Topics:                               EPRI CS-3706
Flue gas desulfurization                     Volume 2
Sulfur oxides                            Project 982-31
Nitrogen oxides                          Proceedings
Wet scrubbers                           November 1984
Dry scrubbers
Pollution control equipment
Proceedings: Eighth Symposium
on Flue Gas Desulfurization
Volume 2

-------
                                 RJE  PORT       SUM   M _A_ R_ Y
                     SUBJECTS  SOX control / NOX control / Solid by-product disposal/reuse / Integrated
                                 environmental control
                        TOPICS  Flue gas desulfurization
                                 Sulfur oxides
                                 Nitrogen oxides
                                 Wet scrubbers
                                 Dry scrubbers
                                 Pollution control equipment
                     AUDIENCE  Environmental engineers / Generation planners
                                 Proceedings: Eighth Symposium on Flue
                                 Gas Desulfurization
                                 Volumes 1 and 2
                                 Timely exchanges of technical and economic information on
                                 flue gas desulfurization (FGD) systems are essential to coal util-
                                 ities that must meet strict emissions standards. These proceed-
                                 ings constitute a valuable resource for utility, architectural-
                                 engineering, and system-supplier personnel who must make
                                 decisions about  the design, installation, and operation of FGD
                                 systems.
                  BACKGROUND
                     OBJECTIVE
                     APPROACH
                    KEY POINTS
Sulfur dioxide (SO2) emissions from coal-fired generating plants must be
carefully controlled to comply with government regulations. Compliance,
however, frequently means that utilities must install expensive and compli-
cated FGD systems. Therefore, utilities faced with limiting S02 emissions
need up-to-date information on this rapidly evolving technology in order to
select the most reliable and cost-effective process.

To provide a forum for exchanging information on the scientific, technical,
and regulatory developments related to SO2 control.

The EPA and EPRI cosponsored a four-day symposium that featured the
presentation of 40 technical papers and a major panel discussion. Utility and
industrial users and representatives of FGD system suppliers, research insti-
tutions, and government agencies were invited to contribute papers empha-
sizing progress in SO2 control, recent experience with installed systems, and
pertinent test results. Some 730 persons attended.

In the keynote address, the executive director of the National Acid Precipita-
tion Assessment Program examined the program's purpose, scope, and
status and its focus on providing Congress with a better scientific basis for
legislative and regulatory decisions. Nine other sessions included such
diverse topics as economics, construction materials, absorbent injection,
dual alkali systems, flue gas treatment (combined SOX/NOX), FGD chemistry,
limestone and organic acid, waste disposal and utilization, and dry FGD
systems (both pilot- and full-scale). The role of the architect-engineer in
EPRI CS-3706S Vols. 1 and 2

-------
       EPRI PERSPECTIVE
                 PROJECT
constructing FGD systems for utilities was the topic of a panel discus-
sion conducted by representatives from seven architectural-engineering
firms. The purpose of the architect-engineer, all agreed, was to serve as
an extension of the utility's own engineering staff.

Of eight flue gas desulfurization symposia that have been held, this is
the second EPRI has cosponsored. The meetings, held approximately
every 18 months, bring together FGD vendors, government regulators,
researchers, and architect-engineers. In this relatively new and contin-
ually changing technology, the symposia present an excellent opportu-
nity for a wide-ranging interchange of FGD information and  experience.
The meetings are well attended, and the published proceedings provide
a comprehensive and useful source of up-to-date happenings in SO2
control. The next symposium is planned for June 1985 in Cincinnati.

RP982-31
EPRI Project Manager: Thomas M.  Morasky
Coal Combustion  Systems Division
Contractor: Research Triangle Institute
                            For further information on EPRI research programs, call
                            EPRI Technical Information Specialists (415) 855-2411.
  ORDERING INFORMATION
             EPRI Members
 EPRI CS-3706 Vols. 1 and 2, Proceedings, November 1984.
 V1, 594 pages. V2, 578 pages.

 If this report is not available from your company libraries or your
 Technical Information Coordinator, you can order it from
 Research Reports Center
 P.O. Box 50490
 Palo Alto, CA 94303
 (415) 965-4081
              Nonmembers  You can order this report in print or microfiche from
                            Research Reports Center.
                            Price: V1 $41.50; V2 $41.50   Overseas price: V1 $83.00; V2 $83.00
                            (California residents add sales tax.)
                            Payment must accompany order.
;>•: Po.'.er Research institute PO BOA 10412 Palo Alto, CA 94303 All rights reserved

-------
Proceedings: Eighth Symposium on Flue Gas
                  Desulfurization
                       Volume 2
                   CS-3706, Volume 2
                Research Project 982-31

                Proceedings, November 1984

                  New Orleans, Louisiana
                   November 1-4, 1983
                       Prepared by

              RESEARCH TRIANGLE INSTITUTE
                     Cornwallis Road
         Research Triangle Park, North Carolina 27709

                        Compiler
                        F. A Ayer
                       Prepared for

              Environmental Protection Agency
             Office of Research and Development
                     401 M Street SW
                  Washington, DC 20460

         Industrial Environmental Research Laboratory
         Research Triangle Park, North Carolina 27711

                    EPA Project Officer
                       J. W. Jones

                          and

              Electric Power Research Institute
                   3412 Hillview Avenue
                 Palo Alto, California 94304

                  EPRI Project Manager
                      T. M. Morasky

             Desulfurization Processes Program
             Coal Combustion Systems Division

-------
                                ORDERING INFORMATION
Requests for copies of fhis reporf should be directed to Research Reports Center
(RRC), Box 50490, Palo Alto, CA 94303, (415)  965-4081. There is no charge for  reports
requested by EPRI member utilities and affiliates, U.S. utility associations,  U.S.  government
agencies (federal, state, and local), media, and foreign organizations with which EPRI  has an
information exchange agreement. On  request, RRC will send a catalog of EPRI  reports.
 Reseaich Calegones  SOV control
                   NO. control
                   Solid by-broduct disposal/reuse
                   Integrated environmental contra

 Copv'ignt -  '9S- Electric Pov.er Research Institute, Inc All rights reserved


                                           NOTICE
 T"is -ecor! >,vas prepared as an account ot work sponsored in part by the Electric Power Research Institute, Inc
  EPRi' N-E- "•?• EPRI rremoers of EPRI, nor any person acting on their behall  (a) makes any warranty, express or
  mpi'ec1 .vth respect 'o :ne use of any information apparatus, method or process disclosed in this report or that
 sucn use mav not infringe privately owned rights, or (b) assumes any liabilities with respect to the use of, or for
 dar-.ages 'esuiting from  ire  use of, any information, apparatus, method, or process disclosed in this report

-------
                                  ABSTRACT
These proceedings are  of the Eighth Symposium on  Flue  Gas  Desulfurization,
held November  1  to  4,  1983, in  New  Orleans,  Louisiana.  The  symposium  was
sponsored by EPA's Industrial Environmental Research Laboratory,  located in
Research Triangle Park,  North Carolina,  and  the EPRI  Coal  Combustion  Sys-
tems Division, located in Palo Alto, California.

The objective  of the symposium  was  to  provide  a forum for  supplier,  user,
service,  and  regulatory groups  to  discuss  the  technical  and  regulatory
aspects  of  SOp  control.    The  emphasis  was on  progress  in SO^  control
technology,  recent   experience,   and test  results,  not  on  future  plans.
Volume  1  contains  24 papers from  days  one,  two, and three, plus  abstracts
from panel  members.   Volume 2 contains 16 papers from  days  three  and  four,
plus 6 unpresented papers.
                                     111

-------
                                  PREFACE
These  proceedings  for  the  Eighth Symposium  on Flue  Gas  Desulfurization
(FGD) constitute  the  final  report  submitted  to the  Industrial  Environmental
Research  Laboratory;   EPA;  Research   Triangle  Park   (IERL-RTP),   North
Carolina;   and   EPRI's  Coal   Combustion   Systems   Division,   Palo   Alto,
California.   The  symposium was  conducted  at  the  Sheraton  Hotel  in  New
Orleans, Louisiana, November 1  to 4, 1983.

The meeting  served  as a forum for the exchange  of  technical and  regulatory
information  and  developments regarding  systems and processes  applicable  to
utility  and  industrial boilers.    At  the opening  session,   the keynote
address examined  the  status and outlook  for  the National  Acid  Precipitation
Assessment  Program:  its present  status  and outlook for  the future.    Pre-
sentations were also  made on the  state of air  quality legislation and  regu-
lations, current  and  projected  regulations of  the Resource  Conservation and
Recovery Act, and trends in  commercial application  of FGD technology.   Sub-
sequent  technical sessions  dealt  with  economics,  construction  materials,
dry furnace  absorbent injection,  dual-alkali,  flue gas treatment  (combined
SOX/NOX).   Other sessions  included  FGD  chemistry,  limestone/organic  acid,
waste disposal/utilization, and dry FGD  systems, pilot plant  test results,
and  full-scale  installations.    Participants  also  discussed  the role  of
architect-engineer as middleman between the utility and FGD suppliers.

Representatives   from  electric  utilities,  state   environmental   agencies,
equipment  and  process  suppliers,  coal  and  petroleum  suppliers,  EPA  and
other federal agencies, and research organizations  attended the sessions.

The following people contributed their efforts to this symposium.

     •    Julian  W.  Jones,  Chemical   Engineer,   Emissions/Effluent
          Technology Branch, Utilities and Industrial Power Division,
          IERL-RTP,  Research  Triangle  Park,   North  Carolina;   EPA
          symposium general chairman and project officer

     •    Thomas  M.  Morasky,   Manager,  Reliability  and  Nonrecovery
          Systems,  Coal  Combustion  Systems   Division,  Palo Alto,
          California;  EPRI   symposium  general   chairman  and  project
          manager

     •    Franklin  A.  Ayer,  Manager,   Conference   Planning  Office,
          Center  for  Technology   Applications,   Research Triangle
          Institute,   Research   Triangle   Park,    North   Carolina;
          symposium coordinator and compiler of the proceedings
Thomas M. Morasky, Project Manager
Coal Combustion Systems Division

-------
                              TABLE OF CONTENTS
                                  VOLUME 1
Section                                                           „
	                                                           Page

 SESSION 1:   OPENING SESSION
      Julian W.  Jones, Chairman

 Keynote Address:   National Acid Precipitation
 Assessment Program:  Status and Outlook	1-1
      J.  Christopher Bernabo

 Remarks	1-21
      Sheldon Meyers

 The Resource Conservation and Recovery Act:   Current.
 and Projected Regulations	1-27
      Stephen A. Lingle

 Trends in Commercial Applications of FGD	1-29
      Bernard A. Laseke,"' Michael T.  Melia,  and Norman Kaplan

 SESSION 2:   ECONOMICS
      Thomas M.  Morasky, Chairman

 Computer Economics of Physical Coal Cleaning and
 Flue Gas Desulfurization	2-1
      Charles R. Wright,"' Terry W. Tarkington, and
      James D. Kilgroe

 Economic Evaluation of FGD Systems	2-27
      Jack B. Reisdorf,* R. J. Keeth, C. P.  Robie,
      R.  W.  Scheck, and Thomas M. Morasky

 Estimating Procedure for Retrofit FGD Costs	2-47
      R.  R.  Mora,  P. A.  Ireland," R.  J. Keeth, and
      T.  M.  Morasky

 Comparative Costs of SO^ Removal Technologies	2-63
      John 0. Milliken

 SESSION 3:   MATERIALS OF CONSTRUCTION
      Charles E. Dene, Chairman

 ""Denotes speaker
                                     VII

-------
Section                                                          Page

EPRI Research on Corrosion and Degradation of
Materials for FGD Systems	3-1
     Barry C. Syrett

Simultaneous Design, Planning, and Materials of
Construction Selection for FGD Systems	3-15
     Alex Kirschner, Norman Ostroff,* R. F. Miller,
     and W. L. Silence

Acid Deposition in FGD Ductwork	3-47
     Daniel A. Froelich,-'" Carl V. Weilert, and Paul N.  Dyer

In Situ Evaluation of High Performance Alloys in
Power Plant Flue Gas Desulfurization Scrubbers	3-61
     R. W. Schutz and Charles S. Young*

SESSION 4:  DRY FUMACE ABSORBENT INJECTION
     Randall E. Rush, Chairman

Results from EPA's Development of Limestone
Injection  into a Low NO  Furnace	4-1

     Dennis  C. Drehmel,* G. Blair Martin, and
     James H. Abbott

Review  of  EPRI Research on Furnace  Sorbent
Injection  S02 Control  	 4-19
     Michael  W. McElroy

Direct  Desulfurization Through Additive
 Injection  in the Vicinity of  the Flame	4-31
     M.  Yaqub Chughtai-" and Sigfrid Michelfelder

 SESSION 5:   DUAL ALKALI
     Norman  Kaplan,  Chairman

 Utility Double  Alkali  Operating  Experience	  .5-1
     Dennis  L.  Clancy, Richard J. Grant, L. Karl  Legatski,*
     James H. Wilhelm, and Beth  A.  Wrobel

 Pilot  Evaluation  of Limestone Regenerated Dual
 Alkali Process	    	5-21
      John C. S.  Chang"" and Norman Kaplan

 SESSION 6:  FLUE  GAS TREATMENT  (COMBINED SO /NO  )
      J. David Mobley,  Chairman             x   x

 Status of the DOE  Flue Gas Cleanup  Program	/. 1
      John E. Williams

 "'-"Denotes speaker
                                     VI11

-------
 Section                                                         Page

 Status  of SO^  and NO  Removal in Japan	6-37
      Jumpei  Ando

 PANEL:   THE  ARCHITECT-ENGINEER - MIDDLEMAN  BETWEEN
         UTILITY AND FGD SUPPLIER	6-43
      A.  V.  Slack, Chairman

      Edward  W.  Stenby, Gene H. Dyer,  Paul R.  Predick,
      Michael L. Meadows,  Douglas B.  Hammontree,
      Christopher P  Wedig, and Richard Rao,  Panel Members

 SESSION 7:   FGD CHEMISTRY
      Dorothy A. Stewart,  Chairwoman

'Influence of Chlorides on the Performance of
 Flue Gas Desulfurization	7-1
      William Downs,"" Dennis W. Johnson,  Robert W. Aldred,
      L.  Victoria Tonty, Russell F. Robards,* and
      Richard A. Runyan

 Effect  of High Dissolved Solids on Bench-Scale FGD
 Performance	7-19
      James  B.  Jarvis,* Timothy W. Trofe, and
      Dorothy A. Stewart

 Pilot Plant  Tests on the Effects of  Dissolved Salts
 on Lime/Limestone FGD Chemistry	7-37
      Dennis  Laslo,* John C. S. Chang,  and
      J.  David Mobley

 Modeling of  SO-^ Removal by Limestone Slurry
 Scrubbing:   Effects of Chlorides	7-57
      Pui K.  Chan and Gary T. Rochelle*

 Influence of High Dissolved Solids on Precipitation
 Kinetics and Solid Particle Size	7-79
      Frank B.  Meserole, Timothy W. Trofe, and
      Dorothy A. Stewart"

 Effect  of Limestone Grinding Circuit on FGD
 Performance  and Economics	7-105
      J.  David Colley,"' 0.  W. Hargrove, Jr.,  and
      Dorothy A. Stewart

                                 VOLUME 2

 SESSION 8:   LIMESTONE/ORGANIC ACID
      J.  David Mobley, Chairman

 "'Denotes speaker

-------
Section                                                           Page

Process Troubleshooting at a Utility Limestone
FGD System	8-1
     J. David Colley, Robert L. Glover,
     Temple E. Donaldson,* and Dorothy A. Stewart

Technical/Economic Feasibility Studies for Full
Scale Application of Organic Acid Technology  for
Limestone FGD Systems .	8-23
     James C. Dickerman* and J. David Mobley

SESSION 9:  WASTE DISPOSAL/UTILIZATION
     James D. Kilgroe, Chairman

Full-Scale Field Evaluation of Waste Disposal
From Coal Fired Electric Generating Plants	9-1
     Julian W. Jones,* Chakra J. Santhanam, Armand
     Balasco, Itamar Bodek, Charles B. Cooper,
     John T.  Humphrey, and Barry K. Thacker

Operations History of Louisville Gas & Electric
FGD Sludge Stabilization	9-25
     Robert P. Van Ness,"" John H. Juzwiak, and
     William  Mclntyre

Coal Waste Utilization in Artificial Reef Construction	9-37
     Jeffrey  H. Parker,* Peter M. J. Woodhead, and
     Dean M.  Golden

Solid  Waste Environmental Studies at Electric
Power  Research Institute	9-49
      Ishwar P. Murarka
      Presented by Karen Summers

 SESSION 10,  PART  I:  DRY FGD:  PILOT PLANT TEST RESULTS
      Theodore G.  Brna, Chairman

 Current Status of Dry SO^ Control Systems  	  10-1
      Michael  A. Palazzolo,* Mary E. Kelly,
      and Theodore G. Brna

 Acid Rain Prevention Thru New  SO /NO  Dry
 Scrubbing Process	*  . *	10-23
      Karsten S. Felsvang,* Per Morsing,
      and Preston  L.  Veltman
 "''Denotes speaker

-------
 Section                                                         Page

Process Characterization of SO^ Removal in
Spray Absorber/Baghouse Systems	10-41
     Eric A. Samuel,"' Thomas W. Lugar, Dennis E. Lapp,
     Kenneth R. Murphy, Owen F. Fortune, Theodore G. Brna,
     and Ronald L. Ostop

Dry Scrubber, Flue Gas Desulfurization on High Sulfur,
Coal-Fired Steam Generators:  Pilot-Scale Evaluation	10-61
     Bryan J. Jankura,* John B. Doyle, and Thomas J. Flynn

EPRI Spray Dryer/Baghouse Pilot Plant Status
and Results	10-81
     Gary M. Blythe* and Richard G. Rhudy

SESSION 10, PART II:  DRY FGD:  FULL SCALE INSTALLATIONS
     Richard G. Rhudy, Chairman

Field Evaluation of a Utility Dry Scrubbing System	10-109
     Gary M. Blythe,"- Jack M. Burke, Theodore G. Brna,
     and Richard G. Rhudy

Overview and Evaluation of Two Years of Operation
and Testing of the Riverside Spray Dryer System	10-131
     John M. Gustke, Wayne E. Morgan,""
     and Steven H. Wolf

Design and Initial Operation of the Spray Dryer
FGD System at the Marquette, Michigan, Board of
Light and Power - Shiras #3 Plant	10-161
     0. Fortune,- T. F. Bechtel, E. Puska, and J.  Arello

Start-Up and Initial Operating Experience of the
Antelope Valley Unit 1 Dry Scrubber	10-181
     Robert L. Eriksen,* Frederick R.  Stern,
     Richard P. Gleiser, and Stanley J. Shilinski

Characterization of an Industrial Spray Dryer at
Argonne National Laboratory	10-199
     Paul S. Farber* and C. David Livengood

UNPRESENTED PAPERS

An Economic Evaluation of Limestone Double Alkali
Flue Gas Desulfurization Systems	11-1
     Gerald A. Hollinden, C. David Stephenson, and
     John G. Stensland

""Denotes speaker
                                  XI

-------
Section                                                          „
	                                                          Paj|e_

Developments and Experience in FGD Mist Eliminator
Application	11-39
     Richard T. Egan and William Ellison

FGD Gypsum:  Utilization vs. Disposal	11-61
     William Ellison

Operating Experience with the Chiyoda Thoroughbred 121
Flue Gas Desulfurization System	11-75
     Seiichi Kaneda, Mitsuhiro Nishimura,
     Hitoshi Wakui, Ikuro Kuwahara, and
     Donald D. Clasen

Operation Experience with FGD Plant II at
Wilhelmshaven Power Plant, West Germany	11-91
     B. Stellbrink, H. Weissert, and P.  Kutemeyer

The SULF-X Process	11-111
     Edward Shapiro and William Ellison

APPENDIX:  ATTENDEES	A-l
                                   XII

-------
SESSION 8:   LIMESTONE/ORGANIC ACID

Chairman:   J.  David Mobley
           Industrial Environmental Research Laboratory
           U.S.  Environmental Protection Agency
           Research Triangle Park, NC

-------
PROCESS TROUBLESHOOTING AT A UTILITY LIMESTONE
                  FGD SYSTEM

         J. D. Colley, R. L. Glover,
       T. E. Donaldson, D. A. Stewart

-------
          PROCESS TROUBLESHOOTING AT A UTILITY LIMESTONE FGD SYSTEM

by:  J. David Colley
     Radian Corporation
     Austin, Texas

     Robert L. Glover
     Radian Corporation
     Austin, Texas

     Temple Donaldson
     Central Illinois Light Company
     Peoria, Illinois

     Dorothy Stewart
     Electric Power Research Institute
     Palo Alto, California

                                  ABSTRACT

     Central Illinois Light Company's Duck Creek Unit No. 1 experienced sig-
nificant reliability and operating problems with its limestone FGD system
following start-up in 1979.  CILCo entered into a testing and evaluation pro-
gram co-funded by the Electric Power Research Institute in late 1981 to
verify the  feasibility of using additives to improve the system's low S02
removal efficiency.  A second objective of the work was to improve the sys-
tem's reliability, which averaged slightly better than 60 percent prior to
testing.  Severe mist eliminator scaling that was causing routine scrubber
outages was primarily responsible for the reliability problems.  The eight
month program that followed involved extensive chemical process troubleshoot-
ing.

     Two types of organic acid buffers were tested along with the addition of
magnesium oxide during the period.  Both dibasic acid (DBA) and the magnesium
proved capable of enhancing S02 removal to levels sufficient to maintain the
unit in compliance with the 1971 S02 NSPS.  An economic analysis was performed
based on the data collected during this testing.  The cost study compared
capital and annual operating and maintenance costs for each option over the
remaining life of the plant.

     Work was conducted concurrent with the additive testing to solve the mist
eliminator  scaling.  The cause of the scaling was identified and effectively
stopped by  switching to a fresh water wash, adjusting the wash sequence, and
improving limestone utilization.  Improvements in limestone utilization were
accomplished by optimizing the operation of the mill circuit to provide a
finer, more reactive product.  Since completion of the program, the FGD sys-
tem has consistently achieved reliability numbers greater than 95 percent
while at the same time lowering operating and maintenance costs.

-------
                                INTRODUCTION

     Duck Creek #1 is a bituminous coal fired unit which is owned  and  oper-
ated by Central Illinois Light Company (CILCo).  The boiler is rated at  400
megawatts and began commercial operation in 1976.  It  fires a high sulfur
Illinois coal and is subject to the 1971 utility NSPS  which limits the units
SOz emissions to 1.2 pounds per million Btu boiler heat  input.   The FGD  sys-
tem consists of four Riley Environeering rod deck scrubber modules each  de-
signed to treat 25 percent of the maximum  gas flow.

     Since startup of its scrubbers, CILCo has  experienced many  of the same
operating problems seen at other utility limestone FGD systems.   The primary
difficulties have been low S02 removal, chemical scaling in the  mist elimi-
nators,  materials and liner failure, and low reagent utilization.   By 1981,
the utility was facing regulatory pressure because of  S02  compliance vio-
lations  and was reporting FGD reliabilities of  only  slightly  more than 60
percent.  In addition, the low reagent utilization was leading  to accelerated
 filling  of  the  FGD waste disposal pond.

      CILCo  contracted Radian Corporation in late 1981  to conduct a test  pro-
gram whose  primary objective was to demonstrate the  feasibility  of using an
 organic  buffer  to  improve the FGD system S02 removal efficiency.  Severe
weather  hampered  the  testing but the initial results were  encouraging.
 Further  test work was delayed until April  following;  a  month long plant outage.
As the testing  progressed after startup in April, the  scope of work expanded
 to include  investigating the mist eliminator scaling problem  and ways to im-
prove limestone reagent utilization.  The  Electric Power Research Institute
became involved at this stage and funded the cost of the remainder of  the
 program.  This  included conducting a test  to determine the  feasibility of
 using magnesium as an alternative to the organic buffers and  completing  a com-
 parison  of  the  economics of  the available  options for  improving  S02 removal
 efficiency.   The  results of  the test work  and cost study are  discussed in this
 paper along with  a description of the FGD  system.  Finally, observations are
 presented from  a  recent Radian sampling trip to document the  systems perfor-
 mance a  year after the end of the Radian/EPRI program.

                       PROGRAM OBJECTIVES  AND SCHEDULE

      The overall  objectives  of the test program were to  improve  both the S02
 removal  efficiency and reliability of the  Duck  Creek FGD system.  The organic
 buffer and  magnesium testing were intended to demonstrate  effective means of
 increasing  the  removal efficiency of the scrubbers.  This  work  gathered  pro-
 cess and chemical data over  a variety of system operating  conditions and pro-
 ceeded in three phases.  The first phase involved gathering baseline data over
 a one week  period to  characterize the performance of the scrubbers before ad-
 ditive testing  began.  The information was collected to  quantify the magnitude
 of the S02  removal problem and to serve as a point of  reference  for data col-
 lected later.

      The organic  acid demonstration test work which  made up phase two began in
 early May,  1982 and  continued for four months until  the  end of  the testing in

-------
August.  Phase three, the magnesium testing, began the first of August  and
concluded four weeks later.

     To accomplish the second objective to increase the reliability of  the FGD
system, work to solve the severe mist eliminator scaling was undertaken in
early May.  This proceeded concurrent with the additive testing and lasted
throughout the 1982 test phase.  Improving limestone reagent utilization was
important in solving the mist eliminator scaling problem.  This work also ran
concurrent with the additive testing and lasted through the program.


                             SYSTEM DESCRIPTION

     The limestone FGD system at Duck Creek was designed to treat the entire
volume of flue gas from the boiler and limit S02 emissions to below 1.2 Ibs
per MMBtu.  At full load operation, approximately 400 MW of power are being
generated resulting in a total flue gas flow of 1.4 million acfm at scrubber
exit conditions (130°F, saturated).  The boilers induced draft fans discharge
into a common plenum to which the inlet ducts of the four modules are con-
nected.  Distribution of the flue gas is determined by the pressure drop
across each module and is not regulated by dampers.   The flow to each may be
adjusted by manually removing rods from or adding rods to any of the seven
decks  inside each tower.  The SC>2 concentration of the 300°F gas as it  enters
the  towers averages about 2500 ppm.  A simplified flow diagram showing  one of
the  four modules and the reagent preparation plant and waste disposal is shown
in Figure 1.  Reagent preparation and waste disposal are common to the  FGD
system.

     Limestone preparation is accomplished with a 10' x 18' wet ball mill which
was  originally designed to grind 40 tons per hour of stone.  Four 6 inch dia-
meter hydroclones are used to control the product size distribution.  Columbia
Quarry supplies the stone which is mined underground from the Kimmswick for-
mation.  Product slurry at about 25 weight percent solids is stored in  a
125,000 gallon tank prior to being fed to the scrubber reaction tanks.

     Limestone feed to each reaction tank is controlled based on a pH feed-
back loop.  The pH setpoint in the reaction tanks can be maintained within
±0.05 units with control of the limestone slurry in an on/off mode.  The
150,000 gallon working volume of the reaction tanks provides the necessary
time for precipitation of the sulfur salts and dissolution of the limestone.
Reaction tank slurry is pumped to the top of the contactors at a rate of
about 15,000 gpm where it passes countercurrent to the flue gas.  At full
load conditions, this results in a liquid-to-gas ratio of slightly more than
40 gals/1000 acf.  Gas-liquid contacting to promote S02 mass transfer is
achieved with the seven rod decks.  Gas side pressure drop across the scrub-
bers is approximately 12 inches H20 at design flow rates.

     A two-stage horizontal mist eliminator (constructed of Hastelloy G) is
located downstream of the last rod deck in each absorber to remove entrained
slurry.  Periodic washing of the mist eliminator removes collected solids.
The majority of the wash water is routed to the disposal pond with a small
amount entering the recycle loop.  Flue gas exits the system without reheat
through a lined wet stack.
                                      5-3

-------
00
i
-p-
   FLUE GAS
 (1.4 x 106acfm
AT 2500 ppm SO2)
       DBA/
   MAGNESIUM"
                                                   HORIZONTAL
                                                 MIST ELIMINATORS
                                                                   PRODUCT
                                            A A A A A
                                                                  __
                                                                  f    FLUE GAS TO
                                                                  (STACK @ 400 ppm so2
                                                                   \
                                                    ABSORBER
                                                                 SCRUBBER
                                                                 RECYCLE
                                                                  SLURRY
                                                                                                   CYCLONES
    I
                                                                        V/   i	^WATER


                                                                         I	1	r-^i
LIMESTONE
 SLURRY
 STORAGE
                                                                                                                BALL MILL
   RAW
LIMESTONE
                                                                                                          - DISPOSAL  -
                                                                                                          ~   POND  -
1
  WATER
                                                                      RECLAIMED WATER
                                                                        TO SCRUBBERS
                                                                       AND ASH SLUICE
                                                     Figure  1.   Duck  Creek  FGD  System

-------
     Solids content in the reaction tank is controlled at  12 weight percent
by adding dilution water reclaimed from the disposal pond.  Waste slurry  is
pumped from each reaction tank to the pond.  This pond also serves as dis-
posal for the boiler bottom and fly ash as well as a final repository for all
other plant wastes.

                            DISCUSSION OF RESULTS

     This section presents the key results of the four month period of
testing in 1982 and the 1983 evaluation trip.  The results of the additive
testing are presented first in the S02 Removal subsection.  This is followed
by results of the system reliability improvement work which are presented in
the Mist Eliminator Scaling and the Reagent Utilization subsections.  The
Cost Study which investigated the economics of the alternative additives  is
discussed next.  Finally, observations made one year following the 1982
testing are presented in the 1983 Evaluation Trip subsection.

SO2 REMOVAL

Baseline Testing

     A characterization of the FGD system was completed at the beginning  of
the program to document the operation and performance of the FGD system, x^ith
particular emphasis on SOa removal.  The testing was done  prior to the ad-
ditive testing to define the baseline characteristics of the scrubbers.   Data
were collected at various reaction tank pH setpoints and over a range of  unit
loads.  The results are plotted so that the effect of pH and gas flow on  the
scrubbers S02 removal can be evaluated.  Figure 2 presents the relationship
observed between reaction tank pH and the S02 removal efficiency of the scrub-
bers.  There  is a  strong correlation between pH and removal up to about
93 percent removal  indicating that liquid phase alkalinity is limiting
the scrubbers removal efficiency-  The data shown in the figure were
taken with the unit operating at or above 90 percent of its maximum to mini-
mize the confounding effect of gas flow on the results.  The figure shows
that controlling at a reaction tank pH of 5.9 results in a removal efficiency
of roughly 85 percent.  This removal efficiency is the approximate level  the
scrubbers are required to operate at to achieve the compliance emission limit
of 1.2 Ibs S02 per million Btu.  As will be discussed in the next two sub-
sections, this high of a pH leads to scaling in the mist eliminators  (ME)  and
poor reagent utilization.

     Figure 3 shows the effect of unit load  (actually gas  flow rate)  on the
stack S02 emission rate.  Above a load of about 320 MW, the emission  rate
begins to exceed the compliance limit.  At a load of 200 MW, the average  S02
removal efficiency approaches 90 percent.  At full load  (400 MW), the removal
drops off to about  75 percent.  The data were taken at an  average pH  setpoint
of 5.7.

Organic Acid Testing

     Pilot scale testing at the EPA-IERL facility, prototype  testing  at the
Shawnee Alkali FGD Test Facility, and the  full  scale demonstration  program at
                                      3-5

-------
     90
o
c
Q)
"o
LU
O
E

-------
     2.0-
     1.5-
 ID
-I— •
GO
CO
o
     1.0-
o
            Duck Creek SO2
            Compliance
            Limit
 o
 05
 H— «
 C/D
0.5-
       0
        200
              250        300        350        400

                      Unit Load, MW
     Figure 3.  Unit load vs. stack S02 emissions - baseline test period
                             3-7

-------
City Utilities Southwest Power Plant documented the feasibility of using
organic acids to enhance the S02 removal efficiency of lime/limestone scrub-
bers.  The organic acids act as buffers when dissolved in the liquid phase  of
the reaction tank slurry.  As buffers, they prevent a large drop  in pH  at the
gas-liquid interface which naturally occurs due to  the absorption of SC>2 •
This improves the mass transfer rate of S02 from  the gas to the  liquid  by re-
ducing the liquid phase resistance.  The higher the dissolved concentration
of  the organic acids, the greater the improvement in mass  transfer and  there-
fore the better the S02 removal efficiency.  The  upper limit  in  improvement
is  set by the gas-liquid contacting efficiency which determines  the  magnitude
of  the gas phase resistance.  Of course, the higher the dissolved buffer con-
centration,  the higher the addition rates and therefore annual costs of the
raw material.  Testing was conducted to determine the concentration of  ad-
ditive required to achieve the desired S02 removal  efficiency.

      To  accomplish this, the organic buffers tested were pumped  directly into
 two of  the  four reaction tanks at measured rates.   The concentration in the
 liquid  phase in each  was measured by an acid-base titration procedure.   At
 the same time,  S02 removal, unit load, and pH data were collected for  cor-
 relation purposes.

      Two types  of buffers were tested.  The majority of the  testing was with
 a material  commonly  referred  to as dibasic acid or DBA, which is a coproduct
 resulting from  the manufacture of  adipic acid.  It  is an aqueous mixture of
 adipic,  glutaric, and succinic acids, all  equally as effective  a buffer in a
 limestone scrubber.   The other  type  of organic additive  tested was composed
 primarily of adipic  acid and  hydroxycaproic acid, a monobasic molecule.
 Testing with this material,  commonly  called acid  water, was  limited due to
 severe reaction tank foaming  caused by a component  in the  mixture.

      Figures 4  and  5  present  the observed  effect  of DBA concentration on S02
 removal for Modules  A and B,  respectively.  All data  shown were  taken at unit
 loads above 360 MW  (90%  of maximum) .  The  data taken while the  reaction tanks
 were being controlled at a pH of 5.4  to  5.7 is separated  from that taken
 during operation  in  the  5.7  to  6.0 range.  For operation  in  the  higher pH
 range,  Module A required approximately 200 ppm DBA  to maintain  removal
 efficiencies above  85 percent.  Less  DBA was needed in Module B   (100 ppm)
 to get 85 percent  removal  in  the higher  pH range.  Data were  collected at
 the lower pH's  for  estimating the  trade-off in limestone versus  DBA consump-
 tion.  For operation in the  lower  pH range, approximately  400 ppm DBA is
 required for Module  A and about  300  ppm  for Module  B  for  85  percent removal.
 The reason for  the  better  performance for  B is unknown  since  liquid recycle
 rate and gas flow measurements  for each  were virtually  identical.

      Because Duck Creek uses  a disposal  pond  instead  of  a  thickener/filtra-
 tion step for dewatering waste scrubber  sludge,  little  of  the buffer added
 is returned to  the  scrubbers.   The pond  has roughly 100  times the volume
 of a typical thickener.   Estimates of the  loss rate of  DBA from  the reaction
 tanks in the waste slurry were compared  with  the  actual  feed rates to deter-
 mine the extent of  DBA losses through chemical  degradation,  coprecipitation
 with the sulfur salts, or in fugitive streams.   The results  showed that

-------
   toon
o
E
CD
CC
c
CD
CD
Q.
    95.
    90
y   85.
    80
    75
•   pH 5.4 to 5.7


•   pH 5.7 to 6.0
          —i	1     i     i	1	1	>     i	'	1	1	1	1	1      i
           100   200   300  400  500  600  700  800  900  1000  1100 1200  1300  1400 1500

                          DBA Concentration (ppm equivalent adipic acid)
   Figure  4.  Effect of DBA concentration on  percent  S02 removal—Module  A
                                           3-9

-------
               100'
                95.
o  90
             o
             CO

             "c
             0)

             a   ss
             0)
             Q_
                 80
                 75
                                                              pH 5.4 to 5.7



                                                              pH 5.7 to 6.0
                   0   100  200  300  400  500  600   700   800  900  1000

                          DBA Concentration (ppm equivalent adipic acid)






Figure  5.  Effect of DBA concentration on percent  SC>2 removal—Module  B

-------
within the accuracy of the measurements all of the material fed could be ac-
counted for in the waste slurry which left the reaction tank.  This  informa-
tion is important because for closed loop FGD systems  (those designed with
thickener/filtration solids dewatering) loss rates over 5 times higher  than
filter cake liquor loss rates have been observed for nearly equivalent  DBA
concentrations.  This difference is probably due to the relatively short
liquid phase residence time in the CILCo system  (10 hours) compared  to  a
closed loop system (weeks).

     Based on these results, annual DBA consumption rates can be estimated
for the Duck Creek FGD system.  Required concentrations for the two modules
not tested (C and D) are assumed to be the same as for Module A.  Assuming
300 days per year online and 12 hours per day operation of the unit  at  or
above 90 percent of full load during which time DBA must be fed, 1.26 x 106
pounds of DBA  (dry basis) are required for a low pH setpoint (5.5) and
630,000 pounds for a high pH setpoint (5,8).  Using a DBA delivered  cost of
29.5 cents per pound, dry basis, the annual cost difference between  the two
is roughly $200,000.  The section on Limestone Utilization discusses the
tradeoffs with annual limestone comsumption.

Magnesium Testing

     Testing to examine the effectiveness of using magnesium to enhance
scrubber S02 removal efficiency was conducted in August, 1982.  Dolomitic
lime was used as the source of magnesium to determine if it was a cost  ef-
fective alternative to magnesium hydroxide.  A detention slaker was  installed
to hydrate the lime (calcium fraction only) prior to its being fed to the
scrubber reaction tank tested.  In the reaction  tank the magnesium oxide
dissolves, increasing the slurry liquid phase magnesium concentration.
Because magnesium sulfite is much more soluble than calcium sulfite, the
liquid phase sulfite concentration increases depending upon how much magne-
sium is present.  At the same time, the magnesium also pairs with the dis-
solved chloride and sulfate ions which reduces its effectiveness.  The  in-
crease in sulfite_improves the liquid phase alkalinity of the slurry by
increasing the SQj species concentration and the MgSOs complex concentration.
Both act as buffers in the same way as the organic acids, promoting  the mass
transfer of 862 from the gas to liquid phase.

     Figure 6 presents the relationship observed between 862 removal and
dissolved magnesium concentration.  No noticeable improvement in S02 removal
was seen below about 1000 ppm of dissolved magnesium.  This is probably due
to the chloride and sulfate association.  Approximately 1500 ppm
of dissolved magnesium was sufficient to maintain Module A at 85
percent S02 removal.  Mass balance calculations verified that all of the
solid magnesium added to the reaction tank dissolved.  The dolomitic lime
appears to be a viable means of adding magnesium to the scrubbers,  therefore.
During the testing, the reaction tank pH was controlled at a relatively high
pH (5.8 to 6.0) to maximize the SOf and MgSOs  complex species concentrations.
Dropping the reaction tank pH to 5.4 would theoretically shift the  sulfite-
bisulfite equilibrium such that about a 30 percent reduction in  concentration
of the S03= and MgSOs  species would result.  Actual test data were not  avail-
able to verify this because of difficulties in operating the slaker  for
                                     8-11

-------
     100  —i
      90 —
  (O
  >
  o

  
-------
extended periods.  However, if accurate, a 30 percent increase in the magnesium
feed rate would be required to offset the reduction.  This would be in ad-
dition to the increase in magnesium required to make up for the loss in scrub-
bing efficiency due to the lower pH operation.  Recall that roughly a two fold
increase in DBA was required in dropping from a pH setpoint of 5.8 down to
5.5.  The reason for this drop in efficiency is a combination of the reduction
in liquid phase alkalinity and of the reduction in solid phase alkalinity
(limestone) due to less excess limestone present in the slurry at lower pH's.
The tradeoff of low pH operation/improved limestone utilization is not favorable
for magnesium as it is for DBA.

     Mass balance calculations also showed that due to the method of solids
disposal at Duck Creek, high feed rates of magnesium would be required until
steady state could be reached in the entire system.  Depending upon the exact
volume of the disposal pond, this could require a number of years.   To im-
prove the economics of magnesium-enhanced scrubbing, recycle of scrubber
blowdown liquor was investigated.  The ability of hydroclone classifiers to
partially dewater the scrubber wastes was tested.  Because of the extremely
fine size of the solids, small diameter classifiers were necessary to make
the desired separation.  Results showed that the slurry could be concentrated
from 10% solids up to about 40% solids, reducing the magnesium consumption by
a factor of 4.  The effect of a recycle step on the operation of the scrub-
bers is unknown and could not be investigated short of major mechanical modi-
fications.  The reduction in DBA consumption would be somewhat less than that
for magnesium since the fraction of organic buffer chemically degrading or
coprecipitating with the sulfur would expect to increase as its residence
time in the absorber loop increased.  The effect of this recycle step on the
economics of these two alternatives is discussed in the Cost Section later.

MIST ELIMINATOR SCALING

     Prior to the test program, mist eliminator (ME) scaling was causing a
severe reliability problem for the Duck Creek FGD system.  In an attempt to
control the scale, each module's mist eliminator was cleaned weekly by a crew
of maintenance men.  Usually the scale growth rate was so fast that within a
few days, solids pluggage would cause a significant increase in the ME gas
pressure drop.  The restriction resulted in a maldistribution of gas flows be-
tween the four absorbers.  Higher gas flows through a recently cleaned tower
would result because of the lower resistance to flow through its ME.  The
higher gas flows would reduce that absorbers SC>2 removal efficiency as well as
carry up greater quantities of mist to the ME.  The increase in removal ef-
ficiency due to the lower gas rates through the restricted towers did not  com-
pletely offset this reduction and overall the removal would drop.  Trouble with
completely isolating the flue gas from the inside  of the towers during cleaning
periods required that some gas bypass the system.  Therefore, the mist elimina-
tor scaling was not only causing reliability problems, but it was also in-
directly causing compliance problems.

     A brief period of time was spent early in the program testing the ef-
fectiveness of adding a gypsum scale inhibitor to  the ME wash water.  Even at
higher than recommended dosage rates, no effect on either the formation rate
or strength of the scale were seen.
                                     8-13

-------
     At the same time, work began to identify the causes of the scaling and
potential solutions for eliminating it.  Regular inspections of the mist
eliminators verified the rapid buildup of solids on  their surface.  Analysis
of the solids showed that they were primarily gypsum with small percentages
of calcium carbonate and calcium sulfite present.  The wash rates were
estimated to be about 1100 pgm with the front and back of  the  first  stage ME
and front side only of the second state being washed.  With the ME's having a
front face area of about 480 ft2, this resulted  in a specific  wash  rate of
about 0.75 gpm/ft2.  Each mist eliminator was washed for 5  minutes  every
twenty minutes.

      During  the plant  outage prior  to  the  test  program each ME was  thoroughly
 cleaned  and  all wash  nozzles and headers were  cleaned or replaced.   Several
weeks into  testing, a significant number of spray  nozzles were requiring re-
 placement each week.   The  nozzles were plugging with scale themselves.   Inspec-
 tion of  the  inside of the  spray headers showed  that  a thick gypsum scale was
 growing  on the  walls.   The pluggage was reducing wash rates and this combined
 with the plugged  nozzles was accelerating  scale formation on the ME's.   The
 wash water which  was  reclaimed disposal pond liquor  was  analyzed for major ca-
 tions and anions  and  the results  input to  the  Radian Inorganic Equilibrium
 Program.  The model predicted  a  gypsum relative saturation of  1.9 which is
 significantly higher  than  the  level at which gypsum scaling usually initiates
 (1.4 relative saturation).   From  this  analysis,  it  was obvious that the re-
 claimed  pond liquor could  no  longer be used for washing  the ME's.

      The only other source of water available was  fresh  water  from  the
 cooling  lake.   Using  it at  a rate of 1100  gpm would  cause  significant over-
 flow of  the  disposal  pond  which was not allowed.  Results  of previous work
 with mist eliminators at the Shawnee Test  Facility  indicated that minimum
 ME wash  rates could be achieved  using  a combination  of fresh water  (or a
 water low in TSS, dissolved calcium, and dissolved  sulfate) ,  good limestone
 utilization, and  sufficient wash  intensity. Based  on the recommended wash
 rates and durations and  on the dimensions  of the Duck Creek ME's, a total of
 only 100 gpm of fresh water would be needed to  keep  them scale free if  the
 limestone utilization was  kept above 85 percent.

      After the necessary piping  changes were made  to switch to fresh water,
 durations and frequencies  were slowly  reduced  to find the minimum required
 to keep the mist eliminators  clean.  To help offset  the  impact on the plant
 water balance, scrubber recycle  pump  seal  water was  closely monitored to
 eliminate excess flush rates  and  hoses used for general  cleaning around the
 plant were  switched from fresh water  to pond reclaim water.  The wash rate  had
 been reduced to  about 300  gpm by the end  of August.   Scaling and pluggage of
  the mist eliminators were virtually eliminated.

      Further optimization  of the washing  was possible, though.  The front of
 the first mist eliminator  was  not being  washed with the recommended intensity.
 More extensive piping and valve changes were required for which there was not
 sufficient  time  during the 1982  testing.    The Shawnee results called for  six
 minutes of washing every four  hours at a specific wash rate of 1.5 gpm per
 ft" of ME face area for the front of the  first ME (nozzle pressure 41 psig).
 The recommendation for the back of the front ME was 4 minutes of wash every
 8 hours at  a specific rate of  0.55 gpm/ft2  (nozzle  pressure 13 psig).  There

-------
was no recommendation for the second bank ME wash, but it was assumed  to be
equal to or less than that for the back of the first bank.  The piping con-
figuration for the Duck Creek fresh water wash resulted in the front and back
of the first ME and the front of the second being washed simultaneously at a
specific rate of 0.7 gpm/ft2 and a nozzle pressure of less than 10 psig.
Later the valving was changed so that the first ME was washed separate from
the second ME.  This increased the pressure available for the front wash but
not up to the recommended level.

LIMESTONE UTILIZATION

     Initial measurements of the limestone utilization in the Duck Creek
scrubbers showed that above a reaction tank pH of 5.4, the utilization dropped
rapidly.   The measurements were made by analyzing a sample of the reaction
tank solids for calcium, sulfite, sulfate, and inert material.  Utilization
is defined as the mole ratio of total sulfur to calcium.   It therefore serves
as a good indicator of how efficiently the scrubbers are using the limestone
fed to the reaction tanks.  Including liquid phase calcium and sulfur
species results in less than one percent change in the reported utilization.
Therefore, all numbers are based on solid phase analysis only.

     The low utilizations resulting from operation above pH 5.4 were hurting
the FGD systems operation and performance in the following ways:

     •    reliability — excess limestone in the slurry contributes
          to mist eliminator scaling,

     •    operating costs — excess limestone could substantially in-
          crease annual limestone costs, and

     •    waste disposal — presence of the excess limestone in the
          scrubber blowdown was consuming a significant volume of
          the disposal pond.

     There were two options available for increasing the limestone utiliza-
tion.  First, the reaction tank pH could be controlled at or below 5.4.  To
do so though, would lower the SOz removal efficiency of the scrubbers sig-
nificantly (see Figure 2) and require higher DBA or magnesium concentrations
to make up the difference.  The second option would be to modify the ball
mill circuit operation to produce a finer and therefore faster dissolving
limestone.

     During the baseline testing, the product from the grinding circuit was
sieved with a 200 mesh and a 325 mesh screen.  The mill circuit was original-
ly designed to grind 40 tons per hour of pebble limestone and produce  a
material of which 90 percent would pass a 200 mesh screen.  Sieve results
during baseline testing showed that on the average 86 percent passed the 200
mesh screen and 70 percent passed the 325 mesh screen.  Review of past sieving
data showed that the product varied anywhere from 75 percent passing 200 mesh
to 90 percent passing 200 mesh (no data on the percentage passing 325  mesh) .

     The first and most obvious change examined to improve the grind was to
decrease the limestone throughput.  Based on calculated 862 absorption rates,
                                      3-15

-------
the mill throughput could be decreased from 40 tons per hour to 20 tons per
hour and enough reagent be prepared to keep up with the scrubber demand.  After
several days of operation at the lower throughput, no measurable change was
seen in the product particle size.  This suggested that the classifiers were
limiting the product fineness.

     A series of operational and minor equipment  changes were  made during June
and July with the classifiers and the mill which  resulting  in  a significant  im-
provement in the product particle size distribution.  The  testing  was not de-
signed as a parametric investigation but instead  it was conducted  as an opti-
mization program.  The objective was to produce the finest  particle  size with
only relatively small changes to the existing circuit.  The  result was a mill
circuit  that consistently produced a material that greater  than 99 percent
passed 200 mesh and  90 percent passed 325 mesh.

     Although quantification of the effect of each change on the produce
particle size  is not possible, the following variables were changed  in the
mill/classifier circuit  optimization:

      «    mill throughput,

      •    mill slurry solids content,

      •     level of ball  charge in mill,

      •     top-size of balls  in mill charge,

      •     raw  limestone  feed  size,

      «     diameter of  classifier barrel,

      •     classifier pressure drop  (or  throughput),

      •     solids  content in  classifier  feed, and

      •     diameter  of  classifier vortex finder  and apex.

      The most  extensive  equipment  change made  during  the  program was to re-
 place the 10 inch  diameter  classifiers  with  six  inch  diameter  models.  De-
 creasing the diameter  of the  cyclone barrel  theoretically would reduce the
 size of the largest  particle in  the product  stream by as  much  as 10  to 30
 microns.  For comparison,  the 200  mesh screen  has square  openings  that are
 74 microns in dimension and the  325 mesh screen has  openings that  are 45
 microns square.

      The effect  of  limestone particle  size  on  limestone utilization in the
 scrubbers was  significant.   Utilization data taken during the  grinding cir-
 cuit testing was  separated  into  three  categories  according to  limestone
 particle size.   Figure  7 presents  the  results.   The  data  show that as scrub-
 ber pH increases,  limestone utilization drops  off dramatically.  It also shows
 that at a constant pH,  decreasing  limestone  particle  size improved limestone
 utilization significantly.   For  example,  at  pH 5.8,  the utilization averaged
 about 58% with the coarse limestone being  fed  to  the  scrubbers, while it

-------
        90
        80-
     c
     o
     (O
     N
     *  70
     o
     0)
     CD
     E
        60.
        50-
        40.
• Coarse Grind

A Medium Grind

• Fine Grind

d = DBA present

m - magnesium present
               5.0
          5.2        5.4
   5.6        5.8
Scrubber pH
                                                               6.0
—i
 6.2
Figure  7.   Effect  of  limestone  grind and  scrubber pH  on limestone utilization
                                         8-17

-------
averaged about 92% with the fine limestone.  Representative particle  size
distributions for the three grinds are shown in Table 1.  The scatter in  the
data presented in Figure 7 is primarily due to the variation  in  the particle
size distribution of the limestone within  these three categories.

     Another important observation made is that the  presence  of  either DBA or
magnesium in the reaction tank had no effect on measured  limestone
utilizations.  Samples taken during either DBA or Mg testing  are identified
in the figure.

            TABLE 1.  REPRESENTATIVE PARTICLE SIZE DISTRIBUTIONS
                     FOR COARSE, MEDIUM, AND FINE GRINDS	
                                           Particle  Size  Distribution

 Limestone  Size                     % Passing 200 Mesh   %  Passing 325 Mesh

     Coarse                                 86                   73
     Medium                                94                   81
     Fine                                   97                   87
      As  discussed  earlier,  the  improvement  in  limestone utilization was
 achieved by relatively minor  operational  and equipment changes.   The primary
 costs associated with these changes  include the  capital for the  six inch dia-
 meter classifiers  which  totaled less than $10,000  installed,  an  additional
 $40,000  per year for electrical power  to  operate the  mill at  lower through-
 puts, and an increase in maintenance costs  which can  be determined only by
 time.  The benefits by comparison  are  much  more  significant.   The Duck Creek
 scrubbers are now  being  operated at  a  pH  of 5.8.  By  controlling at this set
 point the scrubbers not  only  operate at a higher S02  removal  efficiency
 (nearly 82%) compared to previous  operation at pH 5.6 but also achieve
 greater than 90%  limestone  utilization.   In terms  of  limestone consumption,
 the Duck Creek scrubbers will use  almost  30 million pounds per year less than
 before.   And since this  limestone  was  passing  through the scrubbers unused,
 the waste disposal pond  will  see an  equivalent reduction in material sent to
 it, thereby extending its useful life. Over  a 25-year operating period this
 would mean a reduction in volume of  solid wastes of almost 350,000 cubic
 yards.  This is roughly  equivalent to  the settled volume of FGD  waste pro-
 duced by the scrubbers in a 5 year period.

 COST STUDY

      Based on the results of  the 1982 test  program, a cost study was com-
 pleted on the economics  associated with  using  DBA and magnesium to improve
 the Duck Creek scrubbers S02  removal efficiency.  A total of six options were
 studied-three each for DBA and magnesium.  They  investigated the difference
 between automatic and manual control of  additive feed as well as the effect
 of recycling scrubber  blowdown to  reduce  makeup  rates.  To simplify the work,
 a constant reaction tank pH of 5.8 was assumed.

      Capital and annual  costs were estimated  for each case.  Based on  these
 costs, a present worth  analysis of the various options was performed using
 the following equation:
                                     3-18

-------
where S is the cost for any year  n and I equals  the discount rate.   For  this
study, CILCo's -long term escalation factor of 8 percent was used  to  escalate
both capital and operating expenses.  CILCo's long  term discount  rate  of
10.75 percent was chosen for I.

     The results of the present worth analysis which is based on  a fifteen
year period are shown in Table 2.  Also shown in  the table are  the capital
investments for each option.  The numbers indicate  that without the  hydro-
clones to recycle some of the reaction tank blowdown liquor, the  magnesium
options are not competitive.  The reason for this is that much  higher  liquid
phase magnesium concentrations are required than  for DBA.  Therefore the
losses in the waste stream are proportionally higher.

     The lowest cost option involves the use of DBA in conjunction with an
automatic feed controller and hydroclones for recycle.  Both Case 2  and 6
follow next and have about the same present worth cost over the 15 year
period at $2.8 and $2.4 million each.  Case 2 is  the DBA option with auto-
matic feed control.  The automatic system consists  of a microprocessor that
receives a signal from the stack gas SC"2 monitor.   The microprocessor  con-
trols the feed of DBA based on the S02 emission level.  As the  emissions
approach the 1.2 lb/105 Btu standard the microprocessor would operate  a valve
to feed DBA to the reaction tanks.  As the 862 emission level drops  below a
predetermined value, the microprocessor shuts off the feed.  This system
would operate the same for the magnesium option.  The constant  DBA addition
option, Case 1, follows at a cost of $3.0 million.

     From an operation and control standpoint, the  constant DBA addition
option, Case 1, is the easiest to operate.  Addition of an automatic con-
troller complicates the system to a small degree  but would pay  itself  off in
a short time (approximately two years).  Addition of hydroclones  would in-
crease the mechanical complexity of the system.   Because of the large number
of hydroclones which would be required, operating and maintenance related
problems would increase as well.  Operation of a  slaker to prepare  the dolo-
mitic lime prior to addition to the reaction tanks  makes the magnesium auto-
matic feed/hydroclone recycle option the most mechanically complex  system.
The operation of a slaker requires close operator supervision.  For  this
reason, purchase of a hydrated lime would be advantageous.  With  this
approach, the material could be fed dry to the reaction tanks or  could be
mixed with water and pumped as a slurry.  Nonetheless, bulk solids  handling
is inherently more difficult than handling small  volumes of an  aqueous buffer
solution making DBA more attractive from an ease  of operation standpoint.

EVALUATION TRIP

     A return trip was made to Duck Creek approximately one year  following  the
'•ompletion of the 1982 test program.  The objective of the  trip was  to document
the performance of the FGD system for comparison  to the observations made
during the 1982 testing.  The results were used to  evaluate the ability  of  the
plant personnel to maintain the recommended operating setpoints.  They also
served as a basis for recommending any changes in controlling the scrubbers
TO further improve their operation.


                                     8-19

-------
                    TABLE 2.   COST ANALYSIS OF DUCK CREEK S02  REMOVAL ENHANCEMENT OPTIONS^
No.
1
2
3
4
5
6
00
Case Description
Constant DBA Addition
DBA Feed Controller
DBA Feed Controller Plus Hydroclones
Constant Magnesium Addition
Magnesium Feed Controller
Magnesium Feed Controller Plus Hydroclones

Thousands of
Capital Investment
110
150
240
140
180
270

Dollars (January 1983)
Cumulative Present Worth3
3,000
2,800
1,400
10,800
10,000
2,400

ho
o
*3
 Escalation factor of 8% was used for both fuel and non-fuel related  items.
 Discount rate assumed was 10.75%.

-------
     There were several significant observations:

     1)    The scrubbers were needing less DBA than anticipated
          to maintain compliance level S02 removal ef-
          ficiencies .

     2)    No mist eliminator scaling or pluggage had been
          seen in the past year.

     3)    Limestone utilization was not as high at a given
          pH as had been seen in 1982.

     4)    FGD system reliability was 96 percent over the period
          July, 1982,  through June, 1983.

     5)    Limestone product screen analyses consistently
          showed better than 90 percent passing 325 mesh.

     6)    A five ton per hour increase in limestone feed to the
          ball mill had little effect on product size distribution
          and no noticeable effect on limestone utilization.

     The DBA usage and limestone utilization observations were the most
interesting.  The tradeoffs between low pH/high DBA and high pH/low DBA
operation were reviewed.  Based on the data collected during the trip, it
would be to CILCo's advantage to lower the reaction tank pH setpoints to 5.6
and increase the DBA feedrate.  This is because the savings in limestone due
to improved utilization at the lower pH more than offsets the increased DBA
consumption.  Operation below pH 5.6 was not recommended because the pH/
utilization relationship begins to flatten below that point (see Figure 7).

     The screen analyses showed that the mill circuit was consistently pro-
ducing a product with over 90 percent passing 325 mesh.  Mill throughput was
increased from 20 tons per hour to 25 tons per hour to see if this affected
the product fineness or limestone utilization in the scrubbers.  The amount
passing a 325 mesh screen dropped from 92 percent down to 89 percent.  The
affect on utilization could not be measured if there was any.  The resulting
25 percent decrease in mill operating time not only will save on power costs,
but should also reduce equipment maintenance.

     In general, the control of the FGD systems key setpoints was very good.
As mentioned, the limestone mill circuit product fineness could be consis-
tently maintained depending on throughput.  Reaction tank pH's were also
being controlled within a very small range with good accuracy.  Overall, this
kept limestone utilization as constant as possible with a single loop  (reac-
tion tank) design.  Measurements of the solids content in the reaction tanks
indicated that the density control system was generally working well.  Almost
all the measurements fell between 10 weight percent and 13 weight percent
with the average being 12 percent solids.  Partly as a result of this, the
FGD system reliability has averaged slightly above 96 percent since July of
1982 and over 99 percent since January of 1983.  At the same time, FGD sys-
tem operating and maintenance costs have dropped about 30 percent.
                                    5-21

-------
 TECHNICAL/ECONOMIC FEASIBILITY STUDIES FOR FULL
SCALE APPLICATION  OF  ORGANIC ACID TECHNOLOGY FOR
             LIMESTONE FGD SYSTEMS

         J. C. Dickerman, J.  D. Mobley

-------
      TECHNICAL/ECONOMIC FEASIBILITY STUDIES FOR FULL SCALE APPLICATION
            OF ORGANIC ACID TECHNOLOGY FOR LIMESTONE FGD SYSTEMS

      by:   James C.  Dickerman
           Radian Corporation
           3024 Pickett Road
           Durham, NC  27705

           J.  David Mobley
           Industrial Environmental Research Laboratory
           U.  S. Environmental Protection Agency
           Research Triangle Park, NC  27711

                                  ABSTRACT

     The application of organic-acid buffer enhancement to flue gas
desulfurization (FGD) systems is a recent development that has resulted in
lowered costs and improved performance for those systems that have adopted
its use.  A process which uses organic acids as an additive has several
advantages over conventional limestone scrubbing systems.   These advantages
include improved SO. removal, decreased limestone consumption, increased
system flexibility fe.g., ability to respond to unplanned fluctuations in
coal sulfur content), and improved process reliability.  This paper
summarizes the results of several cost analyses which were performed to
evaluate the potential economic benefits of converting operating FGD systems
to organic-acid-enhanced limestone scrubbing systems.  Also, since the last
FGD symposium, two full-scale utility limestone scrubbing systems have
converted to organic acid enhanced operations.  A summary of the first year
of operation for one of those systems—City Utilities Southwest Power Plant
(SWPP)—is also included.

                                 BACKGROUND

     The ability of a limestone FGD system to remove SO- is limited by the
absorptive capacity of the slurry liquid that makes contact with the flue
gas.  Adding an organic acid to the system expands the capacity of the
liquid to absorb S0_ by buffering the pH in the absorber.  These are the
reactions that take place:

     Sulfur dioxide + Water  «-  Acidity + Sulfites                       (1)

     Organic buffer anion (carboxylate ion) + Acidity  +  Organic acid   (2)
                                     8-23

-------
As the buffer reacts with the acidity formed in the SO  absorption reaction,
a weak organic acid is formed (Reaction 2) and SO  is removed (Reaction 1).
The buffer is regenerated in the reaction tank by adding limestone:

     Organic acid + Sulfites + Limestone  •*•  Organic buffer anion
          + Calcium sulfite solids + Carbon dioxide                     (3)

Adding organic buffers can result in increased S02 removal, a decreased risk
of chemical scaling, potentially decreased limestone consumption,  and
potentially reduced mist eliminator fouling because of increased utilization
of limestone.

     When EPA realized that organic buffers could possibly to used to  solve
some of the maintenance and cost problems that utilities were experiencing
with their lime/limestone FGD systems, the Agency began an extensive
research program, designed to provide a thorough evaluation of various
organic acid additives and how they can be used to improve lime/limestone
scrubbing.   It  consisted of four phases—laboratory-, pilot-, prototype-,
and  commercial-scale.

     The EPA began  its research by sponsoring theoretical and laboratory
 investigations  of a number of different acids.  As a result of these
 studies, adipic  acid appeared to be the most promising buffer candidate
because of its  solubility in water, low volatility, chemical stability,
non-toxicity, high  availability, and low cost.  Also, the region of maximum
buffering  for adipic acid, between pH 4 and 6, is ideal for limestone  FGD
 systems.

      Two full-scale test programs were also sponsored by EPA which
 demonstrated  the technical and economic viability of this approach on
 operating  utility and industrial FGD systems.  Results of both of  these
 demonstration programs have been reported in previous FGD symposia.

             TECHNICAL AND ECONOMIC ANALYSES OF OPERATING SYSTEMS

      Technical  and  economic analyses were made for four FGD installations  to
 determine  the nature and cost of converting each to an organic acid enhanced
 limestone  scrubbing system.  The four installations are:  City Utilities'
 Southwest  Power Plant, Central Illinois Light Company's Duck Creek Station,
 San Miguel Electric Cooperative's San Miguel Station, and Big Rivers
 Electric  Corporation's R. D. Green Station.  The cost analysis of  the
 R.  D.  Green  Station was not completed in time for this paper.  The results
 will,  however,  be presented at the FGD Symposium.

      The  approach used in conducting these process evaluations was basically
 the same  for all systems.  First, typical process operating conditions for
 each system  were identified through a series of questionnaires and meetings
 with plant personnel.  Operating parameters such as pH, limestone
 utilization,  liquid to gas  (L/G) ratios, and SO  removal were obtained from
                                   3-24

-------
each system to establish baseline process operations and baseline operating
costs.  Next, technical evaluations were made to identify any process
modifications that would be required to convert each system to an organic
acid enhanced system, and to establish organic acid makeup requirements.
Finally, the capital investment costs for all required process
modifications, and the annual operating costs of organic acid enhanced
operations were determined.  The analysis results for each system are
summarized below.

CITY UTILITIES—SOUTHWEST POWER PLANT (1)

     City Utilities' (CU) major interest in converting their system to an
organic-acid-enhanced limestone scrubbing system was to improve the S0~
removal performance so that they could comply with the applicable S09
removal requirements.  Several process alternatives were considered, each of
which had the potential for achieving the utility's goal of increased S0?
removal.  The process alternatives considered included:  increasing the L/G
ratio to the tray tower, adding adipic acid, and adding a mixture of
by-product organic dibasic acids (DBA).  The DBA used at CU is a by-product
of adipic acid production and consists of a mixture of adipic, glutanic, and
succinic acids.  The higher L/G option represented a capital cost intensive
solution; whereas, the adipic acid and DBA options have lower capital costs
but higher annual operating costs.  A present worth analysis was performed
using escalation factors consistent with CU's long range planning
activities.  The escalation of adipic acid and DBA closely approximated CU's
assumed escalation rate of fuel oil, and the escalation of cost of
electricity (production costs only) closely approximated CU's assumed coal
escalation rate.  The discount factor for the analysis was assumed to be
equal to the general inflation rate.

     The results of this study are shown in Figure 1.  Note that the adipic
acid option is a lower cost option than increased L/G for about 6 years.
The technical uncertainties in making the L/G modification, coupled with the
demonstrated adipic acid system flexibility, make the adipic acid option a
good choice.  However, the DBA option remains the economic choice through
more than 15 years.  Consequently, City Utilities elected to convert their
FGD system to DBA enhanced operations in December 1981.  Results of the
first year of operation of that system are reported later in this paper.

CENTRAL ILLINOIS LIGHT COMPANY—DUCK CREEK STATION (2)

     Central Illinois Light Company's (CILCo's) main interest in considering
a conversion to an organic-acid-enhanced limestone scrubbing system was also
related to improving the system's performance so that it would achieve
compliance with the applicable SO,., regulations.  As with City Utilities,
several process alternatives were considered.  The options identified for
CILCo to achieve compliance included:
                                   3-25

-------
        constant  DBA addition,

        intermittent DBA feed,

        intermittent DBA plus recycle,

        constant  magnesium oxide addition,

        intermittent magnesium oxide  feed,  and

        intermittent magnesium oxide  feed plus recycle.
                                                           Adlplc Add
                                                            Addition
                                                             Increased
                                                             LK3 Ratio
                                                               DBA
                                                             Addition
                                    Time (Years)

               Figure 1.  Cumulative Present Cost for Southwest Power Station
     The CILCo system is significantly  different than the other  systems
evaluated in that a waste disposal pond is  used for solids disposal.   In
this system, a slurry waste stream of approximately 10 percent solids is
sent to the disposal pond, thereby significantly increasing the  DBA or
magnesium makeup requirements.  For  this reason, a process alternative,
which included the addition of a hydroclone to concentrate the blowdown
stream to 40 weight percent solids and  thus reduce the additive  losses, was
also examined.
                                     5-26

-------
     Results  of the present worth analysis over a  15-year period are shown
in Figure  2.   These results show that the DBA options  cost significantly
less than  the magnesium addition options, except for the case in which a
hydroclone is used to reduce  the amount of scrubber blowdown sent to the
waste disposal pond.  Although  the process options using the hydroclone
showed  cost advantages, the use of a hydroclone has not  been demonstrated in
the field,  and thus, technical  uncertainties associated  with its use exist.
For this reason, CILCo has elected to convert their process to a
DBA-enhanced  limestone scrubbing system with intermittent feed controls so
that DBA is added only during high load operations.
       12 -i
        11 -
        10 -
        9 -
        7 -
        s -
         1 -
]= Constant DBA Addition
]= Intermittent DBA Addition
]= Intermittent DBA Addition Plus Recycle
]= Constant Magnesium Addition
]= Intermittent Magnesium Addition
]= Intermittent Magnesium Addition Plus Recycle
                                                                         15
          Figure 2.  Cumulative Present Worth of Converting ClLCo's Duck Creek Station
SAN MIGUEL  ELECTRIC COOPERATIVE (3)

     San Miguel's interest  in  converting their  system to an organic-acid-
enhanced system stemmed primarily from a desire  to  reduce their FGD
operating costs rather than from regulatory pressures.   Normal operating
conditions  at San Miguel required the use of fairly high limestone reagent
                                      3-27

-------
ratios in order to meet compliance.   Reagent ratios on the order to 1.2 to
1.3 are routine,  and as high as 1.4  have been measured.  Besides the obvious
costs associated with excess limestone use, San Miguel also experienced
chemical scaling in their absorber tower which was probably related, at
least in part,  to the excess limestone present in the system.

     This paper presents the results of a cost analysis to determine the
potential cost  savings associated with adding DBA to achieve an SCL removal
level of 85 percent.  Due to the uncertainties in the DBA feed requirement
and the current limestone reagent ratio at San Miguel, a range of annual
operating and maintenance (O&M) expenses was used in the cost analysis.
These expenses  reflect the incremental change in O&M expenses from the
current FGD operation.  Maximum O&M expenses correspond to the maximum
estimated DBA requirement (400 ppm)  and the minimum limestone savings (which
result if San Miguel's current limestone reagent ratio is 1.2).  Minimum O&M
expenses correspond to the minimum estimated DBA requirement (300 ppm) and
the maximum limestone savings (which occur if San Miguel's current limestone
reagent ratio is 1.3).  Average O&M expenses assume a DBA feed requirement
of 350 ppm and a current limestone reagent ratio of 1.25.

     A present worth analysis was performed based upon San Miguel's long
term escalation and market interest  rate assumptions and is presented in
Figure 3.  Taking into account the initial capital investment for the DBA
system, the cost savings for the minimum and average O&M expenses were
estimated over a 15-year period at $1.5 and $0.4 million in terms of present
worth.  Payback periods were calculated to be 1.2 and 2.5 years from the the
beginning of 1983, respectively.  Similarly, if the maximum operating
expenses are realized, the DBA system will cost San Miguel $657,000 over the
15-year present worth period.

     Note that no credit for increased reliability was assumed in any of the
cost calculations.  For San Miguel,  which has experienced chemical scaling
problems, the savings could be very significant.  Worst case assumptions
predict a cumulative cost for a DBA system of $657,000 over a 15-year
period.  If DBA allows San Miguel to remain on line for only 2 extra days
during the 15 years, it will have saved the total present worth cost of the
DBA  system for the entire 15-year period.

     Because of the potential cost savings and system performance
improvements associated with the use of a DBA system, San Miguel has decided
to conduct a test program to verify the results of this engineering
evaluation.  Field testing of DBA-enhanced limestone scrubbing operations
is to be conducted in late 1983.

-------
            1000 -
             500 -
       I
        5
        1
        I
            -500 -
-1000 -
            -1500
            -2000 -
             Figure 3. Cumulative Present Worth of Converting San MlgueTs FGD
                    System to DBA Assisted Operations
           RESULTS OF  THE  FIRST YEAR OF COMMERCIAL OPERATION OF AN
                       ORGANIC-ACID-ENHANCED FGD SYSTEM

     In 1980, Radian Corporation and City Utilities entered into an
agreement to participate in  a  demonstration project which was sponsored by
U.S. EPA's Industrial  Environmental Research Laboratory at Research Triangle
Park, NC.  Adipic acid and dibasic  acid (DBA) were tested as buffering
agents in the Southwest Unit 1  limestone scrubber to determine their
effectiveness in increasing  utilization and S0_ removal.  This testing was
conducted during 1981  and  verified  the effectiveness of organic acid buffers
in a commercial limestone  scrubber.

     It was quite important  to  CU officials that improved scrubber operating
efficiencies be obtained immediately.   In late 1980, Region 7 of the
U.S. EPA cited the operation of Southwest Power Station Unit 1 for failure
to comply with flue gas emission regulations.  By October 1981, upon

-------
completion of Radian's testing, negotiations were ongoing with EPA in an
attempt to resolve the dispute.  In efforts to reach an agreement on a
"Consent Decree" City Utilities agreed to use a material, such as adipic or
dibasic acid, as an additive to enhance the reactivity of the CaCC-  and thus
increase the SO  removal from the flue gas.  City Utilities embarked
immediately on a plan to implement the use of DBA (4).  This section of the
paper describes the results of the first year of DBA operations.

DBA FEED SYSTEM

     A temporary DBA feed system was installed in late December 1981 to add
the DBA solution on a continuous basis.  This temporary system used a
26,000 I  (6,000 gal) stainless steel tanker trailer equipped with thermal
jacketing as a storage vessel.  Two residential-sized electrical water
heaters provided a source of hot water at -\> 60°C (140°F) for circulation
through the  thermal jacket of the truck.  The heater return lines were run
in parallel  with the DBA feed line through a 10 cm (4-in.) PVC conduit in
order  to  prevent feed line cooling.  Maintaining high DBA temperatures was a
major  concern because DBA tends to easily precipitate out of aqueous
solution.  As an example, a 20 percent solution of DBA will begin to show
crystal formation at 22°C (72°F); higher concentrations require even higher
temperatures to maintain solubility.  The entire temporary system was
assembled on-site and tied into the scrubber system in a matter of only a
few days.

     Initially, DBA was pumped from the tank truck to the ball mill sump.
The feed  rate was set manually.  After the DBA was mixed with the freshly
ground limestone slurry from the ball mills, the mixture was pumped to the
limestone storage tank.  Since the limestone handling equipment was common
to both scrubber modules, both scrubbers operated with the same DBA
concentration.  Late in September, the DBA feed system was modified to allow
DBA to be fed separately into the reaction tank of either scrubber. (5)

S02 REMOVAL

     Results of a year's worth of testing using DBA as an FGD additive were
consistent with those anticipated based on the demonstration project
conducted in 1981.  On the whole, FGD performance was enhanced to such an
extent that  Southwest Power Station consistently operated in compliance with
the 1.2 Ib S02/10  Btu (516 ng/J) federal emission limitation on a
day-to-day basis.  The few emission exceedances experienced were generally
associated with start-ups, shutdowns, and other periods of allowable
operational  curtailment.
                                    8-30

-------
     Figure 4  displays a monthly history of FGD performance,  expressed as
overall SCL removal  efficiency, for the 1982 calendar  year.   It can be seen
that efficiencies were consistently greater than  the approximately
82 percent removal generally required at the Southwest Station to maintain
compliance.  For comparative purposes, Figure 4 also displays the
performance history  for calendar year 1980, wherein removal  efficiencies
were erratic and averaged only 26.0 percent.  Figure 5 shows  these data in
terms of average monthly emission rates.  This figure  illustrates the
significant improvement in the ability of the system to comply with the
regulatory emission  standard after the conversion to a DBA enhanced
system. (6)
             100


             90


             BO


             70


             60-


             50-


             40 H


             30


             20


             10
                       Note: Expressed as % Removal Over Alt Hours In the Period.
                                   1982
/\
                                                               1980
                  Jan  Feb  Mar  Apr  May  Jun  Jul   Aug  Sep  Ocl  Nov  Dec
            Figure 4.  Comparison of Southwest Power Station SO2 Removal Efficiencies,
                    1980 vs. 1982
     The addition of DBA also improved the  system's ability to respond  to
various operational changes.  During a 2-week  period in March, the coal
supplied to  the  system contained 4.5 percent sulfur instead of the
3.5 percent  sulfur level for which the system  was designed.  During this
period, the  unit was able to maintain an  in-compliance status by increasing
the amount of DBA feed to the system. (7)
                                     3-31

-------
7.0

S 6.0 •
•o
^
fl 5.0 •

•
8. 4.0-
1 3.0-
1
r« „ ..
O 2.0
en
1.0
/ 1980
x"
^"\
-' \
\ X
\ /»x x x ^x
\ / ^*v ^* x** — •*'^
\ / "
N-y





^^^ ~--^__— 19B2


- 3010

- 2580


• Z150

- 1720
• 1290

• B60


- 430
                Jan  F«b  Mar  Apr  May  Jun  Jul   Aug S«p  Ort  Nov  Dec
               Figure 5.  Comparison of Southwest Power Station Monthly Average
                       S02 Emission Rates, 1980 v». 1982
RELIABILITY

     The overall reliability of the  system was also increased significantly
in 1982 in comparison with other years.   Figure 6 shows the monthly  average
FGD reliability for 1982 compared with  the average reliability for 1979  and
1980. (8)  This reliability increase may  have  resulted in part from  DBA
addition, but was also related to a  number of  process improvements made
during 1981 and 1982, such as the conversion from turbulent contact
absorption towers to tray towers.

CONTINUOUS MONITORS

     The performance of the continuous  emission monitoring system  (GEMS)
used by CU to determine their S0_ emissions was also evaluated to determine
the ability of the GEMS to gather the type of  data required by the 1979
revision to the utility New Source Performance Standard.  To meet EPA
requirements for data capture, at least two data points must be taken each
hour for a minimum of 18 hours each  day for 22 days per month.  The  goal of
the CEMS at SWPP was to collect at least  four  data points per hour for
24 hours.

-------
£
3
i
I
8,
E
I
                100
                40  -
                20
                        o
                        1~

                        o
                                c
                                                                  1982
                                                       Average 1979-1980
                       Jan
                           Feb  Mar  Apr  May  Jun  Jul   Aug  Sep  O=t  Nov  Dec
              Figure 6.  Comparison of Southwest Power Station Process Reliability,
                      1982 vs. Average Reliability for 1979-1980
     Monitor  reliability data  were collected  during the last 6 months  of
1982.  During this time, the GEMS proved to be very reliable, capturing in
excess of  97  percent of the available data for each scrubber module.
Table 1 presents the data  capture percentage  for  each module by month  and
for the program as a whole.  The lowest data  capture efficiency was
94 percent during July on  the  S-l Module.  The majority of the downtime
encountered by the GEMS was caused by power interruptions. (9)

              TABLE 1.  SUMMARY OF CONTINUOUS  MONITOR RELIABILITY
                Average
                                     Module S-l
                             97.3
                                                    Module  S-2
June
July
August
September
October
November
December
98.9
94.0
95.9
99.0
100.0
97.4
96.2
100.0
96.8
96.1
98.4
99.9
97.6
100.0
98.4
                                      3-33

-------
DBA CONSUMPTION

     The DBA consumption rate increased significantly during the last half
of 1982 to where it was nearly twice that observed during the 1981
demonstration tests.  The increased DBA consumption was due to two main
factors:  changes in the recycle slurry pH and deterioration of the  slurry
recycle pumps.  Slurry pH was decreased in 1982 from 5.4 to 5.1 to improve
limestone utilization in an attempt to improve mist eliminator reliability.
(Mist eliminator performance has been greatly improved by this technique and
by revising the system water balance to wash with greater amounts of fresh
water.)  While the lower pH does improve limestone utilization, a higher
concentration of DBA is required to achieve the desired SO- removal. (10)

     The deterioration in the recycle slurry pumps over time resulted in an
actual L/G ratio less than indicated.  In order to maintain the same SO^
removal at a decreased L/G, the concentration of DBA was increased.  The
pumps underwent a complete overhaul at the plant's scheduled outage  in early
1983 to correct this problem.

FUTURE ACTIVITIES

     Several operational improvements were implemented during the 1983
extensive maintenance outage.  An example is a limestone classification
system and a permanent DBA addition system.  Based on the experiences of the
Duck Creek Station of the Central Illinois Light Company in their DBA
program, it was found that by improving the efficiency of the limestone
classification system they could reduce DBA consumption on the order of
magnitude of  10 percent.  During a short period of classifier testing during
November 1982, similar, yet inclusive, results on DBA usage were obtained by
CU.

     With the improved operator control over the DBA feed system and the
direct  feed to each hold tank, a much quicker chemical response will be
obtained which should allow consistent S0? emissions rates at or slightly
below  the required  1.2 lb/10  Btu (516 ng/J) level.  An increased reduction
of  DBA usage  is projected as the operator exercises more precise system
control.  When considering the beneficial effects of the classifier  system
as  well, overall DBA consumption should be reduced significantly as  compared
to  1982 values.

-------
                                 REFERENCES

1.    Hargrove,  0.  W. ,  J.  D.  Colley,  and J.  D.  Mobley.   Adipic Acid-Enhanced
     Limestone  Flue Gas Desulfurization System Commercial Demonstration.
     Paper presented at APCA/ASME Information Exchange, Research Triangle
     Park, North Carolina.   December 8-9, 1981.

2.    Colley,  J. D., P. A. Nassos, and S. T. Litherland.  Field Investigation
     of FGD Chemistry.  Radian Corporation, Austin,  Texas.   Draft EPRI
     Report.  March 1983.

3.    Glover,  R. L.  A Cost Analysis  of the  Conversion of San Miguel Electric
     Cooperative's Limestone FGD System to  Utilize Adipic Acid.   EPA
     Contract No.  68-02-3171, Task No. 56.   Radian Corporation,  Austin,
     Texas.  Draft Report.   October  1982.

4.    Hicks, N.  D.  and D.  M.  Fraley.   Addition of Adipic and Dibasic Acids  to
     a Conventional Flue Gas Scrubber:  Costs Operating and Design
     Experiences.   Paper presented at Twenty-Seventh Annual APPA Engineering
     and Operations Workshops.  San  Antonio, Texas.   February 14 - 17,  1983.

5.    Brown, G.  E., J.  C.  Dickerman,  and 0.  W.  Hargrove.  Results of the
     First Year of Commercial Operation of  an Organic Acid Enhanced FGD
     System.   EPA Contract No. 68-02-3171,  Task No.  53.  Radian Corporation,
     Austin,  Texas.  Draft Report.  May 1983.

6.    Reference  4.

7.    Reference  5.

8.    McMahan, J. L.  City Utilities  Quarterly Compliance Status Reports.
     Submitted  to Missouri Department of Natural Resources.  Calendar
     Quarters 1 - 4, 1982.

9.    Reference  5.

10.  Reference  5.
                                    8-35

-------
SESSION 9:   WASTE DISPOSAL/UTILIZATION

Chairman:   James D.  Kilgroe
           Industrial Environmental Research Laboratory
           U.S. Environmental Protection Agency
           Research Triangle Park,  NC

-------
    FULL-SCALE FIELD EVALUATION OF WASTE DISPOSAL
     FROM COAL FIRED  ELECTRIC GENERATING PLANTS

      J.  W.  Jones, C.  J. Santhanam, A.  Balasco,
I.  Bodek, C. B. Cooper, J.  T. Humphrey, B. K. Thacker

-------
                                ABSTRACT
This paper summarizes results of a 3-year study of current coal ash
and flue gas desulfurization (FGD) waste disposal practices at coal-
fired electric generating plants.  The study was conducted by Arthur D.
Little, Inc., under EPA contract 68-02-3167, and involved characterization
of wastes, environmental data gathering, evaluation of environmental
effects, and engineering/cost evaluations of disposal practices at six
selected sites in various locations around the country.  Results of the
study are expected to provide technical background data and information
to EPA, state and local permitting officials, and the utility industry
for implementing environmentally sound disposal practices.

Data from the study suggest that no major environmental effects have
occurred at any of the six sites; i.e., data from wells downgradient of
the disposal sites indicate that waste leachate has resulted in concen-
trations of chemicals less than the EPA primary drinking water standards.
A generic environmental evaluation based on a matrix of four waste
types, three disposal methods, and five environmental settings (based on
climate and hydrogeology) shows that, on balance, technology exists for
environmentally sound disposal of coal ash and FGD wastes for ponding,
interim ponding/landfilling, and landfilling.  For some combinations of
waste types, disposal methods, and environmental settings, mitigation
measures must be taken to avoid adverse environmental effects.  However,
site specific application of good engineering design and practices can
mitigate most potentially adverse effects of coal ash and FGD waste
disposal.  Costs of waste disposal operations are highly system and site
specific.
                                   9-1

-------
Introduction

This study—of current coal ash and flue gas desulfurization (FGD) waste
disposal practices at coal-fired power plants—was conducted for the
U.S. Environmental Protection Agency (EPA) by Arthur D. Little, Inc.
(ADL) under EPA contract 68-02-3167.  The study involved characterization
of wastes, environmental data gathering, evaluation of environmental
effects, and engineering/cost evaluations of disposal practices at six
selected sites at various locations around the country.  Results of the
study are expected to provide the technical background data and information
needed to help EPA determine the degree to which disposal of these
wastes needs to be managed in order to protect human health and the
environment.  The study results will also assist EPA in preparing a
report to Congress required under the 1980 Amendments to the Resource
Conservation and Recovery Act (RCRA), and should provide useful technical
information to federal, state, and local permitting officials and utility
planners on methods for environmentally sound disposal of coal ash and
FGD wastes.  Results of this 3-year effort are summarized in this
paper.

Background on Waste Generation/Disposal Methods

Coal-fired power plants using conventional combustion technology generate
two major categories of waste materials.  Coal ash (fly ash, bottom ash,
or  boiler slag) and FGD wastes are generated in large amounts relative
to  other wastes generated at these plants and, therefore, are usually
referred  to as "high volume wastes."  Numerous other wastes, generated
in  smaller quantities, are associated with other processes or maintenance
operations in a power plant such as coal pile runoff, boiler blowdown,
cooling tower blowdown, water treatment wastes, maintenance cleaning
wastes, general power plant trash, and plant sanitary wastes.  This
project primarily focused on the high volume wastes.

Fly ash from coal-fired utility boilers is collected by mechanical
collectors and/or electrostatic precipitators, fabric filters, or wet
scrubbers.  By late 1982, approximately 103,000 MW of coal-fired generating
capacity—operational units, units under construction, and units at
various stages of planning—had been committed to FGD systems.  Flue gas
desulfurization can be accomplished by nonregenerable throwaway systems,
which result  in FGD wastes, or by regenerable systems, which produce a
saleable  product  (sulfur or sulfuric acid).  Operational nonregenerable
FGD systems are currently predominated by wet scrubbing technology
?o«r?n«,S°me ^ FG° scrubbinS systems were becoming operational in
1982-1983   The principal types of systems used in utility power plants
are those based on direct limestone, direct lime, alkaline fly ash  dual
alkali, and lime- or sodium-based dry FGD systems/''
                                   9-2

-------
Some projections on the generation of coal ash and FGD wastes [together,
these are designated as flue gas cleaning (FGC) wastes] in the U.S. are
presented in Table 1.  Most of the coal ash and all of the FGD wastes
generated are sent to disposal.(2,3)  Considering the expected increase
in coal consumption in the U.S., this is likely to be the case for many
years.  Utilization of FGC waste is expected to grow but at a slower
rate than FGC waste generation.  A significant fraction of the total
coal ash generated is used for such purposes as soil stabilization, ice
control, and as ingredients in cement, concrete, and blasting compounds;
however, there is currently no utilization of FGD wastes in the U.S.  On
balance, disposal will continue to be the major option for FGC waste
management in the U.S. for the foreseeable future.

Currently, all FGC waste disposal is on land.  At-sea disposal may be a
future alternative if it can be practiced under environmentally and
economically acceptable conditions.  The principal methods of disposal
on land are:  ponding, landfilling (including disposal in surface mines),
and interim ponding followed by landfilling.  Table 2 presents data on
current practices based on data obtained on 176 plants.

Ponding of FGC waste is more widely practiced than any other disposal
method.  Ponding can be employed for a wide variety of coal ash and FGD
wastes including chemically treated FGD wastes.  Ponds can be designed
based on diking or incision, but the construction of dams or dikes for
ponds is usually expensive.  In the future, particularly if chemical
treatment of FGD wastes is widely practiced, ponding will probably be
limited to those sites that would involve minimal construction of dams
or dikes.  One exception could be a special case of wet ponding—FGD
gypsum "stacking."  In this case, gypsum slurry from a forced oxidation
system would be piped to a pond and allowed to settle and the supernate
recycled.  Periodically the gypsum would be dredged and stacked around
the perimeter of the pond, thus building up the embankments.

Landfilling of FGC waste is also widely practiced, and can involve one
or more of a variety of handling operations prior to the disposal
operation.  For example, bottom ash is almost always sluiced from the
plant, so it must be dewatered (e.g., via hydrobins) before it is
transported.  Dewatering must also be applied to fly ash that is sluiced
from the plant or is wet-scrubbed from the flue gas—with or without
significant quantities of SO,.,.  Wet FGD waste must also be dewatered via
thickening, vacuum filtration, and, if necessary, blending with dry fly
ash for stabilization or other chemical treatment ("fixation") additives
such as lime.  On the other hand, fly ash slated for landfill is typically
transported directly from the plant in a dry state, with only enough
moisture added as required for dust control and compaction in the
landfill.  Wastes from a spray dryer FGD system can also be transported
directly; during this project, commercial operation of these systems on
utility boilers was just beginning.
                                    9-3

-------
             TABLE  1.   PROJECTIONS  OF FGC WASTE GENERATION
                       BY  UTILITY PLANTS IN THE UNITED STATES
                       (1980-1995)
     Waste Type
                            Waste  Generation (10  Metric tons/yr)
                            1980              1985            1995
Coal Ash*
FGD Wastes
TOTAL

62.4
8.6
71.0
(78.3)c
83.2
26.9
110.1
(121. 4)c
110.0
48.6
158.6
(174. 8)c
      Coal ash quantities are shown  on  a dry  basis.


      FGD waste quantities are shown on a wet basis  (50% solids).


      10  tons/year.
     Source:   Reference 2
                TABLE  2.  CURRENT FGC WASTE DISPOSAL METHODS UTILIZED AT
                         UTILITY COAL-FIRED POWER PLANTS  IN THE U.S.
                         (Data Base:  176 Plants > 200 MW)a


Type of Waste
Fly ash only
Bottom ash only
Combined fly and bottom ash
FGD waste only
Mixed fly ash and FGD waste
Mixed bottom ash and FGD waste
Mixed fly ash and FGD waste (stabilized)
Mixed fly ash, bottom ash, and FGD waste

Pondb
18
29
69
5
7
1
2
2
Number
Landfill0
46
13
9
-
7
-
6
1
of Plants
Interim Pond/Landf 111°
6
29
16
_
_
1
_
1
 Coal-fired plants on which data  were  available  (>_80%  of  their  power generated from
 coal in 1977)  which have  generating capacities  >_200 MW with the exception of four
 plants employing FGD systems.  Figures  represent  the  number of plants at which each
 waste type/disposal method is  practiced.   (Note:   Many plants  utilize more than one
 method.)


 Includes  direct ponding and interim/final  ponding methods.

 Includes  managed and unroanaged fills  and mine disposal.


Source: Reference 4
                                        9-4

-------
In a landfill disposal site, the wastes are spread on the ground in 0.3
to 1 m (1 to 3 ft) lifts and compacted by wide track dozers, heavy
rollers, or other equipment.  Layering proceeds in 0.3 to 1m lifts in
segments of the site.  The ultimate height of a disposal fill is site
specific, but can range from 10 m (30 ft) to as high as 76 m (250ft).
A properly designed and operating dry impoundment system can potentially
enhance the value of the disposal site after termination or at least
permit post-operational use.

Mine disposal is a variation of landfilling that is receiving increased
attention.  Surface coal mines, particularly those serving "mine-mouth"
plants, offer the greatest capacity and economic attractiveness for
disposal of wastes from power plants.(5,6)  Since the quantity (volume)
of FGC wastes produced is considerably less than the amount of coal
burned, many mines would have the capacity for disposal throughout the
life of the power plant.  Several plants, particularly in the Plain
States (e.g., North Dakota), have practiced this disposal method in
recent years.

Interim pond/landfill has been an important waste disposal method in the
past, but is likely to decline in importance in the future, particularly
since dry ash handling and disposal is being more widely practiced.

SITE SELECTION AND TEST PLAN PREPARATION

Candidate Site Selection Process

The overall objective of the candidate site selection process was to
evaluate available data on coal-fired power plants and recommend a
number of candidate and backup sites.  The selection process consisted
of two steps:

First, the contiguous 48 states were divided into 14 physiographic
regions, and the plants in each region were screened to develop a list
of plants that would be suitable for consideration as candidate and
backup sites.  A total target of 25 to 30 sites, including 18 candidate
and 7 to 12 backup sites, was desired.   Based on an assessment of
present and future FGC waste disposal practices, a preliminary distribution
of the targeted number of candidate sites in each region was agreed
upon.  In screening selections, the investigators remained cognizant of
the targeted number in each region,  but were not absolutely limited by
that number.  The attempt was to choose desirable plants in as many
regions as possible.   The list of plants in all the regions that came
through this filtration process amounted to 26.

Second,  these 26 were then ranked in iterative group discussions leading
to the nomination of  18 as candidate sites and the remainder as backup
sites.
                                   9-5

-------
Final Selection Process
The candidate and backup sites were then subjected  to a more  detailed
evaluation.  These evaluations included one or more detailed  site  visits
by engineering, environmental, and hydrogeologic  specialists  assigned to
the project.  Their findings, together with mid-course evaluations that
were continuously taking place, supported an iterative process  that
resulted in the selection of the final six sites.   Table 3 provides
overall information on the final six sites that were selected for
evaluation under this project; Figure 1 indicates the site locations.

Test Plan Preparation

Detailed test plans providing background information on each  of  the
sites, together with a description of the proposed  program of site
development, physical and chemical sampling, and analysis and engineering/
cost assessments, were developed.   The test plans were reviewed  by EPA
and the utility involved, and their comments were incorporated.  The
finalized test plans guided the work at each site.

SITE DEVELOPMENT AND PHYSICAL TESTING

After approvals from the utility and,  in some cases, from state  regulatory
agencies, site development was begun.   Site development and physical
testing were governed by procedure manuals^ '  developed for this project.
The activities involved in site development included the drilling  of
borings; excavating test pits;  collecting waste,  soil,  and water samples;
conducting field permeability tests;  installing ground  water monitoring
wells and piezometers;  and documenting each activity.   These activities
took place at each of the six sites in time periods of  2  to 4 weeks.
Table 4 indicates the timing  under which the six  sites  were developed
and the extent of the activities  at each site.  The table also gives the
number of physical tests performed;  i.e.,  laboratory soil classification
and permeability tests  on waste samples  from the  sites.   Preliminary
water balances were also developed for  each site.

CHEMICAL SAMPLING AND ANALYSIS

At each site  a program of  chemical  sampling and  analysis was  undertaken


                                                                        '
sampling and  analysis  program.
                                          obtained  from a series of
                                                 relativeiy dr-
                                                   *  BWm&T^ °f  the

-------
TABLE 3.  WASTE DISPOSAL SITES SELECTED FOR EVALUATION
Plant (Utility)
Allen
(Duke Power)
Elrana
[Duquesne Light
(waste disposal
by Conversion
Systems, IDC.))
Dave Johnston
(Pacific P&L)
Sherburne
County
(Northern
States Power)
Powerton
(Commonwealth
Edison)
Smith
(Gulf Power)
UL - Unllned.
CL - Clay-Lined.
AL - Artificially
Nameplate
Generating
Location Capacity
State (County) (MW)
N. Carolina 1155
(Gaston)
Pennsylvania 510
(Washington)
Wyoming 750
(Converse)
Minnesota 1458
(Sherburne)
Illinois 1786
(Tazewell)
Florida 340
(Bay)
Lined (Por-0-Tec).
Startup Date
(mo/yr)
Plant (FGD)
-/57
6/52
(10/75)
V59
5/76
(5/76)
V72
6/65


High Priority Issues
Under Study
Employment
of a
Waste Site Under Study Ground Surface Potentially
Waste Type
Combined fly
and bottom ash
Stabilized
FGD waste
Combined
fly and
bottom ash
Fly ash
Fly ash/FGD
Combined fly
and bottom ash
Combined fly
and bottom ash

Disposal Water Water Mitlgatlve
Method Quality Quality Practice
Pond (UL) XX X
Landfill XX X
(UL; offsltc)
Landfill
(UL; offaite)
Landfill (UL) X X
Pond (CL) X - X
Landfill (AL) XX X
Pond (UL) XX X


-------
DAVE JOHNSTON
                    SHERBURNE COUNTY
                                                       -V"^
                                           POWERTON  /"' \  f .
                                                         \^
                                                /-->;^
                                                             ELRAMA
                                                          AU.CN
FIGUBJE 1. LOCATION Of WASTE DISPOSAL SITES SELECTED FOR EVALUATION
                        9-8

-------
TABLE 4.  SUMMARY OF SITE DEVELOPMENT/PHYSICAL TESTING
Plant
Allen
Elrama
Johnston
Sherco
Powerton
Smith
Date
Development
Completed
(mo/yr)
01/81
03/81
05/81
08/81
11/81
12/81
Number of
Laboratory Physical Tests

Borings
20
20
14
13
11
25
Number
Wells
20
16
12
11
9
24
of
Test
Pits
2
4
10
-
1
-

Soil
Samples
152
199
154
178
112
146
Unified Soil
Classification
Series (USGS)
18
17
12
20
30
15
Permeability
4
13
7
6
8
8

-------
              TABLE 5.   SUMMARY OF CHEMICAL SAMPLING AND ANALYSIS PROGRAM





Samples" Trip 1
Site
Allenc



Elrama



Sherco







Smith





Powerton

Dave Johnston



Trip 1
wells
ash solids
interstitial liquors
soils
wells
waste solids
soil
waste extracts
wells
waste interstitial
liquors
waste solids
liner solids
liner liquor
soil solids
soil extracts
liquids
waste solids
interstitial waste
liquors
soil
soil liquors
wells
waste solids
wells
waste solids
waste extracts
soils
Trips 2, 3, and 4
wells and
surface waters


wells, lysimeters,
surface waters


wells and
surface liquors






wells and
surface waters




wells and
surface waters
wells and
surface waters


I CAP
X
X
X
X
X
X
X

X
X

X
X
X
X
X
X
X
X

X
X
X
X
X
X

X
Analyses
Trips 2, 3, and 4
1C As/Se Field Data Other
X

X

X


X
X
X



X


X

X


X
X

X

X

X X
X
X
X
X X
J
X Xd
X

X X
X






X X





X X

X X
X


aSamples obtained  during  site  development and subsequent sampling and analysis  trips.
 Analyses performed are  abbreviated  as  follows!
     ICAP       -  Ag,  Al,  B, Ba,  Be,  Ca, Cd, Co, Cr, Cu, Fe, K, Mg, Mn, Mo, Na, Ni,  P,  Pb,
                  Si,  Sr,  Th,  Ti,  V,  Zn, Zr.  (Does not  Include B, Ba, and SI
                  for  solids-)
     1C         -  F~,  Cl~, N05,  S0£,  Br~, PO^3".
     As/Se      -  either  or both on  selected samples.
     Field Data   ground  water level,  pH, dissolved oxygen,  conductivity, temperature.
^ther samples were obtained  (boiler cleaning wastes).   Analysis was  limited to ICAP,
 pH,  and bromate.
dlncludes solids characterization for SO^2  , total oxidlzable  sulfur, slurry pH,
 acid insolubles.
                                          9-10

-------
Chemical samples are subjected to several types of analyses:  ion
chromatography (1C) for six anions, inductively coupled argon plasma
emissions spectroscopy (ICAP) for 26 metals; and atomic absorption
spectroscopy (AA) for selected metals.  As shown in Table 5, these types
of analyses were performed on a mix of solid and liquid samples for each
site.  In addition, a limited number of experiments were performed to
assess the attenuative capacity of various soils obtained at the sites.
Furthermore, during the initial phase of this project, 23 grab samples
of wastes from 18 plants were obtained and analyzed using the EPA
Extraction Procedure (EP); a summary of results from these tests is
given in Table 6.  Further details on these tests, as well as results of
radioactivity measurements, are included in Reference A.

SITE-SPECIFIC ENVIRONMENTAL EVALUATIONS

The data and information from site development and sampling/analysis
were subjected to environmental effects evaluation throughout the project.
The individual site evaluations were developed by a series of five
sequential steps.

First, a review and evaluation was made of available background information
on the disposal operation and its environmental setting.  Second, present
disposal-related water quality effects were identified and described
based on evaluation and measured information developed in this project.
Third, apparent cause/effect relationships were hypothesized to explain
the findings at the sites.  Fourth, potential future ranges of water
quality effects were considered to the extent that suitable data were
available.  Finally, industry-wide implications of the findings at the
individual sites were considered in the generic assessment, discussed
later in this paper.

Environmental evaluation of all six sites has generated a significant
amount of data and information.  The following general items can be
reported:

     1.   Data suggest that no major adverse environmental effects have
          occurred at any of the sites.  For example, data from wells
          downgradient of the disposal sites suggest that the contribution
          of waste leachate to the ground water has resulted in concen-
          trations of chemicals less than the primary drinking water
          standards established by EPA.

     2.   The results from the sites are internally consistent.  In
          other words, the analyses of samples taken on different dates
          at the same locations are very similar.

     3.   The total integrated evaluation of data from site development,
          site water balances, physical testing of wastes samples, and
          chemical sampling and analysis is providing a large significant
          data base to explain many of the environmental effects that
          can result from coal ash and FGD waste disposal.

A brief account of the results at each of the sites is presented below.
                                  9-11

-------
                           TABLE 6.  SUMMARY OF RESULTS OF EXTRACTION PROCEDURE (EP) TESTS
                                    OF 20 FLY ASH AND 3 FGD WASTE GRAB SAMPLES
 I
I—'
ro
Metal Overall Range Observed, \i g/1
Fly Ash FGD Waste
Arsenic <2 - 410 <2 - 65
Barium <100 - 700 <150 - 230
Cadmium <2 - 193 <2 - 20
Chromium <8 - 930b <11 - 26b
Lead <3 - <36 <5
Mercury <2 <2
Selenium <2 - 340 8-49
Silver <1 <1
Interim Primary Drinking
Water Standards3, jj g/1 Ratio of Range Observed to Standards

50
1000
10
50b
50
<1
10
50
Fly Ash FGD Waste
<0.04 - 8.2 <0.04 - 1.30
<0.1 - 0.7 <0.15 - 0.23
<0.2 - 19.3 <0.2 - 2
<0.16 - 18.6 <0.22 - 0.52
<0.06 - 0.72 0.1
<1 <1
<0.2 - 34 0.8 - 4.9
<0.02 <0.02
aReference 7 gives these standards  ...for use in determining whether solid waste disposal activities comply
 with ground water criteria."  Standards included in Reference 7,  but not measured in these tests,  are for
 fluoride:  1400-2400 pg/1 (depending on temperature),  and for nitrate (as N):   10,000 yg/1.

"Reference 8 contains an amendment to the chromium criteria- for the EP,  revising it from total chromium to
 Cr(VI);  since the total chromium values were measured  by atomic absorption (AA), the measured ranges
 represent upper limits for Cr(VI) in the samples.

-------
Allen - The results indicate the following:

     1.   Leachate generated within the ash ponds contain elevated (over
          background) concentrations of several waste-related chemical
          constituents (e.g., boron, sulfate, calcium, arsenic).  However,
          the surrounding soils have attenuated significant fractions of
          leachate constituent contaminants within the immediate vicinity
          of the ponds.

     2.   Leachate water from the upgradient portions of the ash ponds
          has not moved sufficiently to create steady-state concentrations
          of unattenuated constituents (e.g., sulfate) in the downgradient
          wells.  However, concentrations of these constituents are
          expected to only reach or barely exceed secondary drinking
          water standards (e.g., for sulfate, 250 ppm).

Elrama - The results indicate the following:

     1.   Prior to disposal of FGC wastes, much of the site was contaminated
          by acid mine drainage, resulting in low pH (4.5 to 5) and high
          concentrations of chemical constituents (e.g., approx. 2000 ppm
          sulfate) in the ground water.

     2.   The landfill and runoff collection ponds may serve as potential
          sources of some constituents via leachate and overflow, including
          chloride and calcium (both at approx. 500 ppm).  However,
          neither chloride nor calcium is at a level to cause major
          concern.  In addition, an elevated level (approx.  0.2 ppm)  of
          arsenic was measured at one waste/soil interface lysimeter;
          however, it does not appear to be a general problem.  In any
          event, substantial attenuation of arsenic by soils at the site
          is expected.

     3.   The relative absence of elevated levels of these waste-related
          constituents in downgradient ground water may be explained by
          the relatively short time the fill has been in operation
          (4 years), chemical/physical attenuation phenomena (including
          the effects of the treatment/disposal process), or a combination
          of these factors.

     4.   The landfill does not appear to alter significantly the local
          concentrations of some constituents (such as sulfate) potentially
          available from both mine drainage and FGC wastes.

Dave Johnston - The results indicate the following:

     1.   The water balance and estimates of plume arrival time indicate
          that the widespread measurement at the site of what might
          elsewhere be considered elevated chemical constituent levels
          (e.g., sulfate, approx. 1000 ppm) is not due to the waste
          landfills.  The estimates of plume arrival time for the
                                  9-13

-------
         peripheral wells downgradient  from  (not  directly under)  the
         active landfill are in excess  of  100  to  300  years  considering
         only travel time in the  saturated zone.   Travel time from the
         20-year old inactive landfill  to  a  much  closer  (to the landfill)
         downgradient well is estimated to be  in  excess  of  20 years,
         accounting for both unsaturated and saturated zone travel.

     2.   Most of the "elevated" measurements reflect  pervasively high
         background levels characteristic  of highly mineralized ground
         water  in many western settings.   However,  lower measured
         values  (e.g., sulfate, approx.  100  ppm)  at one  background and
         one peripheral well indicate that even  in highly mineralized
         arid areas there may be  areas  of  good water  quality that
         require protection in waste disposal  site planning and management.

Sherco - The results indicate the  following:

     1.   Leachate movement from the ponds  has  thus far been sufficiently
         retarded by the clay liner to  preclude development of significant
         elevations of chemical constituents in the leachate at down-
         gradient wells.

     2.   There  is a waste-related influence  reflected in the slightly
         elevated levels  (of boron and  sulfate) measured in the peri-
         pheral/downgradient wells to the  west and southwest, but it is
         not clear whether this is due  to  past leakage from the sheet
         piling/conduit area, to  leachate  that has moved through the
         liner,  or to  a combination of  these two  sources.

     3.   Because of the pervious  soils  in  the  area of the  site, significant
         increases in  concentrations of major  soluble species are
         expected to occur in downgradient wells  in the next few years.
         Secondary drinking water standards  are expected to be exceeded
          in these wells.  However, any  effects of movement  of these
          species off-site will be mitigated  (diluted) by the Mississippi
         River,  which  flows by the plant.

     4.   The higher concentrations of waste  parameters in FGC pond
          supernatant versus underlying  waste interstitial waters may be
         due to two factors:  first, the conversion by the  utility to a
         system involving recycle of the FGC waste transport water
         would  have resulted in increased  concentrations of chemicals
          in the water; and second, the  evaporation of water in the pond
         would  also increase remaining  chemical concentrations.
                                    9-14

-------
Powerton - The results indicate the following:

     1.   Although the completed landfill was supposed to have a 0.25 m
          (8 in.) Poz-0-Tec liner, during the coring operation a general
          absence of liner material was observed.  This observation is
          consistent with the practical difficulty of achieving uniform
          placement of such a relatively thin layer of soil-like material
          over a large area.  Current engineering practice suggests that
          a minimum thickness of 0.45 to 0.60 m  (18 to 24 in.) of liner
          placement would be desired to ensure full effectiveness.

     2.   The surface water analytical results for Lost Creek are consistent
          with the water balance calculations.  Both sets of results
          indicate that the stream has adequate assimilative/dilution
          capacity to lower current concentrations of chemical constituents
          in leachate reaching Lost Creek to insignificant levels.

     3.   The results also suggest that the stream, if an effective
          ground water flow divide, may limit the extent of further
          downgradient ground water contamination by the waste plume.

     4.   The general lack of elevated trace metal concentrations in
          ground water suggests that a combination of chemical attenuation
          (especially for chromium and lead) and dilution is preventing
          the release of significant quantities of these elements and/or
          elevation to significant concentrations at downgradient locations.

     5.   Elevated concentrations of nitrate at various sampling locations
          at the site can be attributed to local agricultural and urban
          nonpoint source activities and not the coal ash landfill.

Smith - The results indicate the following:

     1.   There appears to have been a steady state achieved between the
          concentrations of soluble species in the pond and in the
          immediately adjacent downgradient areas.

     2.   There appears to be little nor no chemical attenuation of the
          major tracer species such as calcium and strontium, but rather
          a progressive reduction in concentrations in the downgradient
          direction.  This is consistent with what would be expected due
          to admixing of leachate with the greater amounts of dilution
          water.

     3.   The use of high total dissolved solids Bay water in the pond
          for makeup and its presence in adjacent downgradient areas
          create a situation where little incremental effect is detectable
          from such typical ash pond "tracer" species as sulfates,
          chlorides, and boron.
                                  9-15

-------
GENERIC ENVIRONMENTAL EVALUATION OF COAL ASH AM) FGD WASTE DISPOSAL

The environmental effects of solid waste disposal practice is determined
by three factors:  waste type,  disposal method, and the environmental
setting.  The data base from this project and other related projects
suggests that present and future practices of coal ash and FGD waste
disposal may be effectively evaluated through a matrix consisting of
four waste types, three disposal methods, and five environmental settings.

The four waste types are:

     1.   Fly ash or fly ash admixed with other materials.  A significant
          body of literature suggests that the majority of trace metals
          available for leaching from utility solid wastes may be associated
          with those containing fly ash.  Thus this category of wastes
          includes fly ash or fly ash mixed with bottom ash and fly
          ash/bottom ash/FGD waste mixtures (excluding chemically treated
          FGD wastes; see item 3, below).

     2.   Non fly ash materials.  In this category are included bottom
          ash (or boiler slag)  and FGD wastes that are disposed of
          separately from fly ash (including forced oxidation wastes).
          This category usually contains lesser concentrations of trace
          metals, but can result in higher concentrations of major
          species (e.g., chlorides from FGD waste).

     3.   Stabilized FGD wastes.  FGD wastes may be processed or stabilized
          for full-scale disposal by a variety of processes; the processes
          presently in commercial practice involve the addition of lime
          and fly ash, or processed slag.  Lime/fly ash stabilization
          for landfill disposal is presently practiced at some power
          plants and is expected to grow in importance.  Processed FGD
          wastes are a separate category because of the differences in
          their physical and chemical properties created by the stabili-
          zation process.

     4.   Dry FGD wastes.  Several dry FGD systems are, expected to come
          into commercial use over the next 3 years.     It appears that
          calcium-based dry FGD systems are anticipated to grow more
          than the sodium-based systems.  By either process, dry FGD
          systems provide a combined waste containing fly ash and the
          sulfur compound in a relatively dry form that is likely to be
          sent for disposal to a managed landfill.  The physical and
          chemical properties of these wastes are expected to be different
          from the other categories discussed above; additionally, there
          is a relative lack of even limited field scale information on
          their leaching characteristics to date.
                                  9-16

-------
There are three disposal methods for coal ash and FGD wastes  that are  in
practice and expected to continue in the future:  (1) pond disposal;
(2) interim ponding followed by landfill disposal; and  (3) landfill
disposal (including disposal in mines, which is considered a  special
case of landfilling).

Three of the five environmental settings for solid waste disposal are
based on major differences in climate and hydrogeology.  These are:
(1) coastal areas, specifically those areas where surface water and
ground water are influenced by the ebb and flow of tides;  (2) arid
areas, characteristic of much of the western U.S. where net evaporation
generally exceeds precipitation by a significant margin; and  (3) interior
areas, characteristic of the non-coastal portions of the eastern U.S.
where there tends to be more of a balance between precipitation and
evaporation and where permanent surface water bodies are in such abundance
as to be near many disposal sites.

Further evaluations during this project suggested that a further breakdown
of two special categories would be useful because of their significant
characteristics.  These are:   (1) arid areas in the west where ground
waters and surface waters are very highly mineralized, and (2) interior
areas subject to acid mine drainage.  Both of these last two  types of
settings and the coastal setting tend to have water quality characteristics
that can potentially show less of an incremental effect from  coal ash
and FGD waste leachates.  This is because the waters in these areas
already contain a number of chemicals found in the leachate.

Table 7 is a matrix of waste types, methods of disposal, and  environmental
settings and indicates combinations for which field-scale and other
information is available.  Sources of data and information other than
this study included the Utilities Solid Waste Activities Group (USWAG),
the Electric Power Research Institute (EPRI), and the Department of
Energy (DOE)-  DOE is currently sponsoring a study of disposal FGD
wastes in a surface mine; the study was originally sponsored by EPA.  As
is clear, some information is available for most of the combinations
that are being practiced today or are likely to be practiced  in the
future.

It appears that, on balance, technology exists for environmentally sound
disposal of coal ash and FGD wastes using any of the modes of disposal.
Potential environmental effects are highly site and system specific.
For some combinations of waste types, disposal methods, and environmental
settings, mitigative measures must be taken to avoid ground water and/or
surface water contamination.  However, site specific application of good
engineering design and practice can mitigate most potentially adverse
environmental effects of waste disposal.
                                   9-17

-------
 I
M
oo
                                                                                  TABLE 7.  SUMMARY OF INFORMATION AVAILABLE  FOR COMBINATIONS
                                                                                            OF WASTE TYPES, DISPOSAL METHODS, AND  ENVIRONMENTAL  SETTINGS
Ponding Interim Ponding/Lsndf i 1 ling
Fly
Ash*
COASTAL SETTING X
Smithc


Non-Fly Processed Dry Fly
Ashb FCD FGD Asha
P NA NA X/Pd
Chistnan
Cr.
(USUAC)
Non-Fly Processed Dry Fly
Ashb FGD FGD Ashb
P P NA P



Landfilllng
Non-Fly Processed Dry
Ashb FCD FCD
P P Pe



                                     ARID  WESTERN  SETTINC-
                                     Not Highly  Mineralized
                                    ARID  WESTERN  SETTING -
                                    Highly  Mineralized

INTERIOR SETTING -
Not Highly Acidic





INTERIOR SETTING -
Highly Acidic
(nine drainage)

Notes: a. Includes


X P X NA X/Pd P
Allen, Bruce Ballly
Sherco, Mansfield (USWAC)
Michigan City
(USWAG) ,
Walllngf ord
(USWAG)


P P P NA P P

co-disposal of fly ash with other wastes.
Johns ton;
Hilton Young
(DOE/EPA)

P NA X
Powerton,
Zue 11 inger
(USWAC)
(USWAG)
Dunkirk (DOE)


P NA P




P X Pe
Coneevllle
(EPRI/USWAC)





P it Pd
Elrama

                                              c.  Plants for
                                                  appropria te poa1tions.
                                                  Either  the interim  pond or landfill ?•*-act of operation studied at  field  scale,  but not both.
                                                  Laborabory da ta  only.
                                              e.  Laborabory da ta  only.
                                    Key:
                                              X - Data available  from full-scale field studies.
                                              P - Data aval lab le  from la bora tory and/ or 11 mi ted- scale field a tudies  for  projec tloo  purposes .
                                             NA = Matrix combination  not  applicable due to lack of present and future practice.

-------
ENGINEERING/COST EVALUATIONS

The first major efforts in the engineering/cost evaluations  involved
development of site-specific conceptual engineering designs  and  costs
(capital and first year operating and maintenance costs) for the current
solid waste handling and disposal operations at the six study sites.  To
facilitate the ultimate use of the cost data, the estimates were developed
by breaking down the waste handling and disposal operations  into five
modules:  (1) raw material handling and storage; (2) waste processing
and handling; (3) waste storage; (4) waste transport; and  (5) waste
placement and disposal (including site monitoring and reclamation).

Based on the site-specific cost estimates and other studies by TVA,
EPRI, and other organizations, generic capital and O&M cost  estimates
were then prepared for individual modules comprising waste handling and
disposal for coal ash and FGD wastes.  Tables 8 and 9 provide a  summary
of the results of this effort.

The range of costs given represents variations in specific plant operations
as well as variations in the several cost estimates used in preparing
these estimates.  For example, the higher end of the range for FGD waste
handling/processing might include thickening, vacuum filtration, and
mixing with lime and fly ash, while the lower end could represent a
simpler operation with little or no processing.  Figures 2 and 3 show,
in graphical form, the estimates for the FGD waste handling/processing
"module."  Similar figures for all the modules listed,in Tables  8 and 9
will be included in the final report for the project.

CONCLUDING REMARKS

Results from this 3-year study of disposal of coal ash and FGD wastes
from coal-fired electric generating plants should provide major  technical
guidance for regulatory bodies and the utility industry.  However,
results from field studies of this type are limited, and predictive
tools (e.g., computer models) for evaluating interactions between these
wastes and site-specific hydrogeologic systems are, in many cases,
inadequate.   For this reason, additional efforts sponsored by the industry
are currently underway to develop more sophisticated tools for predicting
and analyzing the potential environmental effects of coal ash and FGD
waste disposal.   These efforts will be described later in this session.
                                   9-19

-------
                        TABLE 8.  GENERIC CAPITAL COST ESTIMATES FOR FGC WASTE  DISPOSAL
                                              (Late 1982 Dollars)*
Capital Cost Range
($/kH)


Fly

Fly
Ply
Fly

Module


Submodule
ash handling/processing Wet handling
Wet handling

ash storage
ash transport
ash placement/disposal
Bottom ash handling/processing
Bottom ash transport


Bottom ash placement/disposal

Raw
PCD
"FCD

FGD

"Er

>>Rt

materials handling/storage
waste handling/processingc

waste placement/disposal

Dry handling
Dry
Wet sluicing
Dry trucking
Unlined pond
Landfill
Wet handling
Wet handling
Wet sluicing
Dry trucking
Unlined pond
Landfill
Dry (lime and
Wet handling
Dry trucking
Unlined pond
Landfill
igineering News Record (ENR) Index - 3931.11

^lationship between plant size a
• 365.97

250b
w/o recycle 2.3-4.3
w/recycle 3.7-6.8
2
4
3
0
15
4
w/o recycle 2
w/recycle 2
3
0
6
1
fly ash) 2
18
0
0
10
4
(1913 100)
(1967 100)
nd waste generation for typical
.2-4.1
.7-8.8
.5-6.4
.3-0.5
.1-27.8
.3-8.1
.2-4.1
.5-4.6
.0-5.6
.2-0.4
.4-11.8
.3-2.4
.4-4.5
.1-33.6
.7-1 .3
.4-0.7
.0-18.6
.1-7.6


case:

Plant
Size (MU)
500b
1
3
1
4
2
0
12
3
1
2
2
0
5
1
2
15
0
0
8
3



.9-3.5
.0-5.5
.8-3.3
.2-7.7
.7-5.1
.3-0.6
.9-23.9
.3-6.1
.7-3.2
.0-3.7
.4-4.5
.2-0.3
.1-9.6
.1-2.0
.1-3.9
.2-28.3
.5-1 .0
.3-0.6
.9-16.6
.3-6.2



1
2
1
3
2
0
11
2
1
1
1
0
4
0
1
12
0 .
1000b
.5-2.
.4-6.
.4-2.
.7-6.
.2-4.
.3-0.
9
4
7
8
0
5
.0-20.5
.5-4.7
.3-2.
.6-3.
.9-3.
.1-0.
.2-7.
.9-1.
.9-3.
.8-23
. 4-n.
0.3-0.
7
2



.9-14
.7-5.



5
0
6
2
7
6
4
.8
8
5
.7
0



2000b
1.3-2.3
1.9-3.6
1.2-2.2
3.2-5.9
1.7-3.2
0.2-0.5
9.4-17.5
1.9-3.6
1.0-1.9
1.3-2.4
1.5-2.8
0.1-0.2
3.4-6.2
0.7-1.3
1.6-3.0
10.8-20.0
0.4-0.7
0.3-0.5
7.0-13.1
2.2-4.0



Annual Waste Generation Rate








Fly Aah
Bottom Aah
FGD Waste
"Typical Case"
Coal Propertie
Load Factor:
(dry metric



Assumptions
s :

tons/HW of Plant Generati
280
70
240

21 S, 13Z Ash,
70Z




10,500

ng Capacity)




Btu/lb





(24.4 M





MJ/kg)



















                   Heat Rate:
                   S02 Removal:
                   Lime Stoichiometry:
                   Fly Ash/Bottom Ash Ratio:
cAssumed  FGD  System:  Wet Lime  Scrubbing
10,250 Btu/kUhUO,8M MJ/kWh)
90Z
1.1
80/20
                                                   9-20

-------
                          TABLE 9.   GENERIC  ANNUAL  COST ESTIMATES FOR FGC WASTE DISPOSAL
                                                (Late  1982 Dollars)*
Annual Cost Range
($/dry metric ton)
Plant Size (Mw)
Module
Fly ash handling/processing


Fly ash storage
Fly ash transport

Fly ash placement/disposal

Bottom ash handling/processing

Bottom ash transport

Bottom ash placement/disposal

Raw materials hand ling /storage
FGD waste hand ling /pro cess ing
FCD waste placement /disposal

Subtaodule
Wet handling w/o recycle
Wet handling w/recycle
Dry handling
Dry
Wet sluicing
Dry trucking
Unlined pond
Landfill
Wet handling w/o recycle
Wet handling w/recycle
Wet sluicing
Dry trucking
unlined pond
Landfill
Dry (lime and fly ash)
Wet handling
Dry trucking
Unlined pond
Landfill
250b
2.5-4.6
3.7-6.8
2.5-4.7
3.3-6.1
4.2-7.6
1.7-3.1
11.5-21.3
7.0-13.0
11.3-21.0
12.3-22.8
9.2-17.1
3.4-6.3
9.2-17.1
5.4-10.0
4.1-7.6
17.2-31.9
1.1-2.1
2.9-5.4
8.5-15.8
4.0-7.5
500b
1.0-3.7
2.9-5.4
2.1-3.9
3.0-5.6
3.2-5.9
1.5-2.8
9.1-16.8
5.6-10.5
9.0-16.7
10.3-19.1
7.3-13.5
2.8-5.2
7.9-14.6-
4.7-8.8
3.7-6.7
13.8-25.5
0.9-1 .7
2.3-4.3
6.7-12.4
3.4-6.3
1000b
1.6-3.0
2.3-4.3
1.7-3.2
2.8-5.2
2.5-4.7
1.3-2.5
7.2-13.5
4.6-8.5
6.9-12.8
8.4-15.7
5.6-10.3
2.2-4;l
6.5-12.1
4.1-7.6
3.4-6.2
11.0-20.5
0.7-1.3
1.8-3.3
5.2-9.7
2.8-5.3
2000
1.3-2.3
1.8-3.6
1.5-2.7
2.5-4.7
2.0-3.7
1.2-2.2
5.7-10.5
3.7-6.9
5.3-9.9
6.9-12.8
4.3-7.9
1.8-3.3
5.4-10.0
3.5-6.5
3.0-5.6
8.8-16.4
0.6-1.1
1.4-2.6
4.1-7.6
2.4-4.4
afcngineering News Record (ENR) Index - 3931.11 (1913 100)

^Relationship between plane size


Fly Ash
Bottom Ash
FGD Waste
"Typical Case
• 365.97 (1967 100)
and waste generation for typici
Annual Waste Gener;
(dry metric tons/MU of Plant
280
70
240
Assumptions

jl case:
ation Rate
Generating







Capacity)




















                   Coal Properties:
                   Load Factor:
                   Heat Rate:
                   S02 Removal:
                   Lime Stoichiometry:
                   Fly Ash/Bottom Ash Ratio:
cAssuraed FGD  System:  Wet Lime Scrubbing
21 S,  13Z Ash,  10,500 Btu/lb  U4-4" MJ/kg)
70Z
10,250 BtuAWh  (10.8M MJ/kWh)
90Z
1.1
80/20
                                                      9-21

-------
50.000
                      100 OOO
                                          metric tons / year
                                         200,000            3OO.OOO
400.00O
                      100,000          300,000         300,000           4OO.OOO
                           F  0 D  WASTE GENERATION  RATE  ( ton* / y«ar )
    Soviet- Arthur 0. Little,Inc. £iiimot«
           500,OOO
                  FIGURE 2.   FGD WASTE HANDLING AND PROCESSING!
                             ANNALIZED COSTS VERSUS FGD
                             WASTE GENERATION RATE
                                            9-22

-------
    eooo
   60OO
CO
O
O

o
z
   40OO
    2000
z
z
                        100,000
metric tons / year
200,000            soopoo
400,000
             Bads: 1982 Dollar*
                                                                                  I
                                                                           I
        O              100,000         200.OOO         3OO.OOO          4OO.OOO

                           F Q 0 WASTE GENERATION RATE ( tona / year )

    ' Source' Arthur 0. Little,Inc. Eilimotei
                                               500,000
                  FIGORE 3.   FGD WASTE HANDLING AND PROCESSING:

                             ANNALIZED COSTS  VERSUS FGD

                             WASTE GENERATION RATE
                                            9-23

-------
REFERENCES

1.    Smith,  M.P.,  et al. ,  "EPA Utility FGD Survey, July-September 1981,"
     prepared by PEDCo Environmental,  Inc., for EPA, Industrial Environ-
     mental Research Laboratory,  Research Triangle Park, NC, EPA-600/7-
     81-012e (NTIS No. PB 82-23150),  December  1981.

2.   Santhanam, C.J., et al.,  "Waste  and Water Management for Conventional
     Coal Combustion:  Assessment Report-1980."  Prepared by Arthur D.
     Little  Inc., for EPA,  Industrial Environmental Research Laboratory,
     Research Triangle Park,  NC,  EPA-600/7-83-007 (NTIS No. PB 83-163154),
     January 1983.

3.   Santhanam, C.J., et al.,  "Waste  and Water Management for Conventional
     Coal Combustion:  Assessment Report-1979," Vol. I-V, prepared by
     Arthur D. Little, Inc.,  for  EPA,  Industrial Environmental Research
     Laboratory, Research Triangle Park, NC, EPA-600/7-80-012a to e (NTIS
     Nos. PB 80-158884, -185564,  -222409, -184765, -185572), January
     (Vol. I) and March (Vol.  II-V).

4.   Santhanam, C.J., et al.,  "Characterization and Environmental
     Evaluation of Full-Scale Utility Waste Disposal Sites - Final
     Report  (draft), prepared by  Arthur D. Little, Inc. under Contract
     No.  68-02-3167  for EPA,  Industrial Environmental Research Laboratory,
     Research Triangle Park,  NC,  to be published.

5.   Lunt, R.R., et  al., "An Evaluation of the Disposal of Flue Gas
     Desulfurization Wastes in Mines  and the Ocean:  Initial Assessment,"
     prepared by Arthur D. Little, Inc., for EPA, Industrial Environmental
     Research Laboratory, Research Triangle Park, NC, EPA-600/7-77-051
     (NTIS No. PB 269270), May 1977.

6.   Lunt, R.R., et  al., "An Evaluation of the Disposal of FGD Wastes in
     Coal Mines and  at Sea:   Refined  Assessment," (draft) prepared by
     Arthur D. Little, Inc.,  under Contract No. 68-03-2334, for EPA,
     Industrial Environmental Research Laboratory, Research Triangle
     Park, NC,  to be published.

7.   "Criteria  for Classification of  Solid Waste Disposal Facilities and
     Practices:  Final, Interim Final, and Proposed Regulations," Federal
     Register, Vol.  44, No.  179,  September 13, 1979.

y-   "Hazardous Waste Management  System; General and Identification and
     Listing of Hazardous Waste," Federal Register, Vol. 45  No  212
     October 30, 1980.                                     '

9.   Murarka, I.P.,  "Solid Waste  Environmental Studies at the Electric
     Power Research  Institute," to be presented at the EPA/EPRI Symposium
     on Flue Gas Desulfurization, New Orleans, November 1-4, 1983.

-------
OPERATIONS HISTORY OFrLOUISVILLE GAS & ELECTRIC
            FGD SLUDGE STABILIZATION

         R. P. Van Ness, J. H. Juzwiak,
                   W. Mclntyre

-------
                OPERATIONS HISTORY OF LOUISVILLE GAS  & ELECTRIC
                            FGD SLUDGE STABILIZATION
          by:   Robert Van Ness
               Manager of Environmental Affairs
               Louisville Gas and Electric Company
               Louisville, KY

               John H. Juzwiak,  P.E.
               Manager of Utility Operations
               Conversion Systems, Inc.
               Horsham, PA  19044

               William Mclntyre
               Senior Technical Advisor
               Conversion Systems, Inc.
               Louisville, KY

                                    ABSTRACT

    The Louisville Gas  & Electric Company  (LG&E)  has been an  industry  leader
in the deployment  of flue gas desulfurization technology  from  its very  incep-
tion.  LG&E  was one  of  the  first major  utilities  to install  SO2  scrubbers;
and,  at the present  time,  it operates seven scrubbers serving  a combined  gen-
erating capacity exceeding 2200 m.w.  These scrubbers which have been purchased
from several  different manufacturers, represent  a  broad  spectrum of  scrubber
technologies including dual alkali, lime and limestone systems.

    The bleed stream  from  these  scrubbers  is  dewatered,  and  the  resulting
solids are chemically  and physically  stabilized  in  processing plants purchased
from Conversion Systems,  Inc.  (CSI),  an early pioneer  in  the  field.  Unit  #6
at Cane  Run,   a  270  m.w.  unit  with  a  dual  alkali  scrubber,  is served  by  a
stabilization  facility  which began operations in April,  1980. Units #4/5  at
Cane  Run  have  a   combined capacity  of  360  m.w.  and are  equipped with  lime
scrubbers.  The waste  from these scrubbers is combined and  treated  in another
stabilization facility.   The  four  units  at Mill  Creek have a combined capacity
exceeding  1600  m.w.  and are also equipped  with lime  scrubbers.  One  large
stabilization  facility  was  installed to  handle  the  combined  bleed  from all
four Mill  Creek scrubbers.

    The stabilization  plants have  run for approximately  three  years and are
operated and maintained  by LG&E  personnel  with advisory  assistance supplied by
CSI.   The  knowledge gained  from the  experiences of  LG&E  and  CSI  in  the  last
three years of operation  would be  useful  both to current  operators of stabili-
zation facilities  and  to those who are  anticipating  the  procurement of scrub-
bers and stabilization facilities.

    This paper  presents  some  of the  operating  and  maintenance history which
has been obtained  from these plants.  Included will be discussions on reliabil-
ity of  individual  equipment and  discussions  of some  modifications  which were
                                      9-25

-------
made to  improve  reliability.  Operating  problems will  be addressed  including
some of  the  inherent  difficulties  encountered  in processing  scrubber  sludge
and fly ash.

    In addition, a  short  discussion of the  landfill  operation which  is  an  in-
tegral part  of the  stabilization  process will  be presented,  as  well  as  the
results of several years of  landfill  investigation.   Discussions of  the  impact
of  the landfill  operation  and plant control  upon  the environmental  properties
of  the final landfilled material will be offered.

                                   BACKGROUND

    Louisville Gas  &  Electric Company  (LG&E) owns and operates two  coal  fired
generating stations  in Jefferson County,  Kentucky.   These  stations  serve  the
city of  Louisville  and its  surrounding  suburbs.   During the early  1970's, when
sulfur dioxide emissions from coal  burning power  plants became a high environ-
mental priority,  LG&E was required  to  install scrubbers  on  the  newest of  its
coal  fired  generating units.  At that  time  there had been only  a handful  of
scrubbers  installed  on major utility  power  plants.    LG&E became one  of  the
industry  leaders  with  its  early demonstration  scrubber at their  Paddy's  Run
Plant.   It  has continued  in  the lead position by installing  a total of  seven
additional scrubbers at its Cane Run and Mill Creek Generating  Stations.

    During the early operating days of the Cane Run scrubbers,  it became  appar-
ent that some  form  of sludge stabilization  was  necessary to dispose  of  the
scrubber  bleed  and  to prevent the creation  of  a surface or groundwater  pollu-
tion  problem.   It was  for  this  reason that  LG&E  in   the  late  70's  contracted
with Conversion  Systems,  Inc.  (then  known  as IU Conversion Systems)  to design,
engineer  and  assist in the  erection of three  scrubber sludge dewatering  and
stabilization  facilities.   Two are located at Cane Run;  and, one at Mill  Creek.
These  three  stabilization facilities  are operated  by  manpower  provided by LG&E
with  the  help  of  an advisory team provided by CSI.  The disposal of  the  scrub-
ber waste  is   accomplished   by  dewatering  and  stabilization,  utilizing  the
Poz-0-Tec system developed by CSI together with controlled  landfill disposal.

    The  LG&E/CSI  operations team has  acquired  over  three  years  experience  in
operating  these  three  stabilization  facilities  and has encountered many  opera-
tion and  design  problems.  Many  of  these problems  will be  encountered by  other
firms  as  they  begin operation of their own stabilization plants.

    The  first  of these facilities  was a  relatively   low  budget  blending unit
designed  to  handle waste generated  by  the Cane  Run   #6 unit.   This  unit is  a
270 megawatt  generator using coal  with 12 to  16% ash  and 4%  sulfur content.
It  is equipped  with  an electrostatic  precipitator  and one of the  first dual
alkali scrubbers.  The blending  facility  included  fly ash  and  lime storage  and
the feed systems  necessary  to  blend fly  ash  and lime with   dewatered  sludge
from  the  scrubber.  The interface on  this  plant was  somewhat unusual when com-
pared  to other facilities  provided  by  Conversion Systems.  LG&E  had been  de-
watering  thickener  underflow and  disposing  of  this   material  in   their  bottom
ash ponds with no stabilization.  When  it was decided  to install  stabilization,
                                      9-26

-------
a facility  was designed which  placed  the blending  equipment  within the skirt
of the fly  ash silo.   This compact arrangement,  although  economical and expe-
dient, was  later  the  cause  of  many  operational  and  maintenance  headaches.
(See Figure 1)
FIGURE 1   Exterior of Cane Run #6 showing Poz-O-Tec
       being loaded for transport to landfill
                                UNDERFLOW HANDLING

     In order  to properly evaluate the experiences encountered  in  the  operation
of  the Cane Run facility, it is necessary to  consider  the  system  including  not
only  the blending  unit,  but also  the  vacuum filters  and  thickeners.   Those
familiar  with  FGD thickener  operation realize that it  is difficult  to maintain
a  specific underflow  density  in  normal  operation.   Usually  the variance  in
underflow density is alleviated by  a  surge  tank between the  thickener  and  the
vacuum filters  in  which  the  thickener  underflow is blended to  control the den-
sity  and  flow.   In  the Cane Run #6  facility  the  vacuum  filters are fed directly
from  the  thickener underflow pumps.   Since the  thickener  underflow  will vary
from  5%  to 30%  solids,  the  resulting  filtercake solids content is  also highly
variable.   Because the  processing  plant  must handle  upsets  in  the  thickener
operation,  the  quantity  of material is highly variable.  Due  to wide  swings in
flow  and  solids content,  the  control system  is sometimes  unable to  respond,
and the final blended product is not always  consistent.  At  times  the  material
is very dry which  causes  stockpile  and landfill  dusting problems;  and, at other
times,  the material  is  wet and  sloppy   which  causes difficulties  in handling
and can turn  the landfill into  a quagmire.  We  have  been  able to mitigate  the
                                     9-27

-------
effects of this variability by careful  stockpile and landfill management  tech-
niques.  These techniques include  segregation  and  selective blending of  stock-
piled material with different properties.  These practices  enable  us  to produce
a more consistent landfill material which  minimizes handling and  dusting  prob-
lems, thus increasing landfill operating efficiency.

    The absence of a surge tank  has increased problems  in the day-to-day  man-
power allocation and management  of the  facility.  The waste treatment  plant is
a slave to the thickener operation.  The plant must be ready to receive  thick-
ener underflow whenever the torque on the thickener rake is high.   The  facility
also must  sit  idle  or  process at  a minimal rate  while  a  bed  is  developed  in
the  thickener.

     The production  rates  of  the  processing  facility have,  at times,  exceeded
the  design rates due to high torque conditions in  the thickener.  In  most  cases
we  have been  able  to handle  these  additional  quantities of material,  although
the  resulting product has not always been entirely  up to specification  require-
ments, and the overload condition has put an excessive strain on the  equipment.
We  have been able to handle this off-spec material  through  the use of stockpile
and  landfill management techniques.  These  techniques have enabled us  to  main-
tain environmentally safe landfill sites.

     The Cane Run #4/5  and Mill  Creek  Processing  Plants have been  equipped with
underflow surge tanks which have produced a more consistent product and permits
greater flexibility for maintenance scheduling.

                                FLY ASH HANDLING

     We have  installed  a total of  four  fly  ash silos  at  the three LG&E facili-
ties.  All of these silos are designed  with  mass flow bin bottoms and  variable
speed  tapered screw feeders.   Although  this design  was initially more expensive
than the  tradition  air  slide  flat bottom fly  ash  silo  installation,  there  are
almost no  operating costs  since it eliminates the need  for compressed air  for
air  slide operation.  We have also eliminated  the high maintenance cost usually
encountered with star valve installations.   This fly  ash feeder and bin design,
however,  is not  without  its  operating  problems.    The bin  bottoms are  designed
with long, tapered sides which reduce the potential for bridging and  rat-holing
by  improving  the  flow  characteristics  of the  fly  ash.   The result  is a  rela-
tively uniform delivery of fly ash to the tapered  screw feeder.

     The tapered  screw  removes the  fly  ash  at a  predictable  and  controllable
rate and  feeds it  into  the  process.    During  the  start-up of  the Cane Run  #6
facility, we experienced several fly  ash floods;  that is,  the contents of  the
silo flowed through the feed system totally out  of  control.  Subsequent experi-
ments  and  careful  review  of  our  operating  logs  indicated  that this condition
was  observed  only  at times when the fly ash level in the silo was very  low or
when ash  was  being  transferred  into the system.    Since  the ash is transported
via  pneumatic  conveyor, we  theorize  that  air was remaining  entrained in  the
ash  and while in this state the  ash was extremely  flowable.
                                      9-28

-------
    Our  solution  to this  problem was  to  modify our  operating procedures and
our equipment.   Our operating  procedures  were  modified to  require  the slide
gate above  the tapered  screw  feeder to be  closed  whenever  the  silo level  is
below the intersection of the cone and  the  straight  side of the silo, or when-
ever ash is transferred  into  the  silo.   In addition, we installed instrumenta-
tion which automatically closes  the  gate whenever a  flood  condition is detec-
ted.  These measures proved successful  in the Cane  Run #6 plant;  but because
of the size of  the  gates in the two later plants, it  was decided  to install a
smaller gate downstream of  the feeder.  This gate is designed to close whenever
a flood condition is detected.

FLY ASH DUST CONTROL

    A  problem  encountered  early  in  start-up was  the operation of  the mixer
dust collectors.   These  collectors  are  intended to  control fly ash  and  lime
dust (and water  vapor)  which escapes  (the  equipment)  to the atmosphere at the
mixer.   Our original design proved inadequate due to plugging of the duct work.
The failure of these collectors led  to high dust concentration  in the air which
made it uncomfortable for operators  and created  problems for the equipment.  A
revised design  was  installed  on the Mill  Creek  plant which proved successful.
While earlier  installations attempted  to  maintain  a  negative  pressure  in  the
mixer  in  an effort to  contain the  dust  within the mixer  chamber,  this  new
design concentrated its efforts  outside the mixer.  Observation of  the  mixer
system revealed  that the atmosphere of the mixer  is extremely turbulent  with
the turbulent action causing dust particles  to  be suspended.  When this atmos-
phere was pulled  into the  dust collector,  the collector soon became overloaded
and clogged.  We  soon learned  that  the  dust particles  tend  to  settle,  if  per-
mitted,  in  a  still atmosphere.   This  was  accomplished with  a  revised  dust
collection  hood  which was  designed  with  a  long stilling  chamber  immediately
after  the   mixer.   The  dust  collector suction  was  moved  downstream of  the
stilling chamber.   This  design provided sufficient time  for dust  particles to
settle and  become  captured by the  slightly moist product.  Our  success  with
this system was extremely  gratifying,  and efforts are  now  underway  to install
the same system at  the other Louisville plants.

MAINTENANCE

    The  fly ash  laden  atmosphere combined  with  the corrosive  properties of
lime and the water  in the  filtercake all  combine to  make  a maintenance head-
ache.  Those  who have  never  operated  a sludge  stabilization  facility should
not be  lulled  into believing  that  these plants  do  not  require  operators, or
that maintenance will be minimal.   Improvements  in  the fly  ash dust collection
system and other  equipment  modifications have contributed greatly to improving
the  service  life of our equipment.   However,  additional maintenance routines
initiated by LG&E at these  plants have  been  an  even  greater factor in the  ser-
vice life of this equipment.   A good deal of the maintenance problems encoun-
tered in a  sludge  stabilization facility are due to ineffective or  inadequate
cleanup efforts.  The theory of  operation  of the Poz-O-Tec  process  is that the
fly  ash  and lime chemically  combine  to  form  a  matrix which binds  the  FGD
scrubber sludge  into  a  cement-like  material.  This  material will build up and
cure into a hard,  abrasive mass which  will  destroy  bearings,  gears and other
                                      9-29

-------
 mechanical  parts.  This was  evident  in the early days at Louisville when mater-
 ial  had  been allowed to cure in the mixer.   The mixer paddles  began  to hammer
 the  cured material.   The resultant  stress was transmitted along  the  shaft and
 eventually  contributed to the premature  failure of  several  gear teeth  in the
 mixer  speed reducer.  LG&E  has  instituted  a program which  keeps  the material
 build-up under check.  Every day the plant is shut down  for  complete  cleaning.
 This includes  cleaning  of   the  mixers  and hosing  of  conveyor  parts.   (See
 Figure 2)   This  cleaning  usually takes several hours  but has proven worthwhile
 in extending the life of the equipment.   To speed cleanup at  the  Cane  Run #6
 processing  plant,  the mixer  has  been  coated with  a  high density  polyurethane
 coating  which has proved to be  very  useful in preventing  material  build-up.
 It permits the  mixers  to be cleaned  with high  pressure water.  This  coating
 has, however,  become  scored by  the fly  ash  and its  benefit  will  diminish.   A
 reapplication  will hopefully restore the  surface.  If this coating  is success-
 ful, the other processing facilities will have  this coating  installed.
FIGURE 2   Interior of Mill  Creek  During  Clean-Up

    The  lubrication program developed  for  the Louisville  plants has  recently
waf^/H     1° takVnt°  account  some of  the  lessons we  have learned.  As
was  stated  earlier  the  facilities  are  washed down  at least once  a day  with
water  at  about 50-75 psig.  Over  a  period of  time daily  washdown  will  flush
also  conll     ,  bea^^s'   ™e "Y ash  and lime  present in the  atmosphere may
also  contaminate  grease  and oil.  These  contaminated  lubricants  if  left in
Place  act  as  abrasives,  reducing  bearing and  gear life.   To eliminate  the e

?he applcationVof90ne "° * l»bricati™ Sched"le "hich reduces the time  between
the application of grease and increases the frequency of gearbox  oil  changes.
                                    9-30

-------
WINTER OPERATION

    Those  stabilization  plants operating  in  cold climates are  faced with  the
additional hardship of winter operation.  Cold weather hampers processing plant
operation  by  reducing  the filterability of the  thickener underflow.  In addi-
tion, the  problem of belt and  idler  freezing reduces conveyor  belt and idler
life.  The clean-up of  outdoor conveyors is  also  a problem since  the  use of
high pressure water hoses increases the possibility of ice-up.

    Cold weather  further inhibits the  fly  ash/lime reaction.  This  particular
problem  is  resolved  by stockpile management  techniques  which utilize the pro-
duct's  insulating properties.   With  proper   techniques,  even  if  the outside
temperature is below freezing,  temperatures approaching 100°F may be  maintained
within the pile.  Periodic monitoring of these temperatures,  however,  is neces-
sary in  order to  insure  sufficient  reaction before landfilling  the product  and
to guard against  over-reaction  which would result in the  creation of  boulders.

                                    SOLUTIONS

    In the design and construction  of the processing plant  that services  the
Cane Run Unit #4/5,  several modifications were  made  and improvements realized
over the Cane Run #6 design.   This  unit  is  about 3/4  mile from the  thickeners
and  approximately 1 mile from  the  fly ash  silos.   It was decided  to include
100% redundancy for  the  filtration  system,  and a single  surge tank was instal-
led.  No redundant conveyor  systems  or redundant mixers  were installed.   Oper-
ation of the  surge  tank  has enabled LG&E to  schedule operations  so that most
of  the   maintenance  is  performed  during  the  daylight  shift.    It  has  also
smoothed the  flow variations  which have  been seen in  the Cane  Run #6 unit.
Underflow  feed from the  surge tank varies  approximately  2 to 3% over the dura-
tion of  a  pumping shift.  We are able to process at a more consistent rate  and
generate product  with a more consistent solids content.

    The  processing plant  servicing  the  Mill  Creek Units  #1,  2,  3 &  4   was
designed with almost  100% redundancy.  The only common  element  at  this plant
is  the  large  surge  tank.   Redundant fly ash and  lime systems  are included in
this facility as  well  as  separate conveyors  and mixer  trains.  Separate mixers
allowed  the plant to schedule  routine  cleaning  and maintenance.   This insures
efficient  cleaning  of each  mixer since  the  cleaning crews  are able  to take
their time and do a complete job.

    The  Mill  Creek  facility is also equipped  with an on-site laboratory which
is manned  during  the  daylight shift.  This provides  immediate  feedback to  the
processing  plant  operator   so  that  variances  in  design  mix   are   discovered
quickly  and are corrected.   Because  of  this LG&E is assured of efficient util-
ization of lime and material with the proper  in-place qualities.

                              OTHER  RECOMMENDATIONS

    The  utility  just  beginning  to  operate  its  first   FGD  sludge   processing
facility should  realize  that many  elements  will effect  operating  efficiency.
At LG&E  several types of filtercloth were utilized  based  upon filter-leaf tests
performed  by  CSI's on-site  advisors.  The optimum  cloth  was chosen  based upon
                                      9-31

-------
filtercake production  rates  and properties.   At Mill  Creek we  have  utilized
several types of filtercloth on different  filters.   Each type is used  for dif-
ferent properties  of  the  thickener  underflow.  We  have done  this because we
understand that there  are  variabilities in the  coal burned  and  the operation
of  the scrubber  which  must be  taken  care   of by  the  sludge  stabilization
facility.   Since  all  material must  be processed  by the  sludge stabilization
facility,   this facility  must  be  adaptable  to  the variances  which  will be
encountered  in the operation  of the  power  plant,  and  it  must  be  ready to
accept upsets  with-  out causing plant  downtime.  Therefore, we  have  tried to
adapt  the plants  to  accept these  variances.   There  are,  of  course,  some
variances  which go beyond  the adaptability  of  the processing  facility.   We
have been able  to  handle these  through the use  of  stockpiling  and  landfilling
techniques,  most  notably   microencapsulation,   which  we  have   successfully
employed at the Louisville  facilities.

                                    LANDFILL

    The above  topics  discussed the processing  facilities  required to process
the FGD scrubber  sludge  into a  stabilized  material.   However, to complete pro-
perly  the Poz-0-Tec  system,  the processed  material must  be  conditioned  and
placed within  a permitted  landfill.  LG&E  also contracted with Conversion Sys-
tems  to  provide  a Waste  Management  Plan for  the  disposal  of  the  material
produced  by  their  processing  facilities.   This plan   included  all necessary
information and procedures for LG&E  to obtain landfill  permits  from the State
of Kentucky  for both  of  their  generating  stations.   The plan included  proposed
material  characteristics  (unconfined   comprehensive  strengths,  permeability,
leachate), landfill  and process plant  quality control  programs,  and  landfill
site development  plans.  With  this  information, LG&E  was  able  to  obain land-
fill permits at both power stations.

    At  this  time, landfills  are  being developed  at both  sites  in  accordance
with the  Waste Management  Plan with monthly  inspections by  the  State of Ken-
tucky.  The  permitted  sites  include borrow areas (areas where soil  was removed
to construct the station),  flood plain  areas  of  the  Ohio River  and  other areas
not  suitable  for  the  construction  of power plant  structure.   One of the areas
not  suitable for  utility construction  is  currently  being reclaimed  by  landfil-
ling  with Poz-O-Tec  at  the Mill  Creek generating  station.  Plans have been
made to construct  a  limestone  crushing plant upon completion of this  landfill
site.

GROUND AND SURFACE WATER MONITORING

     Included in the Waste  Management Plan  were programs for ground  and surface
water  monitoring.  The  groundwater  monitoring  is  accomplished  by quarterly
sampling  of wells  located near the landfill sites.  These locations were  estab-
lished  by the  groundwater  flow which  flows  toward  the Ohio  River.  Efforts
were made  to  locate one  monitoring  well upstream of each landfill  site and  two
or  three  wells  downstream  of  each  landfill  site   in  order to  determine  the
effect  of the  landfill upon  groundwater  quality.  Initially,  the wells  were
sampled to obtain a  composite standard against  which  subsequent water quality
data could be  measured.  Once  the  landfill operation started at each site,  the
up-gradient  wells have  been used  to   monitor  any  changes in  the quality  of
                                      9-32

-------
groundwater  passing  beneath  the  disposal  area.   Before  each  sampling,  the
static water  level  in each monitoring well  has been recorded.   This  data has
been  used  to determine  the annual  fluctuations  in  the  groundwater elevation
for the  disposal area,  since  the placement  of processed material  within the
disposal area should  not affect  the  existing dominant direction of  the ground-
water  flow.   Up  to  this  time  the  water  quality  analysis  have  indicated no
change  in  the   overall   groundwater  quality  and  dominant   direction  of  the
groundwater flow.

    The  obtaining  of the  monitoring  well  samples  has  been  simplified  as each
well  has had  a  submersible pump  installed.  This pump  is energized  by  a port-
able  generator  which  enables  the  laboratory technicians to  sample each well in
less  than  one  hour.   Originally, the wells were  sampled with a  siphon pump
located  on  the  back  of  a pick-up truck.  Priming  of each well posed problems
due to the wells  isolated location which increased  the  time  required to sample
each  well.   After  obtaining  the samples,  they are preserved  and  sent  to the
CSI's Technical Center for analysis in accordance with the permit guidelines.

    All  landfill  run-off caused  by  precipitation  is  controlled  and  routed
through  a pond  constructed near  each  landfill site.   The ponds  were  designed
to collect the  run-off and discharge through a  sandfilter.  Sizing of the ponds
and associated  draining   channels were  in  accordance with the  Louisville area
one hundred  year rainfall criteria.  Development of  the landfill incorporated
drainage  designs to  control  all  run-off  during  construction.  Efforts  have
been  made daily to control sediment due to errosion  as an  excessive amount of
sediment in the run-off  increases the maintenance  of the pond  and  sandfilter.
Up  to this  point,  the sandfilter has had to  be  cleaned  approximately  twice  a
year.  Run-off  samples have been obtained  at weirs  installed at each sandfilter
discharge, after each rainfall.

QUALITY CONTROL

    In order  to continue  to  meet the parameters of  the  landfill permits, the
Waste Management Plan also provides  a  quality control program for each proces-
sing  facility and  landfill site.   These  programs are maintained by  LG&E's lab-
oratory  personnel  with CSI's  Advisory Staff  personnel  providing  any technical
advice needed to maintain these programs.

    Daily process plant  quality  control  activities  include  sampling of all the
raw materials sent  to  the process  plant  and  determining the  proper addition
rates  for  the  process plant  operators  to  utilize.  The operators use  these
recommendations  to  make   minor set-point changes  on automatic controllers.  At
times, major  set-point changes are required  when one of the  raw materials is
received out  of specification.   This  usually occurs when the filtercake  solids
are low  due to  low  solids received from the thickeners.   Other daily routines
include  sampling of  the  stockpiled  material  to  determine  its  curing   rate.
This  analysis  also  provides  information  on  back-shift  production since the
laboratory personnel  are  only assigned to  the  process plant during the daylight
shift.   This  minimum coverage of  quality control  is satisfactory  as  long as
the process plant  personnel operators  operate the plant in the automatic  mode.
                                      9-33

-------
Whenever non-specification  material  is produced due  to operator error,  equip-
ment breakdowns,  or  raw material  availability,  the  quality  control  personnel
recommend set-point changes to produce an above specification material to blend
with the below  specification material.  Cooperation between both the  operators
and  quality  control technicians  has  resolved most  deviations  smoothly.   (See
Figure 3)
 FIGURE  3  Exterior of Cane Run #4/5 Showing Stockpile Management Practices

     During  the  start-up  of  each  facility,  ratio tests were performed  to deter-
 mine  minimum  addition  rates  of  fly ash and lime.  It was  found that  for  these
 particular  installations, the fly  ash  addition  rate  could be reduced  to 60% of
 the  design  with a small  increase  in the  addition of lime and  still meet the
 minimum material characteristics required  in the landfill.
     The landfill quality control  program  is  also  maintained by LG&E personnel
 with assistance from  the  CSI Advisory Staff personnel.   The  same  technicians
 that perform  the  process  plant  analysis  also maintain  the landfill  program
 since the  frequency of  testing  is dependent  upon the amount of processed mat-
 erial produced.  In-place  landfill density tests are performed  for every 10,000
 to  12,000  tons of  material produced.   These   tests  are used  to control  the
 landfill  placement  technique  and  to  insure  that the  material  is  placed  and
 compacted  to  a  minimum  density   specification.    When   this  minimum  density
 specification has been achieved,  it has  been  found  the  landfill  strength per-
 meability and leachate criteria are met.  These three criteria are continually
 checked as landfill test cylinders  are prepared  routinely  for testing  at  the
 CSI's  Technical  Center.   These  test cylinders  are molded  to  the landfill
 in-place  density test results  by  the  LG&E technicians  and,  after  curing,  are
 sent to  the  Technical Center for  analysis.   Results from these  landfill test
                                      9-34

-------
cylinders have  shown that as long  as  the processed material of the proper mix
is properly conditioned, placed and compacted, all permit guidelines concerning
the materials physical characteristics have been met.

LANDFILL OPERATIONS

    LG&E contracted  a  local  heavy-equipment contractor to prepare the landfill
sites and place the  processed material in accordance with the Waste Management
Plan.   Since  the  landfills  are  located  on the  utility  property next  to the
generating station,  the contractor  has been able to utilize offthe-road equip-
ment.   The  size of  his fleet  of hauling  equipment has  grown as each  of the
three processing  facilities started  up.   At  this  time,  he  is using  both 35
yard and 50 yard capacity trucks.   (See Figure 4)  The larger trucks have posed
a problem as the roads bearing this traffic must be  well-constructed to support
their gross weight.   In  the  landfill  itself,  additional effort must be made in
the site development to eliminate truck  traffic  on recently placed material as
it  will not  initially  support  this  weight.   Since  curing rates  are  slowed
during winter operations, the active landfill must be  increased accordingly.
FIGURE 4   Exterior of Mill Creek showing Poz-O-Tec
       being  loaded for  transport to Landfill

-------
    Low-ground pressure bulldozers place and  compact the material with a  self-
propelled vibratory roller providing additional compaction and surface sealing.
Maintenance repairs  on  the landfill  equipment  is actually  less than the con-
tractor's normal equipment  repairs  on earth projects.   This can be attributed
to the fact that although  the  processed  material has abrasive additives,  it  is
less abrasive  than the  sandy  soil  in  which his equipment  usually operates.
The  contractor  operates  five  days  a week,  approximately  eight hours  a day.
Depending on which site he is actively working,  his hauling  rate varies between
200 to 500 tons per hour.

MATERIAL CHARACTERISTICS

    The  LG&E  material characteristics are  based  upon  the  type of  coal  used,
type of  scrubber  and agent, fly ash  collector  size  and  collection efficiency
and  available  lime supply.  The design  specifications of process  plant  mixes
and  ultimate  landfill characteristics were  achieved without  any major changes
from the original  Conversion Systems,  Inc.  equipment design.  We did, however,
lower the filtercake solids content and production range, as it  was found dur-
ing  start-up  at all  three plants  that  the  filtercake  could be  produced too
dry.  Vacuum filter cloth changes corrected  this problem.

    Annual core  borings of  the  landfills   have  been performed  to  insure that
the  ongoing quality  control results  are  accurate,  and  that what  is actually
placed is meeting  all  the minimum criteria.  All  results have  shown that when
the processed material is properly conditioned,  placed  and  compacted, the per-
mit  guidelines are satisfied  by  landfill disposal using  the Poz-O-Tec system.
Typical data from  these borings are shown in Table #1.

                                    TABLE I
                       TYPICAL 28 DAY MATERIAL PROPERTIES

                                       Mill  Creek           Cane Run

    Unconfined Compressive Strength     75 psi              48 psi

    Permeability                        2.5  x 10~6          1.01 x 10~6
                                      9-36

-------
COAL WASTE UTILIZATION IN ARTIFICIAL REEF CONSTRUCTION




    J. H. Parker, P. M. J. Woodhead, D. M. Golden

-------
         COAL HASTE UTILIZATION IN ARTIFICIAL REEF CONSTRUCTION

           by:  J. H. Parker and P. M. J. Woodhead
                Marine Sciences Research Center
                State University of New York
                Stony Brook, New York  11794
           and
                D. M. Golden
                Electric Power Research Institute
                P.O. Box 10412
                3412 Hi 11 view Avenue
                Palo Alto, California 94303

                                  ABSTRACT

     The technology of coal combustion has improved greatly over the last few
decades.  One result of the elaborate emission control equipment required on
coal plants to meet stringent air quality requirements is the large volumes
of wastes that must be utilized or disposed of safely.  Flue-gas
desulfurization (FGD) sludge and fly ash may be produced at the rate of as
high as 1,000 tons per day at a typical coal fired power plant.  Waste
disposal, especially in urban coastal areas, has become a major obstacle to
conversion to coal combustion for generating electricity.

     To assess a possible solution to the waste disposal problem, 500 tons of
FGD sludge and fly ash were stabilized into blocks and placed in the ocean as
an artificial reef.  Previous laboratory investigations had indicated that no
toxic chemical or physical effects should occur in the marine environment.
After three years in the sea, the coal waste blocks support a diverse
community of reef fish and invertebrates and have maintained their structural
integrity.  No adverse environmental effects have been detected.

     Assuming that the coal waste blocks continue to be environmentally
acceptable in the marine environment, the engineering and economic
feasibility of this method of disposal should now be confirmed.

INTRODUCTION

     The increase in the combustion of coal for generating electricity has
resulted in the improvement in the technology required to burn coal cleanly.
Electrostatic precipitators are able to remove a? much as 99% of the
particulate ash and flue gas desulfurization (FGD) scrubbers reduce the
sulfur oxide emissions to acceptable levels.  A problem, though, is created
-- waste disposal.  With a large coal-burning power plant producing more than
1000 tons of fly ash and FGD sludge each day, coastal metropolitan areas and
areas with shallow ground water tables may lack sufficient land to safely
                                    9-37

-------
dispose of these wastes.   Ocean disposal could become an option if the
materials can be stabilized to prevent any harmful effects to the marine
envi ronment.

     In September, 1980,  500 tons of stabilized FGD sludge and fly ash, in
the form of 15,000 hardened blocks (20 x 20 x 40 cm), were placed in the
Atlantic Ocean as artificial reef substrate.  Two years of prior laboratory
and small-scale field investigations sought to detect any possible adverse
effects to the marine community as well  as any physical deterioration of the
blocks themselves.  With  the manufacture of the 15,000 blocks and their
placement in the sea, a three-year monitoring program continued to
investigate the environmental acceptability of constructing artificial reefs
with coal wastes.

     Several types of biological surveys were conducted to compare the
success of the"coal waste reef with that of an existing reef of rock and
building rubble.  The fish community was sampled with traps to estimate
population densities.  Test organisms were placed at the reef to detect
possible uptake of block  components.  The structural integrity of the blocks
was monitored using ultrasonic and compressive testing.  To understand any
chemical changes occurring and to allow estimation of the lifetime of the
blocks in the sea, the chemical composition of the blocks was determined
throughout the study.

     As these investigations continue to provide positive results, the
engineering and, ultimately, the economic feasibility of disposing of coal
wastes in the ocean as artificial reefs is now being determined.

BLOCK PRODUCTION AND REEF CONSTRUCTION

     The 500 tons of FGD  sludge and fly ash were obtained from two power
plants, the Conesville station of Columbus and Southern Ohio Electric (CSOE)
company and the Petersburg station of the Indiana Power and Light Company
(IPALCO).  The Conesville material was produced with a ratio of fly ash to
FGD sludge of approximately 3:1 while the IPALCO material had a fly
ash:sludge ratio of about 1.5:1.  The chemical compositions of the components
of each material as well  as the stabilized blocks were determined and
reported in Parker _et _al_. , (1981).

     The actual block production utilized standard concrete block industry
equipment at a commercial plant in Pennsylvania (Figure 1).  Very few
modifications had to be made to accommodate the mixture of fly ash, sludge,
lime and cement which was somewhat wetter then conventional concrete
materials.  The 15,000 blocks were all cured for 24 hours in a 'wet steam1
kiln at about 65°C and finally trucked to a dock in New Jersey.  A bottom
opening barge transported the blocks to the project site (Figure 2) and
released them to settle on the bottom in 20 m depth of water.

BIOLOGICAL  INVESTIGATIONS

     To determine the success of the coal waste blocks as an artificial reef
substrate, it was necessary to measure colonization of the block surfaces by
                                    9-38

-------
          F6D
         SLUDGE
         Figure 1.  Schematic of block factory.
           Figure 2.  C-WARP project site.
epibenthic fauna and to determine the population of fish taking up residence
in the "nooks and crannies" provided by the reef.  Most importantly, analyses
were made on reef organisms to detect possible uptake of toxic components
from the blocks into the food chain.  Special test bricks of coal waste were
set out and returned to the laboratory periodically for identification of
species and measurement of coverage (Figure 3 and Table 1).  Close-up
underwater photographs were also taken throughout the entire study period of
the same areas on tagged blocks to document changes in the epibenthic
community.
                                     9-39

-------
          lOOr
                                               TOPS
               SONDJFMAMJJASONDJ  FMAMJJ ASOND
                               .,981	1	1982	
                 -1980-
              o CONESVILLE, D CONCRETE, A IPALCO
          Figure 3.  Coverage  by epifaunal colonizers.
TABLE 1.   EPIBENTHIC SPECIES GROWING  ON SURFACES ON  REEF BLOCKS,  1980-81,
Porifera
Sponqe, unident.
Colenterata
Metridium senile
Clytia sp.
Obel ia dichotoma
Sertularia cupressina
Astranqia danae
Bryozoa
Aetea .sp.
Bugula turrita**
Callopora aurita
Cribulina jui n c t a t a
Electra hastingsae
Electra jrijosa
Schizoporel 1 a unicornis**
Microporella sp.
Mol 1 usca.
Anomia aculeata
Anomia simplex
Mytilus edulis**
Zirphaea crispata
c.pisula sp.
Onchidoris sp.
Acanthodoris pilosa
Eubranchus exiquus
Annel ida
Harmothoe extenuata,
Phyllodoce arenae
Pol vdora social i s *";"
Syllis sp.
Nereis grayi
Sabellaria vulqaris**
Asabellides oculata
Arthropoda
Balanus crenatus**
Caprella linearis
Pokoqeneia inermis
Edotea triloba
Unciola irrorata
Cancer irroratus
harpacticoid copepods
Echinodermata
Asterias rubens
Tunicata
tunicate unident.
                                    9-40

-------
     Fish traps were used to capture, tag, and
coal waste reef and a long established fishing
rubble (Fire Island Reef) nearby.  Because the
was similar to that of the part of Fire Island
comparison of fish populations was conducted.
coal waste reef increased from approximately 2
reef had been established for one year, to 8.5
1981 and 1982 the cunner population densities
                         release fish from both the
                         reef of rock and building
                         area of the coal waste reef
                         Reef being studied, a
                         The density of cunner on the
                         6 fish per m2 in 1981 when the
                         fish per m2 in 1982.   During
                        at the Fire Island Reef
remained steady at approximately 7.5 fish per m2.  These results indicate
that, in less than 2 years the coal waste reef had achieved "carrying
capacities" for fish comparable to that of the existing artificial  fishing
reef.

     Initial toxicity experiments in the laboratory tested sensitive marine
organisms, a diatom and winter flounder eggs and larvae (Parker et al.,
1981), for adverse effects due to exposure to various concentrations of  an
elutriate made from powdered coal waste block material.  The lack of
significant effects was further investigated by field tests at the reef.
Mussels, Mytil us edulis, widely used indicator organisms for toxic uptake
studies, were placed at both the coal waste reef and at a separate control
site.  Again, analysis of mussel tissue showed no significant increases  in
potentially toxic trace elements after 9 months exposure on the reef. A
laboratory experiment using mussels exposed to elevated concentrations of
suspended coal waste material yielded similar results (Table 2).  A spurious
increase in copper was not seen at higher concentrations of coal waste.   A
doubling of iron concentrations in the mussels would not be expected to
create any toxic effects.  Fish were tagged during the three-year study
period and recapture data indicate that these fish remained resident on  the
same reef throughout the complete study.  Fish samples from the reef have
been taken for tissue analysis of trace elements  and should represent the
most realistic case to detect possible uptake of block components.
   TABLE 2.  TRACE METAL CONTENT OF SOFT TISSUES FROM MUSSELS EXPOSED TO
                SUSPENSIONS OF POWDERED CONESVILLE COAL WASTE
CONCENTRATION OF
   COAL WASTE
     (mg/A)	
  Cd
   TRACE ELEMENTS (ppm,  dry weight)

   Cu      Fe      Mn      Ni       Pb
                                   Zn
        0
    (control)

       89

      178

      267
1.28

1.42

1.42

1.24
 8.47

 8.98

13.45

 7.87
 56.4

102.4

101.5

106.4
6.48

4.43

4.03

5.60
1.46

2.29

1.70

5.32
2.31

2.58

2.93

2.76
107.1

101.2

116.5

 97.4
Values are averages of values for 5 replicates.
Underlined values are significantly different than controls at p <_ 0.5  level.
                                     9-41

-------
PHYSICAL AND CHEMICAL STABILITY

     Although the blocks'  surfaces successfully support a wide diversity of
epibenthic colonizers, it is important to determine how quickly the blocks
might deteriorate in the sea.  Previous laboratory experiments indicated a
gradual increase in compressive strength and density over time.  After the
placement of the reef, blocks were tagged for repeated ultrasonic
(non-destructive) testing in the sea.  Results have provided a good
correlation between the series of ultrasonic data and compressive strength
measurements when the blocks were finally returned to the laboratory after
extended exposure at sea.   Both sets of results indicate a gradual increase
in strength seen in Figure 4 (Parker _et aj_., 1982).
              10
                                                               1800
                                                               1500
                                                               1200
                                                               900
                                                               600
                                                               300
                         100
                                    200
                                   SOAK DAYS
                                               300
                                                          400
           Figure 4.  Compressive strength of Conesville (A) and
                      IPALCO (0) blocks in the ocean.

     The coal waste blocks undergo cementitious processes similar to those
 occurring  in concrete.  These reactions normally continue long after initial
 formation  and it is important to understand what changes might occur due to
 protracted exposure to seawater at 3 atmospheres of pressure.  Blocks
 returned from the sea were analyzed for any changes in mineralogy.  As
 exposure time in the sea increased, mineralogical changes were found to occur
 in  the  surface  layer of the blocks.  In Figure 5, mineralogical results
 indicate that calcium sulfite hemihydrate and gypsum are major and minor
 components,  respectively, in the core of the blocks. In the surface (1-2 cm)
 layer,  however, a conversion has occurred making gypsum a major component.
 Although this reaction results  in an overall positive volume change, the
 inherent porosity of the material accommodates any internal pressures that
 would otherwise develop.

     Little, if any, change has occurred in the composition of minor and
 trace elements.  However, significant surface losses of calcium by
 dissolution  have been accompanied by an enrichment in magnesium.  Laboratory
                                     9-42

-------
          CORE
          SURFACE
HEMIHYDRATE
 CaS03 --j-HjC


GYPSUM
 CaS04 -2H20


HEMIHYDRATE
                 GYPSUM
                              m
                              m
                              m
                    CONESVILLE

                   m      M  t    M     M


                          t  -
                                                        m
                   M
                           mm    M
                                                        m
M
                                    14       32 38   49    61
                                       WEEKS OF  SUBMERSION
                                                                80
          CORE
HEMIHYDRATE
 CaS03 --^-HjO


GYPSUM
 CaS04 -2H20
                               M
                              m
          SURFACE
                 HEMIHYDRATE    M
                       IPALCO

                   M       Mm


                   t       t  t



                   M       Mm
M
                 GYPSUM
             M
                                    m
                                            t  m   M    M
                                                                m
                               i—
                               0
                                    14      32 38   49   61
                                       WEEKS OF  SUBMERSION
                                               80
          Figure 5.   Mineralogical changes occurring  in  the  blocks.

experiments have been  conducted to measure the dissolution rate of calcium
and, using these results  in  a  model,  to estimate the overall lifetime of the
blocks.  Preliminary results indicate that,  after 30 years, the diffusion
zone would only penetrate 2-3  cm into the blocks.

ENGINEERING ASPECTS

     After preliminary tests making blocks with machines at research
facilities of the Besser  Company, Alpena, Michigan, the 500 tons of coal
wastes were processed  at  a commercial concrete block plant utilizing standard
"off-the-shelf" equipment.   To move a step closer to the full-time disposal
of coal wastes  from a  typical  power plant using stabilized waste block
production techniques  for reef construction, the capability to handle as much
as 1000 tons of fly ash and  FGD sludge per day must be demonstrated. Figure  6
illustrates a simplified  system for the ocean disposal of coal waste blocks.
The process flow is indicated  in Figure 7.  Calculations were made for  a
hypothetical power station equipped with two 500 MW units, and a 45% load
factor over a 30 year  life.   The details were taken from the EPRI FGD Sludge
Disposal Manual, 2nd Ed.  (1980).  The fuel was an eastern coal with 3.5%
                                      9-43

-------
                     TEMPORARY
                   SLUDGE FILTERCAKE
                     HOLDING BASIN
                                               POWER PLANT
                                                  HrORATED LIME BIN
                    BLOCK PRODUCTION	

                    STEAM KILNS	
                                            BLOCK PLANT
                                   |TUG| BARGE] BARGE LOADING FACILITY
                                                 OCEAN
                                              DISPOSAL SITE
                Figure 6.  Block production design  for  ocean disposal.

sulfur and 12% ash content.  Table 3 presents  the  estimations of the weights
of coal waste requiring disposal daily.

     For estimation of the capacity of machines  to  make blocks from sludge
and fly ash, it has been assumed that 4%  lime  and  3%  cement would be added to
stabilize the waste mix and that the material  would be  consolidated to at
least 1.76 g/cm3 (110 lb/ft3) as delivered from  a  block machine.  The block
size was that of a standard construction  block,  nominally 8 x 8 x 16 in
(20 x 20 x 41 cm), but with an actual volume of  0.017 m3 (0.59 ft3) and
weighing 29.4 kg (65.2 Ib).  Larger block molds  are also readily available
for use in conventional block machines.

     The calculations of materials processing  capacity  (Table 4) indicate
that a single V6-12, or two V3-12 block machines,  operating at a rate of
eight block-forming cycles per minute and with an  efficiency of 90%, have the
potential  to fabricate the entire daily waste  production of the power plant
into blocks for placement in the ocean.
                                    9-44

-------
                                 PROCESS SCHEME

                                   V6 SYSTEM
FLY ASH
SLUDGE •
BOTTOM

to BAT
CEMENT
"uiMr:_-. u r\
^^ LJM 1 ^ i i i i i« w^^ 	 1 Ip W
1 LIME
BLENDING









MOL

1
DING ^<

i
ACCUMULATION

x 	 *- LOA
CONV

i
DING
EYING






, PALLETS
RACKS CURING '
/







CONVEYING

1
^ 	 UNLOADING

HFDAI
i I
1





ITTIMr V
1
CUBING





YAR
DELI\
1
1
DING
/ERY

USE
                  Figure 7.  Process flow for coal waste
                             block production.
     Although the location and layout of an existing power plant would
control how the block facility was constructed, a design that minimizes
handling of the blocks in transfer to the ocean-going barge would greatly
reduce costs.

REEF PLACEMENT ASPECTS

     For a power plant producing 1000 tons of coal wastes per day, as many as
31,000 blocks could require placement in the ocean.  Suitable sites must be
chosen for the placement of the blocks if the benefits of the artificial
reefs are to be fully realized.  Economics factors will dictate the maximum
distance from the loading facility but bottom type and accessibility to
recreational fishermen will also be considered when regulatory agencies are
asked to approve the designation of disposal/construction sites.  The hard,
sandy bottom and number of inlets off Long Island, New York make this area
very suitable for artificial  reef construction.  With several hundred square
miles of ocean bottom between the depths of 40 and 90 feet, sites can be
chosen which would minimize interference with commercial fishing operations,
                                     9-45

-------
Table 3   ESTIMATION OF AVERAGE WASTE QUANTITIES FOR 3.5% SULFUR  EASTERN  COAL
                (FROM EPRI REPORT CS-2009 VOL. 2, NOV. 1981).

ASSUMPTIONS (for a two 500 MW unit generating station)
COAL

     Btu/lb
     Ash content, %
     Fly ash/bottom ash
     Sulfur content, %
     Annual consumption @ 45% capacity factor, metric ton

AIR QUALITY CONTROL SYSTEM

     Upstream fly ash removal efficiency, %
     Overall fly ash removal  efficiency, %
     S0~ scrubbing reagent
     S0? removal, %
     Stoichiometry (CaO/SO? removed), %
     CaS04.2H20/CaS03.±H2(T
     Thickener underflow solids, %
     Dewatered sludge solids, %

ASH PRODUCTION DAILY

     Total  ash produced
     Total  fly ash produced
     Fly ash collected in precipitator

SLUDGE PRODUCTION DAILY

     Fly ash collected in scrubber
     CaSO,.2HnO
     Weignt of excess and unreacted lime reagent
     Weight of dry sludge solids
     Weight of water, at 60% solids content
     Total weight of wet sludge, at 60% solids

 STABILIZATION ADDITIVES

     Lime, 4%
     Cement, 3%
WASTES TO BE PROCESSED DAILY

      Fly ash + sludge solids + lime + cement
      Total dry weight
      Water in sludge at 60% solids
      Total wet weight
                                                                  12,500
                                                                  12
                                                                  80/20
                                                                  3.5
                                                                  1,430,000
                                                                 99
                                                                 99.8
                                                                 Lime
                                                                 90
                                                                 110
                                                                 50/50
                                                                 35
                                                                 60

                                                            METRIC TONS

                                                                 470
                                                                 376
                                                                 373
        3
        132
        398
        108
        642
        427
        1069
        41
        30
   METRIC TONS

373 + 642 +41+30
      1086
       427
      1513
                                     9-46

-------
  TABLE 4.   CALCULATION OF COAL WASTE BLOCK PRODUCTION RATES BY ONE V6-12
                  SUPERPAC OR TWO V3-12 VIBRAPAC MACHINES
                    (FOLLOWING ASSUMPTIONS IN TABLE 1)
Block unit size, nominal:20.3 x 20.3 x 40.6 cm (8 x 8 x 16 in)
Block unit volume:
Block forming cycle:
Block wet density:
Weight of 6 blocks:
Block forming rate:
Hourly block production:
     0.017 m3 (0.59 ft3)
     One 6-12 block machine

or two V3-12 block machines

     1.76 g/cm3 (110 lb/ft3)
     176.7 kg (391 Ib)
     8 cycles per minute
     176.6 x 8 x 60
                                   At 90% efficiency
                    Ash & sludge (dwb) process rate
1 pallet of 6 units
0.100 m3 (3.56 ft3)
2 pallets of 3 units
0.100 m3 (3.56 ft3)
84,805 kg/hr
84.8 metric/hr
(935 short ton/hr)
76.3 metric ton/hr
(84.2 short ton/hr)
76.3 x 71.7% solids
54.7 metric ton/hr
(60.3 short ton/hr)
From Table 1,
Coal wastes to be processed daily (dwb) = 1086 metric ton (1197 short ton)
Block production time per day           = 1086 metric ton/day = 19.8 hours
                                          54.7 metric ton/hr
prevent damage to shellfish areas, and would not "fill  in the ocean."  In
fact, assuming the production of 31,000 blocks each day from a power plant
and allowing a doubling of volume when released on the bottom due to random
stacking, over 30 years of blocks could be contained in a reef covering 1
square mile and rising only 10 feet above the bottom.  Water depth would
control how high the reef could be constructed, due to potential  hazard to
navigation.  Other potential  uses of the blocks include rehabilitation of old
sand mining areas and enhancement of fishing in areas that, historically,
have had soft bottoms.
                                     9-47

-------
THE ECONOMIC VIABILITY OF COAL WASTE REEFS

     This program has demonstrated the technical feasibility of processing
power plant coal  combustion by-products into blocks by using conventional
concrete block making equipment.  The environmental acceptability has been
demonstrated in the laboratory and field investigations.  To answer the big
question on whether this concept is commercially viable, the Electric Power
Research Institute commissioned Michael Baker, Jr., Inc. to do an evaluation
of this technology from a hypothetical coal plant located in the New York
Metropolitan area.  The reef construction system would consist of four parts:
(1)  The block product plant with storage facilities; (2)  a barge loading
facility;  (3) transportation to the reef site for placement, and (4)
monitoring  the site.  The engineering-economic assessment indicated that a
hypothetical 600-MW power plant would produce nearly 300,000 tons of
by-products each year.  The reef construction system would cost about $45/ton
to build and operate.  This cost is double the amount that most power plants
spend on by-product disposal when land is available nearby.  In highly
urbanized  areas where land is not available, the disposal costs are nearly
the  same.   Therefore the potential for reef construction as a viable
alternative to traditional waste disposal technologies looks promising and
can  be  considered by coastal utilities in waste management planning.

     The comparative economics of construction of coal waste artificial
fishing reefs versus conventional land disposal practices would be
drastically changed  if a bill now in the 98th Congress becomes law.  The
proposed legislation  (H.R. 3474) is called the National  Fishing Enhancement
Act  of  1983.  The legislation would establish national standards for the
construction of  siting of artificial reefs in the waters of the United States
in order to enhance  fishery resources and fishing opportunities.  To
stimulate  the construction of artificial fishing reefs,  the proposed law
provides a tax credit equal to  the excess costs of reef  construction over the
usual disposal costs.  This bill if enacted would put artificial reefs on an
equal footing with  conventional disposal costs anywhere.  The reason for the
congressional  interest  in artificial reefs is that fishery products provide
an  important  source  of protein  and industrial products for U.S. consumption,
yet  U.S. production  annually falls $3.2  Billion short of satisfying demand.

 REFERENCES

 Knight, R. G., E. H.  Rothfuss,  and K. 0. Yard,  1980.  FGD Sludge Disposal
Manual, Second Edition,  Electric Power Research Institute, CS-1515,
 Palo Alto, CA.

 Parker, J. H., P. M.  J.  Woodhead,  I. W.  Duedall, and  H.  R. Carleton, 1981.
 Coal Waste Artificial  Reef  Program,  Phase  3,  Volume  2.   Electric Power
 Research Institute,  CS-2009,  Palo Alto,  CA.

 Parker, J. H., P. M.  J,  Woodhead,  I.  W.  Duedall, and  H.  R. Carleton, 1982.
 Coal Waste Artificial  Reef  Program,  Phase  4A.   Electric  Power  Research
 Institute, CS-2574,  Palo Alto,  CA.
                                      9-48

-------
SOLID WASTE ENVIRONMENTAL STUDIES AT ELECTRIC
           POWER RESEARCH INSTITUTE

                I. P. Murarka
         Presented by Karen Summers

-------
                       SOLID WASTE ENVIRONMENTAL STUDIES
                      AT  ELECTRIC POWER RESEARCH INSTITUTE
            by:    Ishwar  P.  Murarka,  Ph.D.
                  Environmental  Physics and Chemistry Program
                  Electric  Power Research Institute
                  Palo  Alto,  CA  94303
                                    ABSTRACT
    Solid Waste Environmental Studies (SWES) at Electric Power Research
Institute (EPRI) is a comprehensive research project aimed at generating
predictive methods and the essential data bases to evaluate the effect of
disposal and reuse of solid wastes produced from fossil fuel combustion and
flue gas cleanup operations on groundwater quality.  EPRI has developed
detailed research plans and has initiated research in leaching chemistry,
chemical attenuation mechanisms, groundwater transport processes, and the
evaluation of existing geohydrochemical models.  For the next three or four
years, fundamental research in geochemistry and geohydrology is expected to
yield quantitative data on release rates, transformation characteristics, and
subsurface transport of inorganic solutes leached from waste.  Results of the
research will be integrated by improving or developing new predictive methods
and by validating the results with data from operating facilities.
                                  INTRODUCTION
    Do constituents released from land disposal and from reuse of solid
residues by electric utilities influence groundwater quality in the surrounding
area?  What factors control leachability of chemicals?  What solid waste types
are of concern?  What is the environmental fate of leached solutes?  When and
to what extent should technologies be applied to contain leachates?  EPRI's
SWES project was initiated in 1982 to develop useful answers to these  and other
questions.

    SWES is a comprehensive research project dealing with fundamental  studies
in geochemistry and subsurface hydrology to produce methods and associated data
                                      9-49

-------
sets for use in evaluating the effects of disposal and the reuse  of  solid
residues on groundwater quality.  The solid residues most emphasized  in  the
SWES research are coal fly ash, coal bottom ash, mixtures of  the  two,  flue gas
desulfurization (FGD) sludges with fly and bottom ash, and oil ash.
                              GOALS AND OBJECTIVES
    The long-term or ultimate goal of the SWES project is to develop and vali-
date methods (namely geohydrochemical models) for predicting the  fate  of inor-
ganic solutes released to the subsurface environments from utility  industry
operations.

    This long-term goal will be met by first conducting fundamental research  to
define the cause-and-effect relationships of greatest importance  and of highest
uncertainty to the development of predictive methods.  Once the fundamentals  in
the research areas are well understood and data sets are properly developed,
the results are to be integrated into predictive methods (models) and  validated
with field data.

    The near-term concurrent research is to accomplish the following:

    (a)  Evaluate existing geohydrochemical models and computer codes  and
         assemble an interim usable geohydrochemical model(s)

    (b)  Develop quantitative data on the leaching chemistry of solid
         residues

    (c)  Develop quantitative data on the chemical attenuation in the
         geologic environment of inorganic constituents

    (d)  Develop quantitative data on groundwater transport of solutes

    (e)  Develop mathematical descriptions of important leaching  process-
         es, chemical attenuation mechanisms, and hydrological transport
         factors in the migration of solutes in groundwaters.


                               RESEARCH PLANNING


    EPRI's initial research into solid waste environmental studies  began in
1977 when projects were initiated on (1) chemical characterization  of  coal fly
ash (RP1061), (2) evaluation of reproducibility of EPA's extraction procedure
as applied to utility solid wastes (RP1487), (3) comparison of the  physical  and
chemical properties of solid wastes from coal combustion and coal gasification
processes (RP1486), and (4) providing a statistical evaluation of the  variabil-
ity of a representative coal ash sample from an individual power  plant (RP1620).
Until late 1980, the degree of emphasis that should be placed on  solid waste dis-
posal and groundwater pollution potential was undetermined.
                                      9-50

-------
    In  1981 a  planning  study  was  conducted  to  define  research for the SWES proj-
ect.  This work was carried out through  four regional meetings,  simultaneous
use of  questionnaires addressed to  about 260 individuals  followed by a research
planning workshop, and  continuing discussions  in  SWES Advisory Committee  meet-
ings.   As a result, research  plans  for individual SWES contracts have been de-
veloped to accomplish its objectives.  The  bases  for  this  research effort are
described in detail in  two EPRI reports  (Murarka  1982 and  SAI 1982).   Figure  1
provides a compartmentalized  sketch defining the  overall problem of estimating
the release, transport,  transformation,  and fate  of leached  constituents.  To
date EPRI has  initiated  research  contracts  in  the high-priority  research  areas
of waste leaching chemistry,  chemical attenuation of  solutes in  the geological
environment, transport  of solutes to groundwaters, evaluation of existing geo-
hydrochemical  models, and evaluation of  groundwater sampling methods and  relat-
ed field measurements.   Present plans call  for an expenditure of about  $21  mil-
lion dollars over the next six years.
                       RESEARCH  METHODOLOGY  AND  MANAGEMENT
    SWES is divided into  the  following  six  parallel  research efforts:

    1)   Identifying and  quantifying  the  solutes present  in the utility
         industry solid residues

    2)   Developing data  on the leaching  chemistry of  these solutes from
         the solids

    3)   Developing data  on the chemical  attenuation of the solutes in
         the geological environments

    4)   Identifying and  quantifying  groundwater transport parameters for
         solute migration

    5)   Evaluating, improving, and validating methods (primarily geohy-
         drochemical model[s]) that can predict the  fate  of solutes in
         land, surface, and groundwaters

    6)   Evaluating and testing groundwater  sampling and  related
         measurement methods  for use  in the  field environments.

    The implementation of research in the six categories  employs a general
strategy of evaluating the existing state of knowledge through critical litera-
ture reviews, a detailed  specification of experimental designs, and laboratory
and field experiments to  develop the  quantitative results.  An integrated ana-
lysis of the data would then  provide mathematical descriptions of cause-and-
effect relationships for  use  in predictive methods.  This analysis would also
provide data values for input parameters  to  the predictive methods.  While  the
fundamental research continues toward achievement of the  long-term goal, an
interim assemblage of geohydrochemical model(s) will provide usable predictive
methods for the near-term.
                                      9-51

-------
I
Ui
r-o
ESTIMATING THE TRANSPORT, TRANSFORMATION, & FATE OF CONSTITUENTS


INPUTS
WASTE CHARACTERISTICS
• Type
• Mass (Volume)
« Physical Properties
phase(s)
bulk density
permeability (T)
• Chemical Properties
ele composition
contaminant speciation
solubility, teachability
volatility, (KH, PC)
gov. chem. variables
" speaation Cu-'1'* OH - — CuOH -
ion-exchange ^^-^ complexation
R COOH - Me - — R - COOMe - H -
Microbially Mediated Rxns. Other
oxidation-reduction precipitation-dissolution
hydrolysis (chemical weathering)
conjugation pH
alkylation r5"1"r'"'~"~
ring fission, etc. t VN.
nw 0 - lOQ C r S\*V ^
ry^rr -^-(TcS \>C
•^ k^-OH V^-C-o- \ /V^
r

I , /'
-]/


OUTPUTS
HYDROLOGIC
• Peizometric surface
h(x,y,t)
• Flow Velocities
v(x,y.z.t)
CHEMICAL
« Solute Concentration
C(x,y,z,t)
• Solute Mass Flux
F(M/L2T)
INPUT TO
MGT. DECISIONS
• Evaluation of
Alternatives
• Estimation of
Acceptable Loadings
• Assimilative Capacities
• Monitoring Guidance
EJS
Evaluating Potential
for Ground Water
Contamination
INPUT TO REGULATORY
RULE-MAKING
PROCESSES



               Figure 1.  Estimating the Transport, Transformation, and Fate  of Waste  Constituents:  Inputs,
                          Outputs, and  Example Processes*
                          *This figure is taken from EPRI Report EA-2415, Figure  1-2.

-------
    The management and technical direction  of SWES  contracts  are  accomplished
primarily through the EPRI project manager.  However, because  of  the  nature  and
magnitude of this project, a technical management contractor  is added to  assist
the EPRI project manager.  In addition, an  SWES  Advisory  Committee  has been
formed to advise the EPRI project manager on the SWES research.   This advisory
committee is composed of seven utility industry  scientists, six scientists from
academia and government agencies, two scientists from other EPRI  divisions,  and
the EPRI project manager as the chairman.
RESEARCH CONTRACTS
    SWES now has seven contracts  to  conduct  the  parallel  research  (fig. 2).  A
short description of the scope of work  for these contracts is provided in the
following sections.
Geohydrochemical Models Evaluation and Development
    This was the first SWES contract  awarded.   The  research  in  this effort has
been divided into three phases:

    •    Phase I: The contractor  is evaluating  several models (computer
         codes) on geochemistry,  geohydrology,  and  microbiological pro-
         cesses for their ability to  predict  the  fate of constituents in
         groundwater-  This phase of  research is  to be completed by March
         1984.

    •    Phase II: The contractor will assemble interim usable  geohydro-
         chemical model(s) by collating  the best  methods identified dur-
         ing the Phase I research.  The  interim models connecting geo-
         chemical and geohydrological processes for modeling the fate of
         selected inorganic constituents will be  tailored  for near-term
         use by the industry.  The Phase II research is to be completed
         by the end of 1986.

    •    Phase III: The results from  research in  chemical  attenuation,
         leaching chemistry, and  groundwater  transport processes will be
         integrated in the improvement and field  validation  of  geohydro-
         chemical models.  Present plans call for Phase III  research to
         be completed by 1990.
Attenuation Rates, Coefficients, and Constants  in  Leachate  Migration
    Chemical interaction of solutes  in  solution  with  the  surrounding soils and
geological materials has profound effect on  the  migration of  the  leached  con-
stituents to groundwater.  Therefore, the  study  of  chemical attenuation has
                                      9-53

-------
VD
 I
Ln
-P-
                          Pre 1982
                                          1982
                                                        1983
                                                                       1984
                     1985
                                                                                                    1986
                                                                                                                  1987
                                                                                                                                Post 1987
                            Project Development and Management
                                      by EPRI Staff
              Technical Management Through Contract and EPRI Staff,
                           Research Results Transfer           2485-1
                          Physical Chemical Characterization
                           and Regulatory Leachate Testing
                                  1061, 1486, 1487
                                        Environmental Setting and
                                     Disposal Systems 1487-13, 2198-7
                                             Initial Evaluation of
                                          Geohydrochemical Models
                                                  1619-1
               Interim Models Development
                     and Testing
                        2485-2
   Models Improvement,
Development and Validation
         2485-2
Short- and Long-Term Chemical Attenuation Studies
               (2198-1), 2485-3
                                                                         Short- and Long-Term Leaching Studies
                                                                                   (2198-2), 2485-4
                                                                                      J	L
                                                                Saturated and Unsaturated Groundwater Transport Studies
                                                                                (2280-1), 2485-5, 2485-6
                                                                    Groundwater Sampling and Measurement Methods
                                                                                   (2283-1), 2485-7
                                                                    t    I
                                                                           Groundwater Data Analysis & Evaluations
                                                                                        (2283-2), 2485
                         Sampling and Temporal Variations Statistics
                                         1620-1
                             Geochemical and Geohydrological Field
                              Measurements              2485-9
                     Figure 2.   Solid Waste Environmental Studies  (SWES)  Project Contracts &  Schedule
                                   Environmental Physics  & Chemistry Program,  EPRI,  September 1983

-------
been considered as the area that can produce  the largest  incremental  improve-
ments in methods for predicting leachate migration to groundwaters.   Thus,  this
research is aimed at fundamental processes at  the onset.

    The research has been divided in two successive phases.   In Phase I  the re-
searchers have compiled and collated the quantitative data  from existing
literature on chemical attenuation rates, constants, and  coefficients for 21
inorganic elements (Table 1).

    The critical evaluation of the literature  in Phase I  has  included data
derived from laboratory and field research.  Emphasis has been placed on the
acquisition of existing data on adsorption-desorption, precipitation-
dissolution, and thermodynamic data on stability constants  for the 21  inorganic
elements.  Thermodynamic data have been used  to evaluate  the  coefficients and
constants for chemical attenuation mechanisms  applicable  to leachates from
utility industry solid residues.  A critical  review of the  experimental  or
observational procedure used behind the literature data has been done prior to
accepting the data values for this research.   This review of  literature  and
assembly of data was completed in July 1983.
          TABLE 1.  INORGANIC ELEMENTS FOR CHEMICAL ATTENUATION STUDIES
Aluminum
Barium
Cadmium
Fluoride
Manganese
Nickel
Sulfate
Antimony
Beryllium
Chromium
Iron
Mercury
Selenium
Vanadium
Arsenic
Boron
Copper
Lead
Molybdenum
Sodium
Zinc
    In Phase II the contractor designed and started laboratory experiments (on
solubilities, solubility products, and kinetics of precipitation-dissolution of
solid phases) to produce data on the solubility-limited concentrations of se-
lected elements in the natural environment. Additional experiments will be
conducted to measure intrinsic adsorption constants, quantify the effects of
competing ions and complexing ligands on the adsorption-desorption characteris-
tics, and to quantify adsorption-desorption on mineral mixtures and natural
soils for single and multiple electrolytes.  These experiments will employ
batch, column, and controlled field set-ups to simulate the complexity of the
environment.  Attenuation rate constants will be established by summarizing the
results.  This contract research is expected to be complete by the end of 198?.
                                      9-55

-------
Solid Waste Leaching Studies


    What factors (e.g., composition of waste and environmental  conditions)  con-
trol leachability?  What constituents can be leached from  wastes?   Under what
conditions?  In which chemical form?  How does one  determine  the  rate of
leaching of a constituent for field conditions?  Answers to most  of these
questions lie in understanding the chemical mechanisms.  Earlier  research by
Hulett et al. (1981) and by Turner et al. (1982) has identified chemical forms
of elements and distribution of trace elements found on the different solid
phases of fly ash samples.

    This work has provided a firm basis for research on chemical  mechanisms for
waste leaching.  Therefore, the objectives of the planned  leaching studies  are
to collate available data, critically evaluate the  methods used in producing
the data, and develop quantitative results on the leaching chemistry of solid
residues.  The research on leaching studies is divided into two phases.  In
Phase I the researchers are compiling quantitative  data from  existing litera-
ture, conducting feasibility experiments, and designing experiments for Phase
II.  All batch, column, sequential, and other leaching methods  used to produce
quantitative data on the chemical composition of leachates are  being critically
examined to identify relevant variables that are particularly important in  con-
trolling the leaching behavior of the fossil fuel combustion  wastes.  These
variables could include factors such as occluded liquors,  pH  and  quality of
natural waters that such wastes come in contact with, solid-liquid ratio,
permeability of wastes, age of wastes, contact time between waste  solids and
liquids, and the solid phases present.  This phase  of research  is  to be com-
pleted by April 1984.

    In Phase II the researchers will conduct laboratory and controlled field
experiments as a means of developing quantitative data on  the leaching rates of
constituents from utility solid wastes.  These experiments will be conducted to
include a broad range of waste materials and environmental conditions such  that
the results can be useful to a large segment of the utility industry.  Phase II
research is expected to be complete in 1986.


Groundwater Transport Studies
    Solutes generally migrate by groundwater  flow.   The  groundwater flow and
 transport calculations are based on mathematical equations  using Darcy's formu-
 lation during the past 50 years.  But the calculations on  solute transport by
 groundwater are relatively new, perhaps less  than  10 years  old.   The solute
 transport problem has two major processes: a  chemical  interaction phenomenon
 and a physical dispersion phenomenon.  The groundwater transport studies relate
 to the physical dispersion phenomenon.  Two contracts  have  been  let out to con-
 duct field experiments on groundwater transport of solutes  in the porous media.

    The research is divided  in two phases.  In Phase I the  researchers are
 compiling and evaluating the existing data on physical-hydrological transport
 factors used to predict the  fate of solutes in groundwaters.   Based on this
                                      9-56

-------
critical review, tracer experiments will be  designed  for  several  field  sites  to
measure the transport parameters in the development of plumes.  The  Phase  I
research is to be completed by June 1984.  In  Phase II experiments are  to  be
conducted for at least three sites for a period of three  or more  years  to  esti-
mate the hydrodynamic dispersion parameters  as a  function of  environmental var-
iables and the solutes of interest.  The groundwater  transport  research is
expected to be complete by the end of 1987.


Groundwater Sampling Methods and Related Field Measurements
    Because the ultimate goal of SWES is to produce methods for predicting the
fate of solutes in the environment,  reliable  input data  (e.g., soil  and waste
permeability, groundwater flow velocity, and  predictive  groundwater  flow direc-
tions) are needed.  The research in  this contract establishes  the  availability
and performance of field methods for sampling geochemical and  groundwater envi-
ronments specifically.

    This research is divided into  two successive phases.  In Phase I the
contractor is to compile and critically evaluate existing geochemical and
hydrologic field sampling procedures.  This evaluation will focus  on at least
the following three categories of  sampling methods:

    1)   Groundwater sample collection, handling procedures, and instru-
         ments

    2)   Soil and waste sample collection, handling procedures, and  in-
         struments specifically for  quantifying hydrologic and geochem-
         ical properties

    3)   Geophysical methods for geologic and subsurface hydrogeologic
         measurements.
    The contractor is to prepare  a manual  containing  details  on  field  sampling
procedures to define the strengths, weaknesses,  and applicability  to the  util-
ity industry.  The Phase I research is  to  be  completed  by August  1984.

    In Phase II the contractor will develop and  field test  procedures  for
groundwater sampling problems identified in Phase  I.  These tests  will  be re-
peated several times to establish reproducibility, accuracy,  and  ease  of  use
for utility industry applications.  The Phase II research is  expected  to  be
complete in 1987.
Groundwater Data Evaluation
    To properly direct the applied  research  in SWES,  a contract has been issued
to analyze data from a full-scale field monitoring  project  conducted for the
U.S. EPA and individual utility  companies.   The information from these studies
                                      9-57

-------
includes data on groundwater quality, hydrology,  soils  and  related geological
measurements, and waste disposal systems.  The  contractor will  analyze these
data to identify hypotheses that should be tested in  the SWES research on chem-
ical attenuation, leaching chemistry, and groundwater transport of solutes.
This data analysis is to be completed by July  1984.


Geochemical, Geohydrological, and Groundwater Quality Measurements


    This research is planned for the later part of SWES.  Intensive measure-
ments at a selected number of sites will be made  to develop a complete data
base for the testing and validation of predictive methods developed in the SWES
project.  This contract is expected to begin in 1985  at the earliest and may
continue for approximately four years.  Detailed  sampling plans for this con-
tract will be specified prior to the beginning  of the measurements.


                                RESEARCH RESULTS
    Since  this research began in  1977 and  the  initial  phase of the current SWES
research started in  1982, reports are now  available  that  offer usable results.
In this section a summary of the  technical  results from several EPRI reports
has been prepared.   For complete  details,  readers are  referred to the reports
sited in this paper.
EXTRACTION  PROCEDURE AND UTILITY INDUSTRY WASTES
    Two EPRI reports  (Rose et al,  1981 and Eynon  et  al.  1983)  deal with the
 topic  of  reproducibility of results  (concentration of  eight  elements:   As,  Ba,
 Cd, Cr, Pb, Ag, Se, and Hg) in the extracts  from  using the extraction  procedure
 (EP) specified by  the U.S. EPA as  applied to fly  ash,  bottom ash,  and  scrubber
 sludge.   The concentration of silver in the  EP  extract was found to be below
 detection limit most  of the time;  therefore,  no statistical  analysis was per-
 formed on the data for silver.

    The solid samples used in this research  were:  (1)  dry fly ash—alkaline
 (DFA--ALK), (2) dry fly ash—acidic  (DFA--AC),  (3) wet bottom ash—alkaline
 (WA--ALK),  (4) wet bottom ash—acidic  (WA—AC), and  (5)  scrubber sludge (SS).
 Four different laboratories were used  to extract  four  samples for each of the
 five waste  types using the EPA's extraction  procedure  of December 1978.  Each
 extract liquid was split into eight  aliquots by each laboratory.  Each aliquot
 sample was  analyzed for the eight  elements using  two different analytic proce-
 dures.  Numerical  results obtained from this research  are summarized in Table 2.
 Based  on  this research the following conclusions  can be  drawn:

    •     Results on the concentration  of the eight elements  were found to
          be hundreds  of percent apart  between different  laboratories.
                                      9-58

-------
                         TABLE 2:  SUMMARY STATISTICS  AND APPORTIONED VARIATION IN LEACHATE COMPOSITION*
I
i_n
'*£>


Element
Arsenic

Barium

Cadmium

Chromium

Mercury
Lead
Selenium



Analysis
FURNACE AA
FLAME AA
FURNACE AA
FLAME AA
FURNACE AA
FLAME AA
FURNACE AA
FLAME AA
FLAME AA
FURNACE AA
FURNACE AA
FLAME AA


Prep
EP
EP
EP
EP
EP
EP
EP
EP
EP
EP
EP
EP


Waste
DFA-AC
DFA-ALK
SS
DFA-AC
DFA-ALK
SS
DFA-AC
DFA-ALK
WA-AC
WA-ALK
SS
WA-ALK
DFA-AC
DFA-ALK
WA-AC
WA-ALK
SS
DFA-AC
DFA-AC
DFA-ALK
WA-AC
WA-ALK
SS
DFA-ALK
SS
DFA-AC
SS
DFA-AC
DFA-ALK
SS
DFA-AC
DFA-ALK
SS

N
Samp
128
128
128
128
128
128
128
128
128
128
128
128
128
128
128
128
128
126
128
128
128
128
128
128
128
128
128
128
128
128
128
128
128


Used
128
128
128
128
128
128
128
128
128
128
128
127
128
128
128
128
128
128
128
128
128
128
128
128
128
128
128
128
128
128
128
128
128

Geom
Mean
(ppb)
7.14
7.07
77.22
6.21
4.58
42.98
69.96
129.32
43.23
181.79
83.44
890.54
95.81
0.19
0.37
1.56
7.09
126.33
9.01
536.74
1 .09
2.72
28.32
590.10
0.73
1.04
0.88
24.11
51.88
39.08
5.25
20.37
18.60

Minimum
(ppb)
0.63
0.63
10.00
1.00
0.17
6.55
1.43
0.63
1 .43
0.49
1.26
24.86
23.00
0.01
0.03
0.28
0.58
38.03
2.50
331.57
0.11
0.16
2.61
195.38
0.05
0.14
0.18
4.63
23.53
15.69
0.05
0.30
0.60

Maximum
(ppb)
46.00
110.00
220.00
140.00
147.23
288.74
650.00
3400.00
1300.00
1600.00
730.00
32000.00
240.00
48.00
11 .00
22.00
62.31
240.00
31.69
830.00
412.37
290.00
110.00
1000.00
59.00
9.00
5.44
70.00
121 .62
73.00
62.00
89.00
68.00
Pei
Inter
Lab
Extr
32
4
47
35
10
29
0
0
0
2
1
6
28
0
12
7
13
29
0
11
0
42
70
3
53
12
6
23
16
0
4
6
2
-cent o
Intra
Lab
Extr
2
44
12
2
33
1
0
3
7
0
0
0
1
29
12
6
5
0
28
39
71
18
24
65
4
0
0
2
34
15
0
15
0
f Total
Inter
Lab
Anal-I
51
31
37
14
4
10
82
73
65
88
90
63
45
41
56
65
70
32
4
13
9
2
3
15
0
54
32
34
11
13
86
58
72
Variance
Inter
Lab
Anal-II
9
7
2
17
36
23
5
18
12
0
0
16
7
2
2
11
4
13
19
13
10
22
0
13
24
7
25
6
22
23
3
14
4

Intra
Lab
Anal
6
13
2
32
17
38
12
6
15
11
10
15
20
28
17
10
8
26
50
24
10
16
3
4
19
27
38
36
17
49
6
7
21
             *   The data in this table are taken from EPRI Report EA-3181,  Tables 3-2 and 3-5.

-------
    •    Reproduciblity differed from element to element, waste type, and
         the analytic technique used.

    •    The interlaboratory variation in results was generally higher
         than the intralaboratory variation.


TRACE ELEMENTS AND THEIR CHEMICAL FORMS


    Two EPRI reports (Hulett et al. 1981 and Turner et al. 1982) deal with the
general subject of ash chemistry.  The objective for Part I in this research
was to establish the distribution of trace elements in the various solid phases
present in fly ash.  By using physical and chemical methods, researchers first
separated the fly ash samples into three phases: (1) glass (approximate compo-
sition Al-Si-0), (2) mullite-quartz (approximate composition 3(Al2Oo).2(Si02)),
and (3) magnetic spinel (approximate composition Fe2A10{j).  The distribution  of
trace elements in each of the three phases was then defined.  The research
results show:

    •    Glass is typically the most abundant solid phase of the non-
         magnetic alluminosilicates.

    •    Trace elements distribution analyses show that a distinct pro-
         pensity occurs in specific phases:

         — The glass phase seems to contain almost all of the alkali,
            alkaline earth, and rare earth elements, as well as the
            majority of As, Pb and Se.

         — Small amounts of trace elements are associated with the
            mullite-quartz phase with the exception of V, Cr, Ga, Zr, Fe,
            and Ti—all of which seemingly occupy the +3 or +4 valence
            sites in the crystalline structure.

         — The magnetic-spinel phase seems to be largely enriched in the
            first row transition elements (V, Cr, Mn, Fe, Co, Ni, Cu, and
            Zn).
    The objective of research in Part II was to identify patterns in  the  coal
ash composition and its leaching behavior in actual and laboratory-simulated
disposal experiments.  Arsenic was selected for a more complete  study on
leachability and chemical speciation in leachates.  The important results
reported by the researchers (Turner et al. 1982) are summarized  below:

    •    Ash sluicing-ponding systems that are neutral to alkaline  in pH
         appear to show appreciable solubilization of elements (e.g.,  S,
         As, Se, Cr, and Mo) with the anionic aqueous speciation.   In
                                      9-60

-------
         contrast, the sluicing-ponding systems with pH in the acidic
         range appear to result in high solubilization of most cationic
         elements (particularly Fe, Mn, Cu, Zn, and Cd).

    •    The study of As indicated that in controlled pH experiments the
         As (v) varied systematically with pH variations and the As (III)
         concentration remained nearly constant over a wide range of pH
         (3 to 12).

    •    A considerable amount of variablility among fly ashes with
         respect to quantity and to the oxidation state of arsenic
         released was observed in the laboratory leaching experiments.
COMPARISON OF SOLID WASTES FROM COAL COMBUSTION AND PILOT COAL GASIFICATION
  PLANTS
    The research results from this work are given in an EPRI report prepared by
Turner and Lowry (1983).  The researchers principally focused on developing
data on the physical and chemical characteristics of solid wastes from conven-
tional coal-burning power plants and solid wastes from coal gasification in the
CE-PDU plant.  Bulk composition results of the waste samples analyzed are shown
in Table 3 and the EP extract composition results are shown in Table 4.  The
important results from this contract research are summarized below:

    •    Slag from the gasifier and slag from the cyclone-fired combus-
         tion unit were morphologically similar and predominantly con-
         sisted of a dense shards of specific surface area of less than
         1 m /g.  In contrast, the bottom ash from a pulverized-coal com-
         bustion plant predominantly contained round vesicular cinders of
         specific surface area between 1 and 3 m /g.

    •    The two slags did not show any crystalline structures, but
         mullite crystals were found in the fly ash from the pulverized
         coal combustion power plant.

    •    Based on aqueous extraction and inorganic elements analyses
         according to the EPA extraction procedure, none of the com-
         bustion or the gasification solid wastes studied would be
         classified as hazardous on the basis of the toxicity criteria
         of the Resource Conservation and Recovery Act regulations.

    •    The gasification slag showed a more acidic reaction in
         continuous-flow column extractions and thus may leach some
         constituents at a higher rate than combustion slag and bottom
         ash.

    •    Although some differences exist between the wastes examined,  the
         gasification slag appears to be nearly as inert chemically as
         slag from the cyclone furnace.
                                      9-61

-------
   TABLE  3.   ELEMENTAL  COMPOSITION OF COMBUSTION AND GASIFICATION SOLID
                    WASTES PRODUCED FROM  PITTSBURGH  COAL*



Si
Al
Fe
Ca
Na
Mg
K
Ti
S
C
Ag
As
Ba
Cd
Cr
Cu
Hg
Mn
Ni
Pb
Se
Zn
Mitchell
Fly Ash

20.5
11.6
10.2
3.09
0.56
0.45
1.19
0.62
0.80
7.04
<1
93.
743.
0.82
113.
68.
0.10
188.
129.
50.
12.
92.
Station**
Bottom Ash

21.6
11.9
14.9
2.38
0.55
0.46
1.10
0.56
0.18
0.58
<1
3.2
551.
<0.1
87.
44.
0.006
176.
122.
8.
N.D. ******
32.
##*
CE-PDU
Slag
	 « 	
19.2
14.9
14.4
1.32
0.87
0.31
1.32
0.59
0.31
0.80
	 MB' 6 	
<2
3.6
1470.
0.18
4750.
68.
0.011
273.
228.
4.9
1.0
28.
Merrimack
Slag

21.4
12.6
12.2
5.52
0.94
1.62
1.03
0.60
0.07
0.16
<2
<1.8
1253.
0.64
126.
69.
0.008
205.
98.
8.1
2.2
32.
Station****
. , ******
Fly Ash

18.0
11.2
13.1
3-72
1.34
0.73
1.58
0.58
0.91
4.6
<2
393.
1400.
4.2
298.
431.
0.04
182.
158.
309.
N.D.
460.
     *   This  table is taken from EPRI  Report EA-2867 Table 3.4.
    **   Pulverized coal furnace.
   ***   Entrained-flow gasifier (low Btu), collected June 3, 1980.
  ****   Wet-bottom slagging furnace.
 *****   Recycled to furnace at this power plant.
******   Not detected.
                                     9-62

-------
TABLE  4.   RESULTS OF EPA'S  EXTRACTION  PROCEDURE  APPLIED TO  SOLID WASTES
              PRODUCED FROM  PITTSBURGH  COAL*   (concentrations  are in
                     yg/L except for SOJj which  is  in mg/L)
Mitchell Station** CE-PDU****

Ag
As
Ba
Cd
Cr
Cu
Fe
Hg
Mn
Ni
Pb
Se
Zn
sojj
Acid,
mL/g
Final
pH
Fly Ash
0.06
9.2
145.
12.
87.
106.
362.
0.014
1700.
188.
4.35
2.8
160.
940.

1.2

5.0
Bottom Ash
<0.01
0.88
85.
0.12
1.3
1.4
144.
0.011
244.
11.
<0.06
<0.05
5.7
16.2

0.15

5.0
Slag
<0.05
0.55
58.
0.06
1.45
5.1
326.
0.058
4.45
180.
0.53
<1
5.2
6.24

0

4.6
Merrimack
Slag
<0.05
0.09
2.5
0.11
<0.1
0.74
275.
0.004
1.3
0.28
0.14
<0.5
0.61
--

0.01

4.8
Station
Drinking
Water
Fly Ash****** Criteria
0.12
63.
100.
69.
1.5
290.
22.
0.04
580.
210.
3.1
1 1 .
1500.
--

1 .2

5.1
50
50
1000
10
50
1000
300
2
50
—
50
10
5000
250




      *   This table  is taken from EPRI Report  EA-2867 Table  3.8.
     **   Pulverized  coal furnace.
    ***   Wet-bottom  slagging furnace.
   ****   Entrained flow gasifier  (Low Btu),  collected June 3,  1980.
  *****   Recycled  to furnace at this power plant.
                                      9-63

-------
STUDY OF TIME-VARIABILITY OF ELEMENT CONCENTRATIONS  IN  POWER  PLANT  ASH
    This research contract was initiated to study the uncertainty  in  ash
composition due to representativeness of the ash samples  subjected to the
extraction test.  This study is critical to answering:  "What  constitutes
a representative waste sample, and how variable is  the  waste  stream over
time?"  The research has been conducted in two parts.   Results  from the
Part I pilot study are given in an EPRI report prepared by  Switzer,
Eynon, and Holcombe (1983).  Part II research has also  been completed and
the report is in preparation.  The Part I research  consisted  of obtaining
samples of coal and coal ashes from a southwestern  power  plant  for a  30-
day period.  The samples were chemically analyzed for 14  inorganic ele-
ments.  The elemental concentration data have been  subjected  to statisti-
cal analyses to estimate the time variability in chemical composition of
coal ashes.  Results from this research are summarized  below:

    •    Time, sampling, sample preparation, and sample analysis are  the
         four components providing sizable contributions  to the total
         variability in ash composition.

    •    The contribution of the four components to total variability
         differed in degree depending upon the chemical element and
         method of analysis.

    •    The correlation between plant process variables  (feed  coal com-
         position, boiler temperature, boiler oxygen, and electrostatic
         precipitator air temperature) and the ash  chemical composition
         showed very little association.  This lack of  correlation may be
         due to the relatively constant chemical composition  of feed  coal
         during this research.
PHYSICAL-CHEMICAL CHARACTERISTICS OF UTILITY SOLID WASTES
    The primary purpose of this research was to compile existing  data on waste
physical properties, chemical composition, and leachate composition for esti-
mating the physical-chemical composition of solid wates.   The  EPRI  report by
Summers, Rupp, and Gherini (1983) provides detail on  the  results.   A large
quantity of the data compiled are on coal ash and its leachates.  These data
are for 19 minor and trace elements.  Relatively small amounts of data are on
physical properties of ash and flue gas desulfurization (FGD)  sludge, chemical
composition of FGD sludge and leachates, and chemical composition of oil ash
and geothermal wastes.  The results of this research  are  summarized in Figures
3 and 4.
                                      9-64

-------
                                                                      PLANT     COAU   ASH
                 PftOLXJCTION £ PI5PO5AU

                                        HICM, I—» pLV
                                     CDU.tcrof>.     ASH
                                 BOTTOM ASH
                 RELATIVE AMOUNTS  AMMOM- 615 A5H
 I
ON
Ln
                                    TOTAL-'. 68 million
                                     tons per year
       : 16 million
•tons per year
PDFD5EP

     | j5oF*et
                    l-AHPFIU-
FmWfj'ft'tttr'ftp

 fonp
                            Prrr51CAU-CH£MICAU
                            PHYSICAL. CHARACTERISTICS
                                         FLY ASH 0.5-IOO^ L PW55
                                                 35   )
                                                                         -x 03)
                                                                                            Co. V
                                                                 Quorti
                                                                 GlaM
                                                                 Carbon
                                                10  j   M^-Ti^Zr
                                                10     Al
                                                35
                                                IO
                                                                             LEACHING
                                                                                                                              EXTRACTION  R£56ILT5
                                                                                                                                AVG. CQNC H9/t
                                                                                                                     SILVER
                                                                                                                     AR5EHIC
                                                                                                                     BAKlOt-1
                                                                                                                     CAPMICJM
                                                                                                                     CHROMIUM
                                                                                                                     MtRCdJRV
                                                                                                                     LEAP
                                                                                             .6
                                                                                             12
                                                                                            200
                                                                                             4,7
                                                                                             J6
                                                                                             O4
                                                                                             5
                                                                                             10
                                                                                                        O.I  — 2OO
                                                                                                        04 - I6OO
                                                                                                        3  - 76OO
                                                                                                        O.I  — I4OO
                                                                                                        I   -  700
                                                                                                       0-—
                                                                                    (i-apid hyarolysis of 5tJrface mctoV
                                                                                 5ubs«)cient pH anop dtrf to wash oat of aikolinrty.
                                                                                 Pnpp apon air eqailibratlon: tV+^OH'+ZCO, (oir)
                                                                                 — fV'+2HCO,-
                                                                                 Acidio Ashes: Innd dissolution of strong odd onions
                                                                                 (e.g 5O^-, C|- ) exceeds tti« dissola+ion of base
                                                                                 cations («jj.Co*+,No')
               Figure  3.    Summary  of Physical-Chemical  Characteristics of Power  Plant  Coal  Ash*
                               *This  figure  is  taken from EPRI  Report  EA-3236,  Figure 7-1.

-------
                                                              PLANT     PGP
                PROPdCf ION ^ PI5PO5AU
VTJ
 1
Co-bostd Met   80
                FWj«r*rnbl< R-CXKCO S
                Pry FVoozsses   2
BA5I5:®IOOO M«4z Plant
-  15% 5 Coal
     LZTtAsh
    AVet Ume 5crabt*r
              '
                                    ©3?OOO MWe
                                     Scrubbed CN»)
                                OS PRORJCtlOn~l
                P15FO5AL-
                            VOLaM& P15PO5EP
                              PhffSICAL-CHEMICAU

                              PHYSICAL CHARACTERI5riC5

                              SHAPES'. 5dlUFrT6R)6H-Platey  5dJUPATeR-&bcJ<,5

                              SEE. RAHGt: l-60/u«qaiv diam. (80% by mass)
                     i-IOO-IIS^i
                   „ . io-*toiO-Jgj-
              fclmilar to sandy clay)
             rixed ^fajdan IO"5 to I0'r£g-
              (simitar roclay)
POROSITY:     Ra« audoe 20-60%
             Fixed Sludge <5 %


CH&hlCAL COMP05ITlOn
   MAJOR COMPOrl&rrr5 (Wet U F&P Sludge 1
    • Calcitim salfate.    Ca5O4'Ci-2)H2O
    • Calcium salflte    CaSOj
    • Calcium cartxpnate CaCOa
    • 5odiam Sijlfate     MaaSG\-
    • "Incrts^CAsh and Impurrlies)
   AVG. -TKAC& ElfcriEJ   _   .
    C5asi5. 7 investigations —OVet FQP 5ladg
   &UeMfirfT   COttC IM ,5OUP5  COriC IM
               mg/hg         ^9A
   Arsenu:          IO          IO
   Boron         134       I4.OOO

   Chromium        13          10
   Copper          27          25
    Flaorine       53O       3,OOO
    Mercery         O.3       '   O.8
    Urad           6          2O
    5?leni£Jm        6         IOO
                                                                  • Umited Pata
                                                                  • Similar to wet systiem sludges
                                                                   wh^n asinq +h« same, absortuznf
                                                                  • Mo trace clement" data available
UACHING CHAr\ACTEf\l5TlC5

EP EAtCh EXfKACTIOH
                                                                                                                5IL-VEP,
                                                                                                                ARSENIC
                                                                                                                BARIUM
                                                                                                                CAPMKJM
              CONCEMTRATIOMS fig/f
            WSf F6P 5U-IPOC PRY fcrP 5UJPQC
                <6O         -
               6O4-4O       HbOO
             <-)00-|oOO       —
               <25O       H5OO
                
-------
GEOHYDROCHEMICAL MODELS FOR  SOLUTE MIGRATION—AN EVALUATION
    An initial evaluation of  existing  geohydrochemical  models  and  codes  is  be-
ing carried out by the  researchers  in  Phase  I.   The  first  volume of the  reports
is in the final stages  of preparation  (Kincaid,  Morrey,  and  Rogers,  1983).  This
volume contains descriptions  of  processes  deemed important in  solute migration
and a selection, for initial  evaluation, of  21  existing codes  (Table 5)  which
deal with these processes.  These codes  employ  one or more of  the  three  compon-
ents in geohydrochemical model(s):  (1) hydrological  flow and transport,  (2)
geochemical interactions influencing the availability of a constituent for  mi-
gration, and  (3) microbial  transformation  which may  alter  the  constituent dur-
ing migration.  The second  and third volumes are to  be  prepared  by early 1984
and will report on the  findings  of  simulations  on the applicability of a subset
of these selected codes for predicting solute migration around the utility
industry disposal sites.
CHEMICAL ATTENUATION  RATES,  COEFFICIENTS,  AND  CONSTANTS  IN  LEACHATE  MIGRATION


    A two-volume report  prepared  under  this  Phase  I  research  contract  is  being
finalized for publication.   In  Volume I of the report  (Rai  and  Zachara,  1983)
data obtained from  the available  literature  have been  summarized  for each of
the elements.  The  quantitative results given  allow  therrnodynamic data-based
predictions of the  relative  stability of solid and aqueous  species,  observed or
hypothesized solubility  controls  in  the geochemical  environments,  and  the rates
of reactions for adsorption-desorption  and precipitation-dissolution processes
for chemical attenuation.  Most of the  data  compiled pertains to  single  element
and pure minerals.

    Volume II of this report (Rai, Zachara,  Schmidt, and Schwab,  1983) is an
annotated bibliography of over  350 publications which  appeared  in existing
literature prior to February 1983 and deals  with research on  chemical  (precipi-
tation-dissolution  and adsorption-desorption)  and  biological  (methylation and
alkylation) attenuation  mechanisms for  Al, As, B,  Ba,  Be, Cd, Cl,  Cr,  Cu, F,
Fe, Mn, Mo, Na, Ni, Pb,  Sb,  Se, SO^, V,  and  Zn.


                             SUMMARY  AND CONCLUSIONS
    In this paper a chronological  summary  of  the  SWES project  at  EPRI  is  devel-
oped.  Although the research  is  in an  early stage of implementation,  it is
expected that thorough and deliberate  planning  has given  the  research  a very
high chance for successful completion.   In the  short-term (the next 3  to  5
years), the research results  will  be in  the form  of data  which can be  used in
assessing the potential  for release, transformation,  and  transport of  inorganic
solutes from utility industry solid residues.   In the long term (the next 6 to
8 years), the research results are to  provide predictive  methods  (mostly geohy-
drochemical model [s]) to quantitatively  estimate  the fate of  solutes in the
groundwaters.
                                      9-67

-------
TABLE 5.  COMPUTER CODES SELECTED FOR EVALUATION IN SWES
I.   UNSATURATED FLOW AND TRANSPORT CODES:

        Code Designation
    (1)   SESOIL
    (2)   NRC-SLB
    (3)   OR-NATURE
    (4)   UNSAT1D
    (5)   FEMWATER/FEMWASTE
    (6)   TRUST/MLTRAN
    (7)   SATURN
II.  SATURATED FLOW AND TRANSPORT CODES:
    (8)   PATHS
    (9)   SWIP2/SWENT
    (10)  TRANS
    (11)  VTT
    (12)  FE3DGW
    (13)  USGS MOC
    (14)  AT123D
III. EQUILIBRIUM GEOCHEMICAL CODES:
    (15)  GEOCHEM

    (16)  MINTEQ

    (17)  PHREEQE
    (18)  EQUILIB
    (19)  EQ3/EQ6
IV.  MICROBIOLOGICAL CODE:
    (20)  BIOFILM

V.   HYDROLOGICAL AND GEOCHEMICAL CODE:
    (21)  FIESTA
                                     Description/Function
                                       Compartmental
                                       One-dimensional
                                       One-dimensional
                                       One-dimensional
                                       Multi-dimensional
                                       Multi-dimensional
                                       Multi-dimensional
                                       Multi-dimensional-analytical
                                       Multi-dimensional
                                       One or two dimensional
                                       One or two dimensional
                                       Two or three dimensional
                                       Two dimensional
                                       Multi-dimensional-analytical
                                       Equilibrium, geochemistry,
                                       includes adsorption
                                       Equilibrium, geochemistry,
                                       includes adsorption
                                       Precipitation-dissolution
                                       Precipitation
                                       Precipitation
                                       Primarily organic compounds
                                       One-dimensional flow and
                                       transport coupled with
                                       equilibrium geochemistry
                         9-68

-------
                                  REFERENCES
Eynon, B. and Switzer, P., 1983.  A Statistical Comparison of Two Studies on
    Trace Element Composition of Coal Ash Leachates.  Palo Alto, CA: Electric
    Power Research Institute.  EPRI EA-3181.

Hulett, L.D.; Weinberger, A.J.; Ferguson, N.M.; Northcutt, K.J.; and Lyon, W.S.
    1981.  Trace Element and Phase Relations in Fly Ash.  Palo Alto, CA:
    Electric Power Research Institute.  EPRI EA-1822.

Kincaid, C.T.; Morrey, J.R.; and Rogers, J.E., 1983.  Geohydrochemical Models
    for Solute Migration: The Selection of Computer Codes and Description of
    Solute Migration Processes. Volume I.  Palo Alto, CA: Electric Power
    Research Institute.

Murarka, I.P., 1982.  Solid-Waste Environmental Studies; The Needs and the
    Priorities.  Palo Alto, CA: Electric Power Research Institute.  EPRI EA-
    2538-SR.

Rai, D. and Zachara, J.M., 1983.  Chemical Attenuation Rates, Coefficients,  and
    Constants in Leachate Migration: A Critical Review.  Volume I.  Palo Alto,
    CA: Electric Power Research Institute.

Rai, D.; Zachara, J.M.; Schmidt, R.A.; and Schwab, A.P., 1983.  Chemical
    Attenuation Rates, Coefficients, and Constants in Leachate Migration: An
    Annotated Bibliography.  Volume II.  Palo Alto, CA: Electric Power Research
    Institute.

Rose, S.J.; Dane, J.; Eynon, B.; and Switzer, P.,  1981.  Extraction Procedure
    and Utility Industry Solid Waste.  Palo Alto,  CA: Electric Power Research
    Institute.  EPRI EA-1667-

Science Applications, Inc., 1982.  Planning Workshop on Solute Migration From
    Utility Solid Wastes.  Palo Alto, CA: Electric Power Research Institute.
    EPRI EA-2415.

Summers, K.V.; Rupp, G.; and Gherini, S., 1983.  Physical-Chemical Character-
    istics of Utility Solid Wastes.  Palo Alto, CA: Electric Power Research
    Institute.  EPRI EA-3236.

Switzer, P.; Eynon, B.; and Holcombe, L.J., 1983.  Pilot Study of Time Vari-
    ability of Elemental Concentrations in Power Plant Ash.  Palo Alto,  CA:
    Electric Power Research Institute.  EPRI EA-2959.

Turner, R.R. and Lowry, P.D., 1983.  Comparison of Solid Wastes From Coal
    Combustion and Pilot Coal Gasification Plants.  Palo Alto, CA: Electric
    Power Research Institute.  EPRI EA-2867.

Turner, R.R.; Lowry, P.; Levin, M.; Lindberg, S.E.; and Tamura, T.,  1982.
    Leachability and Aqueous Speciation of Selected Trace Constituents of Coal
    Fly Ash.  Palo Alto, CA: Electric Power Research Institute.  EPRI EA-2588.
                                      9-69

-------
SESSION 10, PART I:  DRY FGD:  PILOT PLANT TEST RESULTS

Chairman:   Theodore G. Brna
           Industrial Environmental Research Laboratory
           U.S. Environmental Protection Agency
           Research Triangle Park, NC

-------
CURRENT STATUS OF DRY S02 CONTROL SYSTEMS




M. A. Palazzolo, M. E. Kelly. T. G. Brna

-------
                  CURRENT STATUS OF DRY SO  CONTROL SYSTEMS

     by:   Michael A.  Palazzolo               Mary E.  Kelly
          Radian Corporation                 Radian Corporation
          Durham, NC   27705                  Austin,  TX  78766

          Theodore G. Brna
          Industrial  Environmental Research Laboratory
          U.S.  Environmental Protection Agency
          Research Triangle Park, NC  27711
                                ABSTRACT

     This paper provides an update on commercial applications and research
and development (R&D) activities involving three dry SO  control techno-
logies:  spray drying, dry injection, and electron-beam irradiation.

     Spray drying continues to be the only commercially applied dry flue gas
desulfurization (FGD) process, with two additional spray drying systems
having been sold since mid-1982.  To date, 17 commercial utility spray
drying systems have been sold, totalling over 6,800 MWe.  Six of these
systems are currently operational with two of the systems having been turned
over to the utility.  In addition to the utility applications,  there  are
21 industrial spray drying units, 7 of which are operational.  Several
additional utility and industrial systems are expected to start up in the
next 2 years.  Demonstration- and pilot-scale testing of the spray dryer
process is continuing with emphasis on high sulfur applications.

     The first planned commercial application of dry injection technology
has been announced for a 500 MWe utility.  The recent demonstration-scale
testing on a 22 MWe unit has been completed.

     The electron-beam process is also in an early developmental state.
Pilot-scale testing of the electron-beam/lime spray drying version of the
process is scheduled to begin this fall.

                                INTRODUCTION

     This paper updates the commercial and research and development (R&D)
status of dry SO. control processes for utility and industrial boilers.
Four EPA reports on the status of dry S0» control systems have been
published to date (1, 2, 3, 4), the most recent in August 1983.  A fifth
update is currently being prepared, with completion scheduled for early
1984.  The primary sources of information have been contacts with dry SO.
control system vendors and purchasers and government and industry agencies
                                    10-1

-------
involved with dry S09 control in the United States.  Three types of dry SO
control (nonregenerable)  processes'•will be covered in this paper, the latter
one briefly because of coverage elsewhere during this symposium:

       •  Spray drying.

       •  Dry injection.

       •  Electron-beam irradiation.

     Spray drying, which includes a fabric filter or an electrostatic
precipitator (ESP) for collection of fly ash and desulfurization products,
is the only one of the three processes to be commercially applied.  Dry
injection, a process involving injection of dry alkaline material directly
into the flue gas and subsequent collection of the desulfurization products
and fly ash in a fabric filter, is nearing commercial application.  The
electron-beam process, which involves reagent injection into the flue gas
followed by electron-beam irradiation, is in an early developmental state.

     A more detailed description of these processes, including some of the
important process variables and design parameters, is presented below.  The
technology descriptions are followed by a discussion of the current
commercial applications and R&D activities for each of the three processes.

                        TECHNOLOGY DESCRIPTIONS

SPRAY DRYING

     Figure 1 is a schematic diagram of the spray drying FGD process.  The
gas containing fly ash and S0~ enters the spray dryer and is contacted with
a finely atomized alkaline slurry or solution.  During the approximately
10-second residence time in the dryer, the flue gas is adiabatically
humidified as the water in the slurry or solution is evaporated.  Simul-
taneously, flue gas SO  reacts with the alkaline species to form solid
sulfite and sulfate salts.  The solids formed are dried to generally less
than 1 percent free moisture.  The flue gas, which has usually been
humidified to within 11 to 28°C (20 to 50°F) of its adiabatic saturation
temperature, passes through the dryer and into a downstream high efficiency
particulate matter control device.  In some designs a portion of the solids
drop out of the dryer, but the bulk of the desulfurization products are
collected with fly ash in a fabric filter or an ESP.  The reaction between
the alkaline material and flue gas SO  continues as the gas passes through
the ductwork and the fabric filter or ESP.

     Table 1 lists the major design and operating factors for spray drying
FGD.  These considerations are discussed in more detail below.

     Using a spray dryer as a flue gas contactor involves adiabatically
humidifying the flue gas to within a specified number of degrees above
saturation.  With set conditions for inlet flue gas temperature and humidity
and for a specified approach to saturation temperature, the amount of water
which can be evaporated into the flue gas is set by energy balance consid-
erations.  Liquid-to-gas ratios are generally in the range of 0.27 to
                                     10-2

-------
o
I
(-0




















_
BOILER




COMBUSTION />





,
i

^ FLUE G/i






IR





HOT


s












HOT OR WAR
I
/ \
/ \ 1
H

PREHEATER

Ai







M G/
T.
i






R







CLEAN GAS TO
^s BYPASS ATMOSPHERE
'ARM / \
1 (~l FAN GA Q / \

1 1 /? 	 ^-^v '^ / \
1 ff n / \
t I 1 \~^J 1 '

I 1 r
/K ' I~AN STACK
SPRAY ._
DRYER w
\ / * >r
\/ Fi HP RA-; BAGHOUSE
& SOLIDS OR ESP
SOLIDS
PARTIAL RECYCLE OF SOLIDS f
yL
\( *
r PRODUCT SOLIDS &
SoRBENT FLY ASH DISPOSAL
SLURRY
TANK
X
                                                            SORBENT STORAGE
                Figure  1.   Typical spray dryer/particulate collection flow diagr
am.

-------
             TABLE 1.   IMPORTANT DESIGN AND OPERATIONAL FACTORS
                       FOR SPRAY DRYING FGD SYSTEMS
Operational Factors

       •  Approach to saturation at dryer outlet

       •  Sorbent stoichiometry

       •  Inlet SO  concentration

       •  Temperature drop over the spray dryer

       •  Fly ash alkalinity

       •  Gas residence time


Design Factors

       •  Spray dryer design (gas distribution, gas exit configuration)

       •  Atomizer type

       •  Slaker type

       •  Use of solids recycle

       •  Use of flue gas bypass
                                    10-4

-------
            o
0.40 liter/m  (0.2 to 0.3 gal./kef).  The sorbent stoichiometry is varied by
raising or lowering the concentration of a solution or weight percent solids
of a slurry containing this set amount of water.  While holding other
parameters such as temperature constant, the obvious way to increase SO
removal is to increase sorbent stoichiometry.  However, as sorbent stoichio-
metry is increased to raise the level of SO  removal, two limiting factors
are approached:

       •  Sorbent utilization decreases, raising sorbent and disposal costs.

       •  An upper limit is reached for the solubility of the sorbent in the
          solution, or for the weight percent of sorbent solids in a slurry.

     There are at least two methods of circumventing these limitations.   One
method is sorbent recycle, using the solids that have either dropped out in
the spray dryer or collected in the particulate emission control device.
Recycle increases sorbent utilization and can also increase utilization of
any alkalinity in the fly ash.

     The second method of avoiding the above limitations on S0_ removal is
to operate the spray dryer at a lower outlet temperature; that is, a closer
approach to saturation.  Operating the spray dryer at a closer approach to
saturation has the effect of increasing both the residence time of the
liquid droplets and the residual moisture level in the dried solids.   As the
approach to saturation decreases, S0_ removal rates and sorbent utilization
generally increase dramatically.

     The approach to saturation at the spray dryer outlet is set by either
the requirement for a margin of safety to avoid condensation in downstream
equipment or restrictions on stack temperature.  The spray dryer outlet can
be operated at temperatures lower than these restrictions would otherwise
allow if some warm or hot gas is bypassed around the spray dryer to reheat
the dryer outlet gas.  Warm gas (from downstream of the boiler air heater)
entails no energy penalty, but bypassing untreated gas for reheating can
limit overall S0? removal efficiencies.  Significantly less hot gas
(upstream of the air heater) is required for heating, but using hot gas
decreases the energy available for air preheating.   Figure 1 illustrates
these two reheat options.

     Several similarities exist between the designs of the commercial spray
drying FGD systems sold to date.  All of the systems except one use lime as
the sorbent; the exception uses commercial soda ash.  In addition, all but
two of the purchased systems include a fabric filter for collection of fly
ash and waste products.  Two utility systems will have an ESP instead of a
fabric filter.  Reverse-air fabric filters have been selected for utility
systems, while pulse-jet units have been the choice for most industrial
systems.

     Most commercial system designs use rotary atomizers, although nozzle
atomizers are used in some designs.  Solids recycle and flue gas bypass for
limited reheat are site-specific design options.  Recycle, however, is being
included more often than not, primarily because it improves sorbent
                                     10-5

-------
utilization and lowers reagent costs.  The use of recycle and flue gas
bypass depends primarily on the SO  removal requirements.

     Several differences between the commercialized designs of spray dryer
FGD systems are evident.  The differences can be categorized into four major
areas:

       •  Spray dryer design and operation.

       •  Atomization.

       •  Sorbent preparation (lime slaking)-

       •  Fabric filter design and operation.

     Variations in spray dryer design include the shape of the dryer, gas
disperser configurations, multiple or single rotary (or nozzle) atomizers
per dryer, one- versus two-point flue gas admission to the spray dryer, and
solids collection from the spray dryer.  Each vendor claims his design to be
suitable for FGD and to have certain advantages.  Some important
considerations in evaluating the various designs include:

       •  Liquid/gas contact.

       •  Turndown capability.

       •  Potential for plugging or carryover of wet solids to baghouse
          during upset conditions.

     Differences in spray dryer operation include gas residence time and the
approach to saturation at the dryer outlet.  Gas residence times range from
7  to  12 seconds in most commercial designs; most operate in the 10- to
12-second range.  Many commercial system designs call for a relatively close
approach to saturation at the dryer outlet [10 - 14°C (18 - 25°F)].  These
close approaches are common where S0_ removal efficiency requirements
approach 85 to 90 percent, but they may also be used to decrease reagent use
in lower removal efficiency applications.  Some designs, however, are based
on a wider approach to saturation [17 - 28°C (30 - 50°F)].

     Atomization techniques vary with regard to the use of rotary or nozzle
atomizers, atomizer wheel speed in rotary atomizers, atomizing fluid (air or
steam) for nozzles, internal or external mixing of fluids in nozzles, and
the number of atomizers.  Rotary atomizers may be used on a single or
multiple atomizer per dryer basis.  Several commercial utility systems use
three atomizers per dryer, but most sold to date have a single atomizer per
dryer.  Rotary atomizers are generally thought to have certain advantages
over nozzle atomizers.  Specifically, unlike nozzle atomization, the size of
the droplets produced by rotary atomizers is independent of slurry flow
rate, providing for good turndown capability.  Some plugging problems have
occurred in nozzles using steam for atomization.  One potential advantage of
nozzle atomizers lies in the fact that they are mechanically simpler than
rotary atomizers.
                                    10-6

-------
     The method used to slake the lime also varies for commercial system
designs.  Ball mill, paste, and detention slakers are all used, but ball
mill and paste slakers are more common.  Each method is perceived to have
certain advantages.  Ball mill slakers generally produce a more finely
ground and possibly more reactive slurry and, unlike paste slakers, no grit
removal and disposal are required.  Paste slakers, on the other hand, have
lower capital costs, lower noise levels, and lower power requirements.
Also, they generally produce a less abrasive slurry than do ball mills,
which could reduce the potential for wear and maintenance of system pumps
and piping.

     As stated above, fabric filters have been chosen over ESPs in nearly
all the commercial spray drying applications to date.  Fabric filters may
have an advantage over ESPs in that unreacted alkalinity in the solids and
fly ash collected on the fabric surface can react with remaining SO  in the
flue gas as the gas passes through the fabric filter.  Some process
developers have reported that SCL removal across the fabric filter can
account for at least 10 percent of the total system SO  removal.  In some
applications, however, an ESP might be more desirable than a fabric filter.
The factors that are important in making the choice between ESPs and fabric
filters include:

       •  Use of recycle (increases dust loading to ESP, thus increasing ESP
          size and cost).

       •  Fly ash resistivity (high ash resistivity often requires larger,
          more expensive ESPs).

       •  Pressure drop (an ESP has a lower pressure drop and related costs
          than a fabric filter).

     Fabric filter designs for spray dryer FGD systems vary primarily with
regard to bag fabric, cleaning frequency, and cleaning mode.  Other more
subtle variations, too numerous to mention, between the well-established
fabric filter design practices of the different vendors also exist.
Teflon-coated fiberglass bags have been selected for many systems, although
acrylic bags may be a lower cost alternative for low gas temperatures.

DRY INJECTION

     A generalized flow diagram of the dry injection process is shown in
Figure 2.  The process involves pneumatic injection of a dry, powdery sodium
compound into the flue gas with subsequent particulate collection in a
fabric filter.  The point of alkali injection may vary from upstream of the
air preheater to the inlet of the fabric filter.  Reaction between the
reagent and SO  occurs both in the duct and on the filter bag surface (5,6).

     Although other alkaline reagents (e.g., lime and limestone) have been
tested, only certain sodium compounds have shown the capability for high  SO
removal from the flue gas.  Nahcolite and trona ores, which contain
naturally occuring sodium compounds, appear to be the most promising
reagents for dry injection in terms of reactivity and cost.  Nahcolite,
which is usually associated with western oil shale reserves, contains about
                                     10-7

-------
                                                                                                                    CLEAN GAS TO
                                                                                                                     ATMOSPHERE
                                    FLUE GAS
                    AIR PREHEATER
            REAGENT FROM
            BULK STORAGE
o
i
oo
REAGENT
HOLDiNG
  BIN
                  FLUE GAS AND
                     REAGENT
                                                                                                                        STACK
                                   REAGENT PULVERIZER
                                                                                             PRODUCT SOLIDS AND
                                                                                              FLY ASH DISPOSAL
                       INJECTION
                         FAN
          Figure  2.   Dry alkali injection flow diagram.

-------
80 percent sodium bicarbonate (NaHCO»).  Trona ore, which is found in large
deposits in Wyoming and California, contains sodium carbonate (Na CO ) and
sodium sesquicarbonate (Na CO • NaHCO • 2H 0).  Nahcolite has been shown to
be somewhat more reactive with SO  in the flue gas than trona (5,6).

     Major factors that affect SO  removal by nahcolite and trona injection
include stoichiometric ratio and the flue gas temperature at the point of
injection.  Higher S0» removals are obtained at higher normalized stoichio-
metric ratios (equivalent moles Na 0 per mole of inlet SO ) .  However,
higher stoichiometric ratios also result in lower reagent utilization (5).
Injection of the reagent at too low a temperature will reduce the initial
S02 reaction rate and may limit overall SO  removal.  For nahcolite it
appears that SO  removal drops off dramatically at injection temperatures
below 135°C (275°F) (6,7).

     Other parameters that are important in the dry injection process
include reagent particle size, mode of injection (batch, semi-batch, or
continuous), baghouse air-to-cloth ratio, and bag cleaning frequency.

ELECTRON-BEAM PROCESS

     The electron-beam (E-beam) process involves the irradiation of flue gas
containing a reactant, such as ammonia or lime.  The process removes both
SO,, and NO  from the flue gas and results in a dry waste product.
  L.       X

     A schematic diagram of the E-bearn/ammonia process is shown in Figure 3.
In this process, flue gas from a fly ash collection device is cooled and
humidified in a quench tower.  The resulting gas moisture content is about
10 percent.  Ammonia is injected into the cooled gas and the gas is passed
through an E-beam reactor.  In the reactor, oxygen and water are excited and
ionized to form the radicals [OH], [0], and [HO ] by the application of
electrons at a dose of 1 to 3 Mrads (1 Mrad is equivalent to 10 joules/g of
flue gas).  These radicals react with SO- and NO  to form sulfuric acid
(H?SO.) and nitric acid (HNO.,).  The acids are neutralized by ammonia and
water in the flue gas to form solid ammonium sulfate (NH.SO.) and ammonium
sulfate nitrate [(NH.) • SO,* NH.NO ].  The reaction time for formation of
the sulfate and nitrate salts is less than 1 second.  Product solids are
collected in a hopper below the E-beam reactor or in a downstream parti-
culate collector.

     In another version of the E-beam process, the water quench tower is
replaced with a lime-based spray dryer.  S0~ is removed from the flue gas in
the spray dryer and NO  and additional SO,., are removed in the E-beam
reactor.  Reactions inXthe E-beam reactor to form sulfuric and nitric acids
occur in the same manner as in the ammonia process.  The acids are
neutralized by Ca(OH)  to form calcium salts (CaSO , Ca(NO ) , and CaSO ).

     Bench- and pilot-scale studies of E-beam processes have shown that
variables important to performance include gas moisture content, gas
temperature, oxygen content, reagent ratio, and electron dosage.  In
addition, efficient penetration of the gas stream by the beam requires a
unique discharge pattern and other special design considerations.
                                     10-9

-------
o
I
                            TT
                                              QUENCH WATER
TT
                             T T T T T
                                                                   - AMMONIA
                   FLUE

                   GAS
E-GUN




E-BEAM
REACTOR


PARTICU-
LATE
COLLECTOR,
                                                                                              PRODUCT SOLIDS
                                DRAIN
                  Figure  3.   E-beam/anunonia process flow diagram.

-------
                           DEVELOPMENT STATUS

     Of the three dry SO  control technologies considered here, only spray
drying has been commercially applied.  However, dry injection technology is
nearing commercialization with plans for a 500 MWe system.  Research and
development activities are being conducted for all three technologies by
vendors, government agencies, and industry organizations.

SPRAY DRYING

     Through October 1983 there have been 17 commercial utility systems
sales, totaling over 6,800 MWe.  The general characteristics of the utility
systems, including vendor, size, location, design criteria, and projected
start-up date or current operational status, are shown in Table 2.  The
commercial utility spray drying systems have been selected for units firing
relatively low sulfur western coal or lignite (generally less than
1.5 percent sulfur fuel).  S0_ removal guarantees range from 60 to
91 percent, with design controlled emission levels ranging from 77 to
516 ng/J (0.18 to 1.2 lb/10  Btu).  Some of the guarantees stem from the
need to meet the most recent New Source Performance Standards (NSPS) for
utility boilers (8).

     Two commercial utility spray drying systems have been sold since
mid-1982, and a third system was awarded but is currently on hold.  These
awards were all to Joy Industrial Equipment Company and Niro Atomizer,  Inc.
The recently awarded utility systems are for units firing coal/lignite  with
average sulfur contents of less than 1 percent.

     According to the PEDCo survey prepared for EPA,  24 utilities
considering FGD for new coal-fired boilers have not yet made awards (9).  Of
these utilities, two are considering only dry FGD, while nine utilities did
not specify whether they were considering only wet, only dry, or both.
Table 3 shows the utilities, the planned units, and reported coal sulfur
contents for which FGD is being considered.  The data in Table 3 indicate
that dry FGD is not yet being widely considered for utility applications
involving moderate or high sulfur coal.

     The six commercial utility spray drying systems now operational are
listed in Table 2.  Of these the sodium-based system at Coyote Station was
turned over to Montana-Dakota Utilities for operation following performance
tests conducted in August 1982 (10) .  Earlier test results reported for this
system (11) showed 70 percent S0? removal at a soda ash utilization of
75 percent and an approach to saturation of 36°C (65°F).

     The 110 MWe unit at Northern States Power Company's Riverside Station,
initially a demonstration spray dryer coupled with existing ESPs and now
coupled to a fabric filter treating flue gas from two boilers, has also been
turned over to the utility.  Process testing of this system was performed
during the summer of 1983 under EPA and EPRI funding.  Results from these
tests are scheduled to be presented later at this symposium.
                                   10-11

-------
                                         TABLE 2.    COMMERCIAL UTILITY SPRAY DRYING  SYSTEMS
System Purchaser
Otter Tall Power Co.
(Operator: Montana-
Dakota Utilities)
United Power Assoc.
Marquette Board of
Light and Power
Basin Electric Power
Coop .
M
O
1 	 i Colorado Ute
K> Electric Assoc.
Basin Electric Power
Coop.
Basin Electric Power
Coop.
Station/Location
Coyote, Unit 1
(Beulah, ND)
Stanton, Unit 1A
(Stanton, ND)
Shiras, Unit 3
(Marquette, MI)
Laramie River,
Unit 3
(Wheatland, WY)
Craig, Unit 3
(Craig, CO)
Antelope Valley,
Unit 1
(Beulah, ND)
Antelope Valley,
Unit 2
(Beulah, ND)
Size3
MWe Start-up Date/Status
440 Operational. Turned
over to utility.
60 Operational. Not
turned over to
utility.
44 Operational. Not
turned over to
utility.
575 Operational. Not
turned over to
utility.
447 Late 1983. Under
construction.
440 Operational. Not
turned over to
utility.
440 April 1985.
Outlet SO
ng£J
Coal SO Removal Guarantee (lb/10 Btu)
North Dakota lignite; 70% for all coal.
0.78% S average.
North Dakota lignite; 91% for max. S coal.
0.77% S average;
1.94% S maximum.
Western subbitumi- 80% design efficiency.
nous; 1.5% S maximum.
Wyoming subbituminous 82% for avg. S coal;
0.54% S average; 90% for max. S coal;
0.81% S maximum.
Western subbituminous 87% for design coal.
0.7% S design coal.
North Dakota lignite; 62% for avg. S coal;
0.68% S average; 78% for max. S coal.
1.22% S maximum.
North Dakota lignite; 89% for max. S coal.
0.68% S average;
1.22% S maximum.
516
(1.2)
258
(0.6)
404
(0.94)
86
(0.20)
86
(0.20)
335
(0.78)
168
(0.39)
b ,, . c
Vendor
Rockwell/
Wheelabrator-Frye
Cottrell Environmental/
Komline-Sanderson
G.E. Environmental
Services
Babcock and Wilcox
Babcock and Wilcox
Joy/Niro
Joy/Niro
Tucson Electric Power
Tucson Electric Power
Platte River Power
Authority
Springerville,
Unit 1
(Springerville, AZ)

Springerville,
Unit 2
(Springerville, AZ)

Rawhide,  Unit 1
(Fort Collins, CO)
                                            370   June 1985.
                                            370   June 1987.
280   December 1983.
      Under construction.
New Mexico  subbltu-    61%.
minous;  0.69% S
average.

New Mexico  subbitu-    61%.
minous;  0.69% S
average.

Western  subbituminous  80%.
0.3% S average;
0.44% S  maximum.
                                                                                                                        258      Joy/Niro
                                                                                                                       (0.6)
                                                                            258      Joy/Niro
                                                                           (0.6)
  77      Joy/Niro
(0.18)
                                                                       (continued)

-------
                                                                          TABLE  2.    (CONTINUED)
o
 i
System Purchaser
Sunflower Electric
Sierra Pacific Power
Grand River Dam
Authority-State
of Oklahoma
Northern States
Power Company
Cajun Electric
Central and South-
western Services
Northern States
Power Company
Pacific Power and
Light
Station/Location
Holcomb, Unit 1
(Holcomb, K.S)
North Valmy
(Valmy, NV)
Pryor, Unit 2
(Pryor, OK)
Sherburne County
Unit 3
(Becker, UN)
Oxbow, Unit 1
(Coushatta, LA)
Coleto Creek
Unit 2
(Corpus Christ! , TX)
Riverside, Units 6
and 7 (Minneapolis,
MN)
Jim Bridger, Unit 2,
(Rock Springs, WY)
Size3
MWe Start-up Date/Status
319 In start-up
270 1984.
520 March 1985.
860 1990.
563 Projected for 1989
but need for system
being re-evaluated
710 1988.
110 Operational. Turned
over to utility.
100 Shut down. Demon-
stration testing by
vendor completed.
No plans for restart.
Coal SO- Removal Guarantee
Wyoming subbituminous; 80%.
0.34% S average.
Western subbituminous; 76% for all coal.
0.4 to 1.0% S.
Western subbituminous; 85% for all coal.
0.4 to 1.5% S.
Sarpy Creek subbitu- 70% minimum.
minous; 0.9% S average;
0.4 - 2.3% S range.
Louisiana lignite; 80%.
0.93% S average.
Subbituminous; 0.4% S 70%.
average.
Varies with test Varies with test
series. program.
Western subbituminous; Lime: 62-75%
0.56% S average; Soda ash: 74-86%
0.8% S max imum .
Outlet SO,
ng/J .
(lb/10 Btu) Vendor0
116 Joy/Niro
(0.27)
258 Rockwell
(0.6)
NR Flakt
258 Joy/Niro
(0.6)
258 Joy/Niro
(0.6)
NR Joy/Niro
516 Joy/Niro
(1.2)
129 Flakt
(0.3)
          Gross MWe output.

         bSource for SO  outlet emissions:  Ireland, P.  A.  (Sterns-Roger) Status of Spray Dryer  Flue Gas Desulfurization.  EPRI CS-2209.   Final  Report.  EPRI,
          Palo Alto, CA?, January 1982, p. 2-3.

         CRockwell/Wheelabrator-Frye is no longer a joint venture.  Both are, however, selling spray drying systems.  Joy/Niro:  Joy takes the lead  in utility sales
          while Niro takes the lead in industrial sales.    G.E.  Environmental Services was formerly Buell Emission Control Division, Envirotech  Corp.  Spray dryer
          supplied by Anhydro A/S and fabric filter supplied  by  G.E.E.S.  Cottrell Environmental Sciences supplies fabric filter and Komline-Sanderson supplies
          spray dryer.

-------
         TABLE 3.  UTILITIES PLANNING OR CONSIDERING FGD SYSTEM THAT
                   HAVE NOT AWARDED BIDSS  (as of August 3, 1983)

Type of FGD
System Projected Unit
Utility Considered Coal (%S) Start-Up
Atlantic City Electric
Buckeye Power
Central Illinois Light
Central Maine Power
Cincinnati G&E
Colorado Ute Electric
Delmarva Power & Light
Desert Gen & Trans
Florida Power & Light

General Public Utilities

Indianapolis Power & Light

Kentucky Utilities
Louisville G&E

Nebraska Public Power
Nevada Power

Orlando Utilities Commission
Pacific Power & Light


Salt River Project
Seminole Electric

Southwestern Electric Power

Texas Utilities
WA Water & Power

West Texas Utilities
Wet
Wet
Wet
Wet
NR
NR
Wet
Wet
NR

NR

Wet

Wet
Wet

Dry
NR

Wet
Wet (retrofit)

Dry
Wet
NR

NR

NR
Wet

NR
3. £5
NR
3.3
2.23
4.0
0.5
2.5
0.5
NR

3.5

3.5

3.5
4.0

0.36
NR

NR
0.56

NR
0.6
NR

1.5

0.8 (lignite)
NR

0.34
1988
1990
1989
1989
1988
1988
1987
1992
2 units;
1992, 1993
2 units;
1993, 1994
3 units;
1990
1994
2 units;
1986, 1990
1987
4 units;
1984-87
1987
3 units;
1986-90
1986
1989
2 units;
1988, 1990
2 units;
1990
1989
4 units;
1990-94
1996
^Source:   Reference  9.
 NR - Not  reported.
                                  10-14

-------
     Three of the other four operational spray drying systems (Stanton,
Shiras, and Antelope Valley) have all recently met performance guarantees.
However, no test results have been reported yet for any of these systems.

     One other utility system is in the initial start-up stage, and two
systems are scheduled for start-up before the end of 1983.

     In addition to the utility systems, a number of industrial coal-fired
boilers are, or will be, equipped with spray drying FGD systems.  Table 4
gives a summary of the 7 operational and 14 planned industrial spray drying
units.  Coal sulfur contents for the industrial applications range from 0.6
to 3.5 percent, with seven systems designed for coal sulfur contents of
3 percent or greater.  Removal guarantees for the industrial systems range
from 70 to 90 percent.

     Performance test results have been reported for five of the operational
industrial systems.  A continuous monitoring program conducted by EPA on the
Celanese system showed 70 percent average S0? removal over a 23-day test
period (12).  Removal averaged 75 percent on days when the daily inlet S0?
concentrations were 1720 ng/J (4.0 lb/10  Btu) or greater.  It should be
noted that the tests were conducted during a period of fluctuating boiler
operation.

     Performance tests of the Strathmore Paper Company system showed an
average SCL removal of 92.4 percent (90.1 to 96.7 percent range) at an inlet
SO  concentration of about 2000 ppm (13).  Data on lime stoichiometric ratio
or other operating conditions required to achieve these removals were not
included in the performance test results.  This system and the Celanese
unit, which do not employ recycling of solids, have now been turned over to
the system purchasers.

     Compliance tests conducted on the Argonne system showed SO  removals
ranging from 79.7 to 95.6 percent at lime stoichiometries ranging from 0.8
to 2.0, respectively (14).  These tests were conducted on a 3-percent sulfur
coal at an approach to saturation of approximately 13°C (23°F).   In this
system, receiving flue gas from a stoker-fired boiler and described in a
presentation to be made later in this session, solids from the flue gas
cleaning system are recirculated to the spray dryer.

     This system has now been turned over to the Argonne National
Laboratory, and EPA-funded process testing of the system is on-going.
Preliminary results from a 100-hour test run on a 4.6 percent sulfur coal
showed 90 percent SO  removal at a lime stoichiometry of 1.4.

     The Container Corporation and Austell Box Board systems have both met
compliance requirements, and the former system has been turned over to the
system purchaser.  Compliance tests on the Container Corporation system
showed average outlet SO  emissions of 14.3 ng/J (0.03 lb/10  Btu) on a
0.6 percent sulfur coal fl5).  Tests on the Austell Box Board system  (16)
showed an average outlet SO  emission rate of 366 ng/J (0.85 lb/10  Btu).
No data on reagent ratios have been reported for these two systems.  The
system at GM-Buick Motors and one unit at the University of Minnesota have
only recently become operational, and no performance data have yet been
released.
                                   10-15

-------
                           TABLE 4.   COMMERCIAL INDUSTRIAL BOILER  SPRAY DRYING SYSTEMS
System
Purchaser
Celanese Fibers
Co.
Strathmore Paper
Co.
University of
Minnesota
Argonne National
Lab
Container
Corporation
GM-Buick Motors
Fairchild AFB
Austell
Box Board Co.
Puget Sound
Naval Shipyard
Maelstrom AFB
Griffis AFB
Location
Cumberland, MD
Woronco, MA
Minneapolis ,
MN
Argonne , IL
Philadelphia,
PA
Flint, MI
Units 1,2,3
Spokane , WA
Austell, GA
Units 1,2,3
Bremerton, WA
Units 1,2,3
Great Falls, MT
Units 1,2,3,4
Rome, NY
Size, Mg/hr
(Ib/hr)
Steam
50
(110,000)
39
(85,000)
Two units @
40 MW each
77
(170,000)
77
(170,000)
204
(450,000)
3 @ 50
(110,000)
114
(250,000)
3 @ 64
(140,000)
3 @ 41
(90,000)
hot water
4 (3 41
(90,000)
Start-up Date/Status
Operational. Turned over
to purchaser.
Operational. Turned over
to purchaser.
One unit operational.
Second start-up in
September 1983. Not turned
over to purchaser.
Operational. Turned over
to purchaser.
Operational. Turned over
to purchaser.
Operational. Not turned
over to purchaser.
Initial start-up
stages .
Operational. Not turned
over to purchaser.
Late 1987.
Spring 1985.
Late 1984
Coal
1.5% S to 2.5% S eastern
coals .
2.3 to 3% S eastern coal.
Subbituminous coal; 0.6
to 0.7% S.
Bituminous coal; 3.5% S.
1.0% S eastern coal.
Indiana bituminous coal;
1 to 3% S.
1.0% S western coal.
Bituminous coal; 1.0
to 2.5% S.
1.6% S maximum.
Western subbitiuminous
coal; 1.0% S.
Bituminous; 3% S
S02 Removal
Guarantee
70% for 1.5% S coal.
87% for 2.5% S coal.
75%.
70%.
78.8% (516 ng/J or
1.2 lb/10 Btu outlet
so2).
90%.
70 to 90%.
85%.
516 ng SO /J (1.2 Ib
S02/10 Ecu outlet).
84%.
85%.
85% (0,71 Ib
S02/10 Btu outlet).
Vendor
Rockwell/
Wheelabrator-Frye
Mlkropul
Flakt
Niro/Joy
Ecolaire
Niro/Joy
Niro/Joy
Wheelabrator-Frye
G.E. Environmental
Services
Niro/Joy
Ecolaire
Rockwell/Wheelabrator-Frye is no  longer a  joint
industrial system sales, and Joy  takes the lead
venture.  Both are, however, selling spray drying systems.  Niro of Joy/Niro  takes the lead in
in utility system sales.

-------
     In addition to the commercial sales, there are a number of spray drying
research and development activities currently underway.  Table 5 lists the
major current activities.  These programs, which are being funded by EPA,
EPRI, and DOE, are designed to address information needs in the areas of
high sulfur coal applications, alternate spray drying sorbents, and
sustained operation at a close approach to saturation.  Results from several
of the programs are scheduled to be presented at this symposium.  On-going
research by several vendors stresses the further improvement of atomization
techniques with goals of longer service life, lower energy consumption, and
improved slurry droplet size distribution.  Rotary and nozzle atomizers are
being compared on the basis of cost and performance.

DRY INJECTION

     The first planned commercial application of dry injection technology
has been announced by Public Service Company of Colorado for a 500 MWe unit
scheduled for start-up in 1990 (17).  The system, which will use trona ore
as the sorbent, will be designed for 70 percent SO  removal on a 0.4 percent
sulfur western coal.  The trona will be injected just upstream of the fabric
filter into flue gas at approximately 132 to 138°C (270 to 280°F).  A clay-
and plastic-lined landfill will be used for solids disposal.

     The 22 MWe demonstration testing at Public Service of Colorado's Cameo
Station has been completed.  Final reports on the EPRI-sponsored dry
injection program have been prepared (18,19).  Results of the testing showed
70 percent SO- removals for trona and nahcolite at stoichiometric ratios of
approximately 1.3 and 0.8, respectively.  These tests were conducted on a
low sulfur western coal yielding flue gas at 450 ppm S0_.

     A recent cost study based on the Cameo results indicates that dry
injection may be less expensive than spray drying, depending on reagent
cost, coal sulfur content, and S0_ removal requirements (7).   Estimated
costs for nahcolite and trona dry injection systems and a lime spray drying
system are shown in Table 6 for two-500 MWe units firing 0.48 percent sulfur
coal with 70 percent S0_ removal.  These estimates assume nahcolite
(70 percent NaHCO ) and trona (85 percent Na2C03) costs of $100 and $75 per
ton, respectively, and sodium-based waste disposal costs of $7.40/ton.

     The application of trona dry injection had been previously constrained
by questions regarding S09 removal limitations and cost.  However, the
recent demonstration-scale studies have shown that S02 removal efficiencies
of 70 to 80 percent can be achieved with trona on low sulfur coals at
reasonable stoichiometric ratios.  Trona ore is currently mined in large
quantities for conversion to sodium carbonate.

     The commercial application of nahcolite dry injection hinges on the  ,
economic availability of nahcolite.  Nahcolite is not currently mined in the
United States, but at least one firm has announced intentions to develop a
nahcolite mining operation and several other companies are investigating the
possibility of supplying nahcolite through solution mining techniques (20,21)
                                    10-17

-------
                                     TABLE 5.   MAJOR CURRENT  SPRAY  DRYING R&D ACTIVITIES
o
h-'
00

Vendor, Agency, or
Industry Group
Electric Power Research
Institute (EPRI)3
Size
255 m /min
(9000 cfm)
Location
Arapahoe pilot plant;
Denver, CO.
Comments
Test program began in March 1982.
Testing planned through December
                                                                             1983.
            EPRI/EPAC
            EPA
   100 MWe
   77  Mg/hr
(170,000 Ib/hr)
   steam
Riverside Station,
Northern States Power.
Minneapolis, MN.

Argonne National Lab.
Argonne, IL.
Ten weeks of testing beginning in
July 1983, including 3.4% S coal.
Performance test (90-day), start-
ing in November 1983, on 3.4% S
coal.
DOE/Babcock and
•3
Wilcox
25 m /min
(900 scfm)
B&W Alliance
Laboratory.
Alliance, OH.
Two-year test program began in
mid-1982 on eastern high sulfur
coals.

            Papers  on  these  studies will  be  presented  at  this  symposium.
            Actual flue gas  conditions.

-------
     TABLE 6.  COMPARISON OF COSTS FOR NAHCOLITE AND TRONA DRY INJECTION
               AND LIME SPRAY DRYING ON TOO NEW 500-MW COAL-FIRED
               POWER-GENERATING UNITS
    FGD Process
                        Capital
                      Cost ($/kW)
Levelized Costs
  (mills/kWh)
Trona injection
                         89.1
Nahcolite injection
                         90.4
      9.6
Lime spray drying
                        175.9
     10.!
Notes:    All costs include baghouse for particulate control.  Upper midwest
          plant firing 0.48 percent sulfur coal at 8,020 Btu/lb.  1983
          revenue requirements,

          Trona system:  70 percent S0? removal, $75/ton for trona
          (85 percent Na0CO ), $7.40/ton for waste disposal.

          Nahcolite system:   70 percent SO.., removal, $100/ton for nahcolite
          (70 percent NaHCO ), $7.40/ton for waste disposal.

          Lime system:  70 percent S00 removal, $60/ton for lime, $4.60 for
          waste disposal.
Source:
Reference /.
                                    10-19

-------
     Another important consideration for dry injection applications is the
high solubility and leachability of the sodium-based wastes.  Application of
dry injection may be limited in some locations by the costs of disposing of
the wastes in an environmentally safe manner.

ELECTRON-BEAM PROCESS

     The electron-beam process is in an early developmental state.  The DOE
has signed cost-sharing agreements with Research-Cottrell and EBARA/Avco-
Everett to conduct pilot-scale demonstrations of E-Beam processes.
Research-Cottrell will be developing the E-beam/lime process and EBARA/Avco-
Everett will be developing the E-bearn/ammonia process.

     Testing of the E-beam/lime process is scheduled to begin in late 1983
on a 283 to 425 m /min (10,000 to 15,000 acfm) pilot plant at TVA's Shawnee
power plant.  The tests will be conducted on flue gas from firing a 2.5 to
3 percent sulfur coal (22).   Test plans for a 570 Nm /min (20,000 scfm)
pilot E-bearn/ammonia system are being finalized with plans calling for
construction to begin in October 1983 (23) -  This unit will also be tested
on flue gas from firing a 2.5 to 3 percent sulfur coal.

                                   SUMMARY

     Currently over 6,800 MWe of electrical utility generating capacity has
been committed to spray drying SCL control.  Utilities are operating two
systems (550 MWe), and four systems (1,120 MWe) are operational but not yet
turned over for utility operation.  Recent sales of industrial spray drying
have been mainly to Federal government installations, with the industrial
systems now sold totaling 21 units.  Seven of these units are operational,
with four units now being operated by their owners.  While spray drying
currently remains as the only commercially applied dry FGD technology, plans
for a new 500 MWe electrical generating unit specify the dry injection of
trona for S0» control at a Colorado site.

     Research and development (R&D) on spray drying and the E-beam process
is continuing.  Both EPA and EPRI have active spray drying R&D programs.
DOE is supporting the E-beam studies which are currently at the pilot level.

                                 REFERENCES

1.   Blythe, G. M., J. C. Dickerman, and M. E. Kelly (Radian Corporation).
     Survey of Dry S02 Control Systems.  EPA-600/7-80-030 (NTIS PB 80-
     166853), U.S. Environmental Protection Agency, Industrial Environmental
     Research Laboratory, Research Triangle Park, N.C., February 1980.

2.   Kelly, M. E. and S. A.  Shareef (Radian Corporation).  Second Survey of
     Dry^S02 Control Systems.  EPA-600/7-81-018 (NTIS PB 81-157919), U.S.
     Environmental Protection Agency, Industrial Environmental Research
     Laboratory, Research Triangle Park, N.C.,  February 1981.

3.   Kelly, M. E. and S. A.  Shareef (Radian Corporation).  Third Survey of
     Dry^S02 Control Systems.  EPA-600/7-81-097 (NTIS PB 81-218976), U.S.
     Environmental Protection Agency, Industrial Environmental Research
     Laboratory, Research Triangle Park, N.C., June 1981.
                                   10-20

-------
4.   Kelly, M. E. and M. A. Palazzolo (Radian Corporation).  Status of Dry
     SO  Control Systems: Fall 1982.  EPA-600/7-83-041 (NTIS PB 83-247585),
     U.S. Environmental Protection Agency, Industrial Environmental Research
     Laboratory, Research Triangle Park, N.C., August 1983.

5.   Apple, C. and M. E. Kelly (Radian Corporation).  Mechanisms of Dry SO
     Control Processes.  EPA-600/7-82-026 (NTIS PB 82-196924), U.S.
     Environmental Protection Agency, Industrial Environmental Research
     Laboratory, Research Triangle Park, N.C., April 1982,  pp. 97-98.

6.   Muzio, L. J., et al. (KVB, Inc. and Electric Power Research Institute).
     22-MW Demonstration of Dry S0? Scrubbing With Sodium Sorbent Compounds.
     Paper #83-38.5, presented at the 76th Annual Meeting of the Air
     Pollution Control Association, Atlanta, Georgia, June  1983.

7.   Naulty, D. J., et al.  (Stearns-Roger Engineering).   Economics of Dry
     FGD by Dry Sorbent Injection.  Paper #83-38.6, presented at the 76th
     Annual Meeting of the Air Pollution Control Association, Atlanta,
     Georgia, June 1983.

8.   U.S. Environmental Protection Agency.  New Stationary  Sources
     Performance Standards; Electric Utility Steam Generating Units.  In:
     Federal Register, Vol. 44, No, 113, June 11, 1979, pp. 33580-33624.

9.   EPA Utility FGD Survey;  On-line Information System.  U.S.  Environ-
     mental Protection Agency, Industrial Environmental Research Laboratory,
     Research Triangle Park, N.C., August 3, 1983.

10.  Gehri, Dennis (Rockwell International).  Telephone conversation with
     M. A. Palazzolo, Radian Corporation, August 3, 1983.

11.  Lewis, M. F. and D. C. Gehri (Montana-Dakota Utilities and Rockwell
     International).  Atomization - The Key to Dry Scrubbing at the
     Coyote Station.  In Proceedings:  Symposium on Flue Gas Desulfuri-
     zation - Volume 2, Ayer, F.  A. (ed.).  Electric Power  Research
     Institute, Palo Alto, CA, March 1983, pp. 673-688.

12.  Mostardi-Platt Associates, Inc.  Particulate and Gaseous Emission Study
     Performance for Strathmore Paper Company at the Woronco Mill Power
     Plant, Woronco, MA.  Bensenville,  IL, May 7, 1981.

13.  Reinauer, T. V., et al.  (Mikro Pul Corporation).  Dry FGD on an
     Industrial Boiler.  Chemical Engineering Progress, _79_(4) , March 1983.

14.  Farber, P. S. (Argonne National Laboratory).  Startup  and Performance
     of a High Sulfur Dry Scrubber System.  Paper #82-40.5, presented at the
     75th Annual Meeting of the Air Pollution Control Association,
     June 1982.

15.  Air Techniques, Inc.  Compliance Particulate, SO,., and NO Emission
     Testing on Coal-Fired Boiler at Anstell Box Board Corporation, Austell,
     GA.  Marietta, GA, April 1983.
                                    10-21

-------
16.   Glazer,  Norman (City of Philadelphia Air Management).  Letter and
     attachment to M.  A.  Palazzolo,  Radian Corporation, August 19, 1983.

17.   Green,  George (Public Service Company of Colorado).  Telephone
     conversation with M. A. Palazzolo, Radian Corporation, July 27, 1983.

18.   Muzio,  Larry (KVB, Inc.).   Telephone conversation with M. A. Palazzolo,
     Radian Corporation,  August 2, 1983.

19.   Muzio,  L.  J., et  al.  Dry  S0?-Particulate Removal for Coal-Fired
     Boilers -  Volume  1;   Demonstration of SO,., Removal on a 22-MW Coal-
     Fired Utility Boiler by Dry Injection of Nahcolite.  EPRI CS-2894,
     Electric Power Research Institute, Palo Alto,  CA, March 1983.

20.   Shah, N. D.  (Multi-Mineral Corporation).  Dry  Scrubbing of SO .
     Chemical Engineering Progress,  ^7_8_(6), June 1982.

21.   Solution Mining of Nahcolite (Natural Sodium Bicarbonate) May Begin
     by 1986.  Chemical Engineering, _90_(12):20, June 13, 1983.

22.   Williams,  John (Department of Energy, Pittsburgh, PA).  Telephone
     conversation with M. A. Palazzolo, Radian Corporation, August 3, 1983.

23.   Frank,  Norman (EBARA).   Telephone conversation with M. A. Palazzolo,
     Radian Corporation,  August 15,  1983.
                                  10-22

-------
 ACID RAIN PREVENTION THRU NEW SOX/NOX DRY
            SCRUBBING PROCESS

K. S. Felsvang, P. Morsing,  P.  L.  Veltman

-------
     ACID RAIN PREVENTION THRU NEW SOx/NOy DRY SCRUBBING PROCESS
     by:      Karsten Felsvang and Per Morsing
              A/S NIRO ATOMIZER, Soeborg,
              Denmark

              Preston Veltman
              Niro Atomizer Inc., Columbia,
              Maryland, U.S.A.
                                    ABSTRACT
     The issue of acid rain has attracted much attention both in the U.S. and in Europe.
To  cope with  problems  associated  with  acid  rain Niro  Atomizer is developing a dry
scrubbing process for simultaneous removal of SOX and NOX.

     A  description  of  the  process is given in  this paper. The  equipment  used is
essentially  the same  as  used in  the over 5000 MW  utility  dry  scrubbers  currently in
operation, start-up or under construction.

     Pilot  plant test results  from the Copenhagen dry FGD facility and results achieved
during  a full-scale demonstration of the  process  at  one of JOY/NIRO's operating dry
scrubbers are presented.

     Waste product characteristics are  shown and compared with EPA standards.

     Finally,  the  process economics are  analyzed and compared with  other  existing
processes for SOX/NOX removal.


                                 INTRODUCTION

     In  recent years the issue of acid rain has attracted much  attention both in the
United States and in Europe.

     In the U.S.A. signs of acidification of the environment have  become obvious in the
North-East. Discussions focus on  how big a reduction  of SOX and NOX should be  required
(1). The  U.S. EPA is evaluating which technique or combination of techniques would be
the overall most cost effective in meeting the possible requirements (2, 3).

     In Europe the 1982 Stockholm  conference "Acidification Today and Tomorrow" (4)
likewise concluded that  dispersion of SOX and NOX arising  from combustion  of  fossil
fuels has a dramatic impact  on the environment. Early this year a 30% reduction of SOX
                                        10-23

-------
was suggested by Scandinavian countries, and  recently the German conference "Acid
Deposition - A Challenge for Europe" (5) again focused on the effects of acid deposition
and abatement technologies available.

     A  number  of  techniques has  been developed  for  reduction  of SOX  and NOX
separately. Recent developments  on  low NOX burner  technology  look  promising. Con-
siderable progress has  been  made in this  area and  achievable  emission  levels are
approaching the 0.1 - 0.2 Ib/MBtu range (6).

     To encourage development of processes for simultaneous removal of SOX and NOX
the U.S. Department of Energy and  the  U.S. EPA  have  funded a number  of research
programs that should lead to more  cost effective methods for SOX/NOX control (7).

     Dry  scrubbing has in the  recent  years  gained acceptance as an  attractive FGD
technique for many utility companies. Niro Atomizer has developed a  new dry scrubbing
process for simultaneous removal  of SOX and NOX (8). The results of this  development
are presented in this paper.

                    DESCRIPTION OF NEW SOX/NOX PROCESS

     The new dry scrubber process for simultaneous SOX and NOX removal (9)  utilizes
the same equipment  as  currently  supplied for the JOY/NIRO dry FGD system.  Normal
operating conditions  in these  systems are low approach temperatures such as 18°F to
achieve the lowest  lime consumption and the highest SOX scrubbing efficiency. Under
these conditions virtually no NOX absorption is achieved.

     The new SOX/NOX process  is  schematically  shown in Figure 1.
FLUE GAS






DRYER j




Tout
195*F




(
v_>
               (3-1
                   U V
                    FEED TANK
          Figure 1. Niro Atomizer Dry Scrubber SOX/NOX Process
                                      10-24

-------
      A key feature in  the new process  is the existence of a so-called thermal window
where simultaneous SOX and NOX removal can be obtained. The process is operated with
an outlet temperature of  the spray dryer absorber of 195 - 215°F. As  in the normal dry
scrubbing process lime  is  the main reagent. The process is operated in a recycle mode,
i.e. recirculated dry waste product and  slaked lime are mixed in the feed preparation
system to a slurry with a solids content  of 30-40%,  before being atomized in  the spray
dryer absorber module.  An inexpensive additive is required to make the process work in
the mentioned thermal  window. An additive, such as sodium  hydroxide, is simply added
into the  feed  preparation  system in a quantity of approximately 10% of the lime added
In the Niro Atomizer spray dryer absorber (SDA) the efficient atomization and  intimate
mixing with the flue gas create a uniform recycle product which under close control is
impregnated with a minimal amount of sodium sulfite formed by  the selective  reaction
of sodium  hydroxide  with SO2. Hereby  a  effective solid  sorbent is created for  the
simultaneous removal of SOX and NOX. A  major portion of the SOX is removed in  the
SDA, but only minor amounts of NOX is absorbed. The conditioned flue gas and activated
particulates then enter the fabric filter where simultaneous  removal of NOY and SOY is
accomplished.

         RESULTS OF BENCH  SCALE AND PILOT PLANT INVESTIGATIONS

BENCH SCALE INVESTIGATIONS

      As  NO  has  very  little  affinity to water the SOX/NOX research program was
directed  towards developing  active solid materials  that  would promote  an SOX/NOX
gas/solids reaction.

      Bench scale tests were performed using a fixed bed reactor with a layer of reactive
powder through which  the flue  gas was  passed (Figure 2). The  powder used was the
normal dry scrubber waste solids including calcium sulfite, calcium sulfate and calcium
hydroxide. The thickness of the  powder layer and the gas  velocity through the bed was
adjusted to simulate as closely as possible the conditions in  a baghouse filtercake.
                                          REACTOR
                       TEMPERATURE &
                       HUMIDITY CONDITIONING
        TTTT
         AIR  N2 NO SO2
        Figure 2.  Bench Scale Set Up For SOX/NOX Tests
                                                           RECORDER
                                      10-25

-------
Effect of Additive

     Various  additives  including   sodium  sulfite,   Fe++-compounds  and  Ethylene-
diaminetetra-acetic acid, disodium salt (EDTA) were  used  to impregnate  the  bed of
powder. Sodium sulfite  was found, for technical and economical reasons,  to  be the
preferred reagent.

     Tests with a soda ash based dry scrubber waste, which has the sodium sulfite in
situ, showed no substantial NOX absorption.

Effect of Reaction Temperature

     Normal lime based dry scrubber powder  was impregnated with  small  amounts of
sodium  suifite  and  was subsequently exposed in the fixed bed reactor to various dry
scrubber operating  temperatures. The  results  shown  in Figure 3 clearly indicate the
existence of the so-called thermal window where SOX and NOX react simultaneously.
% Removal  (1 h average)


90

80 •

70 •-

60 •-

50


40 ••

30

20

10
                                  THERMAL

                                       WINDOW
                                                     SO,
                                                     NO,
                                                          Legend:
                                                          Inlet SOX = 300 PPM

                                                          Inlet NOX- 300 PPM
                                                   Temp. °F
                     158   176
                      194
212
                                                230
             Figure 3. Influence of Reaction Temperature on SOX/NOX Reaction

Effect of SOY/NOV Ratio

     The  process is a simultaneous removal of SOX and NOX. If no SOX is present no NOX
removal is achieved.  The ratio between SOX and NOX  should  preferably be  1 (one) or
greater. Table 1  shows the influence of SOX/NOX ratio.
                                       10-26

-------
                                    TABLE 1.

              INFLUENCE OF SOX/NOX RATIO ON NOX REMOVAL
sox
PPM
0
100
300
NOX
PPM
300
300
300
SOX/NOX
RATIO
0
1/3
1
NOX
REMOVAL
(1st hour)
0
34
70
SOX
REMOVAL
(1st hour)
_
92
91
Effect of Other Parameters

     Other parameters such as the flue gas oxygen content and the moisture content of
the powder also affect  the  SOX/NOX reaction. Table 2 & 3 show  the  results. Under
normal power plant operating conditions the flue gas will contain 5% ©2 or more, so no
negative  effect on the SOX/NOX reaction is expected.


                                    TABLE 2.

           INFLUENCE OF FLUE GAS OXYGEN ON SOX/NOX REACTION
O2 %
FLUE GAS
0.15
0.80
1.5
5.0
NOX
(1st hour)
0
>47
56
70
SOX
(1st hour)
80
92
92
91
     Table 3 shows that a certain  residual moisture content has to  be  present in the
powder to achive a substantial SOX/NOX removal.

                                    TABLE 3.

           INFLUENCE OF POWDER MOISTURE ON SQXNQX REACTION
MOISTURE
MOIST POWDER
DRY POWDER
NOX
70
30
SOX
91
34
 PILOT PLANT TEST RESULTS

      Following the bench scale investigation a comprehensive pilot plant test program
 was carried out at the Niro Atomizer in-house FGD facility in Copenhagen. This pilot
 plant  is  routinely used to develop dry FGD system design parameters and provides very
 accurate scale-up information.
                                      10-27

-------
    Table 4 (10) shows the correlation between pilot plant and full scale test results.
                                  TABLE 4.

    COMPARISON OF PILOT AND FULL SCALE (RIVERSIDE) TEST RESULTS***
PLANT
PILOT
RIVERSIDE
PILOT
RIVERSIDE
TEST
No.
6110-6115
20189
6120-6123
20186
OPERATIONAL PARAMETERS
TEMPERATURES (°F)
PPM
S02
700
710
768
758
SDA
IN
31*
315
3*3
3*0
SDA'
OUT
1**
1*3
1*6
1*6
BH**
OUT
136
1*2
1*0
1*6
T
Adsat.
18.5
18.7
18.2
18.7
CRT
Sec.
12.9
13.2
10.2
10.5
FEED
SOLIDS
(%)
33.5
31.1
32.0
32.6

AIR
TO
CLOTH
RATIO
1.25
1.15
1.58
1.**

CLEAN-
ING
CYCLE
1 hr
1 hr
1 hr
1 hr
RESULTS
% SO2
REMOVAL
TOTAL
96.3
96.8
97.7
96.*
SR
1.02
1.09
1.15
1.20
  *  SDA: Spray Dryer Absorber  ** BH: Baghouse

  *** Pilot Plant: 5,000 ACFM - Riverside Full-Scale Plant: 500,000 ACFM


     The existence of the thermal window was confirmed during the pilot tests. Table 5
shows results  from  parametric tests that were not performed under optimum conditions
with  respect to other parameters.
                                   TABLE 5.

   PILOT PLANT DEMONSTRATION OF THERMAL WINDOW IN SOX/NOX PROCESS
SPRAY DRYER ABSO
INLET
TEMP.
oF
330
sox
PPM
1500
NOX
PPM
500
RBER
OUTLET
TEMP.
oF
151
194
214
230
266
% NOX REMOVAL
SDA
0
6
8
5
5
SDA+BH
0
21
35
37
30
% SOX REMOVAL
SDA
72
54
35
25
15
SDA+BH
100
83
67
55
34
                                    10-28

-------
     These tests with  the spray dryer absorber / baghouse combination aiso reveal that
the major part of the SOX/NOX reaction takes place in the filter cake on the  baghouse
bags. Only minor amounts of NOX are removed in the spray dryer absorber.

     Pilot  plant  results  confirmed the influence of the  ratio between  SOX  and  NOX
removal. Table 6 shows typical results.
                                    TABLE 6.

                        EFFECT OF SOX/NOX RATIO
SO2 LEVEL
INLET SDA
PPM
NOX LEVEL
INLET SDA
PPM
% NOX
REMOVAL
1000

300

40
1600

300

50
2500

300

65
     As can be seen from  the  table the highest  NOX  removal efficiency is  -  at  the
present state of development - achieved with higher sulfur coals . Pilot plant test  results
for typical high sulfur conditions are shown in Table 7. These tests were performed with
addition of  sodium hydroxide to the feed  tank. A  quantity  of 10  weight  percent in
relation to the lime flow was used.

                                    TABLE 7.

PILOT PLANT  RESULTS WITH SOX/NOX PROCESS APPLIED TO HIGH SULFUR COAL

% Solids in Feed Slurry
T . + Temp. °F
Inlet v
SOX PPM dry
SDA NOX PPM dry
Inlet
BH TemP' °F
% NOX Removal
% SOX Removal
Lime Stoichiometry
TEST
No. 12138
34
371
2814
310
188
65.5
98.8
1.44
TEST
No. 12149
30
336
2762
309
222
62.5
86.2
1.10
TEST
No. 12150
36
334
2638
346
204
57.2
93.3
1.20
                                       10-29

-------
REACTION MECHANISM OF NEW SOX/NOX PROCESS

     Based on the absorption results and chemical analyses of the SOX/NOX dry scrubber
product, the reaction mechanism is illustrated in Figure 4.
                 SPRAY DRYER ABSORBER

                           SOX   NO,
                                                   SO,  NOX
                                                           H2O Monolayers
                            LIQUID DROPLET
                                               HIGHLY REACTIVE
                                               SOLID PARTICULATE
                 FABRIC FILTER
                                            SOX +• O2
                                            NOX + O2
                                            H2O
-a(N03)2
                                                      Catalysis
                                                     at active site
                           SOLID PARTICULATE

                  Figure 4. SOX/NOX REACTION MECHANISM

      The added sodium hydroxide is converted to water  soluble sodium sulfite during
the absorption  process  in the  spray  dryer absorber.  The  sodium  sulfite  is partly
precipitated as amorphous solids and partly kept in solution in the residual moisture, thus
impregnating the spray dried solid particle of recycle solids and  calcium  hydroxide
reaction products. Figure  5 shows  a scanning electron microscope (SEM) photograph of
such a particle.
                       Figure 5.  SEM-PHOTO OF ACTIVE SITES
                                       10-30

-------
     The solid participate  with the active  sites enters  the fabric filter and  forms a
highly porous filter cake. According to Klingspor (11) between one and two monolayers of
water will still exist on the solids surface in the thermal window where the process is
operated. No NOX absorption could be  accomplished if too  many  monolayers were
present. If on the other hand no water was present virtually no SC>2 absorption could be
achieved.

     At  the active site SOX and NOX are oxidized producing calcium sulfate,  calcium
nitrate and calcium hydrogen sulfate, nitrate.

     The proposed reaction  mechanism is based on the following observations:

           Oxygen and water are necessary to obtain NOX removal.

           The amount of NOX absorbed  agrees with the  amount of nitrate found in the
           reaction product.

           The amount of sulfite in the reaction product is increased when compared
           with normal SOX dry scrubber  products.

                   RESULTS OF FULL-SCALE DEMONSTRATIONS

ARGONNE DEMONSTRATION

     Since November 1981,  i.e. for 2 years,  Argonne National Laboratory has operated
the first commercial JOY/NIRO dry scrubber designed for high (3.5%) sulfur coal (12,
13). The  scrubber treats flue gas from a Spreader-Stoker boiler rated at 170.000  Ib/hr of
steam. The coal is from the Illinois basin. A typical analysis is shown in Table 8. System
characteristics with respect to normal operation of the spray dryer  absorber  and  the
baghouse are given in Table  9.

                                     TABLE 8

          COAL  CHARACTERISTICS FOR ARGONNE HIGH SULFUR COAL
COAL PARAMETER
Heating value Btu/lb
Moisture %
Ash %
Carbon %
Hydrogen %
Nitrogen %
Sulfur %
Chlorine %
Oxygen %
VALUE
12,027
9.59
7.40
66.98
4.65
1.48
3.32
0.06
6.52
                                        10-31

-------
                                    TABLE 9

             ARGONNE DRY SCRUBBER SYSTEM CHARACTERISTICS
           SPRAY DRYER ABSORBER

                   Niro Atomizer 2^.5 ft diameter spray dryer
                   with compound gas disperser

                   35% feed slurry

                   20°F approach to saturation in exit gas

                   80 % SC>2 removal with lime stoichiometries of 1.0
           BAGHOUSE
                   JOY four compartment pulse flow baghouse

                   12 ft long woven fiber glass bags

                   Air/cloth ratio 3:1

                   A P: 3 inch W.G.
      Argonne National Laboratories permitted Niro Atomizer to use their dry scrubbing
system for a two week demonstration of the new SOX/NOX process. The demonstration
was  accomplished  in April  1983  just before the  boiler was brought down for its annual
overhaul.

      For  the demonstration period Niro Atomizer  brought in a continuous NOx-monitor
consecutively  taking samples from spray dryer  inlet,  spray dryer outlet  and baghouse
outlet. During the  tests back-up wet sampling of  NOX and SOX were was Additive sodium
hydroxide solution  was stored in  an  intermediate 4000  gallon tank. A  small centrifugal
pump, a  flow  meter and  some  piping were  all  what were  required  to provide  the
capability of a controlled addition of sodium hydroxide to the recycle mixing tank.

      Some mechanical problems  caused  by the handling characteristics of the powder
during the test period  showed  the minor changes that  have to  be done on full-scale
installations operating in SOX/NOX mode.

      The  test period  was commenced with a test under normal  operating conditions for
the dry scrubber. No  sodium  hydroxide  was added. This  test demonstrated that  under
normal dry scrubber operating conditions virtually no NOX  is removed in the system.

      Then the system was  brought into  the SOX/NOX scrubbing mode  by increasing the
spray dryer outlet  temperature  from l^op  to  195OF and starting  sodium hydroxide
addition to the recycle mixing tank. The result was very impressive and  encouraging in as
                                      10-32

-------
much as substantial NOX removal was observed. Table 10 shows the overall performance
with and without NOX removal. In the SOX/NOX operating mode more  than  50% NOX
removal was  demonstrated  while the  scrubber  was  still  removing 90%+  SC^.  The
demonstration shows  full agreement  between  pilot  plant  data and full-scale scrubber
performance.

                                    TABLE  10

            DRY SCRUBBER SOX/NOX PERFORMANCE IN FULL-SCALE
MODE
OF
OPERATION
NORMAL 5O2
REMOVAL ONLY
50X/NOX
MODE
SCRUBBER
INLET
PPM
S02
1800
1500
PPM
NOX
290
280
TEMP.
°F
335
330
GAS-
FLOW
Ib/h
200,000
1*0,000
OUTLET
TEMP.
°F
1*8
195
BACFILTER OUTLET
PPM
S02
90
60
PPM
NOX
281
126
RESULTS
% REMOVAL
SOX
95
95+
NOX
3
55
STOICHIOMETRY
FOR
LIME
l.lf
1.3
RIVERSIDE DEMONSTRATION

     A 3 week test period is planned at JOY/NIRO's Riverside demonstration plant this
fall. These tests will furnish additional full-scale  data for process optimization. Results
will be published as soon as available.

       CHARACTERIZATION OF SOX/NOX DRY SCRUBBER WASTE MATERIAL

     The waste product from dry scrubbing systems has been  thoroughly analyzed and
characterized over the last 3-4 years (14, 15,  16, 17, 18  and 19). These studies have
shown that the dry scrubber waste product is an excellent material for landfill due to its
pozzolanic properties and low permeability. The  leachate quality can meet EPA limits
and therefore no liners would be required in properly designed landfills.

     The waste product  from the SOX/NOX process has the same appearance as other
dry scrubber poducts, i.e. a fine-grained, dry free-flowing powder.

     The chemical composition  is of course  somewhat different. Table 11 shows the
chemical analyses of the waste product from  the Argonne tests with and without NOX
removal. The Argonne product has low ash content as the boiler is Stoker fired and multi
clones  are installed up-stream of the scrubber.
                                       10-33

-------
                                    TABLE 11
              CHEMICAL COMPOSITION (wt%) OF WASTE PRODUCT
                       WITHOUT AND WITH NOX REMOVAL

Sulfite as CaSO3 • ^H2O
Sulfate as CaSO^ -2H2O
Nitrate as Ca(NO3)2
Excess alkalinity
as Ca(OH)2
Ash & Inerts
Moisture
Na2SO3
SOX
REMOVAL
59
16
-0
15
9
1
— 0
SOX/NOX
REMOVAL
16
43
3
21
8
1
9
     As can be seen from the analyses, a major part of the sulfite has been oxidized to
sulfate when  the  scrubber operates in SOX/NOX mode. The absorbed NOX is present in
the product as nitrate. No nitrite has been found.

     To evaluate  the suitability of the waste product for landfill the following important
parameters were investigated:

          Density

          Compressive strength

          Permeability

          Leachate quality

     Figure 6 shows the  moisture-density curve for  the two products.  The SOX/NOX
product has a higher dry density and about the  same  optimum moisture content as the
normal dry FGD product.
                                       10-34

-------
                  kg/m1 Dry Density
                1000--
                900-'
                800
                         40
                                                      SOX/NOX

                                                      Removal
                                   50
                                                   PureSOx

                                                   Removal
                                                      -I-
                                            60
                                                      70 % Water Added
                Figure 6. Moisture-density data of waste product from pure SOX and
                         combined SOx-NOx-removal

     Table  12 shows  a  comparison of compressive  strengths and  permeabilities. The
strength development is the same as for the normal product, and the permeability is very
low for both products.

                                    TABLE 12

                COMPRESSIVE STRENGTH AND PERMEABILITY OF
                PURE SOX AND COMBINED SOX/NOX PRODUCTS *
COMPRESSIVE STRENGTH (kPa)

1 DAY
7 DAYS
30 DAYS
56 DAYS
PERMEABILITY (ICT6 cm/s)
1-2 DAYS
28 - 30 DAYS
PURE SO2
PRODUCT
120
180
480
600

2.5

SOX/NOX
PRODUCT
110
200
340
580

6
-
        All samples are compacted to 960-980 kg/m^ in dry density
                                       10-35

-------
     The ieachate quality obtained  by the EPA Toxic Extraction Procedure is shown in
Table 13. All of the Ieachate values  shown are less than  100 times the Primary Drinking
Water Standards (PDWS), therefore  the  product would most likely be classified as non-
hazardous.

                                   TABLE 13

                   LEACHATE QUALITY OF WASTE PRODUCTS
              FROM PURE SOX AND COMBINED SOX/NOX REMOVAL*


ARSENIC
BARIUM
CADMIUM
CHROMIUM
LEAD
MERCURY
NITRATE (as N)
SELENIUM
SILVER
EPA-
PDWR
mg/1
0.05
1.0
0.010
0.05
0.05
0.002
10
0.01
0.05
PURE S02
REMOVAL
mg/1
0.001
< 0.5
< 0.005
0.010
< 0.05
< 0.002
2
0.004
< 0.01
SOX/NOX
REMOVAL
mg/1
0.001
< 0.5
< 0.005
0.012
< 0.05
< 0.002
260
0.004
< 0.01
             test carried out according to EPA toxic extraction procedure
      The waste product from the Argonne tests  contains little ash.  The dry scrubber
 product from a typical high sulfur application would have a fly ash content of 50  60 96.
 This would change the characteristics in a positive direction.

      It is expected that the SOX/NOX waste product will be  regulated as non-hazardous.
 However, due to somewhat elevated values for nitrate and total dissolved solids (TDS), it
 may be necessary to install a liner before disposal.

      Currently Niro Atomizer is working on methods to improve Ieachate quality and to
 reduce Ieachate quantity.

   COMPARISON OF ECONOMICS WITH STATE OF THE ART SOX/NOX PROCESSES

      A cost comparison has been made between the Niro Atomizer dry SOX/NOX process
 and  a conventional limestone  wet scrubber / selective  catalytic de-NOx (SCR) system.
 The  reason for using SCR for comparison and not the cheaper selective non-catalytic de-
 NOX  system (SNCR) is  that the latter is  able to achieve only  30 -  50% NOX removal.
 Furthermore SNCR has not been demonstrated  on high  sulfur coal. Excessive ammonia
                                       10-36

-------
 carryover is likely to result in severe air preheater plugging problems as well as problems
 with  waste  characteristics. This excludes the  use  of SNCR for high sulfur applications
 (22, 23).                                                                FF

      The basis for the cost comparison is a 500 MW power plant burning 3% sulfur coal
 and emitting 300 ppm NOX. A 65% NOX removal and a 90% SO2 removal are assumed for
 both  systems. The operating  cost  for the SCR system  is critically  dependent  on the
 catalyst lifetime. A two year lifetime has been assumed for this cost comparison.

      The capital requirement and  the operating/maintenance cost for the two systems
 are shown in Table 14. It is seen that the capital requirement is considerably less for the
 Niro  Atomizer dry SOX/NOX system.  Operating and maintenance costs are very similar
 for the two  systems.
                                    TABLE 1*

                 OPERATING AND CAPITAL COST ESTIMATE FOR
                  COMBINED SOX AND NOX REMOVAL SYSTEMS *
COST ITEM
CAPITAL REQUIREMENT
OPERATING AND MAINTENANCE
COSTS (1st Year) (O&M)
LEVELIZED CAPITAL REQUIREMENT
LEVELIZED O&M
TOTAL LEVELIZED COST
WET LIMESTONE
+
SCR
105 M$ FGD
30 M$ SCR
135 M$
17.2 M$ FGD
4.6 M$ SCR
21.8 M$
7.4 mills/kWh
12.7 mills/kWh
20.1 mills/kWh
NIRO
DRY
SOX/NOX
85 M$
19.5 M$
4.8 mills/kWh
11.3 mills/kWh
16.1 mills/kWh
          *  References:  (3,  20, 21 and 22)


     A  single  annual  levelized cost  value  has  been computed based on  a 30 year
economic  plant life  to  represent  the revenue  requirement  for  the  operating  and
maintenance (O&M) cost  and  the capital requirement  . For the O&M cost a levelization
factor of  1.9 is used. For the capital requirement a levelization factor of  0.18 is used.
The total  levelized cost  for the Niro Atomizer dry SOX/NOX system is 16.1  mills/kWh
compared  to 20.1 mills/kWh for the limestone wet scrubber/SCR system.
                                      10-37

-------
     Besides  having  favourable   economics  when  compared  to  a  limestone  wet
scrubber/SCR  system  the Niro  Atomizer dry  SOX/NOX  process has  the additional
advantage of being one integral air pollution control system. It  will be easier to operate
and will not interfere with the  boiler operation since no equipment is installed upstream
the air preheaters.
                                  CONCLUSION

 •    Dry scrubbing  has gained much acceptance within the utility industry due to its
      low cost and ease of operation.

 •    Dry scrubbing  with simultaneous  remoVal of SOX and  NOX  is a  new  process -
      developed by Niro Atomizer - which  uses the same equipment as provided for dry
      scrubbers presently in operation or under construction.

 «    The process is  well suited for high sulfur coal applications where low NOX burners
      have reduced NOX to the 0.25 - 0.35 Ib/MBtu level. The process can further reduce
      NOX emissions  down to the 0.1 Ib/MBtu level together with 90 - 95% SOX removal.

 •    Process performance has been tested in pilot-scale and has been confirmed in full-
      scale at Argonne National Laboratories.

 •    Waste products from the process will most likely be classified as non-hazardous.

 •    Economics of the new process compare favourably with limestone wet  scrubbing
      plus selective catalytic reduction (SCR).
                                       10-38

-------
                                  REFERENCES


 (1)   Inside E.P.A. Weekly Report, by US EPA, Washington.

 (2)   Drehmel, D.C. et al
      "Low NOX Combution Systems with SC>2 Control Using Limestone"
      Industrial Environmental Research Laboratory, US EPA - paper no. 83-38.7-

 (3)   Slack, A. V.
      "Technology for Power Plant Emission Control"
      Report prepared for Niro Atomizer, June 30, 1983.

 (!+)   Swedish study by the Ministry of Agriculture
      "Acidification Today and Tomorrow"
      Prepared for the June  1982 Stockholm Conference  on  The Acidification of  the
      Environment.

 (5)   EEC Symposium
      "Acid Deposition - A Challenge for Europe"
      Karlsruhe, Germany - September 19-21, 1983

 (6)   1982 Joint Symposium on Stationary Combustion NOX Control
      Dallas, Texas, November 1982.

 (7)   EPA Environmental News.
      The October 22,  1982 issue.

 (8)   Felsvang, K.S. et al
      "Status  of  the  JOY/NIRO  Dry FGD System  and Its  Future  Application for  the
      Removal of High Sulfur, High Chloride and NOX from Flue Gases"
      1983 Joint Power Generation Conference, Indianapolis, Indiana.
      September  25-29, 1983.

 (9)   Donnelly, J.R. et al
      "Process for Removal of Nitrogen Oxides and Sulfur Oxides from Waste Gases"
      U.S. Patent Application S.N. 382,968

(10)   Felsvang, K.S.
      "Results from Operation of the Riverside Dry Scrubber"
      JOY/NIRO  Seminar, Minneapolis, Minnesota - June 1981.

(11)   Klingspor, J.
      "Kinetics and Engineering Aspects on the Wet-Dry FGD Process"
      Department of Chemical Engineering II, Lund's Institute of Technology, Sweden.
      June 1983

(12)   Farber,  P.S.
      "Start-Up and Performance of a High Sulfur Dry Scrubber System"
      APCA 75th Annual Meeting, New Orleans, LA
      June 1982
                                        10-39

-------
(13)   Farber, P.S.
      "The Argonne High Sulfur Dry Scrubber"
      JOY/NIRO Seminar on Dry Scrubbing of High Sulfur Coals, Minneapolis, Minnesota,
      June 25-27, 1982

(14)   Radian Corporation
      "Characteristics of Waste Products from Dry Scrubbing Systems"
      EPRI Report CS-2766, 1982

(15)   Thompson, Carol M.
      "Characteristics of Waste Products from Dry Scrubbing Systems"
      EPA/EPRI FGD-Symposium, Hollywood, Florida
      May 1982

(16)   Donnelly, J.R.
      "Disposal and Utilization of Spray Dryer FGD End Products"
      Canadian Electrical Association Seminar, Ottowa, Canada
      October  1981

(17)   Buschman, J.C. et al
      "Disposal of Wastes from Dry  SC>2 Removal Processes"
      Joint Power Generating Conference, Phoenix, Arizona
      September, 1980

18)   Donnelly, J.R. et al
      "Dry Flue gas  Desulfurization End-Product  Disposal  Riverside  Demonstration
      Facility Experience"
      EPA/EPRI FGD-Symposium, Hollywood, Florida
      May, 1982

(19)   Farber, P.S.
      "Leachate of Dry Scrubber Wastes"
      70th APCA Meeting,  Atlanta,  Georgia,
      June 19-24, 1983

(20)   Stearns-Roger Eng. Corp. and Radian Corporation
      "Technical and Economic Feasibility of Ammonia-Based Postcombustion
      NOX Control"
      EPRI Report CS-2713.
      November, 1982

(21)   "Preliminary Economic Analysis of NOX Flue Gas Treatment Processes"
      EPRI Report FP-1253,
      February, 1980

(22)   "NOX Reduction Alternatives  Study"
      Belridge  Field Cogeneration Project. Shell California Production Inc.
      Submitted to the California Energy Commission, July 1983

(23)   Jumpei, Ando
      "NOX Abatement for  Stationary Sources in Japan"
      EPA-600/7-83-027
      May, 1983
                                       10-40

-------
PROCESS CHARACTERIZATION OF S02 REMOVAL IN
     SPRAY ABSORBER/BAGHOUSE SYSTEMS

  E.A. Samuel, T. W. Lugar, D. E. Lapp,
 K. R. Murphy, 0. F. Fortune, T. G. Brna,
                R. L. Ostop

-------
                                 ABSTRACT
A new equation for  correlating  the  SC>2  removal  with stoichiometric ratio,
approach to saturation,  inlet S02 concentration,  inlet temperature,  and
inlet moisture content  is  proposed.   It reveals essential information about
the kinetics of the  SC>2  removal process.   The chemical reaction between the
S02 and lime appears  to  limit the rate  of  S02 removal in the spray
absorber.  The effect of recycle enhancement  of SC>2 removal may be
understood in terms  of  this  model.   The model also permits a classification
of flyash suitable  for  designing dry flue  gas desulfurization (DFGD)
systems.

DFGD pilot plant  results are reviewed from theoretical and mechanistic
viewpoints.  The  SC>2  removal as a function of the gas flow rate (or  average
residence time) and  the  angle of the secondary  swirl vanes in the absorber
gas disperser is  discussed relative  to  the velocity flow field in the
absorber.  The SC>2  removal efficiency is also discussed as a function of
the atomizer angular  speed.

When the gas flow rate  and approach  to  saturation are fixed, the S02
removal efficiency  in the  spray absorber increases with increasing
stoichiometric ratio  and inlet  temperature  and decreasing inlet moisture
content and S02 concentration.   An  explanation of these effects is given
by their impact on  the  droplet  diameter integrated over the lifetime  of
the droplets.  Another  effect that  increases  the time-integrated droplet
diameter is the decrease in  the volume  of  the core of solids (lime)  for
smaller slurry droplets.

Partial recycle of  the  spent products,  in the absence of externally
injected flyash,  results in  S02 removal efficiency being nearly independent
of S02  concentration.   Thus, the enhancement  from recycle is greater  for
the higher S02 concentrations.   S02  removal performance with recycle  was
improved by the injection  of flyash  derived from a low sulfur Texas  lignite
and from high sulfur  eastern coals.   The enhancement with recycle in the
latter  case is dramatic  and  supports DFGD applicability to high sulfur
coal.

                                NOMENCLATURE


              £   =   502 removal efficiency across spray absorber

              a   =   Ys  Cp/(NNuK)

              Ys  =  Product deposition reaction rate constant

              Cp  =   Flue  gas specific  heat at constant pressure

              N   =   Nusselt number  for liquid droplets in gas stream
               Nu
              K   =   Thermal conductivity of  flue gas
                                     10-41

-------
              Xc   =   Correction factor  for  effect of the solids fraction on
                     the  S02  removal  rate

                  =   i  -  Djj  /D|

              D-L   =   Initial  mean droplet diameter

              Df   =   Final mean particle  diameter of evaporated droplet

              S    =   Stoichiometric ratio

              Tsi =   Spray absorber inlet temperature

              XSf =   Spray absorber outlet  temperature

              Tas =   Adiabatic saturation temperature of flue gas
                     entering spray absorber

              A   =   Time-integrated  droplet diameter
                         T
                  =   r  / D(t)dt
                       0
              r   =   Concentration of droplets

              T   =   Droplet evaporation  time

            D(t)   =   Droplet diameter at  time, t

              u   =   Empirical constant characteristic of a given flyash

              F   =   Mass fraction of solids in slurry droplet or atomizer
                     slurry  flow
                     Reduced penetration
                  =  - £n(l - ec) / {X0 S
                                              Tsi ~ Tas
                                      c
                                               sf    as
                               INTRODUCTION

The New Source Performance Standards (NSPS) for S02 require cleanup of
flue gases from utility boilers burning even low-sulfur coals.  The
parallel increase in application of fabric filters to control particulate
emissions from coal-fired boilers has generated considerable technical  and
commercial interest in dry flue gas desulfurization (DFGD) processes.   The
most commercially promising DFGD process is spray absorption.   In  this
process, an alkali sorbent liquid is atomized in a spray dryer  using  flue
                                    10-42

-------
gas as the drying medium.  The DFGD reactions occur between the gas and
droplets in intimate contact during the various stages of drying.

Eventually the reaction products are reduced to a suspended solid par-
ticulate in the gas stream, some solids dropping out in the absorber
hopper.  When a fabric filter is used for removal of remaining spent sor-
bent and flyash from the gas stream, further DFGD reactions occur in the
filter cake by virtue of unreacted alkali and residual moisture.  While a
variety of calcium- and sodium-based alkalis have been evaluated, lime
currently has the greatest commercial importance.

In 1978, the Buell Emission Control Division of the Envirotech Corporation
(now part of General Electric Environmental Services,  Inc.)  and Anhydro
A/S constructed a spray dryer FGD prototype facility at the Martin Drake
Station of the City of Colorado Springs.  A 4.01-m3/s (8,500-cfm) spray
absorber was installed in slipstream configuration in the inlet duct to a
189 m3/s (400,000 cfra) fabric filter serving an 85 MW pulverized-coal-fired
steam generator burning northwestern Colorado coal.  An extensive program
of parametric performance and process demonstration tests was undertaken at
that facility jointly with EPA and the City of Colorado Springs.  As a
result improved models for spray dryer FGD performance were developed.
This paper presents these models and experimental data which support their
validity.
                                  THEORY

Two models for predicting the S02 removal in the spray absorber/baghouse
system are presented in this paper.  The first is an analytical model
restricted to straight-through operating conditions only.  The second is an
empirical model applicable under wet recycle of solids collected from the
baghouse and/or the spray absorber hoppers.  The analytical model provides
a theoretical basis for understanding the fundamental processes which
remove S02 by spray absorption.  The empirical model extends the usefulness
of the analytical model by including the effect of flyash on S02 removal by
spray absorption.

Analytical Model for Straight-Through S02 Removal (No Recycle)

An analytical expression for the S02 removal rate (or efficiency) in a
spray absorber may be derived for the following idealized conditions within
the spray absorber.

1)  The gas velocity is uniform within the spray absorber (the presence of
    vortices, eddies, and recirculation zones is ignored).

2)  The droplet/gas mixing is uniform, with the droplets following gas
    streamlines.

3)  Most S02 removal takes place during the period of water evaporation
    from the droplets (often called Phase I), when surfaces are maintained
    at the adiabatic saturation temperature.
                                    10-43

-------
4)  The rate of S02 removal is determined by the rate of  the  chemical  reac-
    tion between the 862 and lime to form sulfites or sulfates.

Under these assumptions, the simultaneous occurrence of  S02 removal  and
water evaporation is controlled by the rate of S02 removal, the  rate of
droplet size changes, and the conservation of thermal energy  in  the  spray
absorber (1,2).  Consistent with these relationships the  SC>2  removal effi-
ciency, under straight-through conditions, can be expressed by Equation  1.

                                            T • - T
                                            1si   1as
                  - £n(l - £s) = a Xc S  [in	j-^— ]                 (1)
                                             sf    as

The correction factor, XC) arises from the proportionality between the
penetration [i.e., - £n(l - es)] and the time-integrated  droplet diameter.
For pure water droplets, the final diameter, Df, = 0.  When solids are pre-
sent in the slurry droplet, its initial diameter is increased due to volume
displacement by the solids, and its final diameter, at the end of evapora-
tion, is non-zero due to the volume of the solids.  As the solids fraction
increases, the departure of the correction factor, Xc, from unity becomes
progressively more significant.  The correction factor, Xc, depends  on both
the mean diameter, D^, of droplets produced in the atomizer and  the  final
mean diameter, Df( of particles produced at the end of evaporation.  Its
accurate evaluation is complicated by the dependence of  the initial  droplet
diameter on the solids fraction in the slurry being atomized  and by the
dependence of the final particle diameter on the extent to which the dried
particle retains some moisture.  A reasonable expectation is  for Xc  ^  1  for
solids fractions <0.1.  At solids fraction of 0.5, the correction factor,
Xc, appears to be about 0.5 to 0.7.  Table 1 discusses the dependence  of
the model prediction for the spray absorber SC>2 removal efficiency on
selected spray absorber and flue gas parameters.

Empirical Model for SC>2 Removal with Wet Recycle

The S02 removal efficiency in the presence of wet recycle of  offproduct  is
invariably better than that under straight-through conditions.   In addition
to unspent lime, the recycled offproduct contains flyash.  The flyash
appears to have a synergistic effect on the utilization of lime  from both
fresh and recycled sources.  Different types of flyash in the recycle
slurry provide different levels of enhancement of the SC>2 removal.  This
recycle enhancement appears to be related to the coal from which the flyash
originates  and on flyash characteristics such as pH of  hydrolysis,  par-
ticle size, and chemical composition.  Since the flyash  effect on
SC>2 removal through wet recycle is not well understood,  a fundamental  rela-
tionship for the efficiency in the presence of recycle cannot be derived.
The following empirical relationship is suggested, based  on Equation 1,
to be consistent with the observation of improved S02 removal in the pre-
sence of recycle which increases the solids fraction, F  , in  a newly pro-
duced slurry droplet (or in the atomizer feed):
                                    10-44

-------
                                                          TABLE 1.
                                                                    MDDEL PREDICTIONS  FOR OPERATIONAL  EFFECTS  ON
                                                                    STRAIGHT-THROUGH REMOVAL IN SPRAY  ABSORBER
                               Parameter
                                                      How Parameter Enters
                                                      Equation 1
                                   Relationship to
                                   S02 Removal
                                                                                                                      Comment
                    Absorber Residence Time
                    Atomizer Angular Speed
                    Gas Inlet Temperature
O
 I
-P-
Ui
                                                    Does not enter equation
                                                    explicitly.
Does not enter equation
explicitly.
Enters equation explicitly
as Tsl.
                    Gas Inlet Moisture Content
                                                    Enters equation indirectly
                                                    through the adiabatic sat-
                                                    uration temperature,  Tas.
                               None,  provided residence
                               time is long enough to
                               allow complete evaporation.
None, provided droplets are
of such size that they evap-
orate completely.
S02 removal efficiency
increases with increasing
inlet temperature for
fixed approach to
saturation temperature.
                               S02 removal efficiency
                               decreases with increasing
                               inlet moisture content
                               for fixed inlet gas temp-
                               erature and approach to
                               saturation.
Water flow rate and the inlet and
outlet gas temperatures are held
at conditions which give complete
evaporation  and maintain Xc - 1.

For a fixed atomizer water flow
rate, the time-integrated
droplet diameter is independent of
the initial droplet diameter,
provided complete evaporation
occurs  and X  - 1.

The water required to reach the
same approach to saturation
temperature increases with
increasing inlet gas temperatures.
The time—integrated droplet
diameter increases with increasing
inlet gas temperature.

For the same inlet gas temp-
erature, the adiabatic satura-
tion temperature increases with
increasing inlet moisture content.
Consequently, the water required
to reach the same approach to
saturation at fixed inlet gas
temperature decreases with
increasing inlet moisture content.
                    Inlet S02 Concentration
Enters equation indirectly
through the correction
factor, Xc.
S02 removal  efficiency is
independent of inlet S02
concentration for low con-
centrations «1000 ppm).
S02 removal efficiency
decreases with inlet S02
concentration at high con-
centrations O1000 ppm) at
fixed stoichiometry.
Water flow rate and the inlet and
outlet gas temperatures are held
at conditions which give complete
evaporation.   The droplet evapora-
tion time decreases with increasing
inlet S02 concentration.  For in-
let S02 concentrations less than
about 1000 ppm, the lime rates are
low such that  Xc = 1.  At
higher inlet S02 concentrations,
Xc decreases with increasing
concentration.

-------
or
                                  T •  - T
                                   si     as
                 - es)  = Xc S [£n	—r~ ]  (u F + a)                 (2)
                                   sf     as

                     P  = u F + a                                        (3)
where u is an empirical constant which varies with flyash type.  The
constant, a, comes from Equation 1.  Implications of this empirical equation
will be discussed later.

                                EXPERIMENT

TEST OBJECTIVES AND EQUIPMENT


The overall objective of the dry FGD (DFGD) pilot program at Colorado
Springs was to demonstrate and characterize S02 removal in a spray
absorber/baghouse system for low sulfur and high sulfur applications.
Although lime was the main reagent tested, other reagents,  such  as  trona
and dolomitic lime,  which may be  preferable for certain power plant sites,
were also evaluated.  Limestone was also tested because its effective use
in a DFGD system will dramatically shift FGD economic evaluation in its
favor.  Other secondary objectives of the Colorado Springs program were to
study the effects on S02 removal of baghouse reheat, adipic acid addition
to lime, and the use of wastewater for lime slaking.  Details of the tests
conducted in the pilot program are reported elsewhere (3,4,5,6).  Only
those  tests which are relevant to the proposed models are discussed in
this paper.

The primary elements of the Colorado Springs prototype plant are a 3.8-m
(12.5-ft) diameter spray absorber equipped with a top-entry vaned scroll-
type gas disperser; a centrifugal atomizer with a 0.25-m (10-in.) diameter
wheel; a full-scale commercial-type lime storage, handling, and slaking
system; and a reverse-air baghouse fitted with full-size (0.30-m or 12-in.
diameter, 9.14-m or 30-ft length) fiberglass filter bags.  The plant is
also equipped with a four-element cyclone collector in parallel with the
baghouse at the absorber outlet to provide the capability to independently
vary the absorber flow volume and baghouse air/cloth ratio.  An S02  vapor-
izing and injection system enabled increasing the S02 concentration at the
spray dryer inlet from 200 to 400 ppm (the value typical of the Colorado
coal) to values in the range from 1,000 to 4,000 ppm.  The higher values in
this range are characteristic of the flue gas derived from burning high
sulfur eastern coals.  Flue gas characteristic of a given coal other than
the Colorado coal burned at the Martin Drake Station was simulated by the
external injection of S02 as needed and flyash derived from a boiler firing
the coal type of interest.

Major measurement systems included venturi meters for gas flow measurement
at the absorber and baghouse inlets; an extractive S02 monitoring system
(DuPont Model 460) with probes at the absorber inlet, at the absorber
                                   10-46

-------
outlet, and at the baghouse outlet; a dew point hygrometer with a probe at
the baghouse outlet; magnetic flow meters for the measurement of slurry
flows; and extensive pressure and temperature instrumentation.  All equip-
ment and materials required for instrument calibration were available on-
site, as were chemical laboratory facilities for measurement of slurry
reactant concentration, slurry total solids by evaporation, and offproduct
moisture content.

The Colorado Springs prototype plant was sized to accommodate a wide range
of operating parameters.  Following the plant start-up, shakedown, and ini-
tial optimization tests, operation was established within the range of
parameters summarized in Table 2.

                                 RESULTS

OPTIMIZATION TESTS

Average Gas Residence Time in Spray Absorber

Figure l(a) summarizes the results of a series of tests to determine the
variation in the S02 removal efficiency across the spray absorber,
baghouse, and the spray absorber/baghouse system, due to variation in the
residence time (from 7 to 12 seconds) for fixed atomizer angular speed
(12,500 rpm), stoichiometric ratio (1.0), and approach to saturation
(17°C, 30°F).  The spray absorber S02 removal efficiency increases slowly
with decreasing residence time (increasing flow rate).  On the basis of the
analytical model, a constant S02 removal efficiency with respect to resi-
dence time would be expected.  The increase shown by test results appears
to be due to improved gas flow distribution and gas/droplet mixing with
increasing gas flow rate.

The S02 removal in the baghouse was more sensitive to the average gas resi-
dence time in the spray absorber than was S02 removal in the spray absorber
itself.  The baghouse removal shows a maximum at a residence time of about
10 seconds.  The increase in the baghouse efficiency when the residence
time increases from 7 to 10 seconds (decreasing gas flow rate) seems due
to the decreased cyclonic action in the spray absorber and the consequent
increase in the mass loading of the spent product into the baghouse.  The
decreasing baghouse performance with increasing spray absorber residence
times beyond 10 seconds seems to be due to poorer gas flow velocity distri-
bution and the accompanying reduced gas/particle mixing in the spray
absorber.  This baghouse behavior is also consistent with the improved
cyclonic activity and the accompanying decrease in the mass loading to the
baghouse when the angular momentum of the slurry becomes progressively
greater at the low gas flow rates corresponding to residence times greater
than 10 seconds.

Atomizer Wheel Angular Speed

Figure l(b) summarizes the observed variation in the spray absorber,  baghouse,
and spray absorber/baghouse system S02 removal efficiencies with atomizer
angular speed.  The gas flow rate was fixed at 4.01 rn^/s (8500 cfm)
                                    10-47

-------
                                   TABLE  2.   COLORADO  SPRINGS  SPRAY ABSORPTION PROTOTYPE PLANT SUMMARY
                                             OF  OPERATING  PARAMETERS
                        Parameter3
                                       Typical
                                                                 Value
Minimum
Maximum
                               Maximum
                               Minimum
            Variation
            during test,
            percent
o
i
00
ABSORBER

Inlet  flow volume,  cfm"
Inlet  temperature,  °F
Inlet  S02 concentration, ppm
Inlet  saturation temperature,
Outlet temperature, °F
Tp, approach temperature, °F^

BAGHOUSE

Inlet  flow volume,  cfm

ATOMIZER

Wheel  angular speed, rpm

SLURRY SYSTEM
                                                    8,500
                                                     350
                                                    1,000
                                                     120
                                                     140
                                                      20
                                                   3,000
                                                   12,500
  5,000
    250
    250
    110
    122
     12
  1 ,000
  6,300
 10,600
    400
  4,000
    140
    180
     40
   .,000
 14,000
 2.1
 1.6
16.0
 1.3
 1.5
 3.3
 6.0
 2.2
Lime slurry flow, gpme
Total atomizer feed, gpm
Lime slurry total solids, percent
a Units shown apply only to columns
b Multiply cfm value by 4.72 x 10 ~L
c T. °n. = 5fT. °F - 321/9.
1.0 0.5 3.0 6.0
2.0 1.0 6.0 6.0
15 8 50 6.3
under "Value.1
h to convert to m-Vs.
2
2
2

             Tp, °C =  5(Tp,   ..
             Multiply gpm value by 6.3 x 10 5 to convert to mj/s.

-------
                              - SPRflY flBSORBER
                              - BflGHOUSE
                              - SPRftY fiBSORBER/BflGHOUSE
                               SYSTEM
      -\	1	1	1	1	p
        10    11    12
BVERflGE RESIDENCE TIME, SECONDS
O
I
            1(a).  Observed  dependence of
                   straight-through 862
                   removal efficiency on
                   average gas  residence
                   time in the  spray absor-
                   ber.  A residence time
                   of 10 seconds appears
                   to optimize  the system
                   SC>2 removal  efficiency.
                                                                    - SPRflY fiBSOSEER
                                                                   o - BflGHOUSE
                                                                    - SPROT flBSORBER.BflGHOUSE
                                                                     SYSTEH
                                                       - SPRflY flBSORBER
                                                       - BfiGHOUSE
                                                       - SPRflY flBSORBER/BflGHOUSE
                                                        SYSTEM
                                                             9000    11000    13000

                                                             RTOMIZER flNGULflR SPEED, RPfl
l(b). Observed dependence  of
      straight-through  S02
      removal efficiency on
      the  atomizer angular
      speed.   The system S02
      removal efficiency
      appears to be optimized
      for  angular speeds _>_
      10,500  rpm.        ~
                                                                 Kc)
                                                   IB   20    30
                                                   VflNE flNGLE, DEGREES
Observed dependence  of
straight-through  S02
removal  efficiency on
the inlet gas disperser
vane  angle (with  the
vertical).   A vane
angle of 15 to 20°
appears  to optimize  the
system S02 removal
efficiency.
            Figure 1.  Optimization test  results for  pilot lime spray absorber/baghouse system  without recycle
                        of  offproduct.   [For these tests,  the following parameters  were fixed: inlet S02
                        concentration  (1000 ppm), stoichiometric ratio (1.0), inlet gas temperature (177°C,
                        350°F),  and approach to saturation (17°C,  30°F).]

-------
(10-second average residence time), the stoichiometric ratio at 0.95,
approach to saturation at 17°C (30°F) ,  and the inlet temperature at 111°C
(350°F) for these tests.   S02 removal in the spray absorber was relatively
independent of wheel speed in the range tested and, consequently, indepen-
dent of droplet diameter (6,7,8).  This independence is consistent with the
theoretical expression summarized in Equation 1 for the case of low solids
concentration in the atomizer feed.  In this case, Df - 0 and the correc-
tion factor, Xc,~ 1.  Hence the time-integrated diameter of the droplets to
which the S02 removal is proportional is independent of initial droplet
diameter.

Figure l(b) also shows a stepwise increase in the baghouse SC>2 removal effi-
ciency at an atomizer angular speed of  about 10,500 rpm.  The system
S02 removal efficiency reflects the sensitivity of the baghouse S02 removal
efficiency to variations in atomizer angular speed  and appears to be opti-
mized at atomizer angular speeds above  10,500 rpm.

Atomizer Disperser Vane Angle

By tilting the secondary swirl vanes of the gas disperser from the ver-
tical (i.e., parallel to the atomizer wheel axis), it is possible to impart
a tangential component to the gas velocity as the gas enters the spray
chamber and thereby induce a swirl in the gas velocity flow field.  An ini-
tial swirl in the gas velocity flow field is advantageous in extending the
gas residence time by inducing a helical motion to the gas molecules in the
spray absorber and in improving gas/particle mixing by providing a centri-
petal acceleration to the particles.  Figure l(c) summarizes the observed
S02 removal efficiencies with lime as a function of the vane angle
(measured from the vertical) for fixed  stoichiometric ratio (0.92), inlet
gas temperature (177°C, 350°F), approach to saturation (17°C, 30°F) , and
atomizer angular speed (12,500 rpm).

The vane angle appears to improve the SC>2 removal in the spray absorber
noticeably in the range from 0 to 20°.   Increasing the vane angle beyond
20° does not lead to noticeable improvement in the performance of the spray
absorber.  This suggests that a higher  level of turbulence at the higher
vane angle does not allow the swirl to  be completely characterized by the
initial angle of projection and that, as the flow progresses into the
chamber, a spectrum of swirl angles is  produced.  In this way, the tur-
bulence in the flow field removes any identity of the flow with the initial
swirl angle for vane angles above 20° in the range from 0 to 45°.  The
contribution of the angular momentum of the slurry to the swirl of the gas
will also tend to lessen the effect of  increasing the vane angle.

The baghouse SC>2 removal efficiency appears to be independent of the vane
angle.  This suggests that the size distribution of particles produced at
the end of evaporation of the slurry droplets created in the rotary atom-
izer are in such a range that their carryover to the baghouse is unaf-
fected by changes in vane angle.
                                    10-50

-------
Spray Absorber Inlet Temperature at Fixed Inlet Flue Gas Moisture Content

The effect of varying the temperature of the flue gas at the spray absorber
inlet without changing its moisture content was studied in a series of
tests.  In these optimization tests, the atomizer disperser vane angle was
set at 15° and the atomizer wheel speed was maintained at 12,500 rpm.   As
in earlier tests, the gas flow rate was 4.01 m-Vs (8,500 cfm) and
corresponded to an average residence time of 10 seconds in the spray absorber.
The spray absorber inlet temperature was varied by using a surface heat
exchanger with water as the cooling fluid upstream of the spray absorber
inlet.

Figures 2(a) and 2(b) display the experimental results for the S02 removal
efficiencies under straight-through conditions for approach temperatures of
11°C (20°F) and 17°C (30°F).  The spray absorber S02 removal efficiency
increased with increasing inlet temperature, while the baghouse efficiency
displayed the opposite trend.  Thus the system SC>2 removal efficiency
increased only weakly with increasing inlet temperature.  The rate of
increase of the spray absorber efficiency and the rate of decrease of  the
baghouse efficiency with increasing inlet temperature at a 17°C (30°F)
approach to saturation are more than double the rates at an 11°C (20°F)
approach.

Since the thermal energy of the inlet gas decreases with decreasing tem-
perature, the water required to reach a fixed approach to saturation
decreases with decreasing inlet temperature.  When the atomizer is operated
at a fixed angular speed and the lime feed rate is low enough such that the
correction factor, Xc, ~1, the mean size of droplets produced in the atom-
izer may be regarded as being independent of the inlet gas temperature.
The decrease of the atomizer water feed rate with decreasing inlet gas tem-
perature then translates to a decrease in the initial total surface area of
droplets produced in the atomizer because fewer droplets are formed.

The decrease of the spray absorber efficiency with a decrease in the inlet
temperature can be related to this lower total initial surface area of
droplets resulting from a lower water feed rate to the atomizer to reach
the same approach to saturation.  The corresponding increase in the
baghouse removal suggests that the slower evaporation rate induced by the
lower inlet temperature leaves particles at the spray absorber outlet with
a higher moisture content, allowing more effective S02 removal in the
baghouse.  For the same gas flow rate at the same inlet temperature, the
atomizer water requirement to reach the 17°C (30°F) approach is less than
that for the 11°C (20°F) approach.  Thus, the higher rate of decrease of
spray absorber efficiency with decreasing inlet temperature for the 17°C
(30°F) approach may arise from the difference in the water requirements and
the associated difference in the initial droplet surface area.

Spray Absorber Inlet Temperature Variation with Prequench

The previous section investigated the effect on S02 removal of varying the
inlet temperature while holding the absolute humidity fixed.  That set of
                                   10-51

-------
             »	P
             cv:
             UJ
             UJ
             t—t
             LJ
                 100-
                  75—
                  50-
                  25-
o
I
                        EFFECT OF HEAT EXCHANGE (STRAIGHT-THROUGH)
                        INLET S02 CONG. = 1000 PPM
                        TP
                       D - SPRAY ABSORBER
                       x - BAGHOUSE
                       + - SYSTEM
I—I—T
       n—i—i—r
                    i—i—r
                             i—i  r
                                                                         100 •
                                            *_J
                                            LU
                                            I—I
                                            t!   50—|
                                            l_l_
                                            U_
                                            LU
                                                                      OJ
                                                                      d>
                                                                      LTb
                                                                          25—
                                                      EFFECT OF HEAT EXCHANGE (STRAIGHT-THROUGH)
                                                      INLET S02 CONC. = 1000 PPM
                                                       Tp - 17°C (30°F); 5 = 1.0
                                                      D - SPRAY ABSORBER
                                                      x - BAGHOUSE
                                                      + - SYSTEM
                    200        250        300        350        400
                         SPRAY ABSORBER INLET TEMPERATURE, DEC. F
  n—i—i—i—|—i—i—i—i—|—i—i—i—i—|—i  i  i  r
200        250        300        350        400
     SPRAY ABSORBER INLET TEMPERATURE, DEC. F
             2 (a). Dependence of the  straight-through  SC>2
                   removal efficiency on the spray absor-
                   ber  inlet gas temperature varied by
                   cooling the inlet  gas with a surface
                   heat exchanger for fixed approach to
                   saturation (11°C,  20°F).
                                            2(b).  Dependence of  the straight-through SC>2
                                                   removal efficiency on the  spray absor-
                                                   ber inlet gas  temperature  varied by
                                                   cooling the  inlet gas with a  surface
                                                   heat exchanger for fixed approach to
                                                   saturation  (17°C, 30°F).
             Figure 2.  Effect of the  spray absorber  inlet gas temperature on the  SOo removal efficiency in
                        pilot lime spray absorber/baghouse system without recycle  of offproduct.   [For these
                        tests, the following parameters  were fixed:  inlet S02 concentration  (1000  ppm),
                        stoichiometric  ratio (1.0), average gas residence time in  spray absorber  (10 seconds),
                        atomizer angular speed (12,500 rpm), and inlet  gas disperser configuration (20° vane
                        angle with vertical and small vane inserts).]

-------
tests was useful in understanding the effect of inlet temperature on spray
absorber/baghouse performance for bituminous and subbituminous coals.  It
is also of interest to determine the effect of inlet temperature with
moisture content in the range representative of flue gas derived from
burning lignite.  Increased flue gas moisture content and decreased spray
absorber inlet temperature were simultaneously accomplished by cooling the
flue gas entering the spray absorber by adiabatic humidification using
spray nozzles.  Figure 3 summarizes the test results.  Flue gas originally
at 177°C (350°F) and a moisture content of 10 percent by volume will, after
adiabatic humidification to 149°C (300°F), contain moisture at 11.5 percent
by volume.  As in the case of the heat exchanged gas, the water required in
the absorber to reach the same approach to saturation will be lower when the
inlet gas has already been partially prequenched.  The lower water require-
ments and the corresponding decrease in the initial droplet surface area
are reflected in the decreasing S02 removal efficiency with decreasing tem-
perature (increasing extent of humidification) seen in Figure 3.  When a
gas is cooled at constant absolute humidity (by surface heat exchange), its
adiabatic saturation temperature decreases.  When it is cooled by adiabatic
humidification (by prequench), its adiabatic saturation temperature remains
the same, but its absolute humidity increases.  As a result, for the same
inlet gas temperature, more water is required by the flue gas cooled by
surface heat exchange than by the prequenched (humidified) flue gas to
reach the same approach to saturation.  Based on this difference in water
requirement, the SC>2 removal efficiency for the heat exchanged gas would be
expected to be higher than that for the prequenched gas for the same inlet
gas temperature.  The experimental results shown in Figures 2(a) and 3(a)
confirm this theoretical expectation.  Moreover, the rate of decrease of the
spray absorber efficiency with inlet temperature for the prequenched gas is
about triple that for the precooled gas.  At the same time, the SC>2 removal
efficiency of the baghouse is correspondingly greater for the prequenched
gas than for the precooled gas at the same inlet temperature.

STRAIGHT-THROUGH AND WET RECYCLE TESTS

Comparison with Theoretical Model

Parametric tests were performed to  characterize S02 removal in the spray
absorber/baghouse system with respect to variations in stoichiometric ratio
(0.7 to 1.5), approach to saturation (7 to 22°C, 12 to 40°F) , inlet S02
concentration (500 to 4,000 ppm), and recycle flyash (mixtures of flyash
derived from eastern, western, and lignitic coals with the Martin Drake
flyash).  Other spray absorber operational parameters [such as atomizer
angular speed (12,500 rpm), inlet gas flow rate (4.01 nrVs, 8,500 cfm), and
inlet gas temperature (177°C, 350°F)] were fixed for these parametric
tests.  Figure 4 presents the results of the parametric tests.  When the
chemical reaction between the lime and S02 is the rate limiting step, the
theoretical model (Equation 1) predicts that all points belonging to a
fixed inlet concentration should correlate on the same curve.  Figure 4
shows good agreement (as supported by correlation coefficients better than
0.9) between measurements and the theoretical model predictions.  The
measurements and the fitted curves show a concentration-dependent perfor-
mance in the spray absorber, with the lowest concentration (500 ppm) giving
                                    10-53

-------
                   100-
                   75 —
               OJ
               Q
               U1
                   25—
o
i
Ln
-P-
                         EFFECT OF PREQUENCH (STRAIGHT-THROUGH)
                         INLET £02 CONC. = 1000 PPM
                         Tp = 11°C (20°F); S = 1.0
                         D - SPRAY ABSORBER
                         x - BAGHOUSE
                         + - SYSTEM
   i  I  I  I  |  I  T  I  I    I  I  I  I    1  I  I  T
200        250        300        350

     SPRAY ABSORBER INLET TEMPERATURE, DEC. F
                                                                         100-
                                                                          75-
                                                 CVJ
                                                CD
                                                in
                                                                          25 —
                                                          EFFECT OF PREOUENCH <50 PERCENT WET RECYCLE)
                                                          INLET S02 CONC.  = 1000 PPM
                                                          TD = 11°C (20°F); S = 1.0
                                                          a - SPRAY ABSORBER
                                                          x - BAGHOUSE
                                                          + - SYSTEM
                                                                    \  \  T  T
I  I  II  F  1  I
     350
                                                                             200        250        300        350        400
                                                                                  SPRAY ABSORBER INLET TEMPERATURE, DEC. F
               3(a).  Dependence  of straight-through SC>2
                     removal efficiency on the  spray absor-
                     ber inlet gas temperature  varied by
                     incomplete  adiabatic humidification
                     of the inlet  gas (prequench).
                                                3(b). Dependence of the  S02  removal efficiency
                                                      under  wet recycle  conditions on  the spray
                                                      absorber inlet gas temperature varied by
                                                      incomplete adiabatic humidification of
                                                      the  inlet gas (prequench).
              Figure  3.   S02 removal efficiency  in pilot lime  spray absorber/baghouse system  for both straight-
                          through  and wet recycle of offproduct operation with water prequenching of the flue  gas
                          ahead of  the spray absorber.   [For these tests, the following parameters were fixed:
                          inlet SC>2  concentration (1000 ppm), stoichiometric ratio (1.0), average gas residence
                          time in  spray absorber  (10 seconds),  atomizer angular  speed (12,500  rpm),  inlet gas
                          disperser  configuration (20°  vane angle  with vertical  and small vane inserts), and
                          approach  to saturation  (11°C, 20°F).]

-------
                100 •
                 90 —
             OJ
             CD
             in
             CD
             in
             GO
                 70—
                 60 —
                 50-
o
I
                 40-
                         INLET S02
                        CONC., PPM
                                                        (20°F)
                            - Tp = 17°C (30°F)

                          o  - Tp = 22°C (40°F)
1.0
                           \^ |  I I  I I  | I  I  I I  | I  I I  I  | I  I I  I
                            2.0     3.0     4.0     5.0     6.0
                           S X LN«TINLET - TflS>/(TEX1T -
                                                                        100-
                                                     90	
                                                                     n   80 —
                                                                         70 —
                                                  CM
                                                 CD
                                                 in
                                                                         60-
                                                                         50 —
                                                             INLET S02
                                                            CONC., PPfl
                                                                                  s - Tp =
                                                                                            (20°F)
                                                                                                       - TD = 17°C (30°F)
o - TD = 22°C (40°F)
                                                           i i  i i  i  i i  i i  i i  i  i i  i i  i i  i  i i  i i  r
                                                        1.0      2.0     S.0     4.0     5.0     S.0
                                                               S X LN«TINLET - Tfls>/2 removal  efficiency
                                                       data for  operation with wet recycle of
                                                       offproduct  in the absence of external
                                                       flyash  injection compared with
                                                       Equation  1.
             Figure 4.   Comparison of spray absorber S02 removal  efficiency data  (points) from pilot lime spray
                         absorber/baghouse system with predictions of Equation  1  (curves) for straight-through
                         and wet  recycle operation.

-------
the best performance.   Successive curves in Figure 4 are separated by a
change in the inlet concentration of 500 ppm.  For a fixed inlet gas tem-
perature and inlet gas moisture content, the rate of water flow in the ato-
mizer (rate of water evaporated in the spray absorber) is fixed by the
approach to saturation at the spray absorber outlet.  At the same time, the
lime feed rate to the atomizer required to maintain a fixed stoichiometric
ratio increases with increasing inlet S02 concentration.  Consequently, the
fraction of solids (lime) in the atomizer feed (or in a newly produced
slurry droplet) increases with increasing inlet SC>2 concentration for a
fixed stoichiometric ratio and approach to saturation.  The observed
deterioration of the spray absorber S02 removal efficiency under straight-
through conditions with increasing inlet S02 concentration in the range of
1000 to 2500 ppm may be explained by the decreasing evaporation time of a
slurry droplet with increasing S02 concentration.  Figure 4 shows the
straight-through S02 removal efficiency in the spray absorber to be nearly
independent of inlet 862 concentration in the range 0-1000 ppm.  This
concentration-independent performance in the spray absorber for low inlet
S02 concentrations is also predictable by the analytical model.  In this
case, the solids fraction in the atomizer is low enough that it essentially
has no effect on the droplet evaporation time.  Figure 4 also shows that
the introduction of partial wet recycle of the offproduct results in the
near restoration of concentration-independent spray absorber S02 removal.
It appears that this recycle enhancement results from improved reaction
sites and a greater lime availability at the same stoichiometric ratio.

Characterization of the Flyash Effect

The correlation based on Equation 2 for different types of flyash in the
slurry droplets is illustrated in Figure 5.  The pertinent ash charac-
teristics are presented in Table 3.  The notable feature of the correlation
is the inclusion of straight-through as well as wet recycle data for all
combinations of S02 concentration, stoichiometric ratio, and approach to
saturation on a single curve.  It is equally noteworthy that, despite phy-
sical and chemical differences, the data involving all three types of
eastern flyash correlate on a single curve.

The recycle enhancement of 862 removal efficiency is the lowest for the
Martin Drake flyash (no external flyash injection) which gives a nearly
neutral hydrolysis (pH - 7).  The recycle enhancement is higher in the pre-
sence of the flyash derived from the Texas lignite which leads to an alka-
line hydrolysis (pH ~ 8).   The performance enhancement from wet recycle is
dramatic in the presence of flyash derived from eastern coal which yields
acidic hydrolysis  (pH - 4).  This extremely high recycle enhancement sup-
ports application  of the DFGD technique to high sulfur coal.

Under straight-through conditions, the effect of an increased solids frac-
tion was to decrease S02 removal efficiency in the spray absorber due to
the decrease in the evaporation time of the slurry droplets.  In contrast,
under wet recycle  conditions, the effect of an increased solids fraction
appears to be an increased spray absorber S02 removal efficiency.  This
increased efficiency suggests that the beneficial effects of flyash in
                                   10-56

-------
o
 I
Ln
                     NO INJECTED FLYRSH
                     • - STRfllGHT-THROUGH
                     « - HET RECYCLE
                                   SPRflY flBSORBER
n	1	1	]  i   i  i  ]   i  i   r
       0.2        0.4        0.6
 MflSS FRflCTION OF SOLIDS IN flTOMIZER FEED
                                       TEXflS LIGNITIC FLYRSH INJECTED
                                       o - STRfllGHT-THROUGH
                                       O - WET RECYCLE
                                                                          SPRflY flBSORBER
T	1	i	1	1	1	1	1	1	1	T
      0.2        0.4       0.6
 MflSS FRACTION OF SOLIDS IN flTOMIZER FEED
                                       EflSTERN COflL FLYflSHES INJECTED
                                        BOILER  STRftlGHT  WET
                                        TYPE  -THROUGH RECYCLE
                                       CYCLONE    x
                                       TflNGENT.    +
                                       STOKER    >•
                                                                                                                  SPRflY flBSORBER
T	1	1	1	1	1	1	1  I   I  r
      B.2        0.4
 BflSS FRflCTION OF SOLIDS IN RTOHIZER FEED
              5(a). Reduced  penetration  in
                    the spray absorber with
                    no injected flyash com-
                    pared with Equation  2.
                                5(b). Reduced  penetration in
                                      the spray absorber with
                                      injection of Texas lig-
                                      nitic flyash compared
                                      with Equation 2.
                                5(c).  Reduced  penetration  in
                                       the spray absorber with
                                       injection of various
                                       eastern  coal flyashes
                                       compared with
                                       Equation 2.
             Figure  5.   Comparison of spray absorber  reduced  penetration data (points) from pilot lime  spray
                          absorber/baghouse system with predictions of  Equation 2  (straight  lines)  for straight-
                          through and wet  recycle operation with and without  external flyash injection.

-------
                Table 3.   CHEMICAL  AND  PHYSICAL  CHARACTERISTICS OF FLYASH EVALUATED FOR THEIR
                           EFFECT  ON S02 REMOVAL  IN PILOT SPRAY DRYER
Composition, percent by mass
Constituent
or
Characteristic
Si02
A1203
CaO
F6203
Na20
K20
Li20
MgO
P205
S03
LOIf
Initial pH?
Alkalinity,
meq/gh
Mass median
diameter, (am
Log-normal
standard deviation*
Volume-to-surf ace
diameter, yml
Specific surface area,
">2/kg
Typical surface area
at spray dryer inlet
m2/m3 (ft2/ft3) gas
Typical concentration
at spray dryer inlet
1CT3 kg/m3 (gr/ft3)
Stoker a
Eastern
Coal
34.00
14.34
2.19
23.64
0.49
1.90
0.02
1.17
0.38
1.11
19.40
4.7
-0.062

67
4.1


24.8

106

0.62(0.19)

5.7(2.5)
Tangential
Eastern
Coal
45.77
19.26
2.05
25.96
0.41
1.98
0.02
0.90
0.16
0.45
1.42
4.4
-0.038

28.5
4.3


9.8

251

1.15(0.35)

4.6(2.0)
Cyclone c Front-fired d
Eastern
Coal
27.99
13.47
1.49
21.01
0.31
1.75
0.02
0.69
0.24
0.53
31.34
3.2
-0.307

14.3
2.6


9.1

253

0.59(0.18)

2.3(1.0)
Western
Coal
62.44
23.17
3.15
3.70
0.21
1.20
0.02
1.27
1.22
0.45
1.85
>7 .0
0.180

17.3
2.8
2.8

10.2

280

1.15(0.35)

4.1(1.8)
Lignite6
flyash

60.05
21.91
8.87
3.09
0.46
0.89
0.02
2.04
0.10
0.31
0.10
>7 .0
5.12

—
—


—

—

—

8.0(3.5)
a Stoker-fired boiler at  the  University  of  Iowa

° Tangentially fired boiler  in  Units  5 and  6 of  the Conesville  Station operated  by  Columbus and Southern Ohio
  Electric

c Cyclone-fired boilers  in Unit 2  of  the Conesville Station  operated  by  Columbus and  Southern Ohio Electric (inlet
  hopper sample)

d Front-fired boiler in  Unit  6  of  the Martin Drake Station operated by the  City  of  Colorado Springs

e Flyash derived from burning Texas  lignite at the Monticello  Station operated by Texas Utilities Services

f LOI stands for loss on ignition  at  800°C  (1472°F).

S Initial pH of 0.01 kg  fly  ash in 0.1 i water at 52°C  (126°F)

h Negative alkalinity indicates acidity  being titrated  against  base to a final pH of  7.5.

1 Log-normal standard deviation, a,  is obtained  from a  -  0.5 (zj/T +  z7z2)  where 15.9, 50.0, and 84.1 percent of
  the mass of the particle distribution  originate from  particles  whose diameters are  greater than zj, z", and z2.

J Equal to djj exp [-1/2(C n o )2 ] where  dm is the mass median diameter and a is the  log-normal standard deviation
                                                     10-58

-------
enhancing the rate of chemical reaction between the S02 and lime override
the limitation in efficiency due to reduced evaporation time.
                                CONCLUSION
The data obtained from the Colorado Springs DFGD test program strongly
support the models given by Equations 1 and 2.  As a result, S02 removal
performance can now be accurately predicted for a conservatively designed
system when the stoichiometric ratio, approach to saturation temperature,
and slurry solids content are defined.  The predictions can be extended to
systems using ash recycle when an additional empirical constant derived
from the flyash type is available.

An important DFGD feature which was identified is the manner in which the
spray absorber and baghouse work in unison toward maximizing the system
S02 removal efficiency.  This trend is evident in the variation of the
S02 removal efficiencies with inlet temperature.  The decrease in spray
absorber performance with decreasing inlet temperature is seen to be com-
pensated by an increase in the baghouse performance to result in a nearly
constant system removal.  Conditions of high humidity and low inlet tem-
perature are favorable to good S02 removal in the baghouse.  Conversely,
conditions of low humidity and high inlet temperature appear to be
favorable to good S02 removal by the spray absorber.  Significantly,
however, the system performance appears to be only weakly dependent on
either the inlet temperature or the humidity which may occur either as a
result of variations in coal moisture content or variations in the ambient
(combustion) air.  The spray absorber/baghouse system appears capable of
effecting good performance even in the presence of these variations when no
baghouse reheat is provided.  Baghouse reheat removes the compensating
character of baghouse operation and allows the spray absorber sensitivity
to inlet humidity and inlet temperature to predominate.
                                    10-59

-------
                               REFERENCES
1.   Downs,  W.,  W.J.  Sanders,  and  C.E.  Miller.   "Control of S02 Emissions by
    Dry Scrubbing."   (Presented at  the American Power Conference,  Chicago,
    Illinois.   April 21-23,  1980.)

2.   Apple,  C.  and M.E.  Kelly.   "Mechanisms  of  Dry  S02 Control Processes."
    EPA-600/7-82-026 (NTIS No.  PB 82-196924),  April  1982.

3.   Parsons, E.L., Jr., L.F.  Hemenway, O.T.  Kragh, T.G.  Brna,  and  R.L.
    Ostop.   "S02 Removal by  Dry FGD."   In Proceedings:   Symposium  on Flue
    Gas Desulfurization - Houston,  October  1980, Volume  2,
    EPA-600/9-81-019b (NTIS  No. PB81-243164),  April  1981,  pp.  801-852.

4.   Parsons, E.L., Jr., V. Boscak,  T.G.  Brna,  and  R.L.  Ostop.   "S02 Removal
    by Dry Injection and Spray  Absorption Techniques."  in  Third Symposium
    on the Transfer and Utilization of Particulate Control Technology,
    Volume I,  EPA-600/9-82-005a (NTIS  No. PB83-149583),  April  1982, pp.
    303-312.

5.   Samuel, E.A., T.W.  Lugar, D.E.  Lapp,  O.F.  Fortune,  T.G.  Brna,  and R.L.
    Ostop.   "Dry FGD Pilot Plant  Results:   Lime  Spray Absorption for High
    Sulfur Coal and Dry Injection of  Sodium Compounds for  Low  Sulfur Coal."
    In Proceedings:   Symposium  on Flue Gas  Desulfurization,  Hollywood, May
    1982, Volume II, EPRI CS-2897,  March  1983,  pp. 574-594.

6.   Samuel, E.A., K.R.  Murphy, T.W.  Lugar,  E.L.  Parsons, Jr.,  and  D.E.
    Lapp. "Evaluation of Spray Absorption FGD," Draft Report submitted to
    EPA under  Contract  No. 68-02-3119.

7.   Marshall,  W.E. and  W.R. Marshall,  Jr. "Evaporation  from  Drops  (Parts 1
    and 2)."  Chemical  Engineering  Progress  48,  141  and  173.  1952.

8.   Masters, K.  "Spray Drying Handbook."   George  Goodwin  Ltd., London,
    1979.
                                   10-60

-------
DRY SCRUBBER, FLUE GAS DESULFURIZATION ON HIGH SULFUR,
 COAL-FIRED STEAM GENERATORS:  PILOT-SCALE EVALUATION

       B. J. Jankura, J. B. Doyle, T. J. Flynn

-------
            DRY SCRUBBER, FLUE GAS DESULFURIZATION ON HIGH-SULFUR,
             COAL-FIRED STEAM GENERATORS: PILOT-SCALE EVALUATION
                                     By:

                      B. J. Jankura, J. B. Doyle,
                      and T. J. Flynn
                      Babcock & Wilcox Company
                      Research and Development Division
                      Alliance Research Center
                      Alliance, Ohio  44601
                                   ABSTRACT

     This paper describes the pilot-scale investigation of methods for adapt-
ing dry flue gas desulfurization (FGD) to utility steam generators burning
high-sulfur coal.  Development of the dry scrubber for FGD was initially
directed toward reducing  SO  in steam generators burning low-sulfur, western
coals. The reason for limiting dry scrubbing to low-sulfur coals was two-fold:
First, federal New Source Performance Standards (70% reduction) were less
stringent.  Second, western coals generally contain less sulfur and large
amounts of ash alkali.  This significantly contributes to the effectiveness of
the dry scrubbing process.  Several potential drawbacks (both technical and
economic) limiting dry scrubber technology to western coals are discussed.

     The Department of Energy (DOE) conducted dry scrubber FGD tests in 1981.
Preliminary results indicated dry scrubbing could remove more than 90% of the
sulfur released from the combustion of eastern coals at less-than-anticipated
levels of lime consumption.  The Babcock & Wilcox Company (B&W), under DOE
contract, has tested the dry scrubbing process to evaluate the variables that
have a major effect on SO. capture.

     There are potential ways of increasing SO  capture via dry scrubbing. One
method requires limestone injection into the boiler furnace, reducing SO  lev-
els to the scrubber, calcines limestone to more reactive lime, and simulates a
high-alkali-ash, high-sulfur coal.  Other methods covered include recycle and
dry scrubber approach to saturation at temperatures less than 20°F to obtain
high SO  capture.
Prepared for presentation at the Environmental Protection Agency/Electric
Power Research Institute Symposium on Flue Gas Desulfurization, New Orleans,
La., November 1-4, 1983.
                                      10-61

-------
                                 INTRODUCTION
     The use of dry scrubbers for controlling SO  emissions from power  plants
burning low-sulfur coal has gained wide industrial acceptance.  A natural  out-
growth of the success of this technology is to establish its limits  of  appli-
cation in regard to higher-sulfur coals. The interest in developing  dry scrub-
ber technology has been accelerated due to growing concerns about controlling
acid precipitation.

BACKGROUND

     Using dry scrubbers for flue gas desulfurization began in  the  late 1970s.
The first commercial utility units are operating on low-sulfur  coal.   B&W1 s
development  of dry scrubber technology has been a four step process:

     •   A small, field, pilot plant (8000 ACFM) was constructed and  put
         into operation  in  1978 at the Basin Electric Neal  Station, Velva,
         N. D.  This  pilot  provided basic process information but limited
         information  for scale-up and system sensitivity to fuel changes.

     •   To satisfy the  need for additional design data, B&W proceeded
         with construction  and operation of a 120,000-ACFM  field demon-
         stration unit to demonstrate mechanical scale-up of the system.

     •   B&W  then  constructed a smaller  1500-ACFM pilot plant at the  its
         Alliance  (Ohio) Research Center (ARC).  This unit  was designed to
         develop basic process data.

     •   The  net result  of  B&W's development efforts has been construction
         and  start-up a  dry scrubbing  system consisting of four units of
         800,000 ACFM each  (600 MW total) at Basin Electric's Laramie
         River Station,  Wheatland, Wyoming.  (Another 450-MW unit at  Colo-
         rado Ute's Craig Station is scheduled for start-up in 1984.)

 The data obtained  from  the B&W pilot plants and demonstration unit  has been
 used to establish  a  basis  for predicting performance for dry scrubber systems.

                              PROGRAM OBJECTIVES

     The objective of the  B&W-DOE program was to evaluate  dry scrubber tech-
 nology  when  applied  to  high-sulfur coals.  The first step  was to develop para-
 metric  baseline data using standard dry scrubber practices. This would be  fol-
 lowed  by using innovated methods for improving performance.  This  program was
 funded  writ hi n the  DOE Advanced Environmental Control Technology (AECT) plan
 initiated in 1979.   The AECT program will develop the technology base for
 controlling  contaminants produced during coal conversion  [1],

     To accomplish the  objectives, a wide range of process data were obtained.
 We  decided  that the  B&W pilot plant at  (ARC) would be most suitable to meet
 these  objectives.  This unit was selected because it was small  enough
 (1500 ACFM)  to keep  operating costs reasonable.  Also, the unit was  character-
 ized against a larger (120,000-ACFM) field demonstration unit that  provided  a
 sound basis  for scaling up the data.
                                      10-62

-------
                        DESCRIPTION OF TEST FACILITIES

     B&W's 1500-ACFM, pilot dry scrubber  (Figure  1) is highly  flexible  and  de-
signed so a wide range of variables could be investigated.  The  system  starts
with a combustion chamber that has a rating of 5  x  10  Btu/Hr  heat  input.   The
combustor is also capable of firing oil or gas.   The exhaust gas  from the com-
bustor can be directly vented to the atmosphere or  diverted to the  dry  scrub-
ber pilot facility.  The gas stream going to the  scrubber  can  be  conditioned
with fly ash and/or SO .  This capability provides  a way of simulating  flue
gas pollutants without burning actual fuel.  The  flue gas  can  travel directly
to the dry scrubber or pass through a heat exchanger that  controls  the  flue
gas temperature at the scrubber inlet.

     The first compartment of the dry scrubber is shown in Figure 2.  A per-
forated plate is located at the inlet for distribution purposes.  The gas then
passes through a TurboDiffuser™, where swirl energy is introduced to the gas
stream.  The gas is then intimately mixed with a  finely atomized  reagent
slurry as it leaves the TurboDiffuser™.  Here, the  S09 reacts  with  the  spray's
alkali as moisture in the spray is evaporated into  the flue gas.  A portion of
dry, spent product consisting of fly ash, calcium sulfates, calcium sulfites,
and unreacted lime deposits in the hoppers at the bottom of the  scrubber
(Figure 3).  The remainder is carried to a cyclone  collector and/or baghouse,
where the majority of the particulate by-product  is removed.   The gas stream
is then exhausted through an induced-draft fan to the atmosphere.

     The lime slurry reagent preparation system shown in Figure  1 is composed
of a reagent hopper feeding dry bulk reagent through a rotary  valve to  a com-
mercial paste slaker equipped with an automated grit separator.   The lime
putty then discharges by gravity through a screen to a heated  slurry tank.  The
tank is continuously agitated and the slurry is continuously pumped around  a
recycle loop to avoid potential settling  out of the reagent and  associated
plugging problems.  The slurry feed line  to the dry scrubber is  equipped with
a strainer to protect the spray nozzle from plugging.  Feed slurry  is intro-
duced to the scrubber through a pneumatic spray nozzle (Figure 4).

                                  TEST PLAN

     A test plan was developed based on dry scrubber data  obtained  during
earlier development by B&W on low-sulfur coals.   The starting  point was to
establish a list of process and operating variables that were  known to
influence process performance.  (These variables  are listed in order of
significance on Table 1; the test plan is shown on  Table 2.)   Included  in the
test plan were methods for improving the overall  process performance for dry
scrubbers when used in high-sulfur coal applications; these methods are:

     •  Use of recycle.

     •  Use of low temperature at the scrubber outlet.

     •  Use of furnace limestone injection in combination  with the  dry
        scrubber.
                                     10-63

-------
o
 I
                                                       PULVERIZED
                                                       LIMESTONE
                                                       HOPPER
                                                              PREHEATED
                                                              SECONDARY
                                                              COMBUSTION AIR
                               WATER
GRIT
                                   RECYCLE
                                                       STRAINER
                                                                              TEMPERATURE
                                                                              CONTROL
                                                                              WATER
                                  BCTU
                                  STACK
                                                                                                         n
                                               DRY
                                               ASH
                                               INJECTOR
STEAM
HEATER
                                                       LIQUID SO2
                                            WATER
                                            COOLED
                                            HEAT
                                            EXCHANGER



BAG!
                                                                                                                              ASH
                                                                                                    ASH
                                                            Figure 1.   Babcock & Wilcox 1500-ACFM dry scrubber pilot.

-------
                             FLUE  GAS
   DISTANCE PIECE
SLURRY/ATOMIZING
AIR OR STEAM
                        DISTRIBUTION PLATE
                     GAS FLOW
                    GAS FLOW
           *VENT TUBE CLOSED DURING ARC TESTING

         Figure 2.   Dry scrubber plenum with distributor
                    plate and throat configuration.
                             10-65

-------
?
FLUE
GAS
                    PLENUM

                    REGISTER ARRANGEMENT

                    DSR  REACTOR

                    COMPARTMENT  NO. 1 HOPPER

                    COMPARTMENT  NO. 2 HOPPER

                    COMPARTMENT  NO. 3 HOPPER

                    DUPONT S02 ANALYZER
              Figure 3.   Dry scrubber dimensions.
                                         EXIT HOLE
—SLURRY-r-*	J-H-P_
                                      INSERT
                            3 HOLES, 120° APART
                            AIR OR  STEAM HOLES
       Figure 4.   Advanced sprayer plate for pneumatic
                 atomization of slurries.
                           10-66

-------
                       Table 1
     PROCESS AND OPERATING VARIABLES INFLUENCING
           DRY SCRUBBER SYSTEM PERFORMANCE
      Process
•  Reagent type
•  Dryer inlet
   temperature
•  Dryer Inlet S
   concentration
•  Fly ash
   composition
         Operating
•  Stoichiometry
•  Dryer outlet temperature
•  Solid recycle rate
•  Dryer residence time
                       Table  2
             DESCRIPTION OF DRY SCRUBBER
                    TEST VARIABLES
           Baseline  tests :
             Stoichiometric ratio
             Dryer Inlet temperature
             Dryer outlet temperature
             Baghouse  outlet  temperature
             Baghouse  air-to-cloth
             Baghouse  pressure  drop
             Dryer residence  time
             Reagent variation
             Coal  variation
           Recycle tests
           Low-temperature  outlet  tests
           Limestone injection
                        10-67

-------
                        PILOT PLANT TEST RESULTS

     The major emphasis of dry scrubber testing was  to evaluate  utilizing lime
slurry reagent at high concentrations of SO  at the  inlet  under  a variety of
process and operating conditions.  Holding other variables  constant,  testing
began by generating a nonrecycle data base that included a variation  of both
reagent and coal.  The first series of tests were designed to  show the effect
of stoichiometry (over a wide range) on S02 capture.  Subsequent test series
looked at recycle, low temperature at the scrubber outlet,  and limestone
injection as an alternative means of improving lime  utilization.  Baghouse,
dryer residence time, and coal variation test results are  not  yet available.

     Lime slurry reagent was prepared on site with a paste  slaker. The Miss-
issippi pebble lime specifications are listed in Table 3.   Typical chemical
analysis of pebble line samples taken during slaking operation are shown in
Table 4.  The lime slurry has approximately 50% wt/wt below 7-micron-diameter
particles.

     The test furnace was fired with a Western Kentucky bituminous coal; chem-
ical analyses of the coal and ash are shown in Tables 5 and 6.  The sulfur
content  of  this  coal ranged from 2.6% to 3.0%; furnace excess  air ranged from
15% to  20%.

STOICHIOMETRIC RATIO AND OUTLET TEMPERATURE VARIATION

     Once-through dry scrubber plus baghouse SO  absorption values are shown
in Figure 5;  typical inlet conditions were:

     Inlet  SO  level             —  1900 - 2300 ppm
     Inlet  Gas Temperature       —  290° - 311°F
     Dryer  Gas Residence Time    —  9.6 to 10.8 sec
     Atomizing Air Consumption   —  0.17 to 0.21 Ibs air/lbs  slurry

     Baghouse inlet temperature was held at least 20°F above the flue gas
adiabatic saturation temperature to prevent bag blinding.   The approach to
saturation  temperature (AST) is calculated as the difference between  the flue
gas bulk temperature and its adiabatic saturation temperature.

     The significance the dry scrubber's outlet temperature has  on S09 capture
is evident  in Figure 5. At a stoichiometry of 1.15,  approximately 90% SO
capture was measured at a 10°F AST.                                      ^

     Dry scrubber absorbers have been recently tested on high-sulfur  coals by
several researchers.  At a lime feed rate corresponding to a stoichiometric
ratio of 1.2, and inlet S02 concentration of 2000 PPM, Samuel, et al. [2],
obtained 84%  S02 capture with an AST of 20°F.  Yeh, et al. [3], using commer-
cial hydrated lime, obtained approximately 65% SO  capture  at  a 1.0 stoich-
iometry and inlet SO  concentration of 2100 ppm and  a 30°F AST.   These results
are consistent with  the baseline data in Figure 5.
                                     10-68

-------
                    Table 3
TYPICAL MISSISSIPPI PEBBLE LIME SPECIFICATIONS
     Mississippi or Peerless Rotary Plant
               Pebble Quickline
      Size Designation 1/2 - 1 -  2 inches

CaO — available
CaO — Total . .
CaCO 	
CaSO, 	
4
S — equivalent •
SiO 	
2
Al-O., 	
23
Fe o 	
23
MgO 	
LOI 	
Chemical Analysis
	 93.0%
	 96.0
	 03
	 0.04
	 0.01
	 0.75
	 0.15
	 0.08
	 0.53
	 0.30
                                    to
                                    to
                                    to
                                    to
                                    to
                                    to
                                    to
                                    to
                                    to
                                    to
                                           97.0%
                                           98.0
                                            4.0
                                            0.21
                                            0.05
                                            1.10
                                            0.35
                                            0.12
                                            0.75
                                            2.00
                    Table 4
  CHEMICAL ANALYSIS OF MISSISSIPPI PEBBLE LIME
                    022682        042382
                  Negligible    Negligible
                      0.1           0.06
                      1.0           0.17
                     93.3          96.5
                      0.6       not available
Date
Moisture (%)
S as S03 (%)
C03 as C02 (%
Ca as CaO (%)
Mg as Mg (%)
                    10-69

-------
                               Table 5
Dry Scrubber
ACG-S3-4030-31
April 5, 1982
Sample No.
Description
                     TYPICAL CHEMICAL ANALYSIS:
                        WESTERN KENTUCKY COAL
C-16459
                       C-16460
Basis
Total Moisture, %
Proximate Analysis, %
Moisture
Volatile Matter
Fixed Carbon
Ash
Gross Heating  Value,
   Btu per  Ib.
Btu per Ib.  (M&A-Free)
 Ultimate  Analysis, %
Moisture
 Carbon
 Hydrogen
 Nitrogen
 Sulfur
 Ash
 Oxygen (Difference)
   Total
Pulverized
Run 0011,
0700
As Received
3.8
3.8
40.9
45.1
10.2
12450
—
3.8
69.1
4.9
1.33
2.92
10.2
7.75
100.00
Coal
031082
Dry
	
42.5
46.9
10.6
12940
14470
—
71.9
5.1
1.38
3.04
10.6
7.98
100.00
Pulverized Coal
031282, 0830
As Received
3.8
3.8
41.6
44.4
10.2
12420
—
3.8
68.8
4.8
1.38
3.05
10.2
7.97
100.00
Dry
—
43.3
46.1
10.6
12910
14450
	
71.5
5.0
1.43
3.17
10.6
8.30
100.00
                                10-70

-------
                                 Table 6
         TYPICAL  CHEMICAL ANALYSIS: WESTERN KENTUCKY COAL ASH
Dry Scrubber
ACG-83-4030-31
April 5,  1982
Sample No.
Description
Ash Analysis  (Spectrographlc),  %
Silicon as  SiO
Aluminum  as  Al  0
Iron as Fe20,
Titanium  as  TiO
Calcium as  CaO
Magnesium as MgO
Sodium as  Na 0*
Potassium as K  0*
Sulfur as  SO
Phosphorus  as P90.
C-16459
Pulverized Coal
Run 0011, 031082
0700
     48.61
     18.75
     19.26
      0.99
      3.04
      0.92
      0.51
      2.83
      2.72
      0.42
C-16460
Pulverized Coal
031282, 0830
     48.86
     19.26
     19.92
      0.99
      3.12
      0.90
      0.53
      2.71
      3.05
      0.44
       100
     IT
     t
     <
     O  50
     f\i
     O
     O
APPROACH TO SATURATION
TEMPERATURE
       35°F (19.4rC)
       15°F (8.3 = C)
       10°F (5.6°C)
       6°F (3.3°C)
       20°F (11.1°C)
       RESULTS OF
       SAMUEL, ET AL.
       30°F (16.7-C)
       RESULTS OF
       YEH. ET AL.
                                   1.0
                       MOLES CALCIUM/MOLE SULFUR
                                                             2.0
                   Figure 5.   Stoichiometric ratio testing.
                                  10-71

-------
     In general, eastern coals contain relatively small amounts of  low alka
line ash when compared to Western coals.  Previous B&W work on several western
coals indicated dry scrubber S09 capture of 25% - 35% at 0.0 stoichiometry.
For the 2000 ppm SO  flue gas treated at a 35°F AST, total S02 capture was
less than 10%, as expected, at a 0.0 stoichiometry.

TEMPERATURE VARIATION AT SCRUBBER INLET

     Evaporation time of slurry droplets containing relatively insoluble
solids is characterized by the slurries' physical properties (solids  content,
density, initial particle diameter, and temperature) and the dryer  inlet  and
outlet temperature.  Typical dry scrubbers in FGD applications reduce inlet
temperature by 100° - 150°F (38° - 66°C).  The effect of this temperature
difference on SO  capture for western coals was shown to be very  small by
Parsons, et al. [4].  Figure 6 shows the effect of increasing dry scrubber
inlet temperature for a 35°F AST and a 1.0 stoichiometry.  This indirectly
shows how scrubber temperature drop affects SO  capture and clearly indicates
a significant influence of inlet temperature on SO  capture.  Previous B&W
tests with high-sulfur coal also implied inlet temperature can play a more
dominant role on SO  removal for high-sulfur coals in dry scrubber
applications  [5].

REAGENT VARIATION

     Commercial and industrial dry scrubber FGD systems have relied on lime
slurries as the principal reagent.  Limestone slurries have shown acceptable
S0? absorption properties in low-sulfur, wet scrubber FGD.  But in  dry scrub-
bing FGD, these slurries have unacceptable reactivity.  Research  in various
lime and, most recently, sodium-based reagents (such as Trona  [4] and soda
ash [3]) has  characterized relative reactivities.

     Johnson  [6] performed tests with fly ash alkali and concluded  certain
high-calcium  ash, western coals contained enough inherent alkali  to obtain  the
required  SO   capture.  The majority of commercial  dry scrubber systems  today
use lime  reagents;  one exception is Rockwell's Coyote Station  dry scrubber,
which  uses  soda  ash.

      Table  7  shows  chemical  analysis of  two reagents tested for SO   capture.
 The soda  ash  was a  granular, high-quality (99+%),  commercial-grade  sodium car-
 bonate (Na^CO^).   This material is  a much stronger  base than  lime and forms
 hygroscopic solids,  both properties enhancing SO   capture in the  aqueous- and
 dry-phase chemistry.   A thiosorbic  lime  containing 4.8% Mg  as  MgO was also
 tested.  Figure  7  indicates  total  SO  capture for  soda ash, thiosorbic lime,
 and the base  Mississippi lime.  As  expected, the soda ash outperformed both
 lime reagents.   The  thiosorbic  lime tests indicate  poorer SO   capture relative
 to the base lime.   As  noted,  the  thiosorbic lime supplier reported  excess
 carbonate core.   This  was  due to  low kiln temperatures.  While  the  reported
 slaking time  is  well within  the 3-minute-limit slaking time  for high-reactiv-
 ity lime,  these  tests  should be repeated.
                                      10-72

-------
DC

I
 CM
O
(A
    100
    80
    60
    40
    20
                                         35°F |19°C) AST
                                         1.0 STOICHIOMETRY
                                        O
                               O
                                             J_
      200
                   250          300         350

                  SPRAY DRYER INLET TEMPERATURE (°F)
                                                          400
        Figure 6.   Effect of gas inlet temperature on SO2 capture.
                               Table  7
                 REAGENT VARIATION CHEMICAL ANALYSIS
        Sample  No.

        Description



        Basic

        Calcium as CaO,  %
        Magnesium as  MgO,  %

        Carbonate as  CO  ,  %
        Total  Sulfur  as  SO   %

        Total  Insolubles,  %
          (including  Si02)
        Carbonate as  Na  CO.,,  %
        (as-recvd.(  calculated)
                                  M-38236

                                  Soda Ash
                                  7-6-83
                                  1500

                                  As-Received
                                     41.46
                                     99.84
M-38237

Thiosorbic
Lime
7-5-83

As-Received

   88.35
    4.85

    0.39

    0.34
    3.00
        Note:  Vendor reported thiosorbic shipment contained
              3.1% excessive core, 91.9% available lime index.
              and 1-minute slaking time.
                               10-73

-------
               100
            cc
            z>
            0.
            o
            o
            (fl
             <
             o
                50
                   CONDITIONS
                     > 300DF (149°C)
                      INLET TEMPERATURE
                     » 35°F (19.4°C) AST
                     » 10 SEC GAS
                      RESIDENCE TIME
                   O
                                            o
                                            A
REAGENT

SODA ASH
HIGH CALCIUM
LIME
THIOSORBIC
LIME
                                       i.o
                                                           2.0
                              MOLES ALKALI/MOLE SULFUR
                           Figure 7.   Reagent variation testing.
RECYCLE TESTS
     The degree of  reagent  utilization in a dry scrubber  system has a signifi-
cant effect on system  economics.   One significant difference between wet and
dry scrubbing systems  is  the lower reagent utilization  in dry scrubbers.  This
is mainly due to  type  and rate of gas-liquid contact.   Wet scrubbers use rela-
tively  large  liquid-tq-gas  ratios, on the order of  40 - 90 gpm/1000 ft  gas,
treated (53 -  119 L/M  )  to  achieve intimate gas-liquid  contact with counter-
current absorption trays.   Dry scrubbers in FGD have average gas-liquid
ratios  of 0.2 - 0.3 gpm/1000 ACFM (0.267 - 0.40 L/M ) and generally use co-
current gas-liquid contact  via a gas diffuser  and atomizer.

     Recycle  of solid  material collected in the dry scrubber and particulate
collector  can improve  reagent utilization by reusing unreacted calcium hydrox-
ide  and other fly ash  alkali.  The second source of alkaline material — fly
ash  —  originates from calcined carbonates of  calcium,  sodium, and potassium
formed  during coal combustion. Figure 8 shows  the increase in dry scrubber  S09
capture when  the  feed  slurry ash content increases  for  a 33°F (18°C) and  19°F
(10.6°C) AST.   Approximately 7% more sulfur was captured at a 10% wt/wt wet
recycle rate.   Samuel, et al.  [2] conducted similar tests at substantially
higher  recycle  rates.   Note that his work was  at a  1.2  stoichiometry and  a
utilization at  no recycle of 63%.
                                      10-74

-------


CAPTURE
CM
O
CO
CC
LU
OC
Q
EC
Q.





A
O
D
100
90
80
—
^**
&r-
"t
A-^
60^
50
40
30
20
10
0
0
fA
—

-
-
-



TEST
RUNS
BASE,
175. 176
BASE,
183, 184
-
^^<> r
^1. 	 — 	
. 	
A






1 1 1 I
]









10 20 30 40 50
RECYCLE ASH CONTENT IN FEED SLURRY (% WT/WT)
AST INLET INLET
Ca/S (F/C) SO2 (PPM| TEMP. (F/C| SOURCE
1.6 33/18 2200 303/151 B&W
1.6 19/10.6 2150 305/152 B&W
1.2 20/11.1 2000 - REF4
       Figure 8.   Wet recycle spray of dry scrubber system solids.
Description
                               Table 8




             CHEMICAL  ANALYSIS OF RECYCLE TEST SLURRYS




                              Calcium as   Carbonate as   Sulfur  as
Solids ,  %   CaO, % Dry     CO.,
Dry    SOV % Dry
Base Lime
Tests 175 & 176
Test 175
Test 176
Base Lime
Tests 183 & 184
Test 183
Test 184
23.9

25.9
24.2
19.6

23.0
20.2
71.1

67.2
69.0
72.4

62.6
65.6
	 _£ 	
0.29

3.42
2.07
0.46

6.52
4.90
0.04

4.05
2.36
0.05

9.96
4.46
                                10-75

-------
     Test Runs 183 and 184 were at the same AST but, due to the higher
stoichiometry, attained only 47% utilization.  The difference in marginal
increase of SO  capture based on increases in wet recycle between  these  two
test series appears to be a function of reagent utilization with no  recycle.
Also, at the same SO  capture (no recycle), as stoichiometry increases  from
1.2 to 1.6, so does Che marginal effect of wet recycle on S02 capture.   For
dry scrubber FGD on high-sulfur coals, wet recycle will be essential to
maintain acceptable reagent and fly ash utilization.

     Table 8 lists chemical analyses of the base and recycle tests shown in
Figure 8 from this program.  Addition of ash to the base slurry is evident by
the increase of total sulfur in forms of calcium sulfates and sulfites  and
carbonates, principally calcium carbonate.

LIMESTONE INJECTION

     Limestone injection in a multistage burner (LIMB) uses limestone (CaCO  )
as reagent introduced into high-thermal-combustion zones, where calcination  to
lime (CaO) occurs.  The solid lime then adsorbs gaseous S02, forming sulfites
and sulfates of calcium (CaSO  and CaSO ).  The CaC03 dissociation reaction  is
principally a function of temperature and  the flue gas C02 partial pressure.
At atmospheric pressure, limestone will begin dissociation under normal  coal
combustion conditions at approximately 1450°F (788°C)  [7].  Once the C02 gas
has evolved, sulfation may proceed.

     Obtaining a high-quality calcined limestone depends primarily on the
temperature profile existing in the combustor and time profiles from the point
of  reagent injection.  Murray  [8] studied  calcination efficiency on  43 commer-
cial  limestones  and concluded each particular limestone has a  unique time-
temperature profile for  optimum calcination.  Dogu  [9] investigated  initial
limestone  pore structure effect on S09 adsorption.  The average lime particle
pore  radius and  overall  porosity were measured  at various  calcining  tempera-
tures,  with both having  an  optimum value  at  1740°F  (950°C).  Dogu's  studies
also found diffusion  resistance from pore  closure by CaSO  the  controlling
mechanism.

      Recent work to  develop  low-NO  coal  burners has helped  limestone injec-
 tion studies,  since  their  characteristics  of  low flame temperature and long
 burn time are quite  suitable  for  limestone injection.  Dremmel, et al.  [10]
 studied three types  of  coal-fired  furnaces as part  of  EPA's simultaneous SO  ,
 N0x study. Wall-,  tangential-, and stoker-fired combustors were  used to   X
 evaluate primarily coal  firing  rate,  cooling  rate,  and sorbent  injection
 velocity.  A  wall-fired  furnace simulating the  normal  time-temperature profile
 seen by coal  and sorbent particles was fired  in a low NO   mode.   At  sorbent-
 sulfur ratios of 2.0, 40%  to   60%  SO   capture was measured.   Quantitative
 effects of time, temperature,  and  calcination rates  were  still undetermined.

      Figure  9 lists  furnace SO  capture  versus  stoichiometry  for  primary and
 secondary limestone  injection  (also  see  Table  9).   At  the  2000°F  injection
 temperature,  primary  injection (limestone interspersed with  coal)  captured
 slightly less S02  than secondary  injection (13.6%  to  13.3%).   Within the
 accuracy of measuring S02,  however,  these removals  are  essentially identical.
                                      10-76

-------
ou
UJ
IT
< 20
O
FURNACE SO2
O
0
D
O






0- xs
/J \/ PRIMARY INJECTION
/AT 2000°F (1093°C)
D SECONDARY INJECTION
AT 1800°F (982°C)
O SECONDARY INJECTION
/v AT 2000°F
I I I I
1234
STOICHIOMETRY
(1093°C)


                     Figure 9.   Limestone injection results.
                                    Table  9
                       TEST RESULTS:  LIMESTONE INJECTION
Injection
Method
Primary




Secondary




DOE Test
Number
D221
D222
D223
D223
D223
D212
D213
D214***
D215
D216
D217
Load
MBtu/hr
4.0
4.0
4.0
4.0
4.0
3.1
2.0
4.0
4.0
4.0
* 0,*
2.9
2.75
2.75
3.0
3.13
3.75
3.73
3.17
3.78
3.79
SR
3.0
2.0
1.0
1.0
1.0
3.0
2.0
3.0
2.0
1.0
Injection
Temp.
2000°F**
2000°F
2000°F
2000°F
2000°F
1800°F
1800°F
2000°F
2000°F
2000°F
% Removal
in Furnace
20.0
13.6
6.7
4.1
6.3
26.0
13.6
22.4
13.3
7.3
  *As measured at reactor  inlet
 **Furnace exit  temperature
***Insufficient  turndown on  screw feeder
                                    10-77

-------
Primary reagent injection exposes the limestone particles to the highest  flame
temperatures and longest residence time possible.  Sintering of the newly
formed CaO particle pores is a major concern with this type of injection  mode.
For secondary injection at varying furnace outlet temperatures, a small in-
crease was also measured in SO  capture as the furnace temperature was
lowered.  This assumes that an equilibrium between decarbonation and  recarbon-
ation had been established.  Some partial pore closure by sintering is also
suspected.  In general, with a low-NO  burner producing 250 -  400 ppm N0x>
approximately 25% SO  was captured atXa 3.0 stoichiometry.  The inserted  lime-
stone was pulverized to approximately 70% through 200 mesh —  equivalent  in
size to the pulverized coal.  A finer limestone grind would have increased
furnace  SO  capture by producing more solid surface area and  minimizing  the
pore blinding effects of CaSO  formation.

               DRY SCRUBBING TECHNOLOGY FOR FGD OF EASTERN COAL

     Successful adaption of dry scrubber FGD systems to high-sulfur-coal
applications requires first an understanding of SO  capture at various
operating conditions meeting federal New Source Performance Standards (NSPS).
Based on nonrecycle operation, these tests indicated 90% SO  capture  can  be
obtained at a dry scrubber stoichiometry of 1.20 (75% utilization) and an
approximate 15°F AST.  At a 1.6 stoichiometry and 20°F AST, 20% solids recycle
increased SO  capture from approximately 75% to 90%.  Maintaining the higher
20°F AST required recycle and increased lime consumption to meet NSPS.
However, the higher AST will reduce the possibility of a system upset.  A dry
scrubber treating 3500 ppm SO  flue gas at a 20°F AST and a 1.6 stoichiometry
requires approximately 25% lime slurry.  Recycle will increase slurry solids
and is  primarily limited by the maximum atomizer solids for acceptable
atomization.  For an industrial dry scrubber system, adjustments between  AST,
recycle rate, and stoichiometry will define actual process conditions required
to meet NSPS  and high availability.
                                     10-78

-------
                                  REFERENCES

 1.   DOE/METC/SP-194,  Topical  Report,  Advanced  Environmental  Control
     Technology.

 2.   E.  A.  Samuel,  et  al.,  "Dry FGD Pilot  Plant Results:  Lime Spray Absorption
     for High-Sulfur Coal  and  Dry Injection of  Sodium Compounds  for Low-Sulfur
     Coal,"  EPRI-CS-2897,  May  1982.

 3.   J.  T. Yeh,  et  al.,  "Experimental  Evaluation of  Spray Dryer  Flue  Gas
     Desulfurization for use with Eastern  U.S.  Coals,"  EPRI-CS-2897,  May  1982.

 4.   E.  L. Parsons, et  al.,  SO  Removal by Dry  FGD,  Buell Emission Control
     Division, Environtech Corp.

 5.   J.  B. Doyle and B.  J.  Jankura,  "Furnace Limestone  Injection with Dry
     Scrubbing of Exhaust  Gases,  Babcock & Wilcox,"  Spring 1982, Central
     States  Section of  the Combustion  Institute Technical Meeting.

 6.   C.  A.  Johnson, Flyash Alkali Technology-Low Cost Flue Gas FGD, Peabody
     Process Systems Inc.

 7.   R.  S.  Boynton, Chemistry  and Technology of Lime and  Limestone, John  Wiley
     and Sons,  1980.

 8.   J.  A.  Murray,  et  al.,  Journal American Ceramics Society. 37, No.  7,
     323-328,  1954.

 9.   T.  Dogu,  "The  Importance  of Pore  Structure and  Diffusion in the  Kinetics
     of  Gas-Solid Noncatalytic Reactions — Reaction of Calcined Limestone
     with SO , Chemical  Engineering Journal, 21, 213-222, 1981.

10.   D.  C. Dremmel, et  al.,  "SO  Control with Limestone in Low-NO Systems:
     Development Status,"   EPRI-CS-2897, May 1982.                X
                                     10-79

-------
EPRI SPRAY DRYER/BAGHOUSE PILOT PLANT STATUS
                 AND RESULTS

          G. M. Blythe, R. G. Rhudy

-------
           EPRI SPRAY DRYER/BAGHOUSE PILOT PLANT STATUS AND RESULTS

                                Gary M. Blythe
                              Radian Corporation
                             Austin, Texas  78766

                               Richard G.  Rhudy
                      Electric Power Research Institute
                         Palo Alto, California  94303
                                   ABSTRACT

     In February 1982, the Electric Power Research Institute (EPRI) initiated
a 2-1/2 MW spray dryer/baghouse FGD pilot plant program at their Arapahoe test
facility.   The objective of the pilot plant program is to confirm the capabil-
ities of the FGD process and to provide the electric utility industry with
reliable design and operating information for spray dryer/baghouse FGD sys-
tems.  The pilot unit was described and initial results for sodium carbonate
and once-through lime operation were presented at the May 1982 FGD symposium
in Hollywood, Florida.  This paper presents the results of test work conducted
from May 1982 through August 1983.

     The majority of the test work has been conducted with lime reagent in the
recycle, rather than once-through mode.  Effects of a number of variables have
been studied.  Spray dryer inlet SC>2 concentrations have been varied from a
nominal 350 ppm up to 2000 ppm.   Other variables examined have included re-
agent ratio,  recycle rate, system flue gas flow rate, atomizer feed slurry
preparation and feeding configurations, and approach to adiabatic saturation
at the dryer outlet.  A significant result has been the observation that re-
cycle operation greatly improves spray dryer operation in addition to im-
proving S02 removal performance.

     The fabric filter has been shown to contribute significantly to overall
system S02 removal, particularly at higher system removal levels (80 percent
and greater).  No bag/fabric-related problems have been observed.  However,
corrosion of mild steel baghouse walls and mild steel caps on bags near the
walls has in some instances been severe.   The corrosion has been largely
attributed to insufficient insulation of baghouse surfaces, and to the fact
that the pilot-scale compartment shares no common walls with other compart-
ments.   As a result of several bag cap failures, the fabric filter compartment
was re-bagged in April 1983.   The new bags were brought on-line with no condi-
tioning by fly ash-only operation, and after 4 months continue to operate at a
very low pressure drop.
                                    10-81

-------
                                 INTRODUCTION

     Currently, spray dryer based flue gas desulfurization (FGD)  systems  for
16 utility boilers with a total capacity of over 6000 MW have been  installed
or are under contract.  Six of the systems are in start-up or commercial
operation.  However, of the two systems that have been in operation longest,
one is unique in using soda ash reagent, while the other has been operated  by
the vendor until only recently.  Thus, in spite of the large number of  units
sold,  very little information on the design and operation of these  systems  is
available to the utility industry.

     Because of this  lack of information, the Electric Power Research  Insti-
tute has  chosen to pilot a spray dryer/fabric filter  system at  their test
facility  adjacent to  the Public Service of Colorado Arapahoe station.   Numer-
ous pilot-scale programs have been conducted by process vendors,  but detailed
results from these programs have generally not been published.  The goals of
the EPRI  program are  to independently assess the operability of a spray dryer
and fabric filter system in FGD service, and to provide design  data applicable
to typical utility  installation.  Emphasis has been placed on  low to medium
sulfur operation, with S02 removal efficiencies from  70 to 90 percent.  Most
test work has been  in the operating mode where sorbent and fly  ash  have been
recycled  from  the fabric filter back to the spray dryer, as is  typical  of many
utility  systems.

      The  pilot unit began operation in March 1982.  This paper  summarizes
 results  from May  1982 through August 1983.  Work during this period has in-
 cluded both  once-through and recycle operation, inlet S02 concentrations  of
 400  to  2000  ppm,  various modes of recycle, normal (270°F*) and  low  (210°F)
 inlet flue  gas temperature, and varied makeup water compositions.   In  the fol-
 lowing  sections,  the  pilot unit  is described, results are presented regarding
 the operability of  the  spray dryer and  fabric filter,  S02 removal results are
 presented,  and the  conclusions  available as of August 1983 are  summarized.

                           DESCRIPTION OF PILOT UNIT

      The spray dryer  unit  at Arapahoe consists of a  spray dryer vessel, one
 compartment of a fabric filter,  a reagent preparation system,  and reagent
 feed/recycle systems, as well  as a comprehensive  instrumentation  system.
 Figure 1 is a  simplified process flow diagram for the pilot unit  as set up  for
 rotary atomization.
     ^    spray dryer vessel has a  10-foot  diameter,  5.5 foot straight side and
 a 50° cone bottom.   The vessel has interchangeable flue gas inlet vane rings
 (one for normal flow rates and one for low flow rates), can accommodate rotary
 or nozzle atomization,  and can be  rearranged for either bottom or side gas
 exit.   An additional straight side section which doubles the volume of the
 dryer vessel can be installed to facilitate nozzle atomization tests.  At
 ^British Engineering Units rather than SI units  are used in this paper because
  of customary usage in the electric power industry.   An appendix provides
  appropriate conversion factors.
                                     10-82

-------
normal flue gas rates  (6500  to  9000 acfm dryer outlet gas rate),  the vessel
flue gas residence time without the added section is approximately  5 to  7
seconds.   Thus, the vessel is quite versatile, allowing a number  of potential
configurations.  This  versatility was desired to provide the capability  to
include aspects of various vendors' designs in the test program.
     TEMPERATURE
      CONTROL
       WATER   I


     FLUE GAS
    FROM BOILER ^T
               FLUE GAS BYPASS AROUND DRYER
                                        CLEAN
                                       FLUE GAS
                                                                DETENTION
                                                                 SLAKER
                                                                NOT SHOWN
                                                       LIME DILUTION
                                                         TANK
           Figure 1.
Simplified Process Flow Diagram for  Spray Dryer
Pilot Plant (Rotary Atomizer  Configuration)
      The fabric filter used for  this  spray  dryer pilot plant is one compart-
 ment of a 10-MW equivalent four-compartment pilot unit which had previously
 been undergoing characterization for  over a year.   The fabric filter is also
 quite versatile in design, allowing a choice of  bag diameters and bag lengths.
 The compartment is currently equipped with  36 utility-size (12 in. by 34 ft)
 glass fiber bags (tri-coat finish) in a  6 x 6 array.   At the normal 6500 to
 9000 acfm fabric filter module inlet  flue gas rate, the air-to-cloth ratio
 varies from 1.7 to 2.3 ft/min.   The unit has provisions for either reverse
 gas, shake/deflate, or combined  cleaning.   Each  module, including the one used
                                       10-83

-------
for the spray dryer pilot unit, is controlled by a microprocessor.  The micro
processor enables full variation in cleaning cycle duration, cleaning  intens-
ity,  null period durations,  damper opening and closing rate, etc.  So  far  in
the program,  the fabric filter has been operated on a 3-hour cycle, using
reverse gas cleaning.

     A recent modification to the pilot unit allows bypassing warm gas around
the spray dryer to reheat the gas entering the fabric filter.  The controls on
this bypass line will allow either a constant reheat temperature  or a  constant
bypass flow.

     The reagent preparation area contains ball mill, paste-type, and  deten-
tion slakers.  Additionally, provisions are available to mix simulated cooling
tower  blowdown waters for use as slaking and/or dilution water.   Lime  slurry
used by  the  spray dryer has primarily been slaked in the paste slaker, al-
though some ball mill- and detention-slaked lime has been used.

     Slurry  is  fed to the atomizer using progressing cavity pumps from one of
two  covered, well agitated tanks.  Temperature control water is added  at  the
atomizer.  The  system has the capability of recycling fly ash/spent sorbent
from the spray  dryer bottoms catch (during side flue gas exit operation)  and/
or from  the  fabric filter hopper each.  Solids from each location can  be  pneu-
matically  transferred to separate recycle bins.  Recycle material can  either
be slurried  and  fed from a  separate recycle tank, or can be slurried in the
main slurry  feed  tank along with fresh lime.  The feed rates of recycle mate-
rial and makeup water to either  tank are ratioed to the rate at which  fresh
 lime slurry  is  added to the slurry feed tank.  A separate pump and controller
are  available  to  meter recycle material to the atomizer separate  from  the  lime
 slurry if  the  two are not slurried together in the same tank.

     The instrumentation on the  pilot unit was discussed in detail in  a pre-
vious  paper,  "EPRI Spray Dryer Pilot Plant Status and Results,"1  given at  the
EPA/EPRI FGD Symposium in Hollywood, Florida,  in May 1982.

                               OPERATION RESULTS

     The EPRI  pilot unit has provided insight  into the operation  of the equip-
ment used in spray dryer/baghouse systems.  Many improvements in  pumps, tanks,
 agitators,  etc.,  have been made  since start-up to improve the operation of the
 pilot  unit.   However, most  of  these changes were necessary because of  the
 small  scale  of  this pilot plant  and are not directly applicable  to utility
 systems.  However, observations  about the operation and maintenance of the
 spray  dryer  and fabric filter  are more generic in nature and may  be applicable
 to most  spray  dryer/baghouse FGD systems.  The spray dryer and fabric  filter
 operation results are discussed  in the following section.

 SPRAY  DRYER  OPERATION

     Operation of the  system in  the once-through mode, with lime  reagent,  for
 low  inlet S02  levels  (300 to 400 ppm) has been the most troublesome  configura-
 tion tested  to  date.  After an initial check-out of the system with  soda  ash
reagent  and  a  spinning disc atomizer wheel,  the  first  lime  slurry tests were
                                      10-84

-------
conducted in the once-through mode.  The results of soda ash testing and early
once-through lime tests were reported in the previously referenced FGD sym-
posium paper.  The problems encountered after switching to lime slurry were
multiple.  First, the pilot spray dryer had been designed for a flue gas flow
rate that resulted in a residence time of approximately 5 seconds.  When try-
ing to operate at a 20°F approach to adiabatic saturation at the dryer outlet,
as many utility  systems sold to date will, a continued buildup of wet solids
on the dryer walls was encountered.  This buildup caused several problems.
First, wet solids from the walls would intermittently fall to the bottom of
the dryer.  The  dryer is equipped for pneumatic conveying of solids from the
bottom of the dryer, but these wet solids (10 percent moisture and greater)
tended to plug the pneumatic lines.  Once the pneumatic lines plugged, in
spite of efforts to quickly clear the pluggage, wet solids would begin to
bridge across the cone bottom on the dryer.   This bridge of wet solids would
continue to grow until the flue gas exit duct from the dryer was almost
plugged.  At this point, the dryer would have to be shut down,  the rotary
valve and pneumatic line removed from the bottom of the cone, and the vessel
would have to be washed out.  The problems of drop out of wet solids on the
walls of the dryer were substantially reduced, however, when operating a lower
flue gas flow rate (7 to 8 second residence time) or at a 30°F or higher
approach to saturation at the dryer outlet.

     The second major problem during the once-through lime tests was related
to the atomizer  rather than the vessel itself.  For lime reagent tests,  a
nozzle insert-type wheel normally used for abrasive slurry service has been
used.  During the first several weeks of operation, the spray machine was
taken off-line several times due to high vibration.  Invariably, when the unit
was removed for  inspection, it was found that several of the twelve nozzles
around the periphery of the wheel were plugged with lime solids.  The pluggage
problem was lessened but not solved by increasing the wheel speed from 12,000
to 15,000 rpm.    It became apparent that the pluggage of nozzles with lime
solids was caused by flue gas pumping through the wheel.   The 9-inch diameter
wheel with twelve 3/8-inch openings had a hydraulic capacity much greater than
the nominal 2 gallons per minute of slurry normally fed at these flue gas
rates.  Because  it had a greatly increased hydraulic capacity,  it would educt
hot flue gas across the top of the wheel to be pumped through the nozzles.
Because of this flue gas pumping, lime and/or fly ash solids were deposited on
the top side of the wheel,  and lime solids tended to dry within the wheel,
particularly within nozzle openings.   An attempt was made to fit a seal
between the spray machine and the wheel in order to restrict this pumping
effect.   However, the pumping continued,  although perhaps at a reduced rate.
As flue gas was pumped through this low clearance area, lime and/or fly ash
deposited on the top of the wheel caused the seal on the bottom of the spray
machine to rub the wheel.   The resulting imbalances led to early failure of
the lower bearing on the machine.  An alternate approach,  then, was to reduce
the number of active nozzles in the wheel from twelve to six.  The other six
nozzles were replaced with solid inserts.  Although the tendency for the
nozzles to plug was reduced, on numerous occasions, three of the six nozzles
would plug.   If any two adjacent nozzles plugged, an imbalance would lead to
high vibration measurements and cause the operators to shut the spray machine
down.   However,  if every other of the six nozzles plugged, vibration problems
were not encountered.   this led to a further reduction in active nozzles to
                                      10-85

-------
only three,  where now nine of the original twelve nozzles were replaced with
solid inserts.   While this last modification solved the nozzle pluggage prob-
lem, it tended to aggrevate the previously discussed problem with wet product
on the dryer walls.   Empirical correlations in Masters2 for predicting the
droplet size produced by a vaned atomizer wheel suggest that for a nozzle
wheel with a given nozzle diameter,  increasing the number of nozzles around
the periphery of the wheel will produce a finer droplet size.  Based on this
correlation, it would seem that in reducing the number of active nozzles in
our atomizer wheel,  we produce a coarser droplet size which is not as readily
dried.  This no doubt contributes to the buildup of wet product on the spray
dryer chamber walls.

     After once-through operation, lime reagent recycle tests were begun.  In
most recycle tests conducted thus far in the program, recycle material has
been taken from the fabric filter hopper, and mixed with fresh lime and makeup
water  in  the main slurry feed tank.   This fresh lime/recycle slurry is nor-
mally  fed to the atomizer by a single pump.   Recycle ratios are expressed in
this paper as pounds of recycle material per pound of hydrated lime (CaCOH^)
solids  in the fresh lime feed.  The lowest recycle ratios tested were around
2:1, but  even at this level of recycle an immediate improvement in spray dryer
operation resulted.   With recycle, the number of active nozzles in the atom-
izer wheel could be increased back to six from the three required for once-
through operation.  This increase in active nozzles should result in somewhat
finer  atomization, further promoting improved drying.

     Upon the implementation of recycle, problems with buildup of solids
within  the atomizer wheel, buildup of wet solids on the spray dryer chamber
walls,  and plugging of the pneumatic transfer line from the spray dryer bottom
have also been virtually eliminated.  The dryer can be successfully operated
over a  long term at a 20°F approach, at a flue gas flow rate resulting in a
7-second  dryer residence time.  At recycle ratios near 12:1, resulting in a
solids  level at the atomizer wheel of 35 to 45 percent, the dryer has been
successfully operated for days at a time at a 20°F approach at residence times
as  low  as 5 seconds.

     It appears that much of the improvement in spray dryer operability in
recycle operation is related to having an increased weight percent solids
level  in  the slurry fed to the atomizer.  This effect has become apparent in
recent  once-through lime tests conducted at 1000 ppm inlet SC>2 levels.  At
1000 ppm  inlet levels, roughly three times the amount of lime is fed compared
to  low  sulfur operation, at equivalent reagent ratios.   This solids level,
then,  is  equivalent to the solids level at the same reagent ratio for 350 ppm
inlet  SC>2 and 2:1 recycle ratio operation.

      In the baseline 2:1 recycle tests, the weight percent solids in the
slurry  atomized ranged from 6 to 16 percent.  Similarly for the 1000 ppm inlet
S02 once-through lime tests, the weight percent solids level to the wheel
varied  from 12 to 16 percent.  In these once-through tests at a nominal
7-second  dryer residence time, problems of wheel pluggage with the six nozzle
atomizer wheel, and pluggage of the drying chamber with wet solids were not
encountered.  These successful once-through lime tests at a 1000 ppm inlet  S02
                                     10-86

-------
level seem to indicate that at least part of the improved spray dryer oper-
ability during recycle operation comes from the increased solids level in the
slurry atomized rather than from the presence of the recycled material per se.

     Further evidence of this weight percent solids effect on operability is
seen for the high recycle rate tests.  For the nominal 12:1 recycle tests the
percent solids to the wheel ranged from 31 to 37 percent.  Operability of the
spray dryer was improved over that even for nominal 6:1 recycle ratios.   At an
inlet flue gas flow rate of 5800 scfm (5.0-second residence time in dryer)
when recycling to a nominal 35 weight percent solids level to the wheel,  no
tendency to drop out wet solids in the dryer or plugging of pneumatic transfer
lines was noted.   However, some tendencies for wet product formation had been
seen when operating at this flue gas rate at lower recycle ratios (hence lower
weight percent solids to the wheel).

     The effects of operating condition on the weight percent solids in the
slurry fed to the atomizer are summarized in Table 1.  The table shows that
once-through operation at low inlet SC>2 concentration results in the lowest
solids content in the slurry fed to the atomizer.   It was at these conditions
that problems with wet spray dryer products persisted.   Other operating modes
resulting in higher solids content in the slurry feed (recycle,  and/or higher
inlet S02 concentration) greatly improve the spray dryer operation.

  TABLE 1.   SUMMARY OF WEIGHT PERCENT SOLIDS IN SLURRY FED TO ATOMIZER WHEEL
            AT VARIOUS OPERATING CONDITIONS


Operating Mode
Once-through
Recycle (2:1)
Recycle (6:1 to £
Recycle (12:1)
Once-through
Recycle (2:1)
Recycle (2.5:1)*
Nominal
Inlet
S02 (ppm)
350
350
5:1) 350
350
1000
1000
1000
Weight
Percent Solids
Atomizer Wheel
3 to 7
6 to 16
17 to 24
31 to 37
12 to 16
20 to 34
25*
Vessel
Residence
Time (sec)
7
7
7
5
7
7
5
Nature
of S.D.
Product
Wet
Dry
Dry
Dry
Dry
Dry
Dry
*0ne test.

     Theoretical considerations presented by Masters^ would tend to support
these observations of improved operability at higher feed solids contents.
Masters describes the first two phases of drying of droplets in a spray dryer
as the "constant rate" period (the first phase of drying) and the "first
falling rate" period (the second phase of drying).   In the first phase, or
constant rate period, the surface of the droplet behaves as though no solids
are contained, and water evaporates freely from the surface.  In the first
falling rate period (second phase of drying), solids protrude from the surface
of the droplet,  and the rate of evaporation of moisture from the droplet  is
slowed by the rate at which water diffuses through the network of solid par-
ticles.
                                     10-87

-------
     It is important that droplets atomized are beyond the constant rate, or
first drying period before they come into contact with vessel walls.  The
length of the first drying period is labeled as the critical drying time.   In
general, problems with wet product drop out in the dryer can be avoided by
decreasing this critical drying time for the droplets atomized.  Empirical
correlations presented in Masters allow one to predict the impact of variables
such as solids content on critical drying time.  The results of example cal-
culations are summarized in Table 2.  These example results show that for a 50-
micron droplet, the critical drying time is reduced by one-half if the solids
content is raised from 5 to 30 percent.  Calculations for 100-micron droplets
show that critical drying times are 3 to 4 times those for the 50-micron drop-
lets, but that the higher solids level again nearly reduces the critical dry-
ing time by one half.   Of course a higher solids content in the slurry feed to
an atomizer will also impact the atomized slurry particle size distribution to
a minor degree, but it is clear from these example calculations that the net
effect of increased solids content in the slurry feed is reduced critical dry-
ing time, hence potentially improved dryer operation.  Also in Table 2 are
falling rate drying times and total drying times for each of the four cases.
The falling rate drying times and total drying times actually increase at
higher weight percent solids levels.  This increase may account for some of
the  improved SC>2 removal performance seen in recycle operation.  Please note
that the values presented in Table 2 were calculated for example only, and do
not  represent an attempt to accurately predict the critical and/or total dry-
ing  times in the Arapahoe spray dryer.

             TABLE 2.  EXAMPLE DRYING TIMES FOR ATOMIZED DROPLETS

                                   Case 1     Case 2     Case 3     Case 4
Droplet Size, y
Weight Percent Solids
Critical Drying Time, sec
Falling Rate Drying Time, sec
Total Drying Time, sec
50
5
0.4
0.1
0.5
50
30
0. 2
0.8
1.0
100
5
1.4
0.4
1.8
100
30
0.8
2.0
2.8
 BAGHOUSE OPERATION

      Throughout  over  6000 hours of spray dryer operation, no bag fabric prob-
 lems have been observed.  This result is significant, because the operating
 time includes both  sodium carbonate and lime reagent, once-through and recycle
 operation,  low sulfur  to medium sulfur inlet S02 levels, a great number of
 start-ups and shutdowns, and  several spray dryer upset conditions.   Spray
 dryer upsets which  one would  have expected to have a detrimental impact on bag
 performance did  not seem to affect the bags at all.  In one incident, a new
 operator flushed a  slurry feed line into the spray dryer through the atomizer,
 driving the dryer outlet temperature to the adiabatic saturation temperature.
 In spite of such abuse, no bag failures have occurred, outlet opacity remains
 quite low,  and tube sheet pressure drop values have been quite acceptable.

      Although there have been no bag fabric problems due to operation down-
 stream of the spray dryer per se, there has been a problem with corrosion on
                                     10-88

-------
the baghouse walls.  The problem did not become apparent until after the spray
dryer had been in operation for approximately 8 months, in November 1982.
During an outage, the baghouse module was opened and large flakes of rust
scale were noted to be forming along the inside walls.  The corrosion was
attributed to moisture condensation on compartment walls.   Several factors may
be contributing to this condensation and corrosion problem.  First, the com-
partment is part of a pilot unit, so it goes through numerous start-ups and
shutdowns, passing through the dewpoint of the flue gas each time.  In addi-
tion, since the compartment shares no common walls with other compartments,
all four walls of the compartment are exposed to ambient temperatures.   Also,
the baghouse was originally designed to clean hot gas (250°F to 300°F)  rather
than gas which is approaching adiabatic saturation.  Consequently, the bag-
house walls are insulated with only a 2-inch thickness of fiberglass batts.
At the spray dryer outlet, the flue gas is generally 17 to 22°F from adiabatic
saturation.  This corresponds to roughly 20 to 30°F from the dewpoint.   As the
flue gas is cooled in the outlet ductwork and in the baghouse, the bulk gas
temperature may come within 10 to 15° of its dewpoint.  Any signficant cooling
of the baghouse walls, then, may put the skin temperature at or below the
dewpoint of the flue gas.  This effect would particularly be noticeable in
very cold weather, approaching 0°F, where the driving force for cooling the
walls would exceed 100°F.  The explanation that the corrosion is caused by
localized cold spots is supported by the observation that corrosion is  worst
near an entry door to the compartment, where not only is the wall insulation
sparse, but where cold air can leak into the compartment.

     Sulfuric acid condensation is not thought to directly contribute to the
corrosion problem, because no measurable 803 content has been found in the
flue gas downstream of the filter bags.  However, S02 present in the outlet
flue gas would be sorbed into any condensation which forms.  Subsequent liquid
phase oxidation reactions would tend to form sulfuric acid in the condensed
moisture.  Sulfuric acid-laden moisture then would not readily evaporate be-
cause of the azeotropic nature of the sulfuric acid-water mixtures.

     A later outage, after approximately 10 months of operation, showed con-
tinued corrosion.  Additionally,  during this outage the top side of the tube
sheet was found to be covered to a depth of several inches with wet ash and
flakes of corrosion.  The moisture on the tube sheet was attributed to the
severe spray dryer upsets mentioned previously,  where an inexperienced opera-
tor repeatedly overloaded the atomizer with water while flushing a line.  The
ash on the tube sheet had been observed during a previous outage.  It is not
clear when this dry ash was deposited.  The top of the tube sheet was cleaned
and the unit operators were reminded of proper line flushing procedures which
call for blocking out flow to the atomizer.

     During this outage,  thermocouples were installed on two wall  surfaces
within the fabric filter compartment and continuously monitored.  In colder
weather,  it was observed that the temperatures were generally several degrees
below the bulk gas temperature.   In April 1983,  the ambient temperature over-
night in Denver dropped to 9°F.   One of the two wall temperatures dropped  to
96°F for over an hour.   This is several degrees below the dewpoint of the  bulk
gas at the fabric filter outlet.   The pilot unit was shut down and the com-
partment was opened.   As expected,  moisture was found on compartment walls and
                                     10-89

-------
on top of the tube sheet.   More startling was the discovery of three bags
collapsed onto the tube sheet.   The bags had fallen due to corrosion failure
of the mild steel bag caps.   These bag cap failures were not thought to be the
result of this one low temperature excursion, but a cumulative effect of 13
months of operation downstream of the spray dryer and a particularly cold
winter.

     The bags were all removed and replaced, and the removed bags were exam-
ined.  A map of bag cap corrosion was prepared, and is presented in Figure 2.
In general, the caps showing the worst corrosion were around the edges.  Of
the  three failed bags, two were in corners and one was near the compartment
doors.  Top and bottom portions from twenty of the bags were submitted to
Albany International for routine performance evaluation.  No failures or areas
of potential failure were observed on any of the submitted samples except a
top  sample from one of the fallen bags.   This top sample contained two small
holes  in an area which was heavily stained with rust.   The damage likely
occurred when the bag fell.


FLUE GAS
INLET ""^


• O
® O
® o

(X O
0 0
o
o
0

o
o
® ft <»
O O 
-------
                              FLUE GAS
                                OUT
        FLUE
         GAS
          IN
Figure 3.  Cutaway View of Thermocouple Locations
           in Baghouse Compartment
                     10-91

-------
filter enclosure,  while the west and north walls face the inside of the en-
closure.   Table 3  summarizes selected temperature measurements since the pilot
unit was put back  in service.   Case 3 in the table includes the two wall mea-
surements previously recorded for a 9°F ambient temperature.  Note that in
Case 1 and Case 2,  no temperatures are actually below the bulk gas dewpoint,
but for Case 2 (36°F ambient) one of the wall temperatures is approaching that
value.  Also note  that the 15 temperature which reached 96°F in Case 3 is not
even the coolest temperature in the other two cases.   Monitoring of all ten
locations during the upcoming winter is expected to identify a number of ex-
cursions of wall and cap temperatures below the bulk gas dewpoint.

        TABLE 3.  AMBIENT TEMPERATURE EFFECTS ON BAGHOUSE TEMPERATURES
Temperature (°

Ambient
S.D. Outlet
Baghouse Outlet
Approximate Dewpoint




(F.F. Outlet)
TI (Top Interior Wall)
15 (Bottom Interior
Tg (Bottom Exterior
Tg (Bag 16-Cap)
TIQ (Bag 36-Cap)
Wall)
Wall)


Case 1
82
138
132
108
117
125
125
136
133
Case 2
36
135
124
104
107
113
117
130
124
F)
Case 3
9
137
125
104
-
96
109
-
—
     These results certainly indicate that full-scale baghouses downstream of
 spray dryers will have to be designed with ample insulation on exterior walls,
 and should be insulated between compartments to avoid cold walls when adjacent
 compartments are off-line.   Also,  compartment doors must be well sealed, and
 penetrations of supports through the insulation and welding directly to com-
 partment walls should be avoided.

     Replacement of the bags in the compartment downstream of the spray dryer
 has allowed a number of "before" and "after" comparisons.   Since this compart-
 ment is one of four in the same baghouse,  a control compartment which treats
 the same flue gas without the spray dryer upstream is also available for com-
 parison.

     The first comparison (Figure 4) shows the pressure drop across Compart-
 ment A  (downstream of the spray dryer) compared to that for Compartment C
 (control, fly ash only) during the month of January 1983.   In this plot, the
 original bags were still in service downstream of the spray dryer.  The lower
 pressure drop across Compartment A is expected considering Compartment A was
 operated at a lower filtering air-to-cloth ratio of 1.7 ft/minute compared to
 2.0 ft/minute in Compartment C.   This result is somewhat surprising though,
 considering that Compartment A is removing 2 to 3 times the quantity of partic-
 ulate matter downstream of the spray dryer than Compartment C collects from
 the untreated flue gas.
                                     10-92

-------
* 10
i 9

i' •
a
ui 7
oc

  6
  4
  3
  2
  1
  0
           I   I  I   I  I   I  I   I  I   I  I   I  I   I

                          COMPARTMENT C
                                       COMPARTMENT Ax
         3  4  5  6  7  8  9  10 11  12 13 14 15  16 17 18 19 20 21  22 23  24 25  26 27  28 29  30 31
                                      JANUARY 1B83

            Figure 4.   Average  Tube Sheet Pressure Drop History  -
                        Compartments A and C - January 1983

     The bags in Compartment A  were originally filtering fly ash only  for
approximately 5  to 6 months prior  to start-up of the spray dryer (July 1981  -
January 1982), thus establishing a residual fly ash cake on the  bags.   When
new bags were placed  in service after the corrosion problem in April  1983, the
spray dryer was  started up  immediately;  no conditioning on fly ash-only was
conducted.   In Figure  5 a pressure drop  comparison of Compartments A  and C,  4
months after the new bags were  installed in Compartment A, for July 1983,
shows a substantially  reduced pressure  drop across Compartment A relative  to
that of Compartment C.   When the difference in A/C ratios is accounted for,
the Compartment A  Ap  is still significantly lower than that for  Compartment  C.
The difference is explained by  comparing the period of operation between the
two compartments.  The Compartment C bags have been in operation for  over  two
years and have reached a steady-state pressure drop and residual dust  cake
weight (50-60 Ibs).  The Compartment A bags, after operating only 4 months,
have a very low pressure drop and  only  about 15 Ib of residual dust cake.
Whether the new Compartment A bags will  eventually reach a Ap and residual
cake weight similar to Compartment C remains to be seen.   However, short-term
operation without fly  ash conditioning  has not proven detrimental.  In fact,
the AP across Compartment A has risen very slowly and does not appear  to be
substantially increasing with time.   This effect is shown in Figure 6.
       i—i—i—i—i—i—i—i—i—i—i—r~ir
1C
D
Uj a)
X "g
W c
UJ ~.
co a.
D O
I- or
uj O
EC
UJ
                         COMPARTMENTC
                         COMPARTMENT A
          I  I   I  I  I  I   I  I  I   I  I  I
                                         I
                                              I  I   I  I
J_
I  I   I  I
        1  2  34  5  67  8  9  10 11 12 13 14 15  16 17 18 19 20 21 22 23 24 25 26 27 28  29 30 31
                                      JULY, 1983
            Figure 5.
                       Average Tube Sheet Pressure  Drop  History
                       Compartments A and C - July  1983
                                      10-93

-------
                           1.7:1 A/C Ratio, April through July 1983
      Q O
          4-
      _
      O o
      M- C
      o-
      c
      (0
      CD
                                        2.3:1 A/C Ratio
                                   D
                                                            a a
QDDD°
      DaaOQD D
DDDaD       0
QDU  '-'--'-'  Q
         D aO

                                                     D
                                                       a
                                                    a
                             30                 60
                                   Days since Startup
                                                                90
    Figure  6.
  Pressure Drop Profile of a Recent Start-Up  of  Compartment A
  Downstream of Spray Dryer
     Several important findings have resulted from the Arapahoe pilot  baghouse
operation downstream of the spray dryer.   These include:

     1)   Operation downstream of a spray dryer does not appear to  have
          an adverse affect on the bag fabric, even when subjected  to
          occasional upset conditions at or very near adiabatic satura-
          tion;

     2)   A baghouse downstream of a spray dryer operated  at  close
          approach to adiabatic saturation must be well insulated,  with
          close attention paid to installation details in  order to  avoid
          cold spots which lead to localized  corrosion; and

     3)   A baghouse downstream of a spray dryer may operate  at substan-
          tially lower pressure drop than the same baghouse  treating
          normal fly ash-laden flue gas, particularly if the  bags are  not
          conditioned by any lengthy period of fly ash-only  operation.

                             S02 REMOVAL RESULTS

     The  S02 removal results presented here concentrate on the recycle lime
 reagent operating mode.  This mode is typical of the majority of  utility spray
 dryer/fabric filter dry FGD systems  sold to date,  and represents  the most
 operable  configuration for this pilot unit.   The recycle data presented here
 all  represent  recycling solids from  the  fabric filter.
                                      10-94

-------
     In most of the data plots presented  in this section,  S02 removals across
the overall system, across the spray dryer alone, and across the  fabric  filter
are plotted for each test.  S02 removal across the fabric  filter  in  these
plots is calculated by the following equation:

S02 Removal in = (SC>2 Spray dryer outlet) - (SC>2 Fabric  filter outlet) x IQQ%
Fabric Filter                    (SC>2 Spray dryer inlet)

By calculating SC>2 removal across the fabric filter in this manner,  the  SC>2
removal across the spray dryer and across the fabric filter can be directly
summed to get the overall system removal.  The SC>2 removal values for the
fabric filter presented here represent time averaged values, which have  a
number of 3-hour cleaning cycles included in the averaging time.   This time
averaged value is thought to coincide well with the instantaneous values one
would encounter across a multicompartment baghouse, where perhaps ten compart-
ments would be evenly distributed with respect to time throughout their  clean-
ing cycles.

     Each data point for the recycle tests represents the mean value for
approximately 24 to 48 hours of steady-state operation at those conditions.
Steady-state operation is normally achieved in approximately 24 hours.  For
the system to be at steady state, the bags in Compartment A must have been
cleaned,  the Compartment A hopper emptied, the recycle material holding bin
refilled, and the combined lime/recycle slurry feed tank refilled at least two
to three times each.   When steady state is achieved, the chemical composition
of the slurry fed to the atomizer and of  the material collected on the bags in
Compartment A no longer change significantly from batch to batch.

     Baseline SC>2 removal results for recycle operation at normal Arapahoe
inlet S02 concentrations around 350 to 400 ppm are presented in Figure 7.
These results represent an approach to adiabatic saturation at the dryer out-
let within 20°F, and a recycle ratio of 2:1.   The recycle ratio was  defined
earlier as:

                       Ib of material recycle (dry basis)
     Recycle ratio = ^ of fresh Ca(OH)2 makeup (dry basis)


     The results in Figure 7 show that at low reagent ratio values (below  0.8)
fabric filter SC-2 removal contributes little to overall SC>2 removal, because
lime reagent utilization in the spray dryer approaches 100 percent.  At  higher
reagent ratio values (1.3 and greater) fabric filter SC>2 removal  contributes
greatly to system S(>2 removal,  because S02 removal in the spray dryer does not
increase  with any significance as the reagent ratio is increased.   In this
paper,  unless otherwise noted,  reagent ratio is defined as:

                        Ib-mol fresh lime fed to spray dryer	
     Reagent Ratio = ib-mol S02 in inlet flue gas to spray dryer
                                    10-95

-------


80-

^ 60-
03
O
E
03
oc
C\J
0 40-
C/J





20-



0-
/ .-o--°
I/
Cp __ 	 3
i n x id
^
! LEGEND
/
1 O Overall SO2 Removal
/ n Spray Dryer Removal
/ A Fabric Filter Removal

/
/
/
/
/ f__

/ ^A'"'"
/ ^^
1 A'
/
i i
0 0.5 1.0 1.5 2.
                                  Reagent Ratio

            Figure 7.   Baseline  S02  Removal for 400 ppm Inlet
                       20°  Approach,   2:1  Recycle Ratio
     Once baseline performance curves for moderate recycle rates were deter-
mined a series of tests at higher recycle rates was conducted.  These results
are plotted in Figure 8.   Recycle at ratios of approximately 6:1 to 8:1 did
not appear to demonstrate any marked improvement in SC>2 removal over that for
recycle ratios around 2:1.   Recycle ratios of approximately 12:1 did show a
marked improvement,  primarily in SC>2 removal by the spray dryer.  These
highest recycle ratios correspond to solid contents of 30 percent or better in
the slurry fed to the atomizer.   The fabric filter contribution to overall
removal does not appear to be affected by the increased recycle ratio.  The
upper curves in Figure 8 represent a visual fit of the SC>2 removal data across
the spray dryer during 12:1 recycle operation.  These upper curves show virt-
ually complete lime utilization up to a reagent ratio of 0.9.
                                     10-96

-------
                  100-
                   80-
               -  60-
               co
               o
               E
               01
               cc
                   20-
                      Nominal
                     Recycle Ratio
                      6:1 12:1
                       O  •
                       D  •
                       A  A
                                             Overall SO2 Removal
                                             Spray Dryer Removal
                                             Fabric Filter Removal
                            0.5
  i       i
 1.0     1.5
Reagent Ratio
                                    2.0
Figure 8.
                          S0?  Removal  Results for Recycle Tests
                          20 F Approach Temperature
     Mention was made  earlier  in this  paper that  operation of the spray
dryer/fabric filter FGD  system is improved  by recycle  versus  once-through lime
operation.  Both spray dryer operability  and S0«  removal  improvements  have
been noted.  The results  of once-through  lime tests  are  compared  to  the 2:1
recycle, 20°F approach,  350 to 400 ppm inlet S0_  level baseline  test results
in Figure 9.  Because  of  the repeated  operability problems during once-through
operation, many once-through S0_ removal  test results  have been  discarded.
The results in Figure  9  represent a number  of successful  short-term  once-
through tests.  The open  symbols represent  once-through  results  at a 20°F
approach to saturation at normal Arapahoe inlet S0«  levels.   The  dashed curves
are the visual best fit  of the 2:1 recycle,  20°F  approach baseline results.
The figure shows that  at  low reagent ratio  (below 1.0),  the difference in S0«
removal performance between once-through  and recycle operation is difficult  to
quantify because of scatter in the results  of only two successful short-term
tests.  The results at higher  reagent  ratios (nominally  1.5)  show that once-
through operation gives approximately  10  to  15 percent lower  S0_  removal than
for recycle operation.  This difference is  primarily due  to a decrease in
spray dryer rather than fabric  filter  SO- removal performance during once-
through operation.  It may have  been that more successful once-through tests
were completed at higher reagent  ratios because of the beneficial impacts on
spray dryer operability of higher weight percent  solids  in the lime  feed for
these tests.
                                     10-97

-------
                   100-
                    80-
                    60-
                  0
                  E
                  CD
                 o
                 CO
                    40-
                    20-
      LEGEND
•Overall SO2 Removal
• Spray Dryer Removal
A. Fabric Filter Removal
                             0.5      1.0      1.5
                                 Reagent Ratio
               2.0
        Figure 9.  Comparison of Once-Through S0~ Removal Points
                   with Best Fit Curves from Recycle Results
                   for 400 ppm Inlet SO  Levels

     With the performance of the spray dryer/fabric filter system well charac-
terized for a nominal 350 to 400 ppm inlet S0« level, an extensive series of
tests at a 1000 ppm inlet SO  level was begun.  Testing at this inlet SO-
level was of interest because it simulates performance of the system on medium
sulfur coals (1 to 2 percent sulfur).  Also, it was felt that operation at
higher inlet SO  levels would make subtle variable effects on SO- removal per-
formance more pronounced.

     Figure 10 summarizes the results of a number of tests conducted at a 1000
ppm inlet SO  level, with a 2:1 to 4:1 recycle ratio and a 20°F approach to
adiabatic saturation.  Also plotted is a "best fit" line for the 400 ppm
results at similar conditions.  The majority of the test results are in the 80
to 90 percent S02 removal range, which is what current utility boiler New
Source Performance Standards would require for most medium sulfur content
coals.  The figure shows that at moderate overall SO  removal levels (70 to 85
percent), somewhat lower S02 removals result compared to what would result at
an equivalent reagent ratio in the 400 ppm tests.  At around 90 percent S0?
removal, the overall SO  removal curves tend to intersect, with a reagent
ratio of approximately 1.3 required to achieve 90 percent removal in either
case.  The shape of the overall performance curves in each case appears to be
set by spray dryer removal performance.  The fabric filter contribution data
                                      10-98

-------
appear to plot on one line regardless of inlet SC>2 concentration.  The  inter-
section of the two spray dryer SCL removal curves appears to be the result of
approaching an observed maximum for SO  removal across this particular  spray
dryer.  It is not obvious whether this observed maximum, at around 78 to 80
percent, is the result of limitations on the effectiveness of flue gas  SO^ and
slurry contact in this particular vessel, or is perhaps dependent on some
other parameter, such as atomized slurry particle size distribution.
               100-
                80-
              o
              E
              0)
             O
             05
                40-
                20-
                       LEGEND
                 O Overall SO2 Removal
                 D Spray Dryer Removal
                 A Fabric Filter Removal
    0.5      1.0      1.5
          Reagent Ratio
                                                 2.0
         Figure  10.
Comparison of 1000 ppm Inlet S02 Recycle
Results with Best Fit of 400 ppm
Inlet S09 2:1 Recycle Test Results
     Once-through tests at  1000  ppm inlet  S02  levels  were  also  conducted.
 These results are compared  to  the  1000  ppm recycle  results in Figure 11.   Con-
 trary to the 400 ppm  inlet  S02 test results, the  difference between once-
 through and recycle tests at TOGO  ppm inlet S02 levels  is  less  dramatic.   For
 a  given reagent ratio,  SO   removal in once-through  operation is generally
 within 10 percent of  removal in  recycle operation.  A reason for this obser-
 vation may be the greatly increased weight percent  solids  to the wheel in  once-
 through operation when  the  inlet S02 levels are increased  by a  factor of  2-1/2
 to 3.  For example, once-through operation at  400 ppm or lower  inlet S02  re-
 sults in a lime weight  percent of  only  3 to 7  percent at the atomizer.  Simi-
 lar reagent ratios for  once-through operation  at  1000 ppm inlet S02 levels
 result in weight percent solids  levels  of  10 to 16  percent.  These solids
 levels are the same as  solids  levels at 400 ppm inlet S02 for 2:1 recycle
 operation.
                                      10-99

-------
               100-
                80-
             (0  60-
             o
             E
             o>
             QC
             Q'  40-
             crt
                20-
                     LEGEND
                 Once-
                 Through
                   •   Overall SOz Removal
                   •   Spray Dryer Removal
                   A   Fabric Filter Removal
0.5
 I
1.0
 i
1.5
                                               2.0
                             Reagent Ratio
            Figure  11.  Comparison of Recycle  Versus  Once-Through
                        Operation Results  for  1000  ppm Inlet S02

     Improved once-through S02 removal performance  at  higher inlet SO  levels
 is further  seen in  Figure 12, where once-through  SO  removal results at both
 inlet S02 levels are plotted.  In this figure,  it can  be  seen that except for
 one 400 ppm test at low reagent ratio, all of  the 1000 ppm test  results for
 overall and spray dryer S02 removal plot on a  curve above  that  for the 400 ppm
 test results.  The  one 400 ppm inlet S0? test  result  at close to 100 percent
 utilization at an overall removal level"of 75  to 80 percent is  suspected to be
 .invalid.  Current heat balance analyses of all  test results should serve to
 either substantiate the results of this test or indicate that the point should
 be repeated.

     A number of tests were conducted at low inlet  temperatures  of 210°F to
 220°F.  At Arapahoe barometric pressure conditions, this results in a spray
 down temperature of only 100°F to 110°F.  For normal inlet  temperatures of
 260 F to 280°F, the spray down temperature is on the order  of 145°F to 165°F.
 The 400 Ppm inlet S02 tests in recycle operation are of particular interest
 at reduced inlet temperature.  It is for low sulfur content coal,  resulting
 in FGD inlet S02 levels of 400 ppm and less, that low  inlet temperatures are
    ^ ^™be encountered-   F°ur tests were conducted at low inlet  temperatures
 in the 400 ppm,  recycle mode.  These test results are  plotted with the 400
ppm,  2:1 recycle,  normal inlet temperature baseline results in Figure 13.  The
four  low inlet temperature tests  are at a wide range of reagent  ratios and
show  a good  bit  of scatter.   However,  they show that SO, removal performance
is not  substantially reduced  at low inlet temperatures/ In fact,  one point
                                     10-100

-------
 demonstrates  overall SO^ removal in excess of 90 percent at a reagent ratio
 between  1.2 and 1.3.
            100-
o
E

-------
100-
80-
:£ BO-
'S
o
E
cc
0 40-
20
/ ^
i''
9^ n^---13
tif "
®/ 	
, LEGEND
• Normal 215°F
/ Inlet Inlet
/ Temp. Temp.
1 O 9 Overall SO2 Removal
/ D • Spray Dryer Removal
1 A A Fabric Filter Removal
' ^
0 0.5 1.0 1.5 2.
                                 Reagent Ratio

       Figure 13.   Effects  of  Reduced  Inlet Temperature  on  SO^  Removal
                   for  400  ppm Inlet SO., Recycle Operation

filter after the new bags were installed.  However,  the  results in Figure 14
do not support this speculation.   These results  compare  1000  ppm inlet SO-,
recycle test results with the  old  bags to similar tests  after the new bags
were put in service. The open symbols indicate  the  previous  results, while
the darkened symbols represent results with the  new  bags.   The  tests were con-
ducted over a time period from immediately upon  startup  of  the  new bags to
several months afterward.   No  dependence of overall  or fabric filter S0_
removal on old versus new bags or  on operating hours on  the new bags is
apparent in these  results.

     Crucial to the use of  spray  drying technology for many western plants is
the ability to employ makeup waters that are  blowdown streams from other plant
water systems.  Water balance  considerations  make cooling  tower blowdown a
particularly desirable  spray dryer makeup water  source.  A series of tests was
conducted on the Arapahoe pilot unit to determine the suitability of two simu-
lated cooling tower blowdown  (CTB) streams as makeup to  the system.  The com-
positions of the two simulated CTB streams are summarized  in Table 4.
                                    10-102

-------
                  100-
                   80-
                 '. 60-
                 o
                 E
                 o>
                 IT
                 CNJ
                 O
                 C/3
          40-
                   20-
                                            LEGEND
                                    Overall SO2 Removal
                                    Spray Dryer Removal
                                    Fabric Filter Removal
                             0.5      1.0      1.5
                                 Reagent Ratio
                                            i
                                           2.0
Figure 14.
                      Comparison of 1000 ppm Inlet SO  Recycle Results
                      with  Old  and New Bags in Fabric Filter
  TABLE 4.  NOMINAL  COMPOSITIONS  OF SIMULATED COOLING TOWER SLOWDOWN WATERS
            (mg/L)
          Component
                         CTB No.  1
CTB No. 2*
Ca++
Na+
Mg++
S04
ci-
1160
480
-
1000
2050
410
1830
53
5080
170
*Includes 10 ppm AMP, HEDP,  or  polyacrylate scale inhibitors for some runs.

     CTB No. 1 was chosen because  it  represents the sulfate levels of a gypsum
limited, acid treated system.   Originally of interest in this water, then, was
the sulfate content of  1000  ppm.   This  sulfate content has generally been
accepted as being too high for  use as slaking water.  Poor slaked lime quality
has been reported to result  from the  use of such waters in slaking.   In order
to evaluate these impacts for a spray dryer based FGD system, tests were con-
ducted with CTB No. 1 used as slaking water, makeup to the recycle/lime slurry
mix tank, and temperature control  water.  In one test, CTB No. 1 was not used
for slaking, but still  used  in  the mix tank and for temperature control.  The
                                      10-103

-------
results of these tests are  compared  to baseline 1000 ppm inlet SO,, recycle
tests  in Figure 15.  The results  in  Figure 15 are on the surface quite  sur-
prising.  All of the tests  using  CTB No.  1 show elevated S0? removal perfor-
mance  both in the fabric filter and  overall.
             100-
              80-
ra 60-
o
E
CD
rr
040H
to
              20-
                                                LEGEND
                                              CTB
                                     Base- CTB Slaking
                                      line  M/U  & M/U
                                          0
                                          a
                                          A
           Overall SO2 Removal
           Spray Dryer Removal
           Fabric Filter Removal
                        0.5
                    i
                    1.0
 T
1.5
 r
2.0
                             Reagent Ratio
         Figure  15.
         1000 ppm Inlet S02 Recycle Results  for  Baseline
         Conditions When Using Simulated Cooling Tower
         Slowdown No. 1 for Slaking and Makeup Water
     A  further  examination of  simulated CTB No.  1  shows  that the nominal
 chloride  level  is quite high,  at  2050 ppm.  Although not yet confirmed by
 chemical  analyses, material balance calculations  indicate that for the three
 tests where CTB No.  1 was used as slaking, mixing,  and  temperature control
 water,  steady-state  atomizer feed chloride levels were  approximately 5000 ppm,
 and  the fabric  filter solids had  approximately a  1  percent chloride content.
 These levels are very near those  reported for substantial S00 removal en-
 hancement by others.4,5,6  The enhancement effect may be cauied by the deli-
 quescent properties  of CaCl2, resulting in a lengthened  second drying phase
 time period in the spray dryer, and significantly increased residual moisture
 levels  in the solids collected on the fabric filter.  Both impacts would tend
 to improve SO  removal performance.

     The results in  Figure 15 appear to substantiate  a  chloride enhancement
 effect.  The three tests involving CTB No. 1 for  slaking,  mixing,  and tempera-
 ture control show dramatically increased fabric filter  SO, removal contribu-
 tion.  The one test  employing CTB No. 1 as mixing and temperature  control
water only shows a somewhat reduced effect.  This is  expected because the use
                                      10-104

-------
of service water for slaking tends to dilute the steady-state chloride level.
While the presence of this chloride enhancement effect may mask the impact of
high sulfate levels on lime slaking, it does appear that any detrimental efect
of 1000 ppm sulfate levels is small compared to the potential benefits of
elevated chloride levels.  One consideration in evaluating these results,
though, is that the chloride level in CTB No. 1 is artificially high, signif-
icantly higher than might be seen in a typical CTB water.  The high chloride
content in CTB No. 1 is a result of adding CaCl2 to produce a gypsum relative
saturation of near 1.0 for the 1000 ppm sulfate level simulated blowdown
water.

     The second simulated cooling tower blowdown water in Table 4, CTB No.  2,
accurately represents a side-stream softened cooling water system.  This water
contains a significant amount of sodium as the result of softening, and con-
sequently low calcium and high sulfate levels.  Four tests were conducted with
this simulated CTB; one with the constituents in Table 4 alone,  and one each
with 10 ppm as active ingredient of HEDP, AMP, and polyacrylate scale inhib-
itor, respectively, added.  These tests were conducted because most cooling
towers are treated with scale inhibitors.  Scale inhibitors are intended to
delay the precipitation of limited solubility species on heat exchanger sur-
faces.   However, precipitation of these species is the very operation occur-
ring as water evaporates from the droplets in a spray dryer.   Although the
dosage of scale inhibitors encountered in most cooling tower blowdowns is
relatively low (<10 ppm), the net effect of such inhibitors on spray dryer S02
removal or solids drying performance was unknown.   The results of these tests
are presented in Figure 16.   The results show that, in general,  SC>2 removal
across the spray dryer and for the overall system is enhanced by the use of
CTB No. 2 as mixing and temperature control water (slaking water use was not
tested).   The individual tests are not identified because there is some
scatter in the four tests conducted, and individual points are not sufficient
to rank the magnitude of enhancement resulting from the use of the three
generic scale inhibitors.  What is important in Figure 16 is that for all of
the four tests, overall system performance is enhanced by the use of the simu-
lated CTB as makeup.

     Based on the encouraging test results presented in Figures 15 and 16 for
the use of simulated cooling tower blowdowns as system makeup water, a number
of additional tests using other simulated as well as actual CTB waters are
planned for the coming months.

                           SUMMARY AND CONCLUSIONS

     A number of significant conclusions can be made at this point in the
pilot test program.  These include:

     •    Spray dryer operability and S02 removal performance both are
          improved by operation in the recycle rather than once-through
          mode.  The improvements appear to be related to increased
          solids content in the slurry fed to the atomizer as well as to
          the other benefits of recycling in improving sorbent utiliza-
          tion.
                                    10-105

-------
          100-
           80-
         o
         E
         o>
         o:
         CO
           40-
           20-
                                     LEGEND
                     Overall SO2 Removal
                     Spray Dryer Removal
                     FabricFilter Removal
                      I
                     0.5
             1
            1.0
1.5
2.0
                        Reagent Ratio
    Figure 16.
1000 ppm Inlet SO  Recycle Results  for
Baseline Conditions and When Using
Simulated Cooling Tower Slowdown No.  2
for Makeup Water
Operation of a fabric filter downstream  of  a  spray dryer
appears to be acceptable in terms of bag fabric  impacts, and
may actually be beneficial with respect  to  pressure drop per-
formance.  The one area of concern  is  that  of  corrosion of
compartment walls and mild steel bag caps in  cold  spots within
the compartments.  Corrosion problems  can most  likely be
avoided through the use of adequate insulation  thicknesses and
careful design and installation practices to  avoid localized
cold spots.

For low spray dryer inlet S02 concentrations,  SO   removal per-
formance and lime reagent utilization  remains  quite acceptable
in spite of very low dryer inlet temperatures  (as  low as
210°F).

Cooling tower blowdown waste streams available  in  most power
plants appear to be an acceptable and  in some  cases beneficial
source of makeup water to a spray dryer  based FGD  system.  Typ-
ical scale inhibitors present in these cooling  tower blowdown
streams do not appear to adversely affect SO,,  removal or solids
drying performance.                         ^
                           10-106

-------
                                  REFERENCES

1.    Blythe,  Gary M. ,  et al.   "EPRI Spray Drying Pilot Plant Status and
     Results," presented at the Joint EPA/EPRI Symposium on Flue Gas Desul-
     furization,  Hollywood, FL, May 17-20, 1982.

2.    Masters, Keith.   Spray Drying Handbook.  Third Edition.  George Godwin,
     Ltd.,  London,  1979.

3.    Baker,  Robert  J.  and Richard W. Jordan.  "Effect of Dissolved Solids  in
     SC>2 Scrubbed Water Used for Lime Slaking," presented at and part of the
     proceedings of the 3rd WWEMA Industrial Pollution Conference, Chicago,
     IL, April 1975.

4.    Karlsson, Hans T. ,  et al.   "Activated Wet-Dry Scrubbing of SC>2. " Journal
     of the Air Pollution Control Association.  Volume 33, No.  1, January
     1983.   pp. 23-28.

5.    Hansen,  Svend Keis, et al.  "Status of the Joy/Niro Dry FGD System and
     Its Future Application for the Removal of High Sulfur, High Chloride  and
     NOX from Flue  Gas."  Publication No. 83-JPGC-APC-8, The American Society
     of Mechanical  Engineers, New York, NY.

6.    Blythe,  Gary M.  and Richard G. Rhudy.  "Field Evaluation of a Utility Dry
     FGD System."  To be presented at the joint EPA/EPRI Symposium on Flue Gas
     Desulfurization,  New Orleans, LA, November 4, 1983.
                                     APPENDIX

           CONVERSION OF BRITISH ENGINEERING (ENGLISH) UNITS TO SI UNITS

          To Obtain                  From                 Multiply by

          ng/J                    lb/106 Btu                   430

          m                           ft                     0.3048

          m2                          ft2                    0.0929

                                                             0.0283
mg/nm^
°C
gr/dscf
°F
1.83
tc = 5/9(tf-32)
                                    10-107

-------
SESSION 10, PART II:   DRY FGD:   FULL SCALE INSTALLATIONS

Chairman:   Richard G.  Rhudy
           Electric Power Research Institute
           Palo Alto,  CA

-------
FIELD EVALUATION OF A UTILITY DRY SCRUBBING SYSTEM




G. M. Blythe, J. M. Burke, T. G. Brna, R. G. Rhudy

-------
              FIELD  EVALUATION  OF A UTILITY DRY SCRUBBING SYSTEM

                       Gary  M.  Blythe and Jack M.  Burke
                             Radian Corporation
                             Austin, TX  78766

                               Theodore G.  Brna
                 Industrial  Environmental Research Laboratory
                     U.S.  Environmental Protection Agency
                      Research  Triangle Park,  NC  27711
                               Richard G.  Rhudy
                      Electric Power Research Institute
                             Palo Alto, CA  94303
                                   ABSTRACT
     This program,  co-funded by the U.S.  Environmental Protection Agency and
the Electric Power  Research Institute,  has resulted in an evaluation of a full-
scale utility spray dryer/baghouse dry  FGD system.   The system is installed at
the Northern States Power Company's Riverside Station and treats flue gas from
a nominal 100 MW of coal-fired power generation.   This has been the  first
independent  evaluation of a full-scale  spray dryer/baghouse system.

     For the test program, two different  coals were used as boiler fuels.  One
coal was a subbituminous coal and coke  mixture with a nominal 1.2 percent sul-
fur content.  The second was a 3.4 percent sulfur Illinois bituminous coal.

     During  the  test program, S02 removal, particulate emissions, sulfuric
acid removal, and extensive process data  were recorded.  The test program was
conducted from July to October 1983, so only preliminary results are pre-
sented.   Low sulfur coal tests indicated  up to 90 percent S02 removal was
achievable in the short term with slightly sub-stoichiometric amounts of lime
addition. A similar removal was achieved in short  term tests with high sulfur
coal at  reagent  ratios of 1.3 to 1.4.  Calcium chloride addition was found to
reduce the lime  addition requirements for high sulfur tests by approximately
25 percent.
                                     10-109

-------
                                 INTRODUCTION


     This paper presents preliminary results from a program entitled "Field
Evaluation of a Utility Dry Scrubbing System."  The objective of the program
has been to acquire performance data on an operating, utility-scale, spray-
dryer-based, dry FGD system.   The system was evaluated primarily to determine
S02 and particulate removal performance and lime reagent consumption.  The
system chosen for evaluation is the Joy/Niro Demonstration Unit located at the
Northern States Power Company (NSP) Riverside Station in Minneapolis, Minne-
sota.  The Riverside system was chosen for this program because it is the
first lime-based system in operation using a full-size (46 ft* diameter) spray
dryer module.  Testing was conducted with both low and high sulfur boiler
fuels.  The program is being conducted for the Environmental Protection
Agency, Industrial Environmental Research Laboratory in Research Triangle
Park, and for the Electric Power Research Institute under a cooperative fund-
ing arrangement.

     This test program was needed for several reasons.  First, a significant
number of spray-dryer-based dry FGD systems have been sold to the utility
industry.  At least 17 systems representing approximately 6,800 MW of electric
generating capacity have been sold to the utility industry.  No independent
evaluation of the technology is available to aid utilities purchasing dry FGD
systems, since virtually all of the performance data have come from pilot- or
demonstration-scale units operated by system vendors.  Several full-scale sys-
tems are coming on-line, but little performance data on these systems have
been published.  A second consideration is that most of the new systems and
all of the future systems must meet the 1978 New Source Performance Standards
for coal-fired utility boilers, which call for 70 to 90 percent S02 removal on
a 30-day rolling average basis and particulate emissions of 0.03 lb/10" Btu
heat input or less.  Few data exist to confirm the ability of full size spray
dryer-based dry FGD systems to achieve these levels of performance, particu-
larly for high sulfur coal requiring 90 percent S02 removal.  Finally, the
spray dryer-based systems sold to date have been justified using economics
based on vendor guarantees for lime consumption.  However, no independent eval-
uation of lime consumption on an operating full-scale system has been con-
ducted.  The program described in this paper was designed to collect the type
of data that can help determine the suitability of this technology for utility
application.
 ""British engineering units rather than SI units are used in this  paper  because
 of customary usage in the electric power industry.  An appendix  provides
 appropriate conversion factors.
                                    10-110

-------
     The on-site test program was conducted from July 11 to October 8, 1983.
Because little time has been available for data evaluation, this paper does
not include a complete evaluation of results.  Instead, the project is des-
cribed, system operation is discussed, and preliminary performance results are
presented.

                             PROJECT DESCRIPTION

     The project description includes a description of the Riverside Station
and FGD system, a summary of the test program, and a discussion of the limita-
tions of the Riverside system and how they have affected this evaluation.

SITE DESCRIPTION

     The Riverside Generating Station, operated by Northern States Power
Company, is located on the east bank of the Mississippi River in northeast
Minneapolis.  Some parts of the facility are well over 50 years old, and the
two units of interest on this project, No. 6 and No. 7, began operation in
1949.  The combined generating power of Units 6 and 7 is rated at 98 MW.  How-
ever; the pulverized coal, wall-fired units were originally designed to fire
an eastern bituminous coal.  Recently, the units have fired a western (Sarpy
Creek) subbituminous coal.  A small amount (10 to 15 percent) of high sulfur
coke is added to the subbituminous coal to improve its firing properties.  The
units can still be fired with high sulfur bituminous coal and, in fact, were
so fueled for five weeks of this test program.

     Units 6 and 7 are used as peaking units in the NSP system.  During the
cooler months of the year, the units are rarely operated.  During the summer
months, the units are operated at up to 90 MW during daytime hours and at a
minimum load during the night.  The units are normally banked over the weekend
during the summer.

     In 1980, a full-scale, Joy/Niro, spray dryer/fabric filter FGD system was
installed to treat the combined flue gas from the two units.  The fabric fil-
ter was actually purchased by NSP because of the inability of existing ESP
collectors to efficiently collect the ash from the western coal.  The spray
dryer system was installed by the Joy/Niro joint venture under a cooperative
agreement with the utility to serve as a full-scale demonstration of the capa-
bilities of their dry FGD system.  Figure 1 is a simplified flow diagram for
the system.

     The spray dryer is a 46-ft diameter vessel, with flue gas introduced both
above the atomized spray in a roof gas disperser and below the atomized spray
in a central gas disperser.  A rotary atomizer is used, currently employing a
700 hp drive motor.  The spray dryer was sized to treat flue gas corresponding
to a 70 MW boiler load.  This reduced sizing allowed the capability to test
the system at greater than design flow rates.  It should be noted that, be-
cause Units 6 and 7 are over 30 years old, the flue gas flow rate at 70 MW is
equivalent to the flue gas rate from approximately a 100 MW new unit.  A new
unit would experience much less air inleakage and operate at a much lower net
plant heat rate than these units.  The downstream fabric filter contains 12
compartments, in 2 rows of 6 compartments each.  Because the fabric filter was
                                     10-111

-------
sized to treat hot flue gas, it is actually oversized when  the  spray dryer is
in operation because of the flue gas volume shrinkage which results from the
reduced spray dryer outlet temperature.

     Pebble lime reagent is slaked in a Joy/Denver  attrition slaker.  A Joy/
Denver ball mill is also available for lime slaking.  Milk  of lime, dilution
water, and recycle solids are added to a mix  tank at rates  determined by a
Honeywell process control computer.  The mix  tank effluent  is pumped to a sep-
arate atomizer feed tank.  From the atomizer  feed tank,  slurry  is pumped to a
head tank at the top of the spray dryer.  A pinch-type  control  valve regulates
the flow of slurry to the atomizer to maintain  either a  constant spray dryer
outlet temperature or a constant approach to  adiabatic  saturation.  When the
system is operated in an S02 removal control  mode,  the Honeywell process con-
trol computer calculates the amount of lime which must be added upstream at
the mix tank in order to achieve the desired  S02  removal.  Recycle material is
added at the mix tank at a rate required to bring the mix tank solids level up
to a set point, normally 35 weight percent solids.  The  recycle solids are col-
lected from the spray dryer bottom dropout and  largely  supplemented by a por-
tion of the fabric filter catch.
                                                                             Unit?
                                                                             Stack
  Flue Gas
From Unit 7
                 Lime
                 Slurry
                Storage
                Trough
              Figure  1.   Flow Diagram for Riverside Dry FGD System
                                   10-112

-------
     The atomizer feed scheme tested is now an obsolete design.  The mix tank
and atomizer feed tanks used at Riverside each have about 45 minutes of resi-
dence time.  Thus the system is relatively slow to respond to changes in spray
dryer inlet flue gas conditions (i.e., changing SC>2 concentration, inlet tem-
perature, or wet bulb temperature)-  Current Joy/Niro system designs call for
preparing milk of lime slurry and recycle slurry in separate tanks and then
mixing them in a tank with a relatively short residence time.  This should
result in a quicker response to process changes.

TESTING APPROACH

     As described in the introduction, a primary objective of the program has
been to quantify SC>2 removal by the system.  A continuous emission monitoring
system (CEMS) has been installed to quantify S02 removal.  The CEMS includes a
DuPont Model 460 two-point extractive 862 monitor and a Thermox 02 monitor
sampling the spray dryer inlet and spray dryer outlet ducts, and a Lear
Siegler SM810 in-situ point-type S02 analyzer installed in a short run of duct
at the fabric filter outlet.  A second Thermox 02 analyzer is mounted on the
duct exactly opposite the Lear Siegler monitor.

     Other than S02 removal data, lime consumption and other important process
parameters were recorded as hourly averages for each test day-  Lime consump-
tion was measured primarily by determining the lime content of the milk of
lime slurry introduced to the atomizer feed mix tank.  The flow rate of this
slurry was continuously measured with a magnetic flow meter and recorded by
the Honeywell process control system computer.  Other methods of lime consump-
tion measurement have included continuous quicklime weigh belt rate measure-
ments, recording of daily quicklime truckload deliveries, and determination of
the lime content and flow rate of the actual atomizer feed slurry-  Energy bal-
ances have been used to confirm agreement between slurry feed rate and flue
gas flow measurements.

     Other than quantifying S02 removal and lime consumption performance for
the spray dryer/baghouse system, determination of particulate removal
performance for the system was also a primary objective.  This performance was
determined by manual sampling of flue gas streams for particulate
concentrations, using EPA methods.  Particulate loadings were measured at the
spray dryer inlet, spray dryer outlet, and fabric filter outlet locations.

TEST PLAN

     The test schedule is summarized in Table 1.  The schedule shows four dif-
ferent sets of conditions with low sulfur, Sarpy Creek coal/coke blend, and
three different sets with a high sulfur Peabody Illinois coal.  The Sarpy
Creek coal/coke blend has a nominal sulfur content of 1.1 to 1.2 percent and a
heating value of 9300 Btu/lb.  New Source Performance Standards for utility
boilers would require 75 to 80 percent S02 removal for boilers firing a fuel
with this sulfur and heating value.  In some localities, state or local regula-
tions might require as high as 90 percent S02 removal with this fuel.  Conse-
quently, low sulfur tests were conducted at both 75 percent and 90 percent
target S02 removal levels.  Additionally, some tests were scheduled at a fab-
ric filter air-to-cloth ratio higher than the normal value of 2:1 and with  the
                                     10-113

-------
        TABLE  1.   TEST  SCHEDULE  AND TARGET SYSTEM OPERATING CONDITIONS3
Fuel
Sarpy Creek Coal/Coke
Sarpy Creek Coal/Coke
Sarpy Creek Coal/Coke
Sarpy Creek Coal/Coke
Illinois Coal
Illinois Coal
Illinois Coalb
S02
Removal
Level,
%
75
75
90
90
90
90
90
Fabric
Filter
A/C Ratio,
cfm/ft2
2:1
2.2:1
2:1
2:1
2:1
2:1
2:1
Slaker
Type
Attrition
Attrition
Attrition
Ball Mill
Ball Mill
Attrition
Attrition
aAll tests planned to be  conducted  at  an  18°F  approach to adiabatic
 saturation,  35 weight percent  solids  in  the atomizer  feed slurry,  a 70 MW
 daytime boiler load, and with  a  once-per-hour baghouse  cleaning frequency,

^Calcium Chloride  Addition  Tests
                                    10-114

-------
ball mill slaker rather than attrition slaker being used to prepare the lime
slurries.

     The test at the higher air-to-cloth ratio was originally intended to be
conducted at an air-to-cloth ratio near 2.7:1.  However, the air-to-cloth
ratio is increased at a constant boiler load by taking individual compartments
off-line.  After taking compartments off-line to raise the gross air-to-cloth
ratio to near 2.7:1, only 6 of the 12 compartments remained on-line at a 70 MW
boiler load.  The combined effects of taking an operating compartment off-line
for cleaning and reverse gas flow during cleaning momentarily sends the effec-
tive air-to-cloth ratio to 3.5:1 to 4:1.  In a unit with more compartments on-
line, the effects of cleaning one compartment in increasing the effective air-
to-cloth ratio are much smaller.  Additionally, in the event of a temporary
loss of slurry flow to the atomizer while a compartment is cleaning, the
increased flue gas volume due to the much higher spray dryer outlet tempera-
ture would result in a further increase in the effective air-to-cloth ratio to
well over 4:1.  When operated at such a high effective air-to-cloth ratio, the
pressure drop across the fabric filter is increased to several times its nor-
mal value.  Such a combination of events did occur on the first day of the
high air-to-cloth ratio tests, and the resulting momentary high pressure drop
caused the control system to bypass the fabric filter.  In order to provide a
margin of safety, one more compartment was put on-line.  This lowered the
gross air-to-cloth ratio to 2.2:1 to 2.3:1, which is only 10 to 15 percent
higher than baseline conditions.

     The fourth set of conditions for the low sulfur tests (see Table 1) was
to have included ball mill rather than attrition slaked lime.  However, ball
mill operation during this test program was generally not successful.  The
reasons for the ball mill operating problems are further discussed under Oper-
ational Results.  Because of these problems, less than one day of system
operation was observed using ball mill slaked lime.

     Two sets of conditions were tested with the high sulfur coal.  The coal
was an Illinois No. 6 coal with a nominal 3.4 percent sulfur content and
10,800 Btu/lb heating value.  Current New Source Performance Standards for
utility boilers require 90 percent S02 removal when coal of this sulfur con-
tent and heating value is burned.  Consequently, only a 90 percent target S02
removal was tested with this high sulfur coal.  The tests included baseline
conditions of 90 percent removal with attrition-slaked lime and a second test
run employing calcium chloride addition for lime utilization enhancement.
Chloride addition has been reported previously to enhance lime utilization in
spray dryer/baghouse FGD systems^ ' , but this is the first test of chloride
addition at a full-scale utility installation.  The enhancement effect is
thought to occur because of the deliquescent properties of calcium chloride.
This effect is further discussed in the System Performance portion of this
paper.

     Originally, ball mill slaker tests and high air-to-cloth ratio tests were
planned for the high sulfur coal portion of the test program.  However, due to
                                   10-115

-------
the ball mill slaker operating problems previously mentioned and the previous-
ly discussed constraints on gross air-to-cloth ratio, only the last two  tests
listed in Table 1 were conducted during the high sulfur test period.

SYSTEM LIMITATIONS

     Several system limitations combined to restrict the amount and the  type
of data that could be collected.  First, as mentioned earlier, Riverside  Units
6 and 7 are peaking units.  As such, they are rarely operated in the winter
and only operate part-time during July to October.  This part-time operation
involves unit loads of 70 to 90 MW during weekday daylight hours, minimum load
(30 to 50 MW) overnight during the week, and banking the boilers over  the
weekend.  Figure 2 is an illustration of the combined Units 6 and 7 loads dur-
ing a typical week.  Therefore, although the FGD system was operated at
desired S02 removal levels 24 hours per day, only about 12 hours per weekday
of near full-load operation were available for evaluating FGD system perfor-
mance.  At  the beginning of each 12-hour full-load period, the FGD system gen-
erally goes through a transient period due to a large increase in boiler load.
On Mondays  the unit must undergo a cold start-up.  Although this cycling pro-
vides a  severe test of the capabilities of the system, it reduces the  period
of steady state  operation at the desired S02 removal level over which  lime
reagent  consumption can be measured.
                  100-
                   80-
                   60-
                 _
                 ~ 40 H
                 c
                   20^
                        Sun    Mon   Tues   Weds Thurs  Fri    Sat
         Figure 2.  Illustration of Typical Unit Load During Test  Program

      Additionally, the Riverside system was the first utility-scale  system
  designed and built by the Joy/Niro joint venture and was  built as  a  demonstra-
  tion unit.  As the first unit built, the Riverside system has  provided the
  opportunity to refine and modify design features for subsequent  systems.
                                     10-116

-------
Thus, some specifics of the Riverside system are different from what will be
found in later designs.  An example of this is the slurry feed system design,
which has been modified for subsequent system designs to provide quicker
response to transients such as load changes.  At Riverside, lime slurry, make-
up water, and recycle solids are mixed in a single tank.  The lime slurry
addition rate is based on SC>2 removal (for the SC>2 removal control mode of
operation), the makeup water rate is controlled to maintain level in the mix
tank, and the recycle solids rate is controlled to achieve a set weight per-
cent solids (normally 35 percent solids).  The mix tank has a residence time
of approximately 45 minutes.  This slurry mixture then overflows to an atom-
izer feed tank with a similar residence time.  When the inputs to the mix tank
change composition abruptly (such as when the unit load increases), approxi-
mately 3 or 4 hours are required before the feed system stabilizes near steady
state for the new conditions.  Recently designed Joy/Niro systems have a
faster response slurry feed system.  In these systems, recycle solids and
makeup water are mixed in a recycle slurry tank to a consistency of approxi-
mately 50 to 60 weight percent solids.  The lime and recycle slurries are then
mixed in a feed tank with a relatively short residence time of one-half hour
or less.  Because only this feed tank residence time impacts the system
response time, this feed system might stabilize after an abrupt change of
operating conditions in one hour or less.

     The impacts of the slurry feed preparation system design are particularly
important at the Riverside Station, as the normal station operation causes
significant load changes at least twice per day.  In fact, on some days during
this test program the NSP dispatcher called for the unit load to be varied
between 70 MW and 90 MW throughout the day.  On these days, the unit never
operated at one load long enough for the feed system to stabilize.  On most
other days, even if the load was steady all day, only 8 or 9 hours of the 12
hours of full load operation actually represented steady state conditions.

     Other aspects of the Riverside system's status as a demonstration unit
affected the results of this program.  For example, the system contains only
one  spray dryer module, while most utility systems will be multiple module
systems.  In a multiple-module system, equipment problems which affect one
individual module have a smaller impact on overall system performance.  Being
a one-module system, equipment problems tended to cause the entire system to
have to be shut down, or operated at conditions other than those desired.
Even considering that the smaller Riverside system only warrants a single
dryer vessel, some individual components of the system have not been installed
with the redundancy that the vendor would likely install in a commercial  sys-
tem.  These considerations have had a detrimental effect on both the amount of
system downtime and the number of process-equipment-related upsets during the
test program.

     A final consideration which has affected the test program involves the
recent history of the Riverside station.  For nearly the first two years  of
operation, the FGD system was used as a full-scale demonstration and testing
unit by the process vendors.  During this time the Joy/Niro joint venture had
responsibility for the operation of the system, even though NSP actually  pro-
vided operating personnel.  Within the year prior to the test program, NSP had
assumed responsibility for the operation of the FGD system.  Immediately  prior
                                     10-117

-------
to the test  program,  the Units 6 and 7 boilers and FGD system were off-line
for much of  the winter and spring,  as NSP does not need power production from
these units  during this time.   During this long period of downtime, normal
personnel turnover (promotions, retirement, transfers, etc.) resulted in a
number of new operators rotating into the FGD system operating staff.  Early
in the test  program,  then, the FGD system was being operated for the first
time in several months with a staff of operators having little previous exper-
ience with the FGD system.  Consequently, a lot of test days early in the pro-
gram were not productive as far as  representing steady state operation at test
conditions.   As the test program continued, successful test days represented a
higher fraction of potential days of operation.

     Another consideration here is  that, after the Joy/Niro joint venture
turned over  the dry FGD system to the utility, and prior to conducting this
test program, NSP only operated the system at 50 to 60 percent S02 removal
levels.  For such moderate S02 removal levels, they had previously operated
the  system under very conservative  conditions (i.e., higher spray dryer outlet
temperatures).  Thus, operation at  75 or 90 percent S02 removal, with a 20°F
approach to adiabatic saturation, was not a normal operating mode immediately
prior to this test program.  Early  in the program, the operators tended to
revert to conservative operation (i.e., higher spray dryer outlet temperature)
during any minor upset, such as soot blowing in the boilers.  This would move
the  operation away from the desired conditions and would preclude acquiring
desired steady state operating data.  As the program continued though, this
occurred much less frequently as the operators became more comfortable with
operating at test conditions.

     Considering the previous discussions, it was not realistic to report
availability of the system, as the  availability of the Riverside system would
tell little  about that of a commercial utility, multi-module, dry FGD system
on a new base-loaded boiler.  The combined effects of weekly cold startups,
frequent load changes, little redundancy, and a somewhat undertrained operat-
ing  staff at Riverside do little to promote a fair assessment of the potential
availability of a commercial system.

     However, the general operation of the system was closely observed during
the  test program.  Much of the downtime or off-condition operating time was
due  to problems specific to the Riverside system.  Others appear to be more
generic to dry FGD systems.  These  more generic problems are discussed in this
paper, as they are more likely to occur in other systems.

                                   RESULTS

     The results of the program are divided into two areas:  Operational
Results, which includes a qualitative discussion of the operation of the sys-
tem  during the test program; and System Performance, which includes prelimi-
nary S02 removal, lime consumption, and particulate and sulfuric acid removal
data.
                                   10-118

-------
OPERATIONAL RESULTS

     In general, the equipment that comprises the basis of the dry FGD system,
the spray dryer and baghouse,  were relatively trouble-free throughout the
program.  At the conditions tested, the spray dryer did not show evidence of
potential problems such as wheel nozzle pluggage, excessive buildup of solids
on the walls, or formation of wet solids within the dryer.  Some atomizer
problems were observed, but most of these appeared to be caused by circum-
stances specific to the situation at Riverside rather than being generic to
the Joy/Niro system.  These problems will be discussed further later in this
section.

     The baghouse also operated well, with no significant bag/fabric related
problems being observed.  In this short-term test though, long-term effects
such as bag life or compartment wall corrosion rates could not be evaluated.

     Some problems were observed in four specific areas—the slurry feed sys-
tem, the ash handling system, the ball mill slaker, and in atomizer protec-
tion.  The system vendors may have addressed these problems in system designs
subsequent to Riverside, but the problems could be encountered in virtually
any spray-dryer-based, dry FGD system.  Each of these areas is discussed
be 1 ow.

     Slurry Feed System—In a recycle lime system, lime slurries containing
up to 25 percent solids and recycle/lime slurries of 30 to 40 percent solids
are commonly encountered.  When dealing with slurries with a high solids con-
tent and high viscosity, problems such as plugging of pump suction lines,
solids buildup on tank walls, plugging of in-line screens used to remove over-
size material, and loss of flow when switching pumps are commonly encountered.
Such problems were encountered often at the Riverside system.  Years of opera-
tion of wet lime/limestone FGD systems have established means of dealing with
such problems:  always keep slurry in tanks well agitated, never allow slurry
levels to drop to the agitator blade level, never allow slurry to stand in
filled lines, allow for on-line changing of plugged or partially plugged
screens, etc.  These same solutions will need to be applied in spray drying
systems.

     The quantities of these slurries that must be dealt with in a  spray dryer
system are much smaller than what would be encountered in a wet system.  At
Riverside, typical atomizer feed slurry rates are 150 to 200 gpm.   In a  simi-
larly sized limestone wet FGD system, the slurry recirculation rate  could  be
as high as 40,000 gpm.  The point to be made is that while some of  the  slurry
handling problems of a wet FGD system may still be encountered in a  spray
dryer system, they will occur on a greatly reduced size of equipment.  This
should make both problem solving and routine maintenance  easier.  Another
important point to be noted is that no chemical  scaling tendencies were
observed at Riverside.

     Ash Handling System—In comparing the wet versus  dry FGD systems the
spray dryer system has a slurry feed system  that deals with much  lower  flow
rates for a given unit capacity, but the quantities of  dry ash  and  FGD
                                    10-119

-------
by-products that must  be moved around the system are substantially greater
than for a comparably  sized particulate collection device/wet scrubber system.

     For example, at Riverside the amount of solids entrained in the flue gas
at the fabric filter inlet (spray dryer outlet) is 3 to 5 times that in the
spray dryer inlet gas.  For low sulfur coal operation only a small portion of
this increase is attributable to the lime reagent introduced and S02 removed
in the spray dryer.  The majority is material being recycled from the fabric
filter catch to the spray dryer.  Thus, the rate at which solids are removed
from the fabric filter hoppers and transported to recycle storage bins can be
several times greater than the net rate of waste product actually leaving the
system.

     Specifically, for low sulfur coal, 75 percent S0£ removal, and the unit
operating at 80 MW, the steady state rate of fly ash and spray dryer by-pro-
ducts leaving the system amounted to 7 to 8 tons/hr-  However, due to the
effects of recycle, approximately 20 tons/hr were actually being collected by
the baghouse.  The excess between that going to the disposal silo and the 20
tons/hr being collected, or 12 to 13 tons/hr, was being recycled to the slurry
feed system.  In this example, roughly three times the quantity of ash and
spray dryer solids leaving the system must be continually removed from the
baghouse hoppers, and roughly twice the quantity leaving the system must be
continually transported to the recycle storage bin.

     At Riverside, solids collected in the baghouse hoppers and at the bottom
of the  spray dryer are intermittently dumped onto mechanical conveyors through
motor-driven tipping valves.  One conveyor each handles ash from the six com-
partments on the north and south sides of the baghouse, respectively, while a
third conveyor handles spray dryer bottom solids.  These conveyors empty into
surge bins mounted above air conveying blow pots.  The blow pots cycle through
the following five consecutive processes:  filling from the surge bin, isola-
tion from the surge bin, pressurization, pneumatic conveying to either the
recycle or disposal storage silos, and then depressurization.  The recycle
silo empties through a rotary valve onto a weigh belt which meters recycle
material into the mix tank.

     The problems with solids handling which most frequently occurred at
Riverside involved the baghouse mechanical conveyors, blow pots, and the
recycle bin rotary valve.  The problems with the mechanical conveyors appeared
to be related to marginal capacity.  The north or south conveyors frequently
tripped out because their drive motors were overloaded.  Momentary overloading
of the  conveyors is aggravated by the nature of the operation of the baghouse,
where compartments clean intermittently throughout each hour.  That is, the
average rate at which the baghouse collects material might be 20 tons/hr, as
in the  example above.  Each conveyor would see an average of 10 tons/hr if the
same number of compartments were on-line on both sides.  However, this average
includes short periods of high solids flow, immediately after a cleaning com-
partment begins reverse gas flow, and periods of low solids flow rate between
compartment cleaning periods.  Thus, one conveyor may see instantaneous rates
of several times 10 tons/hr.  This problem was apparently corrected after this
test program was completed by a modification which meters the rate at which
ash/spray dryer by-products are emptied from the hoppers onto the conveyor.
                                    10-120

-------
     In addition, the conveyors require occasional maintenance, such as
replacement of failed drive motors and bearings.  If one conveyor must be
off-line for a very long time, then all six of the compartments emptied by
that conveyor must be taken off-line once the hoppers below those compartments
are full of ash.  At Riverside, failure of one conveyor can cause the boiler
load to be limited to that which can be handled by the six compartments
remaining on-line.  Because the rate at which solids are collected in a fabric
filter downstream of a spray dryer is greatly increased, unless the hoppers
are increased in size compared to conventional particulate control configura-
tions, such a conveyor failure can very quickly limit unit load.

     Failure of a blow pot can have a similar effect, as only a small surge
bin separates each conveyor from a blow pot.  NSP has attempted to provide
some redundancy in blow pots, but in an instance where only one blow pot is
available to accept solids from a conveyor, a blow pot outage can cause the
conveyor to be shut down as well.  Common blow pot failures observed included
plugging, leaking isolation valves, and loss of air pressure to the
pneumatically operated valves.

     A third piece of solids handling equipment which rapidly affected the
operation of the system on failure was the rotary valve at the bottom of the
recycle solids bin.  When this valve failed, often due to a piece of tramp
material jamming the blades, the solids content in the atomizer feed slurry
immediately began to drop, and the system began to approach once-through lime
operation.  The large slurry mix tank and atomizer feed tank could be consi-
dered a 1.5 hour reservoir of feed slurry, but in order to avoid splashing
slurry on tank walls and a resulting buildup of dry solids, the tanks must be
continually kept near full.  Consequently, these tanks begin to dilute immedi-
ately when recycle solids flow is interrupted because makeup water must be
added to maintain level.

     The solids handling equipment appears to be a key to successful operation
of the spray dryer/baghouse system.  The individual pieces of equipment must
be sized to handle solids rates much higher than is typical for conventional
particulate removal alone.

     Ball Mill Slaker—As mentioned previously, during this program the ball
mill slaker did not operate successfully for any extended period.  The feed
end of the slaker tended to plug with wet lime solids.  While there are sever-
al possible reasons why the plugging continually occurred, the actual cause
was not identified.  The ball mill operation is discussed below.

     The ball mill slaker is fed by an independently driven screw feeder
inserted directly through the trunion bearing of the mill.  Product lime  flows
out the opposite end, through a cylindrical screen, and into a product
collection trough.  Figure 3 is a simplified cross section of the ball mill
depicting the feed and product ends of the mill.  Two things are  clear from
this simplified illustration:  first, vapors produced by the heat of the
slaking reaction can only leave the slaker with the product or flow back
through the feed end.  Secondly, little hydraulic head is available to promote
the flow of the viscous slaked lime slurry out the product end of the  slaker
without immersing the feed screw in slurry.
                                   10-121

-------
                                                        Product
                                                        End
           Figure 3.   Simplified Cross Section of Ball Mill Slaker

     Either of these  considerations may have  contributed to feed screw plug-
gage.  The ball mill  slaker  was  operated at a very low water-to-lime ratio,
resulting in 30 to 35 percent  solids in the mill product slurry and a product
temperature of 97°C (207°F)  or  greater.  At these conditions,  the product
slurry is very viscous,  having  a paste-like consistency.  Also, with the prod-
uct slurry very near  or  perhaps  actually boiling, vapor production rates are
high.  Both of these  effects of  low water-to-lime ratio can promote feed screw
pluggage.  Another consideration is that the  ball mill slaker  had not been
operated for some time prior to this test program, as NSP prefers to operate
the attrition slaker  at  Riverside.   The mill  had not been completely cleaned
out after its last use,  and  solids  buildup around the discharge end undoubt-
edly contributed to the  slurry  level in the mill backing up into the screw
feeder -

     A final problem  specific  to the Riverside system may also have contrib-
uted to feed screw pluggage.  The SC>2 removal software in the  process control
computer tended to adjust  the  lime  slurry makeup rate to the mix tank after a
major change in the process  (e.g.,  unit load  change) in a classic dampened
controller response.   That is,  after an abrupt process change  the lime slurry
makeup rate to the mix tank  would tend to overshoot, then undershoot, for
several cycles, before stabilizing  at the required value.  During the under-
shoot portion of the  response,  the  lime slurry feed rates might be completely
stopped.  As the Riverside system has little  lime slurry storage capacity, the
slaking rate tended to directly  follow the slurry makeup rate.  Thus, periods
of high slaking rates would  be  followed abruptly by periods where the screw
feeder would shut down.   Shutting down the screw feeder momentarily several
times in succession undoubtably  contributed to the plugging problems.
                                                                          In
     The ball mill slaker problems at Riverside appear to be somewhat
site-specific as ball mill slakers have operated successfully elsewhere,
retrospect, the slaker might have run more successfully at a higher
water-to-lime ratio,  resulting in a less viscous product slurry and less vapor
                                    10-122

-------
release.   However, in the attempts to run the ball mill slaker during the high
sulfur tests in particular, the slaking water piping size did not allow opera-
tion at higher water-to-lime ratio at the lime slaking rates required.  During
the high sulfur tests, lime slaking rates averaged approximately four times
that required for the normal low sulfur fuel at similar unit load and percent
SC>2 removal conditions.

     Atomizer Problems—Although the atomizer motor, gearbox, and nozzle
wheel were generally trouble-free, on several occasions a problem which could
result in atomizer damage was observed.

     Two scenarios for potential damage were observed.  One occurred when the
unit was forced to run from the basic, or less sophisticated, control station
of the computer control system while the more sophisticated supervisory sta-
tion was undergoing repair.  The control software at Riverside does not have
full interlock protection for the atomizer when running from the basic sta-
tion.  Interlocks are software which automatically shut down the atomizer when
given inputs that are indicators of problems which might result in damage to
the atomizer.  While running in this mode, a minor problem involving loose
wires to the atomizer oil circulating pump occurred, intermittently shutting
the pump off.  However, the basic station only gave the control operator an
alarm rather than shutting the atomizer down automatically.  The atomizer con-
tinued to run for several minutes without oil circulation and sustained gear-
box damage.

     This incident may be very site-specific to the Riverside system due to
the limitations discussed earlier.  First, Joy/Niro has reported that their
commercial systems have full atomizer interlock protection in both the super-
visory and basic computer control stations.  Second, an operating staff with
more experience on a spray dryer system might have shut down the atomizer more
quickly when given an indication of a loss of oil flow and would not have
attempted to restart the machine prior to remedial action.

     The second scenario which could result in atomizer damage was observed on
more than one occasion.  This incident involved feeding slurry to the atomizer
wheel when it was not rotating.  Since the non-rotating wheel has a much lower
hydraulic capacity than a rotating wheel, slurry fed to the standing wheel
tends to overflow the wheel and can flow up the spindle to which the wheel is
attached and enter the atomizer oil system.  In such instances the oil sump
can be immediately emptied and flushed to avoid damage, but if the atomizer is
operated before cleaning, the slurry in the oil can eventually cause gearbox
damage.

     Such an instance of feeding slurry to a standing wheel can occur from
operator error in establishing flow to the atomizer when it is not in opera-
tion and/or starting the atomizer from the basic station without full inter-
lock protection.  As for the previous discussion, such an occurrence is much
less likely for a commercial Joy/Niro system, as even the basic station would
have full interlock protection.  The interlock system would normally close a
slurry feed block valve before the atomizer shut down and would not allow this
valve to open until the atomizer had returned to operation at full speed.
                                   10-123

-------
     However,  even with a properly operating interlock system, the previously
described problem could occur.   One such event occurred during this test pro-
gram.  In this instance,  during an atomizer trip, valve seat damage prevented
the atomizer feed slurry  block valve from closing.  In this case, all of the
interlocks operated properly,  but a mechanical problem allowed slurry flow to
the standing wheel.  An alert  operator could catch such a problem very
quickly, but it is still  likely that slurry would get into the atomizer lube
oil unless the flow was stopped immediately.

     It is interesting to note that in the example above, the spray dryer was
hardly upset by full slurry flow to a standing wheel.  Several thousand gal-
lons of slurry were fed in one instance, and the only repercussions were a
mess under the dryer vessel where the slurry flowed out of the dryer onto the
ground, and a minor buildup of dried solids on the vanes of the central gas
disperser just below the  wheel.

     What is clear from these  observations of potential atomizer damage are
two conclusions.  First,  the rotary atomizer on a spray dryer FGD system
should always be operated with the full protection of interlock controls.
Second, the control operators  should be well trained to respect the potential
damage which can result to the atomizer during upset conditions.

SYSTEM PERFORMANCE

     The results of S02 removal and lime consumption measurements during this
test program are summarized in Tables 2 and 3.  Table 2 summarizes the low
sulfur S02 removal results, while Table 3 summarizes those for the high sulfur
test period.

     For several reasons, the  S02 removal results from the program cannot be
expressed as 30-day rolling averages.  First, the unit was never operated at
one set of conditions for that long a period during the test program.  Also,
due to the operating characteristics of the peaking boilers, only 8 to 9 hours
per day typically represent full load, steady state operation.  Thus, the S02
removal and lime consumption results represent values measured during steady
state unit operation on only a portion of a number of successive days.  The
fact that these results are the average of a number of short term tests cannot
be used to determine whether or not the dry FGD system could sustain these
removal levels over a long period of time; instead, they only reflect the fact
that the boiler rarely operated at or near full load for a long period of
time.

     Lime consumption in Tables 2 and 3 is expressed as a reagent ratio.  This
term is defined as:

     Reagent Ratio  =  Moles Calcium in Fresh Lime Fed to System           m
                              Mole S02 in Inlet Flue Gas

(This definition corresponds to that of the term "stoichiometric ratio" in
many other dry FGD papers.)
                                    10-124

-------
     It should be noted that the values in Tables 2 and 3 should be considered
preliminary values which are based on a number of manual calculations for each
set of conditions.  Where possible, each value is supported by alternate cal-
culations.   For example, lime slurry feed rates are compared to lime weigh
belt readings:  essentially a calcium balance on the lime slaker.  Also, weigh
belt readings have been compared against lime truckload delivery inventories.
Flue gas flow rates are checked against slurry feed rates by energy balance
calculations.  In general, the values in Tables 2 and 3 represent the most
accurate values available at this time.  As more data reduction is completed,
these values will be refined but are not expected to be changed substantially.

     Table 2 shows that the desired SC>2 removal levels of 75 percent and 90
percent were achieved for the lower sulfur fuel with lower than or right at
stoichiometric amounts of lime.  This may be attributable to two factors.
First, at 35 percent feed slurry solids in low sulfur operation very high
recycle rates are possible.  This is seen in the recycle ratio values in Table
2.  Recycle ratio is defined in this paper as:
     P    i   P j. •   _  Lb Recycle Material in Atomizer Feed Slurry
     Kecycie Katio -                            [Ca(OH)2]
(2)
            TABLE 2.  PRELIMINARY S02 REMOVAL RESULTS, LOW SULFUR
Nominal
S02
Removal ,
%
75
90
Spray Fabric
Dryer Filter
Removal, Contribution, Reagent Recycle
% % Ratio Ratio
67-69 7-9 0.6 - 0.7 11:1 - 14:1
80-81 9-10 0.7 - 0.8 9:1 - 13:1
            TABLE 3.  PRELIMINARY S02 REMOVAL RESULTS, HIGH SULFUR
Nominal Spray
S02 Dryer
Removal, Removal,
% %
90 75 - 77
90a 67 - 69
Fabric
Filter
Contribution, Reagent Recycle
% Ratio Ratio
13 - 15 1.3 - 1.4 2:1 - 3:1
22 - 24 0.9 - 1.1 3:1 - 4:1
aHigh Chloride Concentration Tests
                                   10-125

-------
Since a large fraction of the baghouse catch and all of the spray dryer bottom
solids are recycled,  from pilot plant results it would be anticipated that
high sorbent utilization would be promoted.  Additionally, analyses of the
Sarpy Creek coal ash are not yet available, but the ash is known to have a
high alkaline earth content (CaO, MgO, Na20, and K20).  These alkaline com-
ponents are believed to have contributed to SC>2 removal.

     The results in Table 2 also show the range of SC>2 removal  in the spray
dryer (S.D.) and fabric filter (F.F.).  SC>2 removal in the fabric filter is
calculated relative to the spray dryer inlet S02 concentration  as follows:

                    [SO  into F.F.  -  SC>2 out of F.F.]
  S02 RemovalF>F> =          [S02 into S.D.](3)


With this definition, spray dryer removal and fabric filter removal can be
summed directly to yield overall removal values.  The results show that at 75
percent overall SC>2 removal, fabric filter removal contributes  little to the
overall removal.  This occurs because sorbent utilization is very high in the
spray dryer itself.  For 90 percent removal, the fabric filter  contribution
increases somewhat, but in an amount roughly proportional to the increase in
overall removal level.

     Table 3 summarizes the high sulfur coal test results.  For the first set
of data, corresponding to normal low chloride operation, an excess of fresh
lime is required.  Several factors probably contribute to this  effect.  One
may  be that, because of the increased lime addition rates, the  recycle ratio
is greatly reduced relative to low sulfur values in order to maintain the
total solids in the slurry at the desired weight percent value.  Also, the
Illinois coal ash being nonalkaline contributes no alkalinity to the S02
removal reactions.  For the first set of high sulfur data, the  fabric filter
contribution to overall S02 removal appears to be more important than for low
sulfur operation.

     Chloride addition has been reported by others to promote increased sor-
bent utilization in spray-dryer-based FGD sy stems-'- >2.  The benefits are
thought to result from the deliquescent properties of calcium chloride, which
delay complete drying of the droplets in the spray dryer and result in higher
residual moisture levels in the fabric filter solids.  The second set of S02
removal data in Table 3 corresponds to the addition of calcium  chloride at
levels which result in a chloride content of 1 percent in the fabric filter
solids collected.  This chloride level in the fabric filter solids was recom-
mended by the system vendor- Niro Atomizer, as being an optimum value for lime
utilization enhancement, based on pilot-scale studies conducted at their
Copenhagen, Denmark, test facility.  At Riverside, this solids  chloride level
required a liquid phase chloride concentration of approximately 7,500 ppm in
the  atomizer feed slurry.  For the recycle rates at Riverside,  fresh makeup of
calcium chloride accounted for about half of this liquid phase  content, and
the  remainder dissolved from the recycle material.  The results in Table 3
show that chloride addition significantly reduced the lime reagent ratio
requirements to achieve 90 percent removal; the lime requirement was reduced
by about 25 percent.
                                  10-126

-------
     The SC>2 removal results for these high chloride tests indicate increased
SC>2 removal across the fabric filter for 90 percent removal overall.  This is
an indicator that the benefits of high chloride level on residual moisture
level in the fabric filter have a greater impact on SC>2 removal than impacts
within the spray dryer.

     At Riverside, with chloride levels in the fabric filter solids at 1 per-
cent, residual moisture levels increased from just below 1 percent to nearly
2 percent moisture.  These moisture levels are still low enough to avoid
problems which result from handling wet solids.  Also, no buildup of wet
solids on the spray dryer walls occurred during testing, and solids collected
at the bottom of the spray dryer contained moisture levels below 4 percent.
These tests were conducted at an 18°F approach to adiabatic saturation at the
dryer outlet, just as were all previous tests.

     Material balance calculations indicate that, for a coal such as the Illi-
nois coal fired in the high sulfur tests at Riverside, a chloride content
of around 0.3 percent would provide the chloride levels of this test.  This
would be an uncharacteristically high chloride level for a typical 3.5 percent
sulfur coal.  However, using published bulk prices, it appears that calcium
chloride could be delivered to a typical plant site for approximately $225/ton
on a 100 percent CaCl2 basis^.  For $70/ton lime, delivered, a net savings
would result from chloride addition at Riverside if only a 15 percent
reduction in lime consumption resulted.  The actual saving observed was well
above 15 percent.  It appears to be economic in this case, disregarding
capital cost considerations, to operate at high chloride levels even if virtu-
ally all of the chloride must be added as calcium chloride.  For a 3.5 percent
sulfur coal with a higher chloride content (0.1 percent or better), the
economics would likely be improved.  Additionally, a makeup water source with
a  significant chloride content, such as some cooling tower blowdowns, would
further improve these economics.

MASS LOADING MEASUREMENT RESULTS

     Table 4 presents mass loading results for both the high sulfur and low
sulfur test periods.  The results show that the spray dryer increases the
grain loading at the fabric filter inlet to 3 to 4 times that of the spray
dryer inlet value.  The data also show that particulate removal levels
remained high throughout the test program.  Removal efficiencies across the
fabric filter varied from 99.95 to over 99.99 percent.  The emission levels in
Table 4 are expressed as grains per dry standard cubic foot.  Calculation of
emission levels in pounds per million Btu cannot be completed until the
results of coal ultimate analyses are available, but the pounds per million
Btu values should be approximately 1.5 times that in grains per dry standard
cubic foot.  Therefore, emission rates varied between approximately 0.002
lb/106 Btu and 0.012 lb/106 Btu, with most values below 0.006 lb/106 Btu.  The
few higher values were related to minor bag problems in one compartment.   In
all cases, particulate emission rates measured were well below the  current
NSPS level for utility boilers of 0.03 lb/106 Btu.
                                   10-127

-------
                   TABLE 4.  FLUE GAS MASS LOADING SUMMARY
     Sampling                    	Mass Loading, gr/dscf	
     Location                    Low Sulfur Tests            High Sulfur Tests
Spray Dryer Inlet                   3.2 to 4.1                  2.8 to 4.1

Spray Dryer Outlet                 11.0 to 13.8                14.9 to 17.2

Fabric Filter Outlet              0.001 to 0.003              0.001 to 0.008




SDLFURIC ACID MEASUREMENT RESULTS

     Flue gas sulfuric acid concentration measurements were also conducted.
For the low sulfur tests, no measurable sulfuric acid levels were detected at
either the spray dryer inlet or fabric filter outlet.  The inability to
measure sulfuric acid at the dryer inlet is apparently related to the alkaline
nature of the Sarpy Creek coal ash.   It is not clear whether the alkaline ash
removes all sulfuric acid upstream of the spray dryer, or whether any sulfuric
acid present is removed on the ash collected on the heated filter in the
sampling train.  At the outlet of the fabric filter though, it is clear that
no sulfuric acid is present.  Measurements could not be made at the spray
dryer outlet due to the high grain loading at that point, as the upstream
filters tended to plug before an appreciable amount of flue gas could be
sampled.

     Measureable levels of sulfuric  acid were found during the high sulfur
test periods.   Spray dryer inlet values were measured at 2 to 6 ppm 803.
Fabric filter  outlet values varied from 0.1 to 0.5 ppm.  On a limited number
of days where  the spray dryer inlet  and fabric filter outlet 803 concentra-
tions were measured simultaneously,  removal efficiencies of 90 percent or
better across  the system were indicated.

SUMMARY AND CONCLUSIONS

     Based on preliminary data reduction for the 3-month test program on the
NSP Riverside  dry FGD system, the following conclusions are apparent:

          •    In general, the Riverside  system ran quite well.  None of
               the problems anticipated for spray dryer systems, such as
               rotary atomizer wheel pluggage, buildup of wet solids on
               dryer vessel walls, or wetting of fabric filter bag sur-
               faces during upset conditions, were observed.
                                    10-128

-------
          •    Some problem areas at Riverside appear to be potential
               sources of problems on other similar dry FGD systems.
               These include typical problems with mixing and pumping
               slurries with a high solids content, solids handling
               equipment which requires continual maintenance, and some-
               times inadequate atomizer protection during upset condi-
               tions.

          •    At sulfur levels up to a nominal 3.5 percent,  high SC>2
               removal efficiencies (90 percent) were readily achievable
               in the  relatively short-term periods of this program.   For
               the low sulfur Sarpy Creek coal/coke mixture,  substoichio-
               metric  amounts of lime were required even at 90 percent
               S02 removal.  This was attributed to the alkaline nature
               of the  Sarpy Creek coal ash.  For the high sulfur Illinois
               coal, 90 percent SC>2 removal required reagent  ratios of
               approximately 1.3 to 1.4 moles lime per mole of inlet  S02.

          •    Calcium chloride addition to the atomizer feed slurry  to
               achieve chloride levels of approximately 1 percent in  the
               fabric  filter solids catch appeared to be successful in
               promoting lime utilization.  For the high sulfur tests,
               the lime reagent ratio to achieve 90 percent SC>2 removal
               was reduced from 1.3 to 1.4 down to a range of 0.9 to  1.1
               moles lime per mole of inlet SC>2.  This chloride level
               would correspond to 0.3 percent chloride in a  nominal  3.5
               percent sulfur coal.  Even for a low chloride, high sulfur
               coal, high chloride levels achieved through calcium chlor-
               ide addition appear to be cost effective for reducing  lime
               consumption.

          •    Particulate control efficiencies were high throughout  the
               test program, maintaining outlet grain loadings well below
               required levels.  In spite of baghouse operation within
               18°F of the adiabatic saturation temperature and very  high
               baghouse inlet grain loadings, no bag-fabric-related
               problems were observed and flange-to-flange pressure drop
               remained acceptably low.

          •    Based on a limited number of simultaneous measurements
               during  the high sulfur test periods, sulfuric  acid removal
               levels  of 90 percent or greater were observed.

                                  REFERENCES

1.   Karlsson,  Hans T., et al., "Activated Wet-Dry Scrubbing  of S02."  Jour-
     nal of the Air Pollution Control Association.    Volume  33, No.  1, Jan-
     uary 1983.  pp. 23-28.
                                    10-129

-------
Hansen, Svend Keis,  et  al.   "Status of the Joy/Niro  Dry  FGD System and
Its Future Application  for  the  Removal of High Sulfur, High Chloride and
NOX from Flue Gas."   Publication No.  83-JPGC-APC-8,  The  American Society
of Mechanical Engineers,  New York,  NY.  October,  1983.

Chemical Marketing Reporter.     Volume 224, No.  15,  October 10,  1983.
Schnell Publishing Company,  Inc., New York, NY.
                              APPENDIX

   CONVERSION OF BRITISH ENGINEERING  (ENGLISH)  UNITS  TO  SI  UNITS
To Obtain
ng/J
m
m2
m3
mg / Nm
°C
KW
joules /kg
m / sec/m
tonnes
kg
m / sec
nr / s e c
From
lb/106 Btu
ft
ft2
ft3
gr/dscf
°F
hp
Btu/lb
cfm/ft2
tons (short)
Ib
cfm
gpm
Multiply by
(or use equation)
430

0.3048
9.29
2.83
1.83
tc (°C) = 5/9[tf(°
0.746
1.33
5.08 x
0.907
0.454
4.72 x
6.31 x
x 10~2
x 10~2

F) - 32]

x 10~4
io-3


10~4
10-5
                             10-130

-------
OVERVIEW AND EVALUATION OF TWO YEARS OF OPERATION
 AND TESTING OF THE RIVERSIDE SPRAY DRYER SYSTEM

     J. M. Gustke, W. E. Morgan, S. H. Wolf

-------
             OVERVIEW AND EVALUATION OF TWO YEARS OF OPERATION
                       OF THE RIVERSIDE SPRAY DRYER SYSTEM

                 by:   John M. Gustke and Wayne E. Morgan, Ph.D.
                       Black & Veatch, Engineers-Architects

                       Steven H. Wolf
                       Northern States Power Company

                                       ABSTRACT

      Initial operation of the  100 MW spray dryer system at Northern States Power Company's
Riverside Generating Station began in November 1980. At that time, a comprehensive test program
was initiated to  demonstrate  the suitability of dry flue gas desulfurization (FGD) for utility coal
fueled boilers. Since initial operation of the Riverside spray dryer began, other publications have
described  individual aspects of this  system's performance. This  paper  provides a comprehensive
analysis and  overview  of the performance  test data collected during the initial  two-year period
of operation of the Riverside spray dryer.

      Overall data  correlations describing important variables in spray dryer operation  and per-
formance  are established from test results obtained during operation of the Riverside spray dryer
system under a wide range of conditions. Correlations between total and fabric filter sulfur dioxide
(802) removal and parameters such as lime stoichiometric ratio, total alkalinity, and approach
temperature are  presented for several different coals. Variations in moisture content of the solids
collected in the spray dryer and fabric filter are evaluated to establish the sensitivity to a wide range
of operating  variables. In addition,  system operation  and control  experiences are described to
illustrate the interaction between  flue gas  flow, feed slurry flow,  absorber outlet temperature,
and SO2 emissions during normal operation, as well as during transient conditions such as start-up,
shutdown, and load swings. The effects of boiler soot blowing on the flue gas saturation tempera-
ture and system control are also discussed.
                                     INTRODUCTION

      Joy Manufacturing  Company, Niro  Atomizer  Incorporated, and  Northern States Power
Company (NSP) jointly funded the 100 MW spray dryer demonstration program at the Riverside
Generating Station. Design support for the spray dryer installation was provided by Black & Veatch,
Engineers-Architects. The  purpose of the Riverside demonstration program was to confirm that a
full size spray dryer installed on a utility coal fueled power station could reliably and economically
meet current SOo  and particulate emission regulations.  A single,  full-size spray absorber module
followed by a fabric filter was installed and began initial operation at Riverside Units 6 and 7 in
November 1980. Testing of the Riverside system began in March 1981 and remains in progress.
                                         10-131

-------
      This  paper presents an overview of all performance test  results obtained  during 1981 and
1982. Overall data correlations are developed and relationships are established describing the sulfur
dioxide removal efficiency and waste solids drying performance of the system. Fairly wide varia-
tions  in some  of the  performance test results were observed over the extended test period. The
observed range of performance appropriately reflects the actual  performance of a full scale utility
dry scrubber system subject to normal variations in operating conditions. Operational performance
and control characteristics of the  system are also presented. Major system parameters which are
monitored  and controlled during start-up, shutdown, and rapid load changes are discussed. The im-
pacts  of steam generator soot blowing on operation of the Riverside spray dryer are also described.

                                  SYSTEM DESCRIPTION

      Riverside  Units  6 and 7  have identical Babcock & Wilcox  steam generators  which began
operation in 1949  and  1950, respectively. The combined  generating capacity  of these units is
approximately  100 MW. The design fuel for the steam generators  is a 10,800  Btu/lb Illinois  coal
with  a  sulfur content of 3.5 per cent. Prior to the installation of the spray dryer and fabric filter
system, each unit had  individual electrostatic precipitators, ID fans, and chimneys.

      The  retrofit spray dryer system at  Riverside treats the combined flue gas flow  from Units 6
and 7.  Figure 1 presents a simplified flow diagram of the spray dryer system. Flue gas at the air
heater outlet from both units is combined and drawn through the single spray dryer module. Design
features of the  46-foot diameter spray dryer module include a compound gas disperser and a single
rotary atomizer. After flowing through the spray dryer, flue gas is directed to either the fabric filter
or the original weighted wire precipitators. The fabric filter uses reverse gas cleaning and has twelve
compartments.  The gas flow at the discharge of the particulate collectors is divided and flows to the
original Unit 6 and 7 chimneys.

      Both a ball mill and an attrition slaker are  installed for lime slurry preparation. Lime slurry is
fed to a mix tank  where it is mixed with recycled material from the  spray dryer and particulate
collector before being pumped to the atomizer. The recycle material is collected from the spray
dryer and  particulate  collector (fabric filter or  electrostatic precipitator) hoppers and consists of
flyash,  unreacted lime, and flue gas desulfurization reaction products.  Material collected in the
spray dryer and particulate collector hoppers which is not  recycled is stored in silos prior to dry
landfill disposal.

                             PERFORMANCE TEST RESULTS

      During  1981 and  1982,  164 performance tests were conducted at  Riverside. Most of the
testing during this time consisted of short term parametric tests which were conducted to determine
the influence of various individual parameters on system performance.  Longer term system per-
formance was evaluated by conducting five sets  of demonstration tests. During the demonstration
test periods, the system was maintained at relatively stable operating conditions for 5 to 10 days.
Figure 2 summarizes  the schedule  for the performance tests which comprise  the  data presented
herein.

      During the two years of testing at Riverside, the operating conditions  varied widely.  Five
different fuels were tested: Colstrip coal, Colstrip/Coke blend, Sarpy Creek coal, Sarpy Creek/Coke
blend, and  high sulfur Illinois coal. The  Colstrip and Sarpy Creek  coals  are low  sulfur  coals from
                                          10-132

-------
                Flue Gas
              From Unit 7
                 Boiler
               Flue Gas
              From Unit 6
                Boiler
o
I
LO
OJ
                    Water


Fabric
Filter
1 1


                                      Flue Gas to
                                      Chimneys
                             Solids to
                             Disposal
                                                       —111— Water
                                                         Mix
                                                         Tank
                          Classifier
K7-
  Slurry
 Transfer
  Pumps
Feed
Tank
K7-
  Feed
  Pumps
                                              Lime Slurry
                                                Pumps
              FIGURE 1   RIVERSIDE SPRAY DRYER SYSTEM

-------
o
Spray Dryer
    System Slart-Up
Parametric Tests,
    Colstrip Coal
Demonstration Tests,
    Colstrip Coal
Electrostatic Precipitator
    Tests, Colstrip Coal
Parametric and
    Demonstration Tests,
    Illinois  Coal
Demonstration Tests,
    Sarpy Coal
Miscellaneous Tests,
    Various Operating
    Conditions
Extended Units 6 and 7
    Outage
Parametric and
    Demonstration Tests,
    Sarpy Coal
                                           1980
                                         O ,  N , D
             1981
F ,M| A |M| 1  , 1  , A , S
                                                                      D
                1982
1IF|M|A|M|||||AISI0,N|D
              FIGURE 2   RIVERSIDE SPRAY  DRYER  PERFORMANCE TEST SCHEDULE

-------
Montana. Table 1 presents some of the typical properties for each coal. Tests were performed with
gas flows at the absorber inlet  ranging from 34 to 114 per cent of design flow (170,000 acfm to
570,000 acfm). Approach temperatures ranged from 16 to 78 F. Feed slurry was atomized at fine,
intermediate, coarse, and very coarse levels.

      This paper focuses  on the effects of a range of variable operating conditions on SOo removal
performance and waste moisture content (dryness of the solid waste products). The test results are
primarily viewed on the whole rather than  as isolated individual groups of data or periods of testing
to establish specific correlations. The results presented in this analysis should reflect the effects on
performance of a wide range of operating  conditions judged representative of those that full scale
units will routinely experience.

                              SO2 REMOVAL CORRELATIONS

LIME STOICHIOMETRIC RATIO

      The ability to economically achieve SO2 removal efficiencies required by New Source Per-
formance Standards (NSPS) is a requirement  for any flue gas desulfurization (FGD) system. As a
measure of the Riverside dry scrubber performance, total SOo removal versus lime stoichiometric
ratio (moles of lime added per mole of SO2 removed) was plotted for all of the 1 64 performance
tests. A multiple linear regression analysis  was used to establish correlations between SOo removal,
lime stoichiometric ratio, and approach temperature. Figure 3 presents the results of this analysis,
applied to all 164 tests. The multiple linear  regression technique of data correlation utilizes test data
to generate a correlation between related parameters. Multiple linear regression is not a curve fitting
technique. The curves presented on Figure 3 were generated using the regression equation estab-
lished  for total SO2 removal and lime stoichiometric ratio. Curves were calculated  for approach
temperatures of 18 F  and 40 F. These approach temperatures are typical of low and high approach
temperatures frequently selected as design points for utility spray dryer installations. Since the tests
were performed at a variety of approach temperatures, the distribution of test results  around  the
regression curves is shown for comparative purposes only.

      As shown on Figure  3, both the lime stoichiometric ratio and approach temperature affect
the SOo  removal.  For a  specific lime stoichiometric ratio, decreases in the approach temperature
resulted in  increased SOo removal.  The form of the regression equation used to generate  correla-
tions between SO2 removal and  lime stoichiometry is as follows:

                       SRE=
                 where
                 SRE = SO2 removal efficiency, per cent
                 A, B = Regression coefficients
                 LSR = Lime stoichiometric ratio
                 AA = Ash Alkalinity constant
                 AT = Approach temperature, degrees F
                 K = Regression constant

      Table 2 lists the coefficients and constants for the regression equation shown above, as well
as the number of tests conducted using each fuel and the average SO2 removal efficiency. The value
                                           10-135

-------
        TABLE 1. TYPICAL PROPERTIES OF COALS AND FUEL BLENDS BURNED
                        AT RIVERSIDE DURING 1981 AND 1982
Fuel


Colstrip

Colstrip/Coke

Sarpy Creek

Sarpy Creek/Coke

Illinois
Heating
Value
Ib/MBtu
8,700
9,200
8,300
9,200
10,900
Ash
Content
per cent
8.3
6.7
10.8
8.9
10.4
Sulfur
Content
per cent
0.8
1.2
0.8
1.2
3.2
          TABLE 2.  REGRESSION EQUATION COEFFICIENTS FOR SO2 REMOVAL
           AS A FUNCTION OF LIME STOICHIOMETRIC RATIO AND APPROACH
                       TEMPERATURE FOR ALL RIVERSIDE TESTS
Fuel


Colstrip/Coke

Sarpy Creek/Coke

Illinois**

Sarpy Creek

Sarpy Creek (no lime)

Colstrip

All Fuels
Tests
Average SO2   	
Removal       A
94
24
22
17
4
3
164
85.9
88.0
92.0
75.4
30.6
69.3
84.3
                                                     Coefficients/Constants
                                               -16.9

                                               -79.3



                                               -29.2
                                                        B
        AA
K
Standard
Error of
Estimate
-0.65    0.42    114.9   9.1

-0.38    0.42    148.5   6.1

-0.36             99.9   3.9

-0.21    0.42    105.0   6.9
                                                        -0.66   0.42     124.8   8.6
       *The value for the ash alkalinity constant (AA) was established by optimizing the "all fuels" regression equation. The
 addition of such a constant to the other equations also improved the data correlations, and the same value was assumed for each case.
       "For the limited range of data available, no significant correlation with lime stoichiometric ratio was found.
                                          10-136

-------
o
I
to
           100-
            80-
         c
         V
         U
O.

_*   60
ro
>
O
E
o»
at


6   40
            20-
                   Approach Temp. = 18 F
                                 Approach Temp. = 40 F
                   The Regression Equation Used to Develop the Specific

                   Curves on This Figure Accounts for Variations in Test

                   Conditions for Each Displayed Data Point.
                                                                   - - Approach Temperature Below 25 F

                                                                    - Approach Temperature Above 25 F
              0.0
                0.2
0.4        0.6       0.8        1.0       1.2        1.4


   Lime Stoichiometric Ratio (Based on SO2 Removed)
1.6
1.8
                                                                                                              2.0
         FIGURE 3  SO2 REMOVAL AS A FUNCTION OF STOICHIOMETRIC RATIO

                    AND APPROACH TEMPERATURE (ALL FUELS)

-------
of the ash alkalinity constant (AA) was selected by optimizing the regression equation using all of
the test data. Inclusion of this same constant in the individual fuel regression equations improved
these  correlations  as  well.  Insufficient data were available to optimize this  value for each case.

      Based on an  analysis of the entire range of data, factors such as degree of atomization, gas
flow,  SO-,  inlet concentration,  gas retention time, and amount of recycle could not be directly
correlated" to SCh removal for  the limited variation in these  values tested at Riverside. However,
pilot testing has demonstrated  that factors such as these are significant in spray dryer design and
performance. The results show that for the full scale Riverside spray dryer system operating over a
range of conditions, stoichiometric ratio, approach temperature, and ash alkalinity are the dominant
process  parameters which control SC>2 removal.

      A separate regression analysis was performed on each fuel to determine individual contribu-
tions  to the overall regression equation.  Figures  4 through 6 show the SC>2 removal efficiency
predicted by regression equations obtained  for each fuel,  plotted along  with the respective data
points.  The Colstrip/Coke test data shown on Figure 4 correspond primarily with the mid-range of
the SO-> removal and lime stoichiometries for the overall  correlation presented on Figure 3. The
Sarpy Creek coal  test data shown on  Figure 5 generally correspond to the lower SC>2 removals and
stoichiometries in the data base. The Illinois coal test data shown on Figure 6 comprise most of the
high SO-) removals and corresponding  high lime stoichiometries in the data base. The regression
equations obtained from the analysis  of test results from each fuel, indicate varying sensitivities to
approach temperature and lime stoichiometric ratio. The Sarpy Creek coal is the most sensitive to
lime stoichiometry  but the least sensitive to apporach temperature. The Illinois coal did not exhibit
much sensitivity to stoichiometry but  was sensitive to approach  temperature. The Colstrip/Coke
blend was  the most  sensitive to  approach temperature and  it also  was moderately  sensitive to
stoichiometric ratio.

TOTAL ALKALINITY RATIO

      The  alkalinity ratio is a measure of the total  alkaline material which is fed to the atomizer. In
addition to the fresh lime, the alkaline  material includes alkalinity in the ash and unreacted lime in
the recycled solids. Since the lime stoichiometric ratio provided a  good correlation with total S02
removal, the alkalinity ratio was expected to provide a better correlation. Total SOn removal and
the corresponding  alkalinity ratio for all 164 tests are shown on Figure 1'. The overall trend of data
on this  figure  indicates that there is a correlation between  SO2 removal and  alkalinity ratio. How-
ever, a  significant  number of tests have alkalinity ratios of less than  1.0. Since alkalinity ratio is
defined  in  this case as equivalent  moles of CaO per mole of SO2 removed, the minimum theoretical
alkalinity ratio is  1.0. During the Riverside test  program  two different techniques were used to
measure the total  alkalinity ratio. The results obtained using  each technique differed and neither
technique appeared to provide consistent  and reasonable results. Thus, no correlations were made
using this parameter.

S02 REMOVAL IN THE PARTICULATE  COLLECTOR

      The  total SO2 removal which occurs across a dry scrubber system consists of SO2 removed in
the spray dryer and SO2 removed in the particulate collector. SO2 removal in the particulate col-
lector results from the additional  contact  of the alkaline solids with the remaining SO2 in the flue
gas. SO2 removal in the  particulate collector can be affected  by process parameters in a different
                                          10-138

-------
o
I
            100-
         «-   80
         c
         
-------
                 100
o
i
             e
             lit
             U
             k.
             01
             Q.
             O
             E
             UJ
             at
                 80
                 60-
                 40-
                 20-
                   0.0
                                        Approach Temp. = 18 F
                                        Approach Temp. = 40 F
                        The Regression Equation Used to Develop the Specific
                        Curves on This Figure Accounts for Variations in Test
                        Conditions for Each Displayed Data Point.
                                                 - Approach Temperature Below 25 F
                                                 - Approach Temperature Above 25 F
0.2
0.4       0.6       0.8        1.0       1.2        1.4

   Lime StoichSometric Ratio (Based on SO2 Removed)
1.6
1.8
            FIGURE 5   SO2 REMOVAL AS A FUNCTION OF STOICHIOMETRIC RATIO
                        AND APPROACH TEMPERATURE (SARPY CREEK COAL)

-------
o
I
            c
            01
            U
            O

            E
            O
            1/5
                100-
                 80 -J
                 601
                 40-
                 20-
                   0.0
                                                           Approach Temp. = 18 F


                                                           Approach Temp. = 40 F
The Regression Equation Used to Develop the Specific

Curves on This Figure Accounts for Variations in Test

Conditions for Each Displayed Data Point.
                                       - Approach Temperature Below 25 F


                                       - Approach Temperature Above 25 F
     0.2
0.4        0.6       0.8       1.0       1.2       1.4


    Lime Stoichiometric Ratio (Based on SOz Removed)
1.6
1.8
             FIGURE 6  SO2 REMOVAL AS A FUNCTION OF STOICHIOMETRIC RATIO

                        AND APPROACH TEMPERATURE (ILLINOIS COAL)

-------
o
I
           100-
         c
         Ol
         U
         >
         o

         E
         OJ
         (V
         *E^

         O
            80-
            60-
            40-
            20-
              0.0
0.2
                                                                                           C CD
                                                                                 3 ,


                                                                                <»^
                                                     - Approach Temperature Below 25 F


                                                     - Approach Temperature Above 25 F
0.4        0.6       0.8       1.0        1.2



         Alkalinity Ratio (Based on SO2 Removed)
1.4
1.6
1.8
2.0
        FIGURE 7   SO2 REMOVAL AS A FUNCTION OF ALKALINITY RATIO

                    AND APPROACH TEMPERATURE (ALL FUELS)

-------
manner than SC>2 removal in the spray dryer. To determine the sensitivity of SOo removal in the
particulate collector to various process parameters, a separate analysis was performed.

      Figure 8 presents  total and fabric filter SC^ removal efficiency as a function of lime stoichi-
ometric ratio. As shown in this figure, fabric filter SC>2 removal efficiencies generally range between
10 to 20 per cent and average approximately 15 per cent. Figure 9 presents total and electrostatic
precipitator  SOo removal as a function of lime stoichiometric ratio. As shown in this figure, electro-
static precipitator (ESP) SC>2 removal ranges from 0  to 20 per cent and averages approximately 6
per cent. Higher average SC>2 removal efficiencies in  the fabric filter than in the ESP are attributed
to more effective interaction between the flue gas and the particulate in the filter cake on the filter
bag surfaces. Although measurements were not consistently available from the  performance test
data,  factors such as fabric filter gas-to-cloth ratio, pressure drop, and frequency of bag cleaning are
expected to  have a significant effect on fabric filter SC>2 removal.

      An analysis was performed  to determine the sensitivity of fabric filter SO-,  removal to the
inlet SC>2 concentration. Figure  10 presents a comparison of these parameters for all the fuels. No
significant correlation could  be found in the Riverside data between  fabric filter SC>2 removal and
inlet  862 concentration. A  similar analysis performed on each fuel individually did not yield an
improved correlation. Figure  1 1  presents a comparison of the Riverside fabric filter  SC>2 removal as
a function of approach temperature for all the fuels.  Flue gas approach temperature to the dew
point establishes the relative  moisture content of the flue gas. High flue gas moisture content should
result in improved SC>2 removal in the fabric filter. Contrary to what was expected, there appeared
to be no correlation between fabric filter SC>2 removal and approach temperature. A similar analysis
was performed on each of the fuels; however, no significant correlations were obtained. Correlations
between fabric filter SC>2 removal and approach temperature have been shown to exist in laboratory
and pilot scale units. However,  on a full scale operating unit, variations in other process parameters
such  as fabric filter gas-to-cloth ratio and frequency of cleaning may diminish the influence of
approach temperature on SC>2 removal.

      An interesting correlation which was apparent from the test data was the correlation between
fabric filter  SC>2 removal and spray down temperature. Spray down  temperature is the difference
between the spray  dryer inlet and outlet temperatures.  Spray down temperature  determines the
total  moisture content of the flue gas but is independent of the dew point temperature.  Figure 12
shows the regression equation curve established  for these parameters. As shown on Figure  12, fabric
filter  SOo removal  appears  to  increase  with  increased spray down temperatures.  The composite
regression equation  developed  for all fuels (all tests with the  fabric filter) for  fabric filter SC>2
removal and spray down temperature is as follows:
                                         A
                             FFSRE =       + K,
                 where
                 FFSRE = Fabric filter SO-» removal efficiency, per cent
                 A = Regression coefficient (-2204.6)
                 SDT = Spray down temperature, per cent
                 K = Regression constant (28.2)

      Analysis of fabric filter SC>2 removal versus spray down temperature for each individual fuel
did not produce improved  correlations. The Riverside spray dryer is operated to maintain a fixed
                                           10-143

-------
O
I
            100
         c
         0)
         U



         I  80
         O

         E
         a;

         ^  60 H
.2
3


"O
c
            40-
        •    20 ^F
        J3
        fO
              0.0
                               dC
                                cS-'c
                                                               ;§;


                                                                CE

  Total SOi Removal Efficiency


  Fabric Filter SO? Removal Efficiency
                             A± A

                                               *
0.2
                          0.4        0.6        0.8        1.0        1.2        1.4


                             Lime Stoichiometric Ratio (Based on SO2 Removed)
1.6
1.8
2.0
        FIGURE 8   TOTAL AND FABRIC FILTER SO2 REMOVAL AS A FUNCTION OF

                    LIME STOICHIOMETRIC RATIO (ALL FUELS)

-------
h-1
O
I
I—"
-P-
L/i
         c
         U  100
         I   80
         n
         QC
         rs
         O
             60
         13
40-
          .
         'w
         0)
         «   20H
         o
         ^^
         u
               0.0
              - Total SO2 Removal Efficiency

              - Electrostatic Precipitator SOa

               Removal Efficiency
                                                                                                CD
            0.2
0.4        0.6        0.8        1.0        1.2       1.4


    Lime Stoichiometric Ratio (Based on SO? Removed)
1.6
        FIGURE 9  TOTAL AND ELECTROSTATIC PRECIPITATOR SO2 REMOVAL

                    AS A FUNCTION OF LIME STOICHIOMETRIC RATIO (COLSTRIP/COKE FUEL)
1.8
2.0

-------
O
I
              c
              0*
              U
              O
              E
                  24-i
                 20-
                 16-
                 12-
                  8-
                  4-
                                O

                               O
                                    «>
                                    CD
    a + c c
1     T>
 O    ^O  CD
 C CD       O<
                                                 CDS
                 CDC"-
                                                        c - Colstrip/Coke

                                                        A- Illinois

                                                        -I-- Sarpy Creek

                                                        «- Sarpy Creek/Coke


                                                        x- Colstrip
                             80
     160       240       320       400       480       560


            Fabric Filter Inlet SCh Concentration, PPM
640
720
            FIGURE 10  FABRIC FILTER SO2 REMOVAL AS A FUNCTION OF

                        FABRIC FILTER INLET SO2

-------
O
I
c
U
              O
              E
              - Colstrip/Coke
                                                •- Illinois
                                                 - Sarpy Creek
                                                 - Sarpy Creek/Coke
                                                 - Colstrip
                                                                                    45
                                                                                50
                                                                55

-------
                  24
o
i
CO
              c
              
-------
proportion of solids and water in the atomizer feed. Thus, increasing the spray down temperature
also increases the amount of particulate matter in the flue gas at the fabric filter inlet. From the
Riverside data it appears that the greater the amount of particulate matter entering the fabric filter,
the greater the amount of SC>2 removed in the fabric filter. Higher particulate concentrations in the
fabric filter may provide a higher ratio of alkaline material on the bags which improves SO-> removal
efficiency.  Other parameters such as gas-to-cloth ratio, pressure drop, and cleaning frequency may
also contribute to this correlation.

      Due  to the small size of the data base, no attempts were made to correlate precipitator SOo
removal with any of the process parameters. As previously discussed, the average SO^ removal was
less in the  ESP than in the fabric  filter. However, the Riverside precipitators are relatively small and
were not designed for spray dryer  service. New dry scrubber installations using precipitators may
achieve slightly higher SO2 removal efficiencies in  the particulate  collector than those achieved
using the precipitators at Riverside.

                            WASTE MOISTURE CORRELATIONS

      The  ability of the spray dryer to produce a waste product which is relatively dry is a critical
requirement  of the  system. As a measure of the ability of the Riverside dry scrubber to achieve a
dry product, the moisture content of the solids discharged from the spray dryer cone and fabric
filter hoppers was periodically measured during each performance test.

      The  average and peak moisture contents of the solids recorded during the performance tests
are as follows.

                                               Average Moisture        Peak Moisture
                                               Content	       Content	
                                               (per cent)              (per cent)

      Spray Dryer Solids                        1.5                    4.0

      Fabric Filter Solids                        1.0                    2.3

      Generally, the  solids  were sufficiently dry to  be  conveyed  by conventional ash handling
equipment.

      To determine the effect of major process parameters on the waste moisture content, a com-
parison of spray dryer waste  moisture and approach temperature was made. Figure 13 shows the
corresponding approach temperature for each waste  moisture content. Although there is a general
trend for decreasing waste moisture with increasing approach temperature, a strong correlation was
not apparent from the Riverside test data. Since approach temperature is a relative indication of the
amount of moisture in the flue gas, it was expected to correlate with spray dryer waste moisture.
Additional comparisons  of  waste moisture with spray  down temperature,  feed solids  content,
atomization, and mass flow rate were made. In each case, no strong correlation with waste moisture
was  apparent. A similar analysis  was performed on each fuel, but no significant  correlations were
found. The range of recorded spray dryer waste moisture at Riverside  is very narrow. Correlations
may exist between these parameters; however, on a full scale operating unit, the range of operating
conditions would not normally fluctuate enough for these correlations to be apparent.
                                           10-149

-------
                  3.0
o
i
Ui
o
              c
              
                 CD © ©
                 CD CD
                 A
                                                CD   CD
                                                        C- •*   CD
                                                        O   CD
                                                                    O
                                                              a   o
                                                                      O
                                                                      CD
                                                                      CD
                                                                      O
                                                    1  o
                                                                         CD CD
                                                                         CD
                                                                                                         CD
                                                                                       CD
                                                                             CD - Colstrip/Coke
                                                                             A - Illinois
                                                                             + - Sarpy Creek
                                                                             «•- Sarpy Creek/Coke
                                                                             x - Colstrip
            15
                           20
25        30        35        40
     Approach Temperature, F
45
50
55
             FIGURE 13   SPRAY ABSORBER WASTE MOISTURE AS A FUNCTION OF
                          APPROACH TEMPERATURE

-------
      One correlation which was apparent from the test data was the correlation between spray
dryer and fabric filter waste moisture. Figure 14 shows this correlation. As presented on Figure 14,
fabric filter  waste  moisture  content increases linearly with spray dryer waste moisture.  The com-
posite equation developed for all fuels (all tests) is as follows:

                             FFWM = A * SDWM,

                  where
                  FFWM = Fabric filter waste moisture, per cent
                  A = Regression coefficient (0.67)
                  SDWM = Spray dryer waste moisture, per cent

      This correlation reflects the longer  drying time of the solids in  the fabric filter. Flue  gas
passing through the filter cake on the  fabric filter bags provides additional drying of the collected
solids.

                     OPERATIONAL PERFORMANCE AND CONTROL

SYSTEM CONTROL

      A discussion of the primary control parameters is necessary before discussing the  effects of
start-up,  shutdown, and rapid load swings on the performance of the spray dryer. Control of  the
Riverside spray dryer  is  based on  maintaining a constant approach temperature  at the spray dryer
outlet. The approach temperature is held constant by maintaining the required evaporation rate in
the spray dryer to quench the flue gas to the desired temperature above the adiabatic saturation
temperature. Moisture is added to the gas stream through the atomizer with the additive feed slurry,
while the fraction  of solids in the atomizer feed is maintained at a constant level of approximately
35 per cent. To decrease the approach temperature, the total  flow to the atomizer is increased.

      To control the  SO2 removal efficiency  of the dry scrubber, the  ratio of lime and recycled
solids in the feed slurry  is changed, but the total amount of  solids and water in  the atomizer feed
remains constant. To increase SOo removal, the ratio of the lime to feed solids is increased  and  the
amount of recycle  material is decreased. Conversely, to  decrease SO-> removal, the amount of lime
in the atomizer feed is decreased and the amount of recycle  material is increased. Due to the resi-
dence time in  the  slurry preparation system, changes in SO2 removal efficiency can not be made
instantaneously unless a change in approach temperature is also desired. Since the approach tem-
perature  has a significant effect  on  SO2  removal, it is difficult to stabilize the absorber outlet
temperature  if approach temperature  is used to control SO->  removal. However,  since only an
average SO2 removal rate is required,  periods of low SO2 can be offset by periods of higher SO2
removal.

SYSTEM START-UP

      The start-up of the Riverside spray dryer can be accomplished whenever the flue gas tempera-
ture and flue gas flow rate are high enough to allow proper drying in the chamber. At Riverside, a
minimum gas flow of 125,000  ft^/min (25  per cent gas flow)  at 200 F is necessary to meet the
minimum requirements for spray  dryer operation. Start-up of the spray dryer is performed by an
automatic control  system which requires only limited operator interface once  feed slurry  flow is
                                          10-151

-------
o
I
               c
               01
               U
               ^
               01
               a.
              '5
               0>
               o
               E
                          The Regression Equation Used to Develop the Specific

                          Curve on This Figure Accounts for Variations in Test

                          Conditions for Each Displayed Data Point.
                                                                                  - Colstrip/Coke

                                                                                  - Illinois

                                                                                  - Sarpy Creek

                                                                                  - Sarpy Creek/Coke

                                                                                  - Colstrip
                                         1.0        1.5        2.0        2.5        3.0        3.5


                                                Spray Absorber Waste Moisture, Per Cent
             FIGURE 14   PARTICULATE REMOVAL DEVICE WASTE MOISTURE AS A FUNCTION OF

                          SPRAY ABSORBER WASTE MOSITURE

-------
established to the atomizer. Figure 15 shows a typical warm start-up of the spray dryer at Riverside.
As shown on Figure 15, the gas flow is increased as the boilers come on line, and 807 is generated
when coal firing begins. The  feed slurry flow to the  atomizer is initiated when the minimum gas
flow and temperature  are reached. SC>2 removal begins immediately when feed slurry flow is estab-
lished. As the boilers continue to gain load, the slurry flow to the atomizer increases accordingly to
maintain the desired absorber outlet temperature. When the absorber outlet temperature has stabi-
lized, the automatic SC>2 removal control system adjusts the amount of lime in the feed to achieve
the desired SC>2 removal. The time required for complete stabilization at the desired SC>2 removal is
dependent primarily on the composition of the feed  stock prior to start-up. The time to reach stable
SC>2 removal also varies for hot, warm, or cold steam generator start-ups, because of differing initial
flue gas temperatures.

SYSTEM SHUTDOWN

      The flue gas  flow and temperature requirements used for the shutdown of the spray dryer are
the same as  for the start-up, but it is easier to reduce transient 862 peaks during a controlled boiler
shutdown because  sufficient gas flow and temperature are present to allow spray drying until after
the steam generator fire  is extinguished. The feed slurry to the  atomizer is stopped when the gas
flow is no longer sufficient to ensure proper drying and the atomizer is flushed and taken out of
service. As shown in Figure 16, the spray dryer follows load as the  boilers are ramping down and
comes off line without SC>2 emissions increasing.

RAPID LOAD SWINGS

      The Riverside spray dryer is often subject to  rapid load swings. As shown in Figure 17, the
system automatically responds to the increased  and decreased gas flow. The feed  slurry flow rate
follows closely to the  gas flow, resulting in an almost constant outlet temperature. The load change
illustrated on this figure was accomplished by bringing one of the boilers on line after the other was
operating at  steady load. This results in some dilution of the SO2 concentration into the spray dryer
and accounts for the reduced emissions during this period.

WET BULB TEMPERATURE FLUCTUATIONS

      To control a spray dryer at a constant approach temperature, it  is necessary to measure the
adiabatic saturation (wet bulb) temperature of the flue gas. At Riverside, wick-and-water reservoir
devices are installed at  the fabric filter outlet to measure and transmit wet bulb temperature to the
control system. It is necessary to locate  these devices downstream of the fabric filter to minimize
the fouling of the wick by particulate matter. Although the wet bulb measurement devices provide
an  accurate  and reasonably reliable  indication  of  the wet  bulb temperature  downstream of the
baghouse, they do  not  provide precise information  on the wet bulb  temperature in the spray dryer
chamber. Fluctuations in wet bulb temperature  at the spray dryer inlet are not transmitted to the
control system until 30 to 90 seconds  after they  occur. In addition, by the time  the gas passes
through the fabric filter and reaches the outlet duct,  the heat transfer and air leakage reduce the wet
bulb temperature by 1  to 3 F from that at the spray  dryer inlet.

      To determine the type of wet bulb fluctuations which may occur at the spray dryer inlet, a
temporary device was  installed and  continuously  monitored. Figure  18 presents  the wet bulb
temperature  fluctuations at the inlet of the spray dryer during a typical soot blowing sequence using
                                         10-153

-------
                       350,000
o
i
              140
              120
              100
           Q.
           U
              80
           _

           1/1 60
           -o
            OJ
              40
                       300,000 -
                       250,000 -
  200,000 -
                    £
                    O
•5 150,000 -
                      100,000 -
              20 -     50,000 -
                           0
                                                                                           Gas Flow
                                                                   Feed Slurry Flow
   Spray Dryer

Outlet Temperature
                                                                               SO2 at Chimney
                                                                                                  J	L
                                                                                             275
                                                                                              250
                          225
                                                                                                                  200   E
175
                                                                                              150
                                                                                                                  125
3
O

£
k.
Q
                                                                                                                       O.
                                                                                                                       (SI
                                                                                                                             700
                                                                                                         600
                                                                                                         500
400  c
     E
     IE
     U

300  ™
                                                                                                         200
                                                                                                                             100
                            0      .5      1.0     1.5     2.0     2.5     3.0     3.5     4.0     4.5     5.0     5.5    6.0

                                                                   Time (hr)


           FIGURE 15  RIVERSIDE SPRAY DRYER SYSTEM  RESPONSE

                         DURING START-UP
                                                                                              100

-------
                           350,000 -
o
I
                 140
                 120
                 100
 S    80
 o
J    60
-c
£    40


     20
                       _,  250,000
                                                             Feed Slurry Flow
10    15     20
                                                             25    30    35

                                                               Time (min.)
                                                                 40
                                                                       Spray Dryer
                                                                         Outlet
                                                                       Temperature
                                                                                          »     T
                                                                                          275
                                                                                                       250  ~
                                                                £
                                                                2
                                                           225  |
                                                                Q.
                                                                E
                                                                0)
                                                           200  T,
                                                                                              O
                                                                                          175   £
                                                                                                       150   "
                                                                                                       125
                                                                                                                 700
                                                                                                                 600
                                                                                                                500
                                                                                                                400
                                                                                                                300
                                                                                                                200
                                                                                                                100
                                                                                                                      U
                                                                                                       100
50    55    60
             FIGURE 16   RIVERSIDE SPRAY DRYER RESPONSE
                           DURING SHUTDOWN

-------
O
I
Ul
ON
                  120
                  100
O   80


_o
"•   60

_3

•o   40
                  20
_o
u.
t/i
fS
                           300,000
                           250,000
                           200,000
                           150,000
                           100,000
    50,000
                                                                        Gas Flow
                                              Spray Dryer Outlet Temperature
                                            I	I
                                     I      I     I      I
                                                                    I	I
                    10   15    20
                                                            25    30    35

                                                              Time (min.)
                                                                            250
                                                                            225 £

                                                                                 OJ

                                                                                5
                                                                                 m
                                                                            200  5
                                                                                 Q-
                                                                                 E
                                                                                      600
                                                                            175
                                                                                                    150
                                                                                                        O
                                                                                 Q
                                                                                 >%
                                                                                 m
400   >p
     c
     _E

300   u

     IS
     rs
     O
200   ">




100
                                                                                                    100
                                                                                                            -JO
                                                              40    45    50    55    60
               FIGURE 17  RIVERSIDE SPRAY DRYER RESPONSE TO RAPID
                            LOAD FLUCTUATIONS

-------
 a
 E
 3
 as

 a!
150



140



130



120



110



100
0
                                                     12
                                                            15
                                                                            18
                                       Time, Minutes
                            A.  Wall Blowers

                                (Steam Flow = 6,000 to 8,000 Ib/hr)
            «J

            3
                01
                a
                E
            £


            05

            a;
                    150
                    140
                130
                    120
                    110
                    100
              a
              E
              3
              as
             150



             140



             130



             120



             110



             100
                    0
                                     Time, Minutes


                          B.  Economizer Blowers

                              (Steam Flow = 7,000 to 9,000 Ib/hr)
                                                                  12
                                       Time, Minutes


                          C.  Retractable Blowers

                              (Steam Flow = 10,000 to 12,000 Ib/hr)



FIGURE 18   EFFECT OF STEAM SOOTBLOWINC ON  SPRAY DRYER

             INLET WET BULB TEMPERATURE
                                      10-157

-------
the steam soot blowers. The economizers and furnace wall blowers raised the wet bulb temperatures
approximately 5 F The retractable furnace soot blowers increased the wet bulb temperature by as
much as 8 F  At low loads, soot blowing increased the wet bulb temperature by as much as 12 F at
the spray dryer inlet.

      Another phenomenon which occurs at Riverside is the stratification of dry bulb and wet bulb
temperatures  in the inlet duct to the spray dryer. This is caused by differing operation of the Unit 6
and 7 boilers. Figure  19 presents examples of stratified wet bulb temperatures at the spray  dryer
inlet during soot blowing with one unit off line. As shown on Figure 19,  a 3 F stratification across
the inlet duct and  an  11 F increase over the initial wet bulb temperature occurs during peak soot
blowing. With both units  on line -  one blowing the furnace walls and the other blowing the air
heater - a wet bulb  stratification of 5  F can occur across the duct. It was initially believed that
the stratification would not be  significant  when the FGD system was treating flue gas from one
boiler. Similar tests were performed at NSP's Sherburne County Unit 1 (750 MW). Measurements in
the two air heater outlet  ducts on this unit demonstrated  that wet  bulb stratification occurred
across each duct and from one duct to the  other. The stratification measured at Sherburne County
Unit 1 corresponds with the number and location of the soot blowers in operation. The magnitude
of the stratification at Sherburne Unit 1 was similar to that experienced  at Riverside. It is important
for successful low  temperature operation of spray dryer systems that the fluctuations and stratifica-
tions in wet bulb temperature be quantified and monitored whenever possible.

                                      CONCLUSIONS

      The Riverside spray dryer system has demonstrated the capability to provide compliance with
New Source Performance Standards for SOo emissions. The average removal efficiency was 84.3 per
cent for the   164  tests, including low lime and no lime additive tests. SOo  removal efficiencies
exceeded 95  per cent  in 30 of the 164 tests. SO7 removal ranging from 70 to 90 per cent removal
was demonstrated for  a variety of fuels.

      The spray dryer system is capable of drying the waste product to a powderlike consistency.
The moisture content parameters in the data base indicate that the spray dryer system produced a
relatively dry waste product in all of the tests. While operating the spray dryer during the two-year
test period, upset conditions did occur occasionally  which raised the moisture  content of the  waste
solids to relatively high levels.  The solids moisture content recorded in the  data base represents
typical steady-state utility operating conditions.

      From an analysis of the performance test data collected during 1981 and 1982 at Riverside,
certain overall correlations were apparent.  A strong relationship between SO2 removal, approach
temperature,  ash alkalinity, and lime stoichiometric  ratio was observed. Other  correlations between
fabric filter SOo removal and spraydown temperature, and between spray absorber and fabric filter
waste moisture content were apparent from the performance test data. Many other correlations
which have been demonstrated on pilot size units  were not observed from  an  evaluation of the
overall data base.

      The Riverside dry scrubber system responds well to changes in operating conditions. It was
possible to maintain absorber approach temperature  at an almost constant condition during changes
in boiler operations.   SO2 removal efficiency typically changed with changing operating conditions;
however, it was usually greater  than  required, such that it  was possible to maintain compliance
                                          10-158

-------

























Q.
E
K
"5
OB
5

V
1
a
E
U
»r
5 140
75
S. 130
E
u
>- 120
JD
3
= 110
S 100

^
/ ^ V
, 	 /v /\ / v A^rNiV ^ 	 _


-
i i i i

Boiler 7
Side of
Inlet Duct
to
Spray Dryer


0 3 6 9 12
Time, Minutes
A. Boiler 6 Off Line; Boiler 7 at Low Load
I3U
140
130

120

110
tnn
>W^

^^^X^^^^X^ 	 /s 	 ^_^^X^-^/ ^^

-

-
1 I 1 1 1 1
Boiler 6
Side of
Inlet Duct
to
Spray Dryer


03 6 9 12 15 18
15U
140
130
120


110

inn

_
c~
_


-

1 1 1 1 1 1

Boiler 7
Side of
Inlet Duct
to
Spray Dryer




B.
   0          3          6          9         12         15

                                Time, Minutes

Boiler 6 Wall and Retractable Sootblowing; Boiler 7 Air Heater Sootblowing
                                                                    18
FIGURE 19   STRATIFICATION OF FLUE CAS WET BLUB
            TEMPERATURE AT SPRAY DRYER INLET
                                    10-159

-------
almost continually. Variations in the absorber inlet wet bulb temperatures were observed during
soot blowing operations. The  maximum recorded wet  bulb temperature variation during soot
blowing was 8 F at full load and 12 F at low load. Sufficient margin must be maintained in the
approach temperature to account for such operational variations.

      The Riverside FGD system exhibits normal variations in performance consistent with a full
scale operating system and, based on a broad range of data, is not as sensitive to variations in some
parameters as are laboratory or pilot systems.

                                 ACKNOWLEDGEMENTS

      The authors  gratefully acknowledge the  contributions of Niro Atomizer Incorporated and
Joy Manufacturing Company. The performance data presented in this paper were collected as a part
of their testing efforts and were released to Northern States Power Company and Black & Veatch
for independent evaluation. In addition, Northern States Power Company's Riverside operating staff
is  to be commended for their  ability to endure the extensive testing program  conducted at their
installation.
                                          10-160

-------
DESIGN AND INITIAL OPERATION OF THE SPRAY DRYER
FGD SYSTEM AT THE MARQUETTE, MICHIGAN, BOARD OF
      LIGHT AND POWER - SHIRAS #3 PLANT

 0. Fortune, T. F. Bechtel, E. Puska, J. Arello

-------
                                 ABSTRACT

    This paper discusses the design issues, design decisions, start-up, and
early operation of the Spray Dry Flue Gas Desulfurization  (SDFGD) system
which went into operation at the Marquette Michigan Board  of Light and
Power Shiras #3 in May 1983.  This forty-four  (44) megawatt unit consisting
of a rotary atomizer reactor, reverse air fabric filter, lime preparation,
and reagent recycle system was engineered in the 1980-82 time period uti-
lizing pilot plant and prototype industrial system results as a design
basis.
The initial operation of
scaleup from pilot plant
the unit is discussed, as is the success of the
to Commercial  size boiler.

     I.  BACKGROUND
    The development of the process and equipment design technology on which
the Shiras #3 Spray Dry Flue Gas Desulfurization (SDFGD) System is based
began in 1978 under a joint venture agreement with Anhydro A/S of Denmark.
It merged the FGD process/equipment and particulate collection system
knowledge of General Electric Environmental Services,   Inc. with the spray
atomization/drying knowledge of Anhydro into a single
which was bench tested in Copenhagen, pilot tested at
Springs Martin Drake Station (CCS), and characterized
testing  (2, 4, 5) partially funded by the USEPA.
                             integrated system
                             the City of Col orado
                             in two years of
    In the late 70's, the Board of Light and Power of the City of Marquette
Michigan commissioned Lutz, Daily and Brain to design the totally new 44 MW
Shiras #3 unit.  The plant could have complied with Ib/MBTU emission
requirements for S02 via the compliance coal approach but was forced by the
% removal clause of the 1977 Clean Air Act to include FGD in the plant
design.  For this size installation the emerging SDFGD technology provided
the best balanced solution.
                             II.  APPLICATION

    Shiras Unit #3 is a 44 MW installation.  Plant equipment includes a
415,000 pounds (188,244 kg) of steam per hour Combustion Engineering  (CE)
pulverized coal fired boiler fed by four CE coal mills using #2 fuel  oil in
the igniters.  The boiler utilizes steam soot blowing and is operated as a
balanced draft system with single FD and ID fans and a CE regenerative air
heater.  The ID fan system, consisting of a Westinghouse Sturdivant radial
tip centrifugal fan, American Standard fluid coupling, and 1250 HP  (932  KW)
/900 RPM (94 rad/sec) General  Electric motor, was provided as part  of the
air pollution control system contract.

    Design conditions utilized for the Air Pollution Control System defini-
tion are summarized in Table 1.
                                  10-161

-------
                      TABLE 1 DESIGN  CONDITIONS
Air Heater Inlet Temperature

Air Heater Outlet:

Temperature


Air Flow
Pressure

Flyash
 852°F  (728°K)
 265°F  (402°K) continuous
 550°F  (506°K) maximum
 205°F  (370°K) @ 25% MCR
226,000 ACFM  (107 ACFM/sec)
723,000 #/Hr  (91 kg/sec)
1500 PPMV S02
12% Vol. H20

-4" W.6. (-1 kPa G)

9075 #/Hr (1.14 kg/sec)
    The basic SDFGD process is defined  in the process flow  diagram (Figure
1).  In most cases, the  flue gas goes directly form the air heater outlet
to the spray atomizer/reactor.  However there are retorfit  applications
where some  flyash removal occurs as  a result of existing  cyclones or preci-
pitators.
                    GRITS
                                                              SLAKING WATER
                                                         LIME
                                                        STORAGE  UNLOADING
                    LJ     VWYYY
                             TEMPERATURE
                               MONITOR

"1 '
URE
R




*
ASH
SILOS



                                                       I. D. FAN(S)
                    STACK
       PUMP
RECYCLE
FLOODED
LOOP
TANK
                                              Y
                   LANDFILL
FIGURE 1.  DRY FLUE GAS DESULFURIZATION SYSTEM
                               10-162

-------
    Aerodynamic devices in the chamber  (inlet  scroll,  pre-swirl  vanes,
inlet annulus) impart appropriate axial  velocities  and  angular momentum to
the gas stream to insure uniform mixing  with the  reagent  stream  and cyclo-
nic action that retains 40-55% of the dried  product  and  flyash in the
reactor vessel hopper.   This  total  vessel  cyclonic  effect  is  unique to
single atomizer/vessel  systems,  although other reactors  (spray nozzle,
multiple rotary atomizer/chamber) exhibit  some collection  of  particulate in
the vessel due to its "drop-out  box"  characteristics.   The reagent slurry
streams ("milk of lime" and recycle material)  are fed  to  the  centrifugal
atomizer wheel with rates controlled  respectively by  system outlet
S02 level  and reactor vessel  outlet temperature.  The  separate streams par-
tially mix on the wheel and chemically  react,  but not  to  physical or chemi-
cal equilibrium because of the short  (millisecond)  residence  times on the
wheel.  Equilibrium of  these  mixtures takes  on the  order  of 5 to 20 minutes
in controlled bench scale experiments.   The  design  intent  is  to  minimize
the reaction between the two  reagent  streams to the  greatest-  degree
possible.   The reagent  stream passes  from  the  central  feed point on the
wheel  (Figure 2) to the tip as a thin Coriolis-force-induced  liquid film in
radial holes distributed about the circumference  of  the  wheel.   When the
film reaches the tip, it is sheared off  by  the gas  to  form a  droplet
stream.  The initial velocity of these droplets has  a  radi al /tangential
velocity ratio that depends on mass flow rate  per hole  and liquid viscosity
but is always less than one.   The droplet  size distribution is affected
weakly by film thickness and  strongly by gas shear  force  (7,  8); hence ato-
mizer wheel  tip speed has the strongest  influence on  droplet  size.
                              FEED  2
      FIGURE 2. DUAL FEED ATOMIZER
                                  10-163

-------
      The droplet streams  form  what looks like an umbrella  to the  stationary
  observer.  The interaction  of  the inlet gas with this  umbrella of  droplets
  is  a critical factor in  controlling the uniform mixing of reagent  and
  S02 laden gas.  The drying  of  these droplets of highly alkaline  (pH  8-12)
  liquid with suspended particles  of  (.5 - 4 micrometer) calcium hydroxide
  and (.5 - 30 micrometer)  flyash  into  (10 - 50 micrometer) porous particle
  clusters proceeds through  a complex series of surface  chemistry  and  dif-
  fusion mechanisms (6, 9)  to produce a relatively dry  (1 - 3% moisture) par-
  ticulate stream at the reactor outlet.  The size of the resultant  porous
  clusters depends on both  the  initial droplet size and  the % solids in the
  feed slurry.  For very high inlet temperature situations  without recycle
  where the % solids in the  slurry varies,  this porous cluster size  could be
  a  significant variable.

      The spray absorber outlet  stream mixes with reheat gas  of a quantity
  dictated by the desired  particulate collector inlet temperature and  pro-
  ceeds to the particulate  collector.  Some further reaction  occurs  between
  the remaining gas stream  SO?  and the particulate material.   In the case of
  a  fabric filter, this additional reaction can capture  20-50% of the
  remaining SO?; in a precipitator, the additional  capture  is  in the 5-15%
  range.
                            SPRAY ATOMIZER/
                            REACTOR
                                         LIME PREP SYSTEM
                                                    FABRIC
                                                    FILTER
                                                       REVERSE
                                                       AIR DUCT
                                                            1DAND
                                                            RAFAN
                                                           NCLOSUR
FIGURE3.SHIRAS#3
                                  10-164

-------
    The particle clusters normally have a central core of unreacted calcium
hydorxide and a substantial amount of calcium carbonate produced by
absorption/reaction of gas stream C02.  While S02 is more reactive than
C02, the high C02/S02 ratio in the gas stream counter-balances the reac-
tivity difference and results in substantial calcium carbonate formation
(4).

    In addition, certain flyashes contain substantial quantities of alka-
line material.  Reslurrying the collected solids will get the residual
alkalinity of the calcium salt/flyash clusters back into solutions while
also providing relatively large nucleation sites for the fresh calcium
hydorxide particles.

    The physical equipment which turns the process flow diagram into
reality is shown in the site photograph of the Shiras #3 unit (figure 3).


    III.  SDFGD SYSTEM DESIGN ISSUES AND INITIAL OPERATING EXPERIENCE

                             AT SHIRAS NO. 3

    There are many sub-system design issues which have to be resolved in
the design of the SDFGD system.  In this section, we will  briefly define
the issues and discuss how the designs have functioned during initial
operation.

A.  REAGENT PREPARATION

    References 1, 4, 11 and 16 deal  in detail  with the effect of various
reagent choices on the performance of the systems.  Materials such as lime,
limestone, sodium carbonate, Nahcolite,  Trona,  ammonia compounds, etc. were
considered.  For a variety of reasons, e. g.,  cost,  performance,  waste sta-
bility, and availability, soft-burned pebble lime has become the favored
material for SDFGD systems.  In the case of Shiras #3 the lime supply is
high calcium with  specified available calcium oxide content of 88%.

    Lime is a generic term for a range of materials which  vary in their CaO
content, may contain significant levels  of MgO,  have varying amounts of
"grits"*,  and may have "hard burned" constituents (caused  by surface
melting in the kiln).  This variation in composition causes a significant
diversity of opinion in how the lime should be slaked to "milk to lime", a
water slurry of Ca(OH)2.

    The slaking process is described chemically as follows:

    CaO(s)+H20(l)   Ca (OH)2 (1)+15,300  K - Cal
                                         Kg - Mole


*"Grits" are composed of  uncalcined limestone,  kiln brick  fragments,
silica, alumina,  ferric oxide,  etc.
                                  10-165

-------
    When this reaction is carried out with water of almost potable qualtiy
and high calcium,  soft-burned,  pebble lime at a water/lime ratio of 3 to 4,
its exothermic nature causes an almost explosive disintegration of the
pebbles and a 70°-80°F (40-45°K) temperature rise in the water in four
minutes resulting  in a slurry of extremely fine (.5-4 micrometer) particles
suspended in water.  This slurry quality level  is highly desirable for the
SDFGD process.

    The well known paste or detention slakers,  which are simply mixing
tanks with agitation, depend on this exothermic reaction to achieve the
desired product.  The larger "grits" are allowed to settle and are physi-
cally removed from the system via a combination of screw classifier,
vibrating screen,  and dilution  water wash lowering the specific gravity of
the lime slurry from > 1.3 to > 1.2.  These commonly used slakers are
simple, reliable,  and relatively inexpensive.

    When poor quality slaking water, hard burned contents, or MgO contents
(almost always hard burned) are a significant factor, different slaker
types such as attrition or ball mill are usually considered.  Poor quality
slaking water has  dissolved solids which react  with the surface of the lime
pebbles, "blinding" their pores and preventing  water infiltration, thus
stopping the slaking process.  Hard burning causes the same problem via a
surface glaze effect, i.e., the pores left by the escaping C02 in the
calcining process  collapse due  to overheat resulting in "blinding" pebbles.
These more aggressive slaking methods use mechanical means to break up the
surface and allow water contact with the internals of the pebbles to
complete the slaking.  Even with these techniques the quality of the
reagent is not as  good as that  produced under the more ideal conditions.
Even though these slaking systems can grind up  the grits, they still should
be  rejected from the system.  Since they are abrasive and require water for
the transport of inert material, their retention results in an ineffective
use of a key resource in this water limited process.

    Shiras #3 uses a pre-assembled Clow-Coffman lime preparation plant.  It
consists of a pebble lime storage silo with an  integral vent filter and
dual vibrating bin discharges,  redundant lime feeders, redundant lime sla-
kers with agitators and grit removal screws, dual  vibrating screen grit
removers with associated screw  conveyors, a lime slurry tank with redundant
agitators,  redundant slurry pumps, and associated flushing, drain and
control systems.  A major factor in the selection of this design was the
fact that a similar system  (without the operational redundancy) designed
and built by the same vendor performed very well in the CCS pilot program
(2).  The lime slurry design range is 15-25% solids controlled by agitator
current and mass/thermal balances.

    In actual around-the-clock  automatic operation we found that several
small practical improvements could be made to the slaking system, and that
one major issue had to be dealt with.  Minor adjustments were made, such as
the installation of a lime feed cutoff device that is activated if slaker
water pressure drops below a given level.   In a small power plant
                                   10-166

-------
water supply line pressures have occasional wide fluctuations.  It
is better to stop slaking for a few minutes when this occurs, rather
than depart from the optimal lime slaking temperature ranges  (180°F
to 190°F).

    A more serious problem is that we found that one lime truckload out of
a dozen had grit contents over 20%, most of which was unburnt limestone.
This problem was later traced by the lime supplier to learning curve
quality control problems caused by the change from a gas-fired to a coal -
fired kiln.  A large percentage of the limestone grit would leave the
slaking system in the form of a fine suspension, and cause pluggage
problems in the lime distribution system.  It was found that a simple
hydroclone was adequate to filter out the calcium carbonate, provided that
one took into account the highly viscous, non-Newtonian nature of the lime
slurry and over-sized the hydroclone accordingly.

B.  SPRAY ABSORBER

    The chamber, where the reagent and recycle slurries are atomized into
fine droplets, where the reagent droplets mix with the incoming flue gas,
where the absorption/chemical reaction of S0£ takes place, and where a
significant amount of the dried particulate is collected, is the most
critical piece of process equipment.  A cutaway schematic of a typical
spray absorber is shown in Figure 4.  Operational experience with hundreds
of industrial  spray drying applications provide the user with a range of
cost-effective reliable alternatives.  The major concerns that utility
S02 service causes for this proven equipment is the abrasive and corrosive
constituents in the inlet flue gas; abrasive chemically reactive slurries
are not new to the spray dryer industry (13).

    Reference 14 gives a credible review of the choice between two fluid
spray nozzle systems and spinning disk rotary atomizer systems.  The main
factors in that choice are cost, power consumption, droplet size and
variability control, reliability, and gas/droplet mixing.  Those vendors
who supply high horsepower rotary atomizers use one centrally located in
each reactor of a multiple reactor system, treating up to 150 megawatts
equivalent gas flow per reactor.  This symmetry allows use of the vessel as
a moderately efficient cyclonic collector.  This can be particularly useful
in protecting  the downstream particulate collector from wet material in
case of upset  conditions.  Our emphasis in this paper is on the single
rotary atomizer system supplied by Anhydro A/S to GE for the Shiras #3
installation.   The spray absorber/reactor consists of a low pressure drop
gas disperser, belt-driven atomizer, and absorber chamber.  The spray
absorber must  be carefully optimized so that the reagent slurry is ato-
mized, contacted with the flue gas, and dried in a manner that promotes
maximum capture of S02, minimum reagnt consumption, and low energy use,
while maintaining stable and reliable plant operation.
                                   10-167

-------
                                                     Atomizer
                                                     Removal
                                                     Monorail
                                                     Penthouse

                                                     Gas
                                                     Distributor
              Axial
              Entry Vanes
               Atomizer
               High Gas
               Exit Duct
              FIGURE 4. SPRAY ABSORBER
    For the dry SDFGD application, we use a standard 400 mm tip  diameter
centrifugal atomizer designed to generate a uniform spray  of  fine  (10-80
micrometer) droplets over a wide range of feed rates.  For larger  gas
flows, the number of radial holes in the atomizer wheel is increased.   The
droplet size model used in our system design has been  confirmed  in detailed
experiments at GE's Corporate Research and Development Center.   Typical
results over a range of tip speeds are shown in figure 5.

    For the Shiras #3 atomizer, this means that the mass mean particle  size
i s about 30 mi crons.

    The two slurry feeds are piped independently (see  figure  2)  to a  co-
annular liquid distributor.  In this way the two chemically reactive
calcium hydroxide and recycle slurries are kept separate until  fed on  to
the atomi zer wheel.
                                  10-168

-------
                            ATOMIZER OUTPUT FOR
                          CIRCULAR SLOT AT 2.5 GPM
                                       DECREASING co,
                                       Acu = 2500 RPM
                                       eu = 30000 RPM
                           20   40   60   80   100  120  140  160
                              DROPLET DIAMETER (MICRONS)

                   FIGURE 5. FOUR INCH WHEEL DATA
    The slurries are introduced to a central cavity on the rapidly rotating
atomizer wheel, and are induced by inertial forces to flow outward through
radial passages in the wheel and then break off to form spherical  droplets
whose size range is chiefly governed by the viscosity and surface tension
of the liquid and the atomizer wheel tip speed.  As the droplets move away
from the wheel and disperse into the gas stream exiting the gas dsiperser
vanes, they form an umbrella-shaped spray pattern that is symmetric about
the chamber axis and serves as the zone of initial contact between the
reagent and flue gas.  Figure 6 shows a stroboscopic photograph of one
ligament of that umbrella.

    The size of the droplets must be controlled to assure optimum reaction;
small  droplets are desirable because they provide a large surface area for
mass transfer, but they must not be so small that they do not penetrate to
the outer diameter of the inlet gas annulus.

    A belt drive system,  which makes it possible to change atomizer wheel
speed by simply changing  the belts and sheaves, is used on Shiras #3.

    Because of the abrasive character of the slurries atomized in the spray
absorption process,  a wheel design is used which features silicon carbide
inserts in the slurry passages.  The inserts may be rotated as local wear
spots  appear to extend their useful  life.
                                 10-169

-------
FIGURE 6. DROPLET STREAM STOP-ACTION


     The Shiras vessel is 36 feet (11 meters)  diameter,  71  feet  (21.6
 meters) tall and uses a 200 HP (150 KW)  motor to drive  the atomizer  wheel
 at  7800 RPM  (817 rad/sec) thru a custom  designed belt/sheave system.

     The absorber gas dispenser is designed  for high  S02 removal  without  high
 pressure drop or flyash abrasion problems.   Anhydro  selected a  top-entry
 vaned scroll-type gas dispenser, which dischanges an annulan vontex  flow of
 flue gas down into the chamben on all sides of the atomizer wheel.   The  gas
 dispensen is equipped with vanes whose angle can be  adjusted to obtain
 mixing of the spnay and the flue gas.  The  initial gas  notation fnom the
 inlet scnoll and vanes is in the same dinection as wheel  notation which
 pnovides an  incneased cyclonic effect but slowen deceleration of the
 dnoplet stneam and langen dnoplets.

     The absonben chamben must be sized in nelation to the gas flow volume
 to  assune that the slunny dnoplets will  have adequate nesidence time (in
 most applictnions 8 to 12 seconds based  on  absonben  outlet volume) in the
 chamben fon  the vanious stages of neaction  with SO?  and dnying  to occur.
 The chamber  design also affects the degnee  of dnopout in the absonben,
 which should be maximized to neduce the panticulate  loading to the fabnic
 filten.  The absonben is designed as an  axial-entny  cyclone to achieve
 appnoximately 50% dust dnopout in the chamben.

     The GE  CR&D Centen has done extensive coupled modeling of the gas flow,
 particle trajectory, drying, absorption  and chemical reaction to provide a
 tool for connelating openating data and guiding futune design optimization.
                                   10-170

-------
    The model  results indicate that the key phenomena occurring in the gas
inlet/droplet injection region of the vessel is mixing of the gas and
reagent; the bulk of the absorption process occurs after this initial
gas/slurry mixing phase.

    If the droplet/gas momentum ratio is too low, the gas at the outer
diameter of the inlet annulus will see no reagent, i.e., bypass occurs.   If
the ratio is too high, a similar bypass core occurs at the inner diameter
and the larger droplets get picked up by a recirculation vortex and get
deposited on the wall of the atomizer.  Since these ratios are constantly
changing as boiler load changes, the Shiras #3 design divides the inlet
into three dampered co-annular rings so that high gas momentum can be main-
tained as gas flow volume changes.  The variable speed drive concept men-
tioned earlier can also be used to advantage in this turndown management
system, i.e., smaller droplets at low gas velocities, bigger ones at high
velocities.

    A practical illustration of the importance of the droplet/gas momentum
ratio was provided early in the Shiras #3 startup.  Persistent boiler soot
blowing system problems resulted in weeks of steady-state gas temperatures
into the spray absorber of between 350°F and 410°F; far above the design
value of 265°F.  The GE/Anhydro atomizer had the reserve power and pumping
capacity to increase slurry flow and maintain constant approach tem-
perature.  However, this resulted in a significantly higher droplet/gas
momentum ratio, and a much broader reaction umbrella, which caused minor
solids deposits on the vessel walls and roof to be built up over a one
month period.  As soon as this was seen, the setting on the gas distribu-
tion vanes  (the "wing vanes") was altered to permit a more vertical gas
flow into the vessel, thus reducing droplet/gas momentum ratio.  After
another month of operation, the vessel was inspected, and the roof and
walls were free of deposits.

C.  PARTICIPATE COLLECTOR

    In a SDFGD system the spray atomizer/reactor is primarily an S02
capture device but doubles as a particulate collector.  Not to be outdone,
the fabric filter  (or precipitator in some cases) particulate collector
also doubles as an S02 absorber.  Effective operation in this secondary
mode requires operation of the fabric filter, within 20-50°F (11-28°K) of
the gas dewpoint.  Although many utility baghouses operate that close to
the sulfuric acid dewpoint, the quantities of liquid potentially produced
are so different that special design attention is rquired in the SDFGD
case, particularly in the areas of reheat control, internal/external insu-
lation, and leakage control.

    References 10 and 15 thoroughly discusses the pros and cons of precipi-
tators and fabric filters and the special design considerations for  each.
                                  10-171

-------
    In summary,  the superior FGD absorption contribution of the fabric
filter makes it  the logical  choice in high efficiency applications or in
any situation where lime costs are high and operating cost savings are
seriously considered.

    The Shiras #2 unit is a  conventional  eight (8) compartment custom GE
reverse air design.  It uses 168,  35 foot (10.7 m) long, 12 inch  (30.5 cm)
diameter, acid-resistant-coated, fiberglass bags  per compartment.  All
fabric filter dampers  are of the low leakage poppet type except the bypass
which is a pressurized double louver selected by  the LD&B engineers because
of their excellent experience at the City of Colorado Springs Martin Drake
Plant.  The fabric filter system is a conventional design with redundant
reverse air fans,  programmable controller control  of the cleaning cycle in
a variety of operator  selected modes (manual, time,  P batch,  P distri-
buted, etc.), and a centrally manifolded  inlet,  outlet,  R/A duct system.

    With the atomizer  in operation, the fabric filter design conditions are
based on gas at  an inlet temperature of 175°F (325°K) and a gross air to
cloth (A/C) ratio of 1.49.   With the atomizer off  line the inlet tem-
perature is 265°F (402°K) and the gross A/C is 1.62.  Note that the design
dust loading changes substantially between these  two situations going from
6.0 to 4.7 grains/ACF.  In  actual  operation this  inlet gas tempreature to
the system has been as high  as 360°F (455°K).

    Baghouse operation at Marquette has been exceptionally trouble free.

    Operation has been at approach temperatures  between  25 and 60°F.  There
have not been any occasions  of condensation on the baghouse walls or the
bag filter cakes (all  baghouse compartments are  kept on  line regardless of
load level).  There has not  been any build up in  average baghouse pressure
drop in over four months of  operation.

    Most critical  and  interesting  of all, the frequency  of the baghouse
batch cleaning cycle actually drops when  the spray absorber is brought on-
line.  Typically,  if the baghouse has been cleaning 50%  of the time while
the atomizer is  being  inspected, it will  drop to  cleaning only 33% of the
time once the atomizer is restarted.

    Temperature  drop across  the baghouse  is minimal  (under 5°F) when the
spray absorber is operating.


D.  REHEAT

    The evidence is conclusive that operation with reactor discharge tem-
peratures as close to  adiabatic saturation as possible optimizes the lime
use - collection efficiency  relation.  However,  temperature variation
within the reactor due to non-uniform mixing of  reagent  and flue gas, wall
deposits,  reactor hopper pluggage  and ductwork condensation,  etc., constrain
this optimization.  Recognizing that reactor and  particulate collector
                                  10-172

-------
constraints on approach temperature are different, most designers have made
provisions for reheating the gas downstream of the reactor.  Several
methods have been considered for accomplishing this reheat, for example:

a.  Bypass of air heater inlet gas.

b.  Bypass of air heater outlet gas.

c.  Reci rculation of fabric filter outlet gas through a water/steam - gas
    heat exchanger.

d.  Injection of ambient air through a water/steam-air heater exchanger.

    The first two cause similar problems.  Because they contain particulate
and S02, clean-up of the reheat system is required.  If the reheat injec-
tion is upstream of the fabric filter, this clean-up is automatic.
However, since the fabric filter is a low efficiency S02 device (the preci-
pitator is even lower) which degrades in efficiency as its inlet tem-
perature is increased, these types of bypass have a "double kicker" effect
on required reactor efficiency.

    Of the two, airheater inlet gas is theoretically more effective since
smaller quantities are required for the same amount of reheat.   However,
the system heat rate effect is worse.  The pilot work at Colorado Springs
(12) raised some serious doubts about the validity of this theoretical
conclusion which warrants further evaluation.  The issue may be related to
heterogenous mixing but even that is not clear.

    Methods c and d use a scrubbed clean gas for reheat.  They  not only
eliminate the "double kicker"  (only the temperature degradation of baghouse
S02 efficiency remains) but also provide the option for reheat  injection at
the baghouse outlet (eliminating the temperature degradation effect).  This
option assumes that reheat was not for baghouse protection purposes but for
ductwork,  stack, and plume buoyancy purposes.  The negative factors in
these approaches involve recirculation fan horsepower and water/steam BTU
requi rements.

    In spite of the concerns raised about the theoretical  conclusions by
pilot results,  Shiras #3 uses the air heater inlet bypass approach which
requires a reactor efficiency of 80% to provide a system efficiency of 83%
with 20°F (11°K) reheat.  Air heater outlet reheat would require the reac-
tor efficiency to be higher than the system efficiency as indicated in
figure 7.   For example a system efficiency of 80% would require a reactor
efficiency of 100% if 22°F (12°K) reheat was required.  This dramatically
shows the impact of the selected reheat approach, particularly  with the
system design point on the high  stoichi ometry/  efficiency portion of the
performance curve where lime savings more than counter balances heat rate
increases.

    Operation of the reheat system at Shiras #3 has been routine.
                                   10-173

-------
             >- 100
             o
             z
             LU
             o
             LL
               80 -
             cc
             LLJ
             CQ



             X  60
             O
             sr
             UJ
             >
             IT
             Q

             Q
             LLJ
             DC
40
             O
             LLJ
             CC
                20
                     EFFECT OF AIRHEATER BYPASS REHEAT
                            ON SYSTEM EFFICIENCY
             25°F REHEAT
           (AIRHEATER OUTLET)
                                0°FREHEAT
                              AIRHEATER
                                INLET
                        20      40      60      80

                           SYSTEM SO, REMOVAL EFFICIENCY
                                                     100
             FIGURE 7. REHEAT EFFECTS
    Baghouse approach temperatures have been varied  between 25°F and 60°F,
and reheat has been used for the upper end of the  temperature range.
E.  CONTROLS

    The important issues in DFGD system controls  are:

a.  Reagent feed control

b.  Temperature control

c.  Turn down control

d.  System response to transients


    The other control aspects such as  lime  slaker control,  recycle slurry
generator control, and baghouse control are not  unique to SDFGD and are not
covered in this paper.


    Each design deals with the specifics  of these control issues in its own
way.  Therefore, our focus will be generic  with  the specifics relating to
Shiras #3.


    It is clear that the reagent feed  control  strategy should be to maxi-
mize solids concentration to  the limits permitted by solids drop-out and
abrasion consideration and to maximize the  ratio of recycle/fresh  "milk of
                                    10-174

-------
lime".  In our system design we use the GE patented (3) Two Loop Control
system which automatically optimizes system efficiency/stoichiometry.  The
lime slurry is prepared and stored at a solids concentration substantially
higher than would normally be required to achieve the required absorber
S02 removal level.  For an optimum outlet temperature to be reached, the
lime slurry is diluted with recycle slurry on the atomizer wheel just prior
to atomizati on.

    The first of the two automatic loops regulates the flow of lime slurry
to the absorber, based on the fabric filter outlet SOg level,  by modulating
the position of the lime slurry control valve.  SOg levels are also moni-
tored at the absorber inlet and outlet using a Dupont extractive system.

    The second loop controls recycle slurry flow based on approach to
adiabatic saturation temperature at the reactor outlet.

    Changes in the adiabatic saturation temperature due to steam soot
blowing, coal moisture changes, ambient moisture changes, etc.,  are pro-
vided to the system via feed forward controls and/or operator input.
Automatic determination of adiabatic saturation temperature is conceptually
attractive but has been inhibited by the unavailability of proven reliable
sensors.  Although not part of the system design contract, we will evaluate
three candidate sensors for this service at Shiras #3.  The sensors are all
extractive and subject to particle contamination if used at spray dryer
inlet or outlet.  It is conceptually possible to use these devices at the
baghouse outlet but the increased time delay substantially reduces its
value.  Also, use at the outlet would be questionable because of the
variable moisture absorption characteristics of the bag cake.

    In operation at Shiras #3, it has been found the Two Loop  Control
system regulates the temperature leaving the spray absorber vessel to with
+ 3°F during normal operation, and to within +_ 7°F during rapid  boiler
excursi ons.

    For multiple (3 or 4) vessel systems, this range is enough to permit
system turndowns of 40-50% and excess flows to 20% via vessel  dampering and
individual vessel shutdown.  In the single vessel Shiras #3 design, the
turndown is accomplished via variable inlet geometry in the vessel.  The
geometry variation involves dampering of flow in three co-annular inlet
rings and variable swirl vane angle.  Variable atomizer speed will also be
tested,  although not part of the basic Shiras design.  The system allows
variation of the gas/slurry momentum relations in the inlet area which dra-
matically affects system performance.

    Turndown control is complicated for the single vessel Shiras design.
Pilot data indicates that operation of a reactor vessel in the +  20% gas
flow range is practical and has no more than a 5% - 1% effect on
stoi chi ometry for a given efficiency.

    System control dynamics issues are related primarily to the time delays
in sensing the effect of slurry feed on S02 levels at the fabric  filter
                                   10-175

-------
outlet and temperatures  at the reactor outlet and fabric filter inlet and
the slow read rate of the S02 monitor.  If not considered properly, system
interaction with these loops could lead to excessive hunting, overshoot,
and offset in the control  system.   To better understand the dynamics of the
system,  GE performed a computer simultion of the Marquette design.  (17)

    Operating experience has proven out that with 20 second effective
response times on the recycle slurry loop and 15 minute response times on
the lime slurry loop, that hunting does not occur.
F.  MATERIAL HANDLING

    The major issues in this area are the handling of potentially wet pro-
duct from the absorber hopper and the sourcing of recycle material.  Based
on pilot experience (1,5),  GE favors  mechanical  conveying (screw, drag
chain,  etc.) of reactor discharge material  in combination with multiple
assists for hopper clearage.  These include vibrating bit bottom, poke
holes,  and an auxiliary side removal  system from the hopper.  In order to
ensure sufficient recycle solids, reactor hopper and/or fabric filter
hopper discharge should be  available.

    Because the Shi ras #3 ash collections system was combined with upgrades
on units #1 and #2,  the City of Marquette purchased a pneumatic system
which combined pressure and vacuum portions.

    Operating experience has been that the pneumatic system needs a
grinding system to break up occasional  loose  clumps,  and such a device is
being added to the ash removal  system.

    Another practical  consideration is that although the solids hopper
catch handles like a dry; western flyash,  the boiler gas is brought to
within 20 to 40°F of its dewpoint.  Thus care must be taken to thoroughly
heat trace and insulate pneumatic vacuum systems,  since they will entrain
some boiler gas which must  not  be allowd to cool and condense.  Another
solution is to use heated (250 to 300°F) ambient air in the pneumatic
system.  The former was done at Shiras #3 for the baghouse hoppers, and the
latter for the spray absorber hopper.


G.  MATERIALS OF CONSTRUCTION

    Because of the non-corrosive nature of the gas stream down-stream of
the reactor inlet, assuming condensation is avoided, carbon steel is used
throughout the reactor, ductwork, and baghouse.   Since the centrifugal ato-
mizer outer shell is colder than the  inlet gas and subject to untreated
inlet gas, 300 series stainless is utilized in this protective cover.

    Slurry piping is unlined heavy wall  carbon steel; valves are rubber
lined.   Acid resistant coating  was used on the fiber glass bags because its
                                   10-176

-------
superior "fiber wetting" characteristic will protect the bag material from
wet alkali  attack; there is no acid attack concern.

    So far the only sign of erosion in the entire system has been the
expected wear at the atomizer wheel ceramic ports.


H.  INSULATION/LEAKAGE/  T

    Because of the low approach temperature/minimum reheat sytem design,
there is a real potential for localized gas condensation on duct surfaces,
reactor or fabric filter walls, etc.  To minimize the risk, extra thick
insulation (8 inch (20.3 cm) on reactor, 6 inch (15.2) plus air gap on
fabric filter) was employed, special leak tests were run on the reactor and
fabric filter, and purge dampers were not included on the fabric filter.
One feature that might have been desirable was internal  wall insulation in
the fabric filter to avoid compartment off-line cold wall effects.  There
are concerns about condensation under the insulation in this approach, but
it could have design merit if impervious insulation is used.  Marquette
doesn't have internal wall insulation.  The fact that all compartments will
be on line at all times  (except for on-line maintenance) reduces the cold
wall  condensation concern.  Units that have had this problem were running
continuously with off-line compartments in an attempt to minimize fabric
filter gas temperature drop.

    Operating exprience  has been that (with the exception of the vicinity
of one leaky door gasket) wall  condensation is not occuring, and that tem-
perature loss across the system is under 5°F.

I.  PERFORMANCE

    Design performance requirements for the Shiras #3 system are:

1.  80% S02 removal at all boiler loads from 20 to 100%.

2.  Outlet particulate of .005 grains per ACF or 99.5% weight efficiency
    which ever is less stringent.

3.  Lime use not to exceed 2874 #/Hr (.36 kg/sec) at design conditions.

4.  System pressure drop of 12.4" W.G. (3.1 kPa G)

5.  Gas temperature at stack inlet of 165°F.

6.  Bypass reheat of 3,600,000 BTU/Hr (1055 KW) based on 852°F  (729°K) gas.

    In September 1983 compliance test of all the above objectives were
met,  even though the system was not yet optimized, e.g., air preheater exit
gas temperatures were from 70 to 80°F above the design condition.
                                    10-177

-------
    A comparison of the compliance  test  data with 1980 data from  the  2  MW
demonstration facility at the  City  of  Colorado Springs, Martin Drake  No.  6
plant, shows that for the same  stoichiometries,  S02 removal efficiency  was
10 to 20% higher in the 44 MW  system than  in the 2 MW system.  Thus,  scale-
up has been properly applied in the Marquette system.

    The overall conclusion from the first  four months of spray absorber
performance is that it is possible  to  operate at S02 removal efficiencies
between 70 and 90% simply by varying lime  stiochi ometry.  This performance,
as well as the lack of scale-up problems,  is attributed to the Marquette
design which was supported by  the R &  D  effort at General  Electric's
Schenectady laboratories.

J.  ON GOING PROGRAMS AT MARQUETTE

    For DFGD vessels handling  from  100MW to 150MW of boiler gas each, 500
HP to 800 HP atomizer motors are required.   Radial  bearing loads  for  either
belted or spur-pinion gear systems  become  excessive for these high hor-
sepower drive trains,  and bearings  B-10  lives  become unacceptibly short
(under 10,000 hours).   As a result, General  Electric has developed a  Direct
Drive System which uses a voltage inverter  to operate a vertical   AC motor
at power frequencies other than 60  Hertz,  so that the motor rotates at  the
atomizer speed.  A flexible metal coupling  is  used  to transmit the torque
from the rotor to the atomizer.

    A Direct Drive system (shown in Figure  8)  has been successfully
operated for several months at the  General  Electric Turbine Technology
Laboratory in Schenectady, New  York, and is  currently being installed at
Shiras #3 for field testing.
                      INDUCTION MOTOR
                      200 TO 800 H.P. AT
                      8000 RPM
                           COOLING AIR


                           FEEDPIPES
                                             5 HP. FAN DRIVE


                                             MOTOR LIFTING LUG
                           INSULATION
                                             MOTOR SUPPORT
                                               BRACKET
                                             CONTOURED FLEXIBLE
                                             DIAPHRAGM COUPLING
                                             ATOMIZER BODY
                                             ATOMIZER WHEEL
                       FIGURE 8. DIRECT DRIVE ATOMIZER ASSEMBLY
                                   10-178

-------
                                  REFERENCES

 1.    Rhudy,  R.  G.  and  Blythe,  6.  M.;  "EPRI Spray Dryer/Baghouse  Pilot Plant
      Status  and Results";  EPRI  Symposium on  Particulate;  April,  1983.

 2.    Samuel,  E.  A.  et  al;  "Dry  FGD Pilot Plan  Results:   Lime  Spray  Absorption
      for  High  Sulfur Coal  and  Dry Injection  of Sodium Compounds  for Low  Sul-
      fur  Coal";  EPA/EPRI  S02 Sumposium;  May  1982.

 3.    Roth, A.;  U.S. Patent #4,  322,  224.

 4.    Parsons,  E.  L., Boscak, V.,  Brna, T.  C.,  and Ostop,  R. L.;  "S02  Removal by
      Dry  Injection  and Spray Absorption  Techniques";  USEPA Third  Symposium on
      Particulate Technology; March 1981.

 5.    Parsons,  E.  L., Hemenway,  L. F.,  Kragh, 0.  T.,  Brna, T.  G.,  and  Ostop, R.
      L.;  "S02  Removal  by  Dry FGD"; USEPA Sixth FGD Symposium,  October  1980.

 6.    Getler,  J.  L., Shelton, H.  L.,  and  Furlong, D.  A.;  "Modeling the  Spray
      Absorption Process for S0?  Removal";  Journal  of APCA, Vol.  29,  No.  12;
      December  1979.

 7.    Snaddon,  R.  W. L.; "Rotary  Atomization  Studies  for  Dry Flue  Gas
      Desulfurization"; GE  CR&D  internal  report;  March 1983.

 8.    Master,  K.;  "Spray Drying  Handbook";  John Wiley; 1979.

 9.    Downs,  W.  et al;  "Control  of S02  Emissions  by Dry Scrubbing";  American
      Power Conference; April,  1980.

10.    Fortune,  0.  F. and Miller,  R. L.;  "Design Considerations  for Baghouse -
      Dry  S02  Scrubber  Systems";  EPA Particulate  Symposium, November,  1982.

11.    Kelly,  M.  B.  and  Shareef,  S. A.;  "Second  Survey of  Dry S02  Control
      Systems";  US EPA-600/7-81-018;  November 1980.

12.    Samuel,  E.  A.  and Brna T.  C. "Final Report  on CCS";  June 1983.

13.    Tuttle,  J.  N.  et  al;  "Neutralization  During Atomization  in  Spray Dryer
      Yeilds  100% Dry Powder";  Chemical  Processing; November,  1979.

14.    Maurin,  P.  G.  et  al;  "Two  Fluid Nozzles vs. Rotary  Atomization for  Dry  -
      Scrubbing  Systems":  Chemical Engineering  Progress;  April  1983.

15.    Campbell,  K.  S. et al; "Economics of Fabric Filters  Versus  Precipitators";
      EPRI FP-775  June, 1978.

16.    Yeh, J.  T.,  Demski,  R. J.  et al;  Experimental Evaluation of Spray Dryer
      Flue Gas  Sulfurization for  use  with Eastern U.S. Coals"; EPA-EPRI FGD
      Symposium;  May 1982.

17.    Forsberg,  C.  H. and  P. S.  MacDonald;  "Flue  Gas  Desulfurization System
      Model";  GEESI-ADAPCO  Report 18-03-001;  May  1983.
                                     10-179

-------
     START-UP AND INITIAL OPERATING EXPERIENCE OF THE
           ANTELOPE VALLEY UNIT 1 DRY SCRUBBER

R. L. Eriksen, F. R. Stern, R. P. Gleiser, S. J. Shilinski

-------
         START-UP  AND  INITIAL OPERATING EXPERIENCE
                    Robert L. Er ikson
             Environmental Control Supervisor
             Basin  Electric Power Cooperative
                   FrederIck R. Stern
                     PI ant Eng i neer
             Basin  Electric Power Cooperative
                   Richard P. Gleiser
                  Field Service Engineer
                 Joy Manufacturing Company
                   Stan ley J . Sh I I i nsk i
                  Field Service  Engineer
                    N i ro Atom i zer Inc.


                        AESIBACI

The first  competitively bid  and awarded utility dry scrub-
bing system utilizing  lime as the scrubbing reagent was for
Basin  Electric's  Antelope  Valley  Station Unit  No.  1.
Awarded In  1978  to  Joy Manufacturing with Niro Atomizer  as
the major  subcontractor, the  system was scheduled to start-
up  in  1981,  however,  due  to  reduced  load  growth, was
delayed until  this year.

The dry scrubbing  system  treats flue  gases from  a  435  MW
lignite fired boiler and consists of reagent preparation
equipment,  five spray dryer  absorbers,  and  two  fabric
f I Iters.

Initial operation  on coal began  in May, 1983 and commercial
operation  of  the  system  Is  scheduled for July, 1984.  This
paper  will  review the start-up procedure,  any problems
which  have  developed  thus  far  and  how they  have been
handled,  as well  as the  results of  the  operation  of the
system  to  date.
                          10-181

-------
In September, 1978, Basin Electric  awarded  a  contract for
a  pollution control system  to  the Western  Precipitation
Divlson  of  Joy Manufacturing with  Niro  Atomizer  as a major
subcontractor.  The contract was  for  the  supply  and  erec-
tion of an SO  dry scrubbing system  and  baghouse  for Unit
#1 at Basin Electric Power Cooperative's Antelope Valley
Station near  Beulah,  North Dakota.   Major equipment in-
cluded  In  this contract  were  the reagent  unloading and
storage system, reagent preparation  equipment,  five spray
dryers,  mechanical conveyors  for a recycle  system,  two
baghouses for  particulate collection, and  a computer
control  system for the operation  of  the dry  scrubber.

Engineering  commenced in November,  1978, with  field mobi-
lization  and  the  first  equipment  arriving  on  site  In
July, 1979.   Field  erection activities  lasted  from then
until May,  1982.  The plant was  originally  scheduled for
commercial operation  in 1981;  however,  due to  lower than
anticipated  regional  power demands,  the  actual  plant
start-up did  not occur until May,  1983.

Joy  and Niro start-up  personnel  arrived  on  site  in
November,  1982, and began pre-start  check-out  of  the spray
dryers  and baghouses, Including  checking motor  rotation,
programming  of the computer  control  system,  etc.   The
actual  system  start-up  began  In  May 1983,  when  the
Combustion Engineering  435MW  lignite boiler  was  first
fired with coal, although the first  flue  gas to  the system
occurred on January 13,  1983, when the  boiler  was  fired on
oil.  Start-up and testing  lasted  until  the  first week  of
October,  1983, when the  performance testing  to verify
operating  parameters  as guaranteed  (lime  consumption,
pressure  drop,  power  consumption,  etc.) was  successfully
compIeted.
The Antelope  Valley Station dry  scrubber  consists of five
46-foot  diameter spray dryer absorbers.  An  inlet manifold
equally distributes  the boiler flue  gas to each of the
five spray  dryers  (Figure 1).   Gas  is  distributed  between
the roof and  central  gas dispersers  of each  spray dryer.
Each dryer  is  equipped with a  single,  F-800  model  rotary
atomizer,  which  is  powered by a flange mounted, vertical
shaft  700 hp motor.
                          10-182

-------
         ROOF  GAS
         DISPERSER
EXHAUST DUCT
             CENTRAL
               GAS
            DISPERSER
                                              INLET
                                              DUCT
             FIGURE 1  -  ABSORPTION CHAMBER
                         10-183

-------
The full design gas flow of 2,055,000 ACFM (at 307°F)  can
be handled  by  any  four of the five spray  dryers with  the
number  of  modules selected for  service based  on  boiler
load.   The  inlet and outlet ducts of  each  absorber have
guillotine  isolation dampers to isolate individual  modules
when they  are  not in  service.   The  scrubber  computer
control  system  signals the operator when spray dryers  need
to be  brought  on or off  line (due to changes In gas  flow),
and the operator then  initiates the appropriate sequence.

The slurry feed to the  atomizers  Is  prepared  In  a con-
tinuous feed  preparation  system with  lime, water,  and
recycle materials  controlled to match the system demands.
Two complete  lime slaking and  feed  preparation  trains
provide  redundant  equipment,  maximizing  system
availability.   Each slaking train Is sized for 100$  of  the
required  lime  consumption at  the  average  performance
conditions.   The  mixing trains  are each  sized  for  100$
capacity under  maximum operating conditions.

Figure  2  shows the  overall process scheme.  Pebble quick
lime is pneumatically  transferred from the  railcar/truck
unloading facility to  the 3,000 ton capacity silo and  from
there, conveyed pneumatically to  a 100 ton  capacity  lime
bin.   Weigh  belt  feeders meter the lime from the lime  bin
into the two  Denver Equipment ball  mills,  each sized  to
slake  6  tons/hr of pebble  lime.  The ratio of the  treated
water  and pebble  lime  fed Into the mills is controlled  to
maintain a 35$  solids  slurry within the mill.  Slaked  lime
slurry  leaving  the ball  mills Is diluted  to 20$ solids,
and classified In spiral classifiers to  100  mesh.   The
grits  separated by the classifier are  recycled back  into
the ball  mill.  The  classified  lime slurry Is pumped by
rubber  lined  pumps to  the mixing tank, where baghouse  and
spray  dryer recycle material along with dilution water  are
added.  The feed slurry  in  the mix  tank  is continuously
pumped  to  the  feed  tank, which overflows back to  the  mix
tank .

From  the  feed  tank,  the feed slurry Is  pumped through
individual  supply  lines  to head tanks  located above each
spray  dryer  module.   There Is a continuous overflow  from
the head tanks  back to the feed tank.  The purpose  of  the
head  tank  is  to supply slurry to each atomizer feed  con-
trol valve at  a constant pressure.   The  control   system
regulates the  amount of  feed delivered through the  control
valve  to the  rotary atomizer to maintain a constant  outlet
temperature at  each spray dryer exit.

Lear Siegler  SM810 S02 monitors,  located  at the scrubber
                           10-184

-------
OO
Ul
                                                                                     FIGURE  2  -  FLOW  DIAGRAM

-------
Inlet duct and  in the stack,  feed  signals  Into the scrub-
ber  controls  (Honeywell  TDC-2000  computer system).  Boiler
flue  gas  flow,  temperature,  and S0_ content  are  con-
tinuously  monitored and  the feed  preparation system inputs
are varied continuously to maintain  a  35$  total solids
content  while obtaining  the  desired S02  removal  at the
required stack temperature.

At the outlet of each spray dryer  is  an  inclined  mechani-
cal  drag  link conveyor  to transport  the dried powder and
ash  collected in each absorber.   Gas  exits  from each spray
dryer  via a side discharge  duct along  with entrained
flyash and spray dried material.   A portion  of  the  par-
ticulate  in  the  gas stream  drops to the bottom of the
spray dryer and  on to the drag  conveyor for reuse in the
feed  system.  The major  benefit  of  this two point collec-
tion concept  Is  that  if  wet material  should exist in the
absorber  due to an  upset condition,  It will fall to the
bottom of  the chamber and  be  col lected by the drag  con-
veyor,  while  the  flue  gas  will  exit  from  the  side
discharge  duct  in the  conical  section  of the  chamber.
Therefore, a  clear flue  gas path  will  always exist.

From the  five spray dryer conveyors,  the collected recycle
material   is deposited  onto  two  horizontal collecting
conveyors, and from the  collecting  conveyors,  the material
dumps Into a single  vertical  lifting  conveyor which dis-
charges into the  recycle  bin.    A  portion  of  the  ash
collected  In the  baghouses is pneumatically conveyed to
the  recycle  bin to  insure that an  adequate supply  of
recycle  powder  is available  to support  process
requirements.  The required  amount of recycle powder  is
metered by a variable  speed drag  link conveyor  into the
mix tank  along with lime  slurry  and  dilution water to make
up  the final feed slurry.  The  use  of  recycled powder in
the process  is to  reduce  the amount of fresh  lime needed
for  SO  scrubbing,  and  to enhance  the drying aspects of
the overall spray  absorption process.
      sie
The two  particulate  collectors  installed  at Antelope
Valley  for  collection  of flyash and  spray  dried product
are Western Precipitation  reverse  air baghouses.   Each
baghouse  consists  of  two parallel  rows of seven compart-
ments  for  a total  of  twenty-eight  compartments.   The
compartments  are  connected to the common inlet, outlet,
and reverse air manifolds  which are located  on  the cen-
terline of each baghouse (Figure 3).
                          10-186

-------
FIGURE 3  - BAGHOUSE  DRAWING
             10-187

-------
Each  compartment contains 288 filter bags.  The bags  have
a 12-!nch  diameter and are 35 feet long.  The bag type is
Menardi-Southern 601T  Teflon coated  fiberglass.   At  the
maximum  design  gas volume of 1,894,000 ACFM, the effective
filter  ratio  is 2.19.  The design operating temperature is
192°F.   The bags are suspended under spring tension from a
hanger  bar  at  the top  of the compartment and retained at
the  bottom by a flared  thimble.  The  thimble  floor is
"stepped"  down  towards the baghouse outer wal Is.   Beneath
each compartment are  two pyramidal hoppers, each equipped
with  an  access  door, poke holes, a hopper heating  system,
and a nuclear  dust  level detector.  Each baghouse compart-
ment  has one  blade type  Inlet damper,  two outlet  poppet
type  valves,  and one reverse air poppet valve.  All valves
and  dampers  are  pneumatically  actuated  by  solenoids
operated from  the control panel.

The operation  of the two baghouses is controlled Independ-
ently by one  Texas  Instruments STI Programmable Controller
utilizing  a Control  Junctions,  Inc.  multiplexer  system.
Reverse air  cleaning  can be  Initiated by  baghouse dif-
ferential  pressure or  be set to clean continuously  with a
variable time  delay between cleaning cycles.  One compart-
ment per  baghouse  is  cleaned  at  a  time.   One  to three
reverse air  cycles  can be  programmed  for  each  cleaning
period.   Reverse air is provided by drawing cleaned  flue
gas  from the  baghouse  outlet via three (two operating,  one
stand-by)  200  hp  reverse air fans.

For  start-up  and  upset conditions,  the baghouses can be
bypassed through double  louver dampers  which connect  the
 inlet  and outlet  flues.   These dampers will  open  for
emergency  bypass on high or  low  inlet  gas temperature or
high differential  pressure.   During baghouse operation,
reverse air is  utilized to purge the area between  the  two
sets of louvers.
Process  and  mechanical checkout  of the various scrubber
sub-systems and  the  baghouses began  in November, 1982, and
continued  through May, 1983.  On May 24, 1983, the first
coaI  firing occur red.

Throughout initial  firing  of  the boiler on  oil  and the
first few days  of  Initial  coal  fire,  the baghouses  were
operated  on  bypass.   During this bypass period, the com-
partment access  doors  were cracked to vent fresh air  into
the  compartments  and  the hopper heaters were turned on to
help  preheat  the baghouse.
                           10-188

-------
On May 27,  1983,  it was felt that coal  firing  was  stable.
The baghouse  access doors were  sealed  and the  bypass
dampers  closed,  bringing the baghouse  on  line for the
Initial  coating  of  flyash.  The boiler  was  operating at
approximately  60% load  and the  flue  gas  entering the
baghouse  was approximately 320°F.  Unfortunately,  within a
few hours after  the  initial bag coating  process  had  begun,
the boiler  control  system  suffered a momentary  power  loss,
causing  a  unit trip.   The baghouses  were put back on
bypass for  relgnlting of the boiler.

Approximately 24 hours later, the baghouses were  brought
back  on  line.  The reIntroductIon of  flue gas  to the
baghouses was  at essential ly the same load and  temperature
as the previous  day.   The  initial  pressure  differential
across both baghouse  inlet to outlet flues was  two  Inches
water.  After  about  six hours, the  pressure  differential
had risen  to  four  inches  water, and the first  reverse air
cleaning  cycle was  Initiated.

On  June  8,  1983,  after  two weeks of  bag  coating with
flyash only, S0? removal with single pass operation  (lime
only) In the scrubber  commenced.   The first scrubber
operation with recycled material being  added  to  the  slurry
occurred  on June 11,  1983.   On  June  12, the  boiler was
taken  off-line for  a one month shutdown  for  turbine fine
screen removal  and boiler chemical cleaning and  was back
on-line on  July  10.  The scrubber operation was  gradual ly
brought  to design  operating conditions (eg.  35% total
solids and  170°F spray  dryer outlet temperature) during
the next  two  months by Increasing the solids  content of
the slurry  with  recycle material and decreasing  the  outlet
temperature with higher slurry feed flow rates.
PROBLEMS_ENCOUNIERED_AND_IH£iR_CURR£NI_5IAIilS_
The Initial  operation  of the scrubber  and  baghouses  fol-
lowed  the  anticipated  schedule and at no time delayed  the
boiler start-up  or  operation.  The overall  success of  the
start-up effort  was  probably the  result of Joy/Niro's
experience  at the Riverside  full  scale  demonstration
plant.   There were, however,  initially  some mechanical
problems with  various  components of  the  absorber product
conveying  system, atomizer wear plates, and the atomizer
head tanks .
                           10-189

-------
The recycle  lifting  conveyor has  horizontal  sections at
the upper  and  lower extremes.  During the  first  weeks of
recycle  operation, a portion of the  paddle  type flights of
the conveyor  experienced fatigue  failure  due  to the wear-
ing of the paddle  tips on the center  plate  in  the  lower
90° bend  of the conveyor casing.   To remedy this, a guide
rail was  added supporting the chain in  the conveyor  bend,
preventing the paddle  flights from  riding on  the center
p I ate .

Dusting  from  the conveyors made it  necessary  to  redesign
shaft and  dump gate seals, and to increase the venting of
the recycle bin.  Operation of  the variable  speed drag
link conveyor which  feeds ash from the recycle bin  Into
the mix  tank  indicated that  the 3  hp  motor originally
supplied  had  insufficient power  at  the  lower  range  of the
operating speed.  A 5 hp drive will  be  Installed.

The  interior of the  atomizer wheel Is equipped with a
ceramic  bottom wear plate.   During  the  early  operation,
several  of these  plates cracked, possibly due to  thermal
shock  as the  wheel came  Into contact with the  flue  gas and
then the  feed slurry.   A procedural change, initiating
atomizer wheel protection water flow simultaneously with
initiation of  chamber gas flow, has  solved  this problem.

The atomizer  head tanks  (1 foot diameter by  6  feet taI I )
are  equipped  with  a  vertically mounted,  internal,
cylindrical screen.  The feed slurry pumped from  the feed
tank  enters  the  bottom  of the  head tank  and  into the
center of the  screen.  The portion of the slurry  which Is
del ivered to  the  rotary atomizer  must  pass through the
cylindrical screen wall  In  order to enter  the  atomizer
feed  line.   The remainder of the  slurry  continues  through
the center of  the head  tank  and  out the open  top of the
screen  to the overflow pipe, which returns  the  slurry to
the feed tank.  The screen self-cleans as this  return  flow
through  the  center of  the  screen washes  It's  inner
surface.   During  the  early  scrubber  operation  at the
design  spraydown  rate and  full  gas flow,  the  head tank
screens  periodically  restricted  flow to  the  atomizers.
The problem  was found to be  insufficient flow  through the
interior  of  the  screen  to  provide the  se I f-c I ean ing
effect.   Using  the  adjustable pitch sheaves  provided on
the feed pumps, the pump speeds  were increased, thereby
providing additional head tank flow.

One of the prime factors for successful  drying  of the  feed
slurry  is maintaining sufficiently high  feed solids.  A
basic  concept  of general spray drying technology is that
                          10-190

-------
It Is easier  to dry  a  thick,  or  higher solids slurry, than
a thin,  or  low solids  one.   During the start-up  operation
of the  scrubber, periodic  tripping of the recycle feeder
to the mix  tank caused  low  feed  solids.  This, along  with
a period of  operation  without the Implementation of the
gas flow feed forward  signal  to  the atomizer  feed control-
lers, caused some  Intermittent  spray  dryer wall  deposits.
The  deposits  did  not  interrupt scrubber  operation,
however, pluggage  of  the  spray  dryer  bottom  discharge did
occur as dry  lumps  fell  from the dryer walls.   An  Inter-
lock  of  outlet temperature to the feed  solids has been
programmed  Into the  controls  to automatical ly  raise the
outlet  temperature  in the  event  of a  low  feed  solids
content, thereby reducing the  possibility that  deposits
will  develop  within  the  dryers.

BaghoiLse

The first and perhaps  most  significant problem encountered
with  the  baghouse  operation was  high  differential
pressure.  From the  time of  the  Initial reverse air clean-
Ing  at  four  inches  pressure  differential,  it became
apparent that the reverse  air  was  not  effectively removing
the dust cake on the bags.   During the first  two  weeks  of
operation,  baghouse  pressure drops ranged  from four Inches
to over  nine  inches  water  gage,  depending  on  boi ler  load.
For  full  load boiler operation, the baghouses had to be
bypassed to assure  that the  I.D.  fans would  not  enter  a
staI  I condIt I on.

Although it was realized that  the  higher gas  temperatures
through  the  baghouses  (since  the spray dryers were still
off  line) were resulting In  higher gas volumes and a  lower
reverse  air  fan efficiency,  these factors  could not ex-
plain the  excessive  pressure  loss.  Inspection of  the
fabric dust layers  showed  heavy  build-ups  of  caked dust on
the bags.  Laboratory  analysts  of  fabric samples confirmed
that  the bags had  experienced  a dew point  encounter caus-
ing soluble sulfate  formations  resulting  in agglomerations
and nodula  formation.   The  level  of agglomeration and cake
pluggage was  not considered  severe  and the dust  cake was
easily removed by manual agitation of  each  individual bag.

Current  baghouse pressure  differential has stabilized at
approximately  five  and one-half  inches water  gage at
desIgn cond111ons .

During the  initial  two weeks of  operation,  stack opacities
of 10-20$ were observed.  The  reason for visible emissions
                           10-191

-------
from the baghouse  was apparent  after the June 12,  1983,
shutdown, when  It  was discovered  that  several  of  the
originally  supplied bypass  damper  blade seals  had torn
loose,  possibly  due  to high boiler load bypass operation.
Velocities through the louver damper were in  excess of 100
ft/sec.   The design  of the blade seals  was  modified  and
new seals were  Installed.
              lie merits.
The dry scrubber  system  is designed  to  remove 62% of  the
S0? when the  average  sulfur coal (0.68?) is burned and  78$
for the  maximum  sulfur  coal  (1.22?).   Unit  2 of  the
Antelope  Valley  Station  is  scheduled to begin commercial
operation in  July,  1986.   At  that time.  Unit 1  SO- removal
requirements  will  increase to 81? for average sulfur coal
and 89? for  maximum  sulfur coal.  The  S02 emissions  from
Unit 1  and Unit  2  combined are  limited to 3,845 pounds per
hour (0.39 pounds  per  million  Btu),  as  a requirement  of
the Permit  to  Construct  issued by the North Dakota State
Department of  Health.

     ia n.ce_Ies±i
Stack emission  tests  were conducted on August 23,  1983,  to
determine  compliance with  the  Permit to Construct.   The
spray dryers  had  only  operated for  a  total  of about  five
weeks  at  the  time of the emission compliance tests.   The
scrubber system had  not  reached design  operating condi-
tions, but  was  determined  to be  capable  of  meeting
emission compliance  requirements, so the tests proceeded.
The  results of  the  compliance tests  are  summarized  in
Tab I e 1 .

Pe_£±or_ma_n.ce_J_ests.

Performance tests to  determine guarantee  compliance  with
the  Joy  contract were conducted  September  26-30,  1983.
During the tests, the  SO  and  particulate removal system
operated  at  the  conditions specified in the contract for
normal operating  conditions  with the average sulfur  coal.

EPA  reference methods  were used  to  determine SO   and
particulate concentrations and gas  flows  at the  inlet  to
                          10-192

-------
           Table 1.   Emission Compliance Test Results
Parameter
Test Result
Permit Limitation
Gross Generation, MW
Stack Gas Flow, ACFM
Stack Temperature, °F
Total Sol ids In Slurry, %
Inlet S02, PPM
Stack S02, PPM
Stack S02, #/hr
Stack S02, #/MKB
SO Remova 1 , %
Stack NO , #/hr
Stack NO , #/MKB
X
Stack Particulates, #/hr
Stack Particulates, #/MKB
468
1,880,000
222
24
668
171
2,308
.46
74
2,222
.45

72.7
.015






3,845


2,465


210

                                 10-193

-------
and  in  the stack,  and  S02  and  gas flow of  the  bypass
reheat  duct.   Lime consumption  was determined  from the
pebble  lime weigh  belt  feeder,  with  the   lime  being
analyzed for available  CaO.   The  lime  consumption  was
verified with the  slaked lime slurry  analysis and flow
rate.   Pressure drop was  confirmed from  static  pressure
readings taken at  several  locations to account for the
pressure drop from the  outlet  of  the air heater  to the
inlet of the ID fan, and from the  outlet of the  ID  fan to
the entrance of the stack.

An enthalpy balance was  used  to  confirm  the gas  flows and
the pressure drop  used  In  the evaluation.   Electrical
energy  consumption was measured  with  a watt transducer.

Samples of coal,  raw  lime,  slaked  lime,  feed slurry,
treated water, dilution water (cooling tower blowdown),
scrubber Inlet flyash, and recycle  products were  collected
and analyzed.  Temperatures  were  taken of feed  slurry, and
spray  dryer  inlets,  outlets, bypass reheat,  and  stack
gases .
L§-S.±_Re-S_u.lts.

Preliminary  results of the  guarantee  performance tests on
the S0? and particulate  collection system are listed In
Tablet.   As of  this  writing, sample  analyses have not
been completed to  determine the  actual sto I chIometr I c
ratio.  An  estimate is  used  here  based  on correlation of
aval I abIe  data.

During  the  performance  tests, the  ash  conveying  system
from the baghouse to  the  recycle  bin was not conveying
adequate quantities of ash.   To  replace  that ash, a truck
was used to transport ash from the  ash storage  silo  to the
recycle bin.  Preliminary  results Indicate that this ash
contained  higher alkalinity  than  the ash collected from
the baghouse.  This  Issue Is  currently under  investigation
to determine  its  impact on stoIchIometrIc ratio,  if  any.

Testing was also conducted under  conditions without  reheat
by-pass to assess  the guarantees  at 62%,  81$, and  89%
removal.   The preliminary  results  Indicate a slight In-
crease  in  stoichiometric ratio when the  reheat  by-pass was
not used for  62% removal.  A  stoIchIometrIc ratio of about
0.9 was used for 8]%  removal,  which was well below the
guarantee  of  1.1.

A  test  at  the high removal  rate  (89$  guarantee)  was
                          10-194

-------
                  Table 2.  Performance Test Results
Parameter                         Test Result    Guarantee Conditions
Spray Dryer Inlet Gas Flow, ACFM
Spray Dryer Inlet, °F
Spray Dryer Outlet, °F
Approach to Saturation, °F
Total Sol Ids In Slurry, %
Baghouse Outlet, °F
Inlet S0_, ppm
Outlet SO , ppm
Outlet S02, Ib/hr
SO Remova 1 , %
Stolen lometr Ic Ratio
(for 63.9$ removal )
Stolen lometr Ic Ratio
(corrected for 62% removal)
Inlet Partlculate, gr/dscf
Outlet Particulate, gr/dscf
Outlet Partlculate, Ib/hr
Partlculate Removal, %
Total Pressure Drop, In. W.G.
Power Consumption, KW
Reheat By-pass, 10 Btu/hr
2,039,641
314
166
29
41
189
655
220
2,973
63.9

.47

.45
2.8496
.0025
24
99.91
14.96
2,341
7.1





185
800
304
3,844
62



.54

.012
210

13.4
2,757
7.2

                                  10-195

-------
conducted using  data  obtained from  the  plant monitoring
equipment, and correlated  to  the  previous  EPA  reference
method tests.   A s t o I ch Iometr Ic ratio  of  1.45  Is
guaranteed for 89$ removal  with the maximum  design sulfur
content of  1,380  ppm  S02  Into the system.  The SO^  inlet
during the  test  was  only  442 ppm.   Re'moval  of  92$ was
obtained  with a stoIchIometric ratio of  about  1.0.
Although some  preliminary optimization  testing has been
conducted on  the scrubbing system,  additional testing will
resume when  the  unit  returns on-line  In December.  The
testing will  focus  on  maximizing  lime  utilization  and
minimizing  differential  pressure  across  the baghouse and
spray dryers.   The parameters  to be studied during this
testing will  be  sto I chIometrIc ratio,  approach tempera-
ture, and reheat usage.

Further testing  of  this  system  will  also be conducted at
81$ and 89$  S0? removal  conditions.   The  testing  at  these
conditions  will  be used  to  confirm operation  for  the
stringent removal requirements  when Unit  2  begins  opera-
tion and when  the  maximum sulfur  content coal  is burned.
Other  tests  will  be performed to  determine  the actual
turn-down capabilities of  the spray  dryers.

Engineering  studies  are  currently planned  to Investigate
the  possibility of decreasing  the system differential
pressure  by  modifying existing  duct  work.   Replacement of
the  existing  slurry feed pumps and  piping is also being
considered  In  order  to  operate  at   lower  approach
temperatures.   Lower approach temperature  operation  should
Improve lime utilization  and  assist in  maintaining  the
plant water  balance.
                          10-196

-------
                       REEEBENCES
1)    Janssen,  Kent E.  and Erfksen, Robert  L . , "Basin
     Electric's  Involvement with Dry  Flue  Gas
     Desulfurization".  Presented at the EPA Symposium on
     Flue Gas  Desulfurization,  Las  Vegas,  Nevada;
     March,  1979.
2)    Davis,  R.A.,  Meyler,  J.A., Gude,  K.E.,  "Dry SO-
     Scrubbing at Antelope  Valley Station".  Presented  ar
     the  American Power  Conference,  Chicago,  I I I Inols;
     Apr I I ,  1979


3)    Eriksen,  Robert L., "The Development  of Dry Flue Gas
     Desulfurization at Basin Electric Power Cooperative".
     Presented at the  Second Conference  on Air  Qual ity
     Management  in the  Electric  Power Industry, Austin,
     Texas;  January, 1980.
                          10-197

-------
CHARACTERIZATION OF AN INDUSTRIAL SPRAY DRYER AT
          ARGONNE NATIONAL LABORATORY

         P.  S. Farber, C. D. Livengood

-------
                CHARACTERIZATION OF AN INDUSTRIAL SPRAY DRYER
                       AT ARGONNE NATIONAL LABORATORY

               by:   P.S.  Farber and C.D.  Livengood
                    Energy and Environmental Systems Division
                    Argonne National Laboratory
                    Argonne, 111.  60439
                                  ABSTRACT

     Argonne National  Laboratory  (ANL)  is  operating  an  industrial-scale,
coal-fired  boiler  with  a flue  gas  cleaning  (FGC)  system  consisting  of a
spray dryer and fabric filter.   This paper presents  a  description  of  the  FGC
system together  with  a   status  report  for  an  EPA-sponsored  project being
carried  out by  ANL  to   characterize  the operation of  the system.   This
project  involves  a  design  and  economic  analysis   of   the   FGC   system,
determination  of  waste  characteristics,  and  analysis of  system operation
through  monitoring  of   inlet/outlet  gas  streams and sampling  of   various
process streams.   Preliminary data  and material  balances  are presented in
the paper,  as  well as a  proposed performance  model  based on an analysis of
key operating parameters.
                                INTRODUCTION

     "Dry" scrubbers consisting of  spray  dryers and fabric filters or elec-
trostatic precipitators have  recently  emerged as a viable pollution  control
option for  coal-firing  electric utilities and  industries.   Such  integrated
flue gas  cleaning  (FGC)  systems  have several potential advantages over  con-
ventional wet scrubbers,  including  higher reliability,  lower cost (for  com-
bined sulfur  oxides and  particulate matter  control),  lower consumption  of
energy and  water,  and  greater ease  of waste  disposal.   The latter point  is
due  primarily  to  the  production  of  a  dry  powdery  waste  that  is  readily
handled,  transported,  and disposed of in landfills.

     Most existing  or  planned dry  scrubbers  have  been designed  for  systems
firing low-  to  medium-sulfur  coals.  However,  Argonne National  Laboratory
(ANL)  has  been  operating  the  first  commercial  spray-dryer  FGC   system
designed   for  high  (3.5%) sulfur  coal since November  1981.    This  system
treats the  flue  gas from a spreader-stoker  boiler rated at 170,000  Ib/h  of
                                    10-199

-------
steam.   Sulfur dioxide (S02) removals of about  80%  are  achieved routinely to
meet the emission limit  of  1.2  lb/10" Btu  mandated by the State of Illinois.

     Because of  the  unique nature of this  system,  the  extensive  provisions
made for  monitoring  and  sampling,  and the ready  availability of  research
facilities at ANL, the U.S.  Environmental  Protection Agency*  (EPA)  is  spon-
soring  a  program for characterization of  the FGC system.   The  objective  of
this program is  to document  design  and  construction experience  with the  ANL
system and  to  achieve a  thorough understanding  of  its operation  so as  to
develop predictive performance correlations.  The first part  of the program
utilized  the  information   compiled  during   the  specification,  bidding, con-
struction,  and  initial operation phases.    The  analysis  of  specifications
revealed items critical to the procurement  of a  spray dryer  FGC system.   The
bid  analysis  reviewed  the  range of  bids  received  on  the  basis  of both
capital and operating costs.  This analysis  of  equipment and installed  costs
will give potential  users  of spray  dryer  FGC systems an insight to  expected
capital and operating expenditures.

     A  second  part  of the  program  involves  characterization  of  the  FGC
system over a 60 to  90 day period.   This characterization  will entail exten-
sive data acquisition and  stream  sampling,  and  will relate  system operation
to process conditions.

     A third program activity  is  determination  of the properties of the  dry
waste  produced  by the  FGC  system.   Various mixtures  of  flyash and  spent
sorbent have  been subjected to  the EPA Extraction  Procedure  (EP)  Toxicity
Test   in  order  to  determine  the extent  and  composition of  any potential
leachate.   The results are useful in  evaluating  disposal characteristics  and
determining the  partitioning of coal-derived trace  elements between ash  and
sorbent particles.

     This paper presents a brief  description of  the  system,  a  summary of  the
EP  test results,  and selected  operating data obtained in  a preliminary test
program.   The  derivation  of material balances  around key system components
is demonstrated,  and  a simple performance model  is proposed.

                           SYSTEM CHARACTERISTICS

ARGONNE BOILER NO. 5

     The  spray dryer system  is  installed  on Boiler No.  5  at the ANL heating
plant.   No.  5  is the largest of  the five  boilers at the plant,  with a  name-
plate rating of  170,000 Ib/h of saturated steam at  200  psig.   It is  a Wickes
*Emissions/Effluent  Technology  Branch,   Industrial  Environmental  Research
 Laboratory,  Research  Triangle  Park,  N.C.,  Theodore   G.   Brna,   Project
 Officer.
                                    10-200

-------
(now Combustion Engineering) spreader-stoker unit  that  was  installed  in 1965
with the  capability  to fire either  coal  or gas.  Oil-firing  capability  was
added as a backup in 1973 when coal use was discontinued at ANL.

     In 1980, as part  of  the DOE  effort to maximize  coal use and  as an econ-
omy measure  for the Laboratory,  the coal handling equipment  at  the  heating
plant  was  refurbished  and Boiler  No. 5  was  reconverted  to coal firing.
However, it was also necessary  to install an FGC system capable of providing
compliance with the  applicable  State of  Illinois  emission  regulations  shown
in  Table  1  since existing  equipment consisted solely  of  cyclone separators
for  flyash  collection.   These  continue  to  be used  and  considerably  reduce
the flyash burden on the FGC system.

     Several  midwestern and eastern coals have  been  successfully  fired  in
Boiler  No.  5 since  the reconversion,  ranging  in  sulfur content  from  about
1.5% to 4.5%.   Table 2 gives  a  typical analysis for  the coal currently  being
used, which is  from  the Illinois  basin and is  representative of the fuel  for
which the FGC system was  designed.   During acceptance  tests using that  coal,
S02 emissions of  0.95   lb/106 Btu and  total suspended  particulate emissions
of  0.007 lb/106 Btu were measured.

FLUE GAS CLEANING SYSTEM

     Although spray dryer  systems  are commonly referred to as "dry,"  they do
involve the  preparation and handling  of  a slurry.  For  this  discussion,  it
is  convenient to divide the system into its wet and  dry subsystems.   The  wet
subsystem  encompasses   pebble   lime  storage,  recycle   powder  storage,  lime
preparation, and slurry preparation.   The dry  subsystem is basically  the  gas
train and associated equipment, and  includes  the spray  dryer, fabric  filter,
ducts,  booster  fan,  and  air  compressor.   The  relationship between  major
components is illustrated in Figure  1.

     The  feed slurry is composed of  a mixture of fresh  slaked lime  and  re-
cycled  "spent"  sorbent.   The recycled material contains  a significant  com-
ponent  of  unreacted  calcium and  thus raises the  effective  Ca/S  ratio  above
that indicated  by  the  fresh lime alone.   Pebble lime  from  the lime  silo is
fed to  a weighbelt feeder,  which  supplies a paste-type slaker where  the lime
is  converted  from  calcium oxide  (CaO)  to  calcium  hydroxide (CaCOH^).   Some
water is  added  at  the   outlet of  the slaker to dilute  the  milk of lime to a
concentration of  approximately  15%  (by weight) solids.   This milk  of  lime
then flows  to the milk of  lime  tank through a  rotary  screen,  which  removes
the "grits," or inerts.

     From the milk of  lime tank,  the lime slurry is  pumped to the slurry mix
tank.   In this  tank,  milk  of lime,  recycled spent sorbent  powder,  and some
additional  water  are   combined  into  an  approximately  35-40%  (by  weight)
slurry.   This  slurry   is  then  pumped  to another rotary  screen  (to ensure
                                    10-201

-------
                 TABLE 1.  EMISSION LIMITS FOR BOILER NO. 5
                      Pollutant                      Limit


                 Sulfur  Dioxide                 1.2  lb/106  Btu
                 Particulate Matter             0.1  lb/106  Btu
                                         and    <20% Opacity
                    TABLE 2.   TYPICAL COAL  CHARACTERISTICS
                       Coal Parameter              Value
                    Heating Value, Btu/lb       12,027
                    Moisture, %                      9.59
                    Ash, %                           7.40
                    Carbon, %                       66.98
                    Hydrogen, %                      4.65
                    Nitrogen, %                      1.48
                    Sulfur, %                        3.32
                    Chlorine, %                      0.06
                    Oxygen, %                        6.52
removal of  any  large particles  which  might clog  the  atomizer nozzles) and
then to the  slurry  feed  tank.   Overflow from the slurry feed tank goes back
into the  slurry  mix tank, so  there  is a  continual  circulation between the
tanks.   Slurry from the feed tank is pumped, at  a  constant high flowrate, to
a  head tank  located  above  the atomizer.   A  control valve  regulates the
amount of slurry  fed  to  the  atomizer,  with the excess being returned to the
feed tank.   This  returned slurry also  passes through the rotary screen that
is filtering the  stream  from  the slurry mix tank.

     Flue  gas, exiting  the boiler's induced draft (ID) fan,  passes into a
modified breeching  at the existing stack.   A  guillotine damper diverts the
flue gas flow into the FGC system ductwork  leading to  the spray dryer.  This
inlet ductwork splits the flue gas into two  streams.   One stream, with  about
60% of the  gas flow,  is  directed into  a roof  gas  disperser,  located on the
top of the spray  dryer (Figure  2).  The remainder  of the gas stream  enters a
central gas  disperser, located  in the  middle of the spray  dryer.    Both gas
                                    10-202

-------
                                                           To
                                                      Atmosphere
o
I
r-o
o
LO
                                                                                Slaking
                                                                                 Water
                                                Dust
                                                Suppression
                                                Water
                                                 -CM
Waste
Disposal
                                           Pebble
                                           Lime
                                           Silo
                                                                                           Slaker
                                                                                           and Milk
                                                                                           of Lime
                                                                                           System
                                   Figure  1.  Argonne National Laboratory FGC System

-------
  Sorbent Flow  into Rotary  Atomizer
                Central
                Gas
                Disperser
    Spent
   Sorbent
    Outlet
Figure 2.  Spray Dryer
          10-204

-------
streams, upon entering the dryer, are  given circular motions  with their  main
directions of flow being opposed to  each  other.   In the spray dryer  the  feed
stream  of  slurry is  atomized  into  fine droplets,  through  the  centrifugal
action  of  the  rotating disk,  and  introduced into  the  gas  stream.  Some of
the dry powder  produced  in the spray  dryer falls  to the bottom of  unit and
is picked  up by  a  drag link  conveyor for  storage in the recycle/disposal
silo.

     The remainder of  the  powder is entrained in the gas stream and is  car-
ried  into the  four-compartment  pulse-jet   fabric  filter.     In  the fabric
filter these particles are separated from the  gas  stream by fiberglass bags.
The cleaned  flue  gas  exits the baghouse  and travels  through  the outlet  duct
to the  booster fan.   From  the booster fan the gas  enters a new breeching
section of the  stack,  and then into the stack  itself  for  discharge to the
atmosphere.

Wet Subsystem

Pebble Lime Silo—
     This  vessel  is  constructed of  carbon  steel,   and  is  12  ft in  diameter
with a  32  ft high straight side.   There  is a partial-cone lower section on
the vessel,  with a  "live-bin" vibratory bottom.   The  silo  is  designed to
contain up to 110 tons of  high-calcium pebble  lime, which is  approximately a
7-day supply at full boiler load.

Recycle/Disposal Silo—
     Standing 46  ft  high with a 20  ft  diameter,  this silo  is the larger of
the two dry  materials  silos  in the FGC system.  As with the  lime  silo,  this
vessel is  constructed of carbon steel.   However,  it has a flat bottom and is
equipped with an  air fluidizing system (air-slide)  to assist  in powder flow.
Two outlets  in  the  bottom are  for  recycle  of  spent sorbent  and for sorbent
disposal.  The  recycle  silo has a capacity  of  180  tons of  dry powder which,
at full boiler load,  is sufficient for 7 days'  accumulation.

Lime Slaking System—
     This  system  consists  of  a Wallace  and Tiernan paste-type lime slaker
(Series A-758),  with a  capacity of  2000  Ib/h, and  a mechanical weighbelt
feeder  (Model  31-120AV).  The feeder  is equipped with  an adjustable-speed
belt drive and a  transmitting  switch for  actuating the totalizing throughput
counter.   The slaker,  made of heavy gauge  steel,  consists  of a  slaking  com-
partment  containing  two sets  of counter rotating intermeshing paddles for
mixing,  a  dilution  chamber with  rakes for  agitation,  and  a   dust and  vapor
arrester.  In the slaker,  water and pebble  lime, in a ratio of approximately
2 to  1  by  weight, are fed continuously into one end  of the slaking  compart-
ment.    Lime  feedrate governs  the  operating rate  of  the entire  system, and
this feedrate is  controlled  on the basis of the liquid level in  the milk  of
                                     10-205

-------
lime tank.   Lime-paste  consistency controls the addition  of  water.  An  in-
crease in torque  on  the mixing shafts  (indicating  a thicker  paste) opens  a
water-control valve to admit  additional  water.

Vibrating Screens—
     These screens are both "Rotex Single Surface Liquatax  Separators," 2 ft
by 2 ft with  4 in. inlet  nozzles  and  3  in.  outlet  nozzles.  One screen,  the
classifier screen, has a 40-mesh stainless steel screen and is  at  the outlet
of the lime  slaker.   The  purpose  of this screen is to separate the "grits"
from the slaked lime slurry.  The overs from this screen flow through a pipe
into a  dumpster  for disposal.  The  slurry  passing  through the screen dis-
charges to the milk of lime collection tank.

     The other screen,  called the  slurry  feed  screen, is a 6-mesh  stainless
steel screen  which processes  the  slurry from  the slurry mix tank as well as
the returned slurry stream from the atomizer head tank.  As with the classi-
fier  screen,  the  overs  are  sent  to  a dumpster  for disposal.   The slurry
discharging from this stream falls  into  the slurry feed  tank.

Tanks—
     There are four tanks  associated with  the wet subsystem.  These are  the:

     •  Milk of Lime Collection  Tank,

     •  Slurry Mix Tank,

     •  Slurry Feed Tank,  and the

     »  Head Tank.

     The milk of  lime  collection tank  is an  agitated, baffled tank with  a
600-gal capacity.  The  tank, constructed  from 1/4-in. thick ASTM-A283 plate,
is  60  in.  in diameter  and  77 in.  high.  Based on  the  maximum milk of lime
flow, the tank residence time is approximately 20 minutes.

     The slurry mix tank and the slurry feed tank are identical in construc-
tion  (920 gal), both being 60 in. in  diameter and 101 in. high.  As with  the
milk  of  lime collection tank,  they are agitated,  baffled  tanks constructed
of  1/4-in.  A283  carbon  steel.   Their  residence times are each  approximately
20 minutes  based  on maximum slurry flows.  In a sense these two tanks may be
considered  as a  single  large  unit.  This  is  due to the fact that  the slurry
mix tank continuously feeds the slurry  feed tank at a rate  of 45 gallons  per
minute (gpm), with the  overflow from  the slurry feed tank  being returned to
the slurry mix tank.

     The head tank, located above the atomizer in the spray-dryer  penthouse,
is 6  in. in diameter  and  4 ft high,  and is  constructed of  carbon  steel.   It
                                     10-206

-------
supplies  a  hydrostatic  head  for  the  slurry feed  to the  atomizer.    Feed
slurry is pumped to the head  tank  at  a constant  rate of 45 gpm.   The flow of
slurry  to  the atomizer  from the  head  tank  is  modulated by a  temperature-
controlled valve in  the  line between the  head  tank and  the atomizer.   This
valve responds to controller  signals  to ensure  a constant gas  temperature at
the spray dryer outlet.  Any  slurry  in the head tank that is not sent  to the
atomizer overflows into  a  return  line  to  the slurry feed screen.   The  con-
stant high  flowrate  to  the  head  tank  is  necessary  to ensure  that,  during
periods  of  low slurry  demand by  the  system,  pipeline velocities  are  kept
high enough  (6-8  ft/s)  to prevent the  feed slurry from  settling out  in the
long vertical feed line.

Pumps—
     All pumps  (milk  of  lime, slurry transfer,  and slurry  feed)  are  Warman
1-  1/2  in.  x 1 in.  Type BM  with  V-belt drives  and overhead motors.   These
pumps are  split-casing  rubber-lined units with  rubber-lined open impellors.
The  pumping  systems  are  all  redundant, with an identical  "spare"  pump  con-
nected with  parallel  piping  and  valves  to the operating  unit.   Each pump is
designed to  deliver  45 gpm  of fluid in order  to  ensure high velocities in
the piping systems.

Agitators—
     These  Galigher  Model  GWO motor-driven  agitators are, along with  the
pumps,  the  only rubber covered equipment  in the  entire  sorbent preparation
system.   Their purpose  is  to ensure that  all  slurries  stay in suspension,
although  the slurry  mix tank agitator (which  is fitted with  two sets  of
impellors)  has  the  additional duty  of mixing  milk  of  lime,   recycled  dry
sorbent, and dilution water into a uniform slurry.

Dry Subsystem

Spray Dryer—
     This is  a  dual-inlet  spray  dryer  of  carbon steel construction designed
by  Niro  Atomizer.   The  unit  is  27  ft  7 in. inside  diameter,  with a  19 ft
straight section and  a  19  ft long cone  bottom.  This  cone  bottom has  a 3 ft
diameter bottom outlet which is  connected to a  Niro-designed  powder cooler,
no  longer  utilized  in  the  process.   With an  interior  volume  (less  the
central  gas  disperser  volume) of 15,000  ft ,  the  spray  dryer has  a gas
residence  time  of 12  seconds at  the maximum boiler  load.   This  residence
time is  based on the  inlet  gas volume to the spray dryer.   Should an average
between  the  inlet  and the outlet  volumes  be  used,  the value rises to almost
14 seconds.

     The spray dryer was constructed  with  a dual gas inlet, such that 60% of
the flue  gas enters  through  the roof gas  disperser,  while the remaining gas
flow  enters  via  the  central gas dispenser.    These two  gas   streams are
                                     10-207

-------
projected  against  each  other,   forming  a  turbulent  swirl  into  which  the
slurry is  injected  by  the  high speed atomizer.   This slurry mixes with  the
hot flue  gas  stream, reacting  with the sulfur  dioxide, and drying into  a
fine  powder.  Some  of this  powder falls to  the bottom  of  the spray dryer,
where it is picked  up  by a drag-link mechanical conveyor.  The remainder of
the powder is  entrained in  the gas stream and exits  the spray dryer.

     The  atomizer  used  in the  Argonne  FGC  system  is  a  Niro  Model  F-160
equipped with a  200 hp  3570  rpm motor.   A  gear drive, within the  atomizer
body, increases  the  rotational  speed to 14,800  rpm.   This high speed  spins
the  8-1/4  in. diameter  atomizing disk  to  a  tip  speed  of  533  ft/s.    The
atomizing  disk  is  constructed  with eight  equally  spaced  3/8-in.  diameter
holes, through which the slurry  is injected  into  the gas stream.

Baghouse Fabric Filter—
     The baghouse fabric filter is a Joy Manufacturing, Model 6012, pulse-
jet  unit.    It  consists of  four  compartments,  each  having  28 rows  of  10
filter bags.   The filter  compartments are  constructed of carbon  steel,  and
equipped with electrical plate-type heaters  on the  hopper  (lower  surface)
walls.   These  hopper  heaters  are thermostatically  controlled,  and ensure
that  no  moisture  condenses  in the compartments.   Dust-laden  gas enters each
filter compartment in the lower plenum.  The  gas travels upward, through  the
bags,  into the  upper  plenum of  the  filter  compartment.   From  the filter
compartment, the gas travels through ductwork to  the booster fan.

      The filter  bags are fabricated  of  a woven 16 oz  fiberglass fabric with
a Teflon coating.  These filter  bags are 6  in. in diameter  by 12 ft  long  and
are fitted  onto  wire mesh  cages.   Each  row  of filter bags has a pipe, which
is  located  over  the bags and  runs  the entire  length  of the  row.  These  pipes
have  holes  in them  above each bag and are connected  through  solenoid valves
to  a  compressed  air  header.   As each solenoid is actuated in turn,  it  sends
a  pulse  of compressed  air  into the pipe.   The compressed air exits through
the  holes  above the bags,  entrains some additional gas with  the aid of  a
venturi  nozzle fastened  to  the  top of the bag cage  combination, and travels
down  the length  of the  bag.   This  pulse  of  air breaks loose some of  the cake
collected  on  the  outside of  the filter bag.   This  broken cake falls to  the
bottom of the filter compartment, and  into another  drag-link conveyor.

Booster Fan—
      This  is  a  belt-driven induced draft fan manufactured  by Chicago Blower
Corporation (Model  5414).   The  fan wheel  is 54-1/4  in. in  diameter and is
driven  (via the  belt drive)  by  a 250 hp,  1800  rpm motor.   The  fan is  de-
signed  for 62,000  ft3/min  of   gas  at  19  in.  W.C.  static  pressure  while
running  at 1335 rpm.   Flue  gas  from  the  baghouse  fabric filter system is
ducted to  this  fan  and, from there, directly to the stack for discharge to
the atmosphere.
                                    10-208

-------
                           RESULTS AND CONCLUSIONS

DESIGN AND ECONOMIC ANALYSIS

     Five vendors  submitted proposals  in  response to  the  Request for  Pro-
posals  (RFP)  issued by  ANL for  installation  of  a dry FGC  system.   In the
spray  dryer,  four  vendors proposed  to use  rotary  centrifugal  devices to
atomize  the  lime   slurry   feed;  the  fifth proposed  two-fluid  nozzles to
achieve the atomization.   The diameter of  the dryer chambers ranged from 23
to 27.6 ft, with straight-wall  height  ranging  from 10  to  22  ft.   Solids  con-
centration  of  lime-slurry  feed varied  from 14  to 35% by weight.  Estimated
pressure  drop across  the  dryer, at  maximum boiler  operating  conditions,
varied from 1.5  to 5 in. W.C., with estimated power consumption  for operat-
ing the dryer  varying from 66 to  111  kW.

     For  collection of  flyash  and spent  sorbent, reverse-air and pulse-jet
fabric  filter baghouses were  proposed, with  Teflon-coated  fiberglass  bags
used  in four   of  the proposed  systems.   The gross  air-to-cloth ratio for
these baghouse systems  ranged  from 1.34 to 4.60, and the resultant net  air-
to-cloth ratio ranged from  1.81  to 6.16.   The  estimated pressure  drop across
the  baghouse  varied from  5.5  to 7.5  in.  W.C.,   resulting  in  an estimated
power requirement for operation of this subsystem of 17 to 30 hp.

     As for overall system chemical  and utility  consumptions, the estimated
lime  (92%  pure pebble  lime)  utilization rate for  the proposed FGC systems
ranged from 1,130  to 1,600  Ib/h at maximum boiler capacity.   Estimated total
water usage varied from 20 to 40 gpm, and estimated  total system pressure
drop varied from 8.75 to 16 in.  W.C.,  with a resultant  estimated  total power
consumption for operation of these FGC systems  ranging from 239  to 545  kW.

     Both the  total system capital costs  and the  estimated  annual operating
costs varied  over  a range of  more  than  two to  one  among  the  various  pro-
posals.   The  final capital  cost for the  system as  constructed  was  $3.5
million.

     After  completion  of the  vendor  selection process,  a  letter of  intent
was  awarded to  Niro Atomizer,  Inc.  in  the  first week of November  1980.
Ground was broken  after  general  facility arrangement drawings were finalized
for construction in the  second  week  of April 1981,  and erection of steel and
platforms was  completed on  August 7,  1981.   Delivery  of major equipment  com-
menced in the first week of June 1981  and  by the  first week  of  October,  both
the baghouse  and  spray  dryer subsystems had been installed.  All mechanical
work, piping,  instrumentation,  and  electrical work were  completed  in  mid-
October, and  final painting and  insulation were  finished in the  first  week
of November 1981.   Construction  was  about  95%  complete at this  time  (about  1
year after  the  start of the project),  and  Argonne and Niro decided to  move
into the start-up phase of  the project.
                                   10-209

-------
WASTE CHARACTERIZATION

     Waste from  the Argonne  system,  which  consists  of  a  mixture of  spent
sorbent and  flyash,  is  currently landfilled  onsite.   Spent sorbent  is  dis-
charged  (see Figure  1)  from  the  recycle/waste  silo  into  a  dump  truck.
Approximately 12% of this material is flyash.   The  bulk of  the  flyash is re-
moved from the  flue gas by cyclones for  reinjection  into the boiler.   That
ash which is not  reinjected  is collected in a separate  silo and  transported
by truck  to  the landfill site.   Various mixtures  of  ash and spent  sorbent
have been subjected to  the EP Toxicity Test   in order to determine  the ex-
tent and  composition  of any potential  leachate.   The results indicate  that
the leachate composition  is  a  function  of both the pH of the leaching solu-
tion and the composition of the waste  (ash/sorbent)  blend.

     Samples of both  spent  sorbent and  flyash were collected under  typical
operating conditions  while  firing the  coal  specified in Table  2.   The S02
removal was  approximately 80% at the time.    The spent  sorbent  (which  also
contains  a  small  amount of flyash) was obtained  from a sampling  tap  at the
base  of  the storage  silo,  while  the  flyash  sample  was  obtained from the
boiler's cyclone separators.

     The test procedures detailed in Ref. 1 were  followed to generate liquid
samples for  elemental analysis.   However, in several cases the pH could not
be adjusted  down  to the target value  of 5.0 ± 0.2 due to the high available
alkali  in the  spent  sorbent.    If  the  pH was below  5.0  (pure  flyash),  no
adjustments were made.

     An  Inductively Coupled Argon  Plasma Spectrometer  (ICAP)  was used for
elemental analyses  of unblended  flyash  and  sorbent  samples.  These measure-
ments involved  32  different  elements.   For the leachate  analyses,  a  Perkin-
Elmer  Model  5000  atomic  absorption  spectrometer was  employed.    Twenty-two
elements  were measured  in these tests.    In  each case, two independent  mea-
surements were made for each sample.

     One  purpose of the analyses was  to compare the waste  streams  from this
system  with those from  other  units  in  operation.    Results  for  several
elements  are shown in Table 3,  and  are  compared to  three  samples  of spent
sorbent from lime-based  spray  dryers  treating flue gas  from low-sulfur coal
combustion,   as  analyzed  by  Radian for  the Electric Power Research Institute
(EPRI).3

     Since the  ANL boiler employs high  efficiency  cyclones for  flyash col-
lection,  it  is  understandable  that ANL's spent  sorbent would contain  less
flyash  than  is  found in  the  waste  from  a utility  system.   Thus,  one would
expect  that  the utility  waste  properties would be  found  between those  of
ANL's  two waste  streams if  all other  factors  were  equal.    This  fact  is
illustrated  by  the  amounts  of  iron  and aluminum  (common constituents  in
                                    10-210

-------
                 Table  3   ELEMENTAL ANALYSIS  OF SAMPLES
Concentration, ppm
Element
Aluminum
Antimony
Arsenic
Barium
Beryllium
Boron
Cadmium
Calcium
Chromium
Cobalt
Copper
Iron
Lead
Magnesium
Manganese
Mercury
Molybdenum
Nickel
Potassium
Selenium
Silver
Sodium
Strontium
Thallium
Tin
Titanium
Vanadium
Zinc
ANL
Flyash
68,100
<500
<500
130
40
500
70
7100
130
80
150
367,000
150
3000
210
ND
<10
460
2600
ND
<10
2800
250
ND
<30
4000
270
330
Research-
Cottrell
Waste3
41,000
<8
14.2
400
5.1
NDa
<1.0
170,000
50
14
84
22,000
4.4
12,000
200
ND
<0.5
26
3100
6.1
<0.5
12,000
1800
<25
<30
4900
70
43
Rockwell-
Joliet
Station
Waste3
71,000
<8
33
190
8.5
ND
<1
97,000
54
15
81
44,000
17
13,000
110
ND
8.7
39
6300
4.7
0.5
15,000
1800
<30
<30
5200
120
92
Joy/Niro
Riverside
Waste3
58,000
<8
30
350
4.3
ND
<1.0
150,000
52
4.9
16
20,000
<20
15,000
630
ND
16
215
4300
<20
<0.5
2700
1900
<25
<36
3100
580
37
ANL
Spent Sorbent
3200
<500
<500
20
4
500
<5
283,000
20
<10
40
6300
130
5500
100
ND
ND
30
1700
ND
ND
1400
170
ND
<30
230
30
120
aNot Detected.
                                 10-211

-------
flyash) in the three utility  sorbents  compared to ANL's flyash  and  sorbent.
Iron comprises over 35% of the total mass  in Argonne's  flyash, with  aluminum
being about 7%.   In  the spent sorbent, they are  found  to be 0.6% and  0.3%,
respectively.   The utility spent  sorbents  averaged 2.9% iron and 5.7%  alum-
inum.   Examining  many  of the  trace  elements  normally  found  in flyashes,
similar comparisons could  be  made  between  the ANL  and  utility wastes.

     On the  other hand,  different  relationships  arise  because  the  utility
sorbents are all the result of combustion  of low-sulfur western  coal whereas
ANL's  flyash  is  from high-sulfur midwestern coal combustion.  For  example,
the presence  in  the  utility  wastes  of  large amounts  (compared to either  the
ANL  sorbent  or  flyash)  of potassium  and  sodium  is  probably  due  to  basic
differences between  eastern  and  western flyashes.   The fact that magnesium
is  detected  in  the  utility  samples at levels ranging from  three   to five
times  that found  in  the Argonne  samples could be due to two factors.  Cer-
tain western flyashes contain large amounts of magnesium compared to eastern
ashes.  Another  factor  could  be  related to the type or quality of lime used
in  the FGC system.   Argonne  uses  a  lime which  contains approximately  2%
magnesium, and it  is possible  that  the lime used  in the utility units had a
higher  magnesium content.   Coal-related   differences in  the  nature of  the
wastes are also  seen in the  analyses of barium and titanium  (more prevalent
in  the utility  sorbents than in either the flyash  or sorbent from ANL)  and
boron  (undetectable  in  the utility  sorbents,  but found  at  the  same  (high)
level  in both the ANL sorbent and  flyash).

     Figures 3-5 show typical results from analysis of  leachate  from the  ANL
sorbent/flyash mixtures.  A  more  complete  presentation  may be found  in Ref.
4.  In general,  three different pH-sensitive trends were observed:   1)  metal
concentration increasing with increasing pH,  2)  a decrease in concentration
with  increasing   pH,  and  3)  specific  sensitivity  (high  concentration)   at
approximately pH 5.

     The  variation of   boron  concentration with  percent  sorbent (and  hence
pH) is given  in  Figure  3.  In  this  case, concentration  rises  with increasing
pH,  although  a  slight  dip  is  noticed at 75% sorbent,  or  about  pH  11.8.
Since wastes  from  a dry scrubber  system will tend to  be  basic in nature when
initially  placed in a landfill or other site,  it  may  be  concluded that  until
the nature of the  waste is modified  by  soil conditions  extensive leaching of
boron  and other similarly  behaving elements may occur.

     Plant growth  experiments using soil treated  with dry  scrubber wastes (0
to  4%  by  weight)  have  been  conducted  at   Argonne.^   For both corn  and soy-
beans, the biomass decreased as the amount of  dry scrubber wastes increased.
Boron  concentrations in the leaves  of  the corn  and  soybeans grown  in soil
treated with  scrubber  wastes were  found to  be up to 20 times  greater than
those found in the leaves  of  the  control (no dry scrubber wastes).6
                                    10-212

-------
                                                                     Legend
                                                                    O PPM
                                                                    X pH
                        0.25
                                     0.50
                               FRACTION SORBENT
                                                  0.75
             Figure 3.  Boron Concentration in Leachate Varies
                        with Fraction Sorbent and pH
     Several  elements,  including cadmium  (Figure  4),  were  found  to  have
lower concentrations  in leachate from the sorbent  and  sorbent  mixtures  than
from the  100%  flyash samples.   This  trend  also follows for pH, in  that  the
concentrations  in  the  leachate  decreased as  pH  increased.   This effect  is
not  unexpected,  as  it has  been  seen  for  these  elements in  many  other
wastes.    If  leaching of cadmium compounds is  actually  retarded  by  high pH,
impoundment of  certain chemical wastes,  which contain these and  like  mate-
rials,  together with  dry scrubber wastes  may  offer distinct advantages  which
should be explored.

     Perhaps most interesting are the cases where response was greatest at a
pH  of  approximately  5.    Antimony   (Figure  5)  showed a concentration  of
approximately 60 ppb  at pH  5  while  levels for the 100% flyash (pH 4) and the
pure waste  sorbent   (pH  12) were virtually  nondetectable.   Similar results
were found  for arsenic, mercury,  and selenium.   In the  toxicity test  pro-
cedure,  pH control  of the  material is carefully  specified.   The  reasons for
the high  leaching  sensitivity of certain metals  at,  or  about,  the pH speci-
fied in the EP  test procedure are not clear.   Further  work  to  determine the
                                    10-213

-------
                                                               1-14
     Q
     <
     O
                       0.25          0.50
                              FRACTION SORBENT
            Figure 4,
Cadmium Concentration in Leachate Varies
with Fraction Sorbent and pH
                                                                    Legend
                                                                   O PPM
                                                                   X pH
reasons for this  occurrence  (e.g.,  the exact  chemical form of the  metallic
compounds  in the waste,  and  equilibrium conditions during the  test)  appears
to be warranted.

     Under RCRA,  if the  EP  filtrate has a concentration of a substance  that
exceeds 100  times  the  Interim  Primary Drinking  Water  Standard,  then  the
waste can be classified as hazardous.   Presently wastes from  flue gas clean-
ing  systems are  excluded from  regulation as  hazardous wastes.  Although  no
element had  a  concentration greater  than  100  times  the standard in  ANL
tests, it  should  be noted  that the chromium  levels  were within 2% of  the
maximum.

HIGH SULFUR COAL  TEST BURN

     During April  1983,  ANL and the Consolidation Coal Company  conducted a
test burn of very  high  sulfur (4.2-4.5%) coal in  Boiler No.  5.   This  test,
complementary   in  concept  to  the  EPA-sponsored  work,  was  undertaken  to
demonstrate:
                                   10-214

-------
           80-,
         o
         OQ
         Q_
         Q_
                         0.25
                                      0.50
                                FRACTION SORBENT
                                                   0.75
                                                                      Legend
                                                                     O PPB
                                                                     X pH
             Figure  5.   Antimony  Concentration  in Leachate Varies
                         with  Fraction  Sorbent and pH
•  Ninety  percent  or  greater
   period,
                                           removal   over  a  100  hour
     •  Ninety  percent SC>2 removal  at  a stoichiometry  (inlet  moles
        of CaO  divided by  inlet moles SC^) of  1.4 or less, and

     •  Seventy percent  SC>2  removal  across  the  spray dryer  alone
        (the balance being attained  in  the baghouse).

     The  test  consisted  of  extensive  sampling  and  analysis  of  process
streams, coupled with  acquisition of process data from plant instrumentation
and manual sampling  (EPA Method 6) of the flue gas streams (inlet, exit from
the spray dryer, and stack).

     Although  much of the data  still  remains  to be  analyzed,  preliminary
findings indicate  that all three  goals  were achieved.   In addition,  a range
of removals was achieved,  which has allowed a comparison of the test results
with  data  obtained  using the  nominal  coal of  3-3.5% sulfur.   A  plot  of
                                   10-215

-------
percent sulfur  removal  as a function  of  external stoichiometry is  given  in
Figure 6 and  shows  that,  while 90% removal was attained for  the  3.5% sulfur
coal at a  stoichiometry  of  about 1.15, a stoichiometry of  approximately  1.3
was required  for  the  same removal when treating flue gas from  a  4.5% sulfur
coal.   It  is important  to  point out  that  the stoichiometries being quoted
are "external"  stoichiometries and  reflect  the amounts of  fresh  pebble lime
entering the system.  From a more  fundamental  standpoint,  it  is important  to
examine  the   internal  stoichiometry as  a  function of  system  performance.
This  is  due  to the  fact  that  the  internal  stoichiometry  also includes  the
available  alkali  in the  recycle  powder.   It  is  this  combined alkali which
actually drives the  S02 removal reaction in  the  spray dryer,  and it should
be  taken  into  account  when  analyzing  system  operation.    Based  on data
analyzed to date, Argonne has  determined  that the  internal stoichiometry  is
(depending  upon  system  operation)  15-20%   greater  than   the   external
stoichiometry-

     Frequent  process  stream  samples  were  taken  and analyzed  during  the
testing period.   For a set of  samples  taken  on April 8, the compositions are
shown  in Table  4.  At that time,  the  average  S02 removal was  79.6%, with a
temperature drop  in  the  spray  dryer of 195°F, and an approach  (to dewpoint)
temperature of  22°F.  Examination of the results  indicates  that the  ratio  of
calcium sulfite to calcium sulfate  is  on the  order  of  9 or  10 to  1.   This  is
significantly  different   from  wet  lime  scrubbing  systems  where  ratios   of
approximately  3 or  4  to  1 are  typical.    It  is felt  that  this difference
exists due to  the relatively  short time (2-4  seconds) that a liquid droplet
actually exists in  a spray  dryer.   These short  droplet  lifetimes,   in a dry
scrubber,  do  not  allow  the  oxidation reaction to proceed to  the  extent that
it does in a  wet  scrubber, where  droplet lifetimes  are an order of magnitude
longer.

     The average  process  flows are shown  in Table 5.  These  flows,  together
with  the  compositions  in Table 4,  allow  acceptable material balances to  be
developed  for system components.  For  example,  taking  a CaO material balance
across the milk of lime tank we have:

    a)  In:   (860.2 Ib/h) (0.946 available  alkali)  = 813.75 lb Ca°
                                                               n

    b)  Out:   (8.7 gpm) (1.09 s.g.) [0.229  fraction  Ca(OH)2]  (500

                       lb Ca(OH)?
              = 1085.8 -         L
              =  821.7
                           h
                       Ib CaO
                                    10-216

-------
    100
 D
 >   90
 E
     80-
(S)


~c
 0)
 o

 0)
Q_
70-
     60
           3.5% Sulfur
                       4.5% Sulfur
       0.7  0.8   0.9   1.0    1.1   1.2    1.3   1.4   1.5    1.6

                  Stoichiometry (CA/S  Ratio)


  Figure 6.  Percent Sulfur in Coal  Affects Performance




          TABLE 4.  PROCESS STREAM COMPOSITIONS

Percent by Weight
Component
Ca(OH)2
CaS03 . -j H20
1
CaC03
H20
Inerts
Feed
Slurry
14.14
9.60
0.23
1.04
71.70
3.29
100
Recycle
P owde r
11.50
60.87
6.71
5.98
2.46
12.48
100
Spray
Powder
11.89
54.42
5.46
8.46
3.25
16.52
100
Milk
of Lime
22.9
-
-
77.1
-
100

                          10-217

-------
    TABLE 5.   AVERAGE PROCESS STREAM FLOWS



Pebble Limea, lb/h                     860.2

Milk of Lime to Mix Tank, gpmb         8.7

Slurry to Spray Dryer, gpmc            14.77

Dilution Water to Mix Tank, gpm        4.6

Recycle Powder to Mix Tank, lb/h       1904

Spray Dryer Bottoms Rate, lb/hd        979

Waste to Disposal, lb/he               3012

Stack Gas Flow, dscfmf                 38,100

S02 in, lb/h                           872

S02 out, lb/h                          178
a@ 94.6% available alkali
b+8% level change in milk of lime storage tank
 over 24 hour period = 59 gallons
cNo change in slurry tank level
 Based on 2.5 minute sample collected
elncludes 1.3 ft drop in silo level = 14,294
 Ib or 596 lb/h.  Net disposal = 1814 lb/h
 based on a settled density of 70 Ib/ft  and a
 solids moisture in the truck of 25%.
 Dry standard cubic feet per minute.
    c)  Accumulation = -52-Sgl (1.09) (8.34 Ib/gal) (0.229)

                            Ib Ca(OH)7
                     = 5.12 	_	2- = 3.9
                                 h
    d)   In = Out plus Accumulation

         813.75 = 821.7 + 3.9 = 825.6
                    10-218

-------
                   e)   Checking for error in the balance,
                            825«6~813.75   n m /     •> / »
                          = -     , - = 0.014, or 1.4%
                                     . o
     Examining the operation of  the  system,  it was determined that the same
fractional removal of S02  is  generally achieved across both the spray dryer
and baghouse.   For example,  if a 50% removal of SC>2  is  detected based on
inlet and  outlet  concentrations at  the  spray dryer,  then  a  50% removal of
SC>2 will be  seen  across  the fabric filter, based on the inlet concentration
to the filter.

     Therefore,  for the overall removal of SC^  of  79.6% on  the 8th of April,
approximately 55%  removal  of S02  occurred  in the spray dryer.   Using this
value in a balance around  the dryer,  it was  found  that  for  the lime reacting
(9.79 Ib moles /h), the products  were:

    CaS03 . y H20  = 66%

            CaC03  = 24%

    CaS04 • y H2°  = 10%


which leads to the conclusion that carbon  dioxide  (C02) pickup does occur in
the spray dryer.

     Performing a  similar balance over the baghouse show that:
     CaS03 .  y H20 = 63.98%

     CaS04 .  y H20 =  7.47%

             CaC03 =  7.23%


           Ca(OH)2 =  4.66%

               H20 =  2.78%

            Inerts = 14.17%

This  gives  the  approximate composition  of  the  powder  exiting the  fabric
filter.   It  is  clear  that  the  system was  not  at  steady  state, since  the
combined  compositions   of  the   spray  dryer   powder   and  the   baghouse
                                    10-219

-------
(calculated)  powder, does  not  equal the  composition of the recycle  powder.
The attainment of steady state is  complicated  by  the residence  time  of  waste
product  in  the  recycle  silo.    This storage  volume  results  in powder  not
appearing for recycle until several  days  following  production  (assuming plug
flow in the silo).

PERFORMANCE MODEL

     Argonne has identified what  are believed to be the critical parameters
for S02 removal in a spray dryer.  These are:


         (Available Alkali]    = /gRv
     ^  V    Inlet S°2   /SDA "
     This  is  the  stoichiometric  ratio  (SR)  based  on the slurry  fed to the
spray  dryer absorber (SDA).


     , v  | Mass Available Alkali\    _ (•o-o)
         I Total Mass of Solids  )
         \                      / o JJA.

     This  recycle  ratio  (RR)  of  available  alkali  injected into  the  spray
dryer  to the  total mass  of  injected  solids  reflects  the  effect  of  spent
sorbent  recycle  on system performance.   The available  alkali  is determined
by a chemical method  (titration)  that  will  detect  the  alkali  whether  it is
on the surface  of  the  particle  or  in  the  interior.   If  recycling of  spent
sorbent  results  in  a  partial   "blinding"   of  available   alkali  by  other
compounds  in  the  recycled  powder,  then  the  stoichiometry   needed  for  a
particular  removal  will  be higher for a  system with recycle than for one in
which  the  operation is on a  "straight-through"  basis.  An  analysis of data
and  samples  gathered in this  program will  allow for  a  determination of the
significance  of  sorbent recycle on system operation.

      c)   [1 - exP(-t/t)]

      This  factor is a  reflection of the effects  of  boiler load (and hence
gas  residence time in the spray dryer)  as well  as  the drying  phenomenon for
particles  in  the spray dryer.   The variable, t,  represents the gas residence
time in  the  spray  dryer,  as  an average  of inlet and outlet gas volumes.  The
time constant,  T,  is  the  time for  drying  an average  sized particle in the
spray  dryer.    This  average  drying  time is  thought  to  be  a  function of
several  factors  including  the   percent  solids   in   the   slurry,   the  gas
temperature drop in  the spray dryer, and  the average droplet size.

           T        -  T          \   T
      , N  I  Gas in    Gas out   1   1
      d)  1 ^	—	
           Gas in    Gas Sat'd
                                     10-220

-------
     Many researchers have noted that the  closer  the  exit  gas  temperature is
to the  dew point,  or  saturation temperature,  then  the lower the  stoichio-
metry for  a  given removal.  If  the  S02 removal reaction  takes place  within
the  droplets  as  they  dry, then  the temperature  gradient  across the  spray
dryer should play  a key role.   The  numerator  of  the factor is the  tempera-
ture drop  of the  flue  gas across the spray  dryer due to  the  evaporation of
the  slurry.   The denominator is  the maximum temperature drop that  could be
taken in the spray  dryer before a wet  product  is produced.   Another way of
viewing this factor is  that it is the  ratio between the operation of  a  dry
scrubber and a wet scrubber, where the gas is quenched to its dew  point.

     Combining the terms;
     Fraction S02 removal = (SR)a (RR)   — I   (1 - exp (-t/T))

where a, b, and  c  are  exponents  which may,  or may not, be equal to  1.   Data
obtained in the  forthcoming performance  tests  on the  FGC  system will be  used
to check and calibrate this relationship.

                                 REFERENCES

1.   Federal Register.  4_5_(98) :33127-33128 (May 19, 1980).

2.   Farber,  P.S.,  Startup and  Performance  of  a High  Sulfur Dry  Scrubber
     System,  Paper   82-40.5,   Proc.   75th  Annual  Air  Pollution   Control
     Association Meeting, New Orleans (June 1982).

3.   Characteristics of  Waste Products From Dry  Scrubbing Systems,  prepared
     by Radian Corporation for the Electric Power Research Institute,  Report
     CS-2766 (Dec.  1982).

4.   Farber,  P.S.,  C.D.  Livengood,   and J.L.   Anderson,  Leachate  of   Dry
     Scrubber Wastes, Paper 83-29.1,  Proc. 76th  Annual  Air Pollution Control
     Association Meeting, Atlanta (June 1983).

5.   Bonk,   L.A.,  Argonne   National  Laboratory,  Unpublished   Information
     (1982).

6.   Knight, M., Argonne National Laboratory,  Personal  Communication  (Feb.
     1983).

7.   Immobilization  and  Leachability  of  Hazardous  Wastes,  Environmental
     Science and Technology 16(4):  219a-223a (1982).
                                    10-221

-------
                                    NOTES
1.   Company Names  and  Products.

    The mention  of  company  names  or  products  is  not to  be considered  an
    endorsement   or  recommendation  for   use   by  the  U.S.  Environmental
    Protection Agency.

2.   Units  of Measure.
    EPA policy is  to  express  all measurements in Agency documents  in  metric
    units.   When  implementing  this  practice will  result  in  undue cost  or
    difficulty in clarity, IERL-RTP provides conversion factors  for the  non-
    metric units.  Generally,  this  paper  uses  British units  of  measure.

    The  following  equivalents  can be  used  for conversion  to  the  Metric
    system:
    British

    5/9 (°F-32)
    1 Btu/kWh
    1 ft 0.3048 m
    1 ft2
    I ft3
    1 ft3/min
    1 gallon per minute (gpm)
    1 gallon (gal)
    1 grain
    1 horsepower (hp)
    1 in. W.C.
    1 Ib (avoir.)
    1 lb/106 Btu
    1 lb/ft3
    1 ton (long)
    1 ton (short)
Metric
1055.056 J/kWh

0.0929 m2
0.0283 m3
4.719E-04 m3/s
6.309 E-05 m3/s
0.003785 m3
0.0648 gram
746 watts
249.08 Pascals
0.4536 kg
429.6 ng/J
16.01 Kg/m3
1.0160 m tons
0.9072 m tons
                                     10-222

-------
UNPRESENTED PAPERS

-------
AN ECONOMIC EVALUATION OF LIMESTONE DOUBLE ALKALI
         FLUE GAS DESULFURIZATION SYSTEMS

G. A. Hollinden, C. D. Stephenson, J. G. Stensland

-------
                  AN ECONOMIC EVALUATION OF LIMESTONE DOUBLE
                    ALKALI FLUE  GAS DESULFURIZATION SYSTEMS
                       by:  Gerald A. Hollinden, Ph.D.
                            Tennessee Valley Authority
                            Chattanooga, Tennessee
                            C. David Stephenson
                            Tennessee Valley Authority
                            Muscle Shoals, Alabama
                            John G. Stensland
                            FMC Corporation
                            Schaumburg, Illinois
                                   ABSTRACT

    Considerable work was done at the EPA Scholz plant facility in defining
the process parameters for limestone double alkali flue gas desulfurization
systems.   In general this study work proved the viability of the process but
uncovered several less than optimum operating parameters that needed further
work.   FMC and others have continued to work with the process and have
defined operating parameter changes necessary to make the system
commercially viable.

    Limestone double alkali is especially appropriate for FGD systems
applied to boilers burning relatively high sulfur (2% and greater) fuel.  A
discussion of site-specific design criteria which impact on the selection of
FGD technology is included with a definition of the optimum parameters for
the applications of limestone double alkali.

    An in depth economic analysis of the system is included with comparisons
to conventional limestone scrubbing technology.  Cost comparisons are made
by subsystems such as absorber system, reagent handling, storage and
preparation system,  solids waste production and disposal.  The economic data
presented is primarily a result of study work done by TVA under contract
from the  Environmental Protection Agency.
                                    11-1

-------
                                  DISCLAIMER

    This paper was  prepared by the Tennessee Valley Authority (TVA).
Neither TVA nor any person acting on its behalf:

    a.    makes any  warranty or representation,  express or implied, with
         respect to the use of any information  contained in this paper;  or
         that the use of any information,  apparatus, method, or process
         disclosed  in this paper may not infringe privately owned rights; or

    b.    assumes any liabilities with respect to  the use of, or for damages
         resulting  from the use of,  any  information, apparatus,  method,  or
         process disclosed in this paper.

    This paper does not necessarily  reflect  the views and policies of TVA.
                                    11-2

-------
                                I.  INTRODUCTION

    For the past five years there have been extensive efforts to develop  a
sodium-based double alkali process in which limestone, rather than  lime,  is
used to regenerate the absorbent, thus retaining the advantages of  using  a
solution as the absorbent liquid while avoiding the higher operating costs
inherent with the use of lime.  FMC Corporation, Combustion Equipment
Associates, Inc. (now Thyssen - CEA Environmental Systems, Inc.—TESl),
A. D. Little, Inc., Accurex Corporation, and the U.S. Environmental
Protection Agency have been active in the development of a limestone double
alkali process.  These efforts have advanced the technology of the  process
to the point that a commercially acceptable limestone double alkali process
seems assured.

    Among the studies of limestone double alkali processes in the past few
years, there has been an EPA-sponsored prototype scale evaluation at the
Gulf Power Company's Scholz steam plant, an economic evaluation by the
Tennessee Valley Authority, and an economic evaluation by the Stearns-Roger
Engineering Corporation.  This paper summarizes these three evaluations and
discusses studies and test work by the FMC Corporation to develop a
limestone double alkali process.

                         II. TESI PLANT SCHOLZ WORK *

    Thyssen-CEA Environmental Systems, in conjunction with Arthur D. Little,
conducted an EPA-sponsored test program at Gulf Power Company's Scholz steam
plant near Sneads,  Florida.  This involved the conversion of a 20-megawatt
lime based dual alkali system to operation with limestone regeneration.

    Budgetary problems caused a reduction in the test program from six
months to two months.  However, even during the shortened program the
process proved to be technically feasible.

A.  SYSTEM DESIGN BASIS

    The limestone dual alkali system at Scholz was based on firing the
boiler with coal containing 2.9 to 3.4 weight percent sulfur.  These sulfur
levels corresponded to approximately 1900 to 2300 ppm S02 in the boiler
flue gas.  Since it was desirable to run tests at the high end of this inlet
loading and to include even higher S02 inlet concentrations,  an S02
injection system was provided.  This allowed operation with both a normal
and maximum load condition with regard to S02»  Normal system operation
was based on a gas  flow of 40,000 dscfm with an inlet S02 concentration of
2200 ppm dry.  The  maximum condition was based on a 45,000 dscfm gas flow
with an S02 concentration of 2650 ppm dry.  Table 1 and Table 2 give
additional details  for the normal and maximum design cases.
    Valencia, J.A.,  et.  al., Evaluation of the Limestone Dual Alkali
    Prototype System at Plant Scholz, Report to EPA, Contract 68-02-3128,
    August 1981.  (The majority of the data contained in the section is
    taken from this  report.)
                                   11-3

-------
             TABLE 1. DESIGN BASIS - NORMAL OPERATION
Inlet Gas:

    Flow Rate
    S02

Removal Efficiency:

    S02

Absorber/Scrubber Feed:

    Absorber Tray Feed
    Scrubber Recycle

Soda Ash and Limestone Feed:

    Soda Ash Feed Rate
    Limestone Purity
    Limestone Feed Rate

Waste Solids:

    Wash Ratio
    Insoluble Solids
40,000  scfm  (dry)
14.9  Ibs/min  (2,200  ppm dry)
95%
2.4 gallons/1000  acf
16 gallons/1000 acf
0.07 mole Na+/mole  S02
95 wt.%
0.986 mole available CaC03/mole   S02
4 displacement washes
55 wt.%
           TABLE  2.  DESIGN  BASIS  - MAXIMUM LOAD  OPERATION
Inlet Gas:

    Flow Rate
    S02

Removal Efficiency:

    S02

Absorber/Scrubber Feed:

    Absorber Tray Feed
    Scrubber Recycle

Soda Ash and Limestone Feed:

    Soda Ash Feed Rate
    Limestone Purity

    Limestone Feed Rate

Waste Solids:

    Wash Ratio
    Insoluble Solids
45,000 scfm (dry)
20.2 Ibs/min (2,650 ppm dry)
95%
2.6 gallons/1000 acf
15 gallons/1000 acf
0.081 mole Na+/mole   S02
90 wt.%

1.010 mole available  CaC03/mole  S02
4 displacement washes
55 wt.%
                                 11-4

-------
B.  DESCRIPTION OF THE SCHOLZ SYSTEM

    The test plant facilities at Scholz consisted of  three  separate  areas:
the boiler, the limestone dual alkali  scrubbing  system, and  the waste
disposal facilities, that were provided as part  of a  parallel EPRI sponsored
program to evaluate landfill disposal  of the waste cake generated.   The
waste disposal facilities will not be  discussed  in this paper.

1.  Boiler

    The dual alkali scrubbing system was installed on Unit  No. 1 at  Scholz
steam plant, a 40 MW capacity (47 MW peak capacity) B&W pulverized coal
fired boiler.  An electrostatic precipitator designed for 99.5% particulate
removal follows the boiler.  A portion of the flue gas, equivalent to a 20-
MW boiler load, was directed to the limestone dual alkali system.  The
remaining gas was exhausted to the main stack.

2.  Limestone Dual Alkali System

    During 1975 and 1976 the Southern  Company and EPA sponsored a jointly
funded test program, conducted by TESI, utilizing a lime dual alkali system
installed at Plant Scholz.  The limestone dual alkali system utilized during
the TESI testwork was a modification of that plant.   The modified dual
alkali system consisted of four basic  sections:  absorption; regeneration;
waste solids dewatering; and reagent storage and preparation.  A process
flow diagram is shown in Figure 1.

a.  Absorption—
    The absorption system consisted of a plumb-bob type venturi scrubber
followed by an absorption tower.  The  variable throat venturi was designed
for both particulate removal and S02 absorption.  Typically, a utility
type dual alkali system would take flue gas from a particulate removal
device (electrostatic precipitator or  fabric filter)  and would not require a
venturi scrubber.  However, rather than removing the  venturi, it was used to
quench and saturate the flue gases.

    The absorption tower was designed  to be used either as  a spray tower or
a tray tower.  The limestone dual alkali testing was  conducted utilizing the
tower with two trays.

b.  Regeneration Section—
    Spent recirculated scrubbing solution was bled from the  absorber to a
multi-stage reactor system.  The reactor system consisted of five separate
tanks—an existing 750 gallon primary  reactor followed by four new 3400
gallon tanks in series to allow a total of approximately 100 minutes
retention time.  All tanks were cylindrical, baffled  vessels with center
mounted agitators.  Tests were run with all five reactors in series  with dry
pulverized limestone fed to the first  reactor.
                                     11-5

-------
                                                    CLEANED GAS
      LIMESTONE
        SILO
       WATER
    NORMAL FLOW-
 ALTERNATE FLOW
Figure  1  Plant  Scholz  Limestone Dual Alkali  System  Flow Diagram

-------
c.   Waste Solids Dewatering—
    An existing 93,000 gallon thickener was utilized for solids settling.
Slurry from the reactor system was pumped to the tank with clear liquor
overflow collected in an existing thickener hold tank.  This tank provided
surge capacity for the regenerated return liquor feed to the scrubbing
system.  Slurry underflow was recirculated in a loop around the thickener to
prevent solids settling and plugging of these lines.  A bleed from this
recirculation loop was fed to the existing rotary drum vacuum filter.
Filtrate was returned to the thickener.  Filter cake was washed on the drum
using two wash water spray banks.  Solids were discharged to a weigh belt
conveyor for handling in the waste processing system.

d.   Reagent Storage and Preparation—
    Ground limestone (97-99.9% less than 325 mesh) was utilized as the
regenerative reagent.  It was received, stored and fed to the system in a
dry form even though the capability existed for feeding it as a slurry.
Soda ash solution was utlized to make up for sodium losses in the system.
It was normally added to the thickener hold tank.

C.  PROCESS PERFORMANCE

    As stated earlier, the original intent of the test program was to
evaluate several aspects of limestone dual alkali process performance.  As
the test program had to be reduced to only a two month period,  a thorough
and complete evaluation of the technology was not possible.  However,
process performance was evaluated for each of the following conditions:

    *    SO2 removal capability

    *    Reagent consumption

    *    Power usage

    *    Waste product properties

    *    Reliability and ease of operation

1.  SO? Removal

    During the test period, inlet S02 concentrations ranged from 1460 to
3240 ppm (dry).  Average removal efficiencies ranged from 93.5% in January
to 96.7% in March.

    Scrubbing liquor pH was the most important variable affecting S02
removal efficiency.  At pH's in excess of 5.7, outlet S02 concentrations
below 100 ppm were obtained.  When operating at a pH of 6.0, outlet
concentrations of less than 50 ppm were obtained.  It was determined  that
operating the absorber at a bleed pH of 5.7 to 6.0 provided the best
compromise between high S02 removal efficiencies and adequate regeneration
of the scrubbing liquor in the reactors.
                                     11-7

-------
    It was also determined that the ratio of bisulfite to  total  oxidizable
sulfur compounds (TOS) concentrations in the liquor could  be  utilized as a
guideline for S09 removal.  The ratio of bisulfite to TOS  during the  test
period typically ranged between 0.60 and 0.75, and this  ratio produced
efficiencies between 94% and 98%.

    Other variables such as pressure drop, inlet  S02, and  active sodium
concentration had a much smaller effect on performance than pH.   Their
effect appeared to be secondary in nature.

2.  Limestone Utilization

    Limestone utilization was very good during the test  period.   Utilization
in the reactor train effluent was between 85% and 95% of available  Ca(X>3
in the raw  limestone.  Final system utilizations  ranged  from  93% to 100%
averaging 97.5%.  Fredonia limestone from Kentucky was used during  the
start-up and initial break-in period and Sylacauga limestone  from Alabama
was used during the end of the break-in period and the entire testing
period.  Table 3 gives the characteristics of these limestones.   An attempt
was made to determine the progress of limestone utilization through the five
vessel reactor train.  Utilization in the first reactor  ranged from 23% with
a level of  104 ppm solids carryover from the thickener to  63% with  1910 ppm
carryover.  The test report hypothesized that this was possibly  due to the
solids entering the first reactor acting as crystals to  facilitate  an
increase in the rate of precipitation of calcium  sulfur  salts.   This,  it was
felt, would promote an increase in the rate at which calcium  was dissolved
and reacted with sodium bisulfite.  It was recommended that further studies
be done to verify this hypothesis.  Testing indicated that regardless of the
extent of reaction achieved in the first reactor, efficient limestone
consumption in the remainder of the train and dewatering system  would occur.

3.  Waste Cake Properties

a.  Settling Characteristics—
    The most significant process limitation encountered  at Scholz was the
generation of solids with good settling characteristics.   During periods of
the test program, solids with excellent settling  characteristics were
generated.  However, in other periods inconsistent and poor quality solids
also occured.   In the months of December and February, solids that  settled
out to 10% of the initial slurry volume in six to eight  minutes  were
generated.  In January poor solids were generated, which took hours to
settle to the same consistancy.  It was determined that  a  significant
contributor to poor solids quality was the carryover of  fine  solids in the
thickener overflow.  These solids, which were ultimately fed  to  the reactor
after passing through the absorber, promoted the  formation of finer and more
difficult to settle solids in the reactor train.  Laboratory  work seemed to
confirm this hypothesis indicating that poor settling behavior was  due in
large part to a high percentage of fine needle shaped solids  present  during
operation with high solids carryover from the thickener.   It  was determined
that solids carryover of 1000 ppm in the thickener overflow could easily be
handled by the system without detrimental effects to the settling
properties.  However, carryover in excess of 5000 ppm resulted  in rapid
deterioration of solids quality.

-------
                        TABLE 3.  LIMESTONE PROPERTIES
                                       Fredonia
Mine Location

Type

Particle Size Distribution
Wet Sieve Analysis (wt.%)

    60 mesh (250 )
    100 mesh (149 )
    325 mesh (44 )
    400 mesh (38 )

Bulk Density (lbs/ft3)

Chemical Analysis
    Alkalinity (wt.% as
    Ca (wt.%) as CaC03>
    Mg (wt.% as MgC03)
    Fe (wt.% as Fe203>
                            Sy lacauga
Fredonia, Kentucky     Sylacauga, Alabama

High-Calcium Calcite   High-Calcium Calcite
        99.7
        99.4
        96.4
        94.4
       55-60
        97.9
        93.6
         4.2
         0.05
 99.9
 99.8
 97.7
 93.9

60-65
 97.8
 95.7
  1.3
  0.04
    It was further determined that waste solids deteriorated in their
settling properties with time.  On various occasions during filter downtime,
solids in the thickener began resuspending as underflow was continuously
recycled to the feed well.  It was determined that it was possible that
agglomerated, quick settling solids had begun to dissolve and recrystalize
into poor settling solids.

b. Filter Cake Characteristics—
    Limitations in equipment size and poor mechanical performance of the
filter contributed significantly to two major shortcomings in filtering
operations.  These were the inability to produce filter cake with 55% or
higher insoluble solids and the inability to wash the cake to reduce sodium
losses to 4% of the insoluble solids content of the cake.

    Leaks in the internal piping of the filter drum caused substantial
losses of vacuum.  This, coupled with the inability of the thickener
underflow pumps to handle slurries containing more than 15% solids and  the
attendant dilution requirement, resulted in the filter cake having insoluble
solids contents ranging between 35% and 45%.

    High sodium losses in the cake were encountered throughout the test
program.  The system was designed with two spray banks to wash the cake
utilizing up to 40 gpm of wash water.  This represented the equivalent  of
displacing four times the final volume of liquid in a 55% solids  filter
cake.  Corresponding sodium losses were anticipated to be approximately 4%
of the insoluble solids content of the cake (equivalent to a Na/Ca ratio  of
0.08 in the final cake).  However, physical limitations in the handling of
the filtrate, limitations in the wash water supply, and operating require-
                                     11-9

-------
ments of the waste disposal system precluded the use of more than  two
displacement washes.  In February the average number of displacement washes
was 1.5 resulting in a high Na/Ca ratio of 0.2, and in March the average
number of displacement washes was only 0.8 with a correspondingly  higher
Na/Ca ratio of 0.4.  However, there were times when up to  three displacement
washes were utilized resulting in sodium losses of as low  as 0.03  to 0.04.

    It was concluded that the problems of low insoluble solids content  cake
and high sodium losses in the cake were not inherent to the  limestone  dual
alkali technology but were rather caused by mechanical conditions  of the
equipment used and operating requirements specific to the  test facility.

4.  Soda Ash Consumption

    Actual sodium losses from the system at Plant Scholz were much higher
than  those anticipated.  This was due not only to excessive  loss in the
filter cake as described above but also to leaks, spills and liquor purges
needed to maintain volume balances.  During the test period  soda ash feed
amounted to 0.29 moles of Na2C03/mole of SC>2 removed.  As  much as  half
of  this amount was needed for cake losses with the remainder associated with
other liquor  losses.  It was felt that sodium consumption  could be held
within design limits of 0.04 moles of Na2CC>3/mole of SC>2 removed on a
production system with little difficulty.

5.  Power Consumption

    Most of the power consumption at Scholz was associated with the system
forced draft fan.  The venturi required pressure drop 2-3  times that of the
absorber.  Thus, in a typical limestone dual alkali scrubbing system where
particulate removal is performed by either an electrostatic  precipitator or
a  fabric filter, the expected power consumption would be much lower than
that  encountered at Scholz.  Power consumption ranged from 2.5% (0.53  MW) at
flue  gas flow rates equivalent to a 21 MW boiler load, to  5.3% (0.42 MW) at
an  equivalent load of 8 MW.  If the power requirement associated with  the
venturi is not included, a requirement of 1 to 1-1/2% of power generated at
full  boiler load was anticipated.

6.  Process Operability

    The system at Scholz was reported to have good process operability after
the first two to three days of stable operation.  The major  problems with
process operability were encountered during the initial days following any
restart of the system.  This was principally due to the inconsistent
settling characteristics of solids generated.  In some cases the solids
settling rate would level off and stable operation would be  achieved.   At
other times settling rates would continue to deteriorate resulting in
significant solids carryover in the thickener overflow which further
exacerbated the problem.  It was hypothesized that the restart problems were
caused by redissolution of solids that had been left in the  thickener  at the
time  of the outage and subsequently recrystalized into fine  crystals with
poorer settling characteristics.
                                    11-10

-------
    Once the system reached stable operaton, the process operability was
considered to be very good.  Variations of inlet SC>2 concentrations of as
much as 500 ppm were handled easily by adjustment of the feed forward rate
of regenerated solution to the absorber.  Variations in boiler load were
handled in the same manner.  In addition, upsets including carryover of fly
ash in the flue gas due to precipitator malfunction, overfeeding of
limestone due to operator oversite, and occasional  limestone and soda ash
feed outages were handled well by the system.

    Some solids deposition was encountered in various portions of the
regeneration system.  Some scale buildup was encountered on the walls of the
reactor and in the overflow pipe connecting the first and second reactors.
Scale buildup in the other reactors was minimal, and it was concluded that a
semi-annual cleaning operation as part of a regular maintenance program
might be adequate to control any reactor scaling.

    Conclusions reached as a result of the Scholz plant testing are as
follows:

    "    A limestone double alkali process is technically feasible.

    "    Additional refinement and further testing  of the system was
         necessary to develop process information required for commercial
         operation.

    *    Laboratory or small pilot plant tests should be conducted to better
         understand the generation of solids with good settling
         characteristic s.

                               Ill FMC TESTWORK

    Since 1977 FMC has pursued the use of limestone for regeneration in its
double alkali process.  They felt that the performance claims made for the
process had been validated by the actual operation of three utility double
alkali systems furnished by three different suppliers.  The operating
benefits of the double alkali technology have been demonstrated but overall
evaluated costs continued to be a concern.

    In 1981 FMC determined that the growing differential in the price of
lime and limestone caused many evaluators to judge  lime double alkali as
excessively expensive when compared with conventional limestone scrubbing
systems.  This caused a renewed emphasis on developmental work for a
limestone regenerated double alkali process.  FMC's goal for a limestone
double alkali process extended beyond technical feasibility.  It was felt
that any system developed must realize the cost benefit of using limestone
while maintaining the advantages associated with lime double alkali
scrubbing, and without increasing other cost components significantly.  A
program was developed to test limestone double alkali, with emphasis on
three general operating parameters:  reaction rates, solids quality, and
absorber performance.  Beginning in September 1981, bench scale experimental
work was initiated on limestone kinetics at FMC's Central Engineering
Laboratories in San Jose, California, and at Illinois Institute of
                                     11-11

-------
Technology Research Institute.  In early 1982 the next phase of the pilot
plant work was started at their Princeton Research Center in Princeton,  New
Jersey.  The intent of the program was to define system operating ranges
that allowed high utilization of limestone and commercially feasible
reaction times while maintaining good solids quality and high  levels of
SO2 removal.

    FMC determined that the primary technical issues concerned the overall
regeneration reaction.  Reaction rate characteristics and the  parameters
that affect it were well understood from previous work done by FMC and
others.  It was determined that reaction rate characteristics  put
constraints on the process in the sense that high utilization  of limestone
and commercially feasible reaction times could only be accomplished with
certain combinations of conditions.

    It was  also found that solids quality is a function of the size and
shape  of individual solid particles.  As was seen at the Scholz test plant,
these  parameters were in turn a function of the chemical and physical
conditions  under which the particles were formed.

    FMC also noted that optimum absorption of sulfur dioxide placed certain
constraints on the composition of the absorbing solution, but  that these
conditions  are not necessariliy compatible with optimum results in the
regeneration reaction and the solids quality.  A plan was developed to
optimize the various competing factors.

A.  REACTION RATES

    The rate of limestone reaction and the degree to which it  proceeds to
completion  were found to be dependent upon several parameters.  First, the
dissolution of limestone creates carbonate and bicarbonate ions which place
an  upper limit on the pH in the regeneration reaction.  It is  not possible
to  regenerate all of the sodium bisulfite without using excessive amounts of
limestone.  This general constraint of limited capacity for bisulfite
regeneration lead to the definition of several variables that  were used  for
defining the feasible ranges of process chemistry.

    The first of these is the regeneration fraction (shown as  "Y" in Figure
2).  In general it was found to be difficult to achieve limestone
utilizations in excess of 90% at regeneration fractions in excess of 0.6.
That is, if more than 60% of the bisulfite was regenerated limestone
consumption became excessive.

    Another parameter defined was the ratio of the active sodium
concentration to the change in bisulfite concentration in the  regeneration
portion of  the process (shown as "Z" in Figure 2).  For a given absolute
amount of regeneration, the ratio is an expression of the active sodium
level  in the system.  A high ratio corresponds to a very concentrated system
and is^a practical limitation based on the total dissolved solids in  the
scrubbing solution.  A low ratio corresponds to relatively dilute solutions
which  lack  the buffering capacity FMC normally attempts to achieve in their
lime double alkali process.  It was determined that scrubber pH,
                                    11-12

-------
regeneration  pH,  change  in  bisulfite  concentration,  fractional  regeneration,
and  ratio of  active  sodium  concentration  to  the  change  in  bisulfite
concentration were all interrelated such  that  classifying  any three  of  these
parameters fixed  the other  two.   These  constraints,  plus additional  ones
discussed below,  set the  optimal  operating range defined by  FMC for  their
limestone double  alkali  process.

     Several other factors which also  affect  the  regeneration reaction rate
and  limestone utilization were identified.   These factors  and their  effects
were as  follows:

1.   Limestone Particle Size

     It was found  that, as with direct limestone  scrubbing, limestone
particle size is  directly related to  utilization.  Larger  particles  tended
to become coated  with the calcium sulfite reaction product rendering a
portion  of the  limestone unavailable  for  reaction.   Grinding the  limestone
to minus 200 mesh was determined  to be  essential for good  utilization,  and
the  relative  small incremental costs  for  grinding to minus 325  mesh was
deemed generally  cost justified.

2.   Limestone Reactivity

     It was found  that limestone varied  considerably  in  its microstructure
with a wide variety  of both amorphous and crystalline characteristics.
Essentially,  reactivity  is  a surface  area phenomenon with  a  limestone having
a high surface  to mass ratio being the  most  reactive.   Therefore,  it was
determined that reactivity  of the limestone  was  not  necessarily defined by
the mesh size.

     The principal effect of limestone reactivity was determined to be on the
regeneration  reaction residence time  requirements.

B.   SOLIDS QUALITY

    FMC  found that solids quality is  a  function  of particle  size  and shape,
which are in turn functions  of the chemical  and  physical conditions  under
which the particles  are generated.  As  a  practical matter, the  chemical
conditions are dictated  largely by the  kinetic and material  balance
constraints.  Poor solids quality were  generally the result  of  excessive
nucleation, or formation of new particles of reaction products.   In  this
sense, the goal of a fast regeneration  reaction  rate is in direct conflict
with a goal of crystal growth.  A reasonable compromise in this regard  was
accomplished with the proper design strategy in  the  reactor  area.  First,
multiple reactors in series  were  utilized with split limestone  feed  so  that
the degree of reaction was  controlled.  Also,  since  the first reaction  step
is the most likely point to  experience  excessive nucleation, some solids
from the thickener underflow were  recycled to  the first reactor,  and the
scrubber bleed pH kept relatively high.   The former  furnished seed crystals
which promote particle growth instead of  nucleation  and the  latter minimizes
the driving force which also favors crystal  growth.
                                     11-13

-------
C.  ABSORBER PERFORMANCE

    FMC found that the overall system material balance put constraints on
the system chemistry.  First, sulfur dioxide absorption began to decline
dramatically for absorber pH's below 6.0.  Very high collection efficiency
typically requires a pH at the top of the absorber of 6.3 or higher.  In
contrast to FMC's lime double alkali system, however, a low pH scrubber
bleed stream was necessary in order to achieve good limestone utilization.
This meant that the regenerated scrubbing solution had to be fed to the top
of the absorber rather than blended with the overall recirculating scrubbing
liquor as is done in their lime based process.  The result of this
constraint was that a spray tower or packed tower proved to be the most
effective absorber for a limestone double alkali process.  The packed tower
was deemed the most effective device because it can operate at very low
liquid to gas ratios and the regenerated solution flow rate is typically
equal to the total absorber flow rate requirement thus eliminating all
recirculation requirements to the absorber.  It was determined that a spray
tower could also be utilized with regenerated solution used directly in the
top stage with lower stage(s) supplied with recirculated liquor from an
absorber sump.

    The other important material balance constraint was found to be the
practical limitation on the absolute regeneration level.  The total flow
rate required to regeneration is inversely proportional to the changes in
bisulfite concentration achieved.  For example, if the bisulfite change is
0.1 M, the flow rate required to regeneration is twice that required for a
typical lime based system using a bisulfite concentration change of 0.2 M.
This means that the limestone double alkali process must operate at higher
active sodium concentrations in order to have regeneration flow rates equal
to lime double alkali.

D.  FMC PROCESS DESIGN

    From their test and process design work, FMC developed an operating
envelope that allows simultaneous achievement of all of the desired
objectives.  This envelope is summarized in Figure 2.  The boundaries of the
operating envelope are sumarized as follows:

1.  Scrubber bleed pH equal to 5.8 - 6.6.  It was found that if the absorber
bleed pH was too low, the reaction rate would improve; but this was offset
by an increased tendency to dissolve magnesium and a requirement for
two-loop absorption.  If the bleed pH was too high, only limited bisulfite
regeneration per pass was achieved and a higher active sodium level was
required.

2.  Regeneration return pH equal to 6.3 - 7.0.  It was determined that
regeneration return pH must be high enough to allow good S02 absorption
and adequate bisulfite regeneration per pass but not so high as to
significantly inhibit limestone utilization.
                                    11-14

-------
I
Q.

I
a

z
DC
D
h-
LLI
DC
     •R
CC
LU


1  °R
LU
DC
                                                     Y = BISULFITE REGEN
                                                        FRACTION
                                                     Z  RELATIVE SYSTEM
                                                        CONCENTRATION
                         D
                           B
                             BLEED  pH (pH)
                                               B

Figure 2  Limestone Double Alkali  Design  Envelope
                                11-15

-------
3.   Regeneration fraction (the percentage of the bisulfite in the scrubbing
bleed solution that is regenerated in the regeneration section):  FMC feels
that this fraction should be in the range of 35 to 70% and preferably
between 35 and 60%.  Higher regeneration fractions led to poor  limestone
utilization,  and lower regeneration fractions required excessive flow rates
to regeneration.

E.   DEMONSTRATION PLANT

    FMC is currently in the process of installing a limestone double alkali
demonstration plant at Northern Indiana Public Service Company  Schahfer
Station.  The plant will be a 3 MWe (9,000 acfm) system installed on a slip
stream  from the existing Schahfer Unit 17 lime double alkali FGD system.
The Schahfer Unit is designed to burn coal to 3.6% sulfur.  The objective of
the demonstration plant program is to confirm the limestone double alkali
test results that have been obtained in the past by FMC and at  the Scholz
Plant.

    The program schedule calls for the completion of erection by early
January 1984.  Testing is divided into three parts:  1) shakedown testing,
2) baseline testing, and 3) stress testing.

    Shakedown testing will involve the characterization of the
responsiveness of the system.  The results of this period will  determine
those variables which most influence system operation and the appropriate
values  to be selected for each of these variables to attain optimal system
operation.  This data will be utilized as the basis for baseline testing
work.   The aim of the baseline test period is to demonstrate the ability of
the limestone double alkali system to sustain performance over  long periods
of  time while responding to normal changes in boiler load and flue gas
conditions.  Stress testing will involve establishing conditions outside the
normal  expected to determine the response of the system to those conditions
and to  determine corrective actions necessary to compensate for any adverse
effects.

    FMC plans to have their demonstration plant program essentially complete
by  the  end of 1984.

                             IV ECONOMIC ANALYSIS

    TVA recently completed an economic analysis of the limestone double
alkali  system for the EPA.  In addition, Steams-Roger has just finished a
study for EPRI which also reviews limestone double alkali economics.  The
design  bases and the economic procedures used in the studies differ
considerably and detailed point-by-point comparisons are difficult.  The
major conditions and criteria used in each study are shown in Tables 15 and
16.  In addition, the TVA limestone double alkali process design is based on
information provided by EPA derived primarily from the tests at the Scholz
steam plant described in this paper and on in-house information.  The
Steams-Roger design is based largely on the process developed  by FMC.  In
general, both studies show that the limestone double alkali process is less
expensive in capital investment and operating costs than those  of a forced-
                                     11-16

-------
oxidation limestone FGD process under the conditions used  in the evaluation.
These results are discussed below.  In addition, an ecomonic comparision  of
the TVA design and the FMC design, both based on TVA premises is discussed.

A.  TVA/EPA STUDY*

    The base case for this study utilized a new single 500 megawatt
pulverized coal fired boiler firing 3.5% sulfur eastern bituminous coal.
The heat rate utilized for the boiler was 9,500 Btu/kWh.  The unit was
assumed to be installed during the construction of the power plant and
utilized a thirty year life with full load operation for 5,500 hours a year.

    The system design included four 125 megawatt operating absorber trains.
A spare absorber train was included to allow utilization of an emergency
bypass.  Hardware spares consisted of spare crushing and grinding equipment
in the limestone preparation area, a spare filter, and spare process pumps
in addition to the spare absorber train.

    It was assumed that construction would begin in early 1981 and cover a
three-year span to late 1983.  Capital investment requirements were based on
1982 costs, and annual revenue requirements were based on mid-1984 costs.

    In the TVA study the economics of the limestone double alkali FGD system
were compared with the economics for a forced oxidation limestone scrubbing
process.  The major absorber process design conditions utilized in the study
are shown in Table 4.

Capital Investment

    The total capital investment for the limestone double alkali process was
$95 million ($190/kW).  The comparable forced oxidized limestone system had
capital investment requirements of $103 million ($206/kW).  The total direct
investment, representing installed equipment costs, was $47 million for
limestone double alkali and $57 million for the limestone process.  A
breakdown of the capital investment requirements for the two systems are
shown in Table 5.

    As can be seen, the largest cost differential between the systems is in
the SC>2 absorption area.  Here the double alkali process has a cost
advantage of $9 million.  While the absorbers in both processes are similar
in size, the higher L/G ratio required for the limestone system resulted  in
a large cost penalty.  The double alkali system required an L/G ratio of  3
with an L/G ratio of 106 required by the limestone process.  This resulted
in a recirculation pump cost comparison of $1.7 million for limestone versus
$80,000 for the double alkali process.  In the areas of raw material
handling and feed preparation, the systems showed similar cost requirements
and the regeneration area required for the limestone double alkali process
did not materially add to the process cost.
    Stephenson, C. D., Burnett, T. A., and Torstrick, R. L., Economic
    Evaluation of a Sodium/Limestone Double Alkali FGD Process, to be
    published.
                                    11-17

-------
                     TABLE 4.  PROCESS DESIGN  CONDITIONS

                                                         Process
                                          Sodium/Lime stone         Limestone
                                           Double Alkali           Scrubbing

Absorber type                             Sieve tray tower        Spray Tower

Superficial Gas Velocity,  ft/sec                   9                   10

L/G, gal/kaft3
  Presaturator/Underspray                         2/1                 4/10
  Absorber                                         5                  106

Stoichiometry, mole Ca/mole (S02 + 2HC1)           1.0                  1.4
  Absorbed

Sulfite Oxidation, %                              10                   95

Thickener Feed Solids, %                           1-4                  8

Thickener Underflow Solids, %                     25                   40

Filter Cake Solids, %                             55                   85

    The dewatering system required for limestone double alkali was evaluated
at  50% higher  than that required for the limestone process.  This was based
on  the high-sulfite waste produced in the double alkali system.  However,
when the costs of forced oxidation were included in the limestone process,
the total cost of preparing waste for disposal were similar for both
systems.  When a  credit was applied to the limestone double alkali process
for flyash disposal,  the landfill disposal costs for that system were about
10% less.

    The other  difference in capital investment between the two processes
consisted of indirect investment and other capital investment, both of which
are based for  the most part on direct capital investment.  Exceptions to
limestone double  alkali's generally lower investment cost were the
contingency  fee and start-up and modification allowance assigned by TVA due
to  the  lower level of technical development of the process.

Annual  Revenue Requirements

    Tables 6 and  7 depict the annual revenue requirements assigned to the
double  alkali  and limestone processes respectively in the TVA  study.  The
limestone double  alkali process showed a first year annual revenue
requirement  of $26 million (9.28 mills/kWh) compared with $29 million (10.58
mills/kWh) for the conventional limestone process.  Annual direct costs for
the double alkali process are shown as $9.2 million (constituting 36% of  the
total first year  annual revenue requirement) and for the limestone process
are shown at $10.6 million (also 36% of the total annual revenue
requirement s) .
                                    11-18

-------
                         TABLE 5.  CAPITAL INVESTMENT
Direct Investment

Materials Handling
Feed Preparation
Gas handling
SC>2 Absorption
Reheat
Regeneration
Oxidation
Solids Separation
Fixation
    Total Process Capital

Services, Utilities and Miscellaneous
    Total Direct Investment Excluding Landfill

Landfill Construction
Landfill Equipment
Landfill Credit (fly ash disposal)
    Total Direct Investment

Indirect Investment

Engineering Design and Supervision
Architect and Engineering Contractor
Construction Expense
Contractor Fees
Contingency
Waste Disposal Indirect Investment
    Total Fixed Investment

Other Capital Investment

Allowance For Startup and Modification
Interest During Construction
Royalties
Land
Working Capital
    Total Capital Investment

Dollars per kW of Generation Capacity
   Capital Investment, $
Double Alkali      Limestone
      2,426
      4,506
     10,800
     11,348
      3,630
      1,506
          0
       ,493
        906
     40,615

      2,437
     43,052

      5,247
      1,454
     (2,312)
     47,441
      3,014
        861
      6,888
      2,153
     11,194
      1,720
     73,271
      6,716
     11,430
        406
        554
      2,775
     95,152

     190.30
  2,528
  4,715
 11,281
 20,288
  3,634
      0
  2,670
  3,679
	0
 48,795

  2,928
 51,723

  3,781
  1,123
	0
 56,627
  3,621
  1,034
  8,276
  2,586
  6,724
  1.707
 80,575
  5,917
 12,570
      0
    614
  3,392
103,068

 206.14
Basis:  500-MW new coal-fired power unit; 3.5% in coal; 89% SC>2 removal;
        fixation and landfill solids disposal; north-central, new power unit
        with 30-yr life at 5,500 hr/yr full-load operation; 1 spare absorber
        train, and provisions for emergency bypass of 50% of the scrubbed
        gas; reheat to 175°F as required.  Mid-1982 cost basis.
                                    11-19

-------
           TABLE 6.  SODIUM/LIMESTONE DOUBLE  ALKALI  PROCESS ANNUAL REVENUE REQUIREMENTS
                                               Annual
                                              Quantity
                                              107,700 tons
                                                6,190 tons
                                               10,580 tons
Direct Costs - First Year

Raw Materials
   Limestone
   Soda Ash
   Lime

      Total Raw Materials Cost

Conversion Costs
   Operating Labor
      and Supervision
      FGD
      Solids Disposal
   Utilities
      S team
      Process Water
      Electricity
   Diesel Fuel
      Landfill Fly  Ash Credit
   Maintenance
      Labor  and Material
Analysis

      Total  Conversion Costs

      Total  Direct  Costs

 Indirect  Costs - First Year

Overheads
   Plant  and  Administrative
       (60%  of conversion  costs
       less  utilities)
      Total First-Year Operating and Maintenance Costs

Levelized Capital Charges
   (14.7% of total capital
   investment)

      Total First-Year Annual Revenue Requirements

Levelized First-Year Operating and
   Maintenance  Costs (1.886
   First-Year 0 and M)
Levelized Capital Charges (14.7%
   of Total Capital Investment)

      Levelized Annual Revenue Requirements
                                                                   Unit
                                                                  Cost,
  8.50/ton
160.00/ton
 75.00/ton
First-Year Annual Revenue Requirements
Levelized Annual Revenue Requirements
                                                      25.52
                                                      35.74
 Mills/kWh

      9,28
     13.00
                 Total Annual
                   Cost, k$
915
990
794
                                                                                        2,699
43,860 man-hr
33,280 man-hr
528,500 klb
262,100 kgal
29,618,000 kWh
123,200 gal


4,990 man-hr


15. 00/man-hr
21. 00/man-hr
2.50/klb
0, 14/kgal
0. 037/kWh
1. 60/gal


21, 00/man-hr


658
699
1,321
37
1,096
197
(312)
2,715
105
6,516
9,215
                                                                                        13,987

                                                                                        25,521



                                                                                        21,753

                                                                                        13,987

                                                                                        35,740
Basis:  One-year,  5,500-hr operation of system as described in capital  investment  table.   Mid-1984
        cost basis.
                                             11-20

-------
                 TABLE 7.  LIMESTONE SCRUBBING PROCESS ANNUAL REVENUE REQUIREMENTS
                                               Annual
                                              Quantity
                                              142,900  tons
Direct Costs - First Year

Raw Materials
   Limestone

      Total Raw Materials Cost

Conversion Costs
   Operating Labor
   and Supervision
      FGD
      Waste Disposal
   Utilities
      Steam
      Process Water
      Electricity
      Diesel Fuel
   Maintenance
      Labor and Material
Analysis

      Total Conversion Costs

      Total Direct Costs

Indirect Costs - First Year

Overheads
   Plant and Administrative
      (60% of conversion costs
      less utilities)
      Total First-Year Operating and Maintenance Costs

Levelized Capital Charges
   (14.7% of total capital investment)

      Total First-Year Annual Revenue Requirements

Levelized First-Year Operating
   and Maintenance Costs (1.886
   first-year 0 and M)
Levelized Capital Charges (14.7%
   of total capital investment

      Levelized Annual Revenue Requirements
                Unit
               Cost,
First-Year Annual Revenue Requirements
Levelized Annual Revenue Requirements
29 .10
41,46
                  .50/ton
Mills/kWh

    10-58
    15-08
             Total Annual
               Cost, k$
                  1,214
                                                                                       1,214
43,860 man-hr
29,120 man-hr
542,200 klb
193,400 kgal
57,657,000 kWh
103,200 gal

4,990 man-hr


15.00/man-hr
21.00/raan-hr
2 .50/klb
0 .14/kgal
0 ,037/kWh
1.60/gal

21 .00/man-hr


658
611
1,356
27
2,133
165
4,285
105
9,340
10,554
                                  3,395

                                 13,949


                                 15,151

                                 29,100
Basis:  One-year,  5,500-hr operation of  system as described  in capital  investment  table.  Mid-1984
        cost basis.
                                            11-21

-------
    Levelized annual revenue requirements are $36 million  (13.00 mills/kWh)
and $41 million (15.08 mills/kWh) for the double alkali and limestone
processes respectively.  TVA utilized a projected value of money at  10%  and
an inflation rate of 6% to determine these figures.

    A comparison of the first year annual revenue requirements was made  and
the major components are shown in Table 8.  The major differences in direct
costs between the systems are in the area of raw materials, electricity,  and
maintenance requirements.  The major direct costs for the  limestone  double
alkali process in order of significance are maintenance, raw materials,
labor, steam, and electricity.  For the conventional limestone process they
are maintenance, electricity, labor, steam, and raw materials.

         TABLE  8.   COMPARISON OF  ANNUAL  REVENUE  REQUIREMENT COMPONENTS

                                             Annual Revenue Requirements, k$
                                            Sodium/Limestone       Limestone
                                             Double Alkali         Scrubbing

Raw Materials
  Limestone                                        915               1,214
  Soda Ash                                         990
  Lime                                             794             	—

Total Labor                                      2,699               1,214

Utilities
  Steam                                          1,321               1,356
  Electricity                                    1S096               2,133
  Other                                            234                192

Maintenance                                      2,715               4,285
Landfill Credit                                   (312)

Direct Costs                                     9,215             10,554

Overheads                                        2 319               3 395

Operating and Maintenance Costs                 11,534             13 949

Levelized Capital Charges                       13,987             15 151

Total First-Year Annual Revenue Requirements    25,523             29,100

    In general, the TVA study concluded  that the limestone double alkali
process has direct costs about 13% lower than those of a conventional
limestone process primarily because of lower maintenance and power costs.
Along with lower overheads which are based on direct costs—and lower
capital charges for the double alkali process,  the first year annual revenue
requirements of the double alkali process are 12% lower than those of the
limestone slurry process.
                                   11-22

-------
B.  STEARNS-ROGER/EPRI STUDY REVIEW*

    Steams-Roger conducted a study under contract to EPRI similar  to  the
TVA/EPA study discussed previously.  The Stearns-Roger study also compared a
limestone dual alkali system with a conventional  limestone scrubbing
system.  Different design and economic criteria were utilized for the
Stearns-Roger and TVA comparisons but similar conclusions were reached.

    The Stearns-Roger study was based on two 500 megawatt PC fired  boilers
firing 4.0% sulfur coal.  Table 9 outlines the specific process design
criteria utilized in the cost evaluation.  Estimates of raw material and
utility consumptions that were utilized in the study are shown in Table 10.

    Stearns-Roger also conducted a relatively subjective technical
evaluation of the advantages and disadvantages of the limestone double
alkali process compared with conventional limestone scrubbing.  This
comparison is shown in Table 11.  Credit was given to the limestone double
alkali process for significant advantages in maintenance as a result of
reduced scaling potential, reagent reactivity, and the ability to follow
changes in flue gas composition and condition.

    Tables 12, 13, and 14 review the base case economic evaluation  of the
limestone double alkali process done in the Stearns-Roger/EPRI report.
Table 12 shows the total system capital cost including the particulate
removal system.  Once the particulate removal system was subtracted from the
total, the capital cost of the limestone double alkali FGD system was
$162/kW.  When compared with Stearns-Roger evaluation of conventional
limestone scrubbing systems, the double alkali process requires 5%  less
initial investment.  This is true even though a higher process contingency
is included in the double alkali cost due to its lack of commercial
development.

    Table 14 depicts the levelized busbar cost for the limestone double
alkali system.  These cost are calculated by the "present worth" method (as
defined in EPRI Economic Premises) and assume a thirty-year plant operating
life.  This levelized bus bar cost is the sum of the fixed and varible
operating costs (Table 13) plus the cost of capital over the thirty-year
plant life.  The fixed operating costs shown for limestone double alkali are
approximately 33% lower than those Stearns-Roger calculated for conventional
limestone systems.  This is principly due to the elimination of the high
costs high maintenance of slurry recirculation pumps and the attendent lower
overall maintenance requirement.  Stearns-Roger also found that the total
varible and capital operating costs are approximately equal for both
systems.

    The costs developed by Stearns-Roger in the same study for a limestone
forced-oxidation process are shown in Tables 15, 16, and 17.
*   Reisdorf, J. B.,  Keeth, R. J., Scheck, R. W., Miranda, J. E., Economics
of FGD Systems, to be published 11/83.
                                   11-23

-------
                    TABLE 9.  LIMESTONE DUAL ALKALI SYSTEM
                      PROCESS  SPECIFIC  DESIGN  CRITERIA*
Flue Gas Handling Area Criteria

    Flue Gas Flow Rate (105% design load)
    Pressure Drop (flange to flange)
    ESP Fly Ash Removal Efficiency
    Spray Tower Outlet Gas Temperature

SO? Removal Area Criteria

    S02 Removal, High Sulfur Coal
    Scrubber Design
    Scrubber Modules

    L/G Ratio
           Makeup Ratio
    Absorber Tower Superficial Velocity
    Reaction Mix Tank Retention Time
    Oxidation Rate of Sulfite to Sulfate
    Solution pH
    Absorber Pressure Drop

Reagent Feed Area Criteria

    Total Soda Ash Storage
    Soda Ash 20% Solution Storage Tank
    Regenerant Limestone Storage
    Limestone Slurry Tank Storage
    Limestone Day Bin Storage
    Limestone Slurry Solids Concentration

Absorber Regeneration Area Criteria

    Limestone Reaction Tank Retention Time
    Limestone Reagent Feed Ratio
Waste Handling Area Criteria

    Thickener Underflow Percent Solids
    Dewatered Sludge Percent Solids
    Sulfite/Sulfate Mole Ratio
    Thickener Size
    Vacuum Filter Design Basis
    Water Washes for Sodium Recovery
2,015,000 acfm
9 in. H20
99.8%
127°F
90% (30-day rolling average)
Spray Tower, 3 Levels
4 @ 33-1/3% capacity ea.
  (3 opp, 1 spare)
20 gpm/1000 acfm
O.Q425 mole
mole S02 removed
10 fps
6 rain
18% available
5.8 - 6.2
6 in H20
30 days
30 hrs
60 days
12 hrs
30 hrs
25%
120 min (total for 4  tanks)
1.05 Ib mole
                                                 Ib mole S02 removed
20%
55%
9:1
22 ft2/tpd day solids
35 Ib/hr dry solids/ft3
3
*For one 500 MW unit
                                     11-24

-------
                    TABLE  10.   LIMESTONE  DUAL ALKALI SYSTEM
                     RAW MATERIAL AND UTILITY CONSUMPTION
                             FOR TWO  500  MW UNITS*

    Item                                                   Quantity

Soda Ash @ 0.0425 Stoichiometric Ratio                       2.5 t/hr

Limestone @ 1.05 Stoichiometric Ratio                       61.6 t/hr

Fixative Lime @ 3% of Dry Sludge and Fly Ash                 4.7 t/hr

Raw Water                                                1,250 gpm

Steam @ 572°F and 418 psig                             221,000 Ib/hr

Power (Operating Horsepower and Equivalent kW)

    Area 10 - Reagent Feed System                        2,040 HP (1,520 kW)

    Area 20 - S02 Removal System                         2,695 HP (2,010 kW)

    Area 30 - Flue Gas System                           11,010 HP (8,215 kW)

    Area 40 - Regeneration System                          150 HP (110 kW)

    Area 60 - Waste Handling System                      2,880 HP (2,150 kW)

    Area 70 - General Support Area                         400 HP (310 kW)

    Area 80 - Particulate Removal System                 5,360 HP (4,000 kW)

    Total                                               24,540 HP (18,315 kW)

Fly Ash (direct disposal to landfill fixated
    with waste sludge)                                      63.2 t/hr



*0peratig at 100 percent load
                                    11-25

-------
                       TABLE  11.   TECHNICAL EVALUATION
                         ADVANTAGES AND DISADVANTAGES
                       OF  THE DUAL ALKALI PROCESS WITH
                            LIMESTONE  REGENERATION*
                         (Compared to  the Conventional
                              Limestone Process)
    Item

1.  Process
    a.  Complexity of Operation
    b.  Oxidation of Sulfite to Sulfate
    c.  Turndown Ratio
0)
to
4-J
C
to


CO
• H
o
60
O
4-J
CO
CO
• r-l
Q
 0)

CO


X
X
                                                                          8P
                                                                             a)
                                                                             SP
                                                                             M
                                                                             C
                                                                             2
                                                                             4-1
                                                                             CO
    d.  Load Following Capability
    e.  Adaptability to Flue Gas Temperature Changes
    f.  Surge Requirements

    g.  Capability of Using Cooling Tower Slowdown
    h.  Stability of Process
    i.  Extreme Vessel Pressure
                                                                     X
                                                                     X

                                                                     X
                                                                     X
                                                                     X
2.
    j.  High Equipment Operating Temperature
    k.  Use of Liquid Fuel or Natural Gas
    1.  Material Handling Characteristics
    m.  Separation/Removal (two phase)

    Operation and Maintenance Requirements
    a.  Labor Requirements
    b.  Equipment Pluggage, Scaling
    c.  Equipment Corrosion
                                                                     X
                                                                     X
                                                                     X
                                                                     X
                                                                         X
                                                                         X
                                                                         X
    d.  Equipment Erosion
    e.  Reagent Reactivity/Makeup Rate

3.  Effect on Net Plant Heat Rate
    a.  Power Consumption - Low Absorption
          Tower Pressure Drop
    b.  Steam Usage

4.  Disposal
    a.  Land Requirements
    b.  Reactivity of Waste
    c.  Waste Handling Characteristics

5.  Use of Exotic Materials of Construction

6.  Operational Hazard
    a.  High Temperature
    b.  High Pressure
    c.  Use of Hazardous Chemicals
                                                                     X
                                                                     X
                                                                     X
                                                                     X
                                                                     X
                                                                         X
                                                                         X
                                                                         X
                                     11-26

-------
                   TABLE  12.   DUAL ALKALI  (LIMESTONE)  SYSTEM
                          TOTAL CAPITAL REQUIREMENT*


Area      Description                                                $ / kW

20        S02 Removal System                                         41.0
40        Regeneration System                                         4.0
30        Flue Gas Handling System                                   24.0
10        Reagent Feed System                                        16.0
60        Waste Handling System                                      22.0
80        Particulate Removal System                                 37.0
70        General                                                     3.0
          Total Process Capital                                     147.0

General Facilities                                                   15.0
Engineering and Home Office Fees                                     18.0
Project Contingency                                                  30.0
Process Contingency                                                   8.0
          Total Plant Cost                                          218.0
Allowance for Funds During Construction (AFDC)                        8.0
          Total Plant Investment                                    226.0

Royalty Allowance                                                     0.7
Preproduction Costs                                                   7.2
Inventory Capital                                                     1.4
Initial Catalyst and Chemicals                                        0.0
          Total Capital Requirement                                 235.0
          Less Capital for Particulate Control System                73.0
          Total Capital Requirement for FGD System                  162.0


^December 1982 dollars, two-500 MW units.
                                     11-27

-------
                  TABLE  13.  DUAL  ALKALI  (LIMESTONE)  SYSTEM
                                OPERATING  COSTS
Fixed Operating Costs

Operating Labor
Maintenance Labor
Maintenance Material
Administration and Support Labor
          Total Fixed Operating Costs

Variable Operating Costs

Lime
Limestone
Soda Ash
Raw Water
High Pressure Steam
Power
Sludge (unlined)
          Total Variable Operating Costs
          Credits for By-products
          Total Variable Operating Costs
            with By-product Credit
                                                1st Year
                                                  $/kW*
 1.3
 1.9
 2.8
 0.9
 6.9
 2.0
 0.3
 4.8
 5.7
 5.5
24.9
 0.0

24.9
              Levelized
              mills/kWh
 0.5
 0.8
 1.1
 0.4
 2.8
 0.7
 2.0
 0.8
 0.1
 2.0
 2.5
 2.2
10.3
 0.0

10.3
*December 1982 dollars
                                   11-28

-------
                   TABLE  14.   DUAL ALKALI  (LIMESTONE)  SYSTEM
                            LEVELIZED BUSBAR COST*
                                                        Mills/kWh

Process Capital                                            3.9
General Facilities                                         0.4
Engineering and Home Office Fees                           0.5
Project Contingency                                        0.8
Process Contingency                                        0.2
          Total Plant Cost                                 5.8
AFDC                                                       0.2
          Total Plant Investment                           6.0

Royalty Allowance                                          0.02
Preproduction Costs                                        0.19
Inventory Capital                                          0.04
          Total Capital Requirement                        6.3

Fixed Operating Cost                                       2.8
Variable Operating Cost                                   10.3
          Total Levelized Busbar Cost                     19.4
          Less By-product Credit                           0.0
          Total Levelized Busbar Costs
            with By-product Credit                        19.4
          Reference Case Levelized Busbar Cost             3.8
          Levelized Busbar Cost Apportioned to FGD        15.6
*Two-500 MW units.
                                    11-29

-------
                      TABLE 15.  FORCED OXIDATION SYSTEM
                          TOTAL CAPITAL REQUIREMENT*


Area      Description                                                 $/kW

20        S02 Removal System                                          67.0
30        Flue Gas Handling System                                    24.0
10        Reagent Feed System                                         14.0
60        Waste Handling System                                       10.0
80        Particulate Removal System                                  37.0
70        General                                                      7.0
          Total Process Capital                                      159.0

General Facilities                                                    16.0
Engineering and Home Office Fees                                      20.0
Project Contingency                                                   34.0
Process Contingency                                                    3.0
          Total Plant Cost                                           232.0
Allowance for Funds During Construction (AFDC)                         8.0
          Total Plant Investment                                     240.0

Royalty Allowance                                                      0.8
Preproduction Costs                                                    7.4
Inventory Capital                                                      0.9
Initial Catalyst and Chemicals                                         0.0
          Total Capital Requirement                                  250.0
          Less Capital for Particulate Removal System                 73.0
          Total Capital Requirement for FGD System                   177.0


*December 1982 dollars, two-500 MW units.
                                    11-30

-------
                   TABLE  16.   DUAL  ALKALI  (LIMESTONE)  SYSTEM
                                OPERATING  COSTS
Fixed Operating Costs

Operating Labor
Maintenance Labor
Maintenance Material
Administration and Support Labor
          Total Fixed Operating Costs

Variable Operating Costs

Limestone
High Pressure Steam
Power
Dry Solids (unlined)
Gypsum
          Total Variable Operating Costs
          Credits for By-products
          Total Variable Operating Costs
            with By-product Credit
                                                1st Year
                                                  $/kW
 1.0
 3.0
 4.5
 1.2
 9.7
 5.4
 4.8
 7.8
 1.7
 1.8
21.5
 0.0

21.5
             Levelized
             mills/kWh
0.4
1.2
1.8
0.5
3.9
2.2
2.1
3.3
0.7
0.7
9.0
0.0

9.0
*December 1982 dollars.
                                    11-31

-------
                      TABLE 17.  FORCED OXIDATION SYSTEM
                            LEVELIZED BUSBAR COST*
Process Capital
General Facilities
Engineering and Home Office Fees
Project Contingency
Process Contingency
          Total Plant Cost
AFDC
          Total Plant Investment
Mills/kWh

   4.3
   0.4
   0.6
   0.9
   0.1
   6.3
   0.2
   6.5
Royalty Allowance
Preproduction Costs
Inventory Capital
          Total Capital Requirement

Fixed Operating Cost
Variable Operating Cost
          Total Levelized Busbar Cost
          Less By-product Credit
          Total Levelized Busbar Costs
            with By-product Credit
          Reference Case Levelized Busbar Cost
          Levelized Busbar Cost Apportioned to FGD
   0.02
   0.20
   0.02
   6.7

   3.9
   9.0
  19.6
   0.0

  19.6
   3.8
  15.8
*Two-500 MW units.
                                   11-32

-------
     Stearns-Roger  concluded  from  their  study  that  the  levelized  bus  bar  cost
apportioned  to  the  limestone double  alkali  process  is  approximately  9%  less
than that of a  conventional  limestone process.

Discussion of the Economic Evaluations

    As the foregoing  summary of the  TVA and Stearns-Roger  economic
evaluations  shows,  the  evaluations are  similar  in  general  approach and
procedure but differ  appreciably  in  both the  design and  economic  assumptions
and  specific methodology.  The differences  occur in three  categories:  the
power plant  size and  operating conditions,  the  FGD  system  design  and
operating conditions, and the economic  criteria and costing  procedures.
These differences preclude comparisons  of the two  evaluations  in  terms of
absolute values or  specific  details.  Both  evaluations,  however,  produce the
same general conclusions and illustrate the same economic  relationships of
the  limestone double-alkali  process  and the limestone  forced-oxidation
process.  This  is best  shown by the  direct  capital  investment  and operating
cost summaries  shown  in the  preceding discussion,  selected values of which
are  summarized  in Table 18.   The  absolute values of the  TVA  and
Stearns-Roger results are not comparable because of the  differences
discussed, but  the  differences between  the  costs for the two processes in
each evaluation can be  compared.
                    TABLE  18.  MAJOR DIRECT COSTS FROM THE
                  TVA AND  STEARNS-ROGER ECONOMIC EVALUATIONS
                             TVA Evaluation
Capital Investment

SO2 Removal
Flue Gas Handling
Reagent Feed
Waste Handling
Regeneration
General

Operating Costs

Labor
Maintenance
Limestone
Soda Ash
Lime
Steam
Electricity
                         Limestone
                          Double
                          Alkali
             Limestone
              Forced
             Oxidation
        $/kW
   23
   29
   14
   22
    3
    5
 46
 30
 14
 17
  0
  6
Levelized, mills/kWh
  0.9
  1.8
  0.6
  0.
  0.
  0.9
  0.7
0.9
2.9
0.8
0
0
0.9
1.4
            Stearns-Roger Evaluation
            Limestone      Limestone
             Double         Forced
             Alkali        Oxidation
                      $/kW
 41
 24
 16
 22
  4
  3
 67
 24
 14
 10
  0
  7
             Levelized, mills/kWh
0.5
1.9
2.0
0.8
0.7
2.0
2.5
0.4
3.0
2.2
0
0
2.1
3.3
                                   11-33

-------
    In direct capital investment, the limestone double alkali process  has  a
much lower cost for SC>2 absorption in both evaluations, while the
additional costs for absorbent regeneration are relatively small.  The
difference in absorber costs is almost wholly responsible for the  lower
capital investment of the limestone double alkali process.   (The Stearns-
Roger evaluation also has a lower waste disposal cost for the limestone
forced-oxidation process that reduces the cost difference beween the two
processes; this, however, is due to the use of stacking of gypsum  disposal,
while the TVA design has a conventional landfill.)

    In operating costs, both evaluations also shown that the maintenance
costs and electricity costs are appreciably lower for the limestone double
alkali process, both the result of the lower absorber costs  and pumping
requirements.  These lower costs more than offset the additional costs for
soda ash and lime.

    Since the size of the absorbers is largely determined by the flue  gas
volume which is the same for both applications, the lower SC>2 absorption
costs for the limestone double alkali process are largely a  result of  the
smaller and less complicated absorbent recirculating system  and lower
pumping costs associated with the lower L/G ratio.  The double alkali
process is inherently flexible in the L/G ratio because the highly reactive
absorbent solution—which also has a low propensity for scale formation—
allows a wide choice of absorber designs.  It is possible that the absorber
design could range from a spray tower to a design similar to a packed  tower,
with corresponding high to low L/G ratios, thus allowing an  optimization of
absorber design and L/G ratio.

    The different absorber design philosophy is seen in the designs used in
the TVA and Stearns-Roger economic evaluations.  The TVA design is based on
information provided by EPA, in part from the limestone double alkali  test
at the Scholz steam plant.  A sieve tray absorber is used and the  L/G  ratio
is 8 gal/acf, 5 gal/acf of which is regenerated absorbent from the
regeneration system.  The Stearns-Roger design is based on information from
FMC Corporation.  A spray tower absorber is used and the L/G ratio is  20
gal/acf, all of which is unregenerated absorbent from the absorber hold
tank.

    The effects of these differences are impossible to quantify by comparing
the two economic evaluations because of the differences in the design  and
economic premises.  To compare the economics of the two designs, TVA
incorporated the FMC design and operating conditions (which  includes a
higher limestone stoichiometry, a longer regeneration system hold  time, and
a lower oxidation rate, as well as, spray tower absorbers and a higher L/G
ratio) into the TVA process and repeated the economic evaluation.  All other
conditions and economic procedures and criteria remained the same.  The
major conditions are shown in Table 19;  other conditions do not differ from
those summarized in the preceding discussion.  The capital investments and
annual revenue requirements are shown in Tables 20 and 21.  The economics  of
the limestone forced-oxidation process are included for comparison.
                                   11-34

-------
    The overall effect of  the changes  is  small;  the  total  capital  investment
is increased 4% and the annual revenue requirements  are  increased  4%,  as
compared to the original design.  The  use of a  spray tower absorber with  a
higher recirculation rate  increased  the capital  investment for  SC>2
absorption by 14%, but electricity costs  were decreased  about 3% because  of
the lower flue gas pressure drop  in  the system.   The only  other capital
investment components affected to a  significant  extent were  the regeneration
area costs, which increased 13% because of the  longer hold time, and the
fixation costs, which increased 10%  because of  the higher  stoichiometry.
The only annual revenue requirement  components  affected  to a significant
extent in addition to the  electricity  costs were the limestone  costs,  which
increased 4% and the maintenance  costs, which increased  slightly because  of
the higher absorber costs.

    Based on this comparison, the effects of the different design  criteria
on the economics of the process are  minimal.  The use of an  absorber such as
a sieve tray that allows a lower  L/G ratio reduces the capital  cost of the
SC>2 absorption area to some degree but it has little effect  on  electricity
costs because of the higher fan costs  associated with the  higher flue  gas
pressure drop.  The changes in limestone  stoichiometry;  regeneration area
hold time, and the oxidation rate have little effect on  the  economics  of  the
process.

                                  CONCLUSIONS

    As a result of the work done  by  TESI  at Plant Scholz and by FMC, it can
be concluded that a limestone double alkali process  is technically
feasible.  The economic evaluations  by TVA and  Stearns-Roger indicate  that
the cost of owning and operating  a limestone double  alkali system  is less
than those of conventional limestone scrubbing.   On  this basis, it appears
that consideration of the  technology is warranted by those faced with  the
installation of flue gas desulfurization  systems, especially on plants
burning relatively high sulfur fuels.
       TABLE 19.   MAJOR OPERATING CONDITIONS FOR THE ECONOMIC COMPARISON
                    OF SIEVE TRAY AND SPRAY TOWER ABSORBERS
Absorber Type
System Pressure Drop, in. H20
Superficial Velocity, ft/sec
Unregenerated L/G, gal/acf
Regenerated L/G, gal/acf
Lb Na2C03/lb S02 absorbed
Mole CaC03/mole S02 absorbed
Oxidation, %
Regeneration Hold Time, Min
Limestone
 Double
 Alkali

Sieve Tray
   10
    9
    3
    5
    0.01
    1.01
   10
  100
                                                    Limestone
                                                     Double
                                                     Alkali
               Forced
              Oxidation
              Limestone
Spray Tower   Spray Tower
  10
  20
   0
   0.01
     05
  1
  5
120
 10
110
  0
  0
  1.4
 95
                                    11-35

-------
                              TABLE  20.  CAPITAL INVESTMENT COMPARISON FOR SIEVE AND SPRAY TOWER ABSORBERS

                                                                              Capital Investment, k$
i
UJ
ON
 Direct Investment

 Materials Handling
 Feed Preparation
 Gas Handling
 S02 Absorption
 Reheat
 Regeneration
 Oxidation
 Solids Separation
 Fixation
         Total Process Capital

 Services,  Utilities,  and Miscellaneous
         Total Direct  Investment  Excluding  Landfill

 Landfill  Construction
 Landfill  Equipment
 Landfill  Credit  (Fly  Ash Disposal)
        Total Direct  Investment

 Indirect  Investment

        Total Fixed Investment

Other Capital Investment

        Total Capital Investment

Dollars per kW of generation capacity
Limestone
Double Alkali
Process
Sieve Tray
2,426
4,506
10,800
11,348
3,630
1,506
0
5,493
906
40,615
2,437
43,052
5,247
1,454
(2,312)
47,441
73,271
21,881

95,152
190
Spray Tower
2,458
4,565
10,615
12,979
3,630
1,700
0
5,533
1,000
42,480
2,548
45,028
5,247
1,454
(2,312)
49,417
76,352
22,747

99,099
198
Forced-Oxidation
Limestone
Process
2,528
4,715
11,281
20,288
3,634
0
2,670
3,679
0
48,795
2,928
51,723
3,781
1,123
0
56,627
80,575
22,493

103,068
206

-------
               TABLE 21.  ANNUAL REVENUE REQUIREMENT COMPARISON
                      FOR SIEVE AND SPRAY TOWER ABSORBERS

                                            Annual Revenue Requirements
Raw Materials
     Limestone
     Soda Ash
     Lime
Total Labor

Utilities
     Steam
     Electricity
     Other

Maintenance
Landfill Credit
Analysis
Direct Costs

Overheads

Operating and Maintenance Costs

Levelized Capital Charges

Total First-Year Annual Revenue
  Requirements

Levelized Operating and
  Maintenance Costs

Levelized Capital Charges

Levelized Annual Revenue
  Requirements

First-Year, mills/kWh
Levelized, mills/kWh
Limestone
Double Alkali
Sieve Tray
915
990
794
2,699
1,357
1,321
1,096
234
2,715
(312)
105
9,215
2,319
11,534
13,987
25,521
21,753
13,987
35,740
9.3
13.0
Spray Tower
951
990
794
2,735
1,357
1,321
1,067
236
2,834
(314)
105
9,341
2,391
11,732
14,578
26,310
22,126
14,578
36,704
9.6
13.3
Limestone
Scrubbing
1,214
1,214
1,269
1,356
2,133
192
4,285
105
10,554
3,395
13,949
15,151
29,100
26,309
15,151
41,460
10.6
15.1
                                     11-37

-------
DEVELOPMENTS AND EXPERIENCE IN FGD MIST ELIMINATOR
                     APPLICATION

              R. T. Egan, W. Ellison

-------
                         DEVELOPMENTS AND EXPERIENCE
                     IN FGD MIST ELIMINATOR APPLICATION

                                     By:

                             Richard T. Egan, PE
                           The Munters Corporation
                          Ft. Myers, Florida  33901

                             William Ellison, PE
                             Ellison Consultants
                          Monrovia, Maryland  21170
                                  ABSTRACT

     The purpose of this paper is to detail available mist-eliminator
technology and to assess U.S. practices and experience in utilization of
these sub-systems of flue gas desulfurization (FGD) processes.   Pertinent
trends in FGD system design and operation tied to the criticality of mist
eliminator performance are identified and discussed and advancements in
eliminator selection and application are reviewed.  Case histories of a
number of significant existing mist eliminator facilities are described and
the importance of selection of internals design and arrangement permitting
use of elevated mist-eliminator inlet-face gas velocity to enhance droplet
separation forces is emphasized.
                                INTRODUCTION

     The mist eliminator is a critical part of the wet flue gas desulfuriza-
tion system and should be selected and effectively designed to avoid in-
crustation and corrosion of downstream system components, as well as to help
limit the amount of carryover to the stack of suspended solids, dissolved
salts and liquid (1).
FGD OPERATIONAL AND MAINTENANCE PROBLEMS RELATED TO MIST ELIMINATOR DESIGN

     Results of the operation of flue gas desulfurization systems in utility
plants indicate that a mist eliminator facility may be a major operational
and maintenance problem, detracting from gas-cleaning system reliability.
Principal difficulties have included inefficient performance,
plugging/scaling, erosion and corrosion of mist eliminators, and deterioria-
tion of downstream components including stack liners.
                                   11-39

-------
Inadequate Eliminator Performance

     Instances of poor droplet collection are the result of system design
deficiencies.   Under ideal service conditions, such as with provision for
uniform gas flow distribution, effective means of keeping elements clean,
and adequate drainoff of collected liquid, eliminators perform well.  In
tests conducted with mist eliminator surfaces clean and gas-flow distri-
bution nearly uniform, extensive measurements of outlet carryover solids
loadings in flue-gas discharges from 10-20 MW demonstration FGD systems with
vertical-flow mist eliminators indicated that eliminator outlet solid
particulate concentration was consistently lower than 0.01 Ib/million Btu,
and thus very much lower than the current New Source Performance Standard of
0.03 Ib total solid particulate/million Btu.   In all cases, scubber outlet
solids concentrations were shown to be less than inlet concentrations, even
for inlet solids loadings as low as 0.004 grains/SCF.  Available data on
commercial systems strongly suggest that poor emission-control performance
by FGD mist eliminators is in some cases the direct result of fouling of the
elements that changes blade profile geometry causing deterioration of
efficiency.

Fouling/Corrosion

     Provisions for chemical scale control designed into the scrubber
slurry-recirculating system best achieve their purpose in the absorber
section where dispersal of liquid flow at high rates continuously irrigates
the exposed internal surfaces.  Because of its lack of such irrigation, and
since carryover liquid collected on and wetting the mist eliminator internal
surface will absorb residual SO  and 0  from the scrubbed gas, the mist
eliminator is the most likely component to be fouled.  Thus the surfaces of
the mist eliminator must be effectively washed to prevent gypsum scaling and
deposition of solids contained in sulfite/sulfate-laden droplets.

Impact on FGD Reliability

     Nationwide FGD reliability data indicate (2) that the scrubbed-gas
handling facilities downstream of the FGD absorbers are the principal
contributors to reduced unit reliability.  Critical corrosion and fouling in
this low-pH wet section have been found to be strongly influenced by the
extent of droplet-carryover flow at the exit of mist eliminators.

CRITICAL DESIGN ASPECTS

     As noted, scrubbing systems can operate without significant deposition
of solids on mist eliminator collecting surfaces.  Furthermore, if the
surfaces are clean and the system is properly designed, entrainment of FGD
liquor does not increase outlet solid particulate emission.

Gas Flow Distribution

     Uniformity of gas-flow .distribution is of major significance in elimi-
nator performance and, thus, use of design and flow-modeling techniques to
equalize gas velocity across the cross-sectional flow area is essential.
                                   11-40

-------
One report  (3) shows that local vertical gas-flow velocities entering the
first level of spray banks in a spray tower ranged from approximately 120%
to 20% of average velocity after the addition of integral gas-distribution
vaning, as compared to a range of 250% to minus 20% without vaning.  While
other adjustments were called for to achieve a more satisfactory distribu-
tion to prevent poor mist-eliminator operation and liquid entrainment, it
can be seen that vaned internal distribution means can be an essential first
step.

In-Situ Washing Means

     Although proper washing of eliminator internals is best done with fresh
water, the material-balance constraints of closed-loop recirculating
scrubbing-slurry systems require close limits on fresh-water input.  Thus,
recycled process liquid must be used in many existing scrubber systems,
particularly those of the vertical gas-flow mist-eliminator type, to achieve
the required mist-eliminator wash-spraying intensity.  By the mechanical
action of wash sprays upstream and downstream of the mist eliminator signif-
icant deposition of soft, sulfite-solids can be prevented.  To control
deposition of CaSO  from a super-saturated scrubber liquid the designer
relies on absorber pH control and the use of CaSO,-unsaturated mist-
eliminator wash liquid.

Prevention of Reentrainment by Controlled Drainage

     High carryover-liquid loading at the mist eliminator inlet and/or
localized, high gas velocity can lead to adverse reentrainment of collected
droplets when the design fails to provide adequate means of removing the
liquid-catch from the eliminator elements.  Through advancements in develop-
ment and application of mist eliminators, improved liquid drainage provi-
sions are available to the FGD designer.

APPLICABILITY OF ALTERNATIVE DESIGN CONFIGURATIONS

Vertical Gas Flow

     Vertical-flow eliminators may be conveniently fitted into the upper
section of the scrubber tower and have been preferred for this reason.  They
have generally used a chevron-type design with horizontal zig-zag baffling
of two to six passes.  Due to the tendency for liquid reentrainment the
chevrons may be positioned on a slant to obtain better drainage.

Horizontal Gas Flow

     The horizontal-flow mist eliminator design, most common in Japan and
Germany, has only recently been applied in utility design in the United
States.  Baffles in a horizontal-flow eliminator are generally oriented in a
vertical position.   In high gas-velocity designs J-shaped phase separation
chambers or disengaging columns extending from the surface of the elements
counter to the gas flow afford drainage paths external to the gas-flow
regime to avoid reentrainment of liquid by the gas stream.  This configura-
tion lends itself to use of higher design gas velocities with resulting
                                    11-41

-------
increased droplet-removal efficiency due to increased inertial separating
forces and better drainage.   It also permits segregation of the eliminator
liquid-catch from the recirculating absorber slurry stream, which is advan-
tageous in eliminator washing-system design, permitting use of a recirculat-
ing wash water circuit.

        AVAILABILITY OF  GENERIC MIST ELIMINATOR TYPES FOR FGD SERVICE

INERTIAL METHODS

     The most commonly applied FGD mist eliminators use gravitational (dyna-
mic) forces caused by change in gas direction to accomplish separation of
liquid droplets from scrubbed gas.  A vertical-flow arrangement has been
regularly used in utility service in U.S.A.  Horizontal-flow design, which
is more common in Japan  and  Germany, has only recently been introduced in
utility applications here.

Chevron/Baffle Type

     Vertical-flow eliminators have historically been of the rudimentary
chevron, zig-zag baffling type.  Per Figure 1, the sweep of the gas flow
             Figure 1.   Reentrainment in Chevron Mist Eliminators

tends to reentrain collected liquid.   An improvement in this design through
slanting of baffles (Figure 2)  provides a vertical directional component for
liquid flow along the length of the baffle to give better drainage of wash
water and collected mist downstream of a bulk entrainment separator of
horizontal orientation designed to collect most of the inlet liquid (3).
Inclining of mist eliminator baffles in this manner at 30 degrees from the
horizontal (4) increases inlet  liquid loading capability free of significant
reentrainment by at least 100%  at high vertical gas velocity.  A general
                                   11-42

-------
review (5) of commercial FGD design in the utility industry reports the
following U.S. practices in eliminator application:
                   Chevron Vinos
                                                      Washer Lance
                                  Bulk Entninment Separator  (BES)
                 Figure  2.   Slanted-Baffle Mist  Eliminator

        Principal use of mist eliminators of chevron multi-pass (continuous-
        vane) construction in a vertical-gas flow configuration is favored
        from strength and cost considerations.  Baffle-type units that
        substitute noncontinuous-slat construction are also commonly used
        and, like the chevron design, offer comparatively simple, open
        geometry with low gas-pressure drop.

        Baffle spacing is normally 1^ to 3 inches except in the second stage
        of two-stage designs which generally use 7/8  to 1 inch spacing.

        Plastic construction is most common due to reduced weight, cost and
        corrosion tendencies.

        Precollection and pre-washing stages are commonly used to improve
        mist eliminator operation.

        Eliminator wash systems typically operate intermittently and use a
        mixture of clear scrubbing liquid and fresh water makeup.
                                   11-43

-------
     A unique and highly efficient vertical-flow baffled arrangement  that
utilizes exceptionally high design velocity without reentrainment  is  illus-
trated in Figure 3.  This unit has advantageous 45 deg. slanting in two
                                        Gas out
                                                   Liquid and
                                                   gas  in
                                   Liquid drain area
                                   against flow
       Figure 3.   Mist Eliminator Drainage by Multi-Directional Slanting

 directions in that the total gas stream is deflected from its straight
 course by a series of zig-zag shaped channel walls.  Slanted chevron-shaped
 airfoils on the eliminator walls collect and drain liquid droplets removed
 from the gas stream.   Installations of this type operating in U.S. for over
 nine years have aptly field-demonstrated the high efficiency of this slant
 design.  In some instances, a bulk entrainment separator of 45 degree
                                    11-44

-------
 orientation (Figure 4)  is  placed  below (upstream  of)  the  zig-zag  channel-
 wall  eliminator.
Figure 4.  Slanted-Baffle Bulk Entrainment Separator

Vane Type

     Elements in the horizontal flow eliminators are typically of vaned
construction and oriented in a generally vertical position.  As noted
earlier,  phase separation chambers (6) extending counter to the gas flow
direction afford drainage paths external to the gas-flow regime for removal
of liquid to avoid reentrainment by the high-velocity gas stream.  (Figure
5).

PACKED TYPE

     Neither loose nor rigid types of packing have been used for mist
elimination in FGD service as they are generally prone to excessive fouling
and plugging when used in services in which the flue gas or recirculating
scrubbing medium have significant concentrations of suspended solids.  Based
on its efficient performance on small micron droplets and reliable service
in difficult industrial applications, a unique layered-mesh dry-bed (7) mist
                                    11-45

-------
Figure 5.  Function of Phase Separation Chamber at Elevated Gas Velocity

eliminator construction is now being considered for trial application in the
utility industry.   It is understood that the packing, an interlaced struc-
ture of thermoplastic monofilaments with essentially all of the filaments
oriented perpendicular to gas flow, may generally be cleaned in-situ by
intermittent spray washing.  In very fouling-prone industrial services the
layered-mesh beds have been found to conveniently lend themselves to perio-
dic mechanical cleaning when temporarily removed from the unit.  It is
contemplated that a utility FGD scrubber module will be backfitted with a
substitute triple-layer mesh-bed mist eliminator of this type to compare its
reliability and performance with that of the original common horizontal-
chevron type.

WET ELECTROSTATIC PRECIPITATION

     Wet electrostatic precipitators, extensively used in the chemical
process industries, offer a highly efficient and effective means of col-
lected carryover droplets from FGD systems and, when present in significant
concentrations, reducing emission of sub-micron sulfuric acid mist parti-
cles.  Engineering and costing studies indicate (8) that the wet electro-
static type offers savings in overall stack gas cleaning system design and
operation when also used simultaneously to control emissions of fine fly-ash
particulate matter in fly-ash scrubbing type FGD applications that use
neither dry fly-ash collectors nor venturi scrubbers.  A 1 mW pilot plant
test of this concept has been completed in 1983 on high-sulfur-coal utility
plant service, and a 50-100 mW size demonstration is now contemplated.  A
unique new 150 mW cogeneration-type powerplant that is to be located at the
Houston Ship Channel and fire petroleum coke has selected a wet-
electrostatic-precipitator type mist eliminator for its wet FGD system in
order to prevent an exceptionally high sulfuric acid mist emission tied to
very high sulfur content of the fuel.
                                   11-46

-------
    ADVANCED DESIGN TECHNIQUES FOR IMPROVED FGD RELIABILITY AND ECONOMICS

     The design of the mist elimination facility is critical to FGD system
dependability and to performance and economics of the utility plant itself.
Available FGD mist eliminator technology has advanced broadly since earliest
designs that used primitive structural angle irons and later chevrons in the
scrubber outlet to reduce the liquid escape.  Thus, while many U.S. system
designers are content to use the simple low-velocity, low-efficiency generic
eliminator configurations, advancements in inertial separation by empiri-
cally designed profiles continue to contribute significantly, worldwide, to
the performance and reliability of many installations.  Technologically
advanced dynamic mist eliminator designs of German origin, (Figure 6 and
Figure 3), utilized only sparingly in the U.S.A. to date, greatly reduce
carryover that can aggravate solid particulate emissions and downstream
corrosion and fouling on low-pH duct walls, flue-gas reheat devices and
Figure 6.  High-Performance Horizontal-Flow German Mist Eliminator Design

stack-liner walls.  As noted earlier, these eliminators provide unique
liquid drainage columns incorporated in aerodynamically designed eliminator
vanes or passages.  This permits use of elevated design superficial gas-
velocity so as to more effectively trap out small micron droplets without
reentraining the liquid that collects on the internal eliminator surfaces.

RELIABILITY EXPERIENCE AND CONCERN

     FGD system reliability experience continues to be generally fair to
poor and it appears, clearly, that greatest utilization of advancements in
mist eliminator technology will be of significant benefit in improving the
dependability of the entire wet scrubbing operation.
                                    11-47

-------
FGD Reliability in U.S.A.

     FGD reliability, the accumulative amount of time the  system was  oper-
ated during any calendar period divided by the duration of time periods
during which the system was called upon to operate, has averaged only 81% in
the U.S.A. during the period 1974-1982 for a large number  of FGD systems
tracked by EPA (9).  For the most recent years the average reliability has
been 83.3% (1980), 82.0% (1981), and 81.3% (1982).  No major improvement
trend is discernable and, thus, there is substantial room  and possible
motivation for upgrading mist eliminator design practice and performance to
yield significant benefits in increased system reliability.  Moreover, in
high-sulfur coal service, very low reliability, often less than 70%,  has
been experienced, due in part to the impact of carryover slurry contributing
to outages through downstream fouling and low-pH corrosion.  Selection of a
high-performance mist eliminator is particularly vital in  such service.

Mist Eliminator Reliability/Performance and Their Influence on FGD and Unit
  Reliability

     Reliability of the mist eliminator depends on its being maintained in
clean condition, which is directly tied to the nature of the absorber
emission and the chemistry of the recirculating scrubbing medium, the
quality of wash water used and the adequacy and timing of washing flows.  A
reliable eliminator installation that achieves high droplet removal effic-
iency will contribute to high FGD and unit reliability.  Conversely,  fouling
of the mist eliminator will ultimately require shutdown of the absorber
module, placing a spare module into operation if available.  Additionally,
FGD failure data by cause code collected from 1978 to the present by  the
National Electric Reliability Council (NERC)  indicates (9) that FGD unreli-
ability caused only a 2-3% forced outage rate of power-generating capacity.
Thus, since as noted earlier, FGD reliability is typically only 81%,  it is
apparent that bypassing of FGD during FGD-shutdown, which is possible  in
most existing installations, has been very common.  The NERC data also shows
that mist eliminators were deemed directly responsible for only 9% of  all
plant outages caused by FGD.  However,  other components that are not  spared,
but can be significantly impacted by unremoved slurry carryover droplets,
caused a substantial amount of the lost plant time.  For example of all
plant outages attributed to FGD failures,  28% was caused by dampers,  19% by
ductwork and 17% by fans.  Thus it appears that a combination of poor
reliability and poor performance in many mist eliminator installations may
be a major cause of FGD unreliability experienced in many FGD installations.
Moreover, as discussed later, through an understanding of of the adverse
chemistry and nature of emissions from U.S.  supersaturated-CaSO mode FGD
installations (e.g., natural-oxidation limestone slurry scrubbing)  and of
critical difficulties in achieving effective washing in this mode,  one can
readily and clearly discern why long-term U.S.  reliability experience  for
high-sulfur FGD installations of this mode/type has been so very inferior
(10)  (11) to that of unsaturated-mode installations,  (i.e., magnesia-
buffered calcium-based slurry scrubbing and sodium-based dual alkali  scrub-
bing.)
                                   11-48

-------
Field Investigative Program by Electric Utility Industry

     Due to widespread mist eliminator problems that have been reported to
it, and a feeling that FGD system suppliers and architect-engineers may have
been neglectful, EPRI has begun a field study program to increase under-
standing of how to size mist eliminators and avoid corrosion and plugging
(12).

Concerns of EPRI — Principal difficulties reported by EPRI that are to be
addressed are as follows:
     -  Half of the existing FGD systems are probably experiencing mist
        elimination system problems.

     -  Malfunction of mist elimination systems has resulted in plugging and
        corrosion, causing significant FGD absorber down-time.

     -  The impact on mist eliminator operations of gas distribution, wash
        water quality and water droplet size distribution needs to be
        quantified.

     -  The effects on wet stacks of mist-eliminator operational inade-
        quacies need to be quantified.

        There is a need for comprehensive information on mist eliminator
        systems so that design and operating guidelines can be established.

Scope of EPRI Work — The EPRI project plan includes:
        Execution of field tests and data collection at specific FGD sites,
        including flue gas measurements, sample collection/analysis and
        historical data collection.

        Evaluation of mist eliminator pressure drop and velocity profile,
        and analysis of the source and type of plugging or carryover.

        Recommendation of improved wash sequences, water sources, mainte-
        nance procedures and design modifications.

        Development of guidelines for testing, evaluation and troubleshoot-
        ing.

TRENDS IN ADVANCEMENT OF FGD SYSTEM DESIGN AND OPERATION

     Broader application of high-performance eliminators is being made in
conjunction with continuing advancements in FGD system design.  Increasing
use of spray tower type FGD absorbers has substantially raised mist elimina-
tor liquid loadings requiring improved performance and liquid handling
capacity.   Droplet removal for protection against downstream fouling becomes
particularly critical in systems that employ wet flue gas heat exchangers of
the direct and regenerative reheat type.  Current general practice in
selection and specifying of mist eliminators calls for limiting droplet
emission to no more than about 0.1 pounds liquid per million Btu boiler heat
                                   11-49

-------
input,  (approximately 0.05 grains/SCF, dry), thus limiting the solid emis-
sion contribution of carryover to about 0.01 pounds/MM Btu (0.005
grains/SCF, dry), which is a small fraction of the current 0.03 pounds/MM
Btu New Source Performance Standard for total solid particulate emissions
from coal-fired utility boilers.   Sustained performance at such high effic-
iency is critically tied to operating conditions in the FGD system, and thus
it is essential that the absorber vessel designer be cognizant of the design
criteria and performance limitations of the mist eliminator.  Moreover, the
successful FGD installation requires closest coordination between the
responsible system designer and the mist eliminator manufacturer.

Regenerative Flue Gas Reheat

     The very high plant lifetime cost of energy required for adequate
reheat of flue gas has led to increasing use of regenerative reheat exchang-
ers downstream of the mist eliminator.  Ljungstrom rotary heaters using hot
flue gas are being employed in Japan and West Germany while tubular heat
exchangers supplied with hot water are operating in U.S.A.  In all such
installations mist eliminator design has been a special concern, and in such
applications it is particularly essential that outlet droplet loading
specifications be met reliably.

Spray-Tower Type Absorbers

     Because of simplified-internals and good load-turndown capability spray
tower type absorbers have come into broad use in FGD service.  The rela-
tively uniform distribution of spray water across the cross-sectional
gas-flow area of vertical towers served by multiple levels of sprays gener-
ally enhances gas flow distribution at the mist eliminator inlet.  However,
the liquid loading is generally high, perhaps higher than 1 to 2 gpm per
square foot in some vertical tower designs, and efficiency potential and
liquid handling capacity of alternative mist eliminator types have become
significant considerations in establishing the design and staging of mist
eliminator facilities.

Mist Eliminator Inlet Loadings

     Inlet loadings are particularly high in limestone scrubbing systems
designed for high SO  removal efficiencies in high sulfur service with
high liquid to gas flow ratios.  With high inlet liquid loading eliminator
performance will generally require the inertial design to be based on use of
elevated superficial gas velocity in order to achieve high-efficiency liquid
removal.  Advanced vertical-gas flow mist eliminator designs are adequate
when meeting the outlet loading specification requires limit drop size no
lower than 35 microns.  Corresponding horizontal-gas-flow designs can be
employed to use gas velocity sufficient for limit drop size as low as 15 to
20 microns.  Moreover, generous provision is made to give adequate drainage
of the liquid catch so as to avoid reentrainment.
                                    11-50

-------
Influence of Gas Velocity on Mist Eliminator Performance

     The mist eliminator removes droplets by use of dynamically-induced
inertial forces that move the liquid out of the conveying gas flow channels.
Thus it becomes more efficient at increasing inlet face velocities.  This is
expressed quantitatively by the expression for the removal efficiency of an
inertial mist eliminator:
                              E=l-e-Z
     ...,,          , x   0.5
     Where z = cd   v
     and where:
        d  = mass median droplet size of the inlet distribution
        v = gas velocity
        c and x are empirically determined values.

Typical Design — The generic chevron arrangements that are extensively used
in FGD service (Figures 1 and 2) are typically designed for nominal inlet
face velocities of about 500 feet per minute yielding a removal efficiency
no higher than about 98.5% at d  =150 microns.  The high-performance
vertical flow designs (Figure 7J, use design gas velocities up to 1100 feet
per minute, and at the same value of d  achieve removal efficiencies
exceeding 99.8%.  The newest horizontal* flow designs (Figure 6),  use design
  Figure 1.  High-Performance Vertical-Flow German Mist Eliminator Design

velocities up to 1300 feet per minute, and comparable removal efficiencies
are as high as 99.99%.

Maximum Feasible Velocities — At a given inlet liquid loading, increasing
inlet face velocity requires increased capability for drainage of the liquid
catch in order to avoid significant reentrainment.  With chevrons reentrain-
ment takes place at 600 to 800 feet per minute velocity.  Because of high-
                                    11-51

-------
velocity pickup of liquid (rather than drainage limitations) the high-
performance, aerodynamically designed elements using drainage columns will
generally encounter reentrainment above 1200 to 1400 feet per minute in
vertical-gas-flow design and above 2000 feet per minute in horizontal-gas-
flow design.

Gas Flow Distribution

     Uniformity of gas flow distribution across the inlet face of the
eliminator is critical to overall droplet collection capability.  Case
histories are replete with instances of poor performance coupled with
mal-distribution, e.g., side-entry absorber towers with insufficient in-
ternal baffling primarily due to a lack of or inadequate gas flow modeling
at the equipment design state.  Inlet gas distribution is critical because
at the comparatively low gas pressure-drop, e.g. generally less than 0.5
inches W.G., at which many mist eliminators operate, the eliminator is,
itself, not an adequate gas-flow distribution means and cannot compensate
for extreme gas-velocity gradients.  Unlike other particulate collection
devices, its performance is adversely affected by both very reduced gas
velocity (e.g., less than 400 feet per minute) and very elevated velocity
(e.g., more than 1200 feet per minute in vertical-gas-flow designs.)  For
optimum design to ensure that the inherent capability of the mist eliminator
is utilized, the RMS (root mean square) deviation of the inlet velocity
profile should ideally be in the 20% to 30% (maximum) range.  Note that
every zone in the mist eliminator face area at which the local velocity
exceeds the level at which reentrainment starts provides a "window" through
which excessive liquid discharge is emitted.  Current stringent standards
for total solid particulate emissions make it necessary for the conceptual
design and geometry of the in-tandem absorber and mist eliminator to be
established in a coordinated manner so that acceptable system performance
may be realistically predicted and achieved.

Washing Technique

     The key element in ensuring high mist eliminator reliability and sus-
tained performance, and in thereby limiting impact on FGD reliability, is
maintenance of day to day cleanliness of eliminator internals.  Eliminator
cleanliness is critically influenced by the chemical analysis of the scrub-
bing slurry carried over to the mist eliminator.  Mist eliminator fouling
with calcium sulfate scale is most effectively prevented in unsaturated-
CaSO -mode scrubbing processes, i.e., those with magnesia buffering or
that use sodium liquor scrubbing in a concentrated-active-alkali dual alkali
system, which have dissolved calcium concentration in scrubbing liquor at
the level of 170-200 ppm or less.  Maintenance of an adequate level of
suspended solids in scrubbing slurry and good regulation of limestone feed
are important considerations in control of scrubbing slurry chemistry
affecting tendencies for eliminator scaling by calcium sulfate and calcium
sulfate, particularly in supersaturated-CaSO -mode limestone scrubbing
systems.  Of special concern in this connection are non-steady state, i.e.
transient, conditions such as during load changes and at startup and shut-
down that lead to elevation of scrubbing slurry pH above the set point.  Use
of the minimum pH in the absorber consistent with SO  removal requirements
                                    11-52

-------
will help ensure the desired high limestone utilization.  In-situ mist
eliminator washing, when carried out effectively, provides an adequate means
of removing deposits that would otherwise occur due to the presence of
collected scrubbing slurry at the surface of the eliminator internals.  This
is best achieved when a maximum amount of FGD system makeup water is allo-
cated to the washing activity.  Constraints on process liquid discharges and
on acceptable rates of FGD liquid effluent-outfall normally prevent generous
use of fresh water for in-situ mist eliminator washing to ensure good
surface cleanup and call for washing schemes tailored to site-specific
conditions..  Based on the authors' review of available details of existing
installations to be assessed in the pending EPRI investigative program noted
earlier, washing effectiveness (together with gas flow distribution) may
well be the key to troubleshooting success.  This is particularly so in
unsaturated-CaSO,-mode FGD systems, which lend themselves to effective
washing due to the less troublesome chemistry of the carryover droplets as
noted earlier and due to the potential for satisfactorily using low-calcium
CaSO,-unsaturated thickener overflow liquor as part of the wash water
supply.  Moreover, it is clear that an assessment of the chemistry and
nature of absorber emissions, unsaturated-CaSO -mode FGD vs. supersatu-
ratedmode, performed in conjunction with an evaluation of comparative ease
of mist-eliminator cleaning by in-situ washing, gives an unusually clear
insight as to the principal reason(s) for poor FGD reliability experienced
through 1982 in natural-oxidation supersaturated-mode high-sulfur FGD
installations throughout eastern U.S.A.

Placement of In-Situ Wash Water — Wash water is applied at the inlet face
but not the final exit face of the eliminator facility.  Systems incorporat-
ing two mist eliminator stages are preferentially washed on both upstream
and downstream faces of the first stage and upstream, only, on the second
stage.

Eliminator Washing Technique for Supersaturated-Mode Absorbers with Low
Limestone Utilization — Extensive experience has shown that absorber towers
with a high concentration of limestone in the scrubbing slurry, i.e., less
than 85% limestone utilization, should be provided with continuous mist
eliminator wash (13).  Vertical gas flow mist eliminators require 0.5 gpm
per square foot, horizontal flow 0.75 gpm per square foot.

Washing When Limestone Utilization is High — With limestone utilization 85%
or greater mist eliminators should be washed intermittently, preferably with
fresh or low total-dissolved-solids makeup water, 1.5 gpm per square foot
for vertical flow, 3.0 gpm per square foot for horizontal flow.

Recycling vs. Once-Through Washing Systems — In mist eliminator systems
that permit segregation of used washwater and the droplet catch from scrub-
bing slurry, such as the horizontal-flow type, recirculated wash systems may
be advantageously used.  Makeup water should be comparatively fresh water,
i.e. with total dissolved solids less than 1,000 ppm, and should ideally be
fed at a rate sufficient to limit recirculating-water dissolved calcium
level to 150 ppm and relative calcium sulfate saturation to 20%.  In using
once-through systems, particularly if employing recycled thickener overflow
in supersaturated-mode systems, treatment with soda ash is effective in
                                   11-53

-------
removing most of the calcium sulfate saturation in the water.  The tolerable
relative saturation in once-through wash liquid supply cannot be explicitly
forecast as it is dependent on the type and amount of total dissolved
solids.

Washing Spray Design — Mist eliminator washing systems normally employ
solid cone sprays spaced so as to provide overlapping spray patterns and
designed for nozzle pressure of 30 to 50 psig.  In systems where intermit-
tent spray cycles are specified, sequence timers should provide the capabil-
ity for "on" cycles of 5 to 10 minutes, "off" cycles of 20 to 60 minutes.
Typically the first stage should be washed 5 minutes on, 20 to 30 minutes
off, the second stage 5 minutes on, 60 minutes off.  The adjustable timers
should permit the cycles to be lengthened or shortened to suit operating
conditions.  Advisedly the wash cycle should be started at the typical rates
and then be adjusted only after the system operation has been stabilized
over several months.  This permits use of sufficient operating experience
needed to properly judge the effects of adjusting the cycle.  Each adjust-
ment should be followed by stabilized operation for several months duration
to afford an adequate observation period.

Droplet Emission Testing

     Measurement of must eliminator performance is of importance both for
the purpose of determining liquid loadings being admitted to flue gas reheat
systems or stacks and to quantify the contribution of liquid-carryover
solids, dissolved and suspended, to the total solid particulate emission
from the stack.  In addition, in many instances the determination of the
droplet size distribution is of equal importance.  However, the flue gas is
water  saturated in the region of the absorber train in which the eliminator
is positioned.  Thus, significant heat losses to the ambient cause water
condensation in the FGD system as well as in any gas sample trains used to
withdraw and analyze gas streams.  This can complicate field testing to
characterize the liquid loading in gas streams entering and leaving mist
eliminators and requires selected techniques to accomplish such measurements
in an  accurate manner.

Candidate Test Methods — For measurement of liquid loading  (14) without
need to determine droplet size distribution the tester may possibly use the
throtting calorimeter, heated probe, double calorimeter, condensation method
or cyclone separator.  When both loading and distribution measurements are
required the impactor method, light scattering probes, or laser probes may
possibly be used.

Criteria for Choice of Test Method — The testing objectives that are most
critical in selecting a method of carryover measurement are as follows:
     -  Drop size measurement discriminates among droplet diameters from 1
        micron to several hundred microns.
     -  The method quantifies droplet concentration in the gas.
     -  Testing can be carried out in water saturated or unsaturated gas.
     -  Liquid loads in the range from a few milligrams per cubic meter to
        several thousand milligrams per cubic meter may be quantified.
     -  The measuring device must be readily portable at industrial sites.
                                  11-54

-------
        It must be possible in a simple way to sample and test gas across
        the cross-sectional flow area of large diameter vessels.

Preferred Test Methods — All criteria, above, for selecting mist testing
equipment cannot be met by any single method.  Experience has demonstrated
the advantageous use of two specific complementary methods/devices for
measurement of size distribution and loading employing compactors — an
impactor probe of special design and a cascade impactor.  The impactor
probe, applicable to the size range of 15 to 2,000 microns, is a plate
coated with a special film such as magnesium oxide that is inserted into the
flowing gas stream via a shield-tube enclosure fashioned with a movable
shutter to permit timed exposure of the plate surface.  A cascade impactor,
applicable to the size range of approximately 0.4 to 10 microns, comprises a
group of staged in-series nozzles behind each of which is a removable
thin-glass baffle (collecting) plate and through which nozzles an isokineti-
cally sampled gas stream is drawn by a vacuum pump.

Tracer Type Testing — In the special case of sodium liquor scrubbing
systems such as dual alkali FGD a dry-filtration type collection in a filter
and impingers of an isokinetically sampled gas stream (15) can be used to
accurately quantify process liquor carryover loading by analysis of sodium
content of the catch and relating it to the sodium ion concentration in the
FGD recirculating liquor.

Materials of Construction
     Mist eliminator users are increasingly emphasizing the need for capa-
bility to withstand upset or transient temperatures ranging as high as from
350  F. to 700  F.  This has brought about rapid development and use of
newer high-temperature plastics as well as high-alloy metals.

Materials with Modest Temperature Capability — The lower temperature
plastics include PVC (polyvinyl chloride), polypropylene and phenyele oxide
based material.  The latter contains amounts of polymers such as styrene
added to make it more extrudable or moldable and is made in various grades
identified by heat deflection temperatures as high as 300  F. and more.
However, when subject to stress and heat in actual service, the maximum
service temperature is considerably less than the heat deflection rating.

Materials for Elevated Temperatures — Particularly as a result of proximity
of heat sources such as flue-gas reheat means, positioned close to the mist
eliminator in some applications, high temperature plastic materials such as
FRF (fiberglass reinforced polyester) are increasingly being specified and
used for mist eliminator internals.  FRP can typically operate up to approx-
imately 350° F., and in some cases, with some charring, has tolerated
temperatures approaching 500  F.  The pultruded blade (profile) forms
typically contain a mixture of premium-grade isophthalic polyester resin,
alumina/silicate filler and approximately 5% antimony trioxide serving as  a
fire retardant.  The addition of a synthetic cloth surfacing veil provides
protection for the glass reinforcing fibers.
                                   11-55

-------
                              PRESENT PRACTICES

OUTLOOK FOR FUTURE FGD SYSTEMS APPLICATION

     The strong trend in U.S. since 1978 toward application of forced-
oxidation type calcium-based FGD systems in preference to previously common
natural-oxidation supersaturated-mode types may be expected to somewhat
reduce absorber-emission related impacts on mist-eliminator and FGD reli-
ability.  Nonetheless the forced-oxidation processes do also, to a degree,
operate in the supersaturated mode.  Those being applied in U.S.A. may be
expected to pose significant fouling problems for the eliminator step and
low-pH corrosion and fouling immediately downstream of the mist eliminator
due to many circumstances not encountered in the high-reliability
unsaturated-mode systems previously discussed.  These include:

     -  Presence of unreacted lime/limestone in carryover liquid causing
        fouling, particularly when the goal of high system limestone utili-
        zation is achieved by use of a precooling or other remote scrubbing
        step
        Comparatively high carryover rates due to the high absorber
        liquid/gas flow ratios required for SO  removal and/or gypsum
        scale control in the absorber causing fouling/corrosion
     -  Comparatively high outlet flue-gas SO  concentration due to inher-
        ently lower SO  removal efficiency causing corrosion
     -  Potentially higher outlet flue-gas HC1, HF, NO , and
        SO  /H SO  concentration (due to a possibly lower absorption
        mass-transfer capability) causing corrosion.

CONTINUED MISAPPLICATION OF  INERTIAL MIST ELIMINATOR TECHNOLOGY

     Notwithstanding the critical need for improved standards of eliminator
performance and reliability, current mist-eliminator specifications of
architect engineers and FGD  system suppliers frequently call for use of very
low  design  gas velocity, e.g., no greater than 480 feet per minute with no
exception allowed, thus sustaining the use of the simple low-efficiency
chevron designs.  In some cases it appears that such requirements are set by
structural  engineering considerations rather than the needs of the process
design.  Further, in some instances where somewhat higher design gas velo-
city is used, chevron-type equipment is employed at design gas velocity
constituting 80 to 85% of the critical (reentrainment threshhold) velocity.
Performance of such an installation is drastically affected by wide devia-
tions  in local gas flow rate.  Thus, the lack of attention to or the disre-
gard for optimal mist eliminator application will continue to result in
operating FGD installations  that fail to meet system performance and reli-
ability objectives.  Success of an FGD system is obviously dependent on
close  coordination between the Engineer responsible for system design and
the mist eliminator applications-specialist/supplier.

IMPROVED INSTALLATIONS FOR REGENERATIVE AND DIRECT STEAM REHEAT

     ^Control of scrubber carryover plays a direct role in all FGD systems  in
helping to  limit fouling and corrosion of downstream components handling wet
                                   11-56

-------
gas.  Moreover, mist eliminator performance and reliability is particularly
critical in achieving dependable operation of flue-gas reheat heat-exchange
equipment that is in direct contact with the scrubbed gas.  Use of high-
efficiency mist eliminator design to minimize droplet carryover (in con-
junction with employment of in-situ high flow-volume steam or air blowing of
the most exposed portions of the gas-wetted heat exchange surface to period-
ically remove deposits) has been successful in limiting exchanger corrosion
and maximizing heat transfer.

Regenerative Reheat

     In current application of FGD in high-fuel-cost nations like Japan and
West Germany growing use is being made of means of utilizing heat in raw
flue gas exitting electrostatic precipitators for reheating of water-
saturated flue gas from wet scrubbers.  Ljungstrom rotary heat exchangers
are applied to achieve major plant lifetime-savings by elimination of use of
reheat heat-sources requiring firing of additional fuel.  A tubular-
exchanger type regenerative reheat system has been utilized in U.S.A. in an
FGD facility now in initial startup.  Like the Ljungstrom units, this
first-of-a-kind facility indirectly cools the raw flue gas below its sulfu-
ric acid dewpoint temperature in extracting "free" heat for flue gas reheat.

TVA, Paradise Steam Plant — The regenerative-reheat-equipped 704 mW Para-
dise Unit No. 1 high-sulfur forced-oxidation FGD system, Drakesboro,
Kentucky, has been undergoing startup and commissioning throughout 1983.  A
two-stage high (1500 feet per minute) velocity horizontal-flow mist elimina-
tor operating at a 120° F. water-saturated condition specified for approxi-
mately 0.07 grains/SCF (dry) outlet liquid loading (16 micron limit drop
size) is utilized immediately upstream of a Type 317L tubular stainless
steel reheat exchanger designed to cool recirculating regenerative-reheating
water to 188° F.  A downstream tubular carbon steel second-stage reheat
exchanger is designed to receive the hot water supply at only 200°F. and
further reheat the flue gas to a reheater outlet temperature of 170° F.
Although only approximately 50% of regenerative flue-gas reheat capacity has
been realized in the 1983 operation to date, the mist eliminator operation
and use of retractable steam blowers at the reheat exchanger has prevented
any deterioration of the stainless steel tubes.  Because of the temporarily
inadequate reheating, the carbon steel reheat exchanger tubes have been
attacked by sulfuric acid deposition for which the system was not designed.
Due to inadvertant overheating of shut-down absorber modules by hot water
allowed to flow through the reheat exchangers, meltdown of mist eliminator
internals has occurred on several occasions, which may require replacement
with FRP.

NWK, Wilhelmshaven Power Plant — The Ljungstrom rotary-regenerative re-
heater equipped 350mW No. 2 FGD module at the low-sulfur Wilhelmshaven, West
Germany, plant has been in sustained high-availability operation since  early
1982.   A two-stage high-velocity horizontal-flow mist eliminator operating
upstream of the regenerative exchanger at a 115° F. water-saturated condi-
tion was specified for approximately 0.09 grains/SCF (dry) outlet liquid
loading (20 micron limit drop size) and has been discharging only 0.03
grains/SCF (dry) by test.  The system design and initial operation provided
                                    11-57

-------
for 10 to 15° F. reheat by flue-gas by-pass upstream of the Ljungstrom,
which heats the flue gas an additional 90° F.  The prior reheating is done
so as to help ensure against excessive sticking of mist eliminator carryover
slurry in the internals of the regenerative exchanger.  In a current trial
run with by-pass reheat shut off, the FGD system had at last report been
operating for eight weeks without any problems with regenerative reheat.

Direct Reheat

     Direct reheat of flue gas is accomplished by use of an in-line tubular
heat exchanger that is installed directly in the scrubbed flue-gas stream
and heats the gas by indirect heat exchange with steam or hot water.  The
heat exchanger is particularly vulnerable at its gas inlet end where the
gas-wetted surface is subject to low-pH attack.  Crevice corrosion is
especially troublesome if these surfaces become continuously fouled with
solids originating from scrubbing slurry carryover droplets.  In a number of
installations improved mist eliminator designs have been successful in
limiting carryover flow and, with periodic tube blowing, avoiding signifi-
cant reheater corrosion.

Arizona Public Service Co., Cholla Station — No. 2 FGD system, low-sulfur
coal service, utilizes a modern-design vertical flow mist eliminator as in
Figure 7 to protect a steam heated flue-gas reheat exchanger that contains
Incoloy alloy 825 tubes.  In conjunction with intensive periodic steam
blowing of the inlet face of the reheater bundle, the mist eliminator has
been fully effective in preventing fouling and corrosion of the reheater.

Montana Power Company, Colstrip Station — Utilizing a wash tray with a very
high  (3" to 5" W.G.) gas-pressure drop ahead of chevron mist eliminators the
low-sulfur FGD systems at Colstrip have operated without significant fouling
or corrosion of the steam-heated direct-reheat exchangers, which are plate
coil type constructed of Inconel alloy 625 (first stage) and Hastelloy alloy
G  (second stage).  As in the case of Cholla Station, intensive periodic
steam blowing of reheater internals are essential to sustained reheater
reliability.

RETROFIT UPGRADING OF COMMON LOW-EFFICIENCY MIST ELIMINATORS

     In numerous instances, and in order to upgrade performance, reliability
and temperature rating the advanced vertical-gas-flow mist eliminator design
(Figure 7) has been retrofitted in FRP to existing modules utilizing space
originally occupied by chevron type installations.

Big Rivers Electric Cooperative, Green Station

     The original chevron elements suffered from meltdown during brief shut-
down periods due to proximity to the flue-gas reheat heat source.  In retro-
fitting the advanced vertical-flow design, area-by-area replacement was
required by the Engineer and, hence, no increase in superficial gas velocity
nor advantageous reduction in limit drop size was accomplished.  However,
field observations indicated that the appearance (opacity) of the stack-
discharge trailoff was improved.
                                   11-58

-------
Duquesne Lighting Company, Phillips Station

     A similar replacement was made at Duquesne.  Operation is understood to
be satisfactory.

Louisville Gas and Electric Company, Cane Run Station

     Retrofitting of the same advanced design together with a compatible
bulk-removal mist eliminator (Figure 4) upstream of it was motivated, in
part, by incidents in which prior chevron elements melted down during
shutdown periods due to proximity to the reheat heat source.  Through an
eliminator washing improvement program a substantial increase in sustained
performance and reliability has been achieved.

                                 CONCLUSIONS

     1.  Mist eliminator selection and design as well as and washing tech-
nique are major factors influencing the performance and reliability of FGD
systems.

     2.  Mist eliminator technology advancements in design and forming of
aerodynamically shaped, extruded profile elements for use in either a
horizontal or vertical-gas-flow regime offer means of gaining potential
performance substantially greater than that of  chevrons.

     3.  Absorber design provisions for gas flow distribution are of criti-
cal  importance  in realizing the performance capability of an eliminator
system.

     4.  Satisfactory mist eliminator performance and reliability will be
best achieved when the designs of the absorber  system and of the mist-
eliminator system are coordinated through the integrated efforts of the wet
scrubber designer and mist eliminator supplier/specialist.

     5.  Traditional use of low-velocity chevron mist eliminators continues
and  is judged to be an important contributing factor in many low-pH-end FGD
reliability problems and in instances of poor or marginal gas cleaning
performance.

     6.  Conservative, high-performance mist  eliminator design and  effective
eliminator washing operation are of critical  importance in  supersaturated-
CaSO -mode FGD  systems, both natural and forced oxidation,  in view  of
serious impact  on FGD system reliability of the absorber-emission species
characteristic  of this type of process.


                                 REFERENCES

  1.  Ellison, W.  Scrubber Demister Technology  for  Control  of  Solids Emis-
     sions from SO  Absorbers.   (Presented at EPA  Symposium on  Transfer
     and Utilization of Particulate Control Technology.   Denver,  Colorado.
     July 24-28,  1978.)
                                    11-59

-------
 2.   Ellison, W., and S. A. Lefton.   FGDS  Reliability:  What's  Being Done to
     Achieve it.  POWER, May  1982.

 3.   Moen, D. A. et al, Coal  Creek  Station Air  Quality  Control  System.
     (Presented at 29th Annual Conference  of  the Association  of Rural
     Electric Generating Cooperatives.  Vail, Colorado.   June 11-14, 1978.)
     P  6,7.

 4.   Calvert, S.  Guidelines  for  Selecting Mist Eliminators.   Chemical  Engi-
     neering, February 27,  1978.  p  110.

 5.   Laseke, B. A., and T.  W. Devitt.   Status of Flue Gas Desulfurization
     Systems in the United  States.   (Presented  at  29th  Annual Conference of
     the Rural Electric Generating  Cooperatives.   Vail, Colorado.   June
     11-14,  1978.)  p 33-35.

 6.   Tennyson, R. P.  Mist  Eliminator Design  and Application.   (Presented at
     70th Annual Meeting of the Air Pollution Control Association.   Toronto,
     Ontario, Canada.  June 20-24,  1977.)  p  77-25.4.

 7.   Pedersen, G. C.  Experiences with  Control  Systems  Using  a  Unique
     Patented Structure.  Kimre,  Inc.   May 1983.

 8.   Bakke,  E., and H. P. Willett.   The Application  of  a  Tubular Wet Elec-
     trostatic Precipitator for Fine Particulate Control  and  Demisting  in an
     Integrated Fly Ash and SO  Removal System  on  Coal  Fired  Boilers.
     (Presented at E.P.A. Symposium on  Particulate Emission Control.  1982.)

 9.   Reference  1.

10.   Henzel, D. S., and D.  H. Stowe.  A Proven  Reagent  for High Sulfur  Coal
     Flue Gas Desulfurization.  (Prepared  for the  EPA/EPRI Flue Gas Desul-
     furization Symposium,  Hollywood, Florida.  May  17-20,  1982.)

11.   Electric Light and Power, Technical Publishing  Division, Dun-Donnelley
     Publishing Corporation.  (Overview of EPA/EPRI  FGD Symposium)  Table 5
     Availability FGD System  Records for Plants Burning 3.0%  or Greater
     Sulfur.  June 1982.

12.   Tampa Electric Company,  letter of  September 2,  1983.

13.   Henzel, D. S. et al, Limestone FGD Scrubbers:   Users Handbook.  PEDCo
     Enviromental Inc. and  Black  and Veatch.  EPA-600/8-81-017. December
     1981.   p 3-37 to 3-41.

14.   Private Communication  from E.  Reinhard,  Euroform GmbH, March,  1983.

15.   Johnson, L. D., and R. M. Statnick.   Measurement of  Entrained  Liquid
     Levels  in Effluent Gases from  Scrubber Demisters.  EPA-650/2-74-050.
     June 1974.
                                   11-60

-------
                     FGD GYPSUM:  UTILIZATION VS. DISPOSAL

                                      By:

                              William Ellison, PE
                              Ellison Consultants
                           Monrovia, Maryland  21770
                                   ABSTRACT

     The purpose of this paper is to give a technical and economic evalua-
tion of alternative means of managing wastes collected in forced-oxidation
(FO) type flue gas desulfurization (FGD) systems.  Worldwide application of
FO-FGD is described and environmental considerations in selection of waste
management alternatives - gypsum utilization vs. disposal - are reviewed.
Feasibility of and barriers to commercial gypsum from FGD in North America
during the 1980s are analyzed.  The impetus, method of implementation, and
potential for gypsum use are addressed for major by-product gypsum FGD
installations now in operation, under construction or in the design phase.
It is concluded that in many instances, the production and sale of usable
gypsum from FGD would be a benefit and a source of direct profit to the
utility plant owner, but that in most cases the gypsum depletion allowance
under IRS Code 613B will continue to be a major disincentive for purchase of
by-product gypsum by vertically integrated gypsum companies in U.S.A.
                                 INTRODUCTION

PRODUCTION OF GYPSUM BY FORCED OXIDATION

     Gypsum, double-hydrated calcium sulfate, is a naturally occurring non-
metallic mineral used as a raw material in the manufacture of gypsum board,
Portland cement, plaster products and in agriculture.  Forced oxidation as
applied to wet lime/limestone flue gas desulfurization (FGD) consists of
forcing air into the spent scrubbing slurry to produce gypsum, the oxidized
form of calcium sulfite.  The gypsum can then be easily dewatered to a cake
containing greater than 80% solids by weight.  This operating technique is
commonly used worldwide.  North America is the last major FGD-using area to
adopt FO operation beginning in 1978, bringing about a major departure from
the natural-oxidation design and operation of FGD facilities.  Since that
time, approximately 20 power plants have purchased provision for FO in FGD
operations constituting 16,000 mW of scrubbed generating capacity and almost
half of all limestone FGD committed to date.
                                    11-61

-------
ADVANTAGES OF FORCED OXIDATION

     This SO  removal method is superior to natural-oxidation FGD commonly
applied in U.S.A.  since it provides improved control of FGD-process operat-
ing conditions, which may result in higher FGD availability in limestone
based SO  absorption.  It results in improved control of internal scaling,
better utilization of limestone chemical reagent and enhanced SO  removal
efficiency.
     Moreover, the solid waste produced from the SO -catch is in the form
of comparatively coarse-grained calcium sulfate solids that may be more
easily dewatered,  transported and stored and which require less than 60
percent of the solid waste disposal area required for direct ponding of
sludge without forced oxidation.  Additionally, gypsum formed by FO-FGD has
potential for use  in gypsum board manufacture, cement manufacture and
agriculture.

MANNER OF WORLDWIDE IMPLEMENTATION

     On a worldwide basis, wet scrubbing using FO of the S0»-catch is the
principal means of desulfurization of flue gas (FGD) from electric utility
boilers.  FGD waste solids, initially generated as a slurry, are subse-
quently dewatered  to a sludge or cake containing both solid and liquid waste
forms.  The decanted liquid and filtrate is generally returned to the
process.  However, build-up of dissolved solids in the scrubbing slurry may
preclude closed-loop operation resulting in a wastewater effluent requiring
disposal.  Production and use of salable gypsum avoids discarding of FGD
solid waste at a permanent disposal site.

     The production and utilization of a usable grade of gypsum from FO-FGD
is routinely done  in Japan and West Germany, the two principal countries
that have applied  flue gas desulfurization overseas.  However, in view of
the limited number of usable-gypsum type limestone-scrubbing systems that
have been purchased in North America (see Table 1), we continue to be
principally oriented to throwaway-solid-waste FGD operation even after the
transition in the  past five years to use of forced oxidation instead of
natural oxidation.

                            DISPOSAL OF FGD GYPSUM

     The disposal  of FGD wastes, both dewatered solids and liquid effluents,
has the potential  for pollution of groundwater and surface water bodies
impacted by waste  disposal activities.
WASTE SOLIDS

     Land-based disposal of scrubber solids from solids-precipitation-type
FGD systems is a problem principally because of the very large waste quant-
ity generated.  At the same time, EPA regulatory activity is concerned with
the possible long-term effects leading to pollution of surface water and
groundwater bodies due to the leaching of constituents such as trace-metals
                                    11-62

-------
                              TABLE 1.  COMPARISON


                  OF COMMITTED FO-FGD INSTALLATIONS YIELDING

                   SALABLE GYPSUM VS.  THROWAWAY SOLID WASTE

Power Plant	mW	% of Total

AESC, Cogen. 1,
Houston, Texas                                 150 (Approx.)           1.0

JEA, SJRPP 1 and 2,
East Port, Florida                           1,200                    7.7

MPW, Muscatine 9,
Muscatine, Iowa                                166                    1.1

Tampa Electric, Big Bend 4,
Tampa, Florida	                    475                    3.0

Sub-Total, Salable Gypsum Producing:         2,991                   12.8

Aggregate of FO-FGD  Throwaway
     Waste Installations:	13,517	87.2
     Total FO-FGD Committed in U.S.A.        16,508                  100.0
from these waste solids.  However, hydrogeology of typical sites that may be
used for storage or ultimate disposal of wastes from such high-volume
sources can be complex.  Thus, for an individual site there may be uncer-
tainties as to the extent of adverse environmental impact due to permeation
of leachate through the discarded waste mass and the underlying strata.   The
two main disposal options for discarding solid waste generated by FO-FGD are
landfilling and wet stacking.

Landfilling

     FO-FGD operation yields typical gypsum crystal size of from 40 to 100
microns, producing an easily dewatered and handled coarse-grained cake that
can, if necessary, be readily transported as an 80-90% solids mass to a
remote disposal site.  However, an implication of this low moisture content,
(and the inability of the discarded coarse-grained waste to retain any
significant amount of FGD liquor over an extended period), is that little of
the scrubber liquor can be disposed of by occlusion in the solid waste mass.
Consequently, the operation of closed loop FO-FGD may not be possible
because of an increase in system corrosion resulting from a greater than
10-fold increase in FGDliquor dissolved solids concentration e.g., to in
excess of 100,000 ppm chloride in the case of a medium/high sulfur coal
application (1).

     Environmental concerns associated with landfilling of FO-FGD waste
solids include:
                                    11-63

-------
     -  Water-soluble material (dissolved metals and salts including calcium
        sulfate)  is present in seepage and runoff.

        After the lifetime of the FGD operation, seepage and runoff of
        leachate will continue unabated with no means of assimilating it by
        recycle operation.  (The substantial solubility-in-water of the
        gypsum solids will yield an endless source of leachate saturated
        with calcium sulfate and containing other components such as sulf-
        ite, dissolved metals and salts.)

Wet Stacking

     Based on field tests (2), FO-FGD solid waste may instead be disposed of
by wet stacking.   This is a technique utilized extensively in the phosphoric
acid industry in Florida for land-based disposal of waste-gypsum, wherein a
10 to 15% solids aqueous-slurry is discharged at a disposal site so as to
permit formation of a sharply-pitched stack over the plant lifetime.  In
this method to be applied at a number of U.S. power plants, reduced acreage
use is achieved through such stacked disposal without need for dewatering by
thickeners and filters or centrifuges.  Overall capital costs and annual
revenue requirements are understood to compare favorable with even simple
ponded-waste natural oxidation FGD systems.  In the so-called "upstream
method" of construction, the wet stacking is begun by first constructing an
earthen starter-dike forming a sedimentation pond.   FO-FGD solid waste is
pumped to the pond in a slurry form, usually at 10 to 20 percent solids, and
allowed to settle and drain (3).  Process water is decanted from the pond
and returned to the plant.  When sufficient gypsum is deposited within the
pond, gypsum is excavated with a dragline to form and step-wise build up a
perimeter dike immediately inside the starter dike.  Decanted liquor is
piped under the starter dike to an annular perimeter ditch and surge pond
formed by the starter dike and a clay exterior dike.  The process of
sedimentation, excavation, and raising of the perimeter dikes continues on a
regular basis during the active life of the stack.   Using the upstream
method of construction, gypsum stacks in the phosphate industry have reached
heights exceeding 100 feet (30.5 m) with slopes of 1.5 horizontal to 1.0
vertical, which is approximately the a.r.gle of repose of some gypsums.  These
steep slopes result from casting the gypsum with a dragline and allowing
some gypsum to roll down the outside of the stack to eliminate need for
shaping.  Therefore, the gypsum perimeter dikes of some stacks have a factor
of safety very close to unity and from a conventional geotechnical engineer-
ing point of view, failures of gypsum stacks sometimes occur.  Fortunately,
gypsum is a very forgiving material, and, unlike most mine tailings, gypsum
does not readily flow.  Therefore, the consequence of a failure is the loss
of process water stored on the gypsum stack.  If process water escapes the
plant property, liabilities from pollution and environmental damage may also
result.  On a day-in, day-out basis, seepage of process liquid to ground-
water and surface water should be controlled by use of an impervious founda-
tion as well as of clay-type exterior and starter dikes, and by collecting
and treating or recycling all seepage and runoff.  Although a greater
occlusion of process liquor in solid waste occurs in wet stacking than in
landfilling, in situ washing of the stacked solids will occur (due to the
high permeability of the stack).  This leachate when subsequently collected
                                    11-64

-------
and recycled would be expected to lead to a salt concentration build-up even
greater than that which would occur using in-plant dewatering and landfill
disposal.  This would most certainly preclude closed loop operation of the
FGD system in the case of wet stacking.

     Environmental concerns in use of wet stacking for a FO-FGD waste
include:

     -  The very large working area of the stack facility will cause the
        liquor recycle to FGD (returning from the stack-pond section) to
        also return large quantities of liquid running off from the entire
        stacking facility.  Resulting water-balance problems in rainy
        weather may lead to intermittent outboarding of substantial dis-
        charges of process-liquor.

     -  All pile run-off will be saturated with calcium sulfate and contain
        gypsum surface-moisture constituents (dissolved metals and salts).

     -  Over a twenty-year lifetime, a large stack in excess of 100 feet
        high may be created, e.g., an increment of 40 mW, 3 percent sulfur
        bituminous coal, will generate a simple truncated conical stack, 400
        feet in base diameter, 100 feet in diameter at the top, 100 feet
        high, with side slopes of 1.5 horizontal to 1.0 vertical.  In no
        country are regulations or experience available as to how to satis-
        factorily decommission a lifetime gypsum stack or how to reclaim it.

        Stack slopes are too great to hold topsoil and vegetation.  Thus,
        normal concepts for "covering" of the waste in this manner do not
        apply.

     -  After the FGD lifetime, seepage of leachate will continue unabated
        with no user to which to recycle for re-use.  The substantial
        solubility-in-water of the gypsum solids will yield an endless
        source of leachate saturated with calcium sulfate plus other compo-
        nents such as sulfite (chemical oxygen demand — COD) plus any
        remaining surface-moisture constituents.

     -  Cold winters may severely restrict dragline operations during cold-
        weather months and jeopardize continuity of FGD operation.  In addi-
        tion, freeze-thaw action is of concern so far as maintaining the
        integrity of dikes.

LIQUID EFFLUENT

     In view of the low surface-moisture content of coarse-grained solid
waste from FO-FGD systems, provisions for management of saline FGD liquor
blowdown are essential for pollution control.  Thus, the FO-FGD  system
designer would be expected to seek all practical design means to manage
waste effluent such that acceptably low scrubber-liquor salinity is main-
tained (4) .
                                    11-65

-------
Significance of Discharge

     Although FO-FGD blowdown may, by volume, be only a small proportion of
the total power-plant wastewater, it contains dissolved-salts concentrations
approaching those in sea water and thus may exceed the assimilative capacity
of available receiving streams.  Such outfall, untreated, can also contain
objectionable concentrations of trace metals such as arsenic, mercury,
cadmium and lead that may originate from the fired coal (5).

     It should be noted that liquid purge outfall such as required from
FO-FGD is not always necessary in a natural oxidation mode because these
latter systems yield a solid waste (primarily calcium sulfite) that is
fine-grained, thixotropic and highly water retentive (containing approxi-
mately 50% moisture).  The encapsulation of a substantial amount of FGD
liquor in the discarded waste mass may provide the liquid purge necessary to
maintain the scrubber dissolved salts at an acceptable concentration.
However, natural oxidation sludges have many disadvantages as compared to
those produced in a forced oxidation mode.  These disadvantages include both
physical properties — the completed disposal site cannot be  reclaimed for
uses requiring the soil to support structural loads — and chemical proper-
ties.  The principal contaminants in the sulfite sludge are dissolved  salts
(principally sulfates), chemical oxygen demand due to sulfites, and soluble
heavy-metal trace elements originating in the coal.  Thus leaching and
runoff from sites used for disposal of such solids have also  been of sig-
nificant environmental concern.  An important remedy is sludge fixation
(stabilization), which consists of chemical processing of scrubber sludge to
facilitate handling, transportation, placement, and consolidation at the
ultimate disposal site.  One of the methods used is pug-mill  blending  and
aging of mixtures of FGD filter cake, dry fly ash and other dry additives,
followed by landfill compaction.  An alternative procedure involves the
addition of silicate-containing materials to thickened FGD slurry before
discharge to a pond site.  Extensive testing indicates that these methods of
chemical treatment significantly improve the load-bearing characteristics of
FGD sludge, reduce the solubility of the major chemical species by a factor
of 2 to 4 and reduce sludge permeability by an order of magnitude or more
(6).  Further, as outlined previously, the encapsulation of a substantial
amount of FGD liquor in the consolidated waste mass from natural oxidation
FGD provides the option for all-in-one, combined disposal of  all solid and
liquid waste without discharge of a liquid outfall.

Alternatives for Treatment/Ultimate-Disposal

     The liquid purge may be treated to precipitate trace-metal constituents
and reduce their concentration to an acceptable level so that the effluent
stream may be discharged to a suitable receiving stream or water body.  Such
treatment followed by outfall discharge is the economically preferred  method
of FGD liquor disposal when it can be permitted.

     Candidate wastewater treatment methods to reduce trace metal concen-
trations include:
                                    11-66

-------
        Chemical precipitation

        Ion exchange  (exchange of complex anionic species)

        Sulfide precipitation

        Chelation/absorption.

     As indicated above, because of the large dissolved solids concentration
of purge-liquor, available surface water receiving streams at inland sites
may be unable to tolerate the environmental impact of saline outfall dis-
 charge of this type.  Blowdown liquor after treatment can, in this case, be
concentrated utilizing commercially available equipment to crystallize
dissolved salts to a  comparatively small mass of salt cake that can be
discarded at a secure landfill disposal site or used as a rock salt by-
product.

                           UTILIZATION OF FGD GYPSUM

INTRODUCTION

     Annual consumption of crude gypsum in the United States is estimated to
be approximately 20 million  tons, with 30 - 40 percent being imported,
mainly from Canada and Mexico.  The average value of crude gypsum was
estimated to be $9.00 per ton (F.O.B. mine) in 1982 (7).

     The sale and utilization of gypsum from FGD would help to offset a
small portion of the  operating cost of the FGD system and would minimize the
environmental impact  of and  use of land for ultimate disposal of FGD solid
wastes.  There are several factors to be considered in determining the fea-
sibility of FGD gypsum sale  and utilization.  These include technical and
economic aspects as well as  incentives, problems and barriers which are
specific to the circumstances in the United States.

TECHNICAL FEASIBILITY

     The gypsum formed by FO-FGD is unlike natural gypsum, the major differ-
ences being:  bulk handling  properties (moisture content, grain size), quan-
tity of impurities, and nature of impurities.  Natural gypsum is mined as a
coarse, low-moisture  rock at approximately 70 - 90 percent purity, with
limestone and/or anhydrite (non-hydrated calcium sulfate) comprising the
majority of impurities.  FGD gypsum is produced as a higher moisture filter
cake of approximately 90-95  percent purity, primarily coarse-grained crys-
tals along with ash,  lime or limestone components and dissolved salts
comprising the majority of impurities.  The technology to make and use
board-grade and cement-grade FGD gypsum exists and is currently in use in
Japan and Europe.   Depending on end use (cement, prefabricated products or
agriculture), the FGD gypsum may require processing (beneficiation) before
it is suitable for use.  For example, gypsum is used in cement manufacture
both as a grinding aid and as a source of calcium sulfate to retard setting.
To be acceptable for this use, FGD gypsum cake would need only minimal
washing (to ensure chloride  content of less than 1 percent), but requires
                                    11-67

-------
drying/agglomeration to achieve suitable grinding properties.  To be used in
gypsum board manufacture, FGD gypsum cake requires washing, (the amount
depending on chloride content, unwashed, and board manufacturers' specifi-
cation) and may require drying/agglomeration.  Additionally, suitable FGD
process design and control is vital to ensure an adequate degree of oxida-
tion of sulfite.

ECONOMIC FEASIBILITY

     There will be costs incurred by the utility in converting the system
design and ultimate operation from disposal to utilization of FGD gypsum.
The net cost to the utility will be the beneficiation costs less the hypo-
thetical disposal cost if the gypsum was not to be used.  However, the
utility should be able to sell the gypsum, which may either offset the
production cost or result in a net credit to the utility.  As described
earlier, natural gypsum has an average value of $9.00 per ton (F.O.B mine).
Thus the apparent value of gypsum to the user is typically $9.00 per ton
plus transportation costs.  Transportation costs for gypsum imported from
Nova Scotia and delivered to the southeastern U.S.A. were estimated to be
$15.00 per ton in 1982 (8).  Thus the value of gypsum to a user can range
from $9.00 to $20 - 30 per ton at site of use, the F.O.B. value in Nova
Scotia being as high as $15.00 per ton.  However, in the United States, a
depletion allowance of 14 percent of the price of gypsum is granted to
mining companies as a tax benefit to compensate for the depletion of mine
reserves (9).  This depletion allowance is applied to the price of gypsum
F.O.B. point of sale, not minesite, and therefore includes transportation
costs  (plus other costs for handling and preparation).  Thus, the tax credit
to the gypsum company effectively reduces the cost of mined gypsum.  Al-
though it is beyond the scope of this study to quantify the benefits of the
depletion allowance, it is suggested that the actual cost of gypsum to a
gypsum mining company/board-manufacturer is substantially less than this
apparent value of $9.00 to $20 - 30 per ton at site of use.

     The costs to the utility of producing usable gypsum are estimated to
range from $5 to $18 per ton for board-grade gypsum and from $8 to $11 per
ton for cement-grade gypsum (8).  The wide range of costs is due to the
diversity of possible methods of processing the gypsum cake (10).  This is
in turn due to the wide variance in board manufacturers' specifications (see
Table 2) and the various options available for dewatering (rotary drum
filter, horizontal belt filter, etc.), agglomeration (pelletization, extru-
sion, briquetting) and drying (fuel-fired vs regenerative).  However, the
cost of producing usable gypsum may be partially or completely offset by
hypothetical disposal costs.  The comprehensive analysis of the costs
associated with landfilling a filtered gypsum waste is a complex procedure.
A preliminary assessment of these costs yields disposal costs ranging from
$10.00 to $25.00 per ton (11).  Costs associated with such disposal include
land costs, site preparation costs, dewatering costs, transport to disposal
site, wastewater treatment and reclamation.  The wide range of costs is due
to variance in land availability and suitability, and uncertainty in waste-
water treatment costs.  The estimate of wastewater treatment cost is uncer-
tain as it depends on the quantification of the flow volume of liquid purge
required to limit the dissolved solids in the scrubber slurry to an accept-
                                   11-68

-------
able level.  This limit and volume depend on coal chloride content,  washed-
cake quality criteria, materials of construction, etc.
                              Table 2.  WALLBOARD
                         MANUFACTURERS'  SPECIFICATIONS
                        FOR FGD GYPSUM FOR USE IN BOARD

National
Gypsum

Georgia
Pacific
U.S.
Gypsum
Gypsum Content, %, Min.
Calcium Sulfite
  Hemihydrate, %, Max.
Na, ppm, Max.
Cl, ppm, Max.
Mg, ppm, Max.
Free Water, %, Max.
PH
Particle size, Min.
   94

  0.5
  500
  800
  500
    1
  6-8
  xy=
2000 sq.
micron
 90

N.S.
200
200
N.S.
 12
3-9
N.S.
   95

    2
   75
  120
   50
   10
6.5-8
   20
 (mean)
 micron
Note:  N.S. = Not specified
     In summary, the cost of FGD gypsum beneficiation is offset by the dis-
posal costs which would have been incurred in a non-utilization landfill
disposal option.  If lower-cost wet stacking is the nominal disposal option,
gypsum beneficiation may be seen to cost more than disposal.  However,
because the gypsum can be sold for use in gypsum board or cement, a net
credit may be realized if arrangements are made for a suitable user.  If
this is so, the economics of production and sale of by-product gypsum can be
expected to be favorable.

PROBLEMS AND BARRIERS

     Many influences within existing gypsum markets affect the potential for
recovery of gypsum from by-product sources.  In each case, many factors
(location, capacity, quality, etc.) will significantly affect the
possibility of such by-product gypsum utilization.  Assessment of these
factors is complex since there has been no extensive by-product gypsum
utilization in North America and because of the vertical integration of
major firms in the gypsum industry (i.e. gypsum mines, transportation
systems and board manufacturing plants are generally all owned and operated
by gypsum companies).
     Principal obstacles to gypsum use by board and cement manufacturers
                                           are
                                   11-69

-------
as follows:

     -  Natural gypsum is a plentiful, inexpensive, raw material.  (Ease of
        transportation of by-product to the potential user(s) as contrasted
        with the conventional source is a crucial factor in its economic
        use.  As with natural gypsum, the transportation cost will generally
        be the most significant part of the delivered cost.)

     -  Gypsum users are reluctant to convert to by-product gypsum unless
        both quality and quantity of the gypsum are assured.  Given the fact
        that electric utilities exist to produce power and are not
        accustomed to serving markets outside the energy sector, warranting
        of by-product quality and quantity may be a significant problem.

     -  When ample land areas for disposal are available close to the power
        plant, the impetus for producing salable gypsum may not be present.

     -  Historically, the gypsum and utilities industries have each been
        very conservative and insular.  The production and use of by-product
        gypsum by these industries will require policy changes that they may
        find difficult to implement.

                ACTIVE BY-PRODUCT FGD PROJECTS IN NORTH AMERICA

     The four plants (Table 1) that have committed to date to the production
 of  salable gypsum vary so far as geographic location, system size and type
 of  utility company, but all are similar so far as:

        The basis for use of forced oxidation including the impetus for
        production of usable gypsum

     -  System design and operation

        The prospect of gaining use of the by-product.

 KEY MOTIVATING FACTORS

     Aside from major process benefits of forced-oxidation mode operation,
 including  improved control of internal scaling and increased reagent
 utilization, by-product gypsum applications to date have incorporated
 forced-oxidation with production of a salable product because of substantial
 constraints preventing use of an extensive area of nearby land for throwaway
 type disposal.  Also, apart from potential sale/use of the gypsum, this
 manner  of  applying forced-oxidation ensures that the SO -catch is rendered
 in  a cake  form that may readily be handled and temporarily stockpiled, and
 thereafter ultimately transported any necessary distance to an off-the-site
 point or points of delivery.  Under such circumstances, and because of the
 comparative ease and limited cost to adjust the design and quality of the
 solid waste to yield a usable product, each of the projects has with
 adequate incentive (but without agreements or contracts for product sales),
 committed  to a design compatible with by-product utilization.  The resulting
 gypsum  yield is a benign and highly acceptable solid waste except for its
                                    11-70

-------
significant solubility in water, and can possible be sold to existing users
or suppliers of natural gypsum located reasonably close to the power plant.

DESIGN AND OPERATION

     In the FGD systems sold for usable gypsum production, the bulk of SO
removal is achieved in counterflow absorbers that discharge slurry effluent
to gas quenchers (precoolers) that yield an exit scrubbing slurry having a
suitably low percent calcium sulfite in the solid phase.  No isolation of
the gypsum product from hydrogen chloride gas or other non-SO  components
of the raw flue gas removed in wet collectors is achieved in the gas
scrubbing section.  After dewatering of this final slurry, a wet cake
product containing as little as 8 percent surface moisture is discharged
from rotary vacuum filters equipped with cake washing accessories.  The
concentration of objectionable water-soluble constituents in the gypsum cake
product is limited both by limiting of concentrations in the scrubber liquor
through discharge of liquid purge from the slurry dewatering system to an
acceptable receiver, and by the washing of the cake after it forms on the
filter drum.  The concentration of objectionable solid-phase constituents
other than calcium sulfite is limited by use of a high-efficiency
electrostatic precipitator upstream of FGD, and of limestone reagent of
adequate quality/purity.

POTENTIAL FOR USE OF GYPSUM PRODUCED

     These FGD systems are capable of continuous production of gypsum that
meets specifications for intended use, wallboard in the case of three plants
in Table 1 and other less critical uses in the case of the relatively small
output of Muscatine Power and Water.  Market demand for natural gypsum in
the regions in which the plants are located is sufficient to assimilate the
FGD by-product yields, but each of the FGD systems has been purchased
without benefit of a gypsum sales agreement with a potential user(s), and no
such contracts have been announced to date.

Advantages of By-Product Substitution

     The principal advantages to potential buyers of use of these FGD gypsum
outputs are:

        The material may be expected to be supplied at a uniformly
        acceptable level of quality.

        The quantity of supply is substantial in relation to the
        contemplated end use(s).

        Particularly in Florida where natural gypsum is presently shipped
        long-distance from Nova Scotia to supply wallboard manufacturers, or
        principally from Spain in the case of gypsum rock for cement set
        retarder, substantial savings in transportation cost may be
        realized.
                                   11-71

-------
Means for Meeting User Requirements

     Provisions for drying and/or agglomerating of the FGD gypsum may in
some instances be necessary since prospective users may incur excessive
costs or inconvenience using raw washed FGD cake instead of gypsum rock.
Presence of surface moisture may also make it difficult to handle and store
the material in existing facilities designed for natural gypsum rock.
Moreover, water adds to shipping weight and unit transportation cost and
requires added energy use by the buyer since drying will generally be needed
for purposes of the end use.  In specific instances, for example supply of
gypsum for use as cement set retarder if relevant, agglomeration of the
product by briquetting, extruding or other suitable means may be essential
for successful marketing of plant output.

Principal Obstacles to Use of By-Product

     Lack of success in sales contract negotiations to date points to the
need for an equitable, generally acceptable, and broadly applicable basis
for pricing of the product.  This may be essential in establishing long-term
by-product gypsum sales contracts between power plants and the major gypsum
users best able to benefit from use of the new source of supply.  The deple-
tion allowance on natural gypsum authorized by the U.S. Internal Revenue
Service Code-613B significantly reduces the financial benefit to the verti-
cally integrated, North American gypsum industry of substitution of by-
product.  If the allowance is to continue to apply in an era of newly and
plentifully available by-product gypsum suitable for principal market uses,
the by-product producer must realistically evaluate the supply economics and
competitive pricing of the product taking into account the purchasing disin-
centive created by the IRS tax allowance.

                                 CONCLUSIONS

     1.   Forced oxidation type flue gas desulfurization (FO-FGD) technology
          offers a design means for forming a usable gypsum by-product
          instead of a throwaway solid waste.

     2.   Widespread application of FO-FGD in U.S.A. since 1978 has been
          almost entirely in a throwaway waste mode with little apparent
          incentive for usable gypsum production.  Only in limited instances
          in which the throwaway-disposal option is very constrained, i.e.,
          13 percent of all U.S. FO-FGD capacity purchased to date, is the
          FGD design predicated on manufacture of usable gypsum, and in none
          of these cases, four in number, has a formal commitment to
          purchase or use the by-product been obtained to date.

     3.   Cost advantages point to possible future use of "wet stacking"
          type disposal in FO-FGD design and to significant environmental
          impacts that are largely circumvented in earlier natural oxidation
          type FGD designs that employ sludge fixation by use of fly ash and
          other additives.
                                    11-72

-------
     4.    Available technology and the successful widespread utilization of
          FGD gypsum in Japan and West Germany indicate that its  use  is
          technically feasible.

     5.    It is likely that,  with or without beneficiation,  the sale  of  FGD

          gypsum to the board and cement manufacturers  would be economically
          preferable for the  utility company in lieu of the  throwaway-
          disposal alternative.

     6.    Institutional barriers to the use of FGD gypsum throughout  North
          America include the high degree of vertical integration of  the
          gypsum wallboard manufacturing industry, the  disincentive for
          by-product gypsum utilization created by the  gypsum depletion
          allowance under U.S. Internal Revenue Service income tax codes,
          and the abundance at diverse locations of natural  gypsum of
          adequate quality.

     7.    Additional major factors influencing the profitability  of
          production and sale of usable FGD gypsum include the magnitude of
          alternative hypothetical throwaway-disposal costs, transportation
          costs to user(s), and costs of transportation of user's nominal
          source(s) of natural gypsum supply.
                                  REFERENCES

1.   Jenkins,  S.D.,  and W.  Ellison.  Uitilization of FGD By-Product  Gypsum.
     (Presented at EPA/EPRI FGD Symposium,  Hollywood,  Florida,  May  17-20,
     1982.)

2.   Pruce,  L.  M.  Evaluating the Newest Disposal Options for Scrubber
     Sludge. POWER,  May 1981.  p. 64.

3.   EPRI Report CS-1579 by Ardaman & Associates, Inc., Orlando,  Florida,
     Evaluation of Chiyoda Thoroughbred 121 FGD Process and Gypsum  Stacking,
     Volume  3:   Testing the Feasibility of  Stacking FGD Gypsum. November
     1980.

4.   Ellison,  W.,  and P.M.  Kutemeyer. New Developments Advance  Forced
     Oxidation FGD.  POWER,  February 1983. p. 43-45.

5.   Disposal of By-Products from Nonregenerable Flue Gas Desulfurization
     Systems:   Second Progress Report. US Environmental Protection  Agency,
     Report  No. EPA-600/7-77-052, May 1977-

6.   Control of Waste and Water Pollution from Power Plant Flue Gas Cleaning
     Systems:   First Annual R and D Report. EPA-600/7-76-018, October 1976.

7.   Mineral Industry Surveys. U.S.  Dept. of Interior, Bureau of Mines.
     Gypsum  in 1982, January 1983.
                                   11-73

-------
8.    Luckevich,  L.M.,  Ontario Research Foundation,  Mississauga,  Ontario,
     letter of September 22,  1983.

9.    Byron, B.B., Member of House of Representatives, Washington,  B.C.,
     letter of September 6, 1983.

10.  Economics of Disposal of Limestone Scrubbing Wastes:   Sludge/Flyash
     Blending. EPA-600/7-79-069,  February 1979.

11.  FGD By-Product Disposal Manual, Third Edition. Michael Baker  Jr.  Inc.,
     Chapter 16, January 1983.
                                   11-74

-------
  OPERATING EXPERIENCE WITH THE CHIYODA THOROUGHBRED 121
             FLUE GAS DESULFURIZATION SYSTEM

S.  Kaneda,  M.  Nishimura, H. Wakui,  I. Kuwahara,  D.  D.  Clasen

-------
                        OPERATING EXPERIENCE WITH THE
          CHIYODA THOROUGHBRED 121 FLUE GAS DESULFURIZATION SYSTEM

                   Seiichi Kaneda and Mitsuhiro Nishimura
                      Mitsubishi Petrochemical Company
                             Yokkaichi, Japan

                      Hitoshi Wakui and Ikuro Kuwahara
         Chiyoda Chemical Engineering and Construction Company, Ltd.
                               Yokohama, Japan

                              Donald D. Clasen
                      Chiyoda International Corporation
                             Seattle, Washington
                                  ABSTRACT

     This paper reviews the design and operating experience of the Chiyoda
THOROUGHBRED 121 Flue Gas Desulfurization System installed and operated by
Mitsubishi Petrochemical Company at their Yokkaichi, Japan, complex.
                                                                  3
     The plant consists of a single scrubber and treats 260,000 Nm /h
(162,000 scfm) of flue gas from a 280 T/h (87 MW equivalent) boiler burning
high sulfur oil.  Plant operation, since startup in May 1982, has been smooth
and trouble free over a wide range of operating conditions.  Plant reliability
has been 100%.  The plant is operated at a S02 removal efficiency of 97 to 99
percent for inlet S02 concentrations ranging between 1000 and 2000 ppm.  Lime-
stone utilization is greater than 99 percent and the dry, gypsum by-product is
sold to a wallboard manufacturer.  Operating and maintenance functions for the
system are minimal and completely absorbed by normal boiler plant operations.
                                INTRODUCTION

     In March of 1981, the Mitsubishi Petrochemical Company awarded Chiyoda
Chemical Engineering and Construction Company, Ltd. a contract to install its
patented,  second generation Chiyoda THOROUGHBRED 121 (CT-121) Limestone/Gypsum
System.  The System was selected due to its low space requirement, simple
operation,  and low plant cost.   In addition, Chiyoda's engineering capability
and expertise were highly regarded based on Mitsubishi's experience with two
first generation CT-101 Systems.

     The CT-121 System was to be designed to remove 97% of the S02 contained
in the flue gas from a new boiler being constructed by Mitsubishi Petro-
chemical Company at its Yokkaichi complex.  The boiler provides steam for
                                     11-75

-------
generating 40 MW of electricity and 47 MW equivalent of process steam.  Fuel
oil containing 3 to 4% sulfur is burned, producing a flue gas S02 concentra-
tion of 1000 to 2000 ppm.

     In addition to removal of SC>2, the System was also designed to remove
particulate matter (mostly carbon dust) from the boiler, thereby taking advan-
tage of the high particulate removal capability of the CT-121 Process and
eliminating the need for installing an ESP.  The CT-121 System described
herein, therefore, incorporates equipment such as a fly ash thickener and
filter press that ordinarily would not be required.

     The CT-121 System, including particulate removal equipment, was routinely
started up on May 11, 1982.  The System has met or exceeded all performance
guarantees and operation has been smooth and trouble free over a wide range of
boiler loads and inlet SO  concentrations.

                         GENERAL SYSTEM DESCRIPTION

     Table 1 is a summary of the major design conditions for the CT-121 FGD
and particulate removal system and Table 2 is an equipment list including size
and materials of construction.

                TABLE 1.  SUMMARY OF DESIGN CONDITIONS OF THE
                          MITSUBISHI PETROCHEMICAL CT-121 PLANT
  Flue Gas Source

  Flue Gas Flow Rate
  Flue Gas Temperature

  S02 Inlet Concentration
  Particulate Loading
  S02 Removal Efficiency
  Particulate Emission
  Reagent

  By-Product
  Total Plot Area*

  Schedule
     Start Construction
     Start Operation
280,000 kg/h (617,000 Ib/h)
oil-fired boiler (87 MW equivalent)
260,000 Nm3/h (162,000 scfm)
140-160°C (284-320°F)
1500-2000 ppm
200 mg/dNm3 (0.09 gr/dscf)
97%
< 50 mg/dNm3 (0.02 gr/dscf)
Powdered limestone
Saleable gypsum
1540 m2 (0.4 acres)

June 1981
May 1982
 Includes gypsum by-product storage, particulate removal equipment  and  120
 meter solid FRP wet stack.
                                     11-76

-------
TABLE 2.   EQUIPMENT LIST

Equipment
Type
Size
Material
Remarks
CT-121 FGD EQUIPMENT
Flue Gas Fan

Precooler
Jet Bubbling Reactor

Air Blower

Overflow Tank
Limestone Silo

Limestone Slurry Tank
Limestone Feeder

Mother Liquor Tank
Centrifuge

Mist Eliminator
Gas Cooling Pump
Gypsum Slurry Pump
Overflow Pump
Limestone Slurry Pump
Mother Liquor Pump

Fly Ash Thickener
Neutralization Tank
Filter Press
Filtrate Tank
Thickener Overflow Pump
Thickener Underflow Pump
Filter Feed Pump
Filtrate Pump
Blow Down Tank
Blow Down Pump
Turbo

Cylindrical
Cylindrical

Turbo

Cylindrical
Cylindrical

Cylindrical
Screw Feeder

Cylindrical
Screw Decanter

Chevron
Centrifugal
Centrifugal
Centrifugal
Centrifugal
Centrifugal
FLY
Cylindrical
Cylindrical
Filter Press
Cylindrical
Centrifugal
Centrifugal
Centrifugal
Centrifugal
Cylindrical
Centrifugal
260,000 Nm3/h
(162,000 scfm)

9.8 m0xlO mH
(32'0x33'H)
5,000 Nm3/h
(3100 scfm)

6.8 m0x!9.6 mH
(23'0x65'H)

2,300 kg/h
(5100 Ib/hr)

3 x 1,330 kg/h
(3000 Ib/hr)






ASH REMOVAL EQUIPMENT










c

c
.s .

.s .
316L


cast

c
c

c
c

c

.s.
.s.

. s .
.s .

.s.
316L

c
c
c
c
c
c

c
c
c
c
c
c
c
c
c
c

. s .
.s.
.s .
.s.
.s .
.s.

.s .
.s .
.s .
.s.
.s .
.s .
.s .
. s .
.s .
.s .


with
s . s.



F1RP lining
FRP




with Agitator

aluminum

with


with


with
s . s.

with
with
with
with
with
with

with
with
with
with
with
with
with
with
with
with

F1RP lining


F1RP lining


F1RP lining


F1RP lining, P
Rubber lining
Rubber lining
Rubber lining
Rubber lining
Rubber lining

F1RP lining
F1RP lining
Rubber lining,
F1RP lining
Rubber lining
Rubber lining
Rubber lining
Rubber lining
F1RP lining
Rubber lining

with Agitator


with Agitator





.P.






with Rake
with Agitator
316L s.s.








-------
     Figure 1 is a plot plan of the entire system showing the arrangement  of
major equipment including the stack.  Figure 2 is an overview of the  FGD
portion of the System.

     Flue gas to the System fluctuates daily between 50-100% of design
capacity due to changes in boiler load.  The S02 concentration also  fluctu-
ates widely due to the variety of fuel oils burned.

     The control of the power plant, particulate and SC>2 removal equipment is
centralized and computer controlled.  Both the boiler  and CT-121 System can
be started up, operated, and shutdown from the central control room.

     The heart of the CT-121 Process is the patented absorber which  is  called
a Jet Bubbling Reactor  (JBR).  This device, illustrated in  Figure  3,  allows
S02  to be absorbed, neutralized, oxidized, and crystallized in a single
vessel.  The overall reaction equation for the process is:

               S02 + CaC03 + %02 + 2H20 -*- CaSO^I^O + CC>2

                            PROCESS DESCRIPTION

     Figure  4  shows the process flow for the CT-121 FGD System including
particulate  removal equipment.  Flue gas from the air  preheater is pressur-
ized by  the  Flue Gas Fan and introduced into the Precooler.  The gas  is con-
tacted with  a  fine spray of recirculated water to humidify  and cool  the gas
to  its adiabatic saturation temperature.  The Precooler is  also designed  to
remove particulates and other impurities contained  in  the flue gas.   The
humidified gas flows to the JBR where it is bubbled into a  shallow layer  of
absorbent.   S02 is absorbed and precipitated as calcium sulfate  (gypsum)  by
the  addition of limestone slurry and oxidizing air.

     The desulfurized gas leaving the JBR flows through a two stage mist
eliminator to  remove carryover mist.  The mist eliminator is washed  once  per
day  with mother liquor  to remove accumulated fine gypsum crystals.

     The level in the JBR is held constant by maintaining a liquid flow over
a weir to  the  overflow  tank.  Overflow liquid is recycled to the JBR.

     Gypsum  slurry is withdrawn from the JBR and is mechanically dewatered
using a  solid  bowl decanter centrifuge.  The by-product gypsum is  essentially
a  dry material and is collected in a temporary storage area directly below
the  centrifuge.  Mother liquor from the centrifuge  flows to the mother  liquor
tank and recycled to  the System.

     Powdered  limestone from the silo is fed to the limestone slurry tank via
a  screw  feeder and mixed with mother liquor  to obtain  a 20  wt.%  slurry.  The
slurry is  pumped to the JBR on flow control based on a feedforward boiler
load signal  and a feedback JBR pH signal.

     A portion of the Precooler cooling water flows to a fly ash  thickener to
concentrate  captured particulates.  The overflow liquid is  returned  to  the
Precooler  and  the underflow is pumped to a neutralization tank. A  small amount
                                     11-78

-------
                                  60.5 m (199 FT.)
                          FROM BOILER
 1 JET BUBBLING REACTOR
 2 PRECOOLER
 3 FLUE GAS  FAN
 4 MIST ELIMINATOR
 5 LIMESTONE  SILO
 6 FILTRATE TANK
 7 FLYASH  THICKENER
 8 SLOWDOWN TANK
 9 CENTRIFUGE (SOLID BOWL DECANTER) AND
  GYPSUM STORAGE HOUSE
10 FILTER PRESS AND FLYASH STORAGE
11 STACK
12 BYPASS DUCT
13 FOUNDATION OF STACK SUPPORT
Figure 1.  ARRANGEMENT OF THE  CT-121  FGD  PLANT FOR MITSUBISHI PETROCHEMICAL CO.

-------
I
CD
o
                               Figure 2.  OVERVIEW OF THE MITSUBISHI PETROCHEMICAL CT-121 PLANT

-------
Flue Gas
Limestone
  Slurry
                                                 Gypsum
                                                  Slurry
              Figure 3.  JET BUBBLING REACTOR
                          11-81

-------
                          . BOILER PLANT! FQD PLANT
i
CO
ro
                                         STACK  PRECOOLER   MIST    JET BUBBLING  OVERFLOW
                                                        ELIMINATOR    REACTOR     TANK
                                                                    (JBR)
 LIMESTONE SILO   CENTRIFUGE  MOTHER LIQUOR
LIMESTONE SLURRY              TANK
     TANK
                     Figure  4.  PROCESS FLOW SHEET  OF THE  CT-121  FGD PLANT FOR MITSUBISHI  PETROCHEMICAL  CO.

-------
of mother liquor is also added to this tank.  The combined stream is neu-
tralized with limestone slurry.  The neutralized liquor, containing approxi-
mately 2 wt.% solids, is pumped to a filter press where the solids are
removed.  The filtrate is collected in a filtrate tank and recycled to the
system via the mother liquor tank.  A small amount of the filtrate (1 to
3 m3/hr (4 to 13 gpm)) is bled from the filtrate tank to maintain system
water balance.

                             OPERATING RESULTS

PLANT OPERATION

     The CT-121 System was smoothly and routinely started up on May 11, 1982.
Tables 3a and 3b are summaries of the operating history of both the boiler
and the CT-121 System.  As shown in Table 3b, System reliability has been
100%.  This is particularly impressive considering the System consists of only
a single scrubbing module and is operated under severe conditions.

DESULFURIZATION EFFICIENCY

     The S02 removal efficiency, as shown in Table 3a, has averaged 97 - 99%
for inlet S02 loadings ranging between 1000 - 2000 ppm.

     Figures 5 and 6 show the effect of boiler load and S02 inlet concentra-
tion changes on desulfurization efficiency.  These two Figures illustrate
the CT-121 Process' capability to maintain stable desulfurization under
widely fluctuating boiler loads and inlet S02 concentrations.

OPERABILITY AND CONTROLLABILITY

     The boiler load changes daily from 50 - 100% of design capacity.  The
System responds well and automatically to such load fluctuations.  Figure 7
illustrates the CT-121 System response to a typical change in load.

     The CT-121 System is extremely simple to operate and requires little
operator attention.  System operation is basically automatic,  being controlled
by a digital computer control system.

     The flow of flue gas to the System is set by boiler load.  Flow is auto-
matically adjusted by a pressure controller installed on the suction side of
the Flue Gas Fan which controls the position of the fan inlet vanes.  A by-
pass damper is installed to protect the boiler against sudden changes in
draft.

     The limestone slurry concentration is maintained constant by adding
limestone in proportion to mother liquor feed by using a proportional control
loop.  Limestone slurry is monitored using an in-line density meter.

     The gypsum concentration in the JBR is maintained at about 15 wt.% by
pumping a small slip stream of slurry to the centrifuge.  The flow rate is
controlled in proportion to the S02 load signal which is the product of the
boiler load signal and the S02 inlet concentration.
                                     11-83

-------
                                         TABLE 3a.   OPERATING HISTORY OF THE MITSUBISHI PETROCHEMICAL CT-121  FGD  PLANT
                                                                 Operating Conditions
                                             Flue gas source
                                             Boiler capacity
                                             Startup date
                                             Gas flow rate
                                             Gas temperature
                                             S02 concentration
                                             Desulfurization efficiency
                                  oil-fired boiler
                                  280 T/h steam (87 MW equivalent)
                                  May 11, 1982
                                  100,000-246,000 Nm3/h (62,000-153,000 scfm)
                                  140-160°C (284-320°F)
                                  700-2,300 ppm
                                  96-99%
 I
oo
-0-
                                                                 Operating Summary
                                                          1982
                                                                                                                  1983
MAY
l,500b
JUN
96
1,500
JUL
98
1,000
AUG
99
1,000
SEP
98.5
1,000
OCT
98.5
1,500
NOV
97.5
2,000
DEC
97.5
2,100
JAN
98
2,000
FEE
98
1,800
MAR
98
1,800
APR
97.5
1,600
MAY
98
2,000
JUN
98
2,000
                              Startup
                              5/11/82
Start of commercial operation
                                                                Cause  of  Shutdowns
                                              Operating
                                              Shutdown
                      ^Scheduled shutdown for boiler/turbine inspection
                     A*Scheduled shutdown for first annual inspection of steam
                       generating plant.
                            Average monthly S02 removal,  %
                            Average monthly inlet  S02  cone., ppm

-------
                           TABLE 3b.  CT-121 RELIABILITY PARAMETERS  (MAY 11, 1982 THROUGH JUNE 30,  1983)
                                      FOR THE MITSUBISHI PETROCHEMICAL FGD PLANT


                Parameter                          Value (%)            Total Time  (hours/hours)

                Reliability j*                       100.0                    9,249/9,249
                Availability                         98.1                    9,816/10,008
                Operabilityc                         99.9                    9,249/9,255
                Utilization Factor                   92.4                    9,249/10,008
!_,             	.	.	_____	.	.	,	,	,	.	
!_,             	,	.	.	,—
I
oo
                 Reliability - hours the FGD System was operated divided by the hours the FGD system
                               was called upon to operate.


                 Availability - hours the FGD System was available for operation (whether operated
                                or not), divided by the hours in the period.

                Q
                 Operability - hours the FGD System was operated divided by the boiler operating
                               hours in the period.


                 Utilization Factor - hours that the FGD System was operated divided by the hours
                                      in the period.

-------
   o
   c
   CD

   'o
   c
   o
   '•£
   03
   N
   3
   CO
100
            0
             20
40
60
80
100
                            Boiler  Load.  %
   Figure 5.   EFFECT OF BOILER LOAD ON DESULFURIZATION EFFICIENCY
^\
o
c
CD
O
LU
C
o
03
N
"5
w
CD
Q
100
99
98
97
96

- — ._
-
,
i —
^^^^

i
-
-

1000 pnnn
            S02 Concentration In The Flue  Gas, ppm



Figure 6.  EFFECT OF S02  CONCENTRATION ON DESULFURIZATION EFFICIENCY
                           11-86

-------
CO
o
—
"o
CD
     100
      50
       0
a   1600
a

 "cvj  1500
O
w   1400
                 0
10
20
                 0
10
20
                 0
10
20
                 0
30
40
30
40
E 98
01 97
C\J
O
CO








/ 	
^



^"X,




v/





30
40
3.7
a
£T 3.5
m
^ 3.3








t
r
\j


\
V


r



\


10      20      30      40

   Elapsed Time, Min.
     Figure 7.  TYPICAL SYSTEM RESPONSE TO CHANGE  IN BOILER LOAD
                           11-87

-------
     The pH in the JBR is controlled by the addition of limestone.   The flow
of limestone slurry to the JBR is controlled by the feedforward signal of the
S02 load, compensated by the feedback signal of the JBR pH.   This control loop
results in only a ± 0.2% perturbation in JBR pH for maximum load changes.
POWER CONSUMPTION
     The power consumption is 1,100 kW for operation at the following condi-
tions:
            Boiler Load                  100%  (87 MW equivalent)
            Inlet S02                    2,000 ppm
            Desulfurization Efficiency   98%
     Power consumption is less than 1.3% of the boiler equivalent generated
MW.  This  consumption includes particulate removal as well as  SOz removal.
CT-121's  low power consumption is due to the elimination of slurry recycle
pumps  characteristic of  first generation limestone  systems.
LIMESTONE UTILIZATION
      Limestone utilization has averaged greater than 99%.  Table 4 shows a
typical analysis  of  the  limestone used.
                     TABLE 4.  TYPICAL LIMESTONE ANALYSIS
                Chemical  Composition
                   CaCO                             97.98 wt.%
                   CaO                              54.87 wt.%
                   CO                               43.80 wt.%
                   SO                                0.050 wt.%
                   MgO                               0.549 wt.%
                   Al 0                              0.059 wt.%
                   Fe203                             0.023 wt.%
                   SiO  + Insol.                     0.86 wt.%
                Particle  Size
                   325 mesh                         90% pass
 BY-PRODUCT GYPSUM
      A typical analysis of  the  gypsum by-product  is shown in Table 5.   The
 gypsum is sold to cement and wallboard manufacturers.
                                     11-J

-------
                TABLE 5.   TYPICAL GYPSUM BY-PRODUCT ANALYSIS

                CaS04«2H20                     99.2 wt.%
                S03                            46.1 wt.%
                CaCO                           0.1-0.6  wt.%
                Combined  water                 20.6 wt.%
                Moisture                        11-12 wt.%
                pH                             6.5-7.5
                Crystal size                   49  microns
                  (average)
WASTEWATER

     A typical analysis of the wastewater,  after neutralization with lime-
stone, is shown in Table 6.   The water is discharged to the ocean.

                   TABLE 6.   TYPICAL WASTEWATER PROPERTIES
                   pH                  6.0
                   Suspended solid     0.5 ppm
                   COD                 less than 7 ppm
OPERATING AND MAINTENANCE REQUIREMENTS

     Operating and maintenance functions for the CT-121 System have been
minimal and are completely absorbed by normal boiler plant operations.   No
scaling, plugging, or other maintenance intensive problems have been encoun-
tered.   The total average personnel requirement, including maintenance, has
averaged less than one-half man per shift.

                   RESULTS OF THE FIRST ANNUAL INSPECTION

     The steam generating plant was shutdown March 1983 for the first sche-
duled annual inspection.   The CT-121 System was also inspected at this time.

     Inspection of the CT-121 System, including the JBR, revealed no scaling,
corrosion, or other problems.  As such, maintenance was limited to routine
servicing of equipment and flushing off some soft deposits of carbon dust and
gypsum crystals that had  accumulated on the upper and lower decks of the JBR.
                                     11-89

-------
                                   SUMMARY

     The CT-121 System has met or exceeded all performance guarantees and
expectations.  The System has routinely achieved greater than 99% desulfuri-
zation using essentially stoichiometric amounts of limestone.  System relia-
bility has been 100%.

     The CT-121 System has also proven to be an exceedingly simple process to
operate and maintain.   Maintenance has been minimal and limited to routine
servicing of equipment.  Total operating and maintenance manpower requirements
have averaged less than one-half man per shift.
                                    11-90

-------
OPERATION EXPERIENCE WITH FGD PLANT II AT
 WILHELMSHAVEN POWER PLANT,  WEST GERMANY

B. Stellbrink, H. Weissert,  P. Kutemeyer

-------
by:
        OPERATION EXPERIENCE WITH FGD PLANT II AT

         WILHELMSHAVEN POWER PLANT, WEST GERMANY
                       (Figure 1)
B. Stellbrink
General Manager
Power Plant Wilhelmshaven
West Germany
H. Weissert
Director of Research
Bischoff GmbH
Essen, West Germany
                                                      & D evelopment
     P. Kutemeyer
     General Manager
     Bischoff Environmental Systems
     Pittsburgh, PA
           Figure 1.  Wilhelmshaven Power Plant
                           11-91

-------
                                  ABSTRACT
     The Nordwestdeutsche Kraftwerke Aktiengesellschaft (NWK) has been
operating a  720 MW coal-fired power plant in Wilhelmshaven, West Germany,
since 1976.  In March, 1982, the second flue gas desulfurization plant, called
REA 2, the German acronym for flue gas desulfurization plant 2, was put into
operation after a two-year construction period.  Startup and checkout proceed-
ed   without major difficulties  and was completed within approximately
three months.  Beginning in June, 1982, NWK accepted the plant.  Since then
REA 2 has been in operation without significant interruptions, effectively
reducing SO  emissions of the power plant,
(Figure 2)
                  Figure 2.   FGD  Plants 1  &  2  at  Wilhelmshaven
                                     11-92

-------
     FGD treatment at Wilhelmshaven  is  accomplished by two plants.  REA 1 is
a co-current scrubber in operation since 1977.   Together REA 1 and 2 treat
about 80 percent of the flue gas  generated.   The FGD  plants are located in
the by-pass to the main flue gas  duct behind  the electrofilters, and are equ-
ipped with their own blowers.  REA 1 has a throughput of  500,000 nm3/h
(294,287 SCFM) and REA 2; 1,500,000 nm3/h (882,862 SCFM) .  (Figure  3)

     Instrumentation for measuring dust content, SO  and NO  concentrations,
and exhaust temperature of the cleaned  flue gas is located at the 27 m (88
ft.) platform elevation of the 275 m (902  ft.)  high exhaust stack.
         Steam Generator
ESP
    REA 1
V =500,000 nnWh
 (294,287 SCFM)
     REA 2
V =1,500,000 nrrvVh
  (882,862 SCFM)
                                                                  Stack
                                                       V = 2,500,000 nrrWh
                                                         (1,471,436 SCFM)
                                                        tg =100°C (212°F)

               Figure 3.   Present Flue Gas Flow at Wilhelmshaven
     Both REA 1 and 2 use a Bischoff wet absorption process.  REA 2, however,
 is a counter-cur rent  scrubber  with regenerative re-heating of the flue  gas.
 A portion of the  flue gas generated by the power plant is injected into  the
 REA 2 by means of an  axial  blower having variable vanes.  From the blower,
 the flue gas enters the  hot side  of regenerative heat exchanger  (REX),  and
 from there enters from the side  into the counter-cur rent scrubber.  In  the
 scrubber, flue gas passes upward  through various levels of lime  slurry  sprays
 and is cooled and desulfurized.   At the top of the scrubber tower the flue
                                     11-93

-------
gas passes through a mist separator, reverses direction,  and enters a venturi
mixer/preheater in which the remaining water droplets in  the flue gas are
evaporated by mixing hot, untreated gas with the desulfurized flue gas.  From
the venturi mixer, the flue gas enters the cold side of the REX and is reheat-
ed to a temperature of 100°C (212°F) .  From there the flue gas exits through
the exhaust stack.  (Figure  4)
                                Scrubber
                                                            Residue Sludge
                                        Water
                            Flow Diagram REA 2

         Figure 4.  Schematic of 415 MW Single-Module Bischoff System



 OPERATING EXPERIENCES, PROBLEMS, AND SOLUTIONS

      Operation and control functions for the two FGD  plants are effected
 from the power plant control room.   To date, no power  plant interruptions
 have been caused by either REA 1 or REA 2.  Both FGD  plants can be shut down
 or re-started on short notice, and desulfurization occurs immediately after
 start-up of the blowers.  In the event longer shut-down periods are antici-
 pated, such as when a low sulfur fuel is used, for example, REA 2 can be
 separated from the main flue gas ducts by means of double valves, and pre-
 served by blowing dried air into the system through both the raw and clean gas
 ducts.
                                     11-94

-------
SCRUBBER

     The scrubber exhibits good operating characteristics and quickly adjusts
to changes in the flue gas volume, maintaining a high degree of desulfuriza-
tion.  Despite a high concentration of solids in the scrubber slurry,  (100
g/1) , slurry liquid circulation can be readily re-started, even after being
shut down over the weekend.  The system does not require the use of recircu—
lating pumps or agitators during the time it is shut down.  During shut-down,
about 150 tons of gypsum settle out in slurry lines, pumps and the sump.  At
start-up, the slurry pumps re-suspend the gypsum by backflushing for five
minutes using a specially-designed system of flushing lines, after which the
pumps are switched to the spray nozzles.  The flushing process operates auto-
matically.

     Due to settled gypsum behind the valves, difficulties  were experienced
initially, in that torque overload sensors shut the automatic flushing pro-
cess down when higher than permitted torques where required to open the valvas.
Increasing the size of the valve drive shafts and motors solved this problem.

     Lime is added below the oxidation zone into the scrubber sump.  The
amount of lime added is controlled by means of an rpm-regulated pump. Ini-
tially, a small amount of scrubber slurry was used as the medium to add the
lime.  After about 2,000 hours of operation, however, the supply linesfor
adding the lime clogged with gypsum.  Since then, lime is added without the
use  of slurry fluid.  The 1,000-hour operating life of the pump stators is
not considered acceptable.  It is hoped that this problem can be solved by
installation of a bypass line directly to the scrubber sump and thus avoid-
ing the upper rpm range of the pumps, which seems to be the cause of the
problem.

     During the acceptance test period, a large amount of water was found
after the mist separator.  The reasons were improperly installed separator
segments and reintrainment of water due to open collection troughs.  The
former problem was corrected; the later was solved by  installing covers
over the collection troughs.  Makeup water for the scrubber is added by using
it to flush the mist separator.

     To date, no scaling has been observed in the slurry systems, the scrub-
ber, or mist separator.  Nozzles have shown no wear after 4,000 hours of
operation.

LIMESTONE

     At the present time there are several FGD plants under construction in
the Federal Republic of Germany (FRG) , which, due to their proximity to lime-
stone quarries, intend to use limestone instead of lime.  In order to deter-
mine the effects in an operating plant, REA 2 was tested using limestone.

     The results were as follows:

     The use of limestone is possible; however, a lower cegree of desulfuriz-
ation is achieved than with lime.   This is a result of the slower reaction
                                     11-95

-------
rate of limestone.  In order to achieve the same degree of desulfurization  a
considerably larger quantity of scrubbing slurry must be used.  This, of
course, means a higher energy requirement and greater wear of the equipment.
Also, the quantity of limestone required is 1.8 times greater than that of
lime due to the lower amount of CaO in limestone.  However, byproduct gypsum
can also be produced without difficulty.

     For the Wilhelmshaven power plant, the use of limestone  is not  an econ-
omically viable alternative, due to the high power requirements and  the
greater transportation costs.

FLUE GAS DUCTING

     The flue gas  ducting is divided into sections.  Each  section is free to
expand or contract.  Corrosion-resistant fabric expansion  joints are used
between the  sections.  During startup and testing, flue gas escaped  through
leaky bolt connections on the expansion joints and caused  damage to  the insul-
ation.  The  bolted connections were improved.  Since, in an FGD plant using
regenerative reheat, sections of the flue gas ducting may  be  operated at
temperatures below the dew point of the acids in gas, special precautions
must be taken with respect to corrosion protection of those sections. Affect-
ed  are especially  the REX, the flue gas ducts between the  REX and the scrub-
ber , and the venturi mixer/dryer.  Also, the flue gas ducts and especially
the "floors  " must be even with a slight slope and have a  gutter to  allow
condensates, which usually have a pH of less than 1, to be drained off.  The
flue aas ducts between the REX and the scrubber should be  sloped toward the
scrubber to  allow  the condensate to drain into the scrubber.  Other  conden-
sates  flow over acid-resistant gutters to a rubber-lined collection basin
from which they are pumped to the scrubber.  This solution prevents  an efflu-
ent problem.  In  case the REX must be flushed, the arrangement also  allows the
drainage of  the flushing water to the scrubber sump.

     Coating of the various  components of the FGD plant presented a  problem,
since no experience with coating materials was available involving applica-
tions of highly concentrated sulphuric acid at temperatures greater  than
80  C  (176 F) .  Various specialty firms in this area were consulted.  Only a
few were willing  to guarantee their coating for a minimum  of  two years.
Several different  materials were selected and applied to different sections
of  components of  the system.  After 4,000 operating hours, no final  evalu-
ation of the various coating materials is possible.  Results, however, appear
encouraging.  It has been determined that all coatings used are satisfactory
on  vertical  walls  and on the "roof" of the flue gas ducting.

     The following results are noteworthy:

     The flue gas  ducting should ba fabricated and assembled as free  of all
stress as possible and should remain stress-free during operations.  This is
important because  movements of the walls will cause the coating to crack or
flake off, resulting in corrosion.  The flue gas ducting should be installed
such as to prevent dents, and should have sufficient slope to prevent acid
puddles from forming during operations and unnecessarily stressing the ccatLng.
                                     11-96

-------
     The  coating material  must remain flexible, even after long periods of
use.  It  is  also necessary to apply the coating a sufficient  number of times,
as well as  during warm weather.   (Figure  5)
                                   Scrubber
               Clean Gas
Rubberized
Kreiwatherm
Keraflake
Plastics
Enameled
                                                Scrubbing Slurry
                                                A Test Coating 3,398 hrs
                                                Colebrand
                                                CXL 2000 Coating
                                                S =1 mm/10 Coats
                                                Clouth
                                                Rubber Application Durabilit 1565
                                                S = 1.6mm
                                                Ceilcote
                                                871 x Lining (Viton)
                                                S =1.4mm/4 Coats
                                                Ceilcote
                                                Flakeline 282
                                                S =1.6mm/4 Coats
                                  Materials

                          Figure 5.  View of FGD Plant


INSULATION

     The  insulation must  be thick enough  and  must be carefully installed.
Improper  insul ation will  result in cold spots, and thus, condensation.  This
in turn will result in corrosion and scaling, which can result in the dis-
charge of solid particles  from the stack.   Special care must  be taken in
insulating support structures.   It is important that in the initial request
for bids,  the flue gas conditions and the  demands placed on the insulation
be clearly spelled out.
                                      11-97

-------
REGENERATIVE HEAT EXCHANGER (REX)
                                  TEMPERATUR 1°C
             Figure 6.   Temperature Profile of REX for Cold Start
      The  concept of reheating the  desulfurized flue  gas has  been   viable
 todate.       Startup of the REX, from a cold  condition, has  not  presented
 any  problems.   Gas discharge temperatures, measured  in the stack,  always
 exceed 100  C  (212 F) ,  even during  periods  of  partial loads.   The heat-stor-
 age  chambers  are made  of plastic material  on  the  cold side and of enamel-
 coated steel  on the hot side.  On  the cold  side,  deposits  form,  caused by
 water droplets  entrained in the clean gas.  These evaporate  at the inlet of
 the  REX.  On  the hot side, deposits form due  to entrained  dust particles
 containing  acidic deposits in the  raw gas.  (Figure  6)

      In case  the electrostatic precipitator does  not function properly, the
 resulting increase in  ash content  of the raw  gas  quickly  increases the pres-
 sure drop across the REX.

      The  hot  surfaces are cleaned by means of pressurized air,  which is
 blown over  the  surfaces by special nozzles.   This is performed after every
 12 hours  of operation.  During normal operations, i.e. without malfunctions
 in the ESP  or the  scrubber, this  cleaning interval has been  found to be
                                    11-98

-------
sufficient.  Should larger deposits form more  quickly due  to malfunctions in
either the EXP or the scrubber, cleaning by means  of high  pressure water is
required.  The same nozzles used for the high  pressure  air are also used for
the high pressure water spray.  For this purpose,  they  are switched to a
high pressure water pump.  On the hot side, the  water pressure before the
nozzles is 80 bar (1160 psi) and 40 bar  (580 psi)  on the cold side.  The REX
continues to turn during the cleaning process  with a reduced velocity.  The
amount of water used is held to a minimum.   This water  is added to the scrub-
ber as make-up water.  The cleaning method  has been used during operation of
the REA 2, as well as during shutdowns.
         Pressure Loss of REX During Operating  Period
                  Figure 7.   Pressure Drop Diagram For REX
     Measurements to determine  pressure losses across the REX from start-up
until the 29th of March,  1983,  show  that the pressure loss increase has been
smaller than anticipated.   The  deposits on the REX have been maintained in
acceptable limits during  the  first 4,000 hours of operation by cleaning regu-
larly with high pressure  air  or water.  The installed reserve capabilities
of the blower, layed out  for  a  pressure drop increase up from 10 mbar (0.145
psi) to 13 mbar (0.216  psi),  has been sufficient.  (Figure 7)
                                   11-99

-------
     Tests will soon be conducted to determine whether deposits on the hot
side can be effectively removed using steam instead of water.   To date, no
noticeable corrosion has been determined in the REX.

     No definite statement can be made regarding  the  lifetime  of the heat-
storage chambers after this short period of operations.

PRESSURE LOSSES

     Measurements of pressure losses across various components of the REA 2
system at start-up showed that the pressure drops were  below those given by
the various suppliers.  Pressure losses across  the scrubber and the flue gas
ducts do not change with time.

ENERGY REQUIREMENTS

     Testshave  shown that in those cases when the SC>2 content of the f^Lue gas
is  in the lower end of the operational  spectrum (S02  le^ss than 2 g/mn );
 (0.8 gr/scf) or flue gas volume less than  1,000,000 nm /h (588,574 scfm) ,
the desulfurization  efficiency does not increase  significantly if two slurry
pumps are operated.  Using only one slurry pump,  the  desulfurization effici-
ency will only  decrease by 2 percent; however the power  saved is 1.2 MW.
(Figure 8)
                                Operation with Two Pumps
                                   SO?>22g/nm3
                                       Operation with One Pump
                                         SO_, ' 2 2g/nm3
                400
                       600     800     1000    1200
                          Flue Gas Volume 103 nnWhr (589 SCFM)
                                                 1400
                                                        1600
                                                               1800
                   Energy Requirements of REA 2
                   Figure  8.   Energy  Requirements  for REA 2
                                     11-100

-------
INSTRUMENTATION AND CONTROLS

Control

     For controlling the REA 2, it was decided for the first time in this
power plant to use a programmable controller.  The controller used is of the
Siemens type 5 and is used in conjunction with a more conventional control
system, ISKAMATIC B.  The S 5 system is used to control the functional groups,
including the safety interlocking systems.  The great advantage offered by
the programmable S 5 system is the ease with which the automatic controls can
be adjusted to the operational process.  This was shown to be of great advan-
tage during the start-up phase of the operation.

     The following group controls are used:

        - flue gas blowers with oil supply and valves

        - slurry pumps with flushing program

        - flushing of mist separator

        - lime addition

     Modifications are presently under way to upgrade the automatic control
system to also handle clarifying and gypsum dewatering.  It is intended to
control the entire FGD plant from the power plant control room and includes
starting the plant and shutting it down.

     Performing the various controlling functions is accomplished with the
process control system TELEPERM C.  Control  is accomplished using the fol-
lowing control loops:

        - raw gas volume

        - drying of the clean  gas

        - controlling amount of lime added

        - solids removal, slurry discharge

        - control of various fluid levels

     Since the REA 2 is a counter-cur rent  scrubber, it  is,  theoretically  at
least, possible to adjust the  pH value at  the slurry discharge to such a
value as to optimize the oxidation of  the  sulphite to  gypsum, without adver-
sely affecting the desulfurization.

     This results in the following  very important advantages.   First of  all,
the lime usage is stoichionB trie, and readily  adjustable  as operating para-
meters change.  Second, it is  possible to  integrate the oxidation stage  in
the scrubbing tower.  This,  of course, does increase the demands placed on
the instrumentation and control system, since, unlike the co-current system,
                                    11-101

-------
the scrubbing fluid is practically unbuffered.

     The pH control loop was modified during  start-up.   Instead of  continu-
ously changing lime dosage as a function of pH,  lime  dosage  is now  adjusted
in discreet  steps.  However, more recent tests  indicated  that, by  also
measuring density, a more accurate mass  flow  rate of  the lime to  be added
is now possible.  Originally, a rough dosage  and  a fine  dosage were used
for the injection of lime required.  The fine dosage  was found to be not
required and was disassembled.
                                        T      OUS  BKB •   Km
                                        A.     mm  •cJiMEK
                                        VJ1     hi' mUHW
              RAW GAS VOLUME
                                        LIME ADDITION
                    Control Diagram of REA 2

                        Figure 9.  Control Diagram


INSTRUMENTATION

    In order to monitor and  automate  the operation  of the FGD plant, as well
as to optimize the dosage  of additives  and the desulfurization, reliable and
maintenance-free instruments must  be  used.  This  is especially true for
determining the amounts of additives  supplied to  the system by measuring the
solids content in the  slurry and the  amount of lime added, the pH value of
the wash fluid and the S02 content  in the raw and clean gas.
                                    11-102

-------
     The following  systems were  selected:

     - solids and density measurements:   radiometric  analysis using  the
                                          absorption principle

     — SO  concentration measurement      spectral photometer using the
                                          absorption principle

     - pH value measurement               electrometric analysis, potentio-
                                          meter

     - flow measurements                  induction measurement

     - scrubber fluid measurement         ultrasonic measurement

     - flue gas volume measurement        pito tube (annular tube)

     Most of these  instruments have proven to be very reliable. Determina-
 tion of SOj concentration using  the absorption principle causes some problems,
 however.    (Figure 9)

 EFFICIENCY OF S02 REMOVAL

     The degree of  SO  removal in the  scrubber is influenced by various
 factors.  These are: the relationship  between scrubbing fluid voJnroe  and gas
 volume;gas velocity; pH value; and SO  concentration  of the raw gas.  During
 acceptance tests, it was found that 95 percent of the SO  in the gas was
 scrubbed out. Under operational  conditions, with S0?  concentrations  of be-
 tween 1 g/nm3 and 3 g/nm3 (0.4 gr/scf  and 1. gr/scf), 91 percent to 95 percent
 of the SO. was removed.  Because the regenerative gas reheater was selected
 in the REA 2, the overall SO  removal  efficiency is lower than in the scrub-
 ber, due to the fact that raw gas is added to the clean gas after the mist
 separator to evaporate any water droplets still in the gas, and also due to
 the fact that some raw gas leakage occurs in the regenerative gas reheater.

     Drying of the clean gas, using raw  gas, was selected to prevent
 deposits from forming on the regenerative gas preheater.

     To evaporate the rest  moisture   in  the clean gas, the clean gas temp-
 erature must be raised.   The rate of  evaporation, due to the increase in
 temperature, is determined by the available transit time of the gas  through
 the dryer.   It must also be considered that the salt  content of the  evapo-
 rated moisture will increase the salt  concentration,  resulting in decrease
 of the water vapor pressure and  thus increasing the dew point.

     The amount of increase in the dew point depends  on the type of  salt
 desolved.   Table salt, magnesium chloride and calcium chloride result in
 dew point increases of about 5°C (9°F),  30°C (54°F),  and 45°C  (81°F) resp-
 ectively, when present in saturated solutions,  Drying, by using raw gas,
 is thus only viable when readily soluble solutions such as magnesium of
calcium chloride are not present.  At  Wilhelmshaven,  where the REA 2  is
operated with sea water, this does not appear to be a problem.
                                    11-103

-------
      A delta  T of 8 C (14.A F)  is  achieved in the  gas preheater, resulting
 in a decrease in the SO  removal efficiency of about  10 percent.  This dis-
 advantage of  clean gas drying can  only be avoided  if  a highly efficient mist
 separator can be used, decreasing  the need for such a dryer by removing a
 sufficient  amount of the residual moisture  and salt.
                                         A T (Dryer)
                                                        Decrease of Scrubber
                                                         Desulphurization
                                                            Efficiency
                              Scrubber
               Stack
                                            Power Plant Wilhelmshaven
                                         8°C (46.4°F)
                                         2°C (35.6°F)
                                         0°C (32°F)
                                                            12.5%
                                                             5 %
                                                             2.5%
             Recirculatmg 1  M^L     ^1
             Blower    V  ^^P     •
                  , '—I  6     w
                                         Drying with Recirculating Blower
                                           2-5 °C
                                         (35.6-41 °F)
2.5%
                                           Blower Using Hot Clean Gas
                                           2-5°C
                                         (35.6-41 °F)
                                                             2.5%
                                           Blower Using Cold Raw Gas
                                                             1%
                                          Blower Using Cold Clean Gas
                                        ~3°C (37.4°F)
                                        Due to Blower
                                                             1%
         Figure 10.  Total Desulfurization Efficiency  of REA 2 with REX


      By  decreasing the delta T  from 8°C (14.4°F) to 5°C  (9°F) , no effects on
the REX  have been determined.  During the winter of 1982,  a delta T  of 2  C
 (3.6  F)  was  measured across the  gas dryer.   In this case,  deposits began  to
build up on  the surfaces of the  cold side of the REX.  These deposits could
not be removed by means of high  pressure air;   they were,  however, removed
with  high pressure water.
            a  2  C  (3.6 F) delta T  in  the dryer is possible if cleaning of  the
                                   „„—4.-T--I -   Increased mist separator effi-
      Thus,              ,    .	  „. _.._ ^ _y
REA  with high pressure water is acceptable.  a_iiuj.fc:asea mist separator
ciencies would increase operating  times between flushing considerably.

      Without  the use of the venturi dryer,  no noticeable difference  in  the
increase in pressure drop across  the REX resulted, as  compared to operation
with  the dryer.   These tests encompassed a  time period of two months of
                                     11-104

-------
effective operation.  In this case, „.._,, „„...„„.  „..-.  „.„„.
intervals  was employed.  Tests are presently  continuing.
i,  only  normal  air cleaning at 12-hour
' /"I  1-V -V- /-I O i-lfl f-T IT /-«|—tT-»f-T1-IT1-1T^ r>
     It is not possible, after the relatively  short  time  since  this system
went into operation, to determine if flushing  with high pressure  water will
cause damage to the REX due to increased corrosion or loosening of  the
regenerative heater packages.

     By reducing the delta T from 8°C  (14.4°F)  to  5°C  (9°F) ,  SO  removal
efficiencies increase from 82  percent  to  89 percent.  By not using the
dryer, the efficiency could be raised  to 92 percent.   If  this is  not possi-
ble the gas would have to be dried using the blower  on the  clean  gas side.
This would not affect the efficiency of the scrubber.

     In such a case, gas leakage in the REX does not decrease the S0? removal
efficiency of the system significantly.  The resulting increase in  the clean
gas temperature caused by the blower minimizes the need  for  a  gas  dryer.
Such a solution is presently being contemplated.  (Figure  10)
              Figure 11.   Magnified  Photograph of the Crystals
                                    11-105

-------
GYPSUM PREPARATION

     The byproduct generated by the Bischoff FGD plant  is high  quality
gypsum containing a high proportion of large gypsum crystals.   Because of
this, NWK and a wallboard manufacturer have signed a contract to  utilize
the FGD byproduct gypsum for wallboard production.  The contract  specifies
the quality of the gypsum.   To meet these specifications, the byproduct
gypsum must be washed and dewatered.  (Figure 11)

     In purchasing the equipment necessary to wash and dewater  the  gypsum,
cost was an important factor, since the profit resulting from the sale of
the gypsum should result in a reasonable payback period.  The equipment
presently being installed or planned is complicated and expensive.   In
general, such a system consists of various intermittently-operating cent-
rifuges, or continously-operating vacuum filters with an additional system
to transport the gypsum to a silo or other storage facility without addi-
tional handling.

     To find a suitable system, extensive tests were conducted  with the fol-
lowing systems:

        - vacuum-band filter

        - vacuum-drum filter

        - two-stage centrifuge

        - four-stage centrifuge

        - screw centrifuge

     All of these systems are, under certain conditions, suitable for use in
washing and dewatering byproduct gypsum.   The following residual moisture   con-
tents are measured.

          vacuum-filter 8 - 10 % (without free water)

          centrifuge    6-8%   (without free water)
                                    11-106

-------
       Transmission
            Oil
          Reservoir
      Rubber Mounts
Fill Tube

Screw Drum

Screw

Wash Tubes
                                                           Filter
                                                           Discharge
                                                           Solids
                                                           Discharge
                            Screw Centrifuge

                        Figure   12.   Screw Centrifuge
     At Wilhelmshaven the screw centrifuge  was selected due  to  the fact that
it was less expensive, required less  space  and energy, could be automated
without much difficulty, and finally,  could be expected to have low-wear
characteristics.   (Figure 12)
                                     11-107

-------
                     / / / / /////////////// // 7
                  Figure 13.   Gypsum Silo with Centrifuge

     The centrifuge is installed  over  a silo,  which has a capacity of
2%   days of gypsum production.   Centrifuge,  silo,  and loading facility are
one structure.   Loading is accomplished by means of a screw conveyor which
is activated by the truck driver.   Before the gypsum slurry reaches the screw
centrifuge, fine particles are removed by a hydrocyclone.  A portion of the
separated material  flows to a thickener.   The main stream is returned to the
scrubber.  The gypsum preparation plant is started  up, or shut down, depend-
ing on the measured solid content in the scrubber  slurry.  Boundary values
are 60 to 120 g/1.   (Figure 13)
                                   11-108

-------
                            Mist
                            Separator
                          Cyclone Water Discharge
                                                                 Thickener
              Figure  14.  Schematic of Gypsum Preparation Plant
     The gypsum generated in 24 hours of desulfurization can be withdrawn
from the scrubber within eight hours.  Gypsum is withdrawn when the solids
content in the slurry reaches 123 g/1.  Withdrawal ends when the solids
content reaches 60 g/1.  To insure that only large gypsum crystals are re-
moved,  the centrifuge filters out undersized crystals and returns them to
the scrubber.   In the time required for the solids content to again reach
120 g/1, the gypsum crystals have ample time to grow.  Thus, an adequate
supply of large crystals is  always assured.  (Figure 14)

     The cost for the REA 2, including the gypsum preparation plant, was
about $25,000,000,00.  The various components, such  as washer, REX, blowers,
gas ducting, instrumentation and controls, as well as installation, were
selected and ordered by the power plant.  Engineering of the plant was
accomplished by an engineering company.
                                     11-109

-------
 SUMMARY

      The availability of the REA 2 meets our expectations.  Scaling the
 REA 2 from REA 1,  which is 1/3 as large, presented no problems.  As a result,
 a F(D plant capable of serving an 800 MW power block was developed.

      Regenerative  reheat has been proven to be possible, and it saves
 energy.   Definitive conclusions regarding corrosion can not as yet be made.
 REFERENCES

 W.  Bosselmann,  H. Weiler
 Erste Betriebserfhrunge mit der Rauchgasentschwefelungsanlage
 im Kraftwerk Wilhelmshaven
 (VGB 1978)

 W.  Bosselmann,  K-R.  Hegemann,  J. Leimkuhler
 Weitere Betriebserfahrungen mit einer Rauchgasentschwefelungsanlage
 in Wilhelmshaven und deren Obertragung auf die Anlagenerweiterung
 (VGB 1981)

 M.  Frauenfeld
Ljunstrom - Gasvorwarmer zur Wiederaufheizung nassentschwefelter
 Reingase
 Technische Losungsmoglichkeiten zur  Verminderung von Leckage,
 Verschmutzung und Korrosion
 (1982)

 J.  Leimkuhler,  H. Weissert
 Erst Betriebserfahrung mit der Rauchgasentschwefelungsanlage 2
 im Kraftwerk Wilhelmshaven und Entsorgung der REA-Ruckstandsprodukte
 (1982)
                                    11-110

-------
  THE SULF-X PROCESS




E. Shapiro, W.  Ellison

-------
                              THE 5ULF-X PROCESS

                                        By:

                                Edward Shapiro, PE
                       Pittsburgh Environmental Systems, Inc.
                           Pittsburgh, Pennsylvania  15236

                                William Ellison, PE
                                Ellison Consultants
                             Monrovia, Maryland  21770

                                    ABSTRACT

     The  purpose of this paper is  to report the status of development of SULF-X
technology for flue gas  desulfurization and to review the design and applicability of the
Process in boiler service  firing  diverse  coal  fuels.   Details are given  of  process
chemistry, design  and operation, economics  and past and future system demonstration
programs.  The flexibility and attractiveness of the technology is shown in  its use either
for SO?-removal-only or for simultaneous SO9/NO  removal.
      ^                                   £    X


                                 INTRODUCTION

     The  SULF-X   Process  is  a wet absorption process  that  utilizes a slurry  of
regenerated ferrous sulfide solids to  achieve  removal of 90 to 99% of sulfur dioxide
from boiler flue gases by wet scrubbing. It is technically feasible for use with all fossil-
fuel  types.  A coal-fired calciner regenerates both spent and oxidized  forms  of  the
ferrous sulfide  reagent,  thereby  converting  collected  sulfur  dioxide  into  salable
elemental sulfur while minimizing formation of non-regenerable waste solids and liquids
requiring disposal.

     Current technical  evaluation and economic analysis of the SULF-X technology by
the Electric Power Research Institute shows that it is substantially less costly/than the
familiar  Wellman-Lord  regenerative flue  gas desulfurization (FGD) process.      The
study also indicates that it may be  more attractive than  limestone-based throwaway
solid-waste type  systems, particularly at  sites  where  waste  management options  are
limited.  The process is primarily applicable in  high sulfur coal-burning regions where
marketing of a commerical grade of by-product elemental sulfur is feasible and where
industrial  by-product  iron compounds such as pyrites (ferrous disulfide) are available for
use as  process  reagent  makeup supply.  With  the benefit of revenues from sulfur by-
product sales of $150 or  more per short ton and availability of inexpensive pyrites or
copperas (ferrous sulfate  heptahydrate), SULF-X is highly competitive with limestone-
consuming FGD systems in common  use.  Pyrites may be derived  from coal cleaning or
from mining  of non-ferrous metals.   Copperas  is commonly available from industrial
waste processing including waste pickling liquor neutralization.
                                        11-111

-------
                              PROCESS CHEMISTRY

     The reaction  steps in the SULF-X Process are schematically  illustrated  in the
simplified block diagram (Figure 1).

WET SCRUBBING

SO., Removal

     An  aqueous   slurry  containing  a  complex mixture of  iron/sulfur compounds,
including finely divided ferrous sulfide (FeS) reagent in stoichiometric excess, is used to
absorb sulfur dioxide in a wet scrubber. Although the solubility of FeS in water  is only 6
mg/liter (at 68°F),  it reacts with  absorbed  SO2 and oxygen converting a portion of the
SO7 to elemental sulfur while being sulfidated and, in part, degraded  to soluble ferrous
sulfate.   Additionally, while  the individual absorption  reactions are  numerous and
complex,  the overall  chemical reaction in  absorption  of  SO2  may  be expressed as
follows:

             (c) FeS +  S02  + 02^ (c-1) FeS(m) + Fe++ + SO4= + (1 -fy 5°

      The quantity  "n"  can  range  from 1 to 10 depending on process conditions.  The
quantity "c" is greater than  1 but less than 2, its  magnitude  varying  depending on
equilibrium  among tne diverse cnemical reactions taking place.  In a typical design, rate
of FeS  feed to the aDsorption  reaction is based  on the assumption that two moles are
required per mole of SO? absorbed (c  = 2).  The quantity "m",  which ranges between 1
and 2, is a measure of th~e  degree  of sulfidation of the regeneated FeS and reflects the
spectrum of possible reaction products between FeS (fresh/regenerated) and FeS« (fully
spent  reagent).   The  SULF-X Process  is typically  designed and  operated  to limit
elemental sulfur formation in the  wet-scrubbing step (to  maximize "n") so that  difficult
to  separate  finely suspended sulfur  will  be   yielded in  the subsequent  thermal
regeneration step where it  is more easily isolated and recovered.

      Intermediate  aosorption  reaction  products include HS™,  SO^,  HSCC and  S.-CC
(thiosultate).   Secondary  reactions not  reflected in the overall absorption  reaction
equation include oxidation  of ferrous ions to ferric and oxidation  of bisulfite  and sulfite
ions to sulfate.

NUX Removal


      The equation tor tne SCu-removal reaction above illustrates the  oxygen absorption
capaoility of  the process.   Through retrofit modification or by provision in the initial
installation,  the  SULF-X  Process  may  be designed  to use  this chemical  reducinq
(deoxidizing) cnaracteriscic of its process  slurry to achieve simultaneous  removal of
NOx  from  boiler  flue gas.   Catalysis  necessary to accomplish this is provided by
operating tne system  so as to maintain  approximately  40 grams per liter ferrous ion
(Fe  ) concentration in the continuously recirculating  scrubbing slurry.  The overall
chemical reaction in this simultaneous absorption of SO? and NO  may be expressed by
the following simplified single equation:                         x

    (c) FeS + S02 + NO + i DZ ^+ (c-1) FeS(m) + Fe++  + SO= +  (1 - JQ) S° + i N2t
                                       11-112

-------
MM 6*5
II
         SOLIDS
                  CLEAI 6AS
                  OUT
                  ACID 6AS
                  •BSORPTIOi
                                    REGENERATED SLURRY
                  SOLIDS
                  DEMATERII6
                                    LIQUOR
                           TRANP-SULFATE
                           CRYSTALLIZATION
 COKE
THERMAL  DMYJI6/
REfEIERATIOl
8Y COAL
                           ELENEITAL  SULFUR
                           It-fRODUCT
                                              LIQUOR
                                               SOLIDS
 Figure  1.  SULF-X  Process Flow Diagram
                           11-113

-------
     Additional intermediate reaction products  include  Fe(NO)++  formed  in  initial
capture of NO.  To remove most  of  the  NO  , substantially greater reaction time is
needed  in  the absorber  than is  required  in  collection of  SO_  alone.   Thus,  for
simultaneous SCL/NO  removal, the installation must provide, either through retrofit
modification or oy initial design,  additional  absorber  capacity to increase the  gas
residence time.

     Note that the absorption-reduction environment in the process slurry chemically
reduces  the  collected  NO  to   innocuous  elemental  nitrogen.    By  contrast, a
simultaneous SO?/NO  removal  process based on absorption-oxidation chemistry that
utilizes an  oxidizing Xagent,  converts  collected  NO  to potentially  water-polluting
oxidized forms such as  nitrites (NO^ and nitrates  (NOy.

     Absorption-reduction can be distinguished from  absorption-oxidation by a compar-
ison among familiar types of calcium-alkali scrubbing  systems.  An absorption-reduction
environment exists in some  of the  common  scrubbing systems such as those  using
magnesia-buffered lime and sodium-liquor scrubbing  (dual-alkali PGD) that  operate in
an  unsaturated-CaSO, -mode  with a significant  ionic  concentration  of  sulfite/bisulfite
(SO^/HSOp, an oxygen scavenger.  On the other hand, lime/limestone FGD systems, as
most commonly applied,  function without appreciable dissolved concentrations of such
chemical deoxidants.   Their scrubbing  slurries typically impose an absorption-oxidation
environment due  to the  presence of dissolved oxygen and are  incapable of chemical
reduction.

SULFIDE/SULFATE REGENERATION

     Suspended solids  in the slurry bleed-off from the acid gas absorption step (Figure
1) are  dewatered prior to  roasting  in  an  indirect-fired calciner.  Most of the process
liquid from  bleed dewatering, principally containing dissolved sodium sulfate (Na^SO.)
and ferrous  sulfate (FeSO^), is used as a quench to cool and slurry the hot solids leaving
the calciner.  The balance of the liquid is diverted to a crystallizer system  to recover
sodium  sulfate solids.  The spent scrubber solids combined with the  crystallizer solids
and a proportion  of coke constitute the feed to the calcining  step.   In the calciner at
1200 to 1400  F, sulfide  is regenerated and sulfur by-product  is formed in accordance
with the equations as follows:


                             FeSl  +        FeS  +   50
                          Na2S04 + 2 C -» Na2S + 2

      To avoid oxidation  of  any of the sulfur  yield, excess coke  is  supplied to the
calciner to provide a protective reducing atmosphere containing carbon monoxide (CO).
During  the quenching  step noted  above, the  sodium sulfide in the calcine  forms FeS
precipitate:

                   Na2S  + FeSO4 + H2O -> FesJ- + Na2SO4 + H2O

THROWAWAY WASTE GENERATED

      The  SULF-X Process produces a minimum of non-regenerable wastes and thereby
minimizes the environmental  impact of acid gas emission control facilities.
                                       11-114

-------
Solid Waste

     Oxidation  of scrubbing slurry results primarily in  the  conversion  of sulfides to
sulfate, which is regenerated to sulfide by calcination.  Thus, the quantity of throwaway
solid waste  is negligible in comparison with the amount of ash and other  solid wastes
generated by coal-fired boiler operations.  A comparatively small proportion of ferrous
iron, estimated to be equivalent to less than 5% of the stoichiometric SO.-, removal rate
in pulverized coal-fired  boiler applications, is oxidized to ferric oxide (Fe^O^) in the
scrubbing step.  Ferric oxide is a tramp solid material that may be discarded from the
system  through  disposal  of  a  small amount  of  dewatered  scrubber-bleed solids.
However, the reducing conditions in the calcination step  convert a large portion of the
ferric oxide solids to reactive ferrous  compounds, thereby decreasing the net amount of
iron oxide waste production.

Liquid  Waste

Chloride Management—
     As in other wet scrubbing processes, hydrogen chloride contained in the inlet flue
gas is efficiently absorbed and must be purged in a liquid effluent.    To prevent loss of
iron salts in this purge stream and to minimize complexity of wastewater management,
the absorbed chloride may be segregated in an isolated pre-scrubbing loop as is done in
application of other regenerative FGD processes.

Other Waste-ion Formation-
     Chloride is the only non-jegenerable  dissolved solid  that tends to accumulate in
the system.  Thiosulfate (S^CO,  a common tramp  compound  that must  be purged in
sodium-base regenerative  FGD systems, is both formed  and reconverted to bisulfite
(HSOQ in the SULF-X scrubbing step.  Thus, purging of thiosulfate from the system is
not required.   Moreover,  unlike  sodium sulfite based absorption-reduction processes
utilized  for  simultaneous  SO^/NO   removal,   the  SULF-X  Process does not form
detectable quantities of  dithionate, imidodisulfonate or other nitrogen/sulfur tramp ion
complexes.

                         SYSTEM DESIGN  AND OPERATION

     The general  flow diagram for a  commercial  SULF-X Process  installation in coal-
fired service is illustrated in Figure 2.

OVERVIEW OF DESIGN BASIS

Gas Absorption

     As in  current typical application of other flue gas  desulfurization processes, an
electrostatic precipitator  (or fabric filter), not shown in  Figure 2, is used to control
stack fly-ash  emission  and  to limit  the amount  of  solid particulates  entering  and
accumulating in the wet  scrubbing operation. In addition, a prescrubber loop as shown is
used upstream  when  required for absorption of  other  tramp materials, including HC1.
This limits  their impact on the operation  of  the process and  the  control of process
chemical losses  (through closed-loop recycle of SULF-X liquid and solid flows). Design
of large-scale SULF-X absorbers is tied closely  to  the prior development and continued
test/demonstration activity in the simultaneous removal  of  SO2 and NO  .  To ensure
mass transfer capacity sufficient  to meet potential NO  control objectives, the packed
                                      11-115

-------
                                                                                           Legend

                                                                                          BFW -    Boiler Fe*d Water

                                                                                          Equipment:
                                                                                               A    -    Pre»erubfo«r
                                                                                               B    -    Absorber
                                                                                               C    -    Thickener
                                                                                               D    -    Centrifuge I
                                                                                               E    -    Heat Exch«nf«e
                                                                                               P    -    CrysUUIzw
                                                                                               G    -    Centrifug* B
                                                                                               H    -    Dryer
                                                                                               1    -    Dust Collector
                                                                                               J    -    Indirect-fired Regenerate*
                                                                                               K    -    Waste Heet Boiler
                                                                                               L    -    Sulfur Condenser
                                                                                               M   -    Incinerator-
                                                                                               N    -    Quench Tank
                                                                                               O    -    Ball MIU
                                                                                               P    -    Reclrculstion Tank
                                                                                               Q    -    Absorber Feed Tank
                                                                                               R    -    Flue Gas Reheater
Figure  2.   SULF-X Process  Flow  Diagram

-------
absorber design utilizes fixed cross-fluted  Munters packing material in counter-flow
scrubbing operation.   In  the design of  the  packed absorber in SO~-only  service, a
recirculating slurry  to  gas flow ratio of  approximately 60 U.S. gallons per thousand
actual cubic feet  is used with an absorber  gas-pressure-drop of  8  in. w.g. at a design
superficial gas velocity of 10 feet per  second.  In such applications concerned  solely
with  SO2  removal  in  the  85-95%  efficiency  range,  a  spray-tower-type absorber,
operated at  a  superficial  gas velocity  comparable to  designs used in lime/limestone
slurry  scrubbing systems,  may be found  to be  preferable.  Due to provisions in the
process for regeneration of sodium sulfate,  the principal product of parasitic oxidation
in the absorber, the SULF-X absorption-reduction system  can, uniguely, be  applied to
emission sources such as bituminous-coal-fired stoker boilers that may contain oxygen
concentration  in  the raw flue gas  as  high as  14%.   Thus,  with this  capability for
processing flue gases with a very high O^/SO^ ratio, the SULF-X Process can be applied
to pulverized coal boilers fueled with coal of all ranks and with either high or low sulfur
content.

Sodium Sulfate Cystallization

      An atmospheric  pressure crystallizer column is used to  preferentially crystallize
the sodium sulfate make for regeneration to sulfide in the calcining operation.  Reasons
supporting this method for recovering and reusing sulfate waste product are as follows:
           Chemical reduction of sulfate is thereby accomplished without liberation of
           SCL that accompanies thermal processing of alternative sulfate solids forms
           such as FeSCh.
           Iron compounds that are present catalyze the reduction of Na^SO. .
           The minimum temperature  for reduction does not  exceed the design  temp-
           erature for simultaneous calcination of the  spent  FeS solids (FeS^ +  I-^Q
           forms).
           Coke  intermixed with  the calciner feed materials can serve as an effective
           reductant.
           FeS  subsequently  precipitated  by  reaction of Na^S solids  with dissolved
           FeSO. is more reactive than the thermally regenerated FeS.
           Precipitated  FeS does not  require milling, thereby minimizing  the size of
           milling equipment required for the total supply of FeS to the absorber.

Thermal Regeneration

      After the feed  has  been dried in  a steam-heated co-current rotary dryer, an
indirect coal-fired rotary drum  kiln is used to roast and regenerate feed solids  at a
temperature as high as 1400°F. Sulfur released during the calcination leaves the  kiln  as
a vapor and is recovered by condensation.  By-product steam  is generated in the sulfur
condensation step, as well as in an unfired waste heat boiler fed by the calciner exhaust
combustion gases.

      Note that indirect calcination isolates the kiln combustion  gas from the  CO/CCL
process atmosphere  laden with the  sulfur product vapor.  Thus, unlike  FGD processes
such  as  the MgO process that  use  elevated temperatures requiring  direct-calcination
operation, the SULF-X  Process by-product yield and  calcined  solids do not become
intermixed with  or  contaminated by  the  combustion  gases from calciner fuel  firing.
                                       11-117

-------
This, in turn, affords the opportunity to use the cheapest available fuel for the calcining
operation.

PROCESS CONTROL

     Principal  provisions for control  of  the process and for isolation and  segregated
shutdown of individual process steps are comparatively simple.

Wet Scrubbing

     The  scrubbing operation  is controlled by  maintaining  a  set-point  pH  in  the
scrubbing slurry through regulation of feed of the  regenerated solids slurry entering the
absorber loop.   Additionally,  feed of  this slurry  and bleed-off  of spent  slurry to the
regeneration step are coordinated to maintain scrubbing slurry inventory  and density in
an  optimum range  to  ensure scrubbing  effectiveness.   Safeguards are provided  to
prevent slurry-pH   excursions that could  impair  gas  cleaning efficiency or cause
objectional decomposition of process chemicals.

Decoupling of Process Steps

     To afford means  of temporarily  shutting down  individual  process  steps for
maintenance without interrupting overall system operation, a  process  inventory and
freeboard capacity are provided for storage of process materials including:
           Dewatered spent scrubber solids in thickener-tank bottom (dryer feed)

           Dried-cake (calciner feed)
           Calciner quench-tank slurry (scrubber feed)

MATERIALS AND ENERGY UTILIZATION

     A current, detailed process economics evaluation by Electric Power Research
Institute (EPRI)   includes a preliminary conceptual design of a hypothetical SULF-X
Process installation for removal  of SO2 from a 500  MW  unit  fired by 4.0% sulfur
bituminous  coal.  The  consumptive  use of  materials and energy and the yield of by-
product elemental sulfur  was  estimated based  on full-load operation  at 500  MW.
Assigning  applicable dollar values to each item as displayed below, it  is seen that the
overall cost of  materials and energy  can be offset by by-product revenue.  Thus, for one
day of 500 MW operation at full load (generating 12  million KWH), the  cost summary is
as follows:

Consumption

     Pryites, 2.8  tons/hr x  24 x $25/ton                   $ 1,680
     Na-S,  0.3 tons/hr x 24 x $470/ton                     3,384
     Water, 727 gpm  x 1440/1000  x $0.30/Mgal               314
     Coke,  5.3 tons/hr x 24 x $51.30/ton                   6,525
     Coal,  1.3 tons/hr x  24 x $35/ton                       1,092
     No.  6  oil, 3 gpm x  1440 x $0.82/gal                   3,542
                                     11-118

-------
Consumption  (Continued)

     H.P. steam*,  100,500 Ib/hr x 24/1000  x  $4/Mlb        9,648
     L.P- steam, 4,500  Ib/hr x 24/1000 x $3  Mlb             324
     Electric  power, 12,800 KW x  24 x $0.045/KWH      13,824
     Total Material  and Energy Cost                      $40,333**

Production

     Elemental sulfur,  9  tons/hr x 24 x $200/ton          $43,200
     Total By-Product  Revenue                           $43,200**

                              PROCESS ECONOMICS

S02 REMOVAL ONLY

     The  detailed  1983  EPRI  economic study  of diverse flue  gas  desulfurization
processes  at the 500 MW scale  in 4% sulfur bituminous coal service estimates the cost
for the SULF-X Process as follows:
                                SULF-X PROCESS
                           (in  December,  1982 dollars)
                                             Levelized  Busbar Cost
    Capital Cost               	(Capital and  Operating)	
(Including contingencies)        No By-Product Credit     $75/ton Sulfur Credit
       $/KW                       mills/KWH                  mills/KWH

       295                             22.8                        20.0

     FGD process selection is highly site-specific and will depend on  items such as the
complexities  anticipated  in throwaway  waste management and  the  actual revenues
available  from  by-product  sales.    The  market  value of  the sulfur  by-product
significantly affects  the operating cost of the SULF-X Process.  For example, with the
sulfur price projected  at  $200/ton, the  levelized busbar cost with by-product credit
decreases  to a very favorable (within the guidelines of the study)  15.4 mills/KWH.

SIMULTANEOUS SO2/NOX REMOVAL

     1979 in-house  studies of similar utility-scale  applications for commercially
available  combined  SO2/NO  removal  facilities achieving 80% NOx removal  confirm
that SULF-X  is substantiallyXless costly than other available wet simultaneous SO2/NO
removal processes.   Moreover,  while capital cost is slightly higher than  for combined
removal  methods  that  utilize  dry catalytic  NO  control,  the  SULF-X  Process  is
estimated to  have lower  annual revenue requirements  than other available SO2/NO
removal methods including those based on dry NO  removal.
                                             /\
     The  levelized  busbar cost  of  a  500 MW SULF-X  installation  designed  for
                                                                  ojected  to be 4
                                                                  This increment,
simultaneous  SO2/NO  removal achieving 80% removal of NO  is projected to be 4
mills/KWH higher  than  that  for  SO?-only service  reviewed  above.
*    50°F indirect-type flue gas reheat
**   Approximately 3.5 mills/KWH
                                     11-119

-------
representing an increase of 20 to 25% over the cost of simple FGD capability, consists
almost entirely of capital costs  for added parallel-operated absorber trains required to
provide increased gas residence time for high-efficiency NOx removal.

INDUSTRIAL BOILER APPLICATIONS

     To  take advantage of the  favorable economics of  large-scale sizing,  small-
capacity  coal-fired boilers in an industrialized region may be equipped with the wet-
scrubbing-system portion of the SULF-X Process  while being served  as a group by  a
single, centrally  located regeneration plant.   The central  plant  would  supply fresh
calcine to each  scrubber  and  receive the  spent  sulfides on the return haul.  The
attractiveness of the economics  of such applications may approach that of utility-plant-
scale installations.

                 SULF-X PROCESS DEMONSTRATION OPERATION

     Development  of the SULF-X Process  through laboratory and field test work was
completed in 1980.  Design and  construction of a 1.5 MW stoker-boiler size integrated
process installation for  the Commonwealth of Pennsyvlania was completed  in 1982.
This demonstration  facility  has been in  round-the-clock operation since  early 1983,
providing  data for process scale-up to larger coal-fired boiler applications.

INTRODUCTION

     The  integrated  SULF-X  Process  installation  completed in  1982   is  now  in
demonstration operation at the bituminous coal-fired boiler plant of the Western  Center
Hospital  of the  Commonwealth of Pennsylvania  at  Canonsburg,  Pennsylvania, near
Pittsburgh.   The system is designed for  the  simultaneous  control of SO2 and NO
emissions from one of the plant boilers at a flue gas flow rate of approximately 7,000
acfm.    Its  current  test/demonstration  operation  in  1983,  under  funding  by  the
Commonwealth of Pennsylvania, constitutes a critical interim stage in the scale-up and
application  of  the  process through  collection  of steady-state  operating  data  in
anticipation of  its  ultimate use in  large  commercial installations.   The continuous,
integrated operation of  this demonstration facility at the same time has  afforded  a
needed opportunity to study the  effects on the  overall chemistry  of any accumulation in
the  system  inventory of reaction products and tramp materials,  and to  evaluate  the
compatability of the installed equipment and instrumentation.

DESCRIPTION OF WESTERN CENTER FACILITIES

      The SULF-X  installation at Canonsburg  is  a  $2 million facility dedicated by  the
Commonwealth  of  Pennsylvania to a  12-month  demonstration period.   The  system
advantageously utilizes  off-the-shelf  process equipment and control instrumentation to
provide a practical replica of the process facilities that are  expected to be ultimately
applied in large commercial installations.   Thus, the system models  a  large-scale
installation,  using  applicable  mass  transfer  devices  and  mechanical  equipment  to
duplicate  primary  unit  operations  such as the gas absorption,  slurry dewatering  and
thermal/chemical regeneration steps, all of which  are sub-processes commonly  used  in
industry.  See  the Flow Diagram, Figure 3.

     Flue gas cleaning  equipment provided includes a  3 feet diameter radial blade
booster fan, a lined carbon steel gas pre-cooler and a 73  feet high by 5 feet diameter
                                       11-120

-------
            4 r *•«.***/*
Equipment
A
a
c
o
E
r
c
H
1
J
K

L

M

N

0 -

Seturetor (Pre-CooO
Abenroer
Thickener
Overflow TV*
Filter Preei
Drrer
CtlclfMT
Sulfur Condeneir
Quench TV*
Attritor
AbKutwr Feed
TV*
Sodium Sulflde
TV*
Preciplmkm
TV*
Precipiutlon
TV*
Sodium Sulfat*
                                                                             Solution TV*
Figure 3.   Flow Diagram - SULF-X System  at  Western Center

-------
FRP scrubber (absorber) vessel using  cross-fluted fixed plastic packing.  Slurry-bleed
dewatering equipment consists of a 10  feet diameter lined thickener with overflow tank,
a horizontal  plate-and-frame filter containing seventeen 1  meter x 1  meter vertical
filter plates and a steam-heated rotary  dryer.  The regeneration system consists of an
indirect gas-fired  rotary calciner, a calcine quench tank and a steam-cooled vertical
shell-and-tube  sulfur  condenser.   For  supply of  regenerated  reagent  to  the gas
absorption system, the system includes  an agitated-media-type calcine slurry crusher
(attritor) and an absorber fresh slurry feed tank.*

      Without adversely affecting the simulation  of  large commercial  system opera-
tions,  the overall  demonstration  program has been simplified by  excluding  sodium
sulfate crystallization facilities.  Instead, dissolved sodium sulfate is purged from the
system by blowdown.  Purchased sodium sulfide is used (in lieu of sodium sulfide that
would otherwise be available from calcination of sodium sulfate) to sustain the quench
tank reaction, which recovers dissolved ferrous sulfate by converting it to precipitated
FeS reagent.  To  further simplify the operation of this small-capacity demonstration
operation, the calciner  is fired with LPG  (propane) instead of coal. Since calcination is
by indirect heat transfer both in  the commercial scale and Western Center designs, use
of  alternative  fuels  does  not affect process  chemistry or  the  validity  of the test
operations.   In  addition,  pre-cooling  and water  saturation  of  the boiler  flue  gas  is
carried out in  a simple gas quenching step utilizing recirculating process  slurry.  In
absence of an isolated  pre-scrubber wastewater purge,  chloride will be purged by the
discard of the  dissolved  sodium  sulfate blowdown  stream.   As  a result of  other
simplifications,  calciner combustion  gases are exhausted without by-product  steam
generation and  sulfur condenser  off-gas is recycled  to the  absorber inlet  instead of
being treated in an incinerator.

PROBLEMS IN COMMISSIONING OF THE SYSTEM

      Principal startup/operating problems encountered with the initial system installa-
tion included difficulties in mechanically transferring the  filter cake to the dryer, in
moving the filter cake through the dryer  and in oxidation of the dried solids  by air that
was entering the process side of  the calciner.  Because  of the small batch-scale nature
of the filtering  operation at Western  Center, movement of wet filter cake is expected
to continue to require  special operator  attention during continuous process runs.  The
original dryer  and calciner were replaced in late  1982  with the  more  compatible
equipment described above and these operations are now considered to be free of major
problems.  Adverse oxidation of wet cake was experienced in early operation of the new
steam-heated rotary dryer, but this problem  has been eliminated.  Retrofit modifica-
tions have also been made to the sulfur  vapor  vacuum and condensing systems to avoid
solidification of sulfur before and after the condenser.

OVERVIEW OF SYSTEM TESTING WORK

      The  Western Center  absorber has successfully demonstrated up to 99+% removal
of SO  and up to 93%  removal  of NO   in 2%-sulfur bituminous coal operation.  At a
superficial velocity no greater than 2 feet per second,  NO removal efficiency of 90% is
obtained with essentially 100% SO2 removal.  Figure 4 shows the graphical relationship
*For  initial ferrous chemical charge  and subsequent Fe makeup supply, the Western
Center operation has used copperas produced from waste pickling liquor neutralization.
                                       11-122

-------
I
h-"

U)
  100
   80
o

lu
QC  60

«0
IU
Q

X
O
lu
O
O
ft
   40
   20
                                                                       I                I

                                                                FERROUS ION CONCENTRATION



                                                                      35 GRAMS/LITER


                                                                    IN SCRUBBING SLURRY
                    100
                                    200
                                                    300
                                                                                     400
                            FLUE GAS VELOCITY (FT/MIN)
 Figure 4.   Nitrogen Oxides Removal
               vs. Superficial  Flue
               Gas Velocity

-------
between  NO   removal and  absorber  superficial gas velocity for the present  absorber
configuratior^ functioning as a simple single-stage counter-current  contactor.  Other
variables affecting NO  removal  rates are  slurry temperature  and viscosity and FeS
particle size.  Figure  £ shows the inlet SO2 flue gas  concentrations and  the percent
absorbed when the system was operated for  simultaneous  SO-and NOx removal. At a
superficial  gas velocity  in  the scrubber  of  approximately  ID feet per second and a
liquid/gas ratio of 60 U.S.  gallons  per thousand cubic feet, typical of operation of
commercial FGD systems, SO2  removal at an  efficiency of  95+% is accompanied by only
token NO  removal.

      The  absorption  system  has operated automatically  with  the  pH controller
regulating fresh slurry addition and spent slurry removal to maintain a constant, pre-set
pK   Spent  slurry  has  dewatered  well  in  the  thickener  and filter, and  the  FeS
precipitation and chemical  makeup systems  have  performed as designed.  Comprehen-
sive data are  being collected for a broad  range of  operating conditions  to achieve
parametric testing that will optimize system design for NOx removal levels up to 80%,
and to yield a data base to be used for scale-up.

CONTINUOUS PROCESS OPERATIONS

Chemistry  of Steady-State Process Streams

      The  iron, sulfur  and  sodium species   in  the  recycled  scrubbing slurry  were
monitored  throughout  the  project  for  both  daily process control and  overall system
performance  evaluation.  Relatively high concentrations  of sodium  sulfate (Na^SOj
were  measured but were anticipated since the sulfate crystallizer had not been included
as part  of  this system and  a low liquid purge rate had been maintained.  This rate of
liquid waste  blowdown was  adjusted to maintain the dissolved Na^SO^ concentration in
the slurry below its saturation point.

      Analyses indicated that steady-state conditions had  been achieved by the fourth
operating  month.   Chemical  consumptions  and product concentrations  have been
consistent  with  earlier studies and  the previously  identified chemical reactions.
Approximately 17% more ferrous sulfide was  reacted in  the absorber than was expected.
This increase is believed to have been caused by  the unusually high oxygen content in
the flue  gas, ranging from 10 to 14%.  A  typical analysis of  process slurry  is presented
in Table 1.

Quantification of Solid and Liquid Waste Discharge

      There was no significant  solid waste discharged from the system.  Approximately
13,600 gallons of sodium sulfate liquid waste were purged  from the system.  Analysis of
this waste liquor is as follows:

      Iron, Fe++      0.6  g/1             Sulfate (SO
      Sodium, Na     34.0
      Other cations
      (Ca, Mg,  Al)    7.6

      pH               6.4
Sulfite (SOp
Thiosulfate  (S-Op
Chloride  (CL~)
76.9 g/1

 0.1
 9.3
 1.0
                                     11-124

-------
ho
Ul
INLET SO CONCENTRATIONS (ppm)
t— •
N) *» 0\ CO 0
O O O O O
O 0 O 0 0 0
-
-
-
so^ cowc.


I


1


1




1


J


1


1


1


EMO
1


ML
1


1


1


1




1
100
99 )?
to
98 i
03
97 m
o
96
                                             6    7    8    9   10   11

                                            OPERATING TIME (WEEKS)
12
    13
        14
            15
                16
                  Fiqure  5.   Inlet Flue Gas S0~  Concentrations and
                                Percent Absorbed  by  SULF-X System

-------
  TABLE 1,
5ULF-X PROCESS SIMULTANEOUS SCU/NO^
REMOVAL
                         Recycled Scrubber
                         Slurry Composition







Iron (Fe )
Sodium (Na )
Other Cations*
(Ca, Mg, Al)
Iron, total
Sulfur, total
% Suspended solids
% Dissolved solids
Specific gravity, slurry
pH, slurry
Dissolved Solids,
49.9 g/1
48.0

0.3
6.42%
7.25%
5 . 72%
22.1 %
1.29
5.4
grams/liter
Sulfate (SOp 164.0 g/1
Sulfite (SOp 0.1
Thiosulfate (S9O^) 20.7
£ J
Chloride (Cl~) 1.7

Calcium and magnesium salts are contained as anti-caking agents in the purchased
iron sulfate makeup; aluminum is leached from the scrubber fly ash accumulating
in the slurry.
                              11-126

-------
     Taking into account the change in system chemical inventory, the quantity of the
dissolved  sulfur  species  in excess  of  that contributed by  ferrous sulfate  makeup
represents  approximately 7% of  the sulfur dioxide absorbed  during the test period.
Sulfur (sulfate) losses in the blowdown can be expected to significantly decrease and the
ratio fo  chloride  to  dissolved sulfur species  correspondingly  increase  for a SULF-X
system installation that includes a sodium sulfate crystallizer.

SUMMARY OF FINDINGS IN WESTERN CENTER DEMONSTRATION RUNS TO DATE

     The installation and  operation of  the demonstration  SULF-X facility  for the
Commonwealth  of Pennsylvania  has been  an important and  successful step  in the
continuing development of the Process.   The data and information collected during all
phases of the project are valuable additions to the engineering and  operating  criteria
needed for building a  larger  SULF-X FGD installation of optimized design.

     During the operation,  equipment and instrumentation compatability  in the face of
variances in process operation was thoroughly evaluated.  The results served to identify
means  of overcoming potential problem areas and confirmed the accuracy of previously
established process design criteria:

           The packed tower functioned  as  an  effective mass transfer unit and should
           be used in future systems  requiring substantial NO removal.
                                                          X
           Use of pH for control of process  chemistry and slurry make-up rate was
           effective  and reliable.
           Dithionates and imidodisulfonates  did  not  accumulate in the recirculated
           slurry nor in the  sodium sulfate solution waste.
           Slurry  bleed from  the absorber was  readily dewatered with a pressure filter
           after  thickening.  This final dewatering could probably be performed with a
           centrifuge.
           Material  transfer of the filter cake  was troublesome and requires additional
           investigation if batch-wise filtration is  to be used reliably in routine system
           operation.
           To ensure efficient  regeneration and sulfur recovery, precautions must be
           taken  to  avoid  excessive oxidation of the solids  during  the  drying and
           calcining steps.
           Iron oxide was converted in the  calciner to an active ferrous  form, thereby
           substantially decreasing the anticipated waste solids blowdown rate.

PROJECTED PROCESS DEMONSTRATION AT  LARGER SCALE

     A new and larger scale demonstration of the process is now planned  in conjunction
with future development of process licensees for  commercial  application of  SULF-X
technology in  the U.S.A.,  Canada, Japan,  West  Germany and  other  industrialized
countries.   In a  new system installation for this purpose,  it is  anticipated that  a
pulverized coal-fired  boiler facility  at  the 10 to  60 MW size range would provide  a
minimum-risk basis  for  ultimate scale-up of  the  process to utility  plant applications
larger  than 300 MW.  At the same time, it should be  possible to reliably quantify the
benefits, simplifications and anticipated economies in the process design resulting from
typically low flue gas oxygen content in pulverized-fired service.
                                       11-127

-------
                EXPANDED WESTERN CENTER DEMONSTRATION
                       FOR U.S. DEPARTMENT OF ENERGY

     A new test/demonstration program is now being carried out at Western Center
under  the  sponsorship  of  DOE  to  define optimum  design/operating conditions  for
achievement of  90%  NO   removal in simultaneous SO2/NOx scrubbing operation.  A
systematic sequence of test runs is being conducted in a continuous integrated closed-
loop mode to evaluate performance,  identify important operating variables and provide
data for assessing  the technical and economic feasibility of high efficiency, simultan-
eous SO-/NO   removal.  Additionally, process runs  are being made to demonstrate
regeneration of sodium sulfate with coke as a reductant.  Alternative fuel and reductant
materials are also being investigated.

90% NO REMOVAL
        X
     The piping of the two in-series  absorber beds, originally arranged to operate as a
simple  single-stage  counter-current  scrubber,  is being  modified  to  provide  two
separately operated scrubbing stages.  A series of test runs to evaluate performance and
study important operating variables will be conducted.  Parameters to be adjusted and
assessed include liquid/gas  flow ratio of the individual absorber stages, superficial gas
velocity and chemistry  of inlet slurry to each stage.  Parametric testing will take into
account the effect of other variables that may significantly affect  the NO  absorption
rate.

RUNS  WITH MODIFIED CALCINER OPERATION

     The calciner  test program will model-test  typical large-scale SULF-X  Process
operation in which  the regeneration system is regenerating all tramp sulfate formed in
the scrubbing step to fresh, usable sulfide reagent.  This will entail  continuous addition
of coke reductant  and sodium sulfate solids at the calciner feed inlet. Individual test
runs will be made  to investigate alternative use of petroleum coke, metallurgical  coke
and boiler-house coal as the reductant.

LONG-TERM RELIABILITY RUN

     After  establishing an  applicable  mode and  optimum  conditions  for SULF-X
operation at 90% NO  removal,  a continuous 30-day system reliability  run will be made
to verify the  design  basis  for  such  commercial operation.  Quantities and quality of
waste  discharges during sustained operation will be monitored  to assess system waste
production characteristics and  to identify any specific provisions that  would be needed
for treatment  and  ultimate disposal of the waste.

CONCEPTUAL COMMERCIAL-SCALE DESIGN FOR 90+% SO?, 90% NO  REMOVAL
                                                          £-          A
     Based on data  generated  and   conclusions  drawn  during  the  field  test work,  a
conceptual SULF-X Process design will be made for a hypothetical 500 MW power plant
burning high sulfur bituminous coal.   This design will serve as a basis for assessment of
the technical and economic feasibility of this mode of operation  of the SULF-X Process.
                                       11-128

-------
                                  CONCLUSIONS

     SULF-X is an effective new wet-scrubbing-type regenerative flue gas desulfuriza-
tion process applicable to all fossil fuels.  It has been demonstrated to advantageously
use and regenerate ferrous sulfide originating from industrial iron by-product wastes to
convert flue  gas  SO2 into  usable elemental sulfur while producing negligible solid and
liquid waste except common chloride liquor purge.

     The  process has  demonstrated an inherent versatility in functioning  as either a
system for SO2 removal only, or for simultaneous SOo/NO  treatment.  Cost projections
indicate that  rt is an economically attractive alternative to throwaway-waste-type FGD
system where a regional market for by-product  sulfur exists, particularly in applications
where solid waste management costs would be expected to be disproportionately  high.
Moreover,  the capital investment for small  to medium-sized SULF-X units can  be
minimized if  concentration of industry in a region allows a central regenerating plant to
serve a number of individual gas scrubbing units.

     Additionally, the SULF-X system is technically and economically  attractive where
nitrogen oxides removal is also necessary at present or in the future.  The system lends
itself to being installed and operated as an SO2-only FGD system and then later being
readily modified to also remove nitrogen oxides.

                                  REFERENCES

1.   Stearns-Roger Engineering   Corporation,  Economics of  FGD;  Electric  Power
     Research Institute, draft report RP1610-1, March 1983.

2.   Pollution Technology  Review No. 82,  New Developments in  Flue Gas Desulfuriza-
     tion  Technology; edited by M. Satriana, Noyes Data Corporation,  Park Ridge, New
     Jersey,  1981.
                                     11-129

-------
 APPENDIX




Attendees

-------
                   EPA/EPRI SYMPOSIUM ON FLUE GAS DESULFURIZATION

                                 November 1-4, 1983
                             Sheraton New Orleans Hotel
                               New Orleans, Louisiana

                                  List  of Attendees
R. S.  Abraham
Engineering Design Coordinator
Florida Power & Light Co.
P.O.  Box 529100
Miami, Florida  33152
305/552-3834

Robert J.  Abrams
Sales Manager
Bishopric Products Co.
4413 Kings Run Drive
Cincinnati, Ohio  45232
413/641-0500

toesim Abuaf
Mechanical Engineer
General Electric Co.
Building 37, Room 647
Schenectady, New York  12301
518/385-3654

Steven J.  Achtner
Sales Representative
M. W. Kellogg Co.
433 Hackensack Avenue
Hackensack, New Jersey  07601
201/646-1000 (ext. 2435)

Radford C. Adams
Program Manager
TRW
P.O.  Box 13000
Research Triangle Park, NC  27709
D. D.  Agarwal
Manager Corrosion Alloys
Cabot Corp.
1U2U W. Park Ave.
Kokomo, Indiana  46901
317/456-6031
Stefan Ahman
Senior Chemical Engineer
Flalt Industri AB
Kvarnvagen, Vaxjo
Sweden S-35187
046/470-87276

0. Ainsworth
Lab Director
Dow Chemical Co.
P.O. Box 150
Plaquemine, Louisiana  70764
504/389-1620

Sy A. Ali
Executive Director, Environmental
  Programs
Public Service Indiana
1000 E. Main St.
Plainfield, Indiana  46168
317/838-1229

David L. Almand
Western Regional Sales Manager
Joy Manufacturing Co.
P.O. box 2744, Terminal Annex
Los Angeles, California  90051
213/240-2300

Harold R. Althen
Marketing Manager
Peabody Process Systems
835 Hope Street
Stamford, CT  06907
203/327-7000

Chuck Altin
Project Manager
Ebasco Services
145 Technology Park
Norcross, Georgia  30092
404/449-5800
                                           A-l

-------
Marlin Anderson
Director, Technical Services
Crestmont Assoc.
P.O. Box 770
Central City, Kentucky  42330
502/754-9460

Gary Andes
Air Pollution Control Engineer
Gi1bert-Commonwealth
P.O. Box 1498
Reading, Pennsylvania  19603
215/775-2600, ext. 2160

Jumpei Ando
Faculty of Science & Engineering
Chuo University
1-13-27 Kasuga
Bunkyo-Ku
Tokyo  112
japan

Amjad  H. Ansari
Mechanical Engineer
Stone  & Webster
16430  Park Ten Place
Houston, Texas  77064
713/492-4148

Roger  Antonie
Utility  Sales Engineer
Warman International Inc.
2701 S.  Stoughton  Rd.
Madison, Wisconsin   53716

Bertrans Anz
Vice President
United Engineer &  Constructors, Inc.
9111 Cross Park Dr.
Knoxville, Tennessee  37923
615/690-8610

C.  William Arrington
President
Crestmont Assoc.
P.O. Box 770
Central City, Kentucky  42330
502/754-9460
Franklin A. Ayer
Symposium Coordinator
Research Triangle Inst.
P.O. Box 12194
Research Triangle Park, North
  Carolina  27709
919/541-6260

Bill Babcock
Sales Representative
Marblehead Lime
Salt Lake City, Utah  84103
601/364-7117

Lothar Bachmann
President
Bachman Ind.
29 Lexington St.
Lewiston, Maine  04240
207/784-2338

Ronald J. Bacskai
Vice-President, Marketing
Conversion Systems, Inc.
115 Gibraltar Rd.
Horsham, PA  19044
215/441-5920

Brian Bahor
Application Engineer
Wheelabrator-Frye Inc.
600 Grant St.
Pittsburgh, Pennsylvania  15219
412/288-7519

Even Bakke
Vice President
Peabody
635 Hope St.
Stamford, Connecticut  06907
302/327-7000

Armand A. Balasco
Engineering Consultant
Arthur D. Little, Inc.
20 Acorn Park
Cambridge, Massachusetts  02140
617/864-5770
                                          A-2

-------
h.  A.  Bambrough
head,  Major Combustion Sources
Environment Canada
351 Blvd.  St.  Joseph, Hull,
Quebec,  Canada
819/997-1220

John H.  Banks
Market Development
Ashland Chemical  Co.
18 Foxhall PI.
Scarsdale, New York  10563
914/472-6953

Bruce P. Bannon
Director,  Marketing
RMI Co.
1000 Warren Ave.
Niles, Ohio  44446
216/652-9951

Jim Barlow
Mechanical Engineer
Benham Holway  Power Group
5300 S. Yale
Tulsa, Oklahoma  74135
918/492-2411

Yves Barthel
Project Leader
Institut Francais du Petrole
1 et 4, avenue de Bois-Preau
92506 Rueil Malmaison
FRANCE

Daniel C.  Beal
Sales Engineer
beneral  Electric  Co.
2015 Spring Rd
Oak Brook, Illinois  60521
312/986-3021

R. T. Beall
Vice President/Sales Manager,
  Aggregates/Cement
Ciifford-Hill & Company, Inc.
P.O. Box 42127
Dallas,  Texas   75247
214/258-7321
Joseph L. Beals
Wallace & Tiernan/Pennwalt
2001 Midwest Rd.
Oak Brook, Illinois  60521
312/620-8820

Earl R. Beaver
Business Development Director
Monsanto
800 N. Lindburgh
St. Louois, Missouri  63167
314/694-8620

Steven Becker
Supervisory Engineer
Southwestern Public Service Co.
P.O. box 1261
Amarillo, Texas  79170
806/378-2441

E. B. Beckman
Marketing Engineer
Stone & Webster Engineering
250 W. 34th St.
New York, New York  10119
212/290-6109

Ann Behl
Research Assistant Tech. Writer
Radian Corporation
8501 Mo Pac Blvd.
Austin, Texas  78758
512/454-4797

Ranier Benninghaus
Thyssen Environmental Systems, Inc.
333 Meadow!and Pkwy.
Secaucus, New Jersey  07094
201/330-2600

Keith Benton
Development Engineer
Combustion Engineering
31 Inverness Center Pkwy.
Bham, Alabama  35243
205/967-9100

Jim Berding
Sales, Marketing Manager
CLOW
P.O. Bix 24
Florence, Kentucky  41042
606/283-2121
                                          A-3

-------
Chris Bernabo
Executive Director
Interagency Task Force on Acia Precipitation
722 Jackson PI., N.W.
Washington, D.C.  20006

Albert H. Berst
Engineering Manager, Pollution
  Control
Zurn Industries, Inc.
P.O. Box 2206
Birmingham  Alabama  35201
205/252-2161

P. A. Bhat
Senior Engineer II
Pennsylvania Electric Co.
1001 Broad St.
Johnstown, Pennsylvania  15907
814/533-8560

Robert T. Bianchi
Sales
Marblehead Lime Co.
300 W. Washington
Chicago, Illinois  60606
312/263-4490

Edward Biedell
Manager, FGD Applications
MikroPul Corp.
10 Chatham Rd.
Summit,  New Jersey  07901
201/273-6360

Cal Billings
President
Advanced Energy Systems
400 Cities Service Hwy
Sulphur, Louisiana  70663
318/625-4816

Thomas B. Blair
Senior Program Manager
Radian Corp.
8501 Mo-Pac
P.O. Box 9948
Austin, TX  78766
512/454-4797
Donald W. Blind
Mechanical Engineer
Tennessee Valley Authjority
400 W. Summit Hill
Knoxville, Tennessee  37902
615/632-4626

Julian Blomley
Environmental Analyst
Middle South Services, Inc.
934 Gravier St., P.O. Box 61000
New Orleans, Louisiana  70161
5U4/569-4741

Gary M. Blythe
Senior Chemical Engineer
Radian Corp.
P.O. Box 9948
Austin, Texas  78766
512/454-4797

Michael L. Bolind
Product Manager
United States Gypsum Co.
101 S. Wacker Dr.
Chicago, Illinois  60016
312/321-4371

Michael F. Bellinger
Environmental Chemist
Union Electric Co.
1901 Gratiot St.
St. Louis, Missouri  63103
314/554-3652

Dr. Arthur Boni
Vice President
Physical Sciences, Inc
P.O. Box 3100, Research Park
Andover, Massachusetts  01810
617/475-9030

Dave Bordson
Mechanical Engineer
Minnesota Pollution Control Agency
1935 W. Co. Read B-2
Roseville, Minnesota  55113
612/296-7780
                                         A-A

-------
Ed Boucher
U.S. Sales
ABCO Plastics, Inc.
45 Accord Park
Norwell,  Massachusetts
617/878-5068
02061
Williard L. Boward, Jr.
Senior Process Engineer
FMC Corp.
231 N. Martingale Rd
Schaumburg, Illinois  60194-2098
312/843-1700

George H. Bowen
President
Electric Utility Services
163 W. Saddle River Rd.
Saddle River, New Jersey
201/337-1788

Warren Bowman
Senior Development Specialist
E.I. du Pont de Nemours & Co. Inc.
1007 Market St., D-13121
Wilmington, Delaware 19898
302/773-3654

Donald E. Boyd
Director Marketing and Sales
Flakt Environmental
  Systems Division
P.O. Box 87
Knoxville, Tennessee  37901
615/693-7550

Paul A. Boyd
Senior Chemical Engineer
U.S. EPA
1200 6th Ave.
Seattle, Washington  98101
26/442-1567

Edward G. Boyer, Jr-
Cromby Station Superintendent
Philadelphia Electric Co.
2301 Market St.
Philadelphia, Pennsylvania  19101
215/933-8995
Richard C. Boynton
Manager, Environmental Systems/Sales
  Development
General Electric Co.
P.O. Box 7600
Stamford, Connecticut  06904
203/357-4975

Wallace H. Bradley
Vice President of Engineering
Austell Box Board Corp.
3100 Washington St.
Austell, Georgia  30001
404/948-3100

Jene Bramel
Vice President Sales
Air Clean Damper Co.
Blue Ash Rd.
Cinciannati, Ohio  45236
513/793-1253

Atly Brasher
Permit Program Manager
Louisiana DNR
P.O. Box 44066
Baton Rouge, Louisiana  70804
504/342-8942

Theodore G. Brna
Program Manager for Dry FGD
U.S. EPA
IERL, MD-61
Research Triangle Park, NC  27711
919/541-2683

Klaus L. Bro
Niro Atomizer, Inc.
9165 Rumsey Ra.
Columbia, Maryland  20145
301/997-8700

Norman I. Brody
Technical Sales
Stebbins Engineering and
  Manufacturing Co.
4830 North River Rd.
Port Allen, Louisiana  70767-3898
504/343-6671
                                     A-5

-------
Clare E. Brown
General  Coordinator of Operations
Pennsylvania Power Co.
1 E. Washington St.
New Castle, Pennsylvania  16103
412/656-5408

F. William Brownell
Lawyer
Hunton & Williams
1919 Pennsylvania Ave.,  N.W.
Washington, D.C.  20036
202/223-8650

Charles K. Bruhl
Engineering Manager
Chiyoda Int. Corp.
13uu Park Place blag.
Seattle, Washington  98101
2U6/624-9350

C.  P- Brundrett
Manager, Market Development
W.  R. Grace & Co.
10  E. Baltimore St.
Baltimore, Maryland  21203
301/659-9192

Doug Brusseau
Senior Mechanical Engineer
Salt River Project
P.O. Box 1018
St. Johns, Arizona  85536
602/337-4131

David P. Burford
Research Engineer
Southern Company Services
P.O. Box 2625
Birmingham, Alabama  35202
205/870-6329

Eugene A. Burns, PhD
Vice President & Manager
S-CUBED
3398 Carmel Mountain Rd.
San Diego, California  92121
619/453-0060
Edbert Buter
Project Engineer
N.V. KEMA
Utrechtseweg 310
Arnhem, NETHERLANDS
3185-457057 ext. 3424

Vernon Butler
Senior Environmental Engineer
U.S. EPA
6th and Walnut St.
Philadelphia, Pennsylvania
215/597-3697

Rob Byrne
Product Manager
Research Cottrell
P.O. Box 1500
Somerville, New Jersey  08876
201/685-4244

Ivor E. Campbell
Consultant
Ivor E. Campbell & Assoc.
P.O. Box 153
New Albany, Ohio  43054
614/855-2183

Bernard S. Camponeschi
Industry Manager
FMC Corp.
231 N. Martingale Ra
Schaumburg, Illinois 60194-2098
312/843-1700

W. R. Cares
Senior Research Chemist
M. W. Kellog Co.
16200 Park Row
Houston, Texas  77084
713/492-2500

Teri L. Carter
Process Engineer
Conoco, Inc.
Ponca City, Oklahoma  74603
4U5/767-3131
                                    A-6

-------
Douglas Carter
Engineer
U.S. Department of Energy
10UO Independence Ave.
Washington, D.C.  24585
202/252-4770

Ray Centa
Sales Application Engineer
Pultrusions Corp.
1331 S. Chillicothe Rd.
Aurora, Ohio  44202
216/562-5201

Denise I. Cessna
Program Coordinator
Pennsylvania Electric Co.
1001 Broad St.
Johnstown, Pennsylvania  15907
714/533-6667

Pui Kun Roland Chan
Research Analyst
University of Texas at Austin
University of Texas
Austin, Texas  78712
512/471-4851

John C. S. Chang
Senior Chemical Engineer
Acurex Corp.
P.O. Box 13109
Research Triangle Park, NC  27709
919/549-8915

Richard Chapman
Senior Engineer
Cor Star
2118 Milvia St.
Berkeley, California  94704
415/540-4100

Art Chappie
Technical Sales Representative
ABCO Plastics, Inc.
45 Accord Park
Norwell, Massachusetts  02061
617/878-5068
Gregory H. Cheng
Director/R&D
Ducon Co., The
147 E. 2nd Street
Mineola, New York  11740
516/741-6100
Suhas T. Chitnis
Manager, Gaseous
General Electric
200 N. 7th St.
Lebanon, Pennsylvania
717/274-7231
Collection Products
Environmental  Services
      17042
Ove B. Christiansen
Niro Atomizer, Inc.
9165 Rumsey Rd.
Columbia, Maryland  20145
301/997-8700

Roger Christman
Consulting Engineer
Christman & Assoc.
11708 Bowman Green Dr.
Reston, Virginia  22090
703/435-3219

M. Yaqub Chughtai
Project Manager
L&C Steinmueller GmbH
P.O. Box 100855
Gummersbach, West Germany
00261-85 2930

Darryl D. Ciliberto
Engineer
Tampa Electric Co.
P.O. Box 111
Tampa, Florida  33601
813/228-4111

Gerald M. Clancy
Vice President, Process Development
Pritchard Corp.
4625 Roanoke Pkwy.
Kansas City, Missouri  64112
816/531-9500
                                     A-7

-------
Curtis W. Clarkson
President
Clarkson Company
3430 W. Bayshore
Palo Alto, California
416/494-1010
94303
Donald D. Clasen
Marketing Manager
Chiyoda  International Corp.
1300 Park Place Bldg.
Seattle, Washington  98101
206/624-9350

Larry Cliver
Regional Sales Manager
Feeco International
3913 Algoma Rd.
kreen Bay, Wisconsin  54301
414/468-1000

Robert  Cmiel
Power Engineer
San Miguel Electric
Box 280
Jourdanton, Texas  78026
512/784-3411

W. James Cole
Program Director
New York State Energy
   Research Authority
2 Rockefeller Plaza
Albany,  New York   12223
518/465-6251

Ed Coleman
President
Herbst  & Assoc.
P.O.  Box 90989
Houston, Texas
 713/440-6090

J. David Colley
Radian  Corp.
P.O.  Box 9948
Austin, Texas   78766
512/454-4797
Marshall F. Conover
Program Manager
Radian Corp.
P.O. Box 9948
Austin, Texas  78766
512/454-4797

Charles B. Cooper
Senior Staff
Arthur D. Little,  Inc.
Acorn Park
Cambridge, Massachusetts  02140
617/864-5770

Frederick M. Coppersmith
Manager, Advanced  Fossil Fuels
Consolidated Edison Co. of N.Y.
4 Irving Place
New York City, New York  10003
212/460-3098

John Coulston
Sales Engineer
Spraying Systems Co.
North Ave. at Schmale Rd.
Wheaton, Illinois  60188
312/665-5000

Bob Cowen
President
Martek, Inc.
85 Research Rd.
hingham, Massachusetts  02043
617/749-6992

Maxwell E. Cox
Senior Technical  Service Representative
Kerr-McGee Chemical Corp.
Kerr-McGee Center
Oklahoma City, Oklahoma  73102
405/270-3373

A. J. Cozza
Manager, Ash Systems Engineering
Combustion Engineering
1000 Prospect Hill Rd.
Windsor, Connecticut  06095
203/688-1911
                                     A-i

-------
Anthony J. Craig
Sales Representative
Anthony J. Craig, Mfg. Rep.
10 Park Place
Butler, New Jersey  07405
201/838-8997

George Cranston
Peaboqy Process Systems
835 Hope St.
Stamford, Connecticut  06907
203/327-7000

William R. Cress
Manager, Engineering Studies
Allegheny Power Service Corp.
800 Cabin hill Dr.
Greensburg, Pennsylvania  15601
412/838-6721

Laird Crocker
Group Supervisor
U.S. Bureau of Mines
729 Arapeen Dr.
Salt Lake City, Utah  84108
801/524-6151

Phillip B. Crommelin, Jr.
Consultant
P.O. Box 38
Stanton, New Jersey  08885
201/236-2324

Richard V. Cross
Supervising Engineer, Mechanical Design
Union Electric Co.
P.O. Box 149
St. Louis, Missouri  63166
314/554-2671

James L. Crowe
Projects Manager
Tennesse Valley Authority
1150 CST2
Chattanooga, Tennessee  37401
615/751-5651

John Cunic
Senior Staff Engineer
Exxon Research & Engineering Co.
P.O. Box 101
Florham Park, New Jersey  07932
201/765-6471
Bob Cunningham
Supt. Production
City Water, Light and Power
7th & Monroe
Springfield, Illinois  62757
217/789-2238

Michael J. Cyran
FMC Corp.
P.O. Box 8
Princeton, New Jersey  08501
609/452-2300

Peter W. Dacey
Team Member, EAS
IEA Coal Research
14115 Lower Grosvenor PI.
London,
England

William C. Daley
Chief Engineer
Virginia Electric & Power  Co.
7th & Franklin Streets
Richmond, Virginia  26661
804/771-6269

George Dal ton
Project Designer
B&R Engineering
620 Wilson Avenue
Downsview, Ontario, Canada 1
416/630-7741

Stuart Dal ton
Project Manager
Electric Power Research Inst.
P.O. Box 10412
Palo Alto, California  94303
415/855-2467

Ashok S. Damle
Research Engineer
Research Triangle Inst.
P.O. Box 12194
Research Triangle Park, NC  27709
919/541-5829
                                     A-9

-------
Earl  H.  Darrough
Senior Sales Representative
Cabot Corp.
1020 West Park Ave.
Kokomo,  Indiana  46901
317/456-6000

Steve Davidson
Project Engineer
brown and Caldwell
1501 M.  Broadway
Walnut Creek, California  94596
415/937-9010

Lawrence Davidson
FGD Specialist
Stone & Webster
245 Summer St.
Boston, Massachusetts  02107
617/589-2381

R. H. Davis
Professor & Chairman
Florida State University
Tallahassee, Florida  32306
904/644-2867
Richard L. Davis
Manager of Engineering
Koppers Co. Inc.
2701 Koppers Building
Pittsburgh, Pennsylvania
412/227-2610
     15219
H. Joel Dean
Sales Manager
Mosser Damper Co.
500 Tilghman St.
Allentown, Pennsylvania
215/395-4900
    18104
Ronald N. Deardoff
Senior Engineer
Dayton Power & Light
P.O. Box 1247
Dayton, Ohio  45401
513/224-6374
Co.
 Edward D. Deboer
 Senior Application Engineer
 General Electric Co.
 607 Tall an Building
 Chattanooga, Tennessee  37402
 615/755-5011

 R. Dean Delleney
 Program Manager
 Radian Corp.
 3401 La Grande
 Sacremento,  California  95823
 916/421-8700

 Christian Demeter
 Assistant Director
 Reyes  Assoc.
 1633 16th Street N.W.
 Wshington, D.C.   20009

 J.  Herbert Dempsey
 Project Manager
 Acurex Corp.
 P.O. Box  13109
 Research  Triangle  Park,  N.C.   27709
 919/549-6915

 Charles  E. Dene,,
 Project Manager
 Electric  Power Research  Inst.
 P.O. Box  10412
 Palo Alto, California  94303
 415/855-2425

 Mark S. Dershowitz
 Marketing Representative
 Shell  Oil Co.
 200 North Dairy Ashford
 Houston,  Texas  77079
 713/670-2835

 Prakash H. Dhargalkar
Manager, Process Engineering
Research-Cottrell
P.O. Box 15uO
Somerville, New Jersey  08876
201/685-4295
                                     A-10

-------
N. N. Dharmarajan
Senior Engineer, Emissions Control
Central  & Southwest Services, Inc.
P.O. Box 220164
Dallas,  Texas  75222
214/754-1373

Jim Dickerman
Senior Program Manager
Radian Corp.
P.O. Box 9948
Austin,  Texas  78766
512/454-4797

William L. Didden
Regional Sales Manager
Conversion Systems, Inc.
115 Gibraltar Rd.
horsham, PA  19044
215/441-5920

David L. Dillehay
Product Line Manager
KVb, Inc.
18006 Skypark
Irvine,  California  92714
714/250-6259

Antonio J. DoVale Jr.
Senior Process Engineer
M. W. Kellogg Co.
433 Hackensack Ave.
Hackensack, New Jersey  07601
201/646-1000

Mark L.  Doane
Generation Sales Engineer
General  Electric Co.
1015 Locust St.
St. Louis, Missouri  63101
314/342-7727

Ir. N. A. Doets
Project Manager
N. V. PNEM
P.O. Box 7
4930 AA Geertruidenberg,
Holland
01621-82542
Temple E. Donaldson
Plant Manager
Central Illinois Light Co.
300 Liberty St.
Peoria, Illinois  61602
309/672-5271

J. R. Donnelly
Assistant Chief Nuclear, Enviornmental
  Engineer
Bechtel Power Corp.
P.O. Box 2166
Houston, Texas  77252-2354645

Van Dostveen
Expert
Ministry Environmental Protection
P.O. Box 450
Leidschendam, Netherlands  2260 MB
70/209367

Angelo Dounoloulos
Project Development Manager
General Electric Co.
P.O. Box 8
Schenectady, New York  12301
518/385-9980

Marty Downey
District Marketing Manager
Pullman Power Products
270 McCarty Dr.
Houston, Texas  77001
713/672-2491

William Downs
Research Specialist
Babcock & Wilcox Co.
1562 Beeson
Alliance, Ohio  44601
216/821-9110

Dennis C. Drehmel
Limb.  Dev.  Branch  Chief
U.S. EPA
IERL,  MD-61
Research Triangle  Park,  NC   27711
919/541-7505
                                    A-ll

-------
F. Carter Dreves
Director - Marketing Communications
Wheelabrator Air Pollution Control Div
600 Grant St.
Pittsburgh, PA  15219
412/288-7325

Charles J. Drummond
Chemical Engineer
DOE (PETC)
P.O. Box 10940
Pittsburgh, Pennsylvania  15236
412/676-6011

Michael J. Du Bois
Project Engineer
Commonwealth Edison
P.O. Box 767
Chicago, Illinois  60690
312/294-8483

Thomas  Dudick
President
Duaick  Corrosion-Proof
576 E.  Highland Rd.
Macedonia, Ohio  44056
216/467-1970

Ronald  G.  Duffy
Peabody Process Systems
835 Hope St.
Stamford,  Connecticut  06907
203/327-7000

David M. Dunkle
Manager, Marketing Services
Conversion Systems,  Inc.
115 Gibraltar Rd.
Horsham, Pennsylvania  19044
215/441-5920

Antonia Duran-Lopez
Chief Environment Department
Empresa Nacional De
   Electricidad, S.A.
Velazquez, 132
f'iadria-6,  Spain
(91)2.61.63.00
 Roger W. Dutton
 Partner
 Black & Veatch
 P.O. Box 8405
 Kansas City, Missouri
 913/967-7265
 64114
 John Dydo
 Santa Fe Energy Co.
 10737 Shoemaker
 Santa Fe Springs,  California

 Gene H.  Dyer
 Manager, Process Technology
   Department
 Bechtel  National,  Inc.
 P.O.  Box 3965
 San  Francisco,  California  94119
 415/768-4201

 John  M.  Ebrey
 Vice  President  - Marketing  & Planning
 Lodge-Cottrell  Operations
 601  Jefferson
 Houston,  Texas   77002
 713/750-2094
Martha  Edens
Dow  Chemical
P.O. Box  150
Plaquemine, Louisiana
504/389-8000
70764
Richard Egan
Manager, Marketing Opertions
Munters Corp.
P.O. Box 6428
Ft. Meyers, Florida  33911
813/936-1555

Joan Ehzeny
Reporter
BNA
1231 25th Street NW
Washington, D.C.
202/452-4423

Henry W. Elder
Assistant Director
Tennessee Valley Authority
501 Chemical Engineering Bldg.
Muscle Shoals, Alabama  35660
205/386-2514
                                    A-12

-------
M. Fred Ellis
Research Economist
Illinois Energy Resources Commission
3U3 Stratton Office Bldg.
Springfield, Illinois  62706
217/782-8220

William Ellison
Director
Ellison Consultants
4966 Tall Oaks Dr.
Monrovia, Maryland  21770
301/86b-53U2

Jack E. Englick
Sales Manager
FMC Corp.
231 N. Martingale Rd
Schaumburg, Illinois  60194-2098

Donald A. Erdman
Manager, Project Engineering
PEPCO
1900 Pennsylvania Ave., N.U.
Washington, D. C.  20068
202/872-3487

Olav Erga
Professor, Chemical Engineering Dept.
Norway Inst. of Tech.
Sem Saelandsv 4
Trondheim
Norway N-7034-NTH Tr. heim
07/594120

Douglas A. Erickson
Operations Engineer
Getty Oil Co.
Box 197X Rt. 1
Bakersfield, California  93308
805/399-2961

Robert L. Eriksen
Environmental Control Supervisor
Basin Electric Power Coop.
1717 E. Interstate Ave.
Bismarck, North Dakota  58501
701/223-0441, ext. 2144
Michael Esche
President, Dipl. -Ing., -Wittsch-ing,
Saarberg-Holter Umwelttechnik GmbH
Hafenstr. 6
D-6600 Saarbruken, West Germany
01681/32-105-06

Thomas F. Evans
Senior Research Specialist
Niagara Mohawk
300 Erie Blvd. W.
Syracuse, New York  13202
315/474-1511

Brent L. Evans
Division Manager
Stebbins Engineering and
  Manufacturing Co.
4831 North River Rd.
Port Allen, Louisiana  70767-3898
504/343-6671

Rita E. Ewing
Environmental Supervisor
Utah International  Inc.
550 California St.
San Francisco, California  94560
415/774-2363

Paul S. Farber
Project Manager
Argonne National Lab.
9700 S. Cass Ave.
Argonne, Illinois  60439
312/972-2000

Bobby Faulkner
Associate Director
Allis-Chalmers Corp.
P.O. Box 512
Milwaukee, Wisconsin  53201
414/475-4624

Thomas J. Feeley, III
Research Scientist
DOE, PGh Energy Res. Ctr.
626 Cochran Mill Rd.
Pittsburgh, Pennsylvania  15236
412/675-6079
                                     A-13

-------
Karsten Felsvang
Niro Atomizer Inc.
Oakland Ridge Industrial  Cntr.
9165 Rumsey Rd.
Columbia, Maryland   21045
301/997-8700

Bryan W. Ferguson
Chemical Engineer
Texas Power and Light
P.O. Box 226331
Dallas, Texas  75266
214/748-5411

John M. Ferraro
Senior Staff Engineer
Exxon Research & Engineering Co.
P.O. Box 101
Florham Park, New Jersey  07932
201/765-1600

Forjtenlehner
Project Manager
Veest-Alpine
Linz, Austria A-hono
0732/585-8626

John R. Field, Jr.
Senior Contract Administrator
Virginia Electric and Power Co.
P.O. Box 26666
Richmond, Virginia   23261
804/771-3625

Joe Finer
Regional Manager
Joy Manufacturing Co.
11707 E. 51st Ave.
Denver, Colorado  80239
303/371-6140

P.  G. Finlay
Head, Electric Power Section
Environment Canada
Ottawa, Ontario. Canada K1A 108
819/997-1220

Gerry Flander
Senior Sales Representative
Cabot Corp.
1800 Place Dunant,  St. Bruno
Quebec, Canada J3V  242
514/653-3312
Timothy 0. Flora
Special Assignment
Ohio Edison Co.
76 S. Main St.
Akron, Ohio  44308
216/364-7959

Ray Fmaddalone
Program Manager
TRW
One Space Park
Redondo Beach, California  90278

Gerald F. Foley
Principal Engineer
burns and Roe, Inc.
185 Crosswaus Park Dr.
Woodbury, New Jersey  11797
516/677-2274

Robert Forbus
Project Manager
Central & Southwest Services
P.O. Box 220164
Dallas, Texas  75222
214/754-1373

Jerry Ford
Manager
Custom Pipe Coating
P.O. Box 3274
Houston, Texas  77253
713/675-2324

George C. Ford
Proprietor
Environmental  Management Assoc.
607 Rosedale Rd.
Princeton, New Jersey  08540
609/924-0601

Owen F. Fortune
Project Manager DFGD
General Electric Co.
200 N. 7th St.
Lebanon, Pennsylvania  17042
717/274-7223
                                     A-14

-------
Landon D. Fox
head Mechanical Engineer
Tennessee Valley Authority
4UO W. Summit Hill
Knoxville, Tennessee  37902
615/632-362:4

Grant Frame
Manager Scrubber
Flakt Canada, Ltd.
P.O. Box 5060, STNF
Ottawa, Ontario
Canada K2C3p9
613/226-3300

Norman W. Frank
President
Ebara International Corp.
1 Northgate Square
Greensburg, Pennsylvania  15601
412/832-1200

Thomas Frank
Manager of Proposals
GE Environmental Services
200 North 7th St.
Lebanon, Pennsylvania  17042
717/274-7146

George Fraunfelder
Eastern Regional Sales Manager
Komline-Sanderson Engineering Corp.
12 Holland Ave.
Peapack, New Jersey  07977
2U1/234-1000

killiam F. Frazier
Staff Engineer
Virginia Electric & Power Co.
P.O. Box 564
Richmond, Virigina  23204
804/771-6147

Paul E. Fredette
Manager, Technical Center/Program
  Marketing
Midland Ross Corp.
900 North Westwood
Toledo, Ohio  43696
419/537-6426
Mark D. Freeman
Supervisory Engineer
Southwestern Public Service Co.
P.O. Box 1261
Annville, Texas  79170
806/378-2184

Donald T. Freese
Research Department
Betz Laboratories, Inc.
Somerton Rd.
Trevose, Pennsylvania  19047
215/355-3300

Steven Frey
Environmental Engineer
U.S. EPA
1860 Lincoln St.
Denver, Colorado  80013
303/837-6131

Dan Froelich
Manager, Air Quality Design
Burns & McDonnell  Engineering Co.
P.O. Box 173
Kansas City, MO  64141

Joseph L. Gaines
Staff Engineer
Rust International Corp.
P.O. Box 101
Birmingham, Alabama  35201
205/254-4472

Ignatius J. Gallo
Assistant Manager, Market Research
Texasgulf Chemicals Co.
P.O. Box 30321
Raleigh, NC  27622-0321
919/829-2810

Marsh Galloway
Manager, FGD Marketing
Ceil cote Company
140 Sheldon Rd.
Berea, Ohio  44017
216/243-0700
                                     A-15

-------
P. R. Gambarani
Manager
General Electric Environmental  Services
5 Penn Plaza
New York, New York  10001
212/613-3203

Bernie Gardey
Enviornmental Manager
C.F. Braun & Co.
R.D. *1
Bangor, Pennsylvania  18013
717/897-5148

John Gaynor
Research Staff Member
United States Gypsum
700 N. Highway 45
Libertyville, Illinois  60048
312/362-9797
Dennis C. Gehri
Program Manager
Rockwell Int.
8900 De Soto Ave.
Canoga Park, California
213/700-4413
91304
D. B. Geottel
Supervisor, FGD Systems
Winyah Generating Station
Santee Cooper
P.O. Box 1275
Georgetown, South Carolina  29440
803/546-4171

Mario A. Gialanella
Regional Sales Manager
Joy Manufacturing Co.
7 Corporate Park Dr.
White Plains, New York  10604
914/694-1364

Elizabeth D. Gibson
Consultant
Arthur D. Little, Inc.
Acorn Park
Cambridge, Massachusetts  02140
617/664-5770
Carl A. Gilbert
President
Dravo Lime Co.
3600 Neville Rd.
Pittsburgh, Pennsylvania  15225
412/777-5559

Tom Gil Christ
Chemist
Colorado-Ute Electric
P.O. Box 1307
Craig, Colorado  81626

Al Giovanetti
Supervising Engineer
Davy McKee Corp.
0. 0. Drawer 5000
Lakeland, Florida  33803
613/646-7311

Dennis L. Glancy
Superintendent FGD Operations
Southern Indiana Gas & Electric
P.O. Box 569
Evansville, Indiana  47741
812/424-6411

Robert J. Gleason
Director Research & Development
Research-Cottrell, Inc.
Somerville, New Jersey  06876
201/685-4884

Donald E. Glowe
Senior Engineer
Texas Research Inst.
9063 W. Bee Caves Rd.
Austin, Texas  78746
512/263-2101

Robert F. Goecker
Eastern Region Sales Manager
Marcona Ocean Industries, Ltd.
1001 N.W. 62nd St., Suite 200
Ft. Lauaerdale, Florida  33309
305/776-4000
                                     A-16

-------
Klaus Goldschmidt
Environmental Protection Dept.
VEBA Kraftwerke Ruhr AG
Bergmannskluckstr- 41-43
D-465U Gelsenkirchen
Federal Republic of Germany D-4650
0209/601-5851

Richard to. Goodwin
Principal
Environmental Engineering Consultant
261 W. 71 St.
New York, New York  10023

Steve R. Graham
Project Manager
Betz Laboratories
3657 Buckskin Tl. W.
Jacksonville, Florida  32211
904/744-0053
Richard J. Grant
Manager, Environmental Affairs
Central Illinois Public Service
607 East Adams St.
Springfield, Illinois  62701
Co,
Jack Greene
Administrative Officer
U.S. EPA IERL RTP
IERL-RTP (MD-60)
Research Triangle Park, NC  27711
919/541-2903

Donald Gross
Supervisory General Engineer
Cdr, Cml R&D Cen
Attn:  DRSMC-CLT-I (A)
APG, Maryland  21010
301/671-4102

K. 0. Groves
Research Manager
Dow Chemical U.S.A.
2020 Dow Center
Midland, MI  48640
517/636-3246

Larry Gruber
Generation Sales Engineer
General  Electric Co.
607 Tallan Bldg., 2 Union Square
Chattanooga, Tennessee  37402
615/775-5009
           Klaus E.  Gude
           Niro Atomizer
           9165 Rumsey Rd.
           Columbia,  Maryland
           301/997-8700
                    20145
Dave Guetig
Assistant Plant Superintendent
Inaaianapolis Power & Light Co.
P.O. Box 436
Petersburg, Indiana  47567
812/354-8801

D. Guidetti
Business Development Manager
SOIMI SPA
5100 Westheimer
Houston, Texas  77056
713/961-0873

Connie Guilbeau
Sales Represenative
Xerox Corp.
417-419 Baronne St.
New Orleans, Louisiana  70112
504/568-9015

Navin K. Gupta
Principal FGD Engineer
Stone & Webster Engineering Corp.
3 Executive Campus
Cherry Hill, New Jersey  08034
609/482-4246

C. Richard Hach
Engineer
Tampa Electric Co.
P.O. Box 111
Tampa, Florida  33601
813/677-9141
           Albert Hackl
           University Professor
           Institut fur Verfahrenstechnik
             Tecnischen Universitat Wien
           A-1060 WIEN,  Getreidemarkt 9
           Austria
           0222/5601-4726
                               der
                                    A-17

-------
Donald K.  Hagar
President
Damper Design,  Inc.
P.O. Box 2045
Bethlehem, Pennsylvania  18001
215/861-0111

Qazi Haider
Chemical Engineer
Industrial Generating Co.
P.O. Box 1111
Rockdale,  Texas  76567
512/446-5861

F. Allen Hall
Field Salesman
Cabot Corp.
4650 S. Pinemone, Suite 130
Houston, Texas  77041
713/462-2177

J. A. Hall
Manager, Henderson Technical Lab.
TIMET
P.O. Box 2128
Henderson, Nevada  89015
702/564-5831

hark Hal pern
Project Licensing Coordinator
Potomac Electric Power Co.
1900 Pennsylvania Ave., N.W.
Washington, D.C.  20068
202/331-6489

David Ham
Manager
Physical Sciences, Inc.
P.O. box 3100, Research Park
Andover, Massachusetts  01810
617/475-9030

Joseph J. Hammond
Fbb Sales Engineer
Zurn Industries
P.O. Box 2206
Birmingham, Alabama  35201
205/252-2181
Doug Hammontree
Mgr. Air Quality Design Section
Burns & McDonnell Engineering Co.
P.O. Box 173
Kansas City, MO  64141
816/333-4375

Svend Keis Hansen
Vice President
Niro Atomizer, Inc.
9165 Rumsey Rd.
Columbia, Maryland  20145
301/997-8700

Frank M. Harbison
Asst. Environmental Analyst
Louisiana Power & Light
317 Barone
New Orleans, Louisiana  70130
504/595-2308

0. W. Hargrove, Jr.
Senior Staff Engineer
Radian Corp.
805 Mopac Blvd.
Austin, Texas  78759
512/454-4797

John Harkness
Chemical Engineer
Argonne National Lab.
EES Division, Bldg. 362
Argonne, Illinois  60439
312/972-7636

Richard C. Harrington
Senior Engineer
Cleveland Electric Illuminating
P.O. box 5000
Cleveland, Ohio  44101
216/622-9800
Thomas L. Hart
Chemical Engineer
American Electric
1 Riverside Plaza
Columbus, Ohio
614/223-3472
Power Service
                                     A-18

-------
John L. Haslbeck
Engineer
NOXSo Corp.
2625 H. C. Mathis Dr.
Paaucah, Kentucky  42001
502/444-6474

John hattrup
Meteorologist
Baltimore Gas & Electric Co.
116 N. Howard St.
Baltimore, Maryland  21201
301/234-6427

Larry E. Haulter
Sales Representative
Cabot Corp.
1020 W. Park Ave.
Kokomo, Indiana  46901
317/456-6073

John R. Hawksworth
Supervisor, Operations Support
Virginia Electric and Power Co.
P.O. Box 26666
Richmond, Virginia  23261
804/771-4633

Kent D. Hedrick
Environmenal Enginer
Kentucky Utilities Co.
One Quality St.
Lexington, Kentucky  40507
606/255-1461 ext. 542

Carl-Rudolf Hegemann
General Manager
G. Bischoff
baertner Str 44, D-4300
Essen, W. Germany
0207-233037

Glen 0. Hein
Vice President of Sales
Marblehead Lime Co.
300 West Washington St.
Chicago, Illinois  60606
312/263-4490
Michael  Heisel
Process  Engineer
Linde
6023 Hoellriegelskrenth
F.R. Germany

Rex Helfant
Manager,  Generation Programs
Con Edison
4  Irving  Place
New York, New York  10003
212/460-3987

Paul D. Hemphill
Scrubber  Project Manager
Dresser Industries
601 Jefferson, 27th Floor, Dressor Tower
Houston,  Texas  77002
713/972-6011

A. Hennico
Technical Sales Manager
Institut  Francais Du Petrole
1-4 Avenue de Bois Preau
Rueil  Malmaison  925u6
France

Robert A. Hentges
Senior Engineer
Procter & Gamble Co.
6105 Center Hill Rd.
Cincinnati, Ohio  45224
513/659-5787

David S.  Henzel
Technical Services Engineer
Dravo Lime Co.
3600 Neville Rd.
Pittsburgh, Pennsylvania  15225
412/777-5556

William D. Herrin
Supervisor, Air
Alabama Power Co.
P.O.  Box 2641
Birmingham, Alabama  35291
205/250-4124
                                     A-19

-------
Robert L. Hershey
Vice President
Science Management Corp.
2101 L St., M.W., Suite 903
Washington, D.C.  20037
202/293-5700

H. Frederick Hess
Operations Manager
FMC Corp.
231 N. Martingale Rd.
Schaumburg, Illinois  60194-2098
312/843-1700

R. A. Hewitt
Senior Engineer
Texas Utilities
2001 Bryan
Dallas, Texas  75201
214/653-4800

Barry R. Hickenbottom
Mechanical Engineer
U.S. Navy-NEESA
Code 111C
Port Hueneme, California  93043
b05/982-5292

Wayne Hickok
Senior Production Engineer
Cooperative Power Assoc.
14615 Lone Oak Rd.
Eden Prarie, Minnesota  55344
612/937-8599

Kent Higgins
Vice President
Koch Process Systems
20  Walkup  Dr.
Westboro,  Massachusetts  01581
617/366-9111

Richard  Hills
Manager  of Research
Pittsburgh Des Moines
Neville  Island
Pittsburgh, Pennsylvania  15225
412/331-3000, ext. 655
Robert G. Hilton
Director of Programs Management
Monier Resources, Inc.
45 N.E. Loop 410, Suite 700
San Antonio, Texas  78216

Maurice Hixon
Plant Superintendent
Board of Muncipal Utilities
138 N. Prarie
Sikeston, Missouri  63801
314/471-5000

Otto H. Hoegberg
Chief Engineer
MW Kellogg
433 Hackensack Ave.
Hackensack, New Jersey  07601
201/646-1000

Art Hoekstra
President
Sirco
4412 Aicholtz Road
Cincinnati, Ohio  45245
513/752-4700

Jerry Hoffman
Director, Project Development
Burns & McDonnell Engineering Co.
P.O. Box 173
Kansas City, MO  64141
816/333-4375

David C. Hoffman
Vice President, Technical  Services
Chemical Lime, Inc.
6000 Western PI., Suite 489
Ft. Worth, Texas  76107
817/732-8164

Kent R. Hoffman
Chemical Engineer
Public Service Co. of New Mexico
Alvarado Square
Albuquerque, New Mexico  87158
505/848-2964
                                     A-20

-------
T. W. Hojnicki
Manager
Argonne National Laboratory
97UO South Cass Ave.
Argonne, Illinois  60439
312/972-5821

Gerald A. Hoi linden
Branch Chief
Tennessee Valley Authority
113U Chestnut St., Towers II
Chattanooga, Tennessee  37401
615/751-3584

Jack Hoi lister
Senior Vice President
Cleveland Cliffs Iron Co.
Huntington Bldg.
Cleveland, Ohio  44115-1448
216/241-2356

Richard G. Hooper
Project Manager
Electric Power Research Inst.
P.O. Box 1U577
Denver, Colorado  80210
303/936-7281

William M. Horton
Chemical Engineer
Lower Colorado River Authority
P.O. Box 220
Austin, Texas  78767
512/473-3544

Robert W. Hospodarec
Environmental Director
Fluor Engineers, Inc.
2310 Kelvin
Irvine, California  92712
714/966-5073

John B. Howard
Manager, Power Production/Construction
Alabama Electric Coop., Inc.
P.O. Box 550
Andalusia, Alabama  36420
205/222-2571
Bill Hughes
Arkansas Lime Co.
P.O. Box 2356
Batesville, Arkansas  72501
501/793-2301

George A. Hugick
Sales Engineer
Kennedy Van Saun Corp.
R.R. St.
Danville, PA  17821
717/275-3050

James Hung
Senior Environmental Analyst
Cajun Electric Power
10719 Airline Hwy.
Baton Rouge, Louisiana  70816
504/291-3060

Joseph Hunt
Market Development Manager
Allegheny Ludlum Steel
Research Center
Brackenridge, Pennsylvania  15014
412/226-2000

James E. Hunt
President
Concord Scientific Corp.
2 Tippett Road
Downsview, Ontario  M3H 2V2
416/630-6331

Thomas B. Hurst
Manager Product Development
Babcock & Wilcox
20 S. Van Buren Ave.
Barberton, Ohio  44203
216/860-2674

Howard Hurwitz
Manager, Process Engineering
General Electric Environ Svs.
5 Penn Plaza
New York, New York  10001
212/613-3175
                                      A-21

-------
Robert C. Hyde
F.G.D. Product Manager
Joy Manufacturing Co.
4565 Colorado Blvd.
Los Angeles, California  90039
213/240-2300 ext. 233

Akitaka Ide
Project Manager
Chiyoda International Corp.
1300 Park Place Bldg.
Seattle, Washington  98101
206/624-9350

Mr. hiaeo Idemura
Managing Director
Chiyoda International Corp.
1300 Park Place Bldg.
Seattle, Washington  96101
206/624-9350

W.  Ijdeveld
ESMIL International
De  Boelelaan 7
Amsterdam, Netherlands
020/541-1054

Paul A. Ireland
Manager, Environmental Control
Stearns-Roger
P.O. Box 5888
Denver, Colorado  80217
303/692-3420

Marty W. Irwin
Research Director
Indiana Department of Commerce
One North Capitol, Suite 700
Indianapolis, Indiana  46204-2248
317/232-8818

Sam Jacobsson
Western Regional Corp.
P.O. Box 6428
Ft. Meyers, Florida  33911
813/936-1555
Bryan J. Jankura
Research Engineer
Babcock & Wilcox
1562 beeson
Alliance, Ohio  44601
216/821-9110 ext. 391

James B. Jarvis
Radian Corp.
P.O. Box 9948
Austin, Texas  78766
512/454-4797

Clarence B. Jeffcoat
Manager, Cross Generating Station
Santee Cooper
P.O. Box 98
Cross, South Carolina  29436
803/351-4586

Frits Jellema
General Manager, APC Division
ESMIL (Intl.)
De Boelelaan 7
Amsterdam, Netherlands
020-5411054

Stephen D. Jenkins
Senior Engineer
Tampa Electric Co.
P.O. Box 111
Tampa, Florida  33601
813/28-4111

Sarah Jenkins
Planning Analyst
Wisconsin Public Service Commission
P.O. Box 7854
Madison, Wisconsin  53707
608/266-5990

Robert M. Jensen
Engineering Specialist
Bechtel Power Corp.
50 Beale St.
San Francisco, California  94105
415/768-1323
                                     A-22

-------
Bob Jewell
District Manager
Research Cottrell
13231 Champion Forest Dr.
Houston, Texas  77459
713/440-0468

Dennis Johnson
Field Process Engineer
Babcock & Wilcox
20 South Van Buren Avenue
Barberton, Ohio  44203
216/860-6324

David I. Johnson
Executive Director
Coal Technology
10703-A Stand iff
Houston, Texas  77099
713/879-8929

Robert M. Johnson
Group Manager, Research
Gold Bond Building Products
1650 Military Rd.
Buffalo, New York  14217
716/873-9750

Lyndon Johnson
Air Quality Supervisor
Muscatine Power and Water
3205 Cedar St.
Muscatine, Iowa  52761
319/263-2631

Howard J. Johnson
Engineer
Ohio EPA
361 E. Broad St.
Columbus, Ohio  43216
614/466-6116

Carl ton A. Johnson
Peabody Process Systems
835 Hope St.
Stamford, Connecticut  06907
203/327-7000
Robert P. Johnson
Research Projects Manager
Pennsylvania Power & Light Co.
2 N. 9th St.
Allentown, Pennsylvania  18101
215/770-5151

C. W. Johnston
Marketing Mgr., Special Products
Badische Corp.
602 Copper Rd.
Freeport, Texas 77541
409/238-6237

Martin A. Jones
Environmental Engineer
Cliffs Engineering,  Inc.
P.O. Box 1211
Rifle, Colorado  81650
303/625-2445

Robert L. Jones
Power Research Sales Engineer
General Electric Co.
4410 El Camino Real
Los Altos, California  94022
415/949-1042
J. R. Jones
Vice President
Gitford-Hi 11  &
P.O. Box 47127
Dallas, Texas  75247
214/258-7330
 Marketing
Co., Inc.
S. M. Jones
Marketing Manager
MPSI
Box M312
York, Pennsylvania
717/843-8671
     17402
Julian W. Jones
Senior Chemical Engineer
U.S. EPA
IERL, MD-61
Research Triangle Park, NC  27711
919/541-2489
                                      A-23

-------
Larry Jones
Environmental  Engineer
U.S. EPA, OAQPS, ESED, SDB
MD-13
Research Triangle Park, N.C,
919/541-5624
       27711
James I. Joubert
Branch Chief
U.S. Department of Energy
P.O. Box 10940
Pittsburgh, Pennsylvania  15236
412/675-5716

Peter Judersleben
General Manager
Knauf/Research-Cottrel1
Iphofen,
West Germany

John H. Juzwiak
Manager, Utility Operations
Conversion Systems Inc.
115 Gibraltar Rd.
Horsham, Pennsylvania
215/441-5900

Yohi Kameoka
Chiyoda Int.
1300 Park Place
Seattle, Washington  98101
206/624-9350
Joel Y. Kamya
Project Engineer
Boston Edison Co.
800 Boylston St.
Boston, Massachusetts
617/424-3250
02199
 Toshio Kanai
 Technology and Engineering Division
 Chiyoda  International Corp.
 1300 Park Place Bldg.
 Seattle, Washington  98101
 206/624-9350

 Ann Marie Kanon
 Program  Secretary
 Electric Power Research Institute
 P.O. Box 10412
 Palo Alto, California
 415/855-2466
Ira E. Kanter
Senior Engineer
Westinghouse R&D Center
1310 Beulah Rd.
Pittsburgh, Pennsylvania  15235
412/256-5808

Marilyn Kaplan
Editor
Mcllvaine Co.
2970 Maria Ave.
Northbrook, Illinois  60062
312/272-0010
                      Steven M. Kaplan
                      Niro Atomizer
                      9165 Rumsey Rd.
                      Columbia, Maryland
                      301/997-8700
                    21045
Norman Kaplan
Chemical Engineer
U.S. EPA
IERL, MD-61
Research Triangle Park, NC  27711
919/541-2556

Steve M. Katzberger
Supervisor, Emission Control
Sargent & Lundy Engineers
55 E. Monroe
Chicago, Illinois  60603
312/269-6672

Ronald H. Kaye
Peabody Process Systems
835 Hope St.
Stamford, Connecticut  06907
203/327-7000

T. J. Kayhart
Manager of Sr. Projects
Allied Chemical Co.
P.O. Box 1087 R
Morristown, New Jersey  07960
201/455-4927

Mike Keckritz
Design Engineer
Illinois Power
500 S. 27th St.
Decatur, Illinois  62525
217/424-6962
                                     A-24

-------
James R. Kendle
Regional Manager
FhC Corp.
231 N. Martingale Rd.
Schaumburg, Illinois  60194-2098
312/843-170U

Thomas R. Kendrick, III
President
Pittsburgh Environmental Systems, Inc.
67 Old Clairton Rd.
Pleasant Hills, Pennsylvania  15236
412/653-75UU

Ernest E. Kern
Staff Engineer
Houston Lighting & Power
P.O. Box 1700
Houston, Texas  77001
713/481-7608

Rick Kesler
Senior Applications Engineer
Mine & Smelter
3600 Race
Denver, Colorado  80205
303/296-8700

James R. Kessling
Senior Engineer
Houston Lighting & Power Co.
12301 Kurland Dr.
Houston, Texas  77034
713/481-7921

John W. Kife
(aroup Vice President
Joy Manufacturing Co.
4565 Colorado Blvd.
Los Angeles, California  90068
213/240-2300 ext. 500
         Kilgroe
         Coal Cleaning
James D.
Manager,
U.S. EPA
IERL, MD-61
Research Triangle Park, NC  27711
919/541-2854
Lawrence P- King
Senior Marketing Specialist
Babcock & Wilcox
1562 Beeson St.
Alliance, Ohio  44601
216/821-9110

Minesh Kinkhabwala
Commercial Manager
Thyssen Environmental Systems, Inc.
333 Meadowland, Pkwy.
Secaucus, New Jersey  07094
201/330-2600

William R. Kins
Supervisor, Air Systems Eng.
Owens Corning Fiberglass
Toledo, Ohio  43659
419/248-6027

Dennis L. Kirchner
Air Quality Engineer
Central 111. Public Service Co.
607 East Adams St.
Springfield, IL  62701
217/523-3600

Noel W. Kirshenbaum
Manager, Mineral Projects
Placer U.S. Inc.
1 California St.
San Francisco, California  94111
415/986-U740

Jonas Klingspor
M. S. Chemical Engineer
Dept. of Chemical  Engineering
P.O. Box 240, S-22007
LUND, SWEDEN
OU46-46-108277

Peter Klose
ERM, Inc.
999 W. Chester Pike
West Chester, Pennsylvania  19380
215/696-9110

G. H. Koch
Projects Manager
Battelle
505 King Ave.
Columbus, Ohio  43201
614/424-4480
                                     A-25

-------
Bernard J. Koch
Technical Coordinator
Conoco Coal Research Div.
4000 Brownsville Rd.
Library, Pennsylvania  15129
412/854-6612

Y. Kogawa
Managing Director
Chiyoda International Engineering
1300 Park Place Bldg.
Seattle, Washington  98101
206/624-9350

M. Koike
Chiyoda International Engineering
1300 Park Place Bldg.
Seattle, Washington  98101
206/624-9350

Nubuo Kojima
Process Engineer
Mitsubishi
520 Madison Ave.
New York, New York
212/605-2006

Ken Kondo
Project Engineer
Mitsubishi
520 Madison Ave.
New York, New York
212/605-2006

Dennis Kostick
Physical Scientist
U.S. Bureau of Mines
2401 E. St.
Washington, D.C.  20241
202/634-1177

Karl M. Kozak
Marketing Specialist
Rockwell Intl., Energy Systems Group
8900 De Soto Ave.
Canoga Park, California  91304
213/700-4013
Brandon P. Krogh
Plant Chemical Engineer
Minnesota Power
30 W. Superior St.
Duluth, Minnesota  55802
218/772-2641, ext. 3311

J. Lee Krumme
President and CEO
Vinings Chemical Co.
2555 Cumberland Pkwy., Suite 2000
Atlanta, Georgia  30339
404/436-1542

Tony Ku
Manager of Technical Services
Cadre Environmental Systems
2845 Clearview Place
Doraville, Georgia  30340
404/458-9527

John Kuehl
Utility Sales Supervisor
Warman International Inc.
2701 S. Stoughton Rd.
Madison, Wisconsin  53716
608/221-2261

Peter M. Kutemeyer
General Manager
Bischoff Environmental Systems
135 Cumberland Rd.
Pittsburgh, Pennsylvania  15237
412/364-8860

Charles R. LaMantia
President
Koch Process Systems, Inc.
20 Walkup Dr.
Westborough, Massachusetts  01581
617/366-9111 ext. 541

Dale Ladd
Regional Manager
Martek Inc.
85 Research Rd.
Hingham, Massachusetts  02043
                                      A-26

-------
G. C. Lammers
Senior Research Engineer
Amoco Chemicals Corp.
Box 400 (C-7)
Naperville, Illinois  60566
312/420-5642
          Land
          Technology Assessment
George W.
Director,
AMAX Coal
105 S. Meridian
Indianapolis, Indiana  46225
317/266-2812

Wes Langeland
Vice President
Martek, Inc.
85 Research Rd.
Hingham, Massachusetts  02043
617/749-6992

William Mike Lankford
Superintendent, Technical Services,
  Cross Generating Station
Santee Cooper
P.O. Box 98
Cross, South Carolina  29436
803/351-4586

Ellen E. Lanum
Conference & Travel Supervisor
Electric Power Research Institute
P.O. Box 10412
Palo Alto, California  94303
415/855-2193

Bernard A. Laseke
Group Supervisor
PEDCo Environmental, Inc.
Chester Towers
11499 Chester Rd.
Cincinnati, Ohio  45246
513/782-4700

Dennis Laslo
Senior Development Engineer
Peabody Process Systems
835 Hope St.
Samford, Conn.  06840
203/327-7000
Peter Lawson
Supervising Engineer, Energy and
  Studies Development
Ontario Hydro
700 University Ave.
Toronto, Ontario
Canada  M4K1A1
416/592-5393

George 0. Layman
Director, Power Supply
Gulf Power Co.
P.O. Box 1151, 75 N. Pace Blvd.
Pensacola, Florida  32520
904/434-8354

John R. Lee
Field Applications Engineer
BF Goodrich
P.O. Box 1010
Tuscaloosa, Alabama  35403
205/752-1521

George C. Y. Lee
Engineering Specialist
Bechtel Group, Inc.
50 Beale St.
San Francisco, California  94119
415/768-3119

Chiun-Chieh Lee
Technical Manager
Union Chemical Laboratories
1021 Kuang Fu Rd.
Hsinchu, Taiwan
Republic of China
035/713131

Yung-Li Lee
Research Assistant
University of Texas at Austin
E.P. Schoch
Austin, Texas  78712
512/471-4851

T. S. Lee III
Research Manager
LaQue Ctr. for Corrosion Techn.,
Auditorium Circle & Hwy. 76
Wrightsville Beach, NC  28480
919/256-2271
                                                                        Inc.
                                     A-27

-------
Jan Lefers
Dr.
Nv Kema
Utretseweg 310
Arnhem, The Netherlands

L. Karl Legatski
Manager Process Tech.
FMC Corp.
231 N. Martingale Rd.
Schaumburg, Illinois 60194-2098
312/843-1700

Jurgen Leimkuhler
Division Manager
G. Bischoff
Gaertner Str 44, D-4300
Essen, W. Germany
0201-233031

Charles C. Leivo
Advanced Product Planning Manager
Dresser Industries
2408 Timberlock, Blag. C
The Woodlands, Texas  77380
713/367-7355

Arnold L. Leriche, Jr.
Environmental Engineer
U.S. EPA, Region I
JFK Federal Bldg.
Boston, Massachusetts  02203
617/223-5137

Clifford J. Lewis
Director of Environmental Services
National Lime Assoc.
3601 N. Fairfax Dr.
Arlington, Virginia  22201
303/237-2948
Jack A. Li bey
Director of Power
City of Lakeland
1000 E. Parker St.
Lakeland, Florida
813/687-3636
Production
 33802
W. A. Liegois
Process Engineer
Stanley Consultants
Stanley Bldg.
Muscatine, Iowa  52761
319/264-6457

Tom Lillestolen
Manager, FGO Systems
Flakt, Inc.
P.O. Box 87
Knoxville, Tennessee  37901
615/693-7550

Nilo Lindgren
R&D Applications Assessment
Electric Power Research Inst.
P.O. Box 10412
Palo Alto, California  94303
415/855-2753

Darol Lindloff
Regional Sales Manager
Wheelabrator APC Div.
12201 Merit Dr.
Dalls, Texas  75251
214/458-8738

Stephen A. Lingle
Chief Technology Branch
U.S. EPA (WH565)
401 M. Street, S.W.
Washington, D.C.  20460
202/382-7917

Michael R. Lintelman
Sales Engineer
Dravo Lime Co.
3600 Neville Rd.
Pittsburgh, Pennsylvania  15225
412/777-5584

C. David Livengood
Environmental Systems Engineer
Argonne National Laboratory
9700 S. Cass Ave.
Argonne, Illinois  60439
312/972-3737
                                     A-28

-------
George C. W. Logan
Energy Applic. Eng.
Middle South Services,
P.O. Box 61000
New Orleans, LA  70161
569-4725
Inc.
Frank Logeat
National Sales Manager
Bachmann Industries, Inc.
29 Lexington St.
Lewiston, Maine  04240
207/784-2336

W. Z. Looney
Project Planning Engineer
Southern Company Services
P.O. Box 2625
Birmingham, Alabama  35202
205/877-7289

D. Antonio Duran Lopez
Empresa Nacional de Electricidad,
Velazquez 132, Madrid
Spain
91/261-63 00

David P. Lovetere
Resident Construction Manager
Burns & McDonnell
Box 1179
Wheatland, Wyoming  82201
307/322-9530

Phillip S. Lowell
Consulting Chemical Engineer
P.S. Lowell & Co., Inc.
4107 Medical Pkwy., #214
Austin, Texas  78756
512/451-3513

Thomas H. Lucy
Northeast Regional Manager
Research - Cottrell
Box 1500
Somerville, New Jersey  08876
201/685-4479
           S.A.
John R. Ludwig
Senior Research Engineer
U.S. Steel Corp.
Research Lab
Coleraine, Minnesota  55722
218/245-2200

Louis A. Luedtke
Central Reg. Manager
Niro Atomizer, Inc.
49U1 College Blvd.
Leawood, KS  66211
913/341-3953

Mario Luperi
Sales Manager, Industrial
  Products Division
Wlm. Steinen Mfg.  Co.
29 E. Halsey Rd.
Parsippany, New Jersey  07054
201/887-6400

Robert D. Lupi
Product Manager, Special  Alloys
Jessop Steel Co.
Jessop Place
Washington, Pennsylvania  15301

Bruce MacDonald
Manager, Sales and Marketing
ABCO Plastics, Inc.
45 Accord Park
Norwell, Massachusetts  02061
617/878-5068

Colin MacDonald
Vice President
Filtres Gaudfrin
4950 Highland Dr.
Salt Lake City, Utah  84117
801/278-2851

Gordon C. MacDonald
Vice President
Mitsubishi International Corp.
520 Madison Ave.
New York, New York  10022
212/605-2006
                                     A-29

-------
Richard Madenburg
Technical Director
htorrison-Knudsen Co., Inc.
P.O. Box 7808
Boise, Idaho  83729
2:08/386-6069

Gary A. Maier
Project Manager
Florida Power and Light Co.
P.O. Box 14000
Juno Beach, Florida  33408
305/663-3608

Jason Makansi
Associate Editor
Power Magazine/McGraw-Hill
122:1 Ave. of the Americas
New York City, New York  10020
212/997-4239

John E. Makar
Regional Manager
FMC Corp.
231 N. Martingale Rd.
Schaumburg, Illinois  6U194-2098
312/843-1700

L. Marnewecke
Process Control Engineer
Armco Autometrics
4946 N. 63rd St.
Boulder, Colorado  80301
303/530-1600

Patrick Maroney
Supervisor
Brown and Calowell
1501 N. Broadway
Walnut Creek, California  94596
415/937-9010

Gregory Martin
Product Engineer
Dresser Industries, Inc.
5323 S. Western Blvd.
Chicago, Illinois  60609
312/471-3040
Peter G. Maurin
Director, FGD Systems
Wheelabrator-Frye,  Inc.
600 Grant St.
Pittsburgh, Pennsylvania
412/288-7323
15219
Michael A. Maxwell
Chief Emissions/Affluent Tech. Branch
U.S. EPA
IERL, MD-61
Research Triangle Park, NC  27711
919/541-2578

T. J. May
Supervisor, Short Range Planning
Illinois Power
500 S. 27th St.
Decatur, Illinois  62525
217/424-6706

Manville J. Mayfield
Branch Chief
Tennessee Valley Authority
1010 Chestnut St., Towers II
Chattanooga, Tennessee  37401

William N. McCarthy
U.S. EPA
401 M. St., S.W.
Washington, D.C.  20460
202/382-2625

Thomas A. McClellan
Production Engineer
Public Service of Indiana
P.O. Box 1009
Mt. Carmel, Illinois  62863
812/386-8491 ext. 580

Charles J. McCormick
Vice President
Dravo Lime Co.
3600 Neville Rd.
Pittsburgh, Pennsylvania  15225
412/777-5553

Michael W. McElroy
Project Manager
Electric Power Research Inst.
3412 Hi 11 view Ave.
Palo Alto, California  94303
415/855-2471
                                     A-30

-------
Edward T. McHale
Manager, Combustion & Physical
  Sciences Dept.
Atlantic Research Corp.
5390 Cherokee Ave.
Alexandria, Virginia  22312
703/642-4088
Robert Mcllvane
President
Mcllvane Co.
2970 Maria Ave.
Northbrook, Illinois
312/272-0010
60062
Marilyn Mcllvane
Editor
Mcllvane Co.
2970 Maria Ave.
Northbrook, Illinois
312/272-U010
60062
John D. McKenna
President
ETS, Inc.
Suite C-103, 3140
Chaparral Dr., SW
Roanoke, Virginia   24018
703/774-8999

D. M. McLane
District Manager
MPSI
Sox M312
York, Pennsylvania   17402
717/843-8671

Brennan McLaughlin
Supervisor  Process  Engineering
United Engineers &  Const.
23 Inverness Way, E.
Englewood,  Colorado  80126
303/79-7310

Joseph McNamara
Manager, Field Services
Thyssen Environmental  Systems, Inc.
333 Meadow!and Pkwy.
Secauscus,  New Jersey 07094
201/330-2600
                Michael  L.  Meadows
                Mechanical  Engineer
                Black &  Veatch
                P.O.  Box 8405
                Kansas City,  Missouri
                913/967-2189
                       64114
Wayne E. Meadows
Black & Veatch
1500 Meadow Lake Pkwy.
Kansas City, Missouri  64114
913/967-2643

Michael Melia
Environmental Engineer, Project Manager
Peaco Environmental, Inc.
11499 Chester Rd.
Cincinnati, OH  45246
513/782-4877

Heinz Merlet
Dip! .-Ing.
Lurgi Unwelt un Chemotechnik GmbH
Gervinusstr. 17/19, D-600
Frankfurt, Federal  Rep. of Germany
611/157228b

Douglas Merrill
Senior Marketing Manager
Church & Dwight Co., Inc.
20 Kingsbridge Rd.
Piscataway, New Jersey  08854
201/885-1220

Frank B. Meserole
Radian Corp.
P.O. Box 9948
Austin, Texas  78766
512/454-4797

Amelia Mesko
Engineer, Generating Engineering
Potomac Electric Power
1900 Pennsylvania Ave.
i.i-v.~Li^»i^i4-x-,k-.  n r   or if tiv 1
                Washington,  D.C.
                202/872-3575
                  20068
                                        N.W.
                Ronald G.  Metz
                Manager, Market Development
                Mixing Equipment Co.
                135  Mt.  Read Blvd.
                Rochester,  New York  14603
                716/436-5550
                                     A-31

-------
S. D. Meyer
Senior Engineer
Combustion Engineering
1000 Prospect Hill  Rd.
Windsor, Connecticut  06095
2:03/668-1911

Chris F. Meyer
Waste Management Systems Division
Koch Process Systems
20 Walkup Dr.
Westboro, Massachusetts  01581
617/3b6-9111

Sheldon Meyers
Director, Office of Air Quality
  and Standards
U.S. EPA
401 M St., S.W.
Washington, D.C.  20460
202/655-4000

John H. Michael
Chief Executive Officer
Princeton Chemical
1810 24th Street N.W.
Washington, D.C.  20008
202/483-0063

Stephen Michel
Project Engineer
PEPCO
1900 Pennsylvania Ave., N.W.
Washington, D.C.  20068
202/872-2436

Samy R. Mikhail
Generating Engineer
PEPCO
1900 Pennsylvania Ave., N.W.
Washington, U.C.  20068
202/872-3487

Robert Miller
Field Sales Representative
Cabot Wrought Products
1020 West Park Ave.
Kokomo, Indiana  46901
317/456-6000
John 0. Milliken
Project Manager
U.S. EPA
IERL, MD-61
Research Triangle Park, NC  27711
919/541-7716
Robert Mingea
Vice President
Lodge-Cottrell
601 Jefferson
Houston, Texas
713/750-2089
- Sales
Operations
 77002
John Minnick
Consultant
Box 271
Plymouth Meeting,
   Pennsylvania  19462
J. E. Mirabel la
Chiyoda International Engineering
1300 Park Place
Seattle, Washington  98101
206/624-9350

P. R. Misra
Utility Sales Manager
Taulman Sales Co.
3312 Piedmont Rd., N.E.
Atlanta, Georgia  30305
404/261-2535

J. David Mobley
Environmental Engineer
U.S. EPA
IERL, MD-61
Research Triangle Park, NC  27711
919/541-2350

Karsten Moller-Jansen
Niro Atomizer, Inc.
9165 Rumsey R.
Columbia, Maryland  21045
301/997-8700

Virginia M. Moon
Development & Testing Engineer
Combustion Engineering Inc.
31  Inverness Center Pkwy.
Birmingham, Alabama  35209
205/967-9100
                                    A-32

-------
Max A. Moore
Vice President, Equipment Division
KVB, Inc.
Irvine, California  92714
714/250-6310

Keith Moore
Marketing
Rockwell Int.
6900 Desoto
Canoga Park, California  91304
213/700-4016

Thomas Morasky
Project Manager
Electric Power Research Inst.
P.O. Box 10412
Palo Alto, California  94303
415/655-2468

Terri E. Morel and
Manager, Coal Development Programs
Illinois Dept. of Energy &
  Natural Resources
325 ]/t. Aaams, Rm. 300
Springfield, Illinois  62706
217/765-2600

David S. Morey
Supervisor, Market Research
Allied Corp.
1411 Broadway
New York, New York  10018
212/391-5149

Robert M. Morford
Vice President, Marekting
Joy Manufacturing Co.
4565 Colorado Blvd.
Los Angeles, California  90039
213/240-2300

Wayne E. Morgan
Manager or Pollution Control
Black & Veatch
P.O. Box 8405
Kansas City, Missouri  64114
913/967-2198
James C. Morgan
Engineer, Mechanical Design
Union Electric Co.
P.O. box 149
St. Louis, Missouri  63166
314/554-2771

Jack Morgenstern
Equipment Engineer
Stone & Webster Engineering
1 Penn Plaza, 250 W. 34th St.
New York, New York  10119
212/290-6668

Michael D. Morris
Generation Equipment Sales Eng.
General Electric Co.
10550 Barkely
Overland Park, Kansas  66212
913/967-6272

Per E. Morsing
Niro Atomizer A/S
9165 Rumsey Rd.
Columbia, Maryland  20145
301/997-8700

Robert E. Moser
Senior Engineer
Bechtel Power Corp.
50 Beale St.
San Francisco, California  94119
415/768-9055

G. A. Mountford
Director of Licensing, Chemicals
Pfizer, Inc.
235 E. 42nd St.
New York, New York  10017
212/573-7405

Hugh Mullen
Director, Government Relations
Conversion Systems, Inc.
115 Gibraltar Rd.
Horsham, Pennsylvania  19044
215/441-5900
                                     A-33

-------
Theo Mullen
Reporter
McGraw Hill
1221 Avenue of the Americas
New York, New York  10020
212/899-9416

Don Muller
Sales Manager
Black River Lime Co.
P.O. Box 1
butler, Kentucky  41006
606/472-7761

Fred Mulloy
Environmental Manager
Phillips Petroleum Co.
7 D 4 Phillips Bldg.
Bartlesville, Oklahoma  74004
916/661-5735
Edward J. Muren
District Manager
Research-Cottrell
85 W. Algonquin Rd.
Arlington Heights,
312/228-6228
Illinois   60005
Andrew J. Murphy
Office Manager
Acurex
3200 Nelson Chapel Hill Hwy.
Research Triangle Park, N.C.  27709
919/549-8915

John Murray
Dow Chemical
P.O. Box 150
Plaquemine, Louisiana  70764
504/389-8000

E. G. Murray
Supervisor, Plant Equipment
  ana Systems
Southern Company Services
P.O. Box 2625
Birmingham, Alabama  35202
205/870-6855
John Mycock
Vice President
ETS, Inc.
Suite C-103
3140 Chaparral Dr., Ski
Roanoke, Virginia  24018
703/774-8999

Amir Nassirzadeh
Design Mechanical Engineer
L.A. Dept. of Water & Power
111 N. Hope St., Rm. 661
Los Angeles, California  90012
213/481-4647

Lewis G. Neal
President
Noxso Corp.
2625 H. C. Mathis
Paducah, Kentucky  42001
502/444-6474

Derek Neely
Pittsburgh Des Moines
Neville Island
Pittsurgh, Pennsylvania  05225
412/331-3000

James A. Nelson
Marketing Manager
Effox, Inc.
10921 Reed Harman Hwy.
Cincinnati, Ohio  45242
513/793-1932

Hans Neukam
Sales Manager
Kraftanlagen Heidelberg
Im Breitspiel 7
D6900 Heidelberg,
West Germany
06221/394378

Thomas H. Newhams
Peabody Process Systems
835 Hope St.
Stamford, Connecticut  06907
203/327-7000
                                     A-34

-------
Joseph T. Newman
Engineering Specialist
Bechtel Group, Inc.
bO Beale St.
San Francisco, California
415/768-3114

T. W. Newton
Senior Vice-President
MPSI
Box M312
York, Pennsylvania  17402
717/843-8671
Yen Nguyen
Engineer
Ontario Hydro
800 Kipling Ave.
Toronto, Ontario,
416/231-4111
         94105
Canada  M8Z 5S4
 Arthur F. Nicholson
 President
 Kentucky Export Resources Authority
 Suite 1505, Vine  Center
 Lexington, Kentucky   40507
 606/233-3545

 Irvin P- Nielsen
 Chairman and C.E.O.
 Nathona Resources,  Inc.
 1600 Broadway, Suite  1120
 Denver, Colorado  80202
 303/839-1600

 Kurt R. Nielsen
 Mineral Economist
 National Resources  Inc.
 1600 Broadway, Suite  1120
 Denver, Colorado  80202
 0303/839-1600

 Gary Niles
 Proauct Manager
 Pathway Bellows,  Inc.
 P.O. Box 1526
 El  Cajon, California   92022
 619/440-1300
Kimio Nishio
Technology and Engineering Division
Chiyoda Intl. Corp.
1300 Park Place Blag.
Seattle, Washington  98101
206/624-9350

jack Noble
Chief Mechanical
C. T. Main
101 Huntington
Boston, Massachusetts  02199
617/262-3200
Heinrich Novak
Chairman
Grosskraftwerk
Nurnberg,
West Germany
                                      Franken AG
                       David M. Novick
                       Product Manager, Utility Env.
                       Combustion Engineering Inc.
                       1000 Prospect Hill  Ra.
                       Winsor, Connecticut  06070
                       205/285-9243

                       John R. Null
                       Manager, Environmental Systems
                       Babcock & Wilcox
                       P.O. Box 61038
                       New Orleans, Louisiana  70161
                       504/587-5636

                       Jack F. O'Donnell
                       Chairman & Chief Executive Officer
                       Advanced Energy Dynamics Inc.
                       14 Tech Circle
                       Natick, Massachusetts  01781
                       617/653-8112

                       John O'Donnell, Jr.
                       Director of Market Development
                       High Performance Tube, Inc.
                       1460 Morris Ave.
                       Union, New Jersey  07083
                       201/964-8520
                                     A-35

-------
Donald O'Hair
Senior Sales Engineer
Joy Manufacturing Co.
1301 k. 22nd Street
Oak Brook,  Illinois  60521
312/654-4090

Robert D. O'Hara
Environmental Engineer
Duquesne Light Co.
One Oxford  Centre, 27-2
301 Grant St.
Pittsburgh, Pennsylvania  15279
412/393-6098

Andrew T. O'Neill
Peabody Process Systems
835 Hope St.
Stamford, Connecticut  06907
203/327-7000

Gary Ochs
Program Manager
York Research Consultants
938 Quail St.
Denver, Colorado  80215
303/233-1513

Leif Olausson
Swedish State Power Board
S-16287 Vaellingby,
Sweden

David G. Olson
Manager, Utility Marketing
General Electric
200 N. 7th  St.
Lebanon, Pennsylvania  17042
717/274-7355

H. Onuma
Manager
Mitsubishi  International  Corp.
520 Madison Ave.
New York, New York  10022
212/605-2663

Sidney Orem
Executive Director
Industrial  Gas Cleaning Inst.
700 North Fairfax St.,  #304
Alexandria,  Virginia  22314
703/836-0480
C. L. Osborne
Santee Cooper
P.O. Box 98
Cross, South Carolina  29436
803/351-4586

Ronald G. Ostendorf
Senior Engineer
Proctor & Gamble Co.
7162 Reading Rd.
Cincinnati, Ohio  45222
513/763-4457

Norman Ostroff
Supervisor, Process Engineering
Peabody Process Systems
835 Hope St.
Stamford, Connecticut  06097
203/327-7000

Delbert M. Ottmers
Assistant Vice President
Radian Corp.
8501 Mopac Blvd.
Austin, Texas  78759
512/454-4797

Buck Oven
Administrator Power Plant Siting
Florida Dept. Env. Regulations
2600 Blair Stone Rd.
Tallahassee, Florida  32301
904/488-0130

Peter Overaick
Establissements Leon Lhoist S.A.
21 Avenue Rogier
Liege,
Belgium B-4000
817/732-8164

David R. Owens
University of Texas at Austin
University of Texas
Department of Chemical Engineering
Austin, Texas   78741
512/471-4851
                                     A-36

-------
Michael A. Ozol
Consultant
Michael A. Ozol Ph.D.
24u3 Ken Oak Ra.
Baltimore, Maryland  21209
301/664-2565

Robert Steven Pace
Bio Environmental Engineer
City of Jacksonville
515 W. 6th St.
Jacksonville, Florida  32206
904/633-3303

Michael A. Palazzolo
Radian Corp.
3024 Pickett Rd.
Durham, North Carolina  277U5
919/493-4574

Knut Papajewski
Manager, Operations
Thyssen Environmental Systems, Inc.
333 Meadow!and Pkwy.
Secaucus, New Jersey  07094
201/330-2600

Rashmi Parekh
Resident Sales Manager, Eastern Region
Dorr Oliver, Inc.
274 Riverside Ave.
Westport, Connecticut  06880
203/358-3800

Jeffrey H. Parker
Marine Science Research Cntr
SUMY
Stony Brook, New York  11794
516/246-5000

Richard W. Patton
President
VFL Technology Corp.
42 Lloyd Avenue
Malvern, Pennsylvania  19355
215/296-2233

R. L. Pearce
Associate Scientist
Dow Chemical Co.
B1605 Bldg.
Freeport, Texas  77541
409/239-1419
Michael Perlsweig
Program Manager
U.S. Department of Energy
FE-23
Washington, D.C. 20545
301/353-4399

Dennis Perrone
Regional Manager
Martek Inc.
1630 Newell Avenue
Walnut Creek, California  94596
415/937-5630

Mogens Petersen
President
Niro Atomizer, Inc.
9165 Rumsey Rd.
Columbia, Maryland  20145
301/997-8700

Vincent Petti
Manager, Application Engr. FkD Sys.
Wheelabrator-Frye Inc.
600 Grant Street
Pittsburgh, Pennsylvania  15219
412/288-7465

Brian R. Phelan
Regional Manager
Joy Manufacturing Co.
4901 College Blvd.
Shawnee Mission, Kansas  66211
913/648-8783

Peter H. Phillips
Manager, I&C
Whellarrator Air Pollution Control
101 Merrit-7
Norwalk, Connecticut  06856
203/852-6846

Allan R. Pike
Consulting Engineer
Self Employed
76 Seir Hill Road
Wilton, Connecticut  06897
203/762-8990
                                     A-37

-------
James E. Pilgrim, Jr.
Section Supervisor
Tennessee Valley Authority
100 IBM Building
Chattanooga, Tennessee  37401
615/751-4455

George Pinheiro
Sales
Research-Cottrell, Inc.
P.O. box 1500
Somerville, New Jersey  08876
201/685-4109

T. Duane Pinson
Southern Region Sales Manager
Marcona Ocean Industries, Ltd.
1001 No. W. 62th Street, Suite 200
Fort Lauderdale, Florida  33309
305/776-4000

Bill Piske
Manager S/W Office
Engineering Science
13740 Midway Drive, Suite 706
Piano, Texas  75075
214/392-0695

Wallace S. Pitts III
Associate
Kilkelly Environmental Assoc.
P.O. Box 31265
Raleigh, North Carolina  27622
919/761-3150

James F. Plappert
Director of Sales
Conversion Systems, Inc.
115 Gibraltar Rd.
horsham, PA  19044
215/441-5920

Fred L. Porter
Section Chief
U.S. EPA, OAQPS, ESED, SDB
MD-13
Research Triangle Park, N.C.   27711
919/541-5624
Tom Potter
Director of Administration
Nat. Lab. Assoc.
3601 N. Fairfax Dr.
Arlington, Virginia  22201
703/243-5463

Geroge Powers
Engi neer-Generati ng Engi neering
Potomac Electric Power Co.
1900 Pennsylvania Ave. - Room 832
Washington, D. C.  2u068
202/331-6250

Paul R. Predick
Acting Head, Mechanical Analytical
Sargent & Lundy Engineers
55 E. Monroe St.
Chicago, Illinois  60603
312/269-6671

Jerry V. Presley
Senior Engineer
Virginia Electric & Power Co.
P.O. Box 564
Richmond, Virginia  23204
804/771-6148

Jack Preston
Senior Engineer
South Carolina Electric & Gas
P.O. Box 764
Columbia, South Carolina  29218
803/478-3849

Tom Priest
Vice President
Ceil cote
140 Sheldon Rd.
Berea, Ohio  44017
216/247-0770

Frank T. Princiotta
Director, Laboratory
U.S. EPA
IERL, MD-60
Research Triangle Park, NC  27711
919/541-2821
Div.
                                   A-38

-------
Edwin J. Puska
Production Systems
Marquette Board of Pwr. & Light
2200 bright St.
toarquette, Michigan  49855
906/226-6900

Irwin A. Raben
President
IAR Technology, Inc.
130 Sandaringham South
Moraga, California  94556
415/376-3951

Alan D. Randolph
Professor
University of Arizona
Tuscon, Arizona  85721
602/621-6051

Richard Rao
Manager, Air Quality Control
Ebasco Services, Inc.
160 Chubb Ave.
Lyndhurst, New Jersey  07071
201/460-1900

Fred Rapp
Sales Representative
Armco Inc.
7000 Roberts Kansas City
Kansas City, Missouri  64125
816/242-5452

Hulic Ratterree
Manager, Technical Services
Blount Energy Resource Corp.
4520 Executive Park Drive
Montgomery, Alabama  36116
205/277-8860

John Reilly
President
Thyssen Environmental Systems, Inc.
333 Meaaowland Pkwy.
Secaucus, New Jersey  07094
201/330-2600
Jack B. Reisdorf
Project Engineer
Stearns-Roger Engineering Corp.
P.O. Box 5888
Denver, Colorado  80217
303/692-3420

Bruce W. Remick
Director of Aglime Marketing
National Crushed Stone Assoc.
1415 Elliot Place, N.W.
Washington, D.C.  20007
202/342-1100

George Rey
Senior Staff Engineer
U.S. EPA
Washington, D.C.  20460
202/382-2626

Richard G. Rhudy
Project Manager
Electric Power Research Inst.
P.O. Box 10412
Palo Alto, California  94303
415/855-2421

David Z. Richards
Kennecty Van Saun Corp.
R. R. St.
Danville, Pennsylvania  17821

Phillip Richardson
Manager, Market Research & Development
Du Pont
1007 Market St.
Wilmington, Delaware  19898
802/774-4930

Ed Riordan
Supv. Waste Treatment
City Water, Light and Power
7th & Monroe
Springfield, Illinois  62757
217/769-2238

Charles J. Rizzo
Vice President/Engineering Service
Air Clean Damper Co.
Kugler Mill & Blue Ash
Cincinnatti, Ohio  45236
513/793-1253
                                     A-39

-------
Russell  F. Robards
Chemical Engineer
Tenn. Valley Authority
1160 Chestnut St., Tower II
Chattanooga, Tennessee  37401
615/751-5662

Gary T.  Roche!le
Assistant Professor
Dept. of Chemical Engineering
University of Texas
Austin,  Texas  78712
512/471-3434

Ron Rohlik
Mechanical Engineering Specialist
Getty Oil Co.
P.O. Box  197X Rt. 1
Bakersfield, California   93308
805/399-2961

Greg A.  Rollins
Sales Representative
Timet
726 Avenue R
Grand Prairie, Texas  75050
214/641-4410

Robert  N. Roop
Product Manager, Spray Dryers/Fabric
  Filters
Research-Cottrell, Inc.
P.O. Box  1500
Somerille, New Jersey  08876
201/685-4451

Kevin C.  Rorke
Manager,  Projects & Construction
FMC Corp.
1501 Woodfield Rd., Suite 300 E.
Schaumburg,  Illinois  60195
312/843-1700

Ruiz-Alsox Rosa
Research  Assistant
University of Texas at Austin
E.P- Schoch  Lab
Austin,  Texas  78712
412/471-4851
Edward C. Rosar
President
Industrial Resources, Inc.
300 Union Blvd., Suite 520
Lakewood, Colorado  80228
303/986-4507

Jean Thomas Rose
Project Supervisor
Betz Laboratories, Inc.
Somerton Rd.
Trevose, Pennsylvania  19047
215/355-3300, ext. 394
Norman A. Rosekrans
Regional Sales Manager
uoy Manufacturing Co.
5775 Peachtree DunwoocHy
Atlanta, Georgia  30319
404/256-2934
Rd., Suite 200E
Harvey S. Rosenberg
Senior Research Engineer
BatteHe-Columbus Lab.
505 King Ave.
Columbus, Ohio  43201
614/424-5010

Clifford A. Rosene
FGD Production Supervisor
No. Indiana Public Service Co.
Route 1 Box 320
Wheatfield, Indiana  46392
219/956-5180

Bud Ross
Industry Manager
Huntington Alloys, Inc.
P.O. Box 1958
Huntington, West Virginia  25720
304/696-3509

Edward S. Rubin
Professor
Carnegie-Mellon University
Schenley Park
Pittsburgh, Pennsylvania  15213
412/578-2491
                                     A-40

-------
Richard A. Runyan
Projects Manager
Tennessee Valley Authority
1160 Chestnut St., Towers II
Chattanooga, Tennessee  37401
615/751-5663

Randall E. Rush
Manager, Flue Gas Treatment
Southern Co. Services
F. Box 2625
Birmingham, Alabama  35202
205/870-6320

James R. Rutledge
Project Coordinator/Operations
Jacksonville Electric Authority
233 U. Duval St.
Jacksonville, Florida  32211
904/633-2220

Stephen B. Ryan
Sales Representative
Marblehead Lime Co.
300 W. Washington
Chicago, Illinois  60606
412/563-1812

George P. Sacco
Environmental Bus. Rep. Dev.
General Electric Env. Services
22 Perimeter Center Ene
Dunwoody, Georgia  30346
404/399-5517

A. Saleem
Vice Presiaent, Int'l. Bus. Dev.
General Electric Environmental
200 N. Seventh St.
Lebanon, Pennsylvania  17042
717/274-7171

Norman C. Samish
Staff Research Engineer
Shell Development Co.
P.O. Box 1380
Houston, Texas  77001
713/493-7944
Eric A. Samuel
Senior Development Engineer
General Electric Co.
200 North Seventh St.
Lebanon, Pennsylvania  17042
717/274-7048

C. J. Santhanam
Manager, Chemical Eng.
A. D. Little, Inc.
20 Acorn Park
Cambridge, Massachusetts  02140
617/864-5770

James S. Sarapata
Director of Process Development
Church & Dwight Co., Inc.
20 Kingsbridge Rd.
Piscataway, New Jersey  08854
201/885-1220

Fred J. Sauereisen
Vice President
Sauereisen Cements Co.
160 Gamma Drive
Pittsburgh, Pennsylvania  15238
412/963-0303

A. E. Saunders
Superintendent, Technical Services
Winyah Generating Station
Santee Cooper
P.O. Box 1275
Georgetown, South Carolina  29440
803/546-4171

Mark Schneider
Project Engineer
Delmarva Power
600 King Street
Wilmington, Delaware  19803
302/429-3616

Donald Schreyer
Peabody Process Systems
835 Hope St.
Stamford, Connecticut  06907
203/327-7000
                                      A-41

-------
Richard Schubert
Sales Manager
General Electric
200 N. 7th St.
Lebanon, Pennsylvania
717/274-7257
17042
David Schulz
Regional Power Industry Expert
U.S. EPA
230 S. Dearborn St.
Chicago, Illinois  60604
312/353-2088

R. W. Schutz
Supervisor, Corrosion Research
  and Development
TIMET, Division of Titanium
  Metals Corp.
P.O. Box 2128
Henderson, Nevada  89015
702/564-2544 - ext. 215

Richard A. Schwartz
President
D. R. Technology, Inc.
Hidden Pines Dr.
Clarksburg, New Jersey  08510
201/780-4664

Joseph A. Schwartz
Marketing Manager
KVB, Inc.
18006 Skypark
Irvine, California  92714
714/250-6258

Tonny C. Schytte
Niro Atomizer, Inc.
9165 Rumsey Rd.
Columbia, Maryland  20145
301/997-8700

William C. Seale
Mechanical Supervisor
Lower Colorado River Authority
P.O. Box 220
Austin, Texas  78767
512/473-3540
David 0. Seaward
Product Manager, FGD Systems
Wheelarrator Air Pollution Control
101 Merritt 7, P.O. Box 5440
Morwalk, Connecticut  06856
203/852-6884

Roland K. Seward
Process Engineer
Kennedy Van Saun Corporation
R. R. St.
Danville, Pennsylvania  17821
717/275-3050

D. Shack!ey
Mineral Resources Director
British Gypsum Ltd.
Gotham, Nottinghamshire
England
0602/830431

Raymond J. Shaffery
Director of Commercial  Development
Church & Dwight Co., Inc.
20 Kingsbridge Rd.
Piscataway, New Jersey  08854
201/885-1220

Navin D. Shah
Consultant
142 Sundance Ct.
Grand Junction, Colorado  81503
303/243-1503

Arvind M. Shah
Vice President
Spraco, Inc.
P.O. Box 3800
Nashua, New Hampshire  03060
603/888-1050

Don Shattuck
Process Engineer
Stearns-Roger Engineering
P.O. Box 5888
Denver, Colorado  80217
303/692-4139
                                     A-42

-------
Charles E. Shelton
Manager, Fossil & Hydro
  Operations Support
Viriginia Electric and Power Co.
P.O. Box 26666
Richmond, Virginia  23261
804/771-4441

Robert Shiely
Research Assistant
University of Texas at Austin
Austin, Texas  78712
512/471-4851

Christopher Shih
Senior Project Engineer
TRW
One Space Park 01/2040
Redondo Beach, California  90278
213/536-4105

G. H. Shroff
Engineering Specialist
Bechtel Power Corp.
15740 Shady Grove Rd.
Gaithersburg, Maryland  20877
301/258-3146

William E. Siegfriedt
Senior Mechanical Engineer
Fluor Engineers, Inc.
200 W. Monroe St.
Chicago, Illinois  60606
312/368-3828

Mark S. Siegler
Chief, Technical Support Branch
U.S. EPA
401 M. Street S.W.
Washington, D.C.  20460
202/382-2835

William L. Silence
Associate Engineer
Cabot Corp.
1020 W. Park Ave.
Kokomo, Indiana  46901
317/456-6201
P. L. Simiskey
Research Associate
Dow Chemical Co.
B1605 Bldg.
Freeport, Texas  77541
409/239-1419

A. P. Sandy Simko
Projects Engineering Manager
Arizona Public Service
P.O. Box 21666 (Station 5760)
Phoenix, Arizona  85036
602/271-7261
Michael Skinner
Engineer
Northern States Power
3100 Marshall St.
Minneapolis, Minnesota
612/33U-5991
Co.
  55418
John H. Skinner
Director, Office of Solid Waste
U.S. EPA
401 M St., S.W.
Washington, D.C.  20460
202/655-4000

A. V. Slack
President
SAS Corp.
Route 1, Box 69
Sheffield, Alabama  35660
205/383-1627

A. G. Sliger
Assistant Manager Process Design
M. W. Kellogg Co.
3 Greenway Plaza
Houston, Texas  77046

John E. Smigelski
Senior Engineer
New York State Electric & Gas Co.
4500 Vestal Pkwy.
East Binghamton, New York  13903
608/279-2551, ext. 4605
                                      A-43

-------
Earl 0. Smith
Project Manager
Black & Veatch
IbUU Meadow Lake Pkwy,
Kansas City, Missouri
913/967-2643
64114
Ted Smith
Director of Design
Burns & McDonnell Engineering Co,
P.O. Box 173
Kansas City, Missouri  64141
816/333-4375

Jeffrey D. Smith
Plant Reliability Engineer A
Ohio Edison Co.
76 S. Main Street
Akron, Ohio  44308

Peter V. Smith
General Sales Manager
Research-Cottrell, Inc.
P.O. Box 1500
Somerville, Mew Jersey  08876
201/685-4221

Robert H. Smith
President
Rob Smith Assoc., Inc.
5355 Knox St.
Philadelphia, PA  19144
215/843-2209

Roger Smith
Plant Results Engineer
South Mississippi Electric
  Power Assoc.
P.O. Box 1589
Hattiesburg, Mississippi  39401
601/268-2083

Norman B. Smith
Chief Chemical Engineer
Stanely Consultants, Inc.
Stanley Bldg.
Muscatine, Iowa  52761
319/264-6299
Robert Smock
Managing Editor
Electric Light & Power Magazine
P.O. box 1030
Harrington, Illinois  60010
312/381-1840

John V. Smolensk!
FGD Specialist
Stone & Webster Engineering
3 Executive Campus
Cherry Hill, New Jersey  08034
609/482-3594

Jeff Snapp
Project Engineer
Public Service Co. of Indiana
1000 E. Main St.
Plainfield, Indiana  46168
317/838-1656

Donald R. Snider
Cincinnati  Gas & Electric Co.
P.O. Box 107
Union, Kentucky 41091
606/586-5600

Robert E. Sommerland
Vice President
Foster Wheeler Develop. Corp.
12 Peach Tree Hill Rd.
Livingston, New Jersey  07039
201/533-3650

James P- Spellman
Senior Business Rep.
General Electric Co.
8101 Stemmons Freeway
Dallas, Texas  75248
214/688-6165

H. Michael  Spence
President
National Resources, Inc.
1600 Broadway, Suite 1120
Denver, Colorado  80202
303/839-1600
                                     A-44

-------
Herbert W. Spenler, III
Manager, Advanced Technology
Joy Manufacturing Co.
4b6b Colorado Blvd.
Los Angeles, California  90039
213/240-2300 ext. 426

Lawrence J. Stanislow
Assistant Power Industry Manager
Pennwalt Corp.
3 Parkway
Philadelphia, Pennsylvania  19102
215/587-7264

Michael J. Stapf
President
Mosser Damper Co.
5000 Tilghman St.
Allentown, Pennsylvania
215/395-4900

Glenn G. Stauffer
Project Engineer
Pennsylvania Power and Light Co.
Two North Ninth St.
Allentown, Pennsylvania  18101
215/770-6563

Harold Steeves
Vice President/General Manager
ABCO Plastics, Inc.
45 Accord Park
Norwell, Massachusetts  02061
617/878-5068

Edward W. Stenby
Assistant Manager, Engineering
Stearns-Roger Engineering Corp.
P.O. Box 5888
Denver, Colorado
303/758-1122

John G. Stensland
Marketing Manager
FMC Corp.
231 N, Martingale Rd.
Schaumburg, Illinois  60194-2098
312/843-1700
Frederick Stern
Plant Engineer
Basin Electric Power Coop.
P.O. Box 1059
Beulah, North Dakota  58523
701/873-4545

Richard D. Stern
Chief, LIMB Applications
U.S. EPA
IERL-RTP (MD-63)
Research Triangle Park, N.C.
919/541-2547
27711
Nicholas J. Stevens
Chemical Process Development Mgr.
Research-Cottrell
P.O. Box 1500
Sommerville, New Jersey  08876
201/685-4887

Jack F. Stewart
Product Specialist
Bablcock & Wilcox
20 S. Van Buren
Barberton, Ohio  44203
216/860-2118

Dorothy A. Stewart
Project Manager
Electric Power Research Inst.
P.O. Box 10412
Palo Alto, California  94303
415/855-2609

Joseph G. Stites
Vice President
Energy Sciences & Services, Inc.
203 Barley Mill Rd.
Old Hickory, Tennessee  37138
615/847-5031

Jim Stone
Resource Specialist
Louisiana ONR
P.O. Box 44066
Baton Rouge, Louisiana  70804
504/342-1206
                                     A-45

-------
Mark R. Stouffer
Associate Engineer
Conoco Coal  Research
4UUU Brownsville Rd.
Library, Pennsylvania
412/854-6639
15234
Don Stowe
Director/Tech Sales & Svc.
Dravo Lime Co.
36UO Neville Rd.
Pittsburgh, Pennsylvania  15225
412/777-5574

Joseph P. Strakey
Director, Process Technology Division
DOE/Pittsburgh Energy Technology Ctr.
P.O. Box 10940
Pittsburgh, Pennsylvania  15236
41Z/675-6125

John Strange
Manager of Mechanical Engineering
bibbs & Hill, Inc.
11 Penn Plaza
New York, New York  100U1
212/760-4162

Harry A. Straw
Product Development Consultant
E. I. Du Pont de Nemours and Co.
1007 Market St.
Wilmington, Delaware  19898
302/744-3673

Joseph J. Stuparich
Senior Business Representative
General Electric Co.
3 Penn Center Plaza
Philadelphia, Pennsylvania  19102
215/241-5240

Ben Y. Su
Environmental Engineer
United Engineers and Constructors
100 Summer St.
Boston, Massachusetts  02110
617/338-6000
Roger P. Summerhays
Product Marketing Engineer
EIMCO
P.O. Box 300
Salt Lake City, Utah  84110
801/526-2367

Karen V. Summers
Senior Hydrogeologist
Tetra Tech, Inc.
3746 Mt. Diablo, Suite 300
Lafayette, California  94549
415/283-3771

Carl E. Swanson
Sr. Metallurgical  Engineer
Newmont Exploration Ltd.
44 Briar Ridge Rd.
Danbury. Connecticut  06810
2U3/743-6784

Donald E. Syler, Jr.
Supervisory Engineer
Consumers Power Co.
1945 West Parnall  Rd.
Jackson, Michigan  49201
517/788-1946

Barry C. Syrett
Project Manager
Electric Power Research Inst.
P.O. Box 10412
Palo Alto, California  94303
415/855-2956

Atsushi Tatani
Process Engineer
Mitsubishi
520 Madison Ave.
New York, New York
212/605-2006

Pamela D. Taulbee
Editor
Coal Technology Report
1401 Wilson Blvd., Suite 910
Arlington, Virginia  22209
703/528-1244
                                    A-46

-------
Charles E. Taylor
University of Texas at
Department of Chemical
Austin, Texas  78712
512/471-4851
Austin
Engineering
Donald P. Teixeira
Senior Engineer
Pacific Gas & Electric
34uO Crow Canyon Rd.
San Ramon, California  94583

Jeffrey A. Telander
Environmental Engineer
U.S. EPA, OAQPS/ESED/TSB/SDS MD-13
Research Triangle Park, N.C.  27711
919/541-5595

Jack T. Thompson
Chief, Technical Services Branch
Tennessee Valley Authority
705 Edney Bldg.
Chatanooga, Tennessee  37401
615/751-2774

Stepen W. Tippet
Product Manager
CHEMFAB
Water St.
N. Bennington, Vermont  05257
802/447-1131

Ed Tomeo
Generation Engineer
Northeast Utilities
P.O. Box 270
Hartford, Connecticut  06104
203/247-0838

Tasia P- Toombs
Conference & Travel Coordinator
Electric Power Research Institute
P.O. Box 10412
Palo Alto, California  94303
415/855-8973

Robert L. Torstrick
Section Supervisor
Tennessee Valley Authority
501 Chemical  Engineering Bldg.
Muscle Shoals, Alabama  35660
205/386-2514
Joseph M. Towarnicky
Research Chemist
United McGill Corp.
One Mission Park
Groveport, Ohio  43125
614/836-9981

Edward C. Trexler
Program Manager
U.S. Department of Energy
Washington, D.C.  20545
301/353-2683

Ronald J. Triscori
Director of Products
Joy Manufacturing Co.
4565 Colorado Blvd.
Los Angeles, California  90039
213/240-2300 ext. 527

Philip C. Tseng
Research Analyst
University of Texas at Austin
E.P. Schoch Lab.
Austin, Texas  78712

J. V. Twork
Engineer, Research Dept.
Bethlehem Steel  Corp.
Homer Research Lab
Bldg. A Rm D 332
Bethelehm, Pennsylvania  18016
215/694-6532

Kiyoshi Urakami
Commercial  Manager
Mitsubishi
520 Madison Ave.
New York, New York
212/605-2006

Robert P. Van Ness
Louisville Gas & Electric Co.
P.O. Box 32010
Louisville, Kentucky  40232
502/566-4011

Ernest Vanhoose
Program Manager
Kentucky Energy Cabinet
P.O. Box 1188, Iron Works Pike
Lexington, Kentucky  40578
606/252-5535
                                     A-47

-------
Joel Yatsky
Manager, Combustion Technology
Foster Wheeler Energy Corp.
9 Peach Tree Hill Rd.
Livingston, New Jersey  07039
2U1/533-2105

Marl in J. Yeesaert
President
beneral Aggregate Corp.
401 North Lindbergh Blvd.
St. Louis, Missouri  63141
314/997-7777

Kurt Veser
Director
Kraftanlagen Heidelberg
Im  Breitspiel 7
D 6900 Heidelberg,
West Germany
06221/394562

Suzanne M. Viale
Marketing Services Coordinator
FMC Corp.
231 N. Martingale Rd.
Schaumburg, Illinois  60194-2098
312/843-1700

Mohr Volker
Process Engineer
Letepro Corp.
Avenue of the Americas
New York, New York

Stephen C. Voss
Noxso Corp.
2625 H. C. Mathis Dr.
Paducah, Kentucky  42001

David A. Wagner
Sales Representative
Marblehead Lime Co.
300 W. Washington
Chicago, Illinois  60606
312/263-4490

J.  Peter Wahlman
Senior Project Manager
Cliffs Engineering, Inc.
P.O. Box 1211
Rifle, Colorado  81650
3U3/625-2445
John R. Walenten
Process Development Engineer
Badische Corp.
50 Central Ave.
Kearny, New Jersey  07032
205/589-1600

Shih-Chung Wang
Engineering Specialist
Bechtel Group, Inc.
50 Beale St.
San Francisco, California  94105
415/768-2873

Dale R. Warner
Regional Manager
Air Clean Damper Co.
645 Ridgemont Dr.
Roswell, Georgia  30076
404/587-5293

Earl J. Weber
Mechanical Engineer
Cajun Electric Power Coop.
P.O. Box 15540
Baton Rouge, Louisiana  70816
504/291-3060

Henry C. Weber
President
HAW Management Science Consult.
415 E. 52 St.
New York, New York  10022
212/355-1448

Greg F. Weber
Research Supervisor
Univ. of No. Dakota Energy
  Research Ctr.
Box 8213, University Station
Grand Forks, North Dakota  58502
701/795-8222

William C. Webster
American Resources Corp.
P.O. Box 813
Valley Forge Pennsylvania  19482
215/337-7373
                                     A-48

-------
Christopher P. Wedig
Supervisor, Process Engineering
  Group Div.
Stone & Webster
P.O. Box 2325
Boston, Massachusetts  02107
617/469-1900

John Weeda
Engineering Superintendent
Cooperative Power Assoc.
  Coal Creek Station
P.O. Box 780
Unaerwood, North Dakota  58576-0780
701/442-3211

Maurice W. Wei
Envir. Eng. Manager Proc. Gas
Aluminum Co. of America
1501 Alcoa Bldg.
Pittsburgh, Pennsylvania  15219
412/553-2085

Ing Helmut Weiler
Director
Grobkraftwerk Franken AG
Rudolphstr. 28
Nuerenberg, GERMANY (W) D-8500
911-5397203

Carl Weilert
Air Quality Consultant
Burns and McDonnell Engineering Co.
P.O. Box 173
Kansas City, MO  64141
816/333-4375

Alex Weir, Jr.
Manager, Chemical Systems, R&D
So. Cal. Edison Co.
P.O. Box 800
Rosemead, California  91770
213/572-2785

Don Welch
Staff Engineer
TRE-ASTECH
800 Hillcrest Rd. #5
Mobile, Alabama  36609
R. Murray Wells
Vice President
Radian Corp.
8501 Mo-Pac Blvd.
Austin, TX  78766
512/454-4797

Robert E. Whetstine
Energy Design Supervisor
Middle South Services, Inc.
P.O. Box 61000
New Orleans, Louisiana  70161
504/569-4730

Michael James Widico
Project Manager
Research Cottrell, Inc.
P.O. Box 1500
Somerville, New Jersey  08876
201/685-4213

David S. Wiggins
Supv. Process Engineering
United Engineers
30 S. 17th St.
Philadelphia, Pennsylvania  19101
609/772-0600

Wilbert W. Wiitala
Director
Marquette Board of Light & Power
2200 Wright
Marquette, Michigan  49855
906/228-6900

W. J. Wijdeveld
Manager, APC Division
ESMIL Intl.
De Boelelaan 7
Amsterdam, Netherlands
020-5411054

Dale S. Wiley
Mechanical Engineer
Wisconsin Public Service Corp.
P.O. Box 1200
Green Bay, Wisconsin  54305
414/433-1274
                                      A-49

-------
James H. Wilhelm
President
Codan Assoc.
2394 Charros Rd.
Sandy, Utah  84092
801/571-6974

Don Wilhelm
Staff Chemical Engineering
Morrison Knudsen Co. Inc.
P.O. Box 7808
Boise, Idaho 83729
208/345-5000

John E. Williams
Chemical Engineer, Project Manager
DOE, Pittsburgh Energy Technical  Ctr.
P.O. Box 10940
Pittsburgh, Pennsylvania  15236
412/675-5727

D. A. Williams
Sales Manager
Davu McKee Corp.
471b S. Florida Ave.
Lakeland, Florida  33803
813/646-7844

C. Bailey Williams
Vice President, General  Manager
Marcona Ocean Ind., Ltd.
1001 N. W. 62nd St., Suite 200
Fort Lauderdale, FL  33309
305/776-4000

R. A. Wilson
Engineering Supervisor
Bechtel Power Corp.
12400 East Imperial Highway
Norwalk, California  90650
213/807-2767

David A. Wilson
Research Leader
Dow Chemical U.S.A.
Blag B-1222
Freeport, Texas  77541
409/236-7737
Gene Winkler
Application Engineer
Munters Corp.
P.O. Box 6428
Ft. Meyers, Florida  33911
813/936-1555

Lloyd Winsor
Asst. Chief Consulting Engineer
Gibbs & Hill Inc.
11 Penn Plaza
New York, New York  10001
212/760-5700

Jozewicz Wojciech
Research Scholar
University of Texas at Austin
E.P. Schoch Lab.
Austin, Texas 78712
512/471-4857
Stanley J. Wojton
Senior Engineer
Cleveland Electric
P.O. Box 97
Perry, Ohio  44081
216/259-3737
Illuminating Co.
Steve Wolf
Engineer
Northern State Power
414 Nicollet Mall
Minneapolis, Minnesota
612/330-5624
     55401
Steve Wolsiffer
Assistant Plant Superintendent
Indianapolis Power & Light Co.
P.O. Box 436
Petersburg, Indiana  47567
812/354-b801

John M. Wootten
Director, Environmental Services
Peabody Coal Co.
301 N. Memorial Dr.e
St. Louis, Missouri  63102
314/342-3400
                                    A-50

-------
R. M. Wright
Product Sales Manager
FMC Corp.
2000 Market St.
Philadelphia, Pennsylvania  19103
215/299-6815

Steve Wright
Senior Engineer
Montana Power Co.
40 E. Broadway
Butte, Montana  59701
406/723-5421

Charles R. Wright
Instrument Design Engineer
Tennessee Valley Authority
501 Chemical Engineering Bldg.
Muscle Shoals, Alabama  35660
205/386-2514

Robert J. Wright
President
Zurn Industries, Inc.
P.O. Box 2206
Birmingham, Alabama  35201
205/252-2181

Charles E. Wright, Jr.
Supervisor Applications
Combustion Engineering
31 Inverness Center Pkwy.
Birmingham, Alabama  35213
205/967-9100

Beth A. Wrobel
Engineer
Northern Indiana Public Service Co.
Rt. #1 Box 320
Wheatfield, Indiana  46392
219/956-5251

Jarnes E, Wuchter
Project Leader
Badische Corp.
602 Copper Rd.
Freeport, Texas  77541
409/238-6284
John Yavorsky
Project Leader
ASARCo
901 Oak Tree Rd., S.
Plainfield, New Jersey  07080
201/756-4800

Robert Yeargan
Power Plant Superintendent
Tennessee Valley Authority
510 Edney Bldg.
Chattanooga, TN  37401
615/751-4909

Charles S. Young
Product Manager, New Products
Astro Metallurgical
3225 Lincoln Way West
Wooster, Ohio  44691
216/264-8639

Dean Young
Vice President
Kissick Corp.
108 Benson East
Jenkintown, Pennsylvania  19046
215/885-6650

George J. Ziegenhorn
Senior Environmental Engineer
Arco Petroleum Products Co.
400 E. Sibley Blvd.
Harvey, Illinois  60426
312/333-3000, ext. 359

Denis M. Zielinski
Environmental Scientist
U.S. EPA
6th & Walnut Streets
Philadelphia, Pennsylvania  19101
215/597-0804

Jan T. Zmuda
Senior Development Engineer
Research-Cottrell
P.O. Box 1500
Somerville, New Jersey  08876
201/685-4915
                                      A-51

-------
Joseph Zuzolo
Project Manager
York Research Consultants
938 Quail St.
Denver, Colorado  80215
3U3/233-1513
John C. deRuyter
Senior Engineer
E. I. du Pont de Nemounrs
  ana Co., Inc.
Engineering Department, Louviers
Wilmington, Delaware  19898
302/366-6442
Bldg.
                                     A-52

-------
EPRI CS-3706, Volume 2




1. EPRI CS-3706
4. Flue Gas Des
— <
-. c
su 3
o ro



33
<§
6





Cross-Referen
o
CD
w













m
-a
33
T)
"TO
o
05
CO
CD
H
S
o
w
<<









Timely exchan
tion (FGD) sys
standards. Th<
architectural-e
sions about th
CD 13
Q.CO
 -*
g3|
£S£
~ W 3.
CQ D °
en — 8L
O W 03
° z ^
| o Q.
^ O n
C 03 O

^ c o
» =3
Proceed!
Novembi
^ 3
(O ^

Contractor
3J
CD
W
CD
to
o

03
CO
CD
5"
CO
c
CD
EPRI CS
Volume :
RP982-3"
"
o
o>
Proceedi
Desulfuri
N 3
•"•• C/3
i'm
CQ
<; ^3"
2. 5
C
M|

O




PROGR
>
2












DESULI
Tl
C
3]
N
H
O
•z.
•u
O
O
m
Below ar
four cros
abstract
is includi


EPRI CS-3706
Volume 2
RP982-31
Proceedings
November 1984






S fflfo £2. £
CO
o
CD
C
-^
jrizatio
D
TJ
0
O

CD
CO

O
(O
a)
3










*r*
EPRI CS-3706, Vol
Flue Gas Desulfui
FT|
sj
o ro
D


K>
3
fo
w


OJ

O
ro
Cfl
c
c"
Q>

o"
13
O
o
CO
CO
CD
CO
?

























o
ross-References


















































m
T)
T)
"CD
o
w
CQ
CD

S
o
£
CO
;*r
•<


















li
a§
•n §
O CD
o ~
& ^
's. °
CD" 3
Pa

3
O1 0)
CO c?
•D a
T3 CD
' O











a) cu
rchitectural-engi
ions about the d<
w 3
2. ro
CQ CD
D I?-
D

= Q.
2. W
5 'w
P CD
= ?
O 73
5 ™
ll

^i. o
-^ ^
O CD
D -
<" ^
"5 0

if 3
3 c
«.
3
Ol Q3
CD J
"0 Q.
P CD
c vu E
cF QJ 3
CD q S.
(ft -^ — -
0 C 0
• c 2. =>
0 q 9
111 5
g S. E
c ^.
— . ^. 03
^i' ^^. ^"
^ CD Q-
3 CD
CO
w E.

o' ^
3 N"
W (jj










w i: H
i mely exchanges
on (FGD) system
tandards. These
"5 m o
o g> =!•
» "" 2
CD CO 0
9:^ f
^ CD =-
CQ 13 °
M - e.
o a; a,
° -» ^
S ° °-
= 8 S
C 03 O
fD c O

|||
CTJO
CD fl) 3

CD 3 S.
o S 9
^ CO =1
o q 2
<" 1 =
~" K CD —
O ;4. c
^ 0) 








3) < m
-0 0 -0
CD c 3D
00 3 —

^ to ^
O
en

n -n
'roceeding
)esulfuriza
zr. en
o ' '
CQ"

2. 5
Q
ro 3
O
C/3
c'
3

o


Tl
C
CD

o
Q)
O>














n
mm
^W
4
^
































mmmm
n
i-r
\
m
CO









1
j
i
4
j
•











T|
1-
m
0
o
m
CO
c

T]
c
DO
N
5
O
z













1
1
J













EPRI CS-3706
Volume 2
RP982-31
Proceedings
November 1984













EPRI CS-3706
Volume 2
RP982-31
Proceedings
November 1984
















Below are five index cards that allow for fMing according io the
              Terences in addition to the title of the report. A brief
abstract describing the major subject  area covered in the report
                                                                   EPRI
            Proceedings:  Eighth Symposium on Flue Gas
            Desulfurization   Volume 2


               Contractor: Research Triangle Institute

            Timely exchanges of technical and economic information on flue gas desulfuriza-
            tion (FGD) systems are essential to coal utilities that must meet strict emissions
            standards. These proceedings constitute a valuable resource for utility,
            architectural-engineering, and system-supplier personnel who must make deci-
            sions about the design, installation, and operation of FGD systems.  578 pp.
               EPRI Project Manager: T. M. Morasky
            Cross-References:
            1. EPRI CS-3706, Volume 2
            4. Flue Gas Desulfurization
          2. RP982-31
                  ELECTRIC POWER RESEARCH
                  Post Office Box 10412, Palo Alto, CA 94303
               EPRI CS-3706, VOLUME 2
                        3. Desulfurization Processes Program
                      N S T I T U T E
                      415-855-2000
                                         EPRI
            Proceedings:  Eighth Symposium on Flue Gas
            Desulfurization   volume 2


              Contractor: Research Triangle Institute

            Timely exchanges of technical and economic information on flue gas desulfuriza-
            tion (FGD) systems are essential to coal utilities that must meet strict emissions
            standards. These proceedings constitute a valuable resource for utility,
            architectural-engineering, and system-supplier personnel who must make deci-
            sions about the design, installation, and operation of FGD systems.  578 pp.
              EPRI Project Manager: T. M. Morasky

            Cross-References:
            1.  EPRI CS-3706, Volume 2
            4.  Flue Gas Desulfurization
                                    2. RP982-31
                        3. Desulfurization Processes Program
                  ELECTRIC POWER RESEARCH INSTITUTE
                  Post Office Box 10412. Palo Alto, CA 94303   415-855-2000
               RP982-31
                                                                    EPRI
            Proceedings:  Eighth Symposium on Flue Gas
            Desulfurization   Volume 2


              Contractor: Research Triangle Institute

            Timely exchanges of technical and economic information on flue gas desulfuriza-
            tion (FGD) systems are essential to coal utilities that must meet strict emissions
            standards. These proceedings constitute a valuable resource for utility,
            architectural-engineering, and system-supplier personnel who must make deci-
            sions about the design, installation, and operation of FGD systems.  578 pp.
              EPRI Project Manager- T. M. Morasky
            Cross-References:
            1. EPRI CS-3706, Volume 2
            4. Flue Gas Desulfurization
                        3 Desullunzation Processes Progran
                  E L E C T R
                  Posl Ollice
C POWER RESEARCH INSTITUTE
Box 10412. Palo Alto. CA 94303   415-855 2000

-------