  SOLIDS ANALYSIS
        Run No.

         72.1
ro
us
co
         72.2
         73


         73.1
	Source	

 Second Cyclone No. 8.1
               No. 8.2
               No. 9
 Bed Overflow No. 3
 Third Cyclone No. 8.1
              No. 8.2
              No
 Bed Probe Sample
. 8
. 8
, 9
No.
No.
No.
No.
                                        1
                                        8
                                        9
                                        10
Second Cyclone No. 12
Third Cyclone No. 12
Bed Probe Sample No. 12
                 No. 13
                 No. 14

Initial Bed
Final Bed

Second Cyclone No. 6
               No. 7
Bed Overflow No. 3
Third Cyclone No. 6
              No. 7
Bed Probe Sample No. 1
                 No. 5
                 No. 6
                 No. 7
                 No. 8
 Ca
^^•M^MM^
15.8
16.2
19.1
20
11
11
11
                                                                         Weight  Percent
25.7
21.0
20.4
19.2

17.4
14.5
22.0
18.8
23.4

27.3
21.8

16.7
16.5
21.9
16.2
16.3
28.1
25.7
23.7
26.4
24.7
S04
27.07
27.60
27.17
27.83
22.64
23.30
20.93
26.90
34.34
37.38
29,65
23.23
15.98
36.65
38.65
32.94
30.29
34.75
26.55
26.33
35.41
24.97
24.78
34.70
24.81
30.73
30.23
35.20
COs
1.34
1.33
0.83
11.49
0.15
0.27
0.99
6.43
0.85
0.88
0.74
1.33
1.71
1.08
0.72
0.53
5.52
1.03
3.33
3.04
3.98
1.60
1.42
3.41
2.18
2.27
1.93
1.01
Total C
1.84
1.58
1.45
0.99
1.18
1.14
1.07
1.37
0.33
0.34
0.40
8.56
4.91
0.93
0.23
0.17
1.21
0.24
2.32
2.28
0.87
1.13
1.09
0.76
0.59
0.61
0.55
0.33
                                                            10.3
                                                             5.93
                                                            13.6
                                                            13.0
                                                            14.1

                                                            16.2
                                                            12.9

                                                             9.75
                                                             8.80
                                                            11.6
                                                                                                7,
                                                                                                7,
                                                               44
                                                               02
                                                                                               15.7
                                                                                               14.3
                                                                                               13.4
                                                                                               15.3
                                                                                               14.1

-------
                         APPENDIX N (CONT'D).  MINI PLANT SOLIDS ANALYSIS
      Run No.

         73.2
           Source
                                                                       Weight Percent
         74


         74.1
ro
vo
        74.2
       75
 Second Cyclone No. 11
 Third Cyclone No. 11
 Bed Probe Sample No. 11
                  No. 12
                  No. 13

 Initial Bed
 Final  Bed

 Second Cyclone No. 3
                No. 4
 Third  Cyclone No.  3
               No.  4
 Bed  Probe  Sample No.  1
                 No.  4
                 No.  5

 Second Cyclone No.  9
               No. 10
 Third Cyclone No.  9
              No.  10
 Bed Probe Sample No. 10
                 No. 11
                 No. 12

 Initial  Bed
Second  Cyclone No.  5
               No.  6
 24.5

 38.6
 47.7

 24.0
 27.2
 19.3
 13.7
 44.8
 47.5
 47.8

 22.8
 20.3
 19.4
 15.6
 52.9
 52.8
 51.9

52.9
13.73
14.2
   S04

 16.88
 19.86
 33.63
 33.04
 31.87

 19.62
 19.7

 14.7
                                                                14
                                                                17,
 16.6
 15.7
 18.0
 15.8

 13.4
 13.6
 19.8
 18.9
 12.8
 14.6
 17.85

17.43
13.11
14.12
COs Total C
2.48
1.63
0.86
1.04
0.48
28.60
1.54
2.62
2.46
3.54
4.56
5.99
2.59
2.08
1.93
1.51
3.93
3.96
2.32
1.84
2.11
1.61
0.67
0.78
2.44
1.06
0.47
0.46
0.20
6.34
0.34
5.44
5.21
3.99
3.04
1.37
0.65
0.47
5.85
6.03
2.84
3.06
0.43
0.53
0.43
0.36
6.87
5.48
10.2
 6.50
13.9
14.5
14.3

-------
                            APPENDIX  N  (CONT'D).  MINIPLANT SOLIDS ANALYSIS
        Run No.
          75
        (Cont'd.)
           Source
          76
ro
vo
01
          77
          78
          80
Bed Overflow No. 5
Third Cyclone No. 5
              No. 6
Bed Probe Sample No. 1
                 No. 6
                 No. 7
                 No. 8
Final Bed

Initial Bed
Final Bed
Second Cyclone No. 6
               No. 7
Third Cyclone No. 6
              No. 7
Bed Probe Sample No. 1
                 No. 7
                 No. 8
                 No. 13

Second Cyclone No. 6
Bed Overflow No. 4
Third Cyclone No. 6

Second Cyclone No. 72
Bed Overflow No. 72
Third Cyclone No. 72
Bed Probe Sample No. 8
                 No. 9

Second Cyclone No. 100
Bed Overflow No. 100
Third Cyclone No. 100
                                                 50.8
                                                 50.7
                                                 45.6
.3
.7
25,
29,
15.0
16.4
12.6
12.0
27.0
27.7
29.2
24.0

13.0
24.8
 9.16

12.9
23.5
 8.41
24.8
22.13

12.30
13.67
24.69
        6.48
      11.07
        5.06
Weight Perceat
$04
24.76
19.19
18.69
16.98
19.23
20.99
19.04
24.72
39.63
40.83
16.53
16.28
14.70
14.07
38.44
40.13
39.14
38.16
20.87
36.00
17.87
9.14
39.10
15.16
39.41
29.53
20.19
31.93
15.12
C03
1.41
-
8.72
2.21
1.89
1.70
0.61
1.36
0.61
0.33
0.30
0.32
1.57
1.34
3.61
0.99
0.93
0.40
0.52
0.49
0.90
0.27
0.68
0.82
1.27
2.65
0.90
3.03
0.57
Total C
0.37
3.05
3.26
0.53
0.43
0.44
0.45
0.29
0.22
0.11
4.38
4.12
2.01
2.39
0.69
0.22
0.25
0.13
3.54
0.15
1.86
0.56
0.27
0.50
0.43
0.62
0.62
0.65
0.39
13,
11,
 5(
 5(
 2
 2
6
9
33
84
02
36
13.1
10.8
11.6
11.7

 6.83
12.3
 2.83

 7.43
12.9
 3.82
13.5
14.35

 7.48
 7.67
14.44

-------
                            APPENDIX N (CONT'D).  MINIPLANT SOLIDS ANALYSIS
        Run No.

          81
           Source
Ca
S04
                                                                         Weight Percent
ro
to
         99

         99.2
        99.3
Initial Bed
Final Bed
Second Cyclone No. 15
No. 51
No. 62
No. 74
No. 86
Third Cyclone No. 15
No. 51
No. 62
No. 74
No. 86
Bed Overflow No. 15
No. 51
No. 62
No. 74
No. 86
25.34
24.47
_
17.18
16.74
-
-
-
10.51
10,15
-
_
_
25.38
23.11
-
-
11.43
13.15
-
7.65
7.92
-
_
-
8.40
8.60
-
_
_
12.65
13.93
_
—
— '
37.24
40.87

32.10
24.16
_
^
—
26.01
26.44
_
_
—
37.78
42.50
—
«.
 Dolomite Blend  "A"           21.67

 Second Cyclone  No. 13        14.76     7.17    20.20
 Third Cyclone No. 13          6.63     4.93    18.81
 Bed Probe Sample No. 1       25.24    14.95    42.25
                 No. 2       24.73    14.33    42.43

Second Cyclone No. 14        12.94     6.30    19.54
               No. 15         9.87     5.52    16.63
Third Cyclone No. 15          5.98     5.93    14.98
Bed Probe Sample No.  3       24.40    15.04    43.76
                 No.  5       25.28    13.50    44.34
C03
1.22
0.10
4.48
4.93
5,43
5.07
3.89
0.44
0.49
0,22
0.11
0.08
0.95
0.40
0.39
0.28
0.33
64.70
4.91
0.24
0.31
0.37
3.73
1.77
0.03
0.37
0.25
Total C
0.31
0.07
2.22
1.64
2.93
1.55
1.77
1.14
0.54
1.57
0.44
0.56
0.26
0.08
0.16
0.11
0.25
-
4.29
2.35
0.09
0.13
2.39
1.57
3.11
0.09
0.12
                                                                                               15.06
                                                                                               14.23

                                                                                               10.52
                                                                                               11.01
                                                                                                6.52
                                                                                                7.25
                                                                             15.18
                                                                             14.41
                                             13.19

                                              8.59
                                              4.19
                                             12.91
                                             13.55

                                              7.96
                                              6.05
                                              3.59
                                             12.94
                                             13.76

-------
                            APPENDIX N (CONT'D).  MINIPLANT SOLIDS ANALYSIS
ro
UD
         Run  No.

          99.4
          99.5
                     Source
                               Ca
                                                                          Weight Percent
 99.7



100

100.1
         100.2
         100.4
Second Cyclone No. 2
Third Cyclone No. 21
Bed Probe Sample No. 11

Second Cyclone No. 22
               No. 23
Third Cyclone No. 23
Bed Probe Sample No. 12
                 No. 14

Second Cyclone No. 36
Third Cyclone No. 36
Bed Probe Sample No. 32

Dolomite Blend "A"

Second Cyclone No, 4
               No. 5
Third Cyclone No. 4
Bed Probe Sample No. 6
                 No. 7

Second Cyclone No. 6
Third Cyclone No. 6
Bed Probe Sample No. 8

Second Cyclone No. 12
               No. 13
               No. 14
Third Cyclone No. 12
              No. 14
Bed Probe Sample No. 13
                                    No.
                                    No.
                               15
                               17
13.46
7.06
24.47
17.21
19.49
6.91
24.49
23.53
18.40
7.90
22.32
21.85
14.16
14.47
7.83
25.16
25.03
10.36
6.36
24.19
10.00
9.31
12.91
6.43
5.03
24.51
24.87
24.74
6.02
7.21
14.09
7.96
7.11
7.78
13.35
14.95
6.81
5.58
14.16
0.05
4.81
4.62
4.65
14.58
13.60
3.59
4.04
13.73
3.45
2.92
4.79
3.37
3.16
13.89
14.12
13.76
18.94
23.14
44.46
21.84
23.07
23.28
44.5
42.38
20.46
18.15
45.03
-
15.10
14.69
15.45
46.44
44.01
11 .93
11.67
44.60
9.20
9.85
13.55
10.75
10.61
44.71
45.33
44.04
C03
1.54
0.01
0.54
2.62
3.69
0.10
0.70
0.32
4.71
0.91
0.39
64.97
2.19
2.16
1.21
0.15
0.28
1.40
0.25
0.14
1.01
1 .08
2.19
0.35
0.00
0.32
0.17
0.13
Total C
1.46
0.75
0.21
3.76
5.59
1.62
0.14
0.08
2.60
1.46
0.09
-
5.48
4.38
1.95
0.04
0.06
4.09
1.66
0.05
3.97
5.56
4.39
1.65
1.86
0.08
0.06
0.07
 6.36
 5.14
13.23

 8.11
 9.21
 5.43
12.95
13.14

 8.71
 4.06
13.02

13.07

 6.19
                                                                                                   .21
                                                                                                   .95
11.99
12.00

 4.52
 2.95
11.60

 3.78
 3.89
 5.61
 2.54
 2.41
11.96
11.71
11.96

-------
                    APPENDIX N (CONT'D).   MINIPLANT SOLIDS ANALYSIS
Run No.

 100.6
            Source
Ca
                                                                 Weight Percent
 102
103
Second Cyclone No. 29
No. 31
Third Cyclone No. 31
Bed Probe Sample No. 32
No. 33
15.61
14.25
6.59
22.99
24.33
5.52
5.33
4.41
13.49
13.32
17.00
15.31
13.16
42.43
42.09
3.26
2.42
0.12
0.06
0.10
3.43
4.09
1.26
0.03
0.02
u
7.24
6.33
3.11
10.86
11.60
 COMBUSTOR
 Second Cyclone No. 30
 Third Cylcone No. 30
 Bed Probe Sample No. 4
 Final  Bed

 REGENERATOR
 First  Cyclone No.  29
 Final  Bed

 COMBUSTOR
 Second Cyclone No. 33
               No. 40
               No. 52
               No. 60
               No. 75
               No. 87
               No. 99
Third Cyclone No. 33
              No. 40
              No. 52
              No. 60
              No. 75
              No. 87
              No. 99
15.61
14.25
6.59
22.99
24.33
29.2
18.6
55.3
47.8
40.0
49.9
23.3
15.8
12.9
18.0
19.0
14.1
6.04
16.3
10.5
8.93
13.5
13.9
12.1
7.46
5.52
5.33
4.41
13.49
13.32
2.63
2.87
2.05
3.43
3.74
1.65
4.32
4.29
3.00
3.58
4.00
2.95
1.58
3.15
3.70
3.14
2.98
2.96
2.34
2.14
17.00
15.31
13.16
42.43
42.09
7.01
8.02
6.87
11.67
9.67
4.24
12.5
11.0
9.19
10.89
11.97
8.83
4.61
9.34
11.08
9.07
9.48
8.99
7.49
6.67
C03
3.26
2.42
0.12
0.06
0.10
1.70
3.13
0.83
0.69
3.91
0.18
1.23
0.52
0.29
0.71
0.93
0.65
0.05
1.24
1.25
0.20
0.55
0.49
0.26
0.05
Total C
3.43
4.09
1.26
0.03
0.02
4.63
3.11
0.25
0.22

0.13
2.60
6.43
8.80
4.69
3.28
5.64
9.60
1.49
3.23
3.52
1.19
0.93
1.57
2.45
Me

-------
                            APPENDIX N (CONT'D).  MINIPLANT SOLIDS ANALYSIS
ro
10
Run No.    	Source	     Ca

 103       Bed Probe Sample No.  28       47.8
(Cont'd.)                    No.  40       48.3
                            No.  52       46.1
                            No.  60       45.1
                            No.  15       44.1
                            No.  87       45.2
                            No.  99       38.8
           Final  Bed                     39.4

           REGENERATOR
           Sand Filters No. 40           35.4
                        No. 50           32.8
                        No. 97           30.1
           Reg. to Comb. Pulse Pot       39.1
           Final  Bed                     40.1

 104       Second Cyclone                12.6
           Third Cyclone                  8.36
           Bed Overflow                  20.8
           Bed Probe                     21  .7
           Final  Bed                     20.7

 105       COMBUSTOR
           Second Cyclone No. 28         14.7
                          No. 44         13.4
           Third Cyclone No. 28          15.9
                         No. 44          13.2
           Bed Probe Sample              44.2
                                         43.1
           Initial Bed                   37is
           Final  Bed                     37.3
8.76
6.10
8.60
10.15
7.67
8.05
11.64
13.60
5.16
4.95
5.61
11.07
9.26
5.92
3.24
11 .50
11.93
10.82
2.70
3.56
2.08
2.21
7.60
7.16
9.42
14.77
24.21
17.88
22.72
29.21
23.48
22.90
33.51
40.5
13.51
13.73
14.29
31.87
27.58
19.43
10.69
33.34
39.64
36.09
19.56
9.84
7.54
7.50
25.31
23.80
16.84
31.66
Weight
COg
0.53
4.33
0.54
1.83
0.91
1.34
1.45
0.88
1.03
0.76
3.78
1.53
0.69
0.04
0.00
0.64
0.39
1.07
0.42
0.24
0.18
0.31
0.33
0.33
9.09
0.70
Percent
Total C
0.14
0.85
0.17
0.44
0.20
0.32
0.33
0.21
1.46
1.34
1.67
0.42
0.21
1.06
0.43
0.19
0.15
0.33
6.10
6.14
1.69
1.95
0.26
0.22
7.99
0.18
                                                                                                   7,
                                                                                                   3,
                                                                                                  12,
                                                                                                  14,
96
22
4
0
                                                                                                  12.5

-------
                             APPENDIX N (CONT'D).  MINIPLANT SOLIDS ANALYSIS
          Run  No.
Source
Ca
S04
                                                                           Weight Percent
o
o
105
(Cont'd.)




107





108


REGENERATOR
Cyclone No. 25
No. 41
Bed Sample
Initial Bed
Final Bed
Second Cyclone No. 27
No. 50
Third Cyclone No. 27
No. 50
Bed Overflow No. 24
No. 44
Second Cyclone - Final
Third Cyclone - Final
Bed Overflow - Final

32.8
35.0
37.7
35.7
44.7
8.35
9.16
5.70
6.10
20.4
22.2
11.5
7.62
21.9

2.46
1.78
3.84
7.01
3.11
4.07
3.42
2.97
3.21
11.07
12.84
.
-
-
^^•••^^^B
7.26
7.44
10.28
24.59
10.47
12.14
10.42
9.57
10.06
34.42
39.72
16.93
10.68
41.90

3.66
3.57
1.58
8.19
0.59
0.45
0.22
0.20
0
7.27
4.81
0.36
0.03
3.39

2.45
1.91
3.80
1.27
0.27
4.62
5.78
1.66
1.33
1.54
1.06
3.33
1.24
0.93
•IMi^MMrt
-
-
-
-
-
4.41
4.88
2.41
2.48
12.5
13.6
6.49
3.32
13.8

-------
                                   APPENDIX 0.   MINI PLANT SAMPLE SHIPMENTS
CO
o
     	Requestor
     Acurex-Aerotherm
     Mountain View, CA
Atlantic  Richfield Corp.
Alexandria, VA
Battelle
Columbus, OH
Colorado State University
Fort Collins, CO


Cornell University
Ithaca, NY
     EFB,  Inc.
     Woburn, MA


     ERE
      Baton Rouge, LA

      Florham Park, NJ

      Linden, NJ


    EPA
      Research Triangle Park, NC
    Sample Description
 3rd Cyclone Flyash
 2nd Cyclone Flyash
 2nd Cyclone Flyash

 3rd Cyclone Flyash


 3rd Cyclone Flyash


 Scrubber  Water
 2nd Cyclone Flyash
 3rd Cyclone Flyash
 2nd  Cyclone Flyash
 Bed  Overflow
 2nd  Cyclone  Flyash
 Bed  Overflow
 3rd  Cyclone Flyash

Bed Overflow
3rd Cyclone Flyash
Fresh Dolomite
                              3rd  Cyclone
                              3rd  Cyclone
                              2nd  Cyclone
                              2nd  Cyclone
                              2nd  Cyclone
                              3rd  Cyclone
            Flyash
            Flyash
            Flyash
            Flyash
            Flyash
            Flyash
                             2nd Cyclone Flyash
                             3rd Cyclone Flyash
                             2nd Cyclone Flyash
                             3rd Cyclone Flyash
                                                         Sorbent

                                                        Dolomite
                                                        Dolomite
                                                            Dolomite


                                                            Dolomi te
 Dolomite
 Dolomite

 Limestone
 Limestone
 Dolomite
 Dolomite
 Dolomite
 Dolomite
 Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite


Dolomite
Dolomite
Dolomite
Dolomite
Run No.
67
67
Combination
of Many Runs
67
80
81
81
81
74
74
78
78
78
78
78
--
67
78
80
93
67
67
50. IB
80
81
81
Amount
15 Gallons
2 Drums
18 Drums

2 Grams
20 Grams
1 Drum
1 Quart
1 Quart
10 Gallons
10 Gallons
10 Gallons
10 Gallons
5 Gallons
4 Drums
1/2 Drum
1-1/2 Tons
10-1/2 Ibs.
1 Drum
1 Gallon
1 Drum
1 Ib.
1 Ib.
1 Drum
20 Grams
5 Gallons
5 Gallons
Date
1/27/78
4/4/78
4/6/78

4/4/78
12/4/78
11/7/78
11/7/78
11/7/78
8/15/78
8/15/78
8/15/78
8/15/78
9/28/78
11/14/78
11/28/78
12/21/78
6/16/78
7/19/78
1/8/79
3/9/79
9/28/78
9/28/78
8/30/77
12/4/78
3/28/79
3/28/79

-------
                             APPENDIX 0  (CONT'D).  MINIPLANT SAMPLE SHIPMENTS
              Requestor
   Sample Descrlption
Sorbent
Run No.
 Amount
 Date
CO
    EPA  (Cont'd)
      Rivesville, WV
    GCA Corporation
    Bedford, MA
    General Electric
      King of Prussia, PA
      Schenectady, NY
      Valley Forge, PA

    MIT
    Cambridge, MA

    N.J. Institute of Technology
    Newark, NJ
    N.Y. University
    Westbury, NY
    Pratt & Whitney Aircraft Co.
    Middletown, CT
    Radian Corporation
    Austin, TX
2nd Cyclone Flyash
3rd Cyclone Flyash
Bed Overflow
Bed Overflow
Bed Overflow
2nd Cyclone Flyash
Final Bed (Combustor)
Final Bed (Regenerator)
3rd Cyclone Flyash
3rd Cyclone Flyash
3rd Cyclone Flyash
Illinois Coal No. 6
Pfizer Dolomite
2nd Cyclone Flyash
3rd Cyclone Flyash
Bed Overflow
Bed Overflow
2nd Cyclone Flyash
Illinois Coal No. 6
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Limestone
Limestone
Dolomite
Dol omi te
Dolomite
—
—
Dol omi te
Dolomite
Dolomite
Dolomite
Dolomite
--
108
108
108
66
67
67
105
105
68
68
81
—
—
50.4
80
78
78
78
-.
1 Quart
1 Quart
1 Quart
1 Drum
12 Drums
5 Drums
10 Gallons
10 Gallons
5 Gallons
5 Gallons
1 Small Jar
1 Quart
1 Quart
1 Drum
5 Gallons
5 Gallons
25 Gallons
25 Gallons
25 Ibs.
8/13/79
8/13/79
8/13/79
2/2/78
2/2/78
2/2/78
5/14/79
5/14/79
4/17/79
1/18/79
11/13/78
3/14/78
3/14/78
9/30/77
9/22/78
8/18/78
10/17/78
10/17/78
10/20/78
Champion Coal

2nd Cyclone Flyash       Dolomite
3rd Cyclone Flyash       Dolomite
Bed Overflow             Dolomite
2nd Cyclone Flyash       Dolomite
Bed Overflow             Dolomite
2nd Cyclone Flyash       Dolomite
3rd Cyclone Flyash       Dolomite
                         1  Drum
               68
               67

               79
               79
               80
               80
               80
           10
           10

            2
            3
           21
           15
            1
Ibs.
Ibs.
Drums
Drums
Drums
Drums
Drum
3/20/79

5/2/78
5/2/78

1/24/79
1/24/79
1/24/79
1/24/79
1/24/79

-------
                         APPENDIX  0  (CONT'D).  MINIPLANT SAMPLE SHIPMENTS
 	Requestor	
 Radian Corporation (Cont'd)
 Sandia Laboratory
 Llvermore, CA

 TVA
 Muscle Shoals, AL

 University of Cincinnati
 Cincinnati, OH
University of Denver
Denver, CO


University of S. California
Los Angeles, CA
     Sample  Description     Sorbent

  3rd Cyclone  Flyash       Dolomite
  Bed Overflow            Dolomite
  2nd Cyclone  Flyash       Dolomite
  Bed Overflow            Dolomite
  2nd Cyclone  Flyash       Dolomite
  3rd Cyclone  Flyash       Dolomite
  3rd Cyclone  Flyash       Dolomite

  2nd Cyclone  Flyash       Dolomite
  3rd Cyclone  Flyash       Dolomite

  Bed  Overflow             Dolomite
 3rd Cyclone Flyash       Dolomite
 3rd Cyclone Flyash       Dolomite
 2nd Cyclone Flyash       Dolomite
 3rd Cyclone Flyash       Dolomite
 2nd Cyclone Flyash       Dolomite
 Granular Bed             Dolomite
   Filter Particulates

 Illinois Coal  No.  6
 Pfizer  Dolomite
 2nd Cyclone Flyash       Dolomite
 3rd Cyclone Flyash       Dolomite
 Bed Overflow             Dolomite
 Champion Coal
 Ohio Coal
 Pfizer  Dolomite
 Grove Limestone
 2nd Cyclone  Flyash       Dolomite
 Bed Overflow             Dolomite
2nd Cyclone Flyash       Limestone
3rd Cyclone Flyash       Limestone
Run No.
81
79
79
80
80
80
81
78
78
78
80
81
78
106
59
59
•• •-
--
78
78
78
—
--
_.
—
50
50
75
75
Amount
1/2 Drum
1 Small Bottle
1 Small Bottle
1 Small Bottle
1 Small Bottle
1 Small Bottle
1 Small Bottle
10 Ibs.
10 Ibs.
6 Drums
5 Gallons
5 Gallons
5 Gallons
10 Gallons
1 Ib.
1 Ib.
1 Ib.
1 Ib.
1 Ib.
1 Ib.
1 Ib.
1 Ib.
1 Ib.
1 Ib.
1 Ib.
1 Ib.
1 Ib.
1 Ib.
1 Ib.
Date
1/24/79
1/19/79
1/19/79
1/19/79
1/19/79
1/19/79
1/19/79
2/6/79
2/6/79
5/8/79
2/6/79
2/27/79
2/27/79
8/9/79
9/7/77
9/7/77
8/15/78
8/1 5/78
8/15/78
8/15/78
8/15/78
10/2/78
10/2/78
10/2/78
10/2/78
10/2/78
10/2/78
10/2/78
10/2/78

-------
                            APPENDIX 0  (CONT'D).  MINIPLANT SAMPLE SHIPMENTS
CO
o
Requestor
University of S. California
(Cont'd)


Valley Forge Laboratories
Devon, PA
Vanderbilt University
Nashville, TN

Westinghouse Research Labs
Pittsburgh, PA





















Sample Description
Bed Overflow
2nd Cyclone Flyash
3rd Cyclone Flyash
Bed Overflow
Bed Overflow
2nd Cyclone Flyash
Bed Overflow
Bed Overflow
Bed Overflow
3rd Cyclone Flyash
3rd Cyclone Flyash
3rd Cyclone Flyash
2nd Cyclone Flyash
2nd Cyclone Flyash
3rd Cyclone Flyash
2nd Cyclone Flyash
Bed Overflow
2nd Cyclone Flyash
Bed Overflow
2nd Cyclone Flyash
3rd Cyclone Flyash
Bed Overflow
2nd Cyclone Flyash
3rd Cyclone Flyash
Bed Overflow
2nd Cyclone Flyash
3rd Cyclone Flyash
Bed Overflow
2nd Cyclone Flyash
3rd Cyclone Flyash
Bed Overflow
3rd Cyclone Flyash
Sorbent
Limestone
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Limestone
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Dolomite
Limestone
Limestone
Limestone
Dolomite
Run No.
75
77
77
77
67
67
26
45
52
67
79
59
79
68
68
69
69
70
70
71
71
71
72
72
72
73
73
73
74
74
74
81
Amount
1 Ib,
1 Ib.
1 Ib.
1 Ib.
25 Ibs.
25 Ibs.
5 Ibs.
5 Ibs.
5 Ibs.
500 Grams
1 Drum
10 Ibs.
10 Ibs.
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Gallon
Date
10/2/78
10/2/78
10/2/78
10/2/78
2/2/78
2/2/78
4/26/78
4/26/78
4/26/78
2/16/78
10/27/78
10/27/78
10/27/78
1/8/79
1/8/79
1/3/79
1/3/79
1/3/79
1/3/79
1/3/79
1/3/79
1/3/79
1/3/79
1/3/79
1/3/79
1/3/79
T/3/79
1/3/79
1/3/79
1/3/79
1/3/79
2/26/79

-------
                             APPENDIX 0  (CONT'D).  MINIPLANT SAMPLE SHIPMENTS
               Requestor
     Westinghouse  Research  Labs
     (Cont'd)
   Sample Description     Sorbent
1st Cyclone (Dipleg)     Dolomite
 Material
3rd Cyclone Flyash       Dolomite
2nd Cyclone Flyash       Limestone
3rd Cyclone Flyash       Limestone
Bed Probe                Limestone
Regenerator Cyclone Ash  Limestone
Regenerator Final Bed    Limestone
Run No.
79
81
105
105
105
105
105
Amount
2-1/2 Gallons
2 Gallons
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
1 Small Jar
 Date

4/28/79

5/11/79
8/17/79
8/17/79
8/17/79
8/17/79
8/17/79
CO
o
en

-------
                               APPENDIX P-l.  BENCH COMBUSTOR RUN SUMMARY
OJ
o
en
 Program

 Run No.                       1.1

 Operating Conditions

 Pressure (kPaa)
 Bed Temperature  (°C)
 Air Flow Rate  (m3/m1n)
 Excess  A1r  (%)
 Coal  Feed Rate (kg/hr)
 Expanded  Bed Height  (m)
 Superficial Velocity (m/s)
 Gas Residence Time (s)
 Ca/S Molar Ratio
 Run Length at Steady State  (hr) 3.7

 Flue Gas  Emissions

 S02 (ppm)
 NT (ppm)
 COX (ppm)
 C02 (X)
/\ » I at \
                                            Initial Checkout
                                       (1)
     Results

     S02 Retention
     Lbs SOp/MBTU
     Lbs NOVMBTU
           rt
                                              1.2
(1)
       2.1
(1)
       3.1
(1)
       3.2
(1)
       4.1
(1,2)
         4.2
(1,2)
770
880
2.9
146
10.3
0.5
1.9
0.28
2.43
•) 3.7
550
200
175
12.0
13
62
2.0
0.50
770
890
2.9
153
10.5
0.5
1.9
0.27
2.71
0.6
450
180
125
12.0
13
69
1.6
0.46
760
910
3.0
111
10.1
0.5
1.9
0.26
3.54
1.9
570
270
500
11.4
11
60
2.2
0.74
760
880
1.6
37
8.0
0.9
1.1
0.83
2.78
2.2
790
190
325
13.6
5.8
57
2.0
0.35
760
780
1.7
41
7.5
0.9
1.1
0.88
2.87
1.5
900
200
350
11.4
6.2
47
2.6
0.42
770
865
2.2
16
12.1
1.1
1.4
0.77
2.87
1.4
1000
120
200
16.8
2.9
54
2.4
0.20
770
870
2.2
13
12.9
1.2
1.4
0.84
2.68
0.5
800
140
225
17.8
2.5
65
1.8
0.22
     (1)  Grove Limestone; ArkwHght Coal

     (2)  1st Cyclone Sol Ids Recycled

-------
                      APPENDIX P-l (CONT'D).  BENCH  COMBUSTOR RUN SUMMARY


Program                                Initial  Checkout

Run NO.                         s.i'1.2'       6.1"'      6.2'1'      r.i'1'      e.i'1'2'       9..

Operating Conditions

Pressure (kPaa)                  760           760         750         750          750          770
Bed Temperature (°C)             880           915         900         950          930          960
Air Flow Rate (m3/min)             1.8          1.9         2.0         1.3          1.6          1.4
Excess Air (%)                    16            41          13          44           37           11
Coal Feed Rate (kg/hr)            10.5          10.2        11.3         4.7          7.1         11.0
Expanded Bed Height (m)            0.9          0.8         1.1          0.5          0.6          0.9
Superficial  Velocity (m/s)         1.2          1.3         1.7         1.0          1.0          1.3
Gas Residence Time (s)             0.76        0.61        0.84        0.54         0.61          0.66
Ca/S Molar Ratio                   0.96        1.09        1.58        3.77         1.30          1.97
Run Length at Steady State (hr)    2.2          0.9         0.4         1.3          1.2          0.3

Flue Gas Emissions
SO, (ppm)                        630           430         100         100          160          218
NCT (ppm)                         95           150         120         180          160          160
CO  (ppm)                        175           125         150          75          250          125
CO, (%}                           15.4          14.2        15.0        12.5         14.8         16.4
Og  (%)                            3.0          6.2         2.5         6.5          5.8          2.1

Results
S02 Retention (%)                 72            81          95          92           91           89
Lbs SOo/MBTU                       1.4          1.0         0.23        0.36         0.47          0.35
Lbs NO /MBTU                       0.15        0.26        0.20        0.46         0.34          0.18
      /\
(1)  Pfizer Dolomite; Arkwright Coal
(2)  1st Cyclone Solids Recycled

-------
                               APPENDIX P-2.  BENCH COMBUSTOR RUN SUMMARY


       Program                       Two Stage Combust1on-NOX Control

       Run No.                                  10.r1^    11.1^     11.2^    12.1^   12.2^   13.
       Operating Conditions

       Pressure (kPaa)                          770        765        765       770       770       770
       Bed Temperature (°C)                     920        870        870       960       950       950
      Air Flow Rate (m3/min)                     0.9        1.9        1.8      1.5       1,4       1.7
      Excess A1r (%}                            39         23        18       25         8        17
      Coal Feed Rate (kg/hr)                     8.3       18.2       18.2     10.9      10,9      13.5
      Expanded Bed Height (m)                    0.9        1.0        1.2      1.1       1,3       1.1
      Superficial  Velocity (m/s)                  0.6        1.2        1.1      1.3       1.3       1.5
      Gas Residence Time (s)                     1.4        0.8        1.1      0.79      1.03      0.70
      Ca/S Molar  Ratio                           6.2        2.9        2.9      3.5       3.5       3.1
      Run Length  at Steady State  (hr)             0.8        0.3        0.2      0.7       0.4       0.6
      Primary A1r (% Stoic.)                    70         66         61       105        75        97
      Secondary/Primary  A1r                      0.37        0.32       0.34     0         0.34      0
o     Secondary A1r Injection (cm Above Grid)    30.7        30.7       30.7      —        30.7

      Flue Gas  Emissions
      SO,  (ppm)                                900       1455       1170       588       960       400
      NO   (ppm)                                120         60        140       250       120       240
      CO   (ppm)                                200      >5000        650       125       150       140
      CO-,  (%)                                   13.0       15.4       18.6     15.4     19.0     17.6
                                                6.0        4.0        3.2      4.3      1.6      3.1
     Results

     SOo Retention (%)                         42         17         37        65        53        79
     Lbs SOo/MBTU                               1.27       1.86       1.47     0.98      1.54      0.62
     Lbs NO^/MBTU                               0.12       0.06       0.12     0.30      0.14      0.27
           rt
     (1)  Pfizer Dolomite; Champion Coal
     (2)  Grove Limestone; Champion Coal

-------
                           APPENDIX  P-2  (CONT'D).  BENCH COMBUSTOR RUN SUMMARY


      Program                       Two  Stage  Combustion-N0x Control

      Run No.                                 13.2'^   14.1^'   14.2^   17.1^    17.2^^   17A.l'2^
      Operating Conditions

      Pressure (kPaa)                         770       760       760       770       770        480
      Bed Temperature (°C)                    950       880       930       850       830        870
      Air Flow Rate (m3/min)                     1.7       1.8       1.5       1.3       1.3        1.5
      Excess Air (%)                            16        19        20        21        31         41
      Coal  Feed Rate  (kg/hr)                    11.4      12.5      11.0      10.5       9.1       10.0
      Expanded Bed Height (m)                    1.0       1.6       1.2       1.2       1.2        1.0
      Superficial  Velocity (m/s)                 1.5       1.5       1.4       1.0       1.1        2.0
      Gas Residence Time (s)                     0.70      1.1       0.87      1.2       1.1        0.5
      Ca/S Molar Ratio                           3.1       2.8       2.9       2.7       3.2        2.9
      Run Length at Steady State  (hr)            1.0       1.0       2.0       0.5       0.5        0.6
w     Primary Air (%  Stolch.)                   91       112        79        92        86        116
S     Secondary/Primary A1r                     0.24      0         0,38      0         0.33       0
      Secondary Air Injector  (cm  Above Grid)     5.9       —        5.9       —        5.9

      Flue Gas Emissions

      S02 (ppm)                               600       840       453       769       829        795
      NO^ (ppm)                               145       180       150       140       160        190
      COX (ppm)                               125       150       150       275       250        400
      C0? (%}                                  16.4      15.2      15.4      14.6      14.0       14.4
      02  (%)                                   2.9       3.4       3.5       3.8       5.0        6.2

      Results
      S02 Retention (%)                        68        60        75        57        50         48
      Lbs S02/MBTU                              1.01      1.48      0.79      1.15      1.52       1.48
      Lbs NOX/MBTU                              0.19      0.23      0.19      0.15      0.21       0.26
      (1)   Grove Limestone;  Champion Coal
      (2)   Grove Limestone;  Arkwright Coal

-------
                      APPENDIX  P-2  (CONT'D).  BENCH COMBUSTOR RUN SUMMARY
  Program                       Two Stage Combustion-N0x Control
  Run  No.                                  17A.2      17A.3^   20.1^   20. 2^    20.3^    21
(1)  Grove Limestone; Arkwrlght Coal
(2)  Grove Limestone; Champion  Coal
  Operating Conditions
  Pressure  (kPaa)                           480        480        490       490       490        480
  Bed Temperature jf°C)                      810        820        900       890       900        840
  Air Flow Rate (m3/m1n)                      1.3        1.4        1.6       1.5        1.5        1.2
  Excess Air (%)                             18         18         15        18        19        18
  Coal Feed Rate (kg/hr)                      9.4        9.5       10.8      11.2       11.3      10.0
  Expanded Bed Height (m)                     1.4        1.2        1.1        1.2        1.2        1.2
  Superficial  Velocity (m/s)                  1.7        1.8        2.2       2.0        2.1        1.6
 Gas Residence Time (s)                      0.76       0.67        0.52       0.59       0.57       0.77
 Ca/S Molar Ratio                             3.1         3.1         3.0       2.9        2.8        2.9
 Run Length at Steady State (hr)             1.9        2.0        1.3        1.5        2.5        1.3
 Primary Air  (% Stolen.)                    89         72        115         85        74        93
 Secondary/Primary A1r                       0.24       0.57        0          0.22       0.42       0
 Secondary A1r Injector (cm Above  Grid)      15         15          —        30        30
 Flue Gas  Emissions

 S0? (ppm)
 N(T (ppm)
 COX (ppm)
 COo
 °2
 Results
 SO? Retention  (%)                          33         25         43        42        42        19
 Lbs SOo/MBTU                                2.16       2.53       2.02      1.79       1.78      2.25
 Lbs NO /MBTU                                0.23       0.25       0.32      0.23       0.20      0.20
1230
185
500
15.0
3.3
1390
190
550
15.0
3.2
1080
240
300
16.4
2.8
1080
190
300
15.8
3.2
1060
190
300
15.8
3.4
1490
170
540
14.4
3.3

-------
                    APPENDIX P-2 (CONT'D).   BENCH  COMBUSTOR  RUN SUMMARY


Program                      Two Stage Combustion-N0x  Control
Run No.                                    21.2^)     21.3^     22.1^     22.2^     22.3^
Operating Conditions

Pressure (kPaa)                             480          505         515         505         505
Bed Temperature (°C)                        860          840         950         880         890
A1r Flow Rate (m3/m1n)                        1.3          1.2         2.0         1.7         1.6
Excess A1r (%)                               16           15          35          37          31
Coal Feed Rate (kg/hr)                       10.6          9.6        14.2        11.5        11.4
Expanded Bed Height (m)                       1.3          1.4         1.3         1.1         1.2
Superficial  Velocity (m/s)                    1.7          1.5         2.7         2.2         2.1
Gas Residence Time (s)                        0.74        0.91        0.48        0.52        0.56
Ca/S Molar Ratio                              2.7          2.93        3.20        3.13        3.15
Run Length at Steady State (hr)               1.3          1.4         1.5         2.0         1.7
Primary A1r (% Stolch.)                      80           73         111          91.2        80
Secondary/Primary A1r                         0.23        0.33        0           0.24        0.39
Secondary A1r Injector (cm Above Grid)       30           30

Flue Gas Emissions
S02 (ppm)                                  1410         1020         500         900         880
N(T (ppm)                                   190          140         245         230         210
CO  (ppm)                                   425          500         225         350         350
CO, (%)                                      15.0         15.6        14.2        15.0        15.0
02  (%)                                       3.0          2.8         5.5         5.7         5.0

Results
SOo Retention (%)                            25           46          67          43          47
Lbs SO-/MBTU                                  2.23        1.57        0.89        1.63        1.56
Lbs NO^/MBTU                                  0.22        0,15        0.31        0.30        0.27
      3\
(1)  Grove Limestone; Champion Coal
(2)  Pfizer Dolomite; Champion Coal

-------
                                 APPENDIX P-3.  BENCH COMBUSTOR RUN SUMMARY
co
H->
ro
  Program

  Run  No.
  Operating Conditions
  Pressure (kPaa)
  Bed  Temperature  (°C)
  Air  Flow Rate (m^/min)
  Excess Air (%)
  Coal  Feed Rate (kg/hr)
  Expanded Bed Height (m)
 Superficial  Velocity (m/s)
 Gas Residence Time (s)
 Ca/S  Molar  Ratio
 Run Length  at Steady State (hr)
 NH3/NOX  Molar Ratio
 H2/NH3 Molar  Ratio
 NH3 Injection Location
 Flue  Gas  Emissions

 SOp (ppm)
 NO* (ppm)
 COX (ppm)
 C02 (X)
 022 (X)
 Results
SO? Retention  (%)
Lbs SOo/MBTU
Lbs NO /MBTU
                                          NH3 Injection-N0x Control
                                            1.1
                                          61
                                           1.25
                                           0.18
  1.2
  1.4
  3.1
  3.2
  4.1
63
 1.24
 0.16
65
 1.17
 0.23
68
 1.08
 0.17
75
 0.81
 0.13
70
 0.98
 0.13
  4.2
760
885
1.5
19
8.4
1.3
1.3
0.98
3.05
1.5
--
—
--
760
890
1.5
16
8.4
1.3
1.3
0.95
3.07
2.2
0.88
0.0
LI
760
885
1.5
16
7.9
1.2
1.3
0.93
3.22
0.9
0.88
2.00
LI
760
890
1.5
15
9.4
1.5
1.3
1.14
2.94
2.8
--
--
--
760
890
1.5
13
10.7
1.4
1.3
1.09
2.57
1.3
2.27
1.48
LI
760
895
1.4
16
11.5
1.4
1.2
1.12
2.08
1.8
--
—
—
760
900
1.4
14
11.9
1.4
1.2
1.14
2.04
1.4
2.00
1.00
LI
700
140
200
16.0
3.8
675
117
550
16.0
3.1
600
160
450
16.0
3.0
675
145
200
16.0
2.9
588
133
500
16.0
2.7
825
155
200
15.2
3.3
800
143
200
15.4
2.9
72
 0.90
 0.12
     Grove Limestone; Champion Coal

-------
                            APPENDIX P-3 (CONT'D).  BENCH COMBUSTOR RUN SUMMARY
CO
t—•
CO
Program

Run No.

Operati ng Cond i t1 ons

Pressure (kPaa)
Bed Temperature (°C)
Air Flow Rate (m^/min)
Excess Air (%)
Coal Feed Rate (kg/hr)
Expanded Bed Height (m)
Superficial Velocity (m/s)
Gas Residence Time (s)
Ca/S Molar Ratio
Run Length at Steady State (hr)
NH3/NOX Molar Ratio
Ho/NH3 Molar Ratio
N&3 Injection Location

Flue Gas Emissions

SOo (ppm)
NO^ (ppm)
CO* (ppm)
C0
      Results

      S02 Retention (%)
      Lbs S02/MBTU
      Lbs N
                                         NH,  Injection-NO,, Control
                                           «3             X
                                           4.3
5.1
5.2
5.3
5.4
5.5
5.6
760
900
1.4
17
11.9
1.4
1.2
1.13
1 .99
1.0
__
_—
__
760
895
1.5
19
7.9
1.3
1.3
1.01
3.07
1.5
_-
-_
__
760
905
1.5
19
10.0
1.1
1.3
0.89
2.46
0.8
2.18
2.94
L2
760
915
1.5
18
10.0
1.4
1.3
1.06
2.46
1.7
1.27
2.50
L2
760
905
1.5
18
9.0
1.4
1.3
1.03
2.67
1.3
--
_-
--
760
900
1.5
18
8.8
1.3
1.3
0.99
2.78
1
3.41
0.0
12
760
890
1.5
17
8.8
1.4
1 .3
1.06
2.70
1.1
8.05
0.0
12
725
170
450
15.4
3.4
725
145
350
14.8
3.3
625
210
475
14.8
3.4
625
200
375
14.8
3.3
600
160
325
15.2
3.3
575
210
300
15.2
3.2
550
250
375
15.2
3.1
75
0.79
0.13
58
1.49
0.22
71
0.99
0.24
71
0.90
0.23
69
1.04
0.20
70
1.03
0.27
71
0.99
0.32
      Grove Limestone; Champion Coal

-------
                        APPENDIX  P-3  (CONT'D),  BENCH COMBUSTOR RUN SUMMARY


  Program                            NH3  Injection-N0x Control

  Run  No-                             5.7       6.1       6.2       6.3       7.1        7.2       7.3
  Operating  Conditions

  Pressure (kPaa)                     760       660       660       660       660       660        660
  Bed  Temperature  (PC)                890       890       880       890       890       890        890
  Air  Flow Rate  (m3/min)               1.5       1.4       1.4       1.4       1.4        1.4       1.4
  Excess  Air  (%)                      17        16        18        17        19        18         17
  Coal Feed Rate (kg/hr)               8.8       9.2       8.4       8.5       9.2        9.2       9.2
  Expanded Bed Height (m)              1.3       1.3       1.3       1.4       1.3        1.3       1.3
 Superficial Velocity (m/s)           1.3       1.4       1.4       1.4       1.5        1.4       1.4
 Gas Residence Time (s)               1.0       0.93      0.92      0.96       0.90       0.92       0.90
 Ca/S Molar  Ratio                     2.70      2.77      3.03      2.97       2.80       2.80       2.80
 Run Length  at Steady State (hr)       0.9       1.2       1.0       1.0       1.2        1.1        1.3
 NH3/NOX Molar Ratio                  —         —         0.82      --         --         1.36       2.25
 H2/NH3  Molar Ratio                   —         --         0.0       --         —         0.0        0.0
 NH3 Injection Location               —         —         L3         --         —         L3         L3
 Flue  Gas Emissions
 S02 (ppm)
 NOX (ppm)
 COX (ppm)
 C02 (*)
 02  (X)
 Results
S0£ Retention (%)                    75        70        71        71        68        75        82
Lbs S02/MBTU                         0.85      0.99      0.89      0.88      1.11       0.84      0.58
Lbs NO^/MBTU                         0.22      0.16      0.12      0.19      0.18       0.12      0.09
475
170
400
15.4
3.1
650
150
475
16.0
3.0
563
103
650
16.8
3.5
575
170
675
16.8
3.5
675
155
525
14.8
3.3
525
100
625
15.2
3.3
375
83
825
16.0
3.3
Grove Limestone; Champion Coal

-------
                       APPENDIX P-3 (CONT'D).   BENCH  COMBUSTOR RUN SUMMARY


Program                             NH3 Injection-NO   Control

Run No.                                7.4         8.1         8.2         8.3         8.4         8.5

Operating Conditions
Pressure (kPaa)                      660        660          660         660         660         660
Bed Temperature (°C)                 890        890          890         890         885         880
Air Flow Rate (m^/min)                 1.4        1.3         1.3         1.3         1.3         1.3
Excess Air (%)                        16         19          18          19          17          15
Coal Feed Rate (kg/hr)                 9.2        7.9         7.9         7.9        10.1        10.1
Expanded Bed Height (m)                1.2        1.3         1.3         1.3         1.3         1.3
Superficial Velocity (m/s)             1.4        1.3         1.3         1.3         1.3         1.3
Gas Residence Time (s)                 0.84       0.95        0.98        0.97        0.96        0.93
Ca/S Molar Ratio                       2.8        3.12        3.12        3.12        2.44        2.48
Run Length at Steady State (hr)        0.8        0.9         1.1         0.7         0.8         1.1
NH3/NOX Molar Ratio                    —         —           1.24        2.29        1.39
H2/NH3 Molar Ratio                     —         —           0.0         0.0         1.46
NH3 Injection Location                 —         --           LI          LI          LI

Flue Gas Emissions
SO- (ppm)                            550        625          600         563         513         450
NO, (ppm)                            150        133          115         100         115         132
CO  (ppm)                            425        225          363         550         550         500
CO, (%)                               15.8       15.8        15.4        14.8        15.2        15.8
                                       3.0        3.6         3.3         3.4         3.2         3.0
Results
S0? Retention (%)                     74         68           70          71          79          82
Lbs S02/MBTU                           0.87       1.04          1.02        1.00        0.68        0.58
Lbs NOx/MBTU                           0.17       0.16          0.14        0.13        0.11        0.12
Grove Limestone; Champion Coal

-------
                      APPENDIX P-4.  BENCH COMBUSTOR RUN  SUMMARY


  Program             Simulated Flue Gas Reclrculatlon-NO   Control

  Run No.                           1A.1         1A.2        1A.3          2.1          2.2

  Operating Conditions
  Pressure (kPaa)                  810         810          810          810          810
  Bed Temperature (°C)             875         890          880          875          890
 Air Flow Rate (m3/min)             1.4         1.4         1.4          1.4          1.5
  Excess A1r (%)                    19           13           17           18            7
 Coal  Feed Rate (kg/hr)             9.0         9.0         9.0          7.9         10.4
 Expanded Bed Height (m)            1.4         1.3         1.3          1.3          1.4
 Superficial  Velocity (m/s)          1.2         1.3          1.2          1.2          1.5
 Gas Residence Time (s)             1.20         1.02         1.15         1.13          0.92
 Ca/S Molar  Ratio                    2.97         3.10         3.10         3.36          2.85
 Run Length  at Steady State  (hr)     1.5         2            0.7          0.5          1.0
 N2  Flow  Rate (m3/m1n)               --           0.16         --           —            0.35
 Redrculatlon Ratio  (%)             --          11.2          --           --           23.4

 Flue Gas  Emissions
 SO, (ppm)                        550          600          450           419           525
 NO^ (ppm)                        160          117          170           155            70
 COX (ppm)                        200          400          475           200           300
 CO, (%)                           15.4         15.4         16.0         14.8          14.5
 02Z (%)                            3.5          2.1          3.3           3.2           1.1

 Results
 S02 Retention  (%)                 73           67            78           77           72
 Lbs S09/MBTU                       0.89         1.17        0.72          0.81         1.09
 Lbs NO^/MBTU                       0.19         0.16        0.19          0.22         0.07
Grove Limestone; Champion Coal

-------
          APPENDIX P-4 (CONT'D).   BENCH  COMBUSTOR  RUN SUMMARY


Program      Simulated Flue Gas Reclrculation-NO   Control
                                               /V

Run No.                            3.1          3.2          4.1          4.2

Operating Conditions
Pressure (kPaa)                  810          810          810          810
Bed Temperature (°C)             910          900          820          820
Air Flow Rate (m3/min)             1.8          1.8          1.3          1.3
Excess Air (%)                    27           25           17           13
Coal  Feed Rate (kg/hr)             9.6          10.9          8.4          8.7
Expanded Bed Height (m)            1.3          1.2          1.3          1 .4
Superficial  Velocity (m/s)         >.5          1.7          1.0          1.1
Gas Residence Time (s)             0.86         0.73         1.39         1.29
Ca/S Molar Ratio                   3.05         2.87         2.65         2.87
Run Length at Steady State (hr)    1.5          2.0          1.0          2.0
N2 Flow Rate (m3/min)              —           0.20         —           0.14
Redrculation Ratio (%)            —           11.2          --           5.0
Flue Gas Emissions
S0? (ppm)                        575          575          750          938
NO* (ppm)                        160          130          260          150
COX (ppm)                        175          550          425          700
CO- («)                           16.0          15.4         14.8         14.2
02  (%}                            5.0          4.0          3.0          2.0

Results
S02 Retention (%)                 67           67           66           54
Lbs S02/MBTU                       1.01         1.07         1.18         1.72
Lbs NOV/MBTU                       0.20         0.17         0.29         0.13
      f\
Grove Limestone; Champion Coal

-------
                               APPENDIX P-5.  BENCH COMBUSTOR RUN SUMMARY


           Program                     Combined NOX Control  Methods

           Run  No.^                         1.1          1.2          1.3           1.4          1.5

           Operating Conditions
           Pressure (kPaa)                  500          500          500          500          500
           Bed  Temperature (°C)             890          900          890          900          900
           A1r  Flow Rate (m3/m1n)             1.5          1.5          1.5           1.5          1.5
           Excess A1r (%)                    25           23           25          23           22
           Coal Feed Rate (kg/hr)             9.0          8.7          9.0         10.4          9.0
           Expanded Bed Height (m)            1.2          1.3          1.2           1.2          1.3
           Superficial  Velocity (m/s)         1.9          1.9          2.0           2.0          2.0
           Gas Residence Time (s)             0.63         0.65          0.63          0.60         0.64
           Ca/S Molar Ratio                   3.08        3.17          3.08         2.64         3.08
           Run Length at Steady State (hr)     0.8         0.6          0.4          0.6          0.7
2          Primary A1r  (% Stolch.)          105          109         106           91           72
00         Secondary/Primary Air (2)          0           0            0            0            0.53
          NH3/NOX Molar Ratio (3)            —           0.89          --           1.43

          Flue Gas  Emissions

          S02 (ppm)
          MT (ppm)                         180          120          125           95           82
          CO  (ppm)                         325          550          650          740          525
          CO, (%}                           14.8         14.8         14.8         14.8         15.2
          Og  (%)                            4.5          4.0          4.5          4.3          4.0

          Results

          S0? Retention  (%)
          Lbs SO-/MBTU
          Lbs MT/MBTU                       0.21         0.15         0.15          0.09          0.10
          (1)  Grove Limestone; Champion Coal
          (2)  Supplementary air Injected 33 cm above grid.
          (3)  NH3 Injected 290 cm above grid.

-------
                           APPENDIX P-5 (CONT'D).  BENCH COMBUSTOR RUN SUMMARY
CO
>->
10
                 Program

                 Run No.
              Combined N0  Control Methods
(1)
Operating Conditions
Pressure (kPaa)
Bed Temperature (°C)
Air Flow Rate (m3/min)
Excess Air (%)
Coal Feed Rate (kg/hr)
Expanded Bed Height (m)
Superficial  Velocity (m/s)
Gas Residence Time (s)
Ca/S Molar Ratio
Run Length at Steady State (hr)
Primary Air (% Stoich.)
Secondary/Primary Air (2)
NH./NOV Molar Ratio (3)
  »5   X
Flue Gas Emissions
S02 (ppm)
NO  (ppm)
CO  (ppm)
C02 (%)
f\ *•* f a/ \
Results
S0? Retention (%)
Lbs SOp/MBTU
Lbs NCT/MBTU
                            2.1
500
880
  1.5
 31
  9.7
  1.1
  2.0
  0.56
  3.02
  0.4
101
  0
                                                  575
                                                  153
                                                  400
                                                   14.2
                                                    5.5
                                  73
                                   0.87
                                   0.17
               2.2
                                       500
                                       880
                                         1.6
                                        25
                                         9.7
                                         1.2
                                         2.1
                                         0.58
                                         3.02
                                         0.4
                                        72
                                         0.50
                                       650
                                       120
                                       375
                                        15.2
                                         4.5
                                                                68
                                                                 1.03
                                                                 0.14
                                                             2.3
                          675
                          105
                          425
                           14.8
                            5,0
                                                     65
                                                      1.15
                                                      0.13
                                                                         2.4
510
890
1.6
25
9.3
1.3
2.1
0.61
3.15
0.5
75
0.50
1.61
510
880
1.6
24
9.0
1.2
2.1
0.60
3.28
0.3
78
0.50
—
                                                                        675
                                                                        113
                                                                        440
                                                                         14.8
                                                                          4.3
                                        63
                                         1 .22
                                         0.15
                 (1)   Grove Limestone; Champion Coal
                 (2)   Supplementary air Injected 33 cm above grid.
                 (3)   NH3 injected 290 cm above grid.

-------
                              TECHNICAL REPORT DATA
                        (Please read Instructions on the reverse before completing)
1 REPORT NO.
 EPA-600/7-80-013
                                                     RECIPIENT'S ACCESSION NO.
4. TITLE ANDSUBTITLE
Miniplant and Bench Studies of Pressurized Fluidized-
 bed Coal Combustion: Final Report
                                                     REPORT DATE
                                                    January 1980
                                                     PERFORMING ORGANIZATION CODE

                                                     EXXON/GRU.18GFGS.79
7.
          t c. Hoke ,E.S. Matulevicius ,M. Ernst,J. L.
Goodwin,A.R.'Garabrant,I.B.Radovsky,A. S. Lescar-
ret.R.R.Bertrand.L.A.Ruth.V.J.Siminski  (see blk 15)
                                                     . PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING OFIOANIZATION NAME AND ADDRESS
Exxon Research and Engineering Co.
P.O. Box 8
Linden, New Jersey 07036
                                                     0. PROGRAM ELEMENT NO.
                                                    INE825
                                                    11. CONTRACT/GRANT NO.

                                                    68-02-1312
 12. SPONSORING AGENCY NAME AND ADDRESS
 EPA, Office of Research and Development
 Industrial Environmental Research Laboratory
 Research Triangle Park, NC 27711
                                                    13. TYPE OF REPORT AND PERIOD COVERED
                                                    Final;  8/77 - 8/79	
                                                    14. SPONSORING AGENCY CODE
                                                      EPA/600/13
 15. SUPPLEMENTARY NOTES IERL-RTP project officer is D. Bruce Henschel, Mail Drop 61,
 919/541-2825. (From blk 7: M.S.Nutkis,M.D. Loughnane,H.R.Silakowski, M.W.Gre-
 gory, and A.Ichel.) EPA-600/7-78-069, -77-107, and -76-011 are related reports
 16. ABSTRACT
          The report gives further results of studies on the environmental aspects
 of the pressurized fluidized-bed coal combustion process, using the 218 kg coal/hr
 continuous combustion/sorbent  regeneration Miniplant  (0.63 MW equivalent), and a
 13 kg coal/hr bench-scale system. Tests on the Miniplant combustor confirmed its
 ability to achieve over 90% SO2 removal with either limestone or dolomite sorbent.
 Studies of dynamic response indicated that the combustor responds much more
 quickly to changes in coal sulfur content than to changes in sorbent feed rate.  High
 temperature/pressure particle  control device testing on the Miniplant addressed:
 conventional and alternative-design cyclones, a ceramic fiber filter, and a granular
 bed filter. Three stages of cyclones may reduce dust loading sufficiently to protect
 a gas turbine. A conventional low-pressure  electrostatic precipitator and fabric
 filter were also tested.  Further tests on the Miniplant regenerator confirmed that
 regeneration can reduce fresh sorbent feed requirements by a factor of 3 to 4. Addi-
 tional sampling was completed  on the Miniplant combustor and regenerator for com-
 prehensive analysis of emissions. NOx control studies in the bench combustor sug-
 gested that NOx emissions might be reduced by 20 to 50% through two-stage combus-
 tion or ammonia injection; flue gas recirculation had little effect.
 17.
                             KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
Pollution
Coal
Combustion
Fluidized Bed
Processors
Flue Gases
Calcium Carbonates
Dolomite
Sulfur Dioxide
Dust
Cyclone Separators
Ceramics
Fabrics
Release to Public
b.lDENTIFIERS/OPEN ENDED TERMS
Pollution Control
Stationary Sources
Particulate
Fabric Filters
19. SECURITY CLASS (THil Report)
Unclassified
20. SECURITY CLASS (This page)
Unclassified
c. COSATi Field/Gioup
13B 07B
2 ID 08G
2 IB
11G
131, 07A
11B
11E
21. NO. OF PAGES
332
22. PRICE
 EPA Form 2220-1 (9-73)
                                        320

-------