EPA-650/2-73-038
December 1973
ENVIRONMENTAL PROTECTION TECHNOLOGY SERIES

                        irljXlKijIWijKx-xjiwI^'JlSkiJlIf

                                                                   :::;



                                                                   -:
                                                                   S


-------
                                    EPA-650/2-73-038
          PROCEEDINGS:
FLUE GAS  DESULFURIZATION
      SYMPOSIUM  -  1973
             May 14-17, 1973
               Jung Hotel
           New Orleans, Louisiana
           E. L. Plyler, Chairman
        M. A. Maxwell, Vice Chairman
          Control Systems Laboratory
     National Environmental Research Center
   Research  Triangle Park, North Carolina  27711
          .  ROAP No. 21ACY-30
         Program Element No. 1AB013
  NATIONAL ENVIRONMENTAL RESEARCH CENTER
    OFFICE OF RESEARCH AND DEVELOPMENT
   U.S. ENVIRONMENTAL PROTECTION AGENCY
    RESEARCH TRIANGLE PARK, N.C. 27711
              December 1973

-------
                               PREFACE
     A Flue Gas Desulfurization Symposium was held May 14-17, 1973, in
the Jung Hotel, New Orleans, Louisiana, under the sponsorship of the
Environmental Protection Agency, Office of Research and Development,
Control Systems Laboratory.  The primary purpose of the symposium was
to present the current status of "throwaway" and "regenerable" flue
gas desulfurization processes as applied to controlling SO^ emissions
from full-scale facilities.

     The symposium included sessions on technology and application of
first generation processes such as lime/limestone scrubbing, magnesia
scrubbing, catalytic oxidation and sodium sulfite scrubbing as well
as a session on second generation or advanced processes.  In addition,
a panel discussion was held concerning the disposal and uses of by-
products from flue gas desulfurization processes.

     Over 430 representatives of government and industry were in
attendance during the 4-day symposium.

     These proceedings have been reviewed by the Environmental Protec-
tion Agency and approved for publication.  Except for minor editing
for consistency of presentation, the contents of this report are as
received from the authors.  Approval does not signify that the contents
necessarily reflect the views and policies of the Agency, nor does men-
tion of trade names or commercial products constitute endorsement or
recommendation for use.  All papers presented (as well as transcriptions
of the panel discussions and the symposium summary) are included in
these proceedings.

     Copies of this report are available free of charge to Federal
employees, current contractors and grantees, and nonprofit organiza-
tions -- as supplies permit -- from the Air Pollution Technical Infor-
mation Center, Environmental Protection Agency, Research Triangle Park,
North Carolina  27711.
                   Publication No. EPA-650/2-73-038

                                  iii

-------
                              CONTENTS


TITLE                                                               PAGE



                           OPENING SESSION

     Keynote Address - THE HEALTH RATIONALE FOR STRICT CONTROL
     OF SULFUR OXIDE EMISSIONS
          V. A. Newill, Environmental Protection Agency,
          Washington, D. C	       1

     TECHNOLOGICAL ALTERNATIVES TO FLUE GAS DESULFURIZATION
          S. J. Gage, Council on Environmental Quality,
          Washington, D. C	      13

     STATUS OF FLUE GAS DESULFURIZATION TECHNOLOGY
          R. E. Harrington,  Environmental Protection Agency,
          Washington, D. C	      57

     STATUS OF JAPANESE FLUE GAS DESULFURIZATION TECHNOLOGY
          J. Ando, Chuo University,  Tokyo, Japan  	      69

     ECONOMICS OF FLUE GAS DESULFURIZATION
          G. T. Rochelle, Environmental Protection Agency,
          Research Triangle Park, North Carolina  	     103

                    THROWAWAY PROCESSES - PART I
          Session Chairman - F. T. Princiotta, Environmental
          Protection Agency, Research Triangle Park,
          North Carolina

     STATUS OF TECHNOLOGY OF COMMERCIALLY OFFERED LIME AND
     LIMESTONE FLUE GAS DESULFURIZATION SYSTEMS
          I. Raben, Bechtel Corporation, San Francisco,
          California	     133

     WASTE PRODUCTS FROM THROWAWAY FLUE GAS CLEANING PROCESSES  -
     ECOLOGICALLY SOUND  TREATMENT AND DISPOSAL
          J. W. Jones, R. D. Stern,  and F. T. Princiotta,
          Environmental Protection Agency, Research Triangle
          Park, North Carolina	     187

     TEST RESULTS FROM THE EPA LIME/LIMESTONE SCRUBBING TEST
     FACILITY
          M. Epstein, Bechtel Corporation, San Francisco,
          California	     235

-------
TITLE                                                               PAGE

     OPERABILITY AND RELIABILITY OF THE EPA LIME/LIMESTONE
     SCRUBBING TEST FACILITY
          H. W. Elder and P. E.  Stone,  Tennessee Valley Authority,
          Muscle Shoals,  Alabama
          L. Sybert, Bechtel Corporation,  San Francisco,
          California
          J. E. Williams, Environmental Protection Agency,
          Research Triangle Park, North Carolina	     333

     SCRUBBING EXPERIMENTS  AT THE MOHAVE GENERATING  STATION
          A. Weir, Jr., and L.  T. Papay, Southern California
          Edison, Rosemead,  California  	     357

                    THROWAWAY PROCESSES -  PART II
          Session Chairman - F.  T.  Princiotta, Environmental
          Protection Agency, Research Triangle Park,
          North Carolina

     A REVIEW OF BABCOCK AND WILCOX AIR POLLUTION CONTROL
     SYSTEMS FOR UTILITY BOILERS
          J. Stewart, Babcock and Wilcox Co., Barberton, Ohio .  .     393

     ONE YEAR'S PERFORMANCE AND OPERABILITY OF THE CHEMICO/
     MITSUI CARBIDE SLUDGE  (LIME) ADDITIVE SO? SCRUBBING SYSTEM
          J. Sakanishi, Mitsui Aluminum Co., Omuta, Japan
          R. H. Quig, Chemical Construction Corporation,
          New York, New York	     419

     BRIEF PANEL DISCUSSION - SIGNIFICANCE OF OPERATION TO DATE
     OF 156 MW CHEMICO/MITSUI LIME SCRUBBING SYSTEM
          P. Wechselblatt, Chemical Construction Corporation,
          New York, New York
          J. Craig, Southern Services,  Birmingham, Alabama
          H. W. Elder, Tennessee Valley Authority, Muscle Shoals,
          Alabama
          F. T. Princiotta,  Environmental  Protection Agency,
          Research Triangle Park, North Carolina  	     451

     THE TVA WIDOWS CREEK LIMESTONE SCRUBBING FACILITY
     PART I - FULL  SCALE FACILITY
          B. G. McKinney, Tennessee Valley Authority,
          Chattanooga, Tennessee
          A. F. Little, Tennessee Valley Authority,
          Muscle Shoals,  Alabama
          J. A. Hudson, Tennessee Valley Authority,
          Knoxville, Tennessee	     475

                                   VI

-------
TITLE                                                               PAGE

     PART II - PILOT PLANT AND PROTOTYPE OPERATING EXPERIENCE
          J. J. Schultz, Tennessee Valley Authority,
          Muscle Shoals, Alabama
          T. M. Kelso, Tennessee Valley Authority,
          Muscle Shoals, Alabama
          J. L. Graham, Tennessee Valley Authority,
          Muscle Shoals, Alabama
          J. K. Metcalfe, Tennessee Valley Authority,
          Muscle Shoals, Alabama
          N. D. Moore, Tennessee Valley Authority,
          Chattanooga, Tennessee  	     495

     STATUS OF C.E.'S AIR QUALITY CONTROL SYSTEMS
          M. R. Gogineni, J. R. Martin, P. G. Maurin,
          Combustion Engineering, Inc., Windsor, Connecticut  .  .     539

                   REGENERABLE PROCESSES - PART I
          Session Chairman - G. G. McGlamery, Tennessee Valley
          Authority, Muscle Shoals, Alabama

     MAGNESIA SCRUBBING
          G. G. McGlamery, Tennessee Valley Authority,
          Muscle Shoals, Alabama  	     553

     OPERATIONAL PERFORMANCE OF THE CHEMICO BASIC MAGNESIUM
     OXIDE  SYSTEM AT THE BOSTON EDISON COMPANY
     PART I - G. Koehler, Chemical Construction Corporation,
              New York, New York	     579

     Part II- C, P, Quigley, Boston Edison Company,
              Boston, Massachusetts    	     605

     DESIGN AND INSTALLATION OF A PROTOTYPE MAGNESIA SCRUBBING
     INSTALLATION
          B. M. Anz, C. C. Thompson, and J. T. Pinkston, United
          Engineers and Constructors,  Philadelphia, Pennsylvania.     619


                   REGENERABLE PROCESSES - PART II
          Session Chairman - N. Plaks, Environmental Protection
          Agency, Research Triangle Park, North Carolina

     APPLICATION OF THE WELLMAN-LORD S02 RECOVERY  PROCESS TO
     STACK  GAS DESULFURIZATION
          R. T. Schneider and C. B. Earl, Davy Powergas, Inc.,
          Lakeland, Florida  	    641

                                  vii

-------
TITLE                                                               PAGE

     APPLICATION OF S02 REDUCTION IN STACK GAS DESULFURIZATION
     SYSTEMS
          W. D. Hunter, Jr., Allied Chemical Corporation,
          Morristown, New Jersey  	     657

     THE CAT-OX PROJECT AT ILLINOIS POWER
          W. E. Miller, Illinois Power Company, Decatur,
          Illinois	     673

     MITRE TEST SUPPORT FOR THE CAT-OX DEMONSTRATION PROGRAM
          G. Erskine and E. Jamgochian, Mitre Corporation,
          McLean,  Virginia  	     687

     DISPOSAL AND USE OF BY-PRODUCTS FROM FLUE GAS DESULFURIZATION
     PROCESSES:  INTRODUCTION AND OVERVIEW
          Panel Chairman - A.  V. Slack, Tennessee Valley Authority,
          Muscle Shoals, Alabama
          J. M. Potts,  Tennessee Valley Authority, Muscle Shoals,
          Alabama	     747

     STUDY OF DISPOSAL  AND UTILIZATION OF BY-PRODUCTS FROM
     THROWAWAY DESULFURIZATION PROCESSES
          J. Rossoff, R. C. Rossi,  and J. Meltzer, Aerospace
          Corporation,  El Segundo,  California   	     77S

     EXPERIENCE IN THE  DISPOSAL AND UTILIZATION OF SLUDGE FROM
     LIME/LIMESTONE SCRUBBING PROCESSES
          W. C. Taylor, Combustion  Engineering, Inc.,
          Windsor, Connecticut  	     799

     FIXATION AND  DISPOSAL OF FLUE  GAS WASTE PRODUCTS:   TECHNICAL
     AND ECONOMIC  ASSESSMENT
          L. J. Minnick, IU Conversion Systems, Inc.,
          Plymouth Meeting, Pennsylvania  	     835

     UTILIZING AND DISPOSING OF SULFUR PRODUCTS FROM FLUE GAS
     DESULFURIZATION PROCESSES IN JAPAN
          J. Ando, Chuo University,  Tokyo,  Japan  	     875

     LONG RANGE MARKET  PROJECTIONS  FOR BY-PRODUCTS OF REGENERABLE
     FLUE GAS DESULFURIZATION PROCESSES
          M. H. Farmer, Esso Research  and Engineering Company,
          Linden,  New Jersey	     891

     NEW USES FOR  SULFUR - THEIR STATUS AND PROSPECTS
          H. L. Fike and J. S.  Platou,  The Sulphur Institute,
          Washington, D. C	     921

     PANEL DISCUSSION:   DISPOSAL AND USE OF BY-PRODUCTS
     FROM FLUE GAS DESULFURIZATION PROCESSES	     931

                                viii

-------
TITLE                                                               PAGE

                          ADVANCED PROCESSES
          Session Chairman - J. H. Abbott, Environmental
          Protection Agency, Research Triangle Park,
          North Carolina

     REMOVAL OF SO- FROM STACK GASES BY SCRUBBING WITH AMMONIACAL
     SOLUTIONS:  PILOT SCALE STUDIES AT TVA
          G. A. Hollinden and N. D. Moore, Tennessee Valley
          Authority, Chattanooga, Tennessee
          P. C. Williamson, Tennessee Valley Authority,
          Muscle Shoals, Alabama
          D. A. Denny, Environmental Protection Agency,
          Research Triangle Park, North Carolina  	    961

     AN EPA OVERVIEW OF SODIUM BASED DOUBLE ALKALI PROCESSES
     PART I - A VIEW OF THE PROCESS CHEMISTRY OF IDENTIFIABLE
     AND ATTRACTIVE SCHEMES
          D. C. Draemel, Environmental Protection Agency,
          Research Triangle Park, North Carolina  	    997
     PART II - STATUS OF TECHNOLOGY AND DESCRIPTION  OF
     ATTRACTIVE SCHEMES
          N. Kaplan, Environmental Protection Agency,
          Research Triangle Park, North Carolina  	   1019

     STONE AND WEBSTER/IONICS SO  REMOVAL AND RECOVERY PROCESS
          N. L. Foskett and E. G. Lowrance, Stone and Webster
          Engineering Corporation, Boston, Massachusetts
          W. A. McRae, Ionics, Inc., Watertown, Massachusetts .  .   1061

     FOSTER WHEELER/BERBAU-FORSCHUNG DRY ADSORPTION  SYSTEM FOR
     FLUE GAS CLEANUP
          W. F. Bischoff, Foster Wheeler, Livingston, New Jersey.   1081

     THE ATOMICS INTERNATIONAL MOLTEN CARBONATE PROCESS FOR SO
     REMOVAL FROM STACK GASES
          W. V. Botts and R. D. Oldenkamp, Atomics International,
          Canoga Park, California 	   1101

     SYMPOSIUM SUMMARY
          F. T. Princiotta, Environmental Protection Agency,
          Research Triangle Park, North Carolina  	   1133

                                 ix

-------
                HEALTH RATIONALE
FOR STRICT CONTROL OF SULFUR OXIDE  EMISSIONS
                         by

        Vaun  A. Newill,  M. D. ,  S.  M. Hyg,
            Jean D. French, Dr. P. H.
             Office  of the Administrator
          Environmental Protection Agency
                Washington,  D.C.

-------
     .Man  has  been  subject  to air  pollution since his primordial ancestor
 lit  the first fire.   It was not,  however, until people became crowded
 together  in cities that pollution was more than a family problem from
 the  hearth.   Today with the phenomenal growth of both the population
 and  use of power in the technologically advanced countries, pollution
 has  reached such magnitude that it threatens the health and well being
 of the population on  a global scale.
     Even though there had been occasional speculation and complaints
 about the health effects of air pollution since the days of Edward I,
 it was not until the  occurrence of certain air pollution episodes in
 the  Valley of the Meuse, in London and in Donora, Pennsylvania that
 people finally came to realize that it was a threat to human health.
 These episodes made it obvious that under certain conditions air
 pollution could kill.
      Despite the drama of the episodes, the greatest impact of air
 pollutants on human health results from day to day exposure under
unexceptional  conditions.
      Environmental pollutants can effect the health of individuals
or communities over a broad range of biological  responses as shown
 in Figure 1.   At any point in time more severe effects  such as death
will  be manifested in relatively small proportions of the population.
Mortality studies have shown that elevated levels of S02 contribute
to approximately 1  percent of excess deaths.

-------
     Illness from sulfur oxides may be either short-lived (acute) or
relatively permanent and irreversible (chronic).
     Bronchitis is an example of a long-term or chronic effect.
Investigators in Japan, Britain, and the United States have demonstrated
excess bronchitis morbidity in areas with moderately elevated 862
exposure.  A significant and consistent excess bronchitis morbidity
                                                  3
occurred over an exposure range of 100 to 350 yg/m  S02 with associated
total suspended particulate levels of 66 to 365 yg/m3.  The following
slides show results from several studies conducted by the Environmental
Protection Agency in different areas of the United States.
     Table 1  demonstrates that nonsmokers as well  as smokers from high
exposure communities of four areas consistently reported more chronic
bronchitis than their smoking counterparts in low exposure neighborhoods.
     As shown in the last row of Table 2, the contribution of air pollu-
tion towards the prevalence of chronic bronchitis  is one-fifth to equal
that of cigarette smoking among males in the four areas.
     A remarkably consistent excess of acute respiratory disease ranging
from 14 to 64 percent was found in children exposed to S02 levels of
50 to 275 yg/m3.  These findings pertained largely to children who had
lived in the polluted areas for three years or longer suggesting a
chronic alteration in their defense mechanisms.
     Figure 2 shows excess  croup frequency among children from high
pollution neighborhoods in  the Salt Lake Basin who had lived in that
neighborhood  for three or more years.   Similar findings in children
in Idaho-Montana are presented in Figure 3.  Aggravation of symptoms

-------
in diseased subjects such as asthmatics and cardie-pulmonary subjects
have been repeatedly demonstrated at daily S02 levels well below the
indicated threshold for excess mortality.  However, some studies have
demonstrated illness aggravation to be most strongly related to
suspended sulfates.  Figure 4 reveals that at the same total suspended
particulate level, asthma attack rates were higher on days with high
sulfate concentrations.  These data are from a study conducted by EPA
in the Salt Lake Basin.
     As shown in Figure 5, when symptom aggravation in heart and lung
patients in New York were computed against exposure within a defined
temperature, daily levels of suspended sulfates more so than SC>2 or
total suspended particulates were closely associated with these attacks
of illness.  The threshold for the adverse effect occurred at sulfate
concentrations of 9.2 ug/m , a level below the average daily sulfate
concentrations of most northeastern cities.
     These findings confirm results from animal studies in which metallic
sulfates were shown to exert adverse biological effects at concentrations
below those of SOn.  These results are physiologically coherent since
S0£ alone tends to be absorbed high in the respiratory tract while
sulfates can be delivered deep into the lung.   It is important for
environmental  scientists to determine the relationship between $©2 and
suspended sulfates.  Although St^ serves as a  precursor to the formation
of sulfates. the relationship is nonlinear; there is reason to believe
that S02 reacts in a complex manner with particulate matter.  Certain

-------
metal  particles, water vapor and oxidizing chemicals in the atmosphere
may promote the formation of more potent respiratory irritant sulfate
aerosols.
     The section of the biological response spectrum entitled "functional
changes preceding disease" can be exemplified by pulmonary function
tests.  A diminution in the pulmonary function test indicates some
resistance in air flow to the lungs and serves as a sentinel  that a
disease process may ensue.  Studies have shown lung function of school
children living in areas of moderate S0« and TSP pollution was impaired
compared with children living in low exposure communities.  In a study
conducted in Japan, lung function of school children responded rapidly
to seasonal changes in air pollution.  Function became worse in polluted
seasons and during less polluted months returned to levels of children
in low exposure areas.  These data suggest that the lung function may
be reversible and improved air quality will enhance lung function in
children.
     Experimental  laboratory studies on humans and on lower animals
exposed to artificially produced S02 and suspended sulfates have shown
subtle physiologic changes,  the significance of which is still  obscure.
Russian investigators have demonstrated neurobehavioral  effects of
sulfur oxides including changes in the electroencephalogram.   Histo-
pathological  studies in animals showed marked histopathological  changes
in nasal  epithelium at concentrations of S02 which had  no direct effect
deeper in the lungs.

-------
     There are no existing data on the relationship between sulfur
oxides and pollutant burdens.
     Table 3 summarizes the levels of S02 and suspended sulfates
associated with biological responses over the entire spectrum.  As
shown in the table, adverse health consequences ranging from functional
changes preceding disease to death itself have been associated with SC^
                                        q
exposures in the range of 80 to 120 ug/nr for one or more days.  Even
in communities where national primary standards for S02 have been
achieved, daily sulfate levels in the range of 7 to 14 ug/tn3 have been
associated with aggravation of symptoms in particularly vulnerable
population groups such as asthmatics and cardiopulmonary subjects.
Although pressure is currently being exerted to relax our existing
national air pollution standards because of the energy crisis, we must
bear in mind that such action could be costly in terms of safeguarding
the nation's health.  We must pursue a prudent course whereby the
energy needs of the country are met in a manner which poses a minimum
threat to the public health.

-------
                           ACKNOWLEDGEMENT



     The authors gratefully acknowledge the Human Studies Laboratory



of the U.  S.  Environmental  Protection Agency for providing from their



research much of the data used in the figures and tables.

-------
Table 1.   RATIOS OF CHRONIC RESPIRATORY DISEASE AREA PREVALENCE  RATIOS
            (MALES) BY GEOGRAPHIC AREA AND BY CATEGORIES
                   OF AIR POLLUTION AND SMOKING
 Exposure
                              CRD prevalence ratios  (males)
Utah
 Idaho-
Montana
                                                 New
                                                York
Chicago
Low
  Nonsmoker

  Smoker
High
  Nonsmoker
  Smoker
(3.0)'
 6.5

 2.3
 8.9
              1.0
             (1.25)
             13.6
              2.8
             14.7
                1.0
               (4.6)
                3.0
                3.4
                4.7
  Smoker	8^9	[     14.7      j     417__	

aBase period prevalence rate  per 100 people in  parentheses
                                                              1.0
                                                             (4.0)
                                                              3.8
  1.3
  4.5
 Table 2.   RATIOS OF CHRONIC  RESPIRATORY  DISEASE  AREA  PREVALENCE  RATIOS
          (MALES) BY POLLUTION  AND  SMOKING  SOURCES  OF  RISK


Exposure
Pollution3
Smoking
Pollution
Smoking
CRD relative prevalence (males)

Utah
2.27
6.53
1
3
Idaho-
Montana
2.78
13.64
1
5
New
York
3.45
3.02
1
1

Chicago
1.33
3.80
1
3
 Nonsmokers  of  high  exposure/nonsmokers of  low exposure area.
'Smokers  of  low exposure  area/nonsmokers of same area.

-------
     Table 3.   EFFECTS OF SULFUR OXIDES  ON HEALTH
                  BY LEVEL OF RESPONSE
Level of
resoonse
                                      SQ2,
                                         (ppm)
                       500-1000 (0.20-0.40)3

                        80-275  (0.03-0.11)a


                        90-120  (0.035-0.45)1


                       500-1000 (0.20-0.40)c


                            No data
Sulfates,
  jjg/m3
Death
Illness  (acute and
 chronic)
Functional  changes
 preceding  disease
Changes  of  uncertain
 significance
Pollutant burdens	

 24 hour average.
bn
 Annual  average.
Experimental  studies with artificial  S02 or other sulfur oxides,
 No data

  7 14a


  9-lla

  25QC


 No data
                                                   ADVERSE
                                                   HEALTH
                                                   EFFECTS
                          PHYSIOLOGIC        \        T
                          HANGES             \
                   PHYSIOLOGIC CHANGES OF
                   UNCERTAIN SIGNIFICANCE
                    POLLUTANT BURDENS
                                                         \
              PROPORTION OF POPULATION AFFECTED •
  Figure 1. Spectrum of biological response to pollutant exposure.1

-------
30
25
£ 20
0
c/o
K
UJ
Q_
O
o
S 15
UJ
ca-
ll.
•=2
0
cc
o
10
5
COMMUNITY EXPOSURE








•
15.7


LOW
INTERf
i MT ror
m 1 tnl
HIGH
11.7

MEDIATE
MEDIATE
11.7
v:::::v::::::;

9.4
•
II
•
•Xv.vXv.
X'X'I'X'X'I
mm

16.9

14.5

16.5
:::X::;X::j:v
26.4

1 ^
I
'XvIvivX
.••••"•"•"•*» «V
1
^R%%
•x-x-x-x-:
x'$Sx$§
"**••**• * i






                                                       >3  (n-7763)
                              RESIDENCE DURATION, years

Figure 2.  Salt  Lake Basin:  incidence rates for croup in children by residence
duration and sulfur oxides exposure.^
                                    10

-------
   12
           COMMUNITY EXPOSURE

                        POOLED LOW


                        POOLED HIGH
o





J
•
               8.1
                                                              6.2
                                                                           11.2
                <3(n=1468)         RESIDENCE DURATION, yea.
         Figure 3.  Idaho-Montana:  Incidence rates for croup in children by residence
         duration and sulfur oxides exposure.2
                                          11

-------
     25
£    20
£E
UJ
D_
«ar
OS
£
     15
     10
                                       WITH HIGH SULFATES
                                                               WITHOUT HIGH SULFATES
                  (26)
                                     I


                                    (71)
                                     I
                                     I
                                     I
                                     i
                            50
                                                                     150
                                      100

                            TSP CONCENTRATION,

Figure 4.  Salt Lake Basin:  asthma attack rates by total suspended particulate

level and  by category of sulfate content.2  (Minimum temperature >51 °F.)
200
oc.
CD
o

t
a.
uu
oc.
UJ
CD
•f
o
K.
                                        10               15               20

                                    DAILY SS CONCENTRATIONS, jig/n>3

        Figure 5.  New York:  Percentage of cardiopulmonary subjects reporting sympton

        aggravation by daily suspended sulfate (SS) concentrations.2
                                                                                          25
                                            12

-------
TECHNOLOGICAL  ALTERNATIVES
TO FLUE GAS DESULFURIZATION
              by

        Stephen J .  Gage
Council on Environmental  Quality
       Washington, D.C.

-------
              TECHNOLOGICAL ALTERNATIVES

             TO FLUE GAS DESULFURIZATION
                   Stephen J. Gage
           Council on Environmental Quality
                   Washington, D. C»
                     Introduction

     The single roost important factor today in controlling
air pollution from stationary industrial sources is the
reduction in emissions of sulfur oxides.  The omnipresent
sulfur in domestic coals and in both domestic and imported
oils poses a serious challenge to the efforts of Federal,
state, and local environmental agencies to significantly
diminish the release of over 30 million tons of sulfur
oxides into the atmosphere each year.

     Natural gas is, of course, the cleanest fuel but it is
already in short supply and, even with accelerated
domestic development and importation of LNG, demand will
undoubtedly outstrip supply.  Also new supplies are likely
to be bid away from industrial and utility uses for
residential and commercial uses.  Domestic oil production
has leveled out and it is expected that much of our future
oil needs will be met by imports of foreign, mostly Mid-
Eastern, oils.  But these oils generally contain sulfur in
concentrations which will, in many cases, preclude use of
the residuum without some measure of hydrodesulfurization.

     Desulfurization of imported oil certainly represents one
strategy for meeting the.need for low sulfur fuels.  But
such a solution would spawn a set of new problems.  There
are the fundamental problems of national and economic
security arising from reliance on large imports of a vital
fuel.  Since new refineries equipped for desulfurization of
residual fuel would be required, siting and environmental
problems of supertanker ports and refineries emerge.  If
such facilities are located offshore, there is further
aggravation of the balance-of-payments problem.  However,
importation of crude oil for desulfurization in this country
and of desulfurized residual fuel will be important in
meeting our clean fuel needs.  The technology for

                           14

-------
desulfurization of oil is reasonably well established so it
will not be covered in this analysis of technological options.

     Coal, our most abundant fossil fuel resource, presents
the serious environmental challenge.  As the traditional fuel
used in utility and industrial power plants in large sections
of the nation, particularly in the industrial Midwest, coal
is the source of a significant fraction of the sulfur oxides
emitted in those areas.  Nearly all of the coal in the
Central coal region has sulfur content exceeding 3 percent
and much of the coal in the Appalachian coal region has
sulfur content in the 2 to 3 percent range.  In the eastern
half of the country, only Central Appalachia* has significant
deposits of low sulfur coal.  And, unfortunately, the terrain
in this area is very mountainous so the a11-too-common
ravages of surface mining—landslides, slumps, massive
erosion, and acid mine drainage—are even more severe.

     Most of the coals in the vast sub-bituminous and lignite
fields of the West have lower sulfur content but their use
to solve air pollution problems in the Midwest appears to be
limited to the western fringe of the industrial region
(Minneapolis, Chicago, Kansas City) because of the high cost
of transportation from Montana and Wyoming.  Although Western
coals are already being used locally, limited water avail-
ability and problems with re-establishing vegetation on
mined lands are potential constraints.

     Technology which removes the sulfur from the coal before
combustion or from the gases during or after combustion is
therefore essential if air quality is to be improved and if
coal production is not to be displaced by a flood of imported
oil.  Maintenance of a viable coal industry is also essential
if part of our clean fuel needs are to be met in the 1980's
with synthetic fuels produced from coal.
*Southern West Virginia, western Virginia, eastern Kentucky,and
 northeastern Tennessee
                            15

-------
     One technology for reducing sulfur oxide emissions from
coal-burning plants is the subject of this symposium—flue
gas desulfurization or stack gas cleaning.  The past year has
seen both significant progress in the development of commercial
stack gas control systems and a rising chorus of concerns
about the application of such systems.  Questions about the
reliability, costs, and associated environmental effects of
stack gas control systems have generated a more fundamental
question:  Because of the current and potential problems with
stack gas cleaning devices, and the promise of precombustion
coal treatment processes, shouldn't application of sulfur
oxide abatement technologies be delayed until the new coal
cleaning processes become available?  Even if the premises to
this question are conceded, the answer is not a simple yes
or no; rather the answer may be both no, in the near-term,
and yes, in the intermediate-term.

     This paper will attempt to place into perspective the
current status of stack gas cleaning systems relative to the
status of coal cleaning processes.  It will review the status
of the coal precombustion cleaning processes, estimates of
their costs, and forecasts of their application.  The focus
will be on liquefaction and gasification processes which
will convert high sulfur coals into usable clean fuels in
either liquid or gaseous form.
         Status of Coal Liquefaction Processes

     One approach to precombustion cleaning of coal is to
liquefy the coal and to remove the ash and sulfur from the
liquid phase.  There has been considerable experimental work
over the past decade to produce a synthetic crude oil which,
in turn, could be used in the manufacture of gasoline.
Conversion of coal to a liquid fuel requires the addition of
hydrogen, but considerably less hydrogen is required to
produce a power plant fuel than is required to produce feed-
stock for a gasoline plant.  In fact, a liquid product is not
required for power plant application, although a liquid
phase is required for the hydrogenation and separation of ash
and sulfur.

Solvent Refined Coal

     One of the most advanced process for producing an
ashless, low-sulfur power plant fuel is known as solvent
refining [1,2,3].  In this process, pulverized coal is first
                           16

-------
mixed with  a heavy  aromatic solvent  (derived from coal)  and
then passed through a reactor under  a  hydrogen atmosphere  at
about 1000  psi  and  800°F.   Small  quantities  of hydrocarbon
gases, hydrogen sulfide,  and light liquids  are formed  along
with a heavy organic material called Solvent Refined Coal
 (SRC).   The gases are separated and  treated;  the  solution
is  filtered to  remove the  ash and some undissolved coal; the
solvent  is  recovered through flash evaporation.   A schematic
diagram  of  the  SRC  process  is shown  in Figure 1.

     The SRC is a black pitch-like substance with a melting
point of about  350°F containing about  0.1 percent ash.   Its
heating  value is about 16,000 Btu's  per pound regardless
of  the quality  of the coal  feedstock.   Most  of the
inorganic sulfur is  removed in this  process  and as high  as
60-75 percent of the organic sulfur.   For instance, it
appears  possible to  convert Illinois coal with 3.5 percent
sulfur to SRC with  0.6 percent sulfur.  With the  higher heat
content,  SRC with 0.6 percent sulfur would release about 0.4
pounds of sulfur per million Btu.

     A 50 ton/day pilot plant is  now being constructed near
Tacoma,  Washington by Pittsburgh  and Midway  Coal  Mining
Company, one of the  developers of the  SRC process  [4,5,6,].
'This project is being  sponsored by the  Office of  Coal Research,
Scheduled for completion in  early 1974, this  plant should
Provide  the engineering data necessary  to build a commercial
Plant.   A smaller pilot plant, 6  ton/day, is  being constructed
under the sponsorship  of the  Southern  Company and the Edison
Electric Institute and is scheduled  for completion in late
1973.  This plant, which will  use  a solvent refining process
developed by Consolidated Coal Company, will be built near
Wilsonville, Alabama  [4,5].

H-Coal Process

     The H-Coal process,  developed by Hydrocarbon  Research,
Inc. (HRI),  introduces a coal  slurry into a ebullating bed
reactor where,   in the presence of  a catalyst  (cobalt
molybdate),  hydrogenation occurs  [2,3,5].  The upward flow
of the feed—which consists of crushed  coal mixed with
recycle oil and hydrogen—maintains the catalyst  in a state
of rapid motion and permits the continuous passage of
Unconverted coal and ash from the reactor.  The H-Coal
reactor operates at 2700 psi and 850°F.  The problems of
introducing feed into the pressurized reactor vessel and
separating the unconverted coal and ash from the synthetic
                            17

-------
Mining
                         Fuel Gas
           Hydrogen (30-40  Ibs/ton Coal)
Coal
Prep'n
                                               Vent Gas
       Power
                         Vent
                         gas

                         Solv.
                         Ret.
                      Energy.
1
RESIDUE
DRYING^
j

FURNACE


Stack
gas
cont. '
en

1
ACID GAS
REMOVAL

H2S
SULFUR
PLANT
                                 Ash
                                               Sulfur
         Solvent Recycle
                                                                                     Chem. Prod.

                                                                                     Light Oil

                                                                                     Gas  Oil
                                                                                     De-ashed
                                                                                     Fuel, Lig.
                                                                                     or Solid
            FIGURE  1.   Solvent  Refined  Coal  (SRC)  Process

-------
oil await  final  solution.  HRI has  stated that the processing
conditions can be varied to obtain  1.5 to 1.0 percent sulfur
without solids removal, 0.5 percent sulfur with ash removal
and increased conversion severity,  and 0.25 percent sulfur with
secondary  coal conversion  [3].  The H-Coal process has been
tested in  a 2.5  ton/day pilot plant with high conversion and
good  catalyst recovery and without  serious difficulties  [1,5].
HRI has stated interest in building a 250 ton/day demonstration
plant, but the financing for such a plant has not been
secured [5]. A schematic diagram is shown in Figure 2.

COED  Process

      Another process which produces a liquid product from coal
is the COED (Coal Oil Energy Development) process [2,3,4],
Developed by Food Machinery Corporation  (FMC) under sponsor-
ship  of the Office of Coal Research, this multiproduct process
uses  fluidized-bed pyrolysis.  A synthetic crude oil, a gas
stream, and a char product are produced.  Powdered coal  is
subjected  to increased temperatures in four successive beds.
Volatile liquid  products are withdrawn at every stage and not
subjected to the higher temperature where further cracking
would occur.  The liquid fuels yield is lower than for coal
hydrogenation.  Further, the sulfur content of the product
char may be too high to meet air quality standards in many
areas, without further desulfurization.  The COED process can
t>e modified somewhat to yield different product mixes, i.e.,
the hydrogen-rich gas stream can be used to convert more of
the char to fuel oil.  Pilot plant  experience has indicated
yield of about 1.2 barrels of fuel  oil and 9000 standard
cubic feet of gas per ton of coal.  FMC has been operating a
36 ton/day pilot plant near Princeton, New Jersey, for
several years to obtain process information on various types
of coal [4].  No plans have been announced for a scale-up
to a  demonstration plant.

.Coal  Hvdrodesulfurization

      Another promising coal liquefaction process being
developed by the Bureau of Mines (BOM) promotes
hydrodesulfurization of the coal slurry by turbulent flow
with hydrogen through a fixed bed of catalyst [1,3,7,8].
The coal,  suspended in coal - derived product oil,  is pumped
with hydrogen through a packed bed  of pelletized cobalt
ttolybdate alumina.  Turbulent flow  through the bed prevents
plugging (as the coal passes through its stickly phase prior
to becoming liquid), keeps the catalyst surface clean for
                           19

-------
                           Vent Gas
   Hydrogen
Mining
Preparation
         ©HSLUE
               PREHEATER
                           OVERHEAD

                           SEPARATOR
H-COAL
REACTOR

2700 psi
850 F
CCTOolybdateCat
                           Cat.I   F   I
                           Liq.L, ,_!
                           Disengaging
                           EBULLATED
                           CATALYS3J
           Note: P=Pump

          	  Recycle Slurry Oil
         Hot Oil
         Recycle
_». Light Liq. Hydrocarbon

   Atm. Bottoms Product
                                                                      Overhead Product
 >*• Vac^ jgttoma Slurry

   fetin. Flashed Slurry
                                                               i-Char-Oil
               FIGURE 2.   H-Coal  Process

-------
good contact, and promotes transport of the hydrogen into
the slurry.  The reactor operates at 2000 psi and 850°F.
The goal of this design is to do just enough hydrogenation
to remove the sulfur with the liquefaction of the coal
essentially being a side effect, therefore use of expensive
hydrogen is minimized.  Sulfur is removed as H2S which can be
easily converted in elemental sulfur.  Experiments have
been conducted with a 0.1 ton/day reactor on 5 different
coals.  For example, a Kentucky coal having 4.6 percent
sulfur and 16.0 percent ash was converted into a synthetic
fuel oil having only 0.19 percent sulfur and 1 percent ash.
BOM is now scaling up the size of the operation and is de-
signing a 5 ton/day pilot plant.
        Status of Coal Gasification Processes

     One promising approach to precombustion cleaning of
coal is gasification.  Because of growing shortages in natural
gas supply, interest is running very high in gasification
processes which can produce a pipeline quality gas, i.e., a
synthetic gas composed primarily of methane with heat content
approaching 1000 Btu per standard cubic feet.  More recently,
interest in producing  a low-Btu content gas as a clean-
burning fuel for power plants has risen sharply.  Because
high-Btu synthetic gas would be too expensive, e.g., $1.00-
1.50 per million Btu, for power plant use, this review will
concentrate on the direct use of low-Btu synthetic gas in
power plants.  However, in order to assess the status of the
several gasification technologies, it will be necessary to
Review high-Btu gasification research.

     For the sake of simplicity, coal gasification to low-
Btu gas can be divided into two categories: present
generation gasifiers (PGG) and advanced gasifiers (AG).
Present generation gasifiers include coal gasification
processes—Lurgi, Koppers-Totzek and Wellman Galusha—that
are now commercially available.  In this country, industrial
interest is focused on the Lurgi process.  Plans for
construction of full-scale commercial coal gasification
plants for production of pipeline quality gas have been
announced by two companies.  In addition, one utility has.
initiated construction of a combined-cycle (gas turbine-
steam turbine)  demonstration plant using a Lurgi gasifier to
produce the gas.
                          21

-------
     As mentioned above, research on the production of
pipeline quality gas from coal has been receiving con-
siderable governmental and industrial support.  A number of
advanced gasification processes are under development and a
cooperative government-industry effort, comprised primarily
of the Interior Department and the American Gas Association,
is currently evaluating four processes.  Pilot plants are in
operation for two of the processes and under construction for
another two.

     With utility interest in low-Btu gas for combined-cycle
plants growing, several study efforts have been initiated to
evaluate the possibility of modifying the advanced gasifiers
now under development to produce low-Btu gas.  Since low-Btu
gas is produced in the initial stages of all of the advanced
gasifiers (prior to methanation of the hydrogen and carbon
monoxide), the proposed modifications appear to have merit.

Present Generation Gasifiers  (PGG)

     Coal gasification processes involve reacting coal,
steam, and oxygen under heat and pressure in a specially
designed reactor to form a synthesis gas of hydrogen, carbon
monoxide, carbon dioxide, sulfur componds and varying amounts
of methane [1,2,3,9,10].

     The Koppers-Totzek and Wellman Galusha processes operate
at atmospheric pressure, produce no methane directly, and
require a rather high oxygen consumption [1*4].  The Lurgi
gasifier is the only high pressure gasifier (ca.300-450 psi)
commercially available [1,10].

     The Lurgi Gasification Process.  In the Lurgi process,
the synthesis reaction takes place in a water-jacketed vessel
[2,11].  The crushed coal (with the fines briquetted with
process tar)  is fed through a coal lock chamber downward
into the gasifier.  Steam and oxygen are introduced at the
bottom of the gasifier, heated by the high temperature ash at
the bottom,  and flow upward through the reaction zone.  The
crude gas and unreacted steam are quenched upon leaving the
gasifier to cool and remove dust, tars, ammonia, and
phenols.  A cross-sectional view of the Lurgi gasifier is
shown in Figure 3.
                           22

-------
FIGURE 3.  Lurgi Gasification Process
                 23

-------
     For the production of pipeline quality gas, relatively
pure (>98 percent) oxygen must be supplied to the gasifier,
since any nitrogen dilutes the pipeline gas [11].  The crude
gas leaving the gasifier at 700 to 900°F will, in this case,
have a heat content, above 400 Btu per cubic foot.  For
pipeline quality gas, the crude gas is subjected, in turn, to
a catalytic shift conversion process to establish the H2~to-
CO ratio required for the subsequent methanation step, a
physical absorption process to remove sulfur compounds and
carbon dioxide, and a methane synthesis process which
catalytically combines carbon monoxide and hydrogen to form
methane and water.  The resulting gas, when compressed and
dried,  has a heat content of 950 to 970 Btu per cubic foot
[11] .

     On the other hand, for the production of low-Btu gas,
air may be fed to the gasifier, yielding a gas with a heat
content of less than 200 Btu per cubic foot [1,2,10,12,13].
Before using in either a conventional gas-fired boiler or in
a combined-cycle power plant, the sulfur componds (largely
H2S with some COS) must be removed.  Although there are a
variety of available gas cleaning processes (Lurgi's pro-
prietary Rectisol process which uses low temperature methanol,
Stretford process, potassium carbonate scrubbing, etc.), all
require cooling the gas to 100 to 200°F [1,2,13].  This
requirement reduces the overall thermodynamic efficiency of
the gasification-power plant system.  This represents a
serious fuel and cost penalty in the design of a combined-
cycle plant.

     At the present state of technology, sulfur compounds
must be removed from the gas before it is introduced into the
gas turbine because of the sensitivity of the turbine blades
to sulfur exposure.  However, the loss of efficiency in
cooling and reheating the gas for scrubbing significantly
degrades the attractive theoretical efficiency of the
combined-cycle.  Research on processes to remove the sulfur
compounds in a high temperature environment is underway.
Development of a high temperature gas cleaning process is
probably essential for widespread application of both PGG and
AG systems.

-------
      Commerical  Applications  of Lurgi  Gasifiers.   Abroad,
 the  Lurgi  gasifiers  have  found  commercial  use  in  India,
 Australia, West  Germany,  South  Africa,  USSR, and  UK  [9,11].
 Recently,  Lurgi  gasifiers have  been  installed  to  provide
 synthetic  gas  for  a  165 MWe combined-cycle power  plant  in
 Luenen,  Germany  [5,10].

      Two full-scale  commercial  coal  gasification  plants have
 recently been  announced  [11,14].  Both are to  be  located near
 Burnham, New Mexico,  and  are  to use  strip-mined coal  from  the
 Navaho Indian  Reservation.  One will use 8.5 and  the  other
 9.7  million tons per year of  coal.   Each plant will employ
 approximately  30 Lurgi gasifiers to  produce 250 million cubic
 feet of  970 Btu  per  cubic foot  gas.  One plant is planned  by
 El Paso  Natural  Gas  and the other by the combination  of
 Transwestern Coal  Gasification  Company, Pacific Coal
 Gasification Company, and Western Gasification Company.  The
 Plants are scheduled to go on line in  the  1976-1978 period.

      Commonwealth  Edison  is assessing  the  feasibility of
 using Lurgi technology to gasify Illinois  coal for direct
 firing or  in a combined cycle at its Powerton  Station [4,12],
 Called the "Clean  Power Fuel Demonstration Plant," this
 project  seeks to demonstrate Lurgi technology  with American
 coals for power plant operation.  A  schematic  of  the
 gasification plant for the Powerton  Station is shown  in
 Figure 4.  The project, sponsored in part  by the  Edison
 Electric Institute, will  involve 3 Lurgi gasifiers and is
 slated to be in operation in late 1975.

j^dyanced Gasifiers (AG)

     As  indicated above,   a number of advanced  gasifiers
 are under development.  Four of  the processes  are now at the
Pilot plant stage while a number of other promising systems
 are  still in the experimental stage.  Because  of  the
Possibility that some of these advanced gasifiers can be
 adapted  for production of low-Btu utility  gas,  they will be
briefly  reviewed below.

     Advanced Gasifier Pilot Plant Studies.  The AG processes
which are at the pilot plant stage are indicated  in Table 1,
 along with the developmental and sponsoring agencies
 and  the pilot plant status [1,2,3,4,9].
                           25

-------
                   FIGURE 4.  Commonwealth Edison's Clean  Power Fuel
                              Demonstration Plant
                                                                  EXPANDER /COMPRESSOR /
                                                                       GENERATOR
                                          DESULFURIZATION
                                              SYSTEM
WATER
                                                                             EXISTING WWERTON *4
                                                                             120 MW STEAM TURBINE
                                   -13-

-------
     All of the  four employ  fluidized bed reactors with the
synthesis gas glowing through a coal bed.  All processes use
the shift reaction, purification, and methanation steps
described above  for production of pipeline quality gas with
the Lurgi gasifiers.  Schematic diagrams of the  four pilot
plant projects are presented in Figure  5.
                       TABLE 1

    Summary of Advanced Coal Gasification Processes
                 in Pilot Plant Stage
Process

1•   HYGAS
    (Electrothermal)

2.   CSG (CO2 Acceptor)

3•   BI-GAS


4•   Synthane
Development
Agency	
IGT
Consol
BCR
BOM
Sponsoring
Agency	
OCR-AGA
OCR-AGA
OCR-AGA
BOM
Pilot Plant
   Status
In Operation

In Operation

Under Con-
 struction
Under Con-
 struction
     IGT:   Institute for Gas Technology
     BCR:   Bituminous Coal Research
     Consol:Consolidated Coal Company
     BOM:   Bureau of Mines, Department of Interior
     OCR:   Office of Coal Research, Department of Interior
     AGA:   American Gas Association

     Methane is produced in the AG by (a) devolatilization
°f the coal and (b) reaction of freshly devolatilized coal
with H2 and CO present in synthesis gas and steam introduced
into the reactor  [2].  The four AG systems now in the pilot
Plant stage have a number of similarities but basically
Differ on the mode of production and composition of the
synthesis gas supplied to the reaction of H2 and CO with
freshly devolatilized coal.
                           27

-------
Hygas-electrothermal process
      Fuel
   iteom ond powei <;eneiolion
                             High BID goi
                                   OGJ
Bigas process
                                                         low got
                                               Stjam
                                                                           Hijh-BIUiat
Carbon dioxide acceptor process
      Synlheiii
        ]«
 Dry lignilt
  Steam
Synthane process
                                                                           Hoi corfconoti
                                                  IttUui
        FIGURE  5.   Schematic Diagrams of Advanced Gasification
                     Pilot  Plants
                                            28

-------
     Three of the processes—HYGAS, BI-GAS, and Synthane—
are fairly similar, in that ground coal is reacted with
steam and oxygen to produce a synthesis gas [1,2].  If the
coal is of the caking variety, it is typically pretreated
with oxygen to render it noncaking.  The coal is  then
introduced  into the reactor vessel which operates around
1000 psi, the higher pressure shifting the equilibrium mixture
to higher concentrations of methane.

     In the HYGAS-Electrothermal process, the hydrogen-rich
synthesis gas is produced by electrically-heated
gasification of char.  In the BI-GAS process, oxygen is fed
to a high-temperature stage with recycled char and steam to
provide energy for production of the synthesis gas.
Similarly, in the Synthane process, synthesis gas is produced
by oxygen and steam reaction with char.

     The C02-Acceptor process uses quite a different method
for supplying heat to the gasifier to generate the
synthesis gas.  Process heat is carried to the reactor by
calcined dolomite (MgO-CaO) which releases both sensible and
chemical energy as it absorbs C02.  The calcined  dolomite is
regenerated with char in a separate fluidized-bed reactor.
This allows elimination of the electrothermal gasifier or
oxygen plant but reduces the gasification temperature so that
only the most reactive coals such as lignites can be used and
increases the complexity of the system.

     Other Advanced Gasification Processes. In addition to
the four processes just described, there are a number of
other processes for which data has been obtained  in
experimental units [1,2,9].  The Institute of Gas
Technology is developing two other process, both of which
use the same gasifier as employed in the Hygas-Electrothermal
process.  In the HYGAS-Oxygen process (similar to the
Texaco process), oxygen is supplied to a separate char
gasifier, the synthesis gas from which is purified by the
removal of CC>2, the addition of steam, and catalytic shifting
before being fed into the hydrogasifier.  In IGT's Steam-
Iron process, synthesis gas produced in an air-blown char
gasifier is used to reduce iron ore Fe304 to Fe + FeO.  The
latter is used to decompose steam to hydrogen, reforming
Fe304 which is recycled.  The steam is fed to a hydrogasifier.
                           29

-------
     The Bureau of Mines is developing a second gasification
process called the Hydrane process which offers the unique
possibility of directly converting raw coal to methane by
reacting the coal with hydrogen  [1,15].  While the economic
advantages of such an approach are well known, the tendency
of most American coals to agglomerate has led to a
pretreatment stage in which the coal feed stream is partially
oxidized.  Such pretreatment also reduces the coal's
reactivity for methane production, thus requiring the indirect
steps of steam-oxygen gasification, water-gas shift reaction,
and methanation.  In the Hydrane process, almost all  (up to
95  percent) of the methane is produced by direct hydro-
gasification of raw coal in a "falling dilute phase" reactor
[15].  The remainder of the pipeline methane is produced by
methanation of carbon monoxide which constitutes 3 to 4
percent of the synthetic gas.  The hydrogen for the hydro-
gasification is produced by gasification of the residual char.
The Hydrane process has been tested at bench-scale in a 3-
inch diameter reactor.  The process offers the promise of both
high thermal conversion efficiency and low gas production costs,

     Two other processes, both of which gasify coal with
steam and oxygen in molten baths, deserve mention.  The
Kellogg process uses a molten sodium carbonate bath and the
ATGAS process uses a molten iron bath with a limestone flux
[1,2,4,16].  In the Kellogg process, the sodium carbonate
strongly catalyzes the steam-coal reaction, permitting
complete gasification at temperatures low enough to allow
methane production in the gasifier [1,2,16],  The
gasification temperatures can be lowered from 1900°F to
1700°F.  The molten carbonate also supplies heat to the
reaction, disperses the coal and steam uniformly throughout
the reactor, permitting the direct gasification of
agglomerating coal, and reacts with the sulfur to release
H2S in the product gas.

     In the ATGAS process, the coal is dissolved in the molten
iron bath, releasing organic and inorganic sulfur
constituents [1,4].  Because of the high affinity of sulfur
for iron, iron sulfides are formed which migrate to a slag
floating on the molten iron bath.  The iron sulfides react
with the limestone flux, releasing the iron and fixing the
sulfur as calcium sulfate which must be discarded.  The
carbon which is dissolved in the molten iron is reacted with
steam and air to produce a sulfur-free synthesis gas of CO,
H2/ and nitrogen oxides.  Little methane is produced directly


                           30

-------
so the indirect steps of shift reaction and methanation are
required to increase the heat content of the gas.  It may be
possible to use the synthesis gas directly in a combined cycle
plant.  The process operates at atmospheric pressures and
agglomerating coals can be gasified without devolatilization.
Outstanding problems with this process include lance con-
struction materials, refractory lining, and control of
particulate emissions.  Experimental design data has been
obtained on a gasifier of about 2MWe capacity [4].

     Adaptation of AG Processes for Low-Btu Gas Production.
The possibility of adapting AG processes for the production
of low-Btu utility gas is being investigated in a number of
laboratories.  Early efforts have already verified the
potential for such adaptation as well as have identified some
of the unresolved problems [13,17].  Because of the loss of
efficiency associated with cooling the synthesis gas for sulfur
removal (as described above),  probably the major unresolved
problem is the yet undeveloped technology for hot cleanup of
tar,  dust,  and sulfur from the synthesis gas.

     It should be noted that,  while the Office of Coal
Research has been given the responsibility for developing low-
Btu gasification processes as well as those for pipeline gas,
the trend in both OCR and industry funding is not supportive
°f adaptation of the largely developed AG processes.  Rather,
OCR and industry appear to be pursuing new AG processes
specifically designed for low-Btu gasification.

     Design of New AG Processes for Low-Btu Gas Production.
within the past two years, several new AG systems have been
Proposed for production of low-Btu utility gas production.
The major difference between these designs and adaptation of
existing high-Btu.systems is that the new designs are
optimized for production of low-Btu gas.  Although several
e£forts are underway,  probably the most advanced is the team
e£fort headed by Westinghouse and partially funded by OCR.
^e team also includes Public Service Indiana, Amax Coal
Company, and Bechtel [18,19],   This effort would be oriented
toward developing a total combined-cycle system using a
fluidized combustor to produce a low-Btu (150 to 200 Btu
Per cubic foot)fuel gas.  Lime would be added to the de-
volatilization stage where it reacts with the sulfur, producing
calcium sulfide particles.  The synthesis gas emerging from
the reactor still has to be cleaned of particulates but the
9as is relatively free of sulfur compounds.  The cleaned hot
9as can then be combusted to drive first a gas turbine and
                           31

-------
 then give  up the  remainder of its  heat to steam boiler.

      Plans call  for a  pilot plant  to be constructed in  the
 Pittsburgh area by mid-1974 to be  followed by a commercial
 plant to be constructed  at a public  service Indiana plant
 near Terre Haute  by late 1977.

     Other  industry teams  are also working with the Office
of Coal  Research  to define design characteristics for low-Btu
AG and combined cycle systems  [4].  One team  is composed of
Combustion  Engineering and Consolidated Edison.  Another team
is made  up  of Pittsburgh and Midway, Poster-Wheeler, and
Pratt and Whitney.

     Outstanding problems  for all of these processes include
advanced gasifier design, hot gas clean-up, and advanced turbine
materials technology.

      There are a  number  of other coal  gasification  processes
 in various stages of R&D which have  not been mentioned  in
 this review.  Also the classification  of the processes  reviewed
 has  been somewhat arbitrary,  e.g.,  the ATGAS process could
 have placed in the last  category considered since it appears
 to have  greater potential for production of utility fuel than
 for  pipeline gas.   The omissions and commissions reflect only
 the  need for brevity and the judgment  of the author.


             Status of Advanced  Combustors

     There  is another class of emerging technology which
should be mentioned because it bridges the gap between con-
ventional coal-fired boilers and coal gasification process.
This technology involves the combustion of coal in the
presence of a sulfur acceptor such as limestone.  Because
of the necessity  to continuously remove the ash and the
reacted  and unreacted limestone, it is necessary to use a
moving or fluidized-bed combustor.  The first generation
combustors  operate at atmospheric pressure.   In addition to
reducing sulfur emissions, such  combustors may, because of
lower flame temperatures than in pulverized coal boiler, re-
duce nitrogen oxide formation and may offer cost savings.

     One systems which should be mentioned is that developed
by Pope, Evans,  and Robbins  [2].  This fluidi«ed-bed boiler
has operated successfully  in the pilot plant  stage and
is ready for scaling to a demonstration plant.  The Office
                           32

-------
of Coal Research has recently granted Pope, Evans, and
Robbins a contract for the design, construction and
operation of a 30 MWe fluidized-bed boiler.  In England, the
National Coal Board is developing an atmospheric, fluidized-
bed combustor.  Partial funding for these projects has been
provided by the Environmental Protection Agency and the
Office of Coal Research.

     These systems do involve changes in boiler design and
manufacture and boiler operation which, although much less
extensive than with coal gasification, have represented a
barrier to acceptance by manufacturers and utilities.  The
general opinion seems to be that, if significant departures
from conventional boiler design are to be made, then it is
logical to go the extra step to gasification with potential
for much improved thermal efficiency and decreased environ-
mental problems or to liquefaction with potential for
decreased environmental problems.  Similar reasoning seems
to apply in the case of stack gas cleaning which is viewed
t>y many utility managers as a troublesome, short-lived
solution.

         Comparison of Alternative Technologies

     The alternative technologies described above embody
different physical processes which are introduced into the
utility fuel cycle in different ways, e.g., liquefaction may
produce a new fuel to substitute directly for coal while
gasification may require an integrated gasifier-gas turbine-
steam boiler system.  Further, the various technologies are
at considerably different stages in the R&D process.  In a
number of. cases, engineering studies of systems incorporating
the technologies have not been conducted.  These factors
niake it quite difficult to compare the alternative
technologies on a common basis.  Insofar as possible, the
technologies will be compared on the basis of  (1) applic-
ability for power plant use;  (2) capital and operating
costs; and (3) associated environmental factors.  Because
of the limited utility interest in advanced combustors, the
comparison will be limited to coal liquefaction, present
generation gasifiers, and advanced gasifiers.

Applicability for Power Plant Use.

     The major difference between the coal liquefaction and
gasification processes for utility boilers is that liquefield coal
products have high energy content (16,000 Btu per pound) and are
in a physical form convenient for both transportation and storage

                          33

-------
while  the  low heat content  and  physical  form of  low-Btu  gas
from coal  is  such  that it can be  neither transported (by
pipeline)  nor stored.

     Coal  Liquefaction. In  terms  of physical  suitability,
liquefied  coal products  (LCP's) could,with minor boiler
modifications, be  used in nearly  all utility  and industrial
boilers.   For instance, SRC could be used in  either  the
pelletized solid form  or, if  heated, could be handled as a
viscous fuel  oil.   Coal liquefaction would probably  be
accomplished  at a  large centralized, possibly mine-mouth,plant
and  could  probably thereby  achieve economies  of scale.   The
ash  and sulfur disposal problems  at the  processing plant could
be quite significant and may  severely limit the number of
potential  plant sites.  The sulfur would probably be converted
to elemental  sulfur which could be more  easily stored than
many sulfur compounds.

     LCP1s,largely ash-free,  still contain some sulfur.  SRC
with 0.6 to 0.7 percent sulfur could be  used  in nearly all areas
of the country under the New  Source Performance Standards and
in most areas under standards established by  State
Implementation Plans or by  state  and local regulations.  But
the  sulfur content in  SRC just slips under the requirements
in many areas and  is too high in  some such as New Jersey and
New  York.  If allowable sulfur concentrations are reduced in
the  future, SRC would  probably not be adequate.  By
comparison, the sulfur emissions  with SRC are as much as twice
those apparently achievable with  stack gas cleaning.  Other LCP
processes, such as the Bureau of  Mines turbulent catalytic
hydrodesulfurization,  have  demonstrated  the ability  to
reduce the sulfur  content by  one-half or even two-thirds
below that achievable  with  SRC.   Supposedly such low con-
centrations can be  achieved with  H-Coal  although the matter is
somewhat confused  at this point [5 ].

     Other factors which may  affect the  applicability of LCP's
are  supply and cost.  Much  of the particulate and sulfur
emissions  in urban  and industrial areas  comes from industrial
rather than utility boilers.  Many of these boilers, because
of space limitations and highly unfavorable economics, cannot
be retrofitted with SGC or gasification  systems.  LCP's
represent,  therefore,  the only technological  alternative to
low  sulfur oil or  natural gas.  As LCP's become commercially
available,  in limited  supply  at first,  industrial boilers
may pay the premium price for the clean  fuel  Il8l•  At the
same time,  utilities may elect to use LCP's for their older,
                           34

-------
 less efficient,  lower  load capacity boilers which cannot
 be  retrofitted with SGC systems and, under regulatory and
 market pressures, retrofit their newer boilers with SGC
 units or use low sulfur coal.  LCP's may provide the option for
 utilities to continue  to use their older and smaller units
 with reduced air pollution to meet peak loads—this could be
 extremely important if siting problems, new plant startup
 difficulties, and regulatory delays persist.

     Present Generation Gasifiers.  If the synthesis gas is
 only to be combusted in a modified conventional boiler, then
 PGG systems can be used with both new and existing plants.
 However, to offset the increased cost of the Lurgi (or other)
 gasifiers, a combined-cycle must be used to increase the
 thermal efficiency.  Commonwealth Edison plans to install a
 gas turbine as part of the gasification demonstration
 project at its existing Powerton Station.  It appears then
 that the application of PGG system will be restricted
 essentially to new plants.  There may be some limited
 application of PGG systems for conventional boilers until hot
 sulfur and ash clean-up processes and improved turbine blades are
 developed but PGG systems will generally be applied on new systems,

     Advanced Gasifiers. For the reasons just given,  AG
 systems will be applied only to new power plants.  All AG
 systems, either as adaptations of high-Btu content gasifiers
 or as new designs involve the integration of a gasifier with
 a combined-cycle unit.

.Comparative Costs of Alternative Technologies

     A most important  comparison,  and the one most fraught
with uncertainties,  is that of costs,  both capital and
operating.  Cost analyses of emerging technologies are
hazardous enough, but  estimates of capital and operating
costs for processes only at the level of bench-top experiments
approach speculation.   Also because of recent cost escalations
it is very important to compare costs calculated in the same
period and preferably  as recently as possible.   Cost analyses
as recent as 1970 and  1971 may be quite out-of-date.   Some
recent cost data will be presented to indicate,  at least,
relative economic attractiveness.
                           35

-------
     One recent comparison based on conservative assumptions
casts an equally jaundiced eye at all technological alter-
natives [5].  This sobering comparison prepared by Chem Systems
Inc. should bracket the maximum costs that may be anticipated.
These costs estimates are given in Table 2.

                       Table 2

     Cost Comparison of Technological Alternatives
                  for Sulfur Control

Technological     Total Annualized      Total Capital      ,
Alternative       Energy Costs,=*        Investment Costs,—'
                    jzf/106 Btu              $/KWe

Stack Gas
  Scrubbing*          80-85                104
Low-Btu Gas           90-95                132
Liquefied Coal
  Product             85-90                (?)


^•Interior coal priced at $7/ton (or 30^/106 Btu) and a load
 factor of 0.70
2Includes contingency of 15% of capital expenses.
     *The costs of stack gas  scrubbing quoted by Chem Systems
were based on an analysis by  the Federal Power Commission.
The system analyzed was a 1000 MWe boiler using high  sulfur
Interior coal.  The system economics were penalized by
inclusion of  an electrostatic precipitator,  coal storage and
preparation facilities and ash, dust,  slag systems — the  entire
waste disposal system — with the wet lime/limestone scrubber,
giving $70.5 per KWe.  $17 million was  included  for utilities
and $13.5 million was included for contingency, yielding
$103.5 per KWe for total investment.   The individual  costs
were broken down as follows:

               Coal                     30>z< /!06Btu
               Operating Costs          14^
               Waste Disposal           lljzf /106Btu
               Capital Charges          3ljz? /106Btu
                                            /106Btu
Chem System indicates that the "relatively high estimated
costs may have been unfavorably influenced by Commonwealth
Edison's recent experience on the Will County stack gas unit" [5]
                            36

-------
     The estimates of low-Btu gas costs were developed by
the Federal Power Commission Coal Gasification Task Force
for a 1000 MWe plant using a combined cycle.  The  onsite
process costs were $71.8 per KWe and the total plant invest-
ment including contingency was $132 per KWe.  The total energy
costs (in terms of gas) were 92ff per million Btu.

     The economics of LCP's are somewhat more complicated in
that the cost of the product is quite sensitive to size of
plant.  To illustrate the trends for SRC, the cost of SRC as
a function of coal feed to the plant is shown in Figure 6.
For a plant producing 7,700 tons of SRC per day* required
to fuel a 1000 MWe generator, the energy cost (in terms of
SRC) were 88^ per million Btu.  As envisioned, one LCP plant,
probably located at  the mine-mouth, would produce  enough
fuel  for several power plants.  It is suggested  that central
plants processing 40,000 to  50,000 of coal per day may eventu-
al ly be developed.

     There are a number of other cost estimates  for the
various technologies, however, it is indeed fortuitous when
the costs have been  computed on a common basis.  Three recent
publications gave cost estimates of electrical power production
from low-Btu gas, although not in directly comparable terms.
Lurgi indicates that, for a  330 MWe plant using  30jzf per million
Btu coal, the energy production costs would be about 10 mills
per Kw-hr and the capital costs would be about $200 per KWe  [19]
IGT has estimated that, for  a 1000 MWe plant using 24jz< per
million Btu coal, the capital costs would be $200 per KWe and,
assuming a 90 percent load factor, the total energy production
costs would be 6.9 mills per Kw-hr [20].  Finally, ATC has
estimated that, for  a 1000 MWe plant using 30^ per million
Btu coal, the capital investment is $163 per KWe and, assuming
a 90 percent load factor, the energy costs would be 6.8 mills
Kw-hr [21] .
     As mentioned above, cost analyses of LCP plants are
considerably more difficult because the effect of plant
size on the economics.  Several recent publications have
estimated the costs  of SRC and H-Coal, although  again the
costs are not on a comparable basis.  On the basis of an
analysis on a SRC plant processing 10,000 tons of coal per
day, Pittsburgh and Midway suggests that, as a rule of
     *In the Pittsburgh and Midway SRC process, this output
would require 13,300 tons of Interior coal.

                           37

-------
S.R.C. SELLING  PRICE
        /MM BTU
 i.oo- -
  .90
  .80
              10,000
20,000
30,000
40,000
50,000
                      COAL   FEED - TONS/DAY
          FIGURE  6.   SRC Economics as  a  Function of Plant
                      Capacity and Coal Price
                                 38

-------
thumb, SRC will sell at a price equal to the cost of raw
coal plus 35jz? per million Btu [ 6 ] .  For 3Q& per million
Btu coal, this gives an energy cost of only 65^ per million
Btu, considerably below the adjusted Chem Systems estimates.
Analysis by Chem Systems indicated for an SRC processing
44,000 tons of 35j£ per million Btu coal per day would yield a
fuel oil  (produced by blending enough of the lighter hydro-
carbons formed in the SRC process with the SRC to make
127,000 barrels per day of  a pumpable product) which costs  about
7Q& per million Btu.  The total plant investment would be
$209 million which,reflected in incremental capital costs to
electrical plants, would be about $63 per KWe [5].  For the
H-Coal process, two feed coals—both priced at 35jt* per million
Btu—were analyzed.  Processing Illinois seam coal with
3.4 percent sulfur at the rate of 29,000 tons per day resulted
in an energy cost of 83.5^ per million Btu while a Pittsburgh
seam coal with 4.2 percent sulfur at 25,000 tons per day
resulted in 88.8/z? per million Btu.  Both plants would
produce about 70,000 barrels per day of a pipeline product.
The incremental capital investments for the two plants would
be $78 per KWe and $90 per KWe, respectively.  It should be
noted, however, that the interest rates used for the LCP
plants are those for normal refinery financing, rather than
utility financing, as usually used for stack gas scrubbing
and for low-Btu gas production.

     Although not really germane to these comparisions, an
excellent common base cost comparison for pipeline coal
gasification was recently published [22].  To the extent
that some of the PGG and AG processes may be adapted for
low-Btu gas production, this is an informative comparison.
The economics of five processes are compared in Table 3.
The study also computed costs for coal at 35^ per million
Btu and for 9 and 15 percent gross return on rate base.
This study did not directly take into account the differences
between lignite and bituminous coals, although some of the
processes are operable only with specific grades of coal,
i.e., the CSG process must use lignite.

Associated Environmental Factors

     While it is too early to characterize all of the
environmental effects associated with the alternative
technologies, it is possible to describe them in general and
to give a limited amount of specific information.  Most of
the environmental factors, except for those associated with
mining, are most easily discussed by considering each
technology generically.
                            39

-------
                       Table 3

 Comparison of Pipeline Gasification Process Costs  [22]



                           Total Capital     20-year Average  2/
                           Investment,—      Gas Selling PriceT
Process                      $106               jrf / 106 Btu

CSG (CO2-Acceptor)              112                   73.5

BI-GAS                          179                   88.9

HYGAS (Electrothermal),          165                   89.3

Kellogg Molten Salt             167                   87.4

Lurgi                           297                  119.6
 Plant capacity 250 million cubic foot per day; 20-year
 plant life; 7.5 percent interest;  taxes and insurance at
 3% of total fixed investment

•j
 Coal price 25jzf per million Btu; 12 percent gross return
 on rate base; load factor 95 percent; by-product credit:
 sulfur $5 per ton, ammonia $20 per ton, phenol $80 per ton,
 char $0.15 per million Btu.

-------
      Probably  the major  environmental  abuse,  one  for which
we have not discovered a technological fix,  is  that due  to
strip mining for coal extraction.   If  we  increasingly  rely on
strip mining for more of our  coal production  and  do encounter
increased need for  coal  to  feed  coal liquefaction and
gasification plants, much improved  surface mining regulation
and enforcement and real innovation in mining and reclamation
technology will be  required.  Because  of  the  losses associated
with  the conversion techniques,  the land  disturbed by  mining
will  correspondingly increase unless offsetting improvements
are made in the power generation cycles  (such as  combined-
cycles) .

      Coal Liquefaction.  The land impacts  are  those associated
with  clearing  and preparing the  site for  the  plant, including
coal  storage,  handling and preparation facilities, ash and
sulfur disposal facilities, product storage facilities,  and
transportation facilities.  If the  plant  is located at the
mine-mouth, coal transportation  would  be  minimized but LCP's
would have to  be transported by  train  or  pipeline  to the power
plant.  For many LCP plants,siting  of  both the  plant and the
transportation corridor  may be a significant  problem.  Disposal
of the ash, and possibly of the  sulfur, may represent  another
serious problem area.  If the plant is located  at  the  mine-
mouth, then disposal in  the mined area might  be feasible,
otherwise the  solid wastes generated in a large LCP plant would
represent a massive landfill operation.   A plant processing
40,000 tons of coal per  day would generate 5,000 tons  of ash
(dry weight) -per day using Eastern  and Interior coals  and up
to 10,000 tons  per  day using Western coals .

     The air and water impacts of coal liquefaction plants
are not well characterized at this  point.  Most of the sulfur
removed from the coal is  converted  to  H2S and then converted
to elemental sulfur by various processes.  Light hydrocarbons
are generated  in the liquefaction processes and some of these
will be vented  to the atmosphere.   Emission standards  for these
and for H2S (not converted to sulfur)  will have to established
for these large sources.  A major variation in  the air
emissions may  depend on whether  the char  or other mineral
residue is fired for process heat or sold for combustion
elsewhere.  Particulates  and sulfur oxides from such combustion
will be difficult sources to  control.   Water clean-up does
not appear to  represent  a major problem,  at least  at this time.

-------
     From an environmental point of view, LCP plants do hold
considerable promise for shifting much of the environmental
burden from the dispersed power plants to a centralized
location.  This is quite attractive to the utilities.
Emissions to air and water may be more effectively controlled
at a larger central plant, however, this has not yet been
demonstrated.  The concentration factor will undoubtedly
exacerbate other impacts such as those on land.

     Coal Gasification. The environmental impacts of gasifi-
cation plants are better known, primarily because environ-
mental assessments have been completed for both commercial
high-Btu gasification plant proposals.  The Federal Power
Commission's Natural Gas Survey  [1] estimated the following
potential pollutants would be generated in a 250 million
cubic foot per day plant, assuming use of Interior coal with
3.7 percent sulfur and 10 percent ash:

                                           Tons per day

Sulfur (Mainly as hydrogen sulphide)           300-400

Ammonia                                        100-150

Phenols                                         10-70

Benzene                                         50-30

Oil and Tars                                 trace to 400

Ash (based on coal with 10% ash)                1500
     Control processes are known for many of these potential
pollutants and are being incorporated into the engineering
design of gasification plants.

     As an example of the type of control planned for a high-
Btu gasification plant, the sulfur balance diagram for the
proposed Transwestern gasification plant is shown in
Figure 7.  The plant will use 25,600 tons per day of low
sulfur coal for the gasifiers and process heat boilers.  The
sulfur input is 233 tons per day and only 7.5 tons are to be
emitted, over 60 percent of which are from the coal-fired
boiler.  The maximum 24-hour  average SOX concentration expected
is about 0.03 ppm.  Concentrations of NOX,  K^S, COS, and
particulates are a fraction of the allowable New Mexico
standards [11].

-------
-tr
U)
                                          S* 3.0 (SULFUa LEAU GA5)
                          ANALYSIS, WT%
                            MOISTURE    IZ.-4
                            ASH        25.6
                            FIXED CAR&Oti  33.6
                            VOL. H&TTEO.   28-2
                                      1OO.O
                                                                           =0.336
                                                                            L8S/UMBTU
                                                                                     SULFUR WT%
                                                                                     HHV,BTU/LB   63 IO
                                                                                     LHV, BTU/LB   7B6O
                                             5-0.0077
                                               LB5/MMBTU
                                             50^=0.0154-
                                                LSSfMMBTU
BOILES
 COAL

 123
 14.1
 40.Z
 334
 too.o

 0.87
987O
942O
                                                      0.28
                                                    H.49Q
                                                    16, 500
                                                                                      Total Sulfur Emissions
                                                                                         7.46 Tons/Day
           NOTE'
           ALL QUANTITIES ABE in SHOUT rows
              EXCEPT SHOWU OTHEBWISE.
                                           FIGURE 7.
Sulfur  Process  Diagram for  Transwestern
Coal  Gasification  Plant

-------
      Water impacts  of coal gasification plants  of two types:
 pollution potential and consumption.   For instance,  in the
 Transwestern  proposal,  all the  gas  liquor streams are
 treated  to remove the phenols,  ammonia,  and I^S and  all
 waste water streams treated to  remove  suspended oils and
 solids and residual organics.   Water consumption,  minimized
 in  the Transwestern proposal by recycling,  is 5100 gallons
 per minute (or  8,260 acre-ft per year)   Twenty  percent of
 this  is  consumed directly in the gasifier and 50 percent is
 evaporated in cooling towers and settling ponds.   Without
 recycle,  the  consumption could  have been 8,000  to 10,000
 gallons  per minute.  In water-scarce areas,  such water
 consumption for both PGG and AG systems may become an
 important factor.

                Forecasts  of  Application

      Forecasting application of emerging  technologies  is  a
 difficult matter. However,  because  of  their relevance to policy
 development for air pollution control, even  rough  forecasts
 are infinitely  better than none  at  all.   This section  will
 review the major features of several recent  forecasts  of  the
 application of  technological options for  clean  fuel.

 CEQ Forecast

     A forecast of  the  application  of all of the major
 technological alternatives for  the  use of high  sulfur  coals
 for steam-electric  plants has been  developed by the  Council
 on Environmental Quality  [18] .    This forecast include  four
 options:  staclc  gas  cleaning  (SGC),  present generation
 gasifiers  (PGG), advanced gasifier  (AG),  and coal
 liquefaction  (CL).  The predictions assume that  the  forecasts
 for SGC installations in  the "realistic"  scenario  of the
 SOCTAP report are valid  from 1974 to 1977.  In  the period
 1978 to 1980,  it is assumed  that only new coal-fired plants are
 equipped with SGC technology; beyond 1980 it is  assumed that
 no plants are outfitted with SGC systems.  These conservative
 assumptions would mean that  84,000  MWe, of which 61,000 MWe
 represent new plants, would have SGC devices by the  end of
 1980.  Limited  application of PGG processes would begin in
 1975 and application of AG system would begin in 1980.
Application of  CL plants would begin in 1979 and would expand
 rapidly.   The allocation of  these processes to  fuel  new or
 existing plants are predicated on many factors which have
been discussed  above.  The results  of these assumptions are
presented in Table  4 and are shown  graphically  in  Figure 8.
                           44

-------
                                                  TABLE  4.
                                      PROJECTIONS OF UTILIZATION OF CLEAN FUEL OPTIONS
                                          WITH LIMITED APPLICATION OF SGC SYSTEMS
Ul
Tech- Type of
nology Install,
new
SGC retro
new
PGG retro
new
AG retro
new
CL retro
Annual Capacity
Addition, KWe
Cumulative Capacity,
KWe
Coal Usable,
105 tons
Total Steam-Electric
Coal Use. 106tons

Percentage with Control
Coal-Fired Steam Electric Capacity with
74 75 76 77 78
900 3.800 8,200 12,000 12,000
1,600 3.600 5,600 12.000
1.000 1.000 1,000 1,000





2,500 8,400 14,800 25,000" 13,000

2,500 11.000 26.000 51,000 64,000

6.4 28 66 129 162

384

7.3%
Clean Fuel Technological
79 80 81
12,000 12,000



1.000 2,000
10,000
1,000 5,000 8,000

13,000 18,000 17.000

77,000 95,000 112.000

195 240 283

460

52%
Options.
82




4.000
12,000
8,000

24,000

136.000

344




MWe
83 84 85




8,000 12.000 12,000
10,000 10,000 10,000
8.UOO 10,000 10.00C

26,000 32,000 32,00n

162.000 194,000 226,000

410 491 572

613
93%


-------
      300 -
Elec.
       20O -
Capacity,
1OOO MWe
       100 -
                 FIGURE 8.  CEQ  Projections of Utilization
                            of Clean Fuel Alternative  Technologies
        YEARS

-------
     This forecast shows dramatically that the only viable
technological alternative for reducing sulfur oxide emissions
until the early 1980's is stack gas cleaning.  Even with rapid
expansion in LCP availability, the first priority for use of
LCP's will be for existing plants.  Competition from
industrial boilers also could limit availability of LCP's.
In this forecast, equal amounts of coal cleaning would be
achieved by SGC and CL in 1985, since AG and CL supply all
clean fuel needs after 1980.

Hittman Forecast

     The application of stack gas cleaning and solvent refined
coal for power plants has been forecast by Hittman
Associates [4].  Hittman emphasized only coal liquefaction
because,in their opinion, SRC is the most advanced clean
fuel option and has the greatest likelihood of success.
Hittman assumed three cases: Base, Higher SGC, and Sooner SRC.
In the Base case, it is assumed that commercial application
of SGC begins in 1976 and the maximum supply capability is
limited to installation on 20,000 MWe per year.  SRC is
commercially available in 1981 and is also supply limited for
the first decade.  The Higher SGC case differs from the
Base case only in that the maximum supply capability is
assumed to be 40,000 MWe per year.  The Sooner SRC case
differs from the Base case only in that SRC is commercially
available in 1979.  The projected application of SGC and
SRC technologies for these three cases is shown in Figure 9.
Some features of this forecasts are well-illustrated in
Figure 9.  SGC application is supply constrained until the
1983-1986 period; after then SGC application is demand
constrained.  SRC application is supply constrained and
since, from the utilities' point of view, is the desirable
approach to coal cleaning, SRC substitutes directly for
SGC to the extent possible.  Utilization of SGC continues,
in this forecast, not only after 1980 but even after the
1983-1986 period.  This application continues, although only
on new plants,  because Hittman feels that SGC will offer
the least expensive alternative in many cases.  The effect
of the technological alternatives in the three cases on the
reduction of sulfur oxide emissions is shown in Figure 10.

-------
    16 i—
    14
£H

O1
    12
Ho
J~
O '
    10
O
U
u
2
  w
fn Q
OW
                        BASE CASE - SGC
            HIGHER SGC-SGC
                                               SOONERSRC-SGC
                                    SOONER SRC-SRC
        1975
                                                   BASK C A SI-]- SRC
                                                   HIGHER SGC-SRC
                   1980
                       ,-^T'   '  -J—I	«—i
1885
  YEAR
                                             1
                                              1U90
                 -t.  i  t	L
1E95
2000
        FIGURE 9.  Hittman Projections  of Utilization of Clean Fuel
                   Alternative  Technologies

-------
     32
     28
u
<;
AH

                  I  .1.
                            J	L
                                 J	L
        FIGURE  10.   Reduction  in Sulfur Oxide  Emissions Associated

                     with Hittman Projections

-------
MITRE Forecast

     Application of SGC and SRC processes as well as synthetic
crude oil production and gasification were also projected by
MITRE [23].  While the projection given in Figure 11 covers total
coal utilization, the SGC and SRC components would probably
be used mostly by the electric utilities.  MITRE projects an
optimistic expansion of coal utilization which triples  over
the next 15 years.  Interestingly, MITRE assumes that 250
million tons of coal could be burned in 1980 with SGC which is
roughly 100,000 MWe.  The MITRE projection for SRC and
synthetic crude production accounts for 150 million tons of
coal which is much larger than that projected by any other source,

Chem Systems Projections

     Chem Systems has made predictions when the CL and PGG/AG
systems might become available [5 ].  Chem Systems believes that
CL processes will become available by 1978 with construction
and initial operation of both SRC and H-Coal plants.  Chem
Systems predicts that PGG plants will become commercial in
the U.S. by 1976-1977 and application will start before the
end of the decade.  AG power cycles will be 3-4 years behind
the PGG systems.  Chem Systems suggests the following
scenario for PGG system application:

          1976    Commonwealth Edison 3-gasifier system
                  successfully in operation.

          1978    Installation of gasifiers on two 300-500 MW
                  retrofit generators

          1980    Gasifier systems in operation on three new
                  1000 MW generators and on five additional
                  300-500 MW retrofit installations.

For this optimistic scenario, Chem Systems predicts a total
of 5000-6000 MW of generator capacity on low Btu coal gas by
1980.  Thus, low Btu gas could account for 2-3% of installed
coal-burning generator capacity by 1980, or more significantly
4-5% of installed coal-burning utility generators feeding
high sulfur Interior coals.  These predictions agree
reasonably well with the application of PGG and AG systems
forecast by CEQ.
                            50

-------
                  FIGURE 11.  MITRE Projection of U.S. Coal  Production  and  Utilization
            1800-1
             1600
             1400
 COAL
 PRODUCTION,
 TREATMENT,
 AMD USE
(MILLION TONS
PER YEAR)
                                                                        100  SOLVENT REFINED
                                                UNCONTROLLED
                                                HfGH SULFUR
             400-
             200-
                                                                                             1825  TOTA L U.S. COAL PR OO UCTI ON
                                                                                                 280  CONVERTED TO GAS
                                                                                                 330  CONVERTED TO OIL
325 HIGH SULFUR COAL BURNED WITH
   STACK GAS CLEANING
225  SOLVENT REFINED

ZERO HIGH SULFUR COAL BURNED
     WITHOUT POLLUTION CONTROLS
                                                                                                      NATURALLY
                                                                                                 66P  LOW SULFUR
                1970
                                          1975
                                                                   1980
                                                                                             1985
            [SOURCE: MITRE ANALYSIS]

-------
                       Summary
     In summary, this review of coal liquefaction, gasification,
and advanced combustion processes indicates that technological
alternatives to stack gas cleaning are just beginning to
enter the pilot plant stage.  Years of development remain
before any of these technologies will see widespread commercial
application.  Low-Btu gas production for utility boilers using
Lurgi gasifiers and solvent refining of coal appear to be the
first technologies that will be applied.  Two coal liquefaction
processes—H-Coal and COED—and high-Btu gas production appear
to lag somewhat behind the PGG and SRC technologies.  Many other
CL and AG processes with promising characteristics are only at
the bench-top experimental stage.

     All of the processes are expensive and it appears that
electric power production costs will increase significantly
as these new technologies are introduced.  Although it is
nearly impossible to generalize, it appears that stack gas
cleaning will be significantly less expensive than the
technological alternatives for many new and at least some existing
power plants.  The relative economics of the various
technologies, of course, must be evaluated for the conditions
of a given plant.  Among the various alternatives, SRC appears
to provide the least expensive fuel which should meet air
quality standards in most areas. SRC may also enjoy economics
of scale not available to PGG, AG and some CL processes.  If,
and there are a number of if's, AG combined-cycle systems can
be successfully developed, they may compete with CL processes.

     Environmentally, CL processes take the burden off the
utilities to control air emissions and manage solid wastes
and shift it to the processing plant.  But the same ash and
sulfur has to be disposed of at the processing plant.  PGG
and AG processes remove the sulfur from the synthetic gas
stream and may, if hot clean-up processes are developed, do this
without loss of thermal efficiency.  But disposal of the ash
and sulfur remains at the power plant.  Air and water emissions
from CL, PGG, and AG processes are not well characterized yet but
it appears that some new problems may arise through the
chemical processing of coal.  Developmental work on advanced
clean-up processes remains underway.
                           52

-------
     Concluding with the forecasts of application, a common
element in all the forecasts is the 5 to 8 year lag between
the commercial availability of SGC and that of the alternative
clean fuel technologies.  To the degree that sulfur oxide
emissions must be controlled to meet primary air quality
standards, it is quite clear that SGC is now and will continue
to be for some years the only viable alternative to fuel
switching.  A delay in the application of SGC to plants
requiring control until another technological alternative is
available places too much burden on the air quality and too
much reliance on yet-unproved technologies which, if
successfully developed, many cost significantly more.
                           53

-------
                     References
 1.   Department of Interior, "Environmental Statement
      for the Proposed Prototype Oil Shale Leasing Program,"
      Volume II, unpublished  [March, 1973].

 2.   Hottel, H.C., and J.B. Howard, Kew Energy Teclmolog-y—
      Some Facts and Assessments, MIT Press, Cambridge  [1971].

 3.   Mills, G.N., and H. Perry, "Fossil Fuel	> Power +
      Pollution,"  Chemtech 53-63 [Jan. 1973].

 4.   Hittman Associates, Inc.,  "Assessment of S02 Control
      Alternatives and Implementation Patterns For the
      Electric Utility Industry," Contract for Office of
      Science and Technology  [March 1973].

 5.   Chem Systems, Inc., "Desulfurization Strategy Options
      Including Fossil Fuel Supply Alternatives For the
      Electric Power Industry,"  Contract for Office of Science
      and Technology [March 1973].

 6.   "Big SRC Pilot Unit Under Construction." Oil and Gas
      Journal [Nov. 6,  1972].

 7.   Akhtar, S., S. Friedman, and P.M. Yavorsky, "Process for
      Hydrodesulfurization of Coal in a Turbulent Flow Fixed-
      Bed Reactor," American Institute of Chemical Engineers
      national Meeting, Dallas [Feb. 20-23, 1972],

 8.   Yavorsky,  P.M.,  S. Akhtar, and S. Friedman, "Process
      Developments: Fixed-Bed Catalysis of Coal to Fuel Oil,"
      American Institute of Chemical Engineers Annual
      Meeting, New York [Nov. 26-30, 1972].

 9.   Mills, G.  A., "Gas from Coal:  Fuel of the Future,"
      Environmental Science and  Technology 5:12,  1178-1183
      [Dec.  1971].

10.   Squires, A.M., "Capturing  Sulfur During Combustion,"
      Technology Review, 52-59 [Dec. 1971].

11.   Transwestern Coal Gasification Company et al,  "Detailed
      Environmental Analysis Concerning a Proposed Coal
      Gasification Plant and the Expansion of a Strip Mine
      Operation Near Burnham, New Mexico," submitted to Federal
      Power  Commission [Feb. 1,  19731.
                            54

-------
12.   Agosta, J.,  et al.,  "Status of Low Btu Gas as a
      Strategy For Power Station Emission Control,"
      American Institute of Chemical Engineers Annual
      Meeting, New York [Nov.1972].

13.   Matthews, C.W., "A Design Basis for Utility Gas
      from Coal,"   Third International Conference on
      Fluidized Bed Combustion, Hueston Woods State Park,
      Ohio [Oct. 30-Nov.l, 1972].

14.   El Paso Natural Gas  Company,  "Report on Environmental
      Factors Burham Coal  Gasification Project," submitted
      to Federal Power Commission [Nov.1972].

15.   Feldman, H.F., J.A.  Mima, and P. M. Yavorsky, "Pressurized
      Hydrogasification of Raw Coal in a Dilute-Phase
      Reactor," American Chemical Society, Dallas
      [April 8-13, 1973].

16.   Cover,  A.E., W.C. Schreiner,  and G. T.  Skaperdas,
      "The Kellogg Coal Gasification Process: Single Vessel
      Operation,"  Synthetic Gas Production Symposium,
      New York [Nov. 1972].

17.   Bituminous Coal Research, Inc., "Economics of
      Generating Clean Fuel Gas from Coal Using an Air-
      Blown 2-Stage Gasifier,"  Office of Coal Research
      R&D Report No. 20 [Aug. 5, 1971].

18.   Council on Environmental Quality, "Technological
      Alternatives for Using Domestic Coal Resources,"
      Appendix A,  The Potential for Energy Conservation;
      Substitution for Scarce Fuels,  Office of Emergency
      Preparedness [Jan. 1973] .
                            55

-------
19.  American Lurgi Corporation, "Clean Fuel From Gas,"
     Lurgi Quick Information, New York, Publication
     No. 01007/10.7.

20.  Matthews, C.W., "Coal Gasification for Electrical
     Power,"  Institute of Gas Technology, presented at
     American Power Conference, Chicago [April 18-20, 1972]

21.  Advanced Technology Corporation,  "SO2 Free 2-Stage
     Coal Combustion Process,"  prepared for Environmental
     Protection Agency [Aug. 1972].

22.  Mehta, D.C., and B.L. Crynes,  "How Coal Gasification
     Common Base Costs Compare," Oil and Gas Journal
     [Feb. 5, 1973].
                            56

-------
STATUS  OF FLUE GAS DESULFURIZATION TECHNOLOGY
                          by
              R. E.  Harrington, Director
             Air Pollution Control Division
          Office of Research  and Development
           Environmental Protection Agency
                 Washington, D.C.
                            57

-------
              THE STATUS OF DESULFURIZATION TECHNOLOGY
     Implicit in our efforts to assess the status of sulfur oxide
control technology is the assumption that there is either a well-defined
measure of the problem that the technology is designed to address or
that there is a well-defined market for the technology, or both.  This
paper deals largely with the factors that must be considered in the
development of a yardstick for measuring the status of sulfur oxide
control technology. ,
     In this conference, several papers will be given which will
specifically address the technical, economic and regulatory factors
relating to specific S02 control processes.  It is the purpose of
this paper to help provide a yardstick against which you can better
judge for yourself the status of each process and the overall status
of the technology.
     To help put the sulfur oxide problem in perspective in the United
States, let us first compare the mass emissions of the five principal
pollutants.  This is done in Figure 1.   Using 1970 mass emission data,
the pie-chart on the left shows the total  mass emission of the five
major ambient air quality pollutants:  SOX, NOX, Particulate, Hydro-
carbons, and Carbon Monoxide.  For comparison, the pie-chart on the
right replots the 1970 mass emission data but excludes emissions from
transportation sources.   The chart on the right, therefore, shows the
distribution of these five pollutants which are emitted from stationary
and some natural  sources.  Comparing the mass emissions,  we see that
stationary sources in the United States account for most of the sulfur
                                  58

-------
 oxide  and  particulate emissions.  They account for roughly half of the
 nitrogen oxide and  hydrocarbons and about 1/5 of the carbon monoxide.
     Mass  emissions, however, do not provide an adequate index of the
 impact of  a source  on the ambient air quality.  Other major factors
 influencing the  impact of pollutants on ambient air quality include:
     1.  The geographical point of discharge
     2.  The height of the source above the receptors
     3.  Topography surrounding the discharge point
     4.  Proximity  of source to other sources
     5.  Other factors affecting distribution such as
         a.  Wind Direction
         b.  Wind Speed
         c.  Ambient Temperature
     To provide a measure of the relative impact of sources on ambient
 air quality, we have developed and applied a dispersion model  which
 takes  into account most of the factors mentioned above.  This  model
 assumes a rectangular dispersion of pollutants downstream from the
 source and considers the relative location of sources,  height  of dis-
 charge, wind direction and velocity.   Using the model,  it is possible
 to predict the ambient exposure of a  receptor in the impacted  area.
     Using this model,  we have calculated the average impact of various
 kinds of emitting sources.  Figure 2  shows the relationship between the
mass emissions and their impact on ambient air quality.  In this figure
we have classified sources into four  major categories:
                                  59

-------
     PP--Power Plants
     IC--Industrial Combustion
     IP--Industrial Processes
     AS-Area Sources
     Measured in terms of tons of pollutants emitted, the fossil fuel
fired power plant accounts for over 50 percent of S02 emissions,
industrial combustion accounts for 24 percent of the S02 emissions,
10 percent are from industrial processes (such as sulfuric acid plants,
pulp and paper plants, refineries and smelters) and 12 percent from
area sources (residential and commercial).
     When the factors affecting ambient air quality other than mass
emissions are applied (location, height, proximity to other sources,
etc.), the impact of the sources are as reflected in the bar-graph on
the right of Figure 2.  Industrial combustion is seen to be by far the
most important single source accounting for about 40 percent of the
ambient air quality impact.  Power plants and area sources each account
for about 25 percent of the ambient air quality impact while industrial
processes continues to be the smallest impactor accounting for only
about 14 percent.
     The data depicted in Figure 2 are a composite of 8 U. S. industrial
cities.   The data represent calculated annual averages which have been
verified against measured annual averages.   The verification or corre-
lation with measured annual averages is quite good.
                         Combustion Sources
     Since combustion sources represent the principal  source of sulfur
dioxide impacting ambient air quality, it is useful  to look more
                                  60

-------
closely at this problem.  Industrial combustion sources are characterized
by being a large number of individual sources.   They span a relatively
large size range and relatively large range of different types of
combustion from small gas burning systems to large residual oil and
coal burning units.  In general, they are further characterized as
being too small to permit the economic use of desulfurization processes
developed for larger power plants.
     One obvious possible solution to this segment of the problem is to
use low sulfur fuel thereby obviating the sulfur oxide emissions.
     Figure 3 shows the relationship between the demand for fossil  fuels
by the utility industry in the United States over the next 16 years.
Also shown is the source of supply of clean fuels supplemented by flue
gas cleaning control technology required to achieve Ambient Air Quality
Standards.
     The principal observations that can be made from this figure are
as follows:
     1.  Naturally occurring low sulfur fuel located in areas where
         it can be reasonably and economically  used is inadequate.
     2.  Cleaned fuels will  not make a significant impact on the
         supply pattern until the latter half of this decade.
     If these projections are correct, they lead to several important
conclusions.  First, clean fuels is not the answer—at least not for
the next 10 years.  Second,  the shifting of the limited supply of low
sulfur fuels from utilities  to industrial  combustion will  help in
certain cases, but it is not likely to be a broadly applicable solution.
                                  61

-------
A key factor limiting the fuel switching approach will be the present
ownership and the existence of extant contract arrangements surrounding
these low sulfur fuels.  Much of the low sulfur coals are captive to
industrial processes and will probably not be dedicated to electric
power generation.  Many utilities have long-term contracts for low
sulfur fuel and will understandably not be willing to give up their
solution to the air pollution problem to make their low sulfur fuels
available to other fuel users.  Third, flue gas cleaning will play a
principal role in the near future control of sulfur oxides.  Flue gas
cleaning will undoubtedly be the work-horse of the industry for at
least the next ten years.
                   What Level of Control is Needed
     So far, we have considered the average pollution problem—the
average source category and the average solution.  By considering only
the average pollution levels in industrial  cities, we can conclude that
average reductions in pollutants of 10, 30, or 50 percent can result in
meeting Ambient Air Quality Standards.
     How meaningful, however, are the averages?  For the individual
source that must control its emissions to meet regulations, is it
possible to control to the average level of reduction needed by the
geographical area in which the source is located and meet his emission
control obligations?  Let's consider this question further.
                                  62

-------
      In order  to get an estimate of the level of control that a given
 individual source would have to achieve, the dispersion model described
 above was again exercised.  The objective was to estimate the control
 capability that a specific control system would have to be able to
 achieve when installed on a single source to permit that source to
 meet applicable regulations.  This, after all, is the real  question
 that must be answered by the regulatory agency, the polluter and the
 research and development community charged with developing control
 technology to meet the regulatory goals.
     Three cities, New York, Philadelphia, and the Niagara Frontier
 were selected largely because of the availability of data.   In each
 city, the ten most seriously impacted receptor points based on ambient
 air quality annual  averages were used.
     For simplicity, the assumption was made that the contributions of
 electric power generating plants would be eliminated (that is 100
 percent control of power plants could be achieved—this permits the most
conservative control  estimate for other sources).   The level  of control
 that would be required  by other impacting sources to achieve ambient
air quality standards were then calculated.   In each case,  it was found
 that greater than 90 percent control  would be required for  many of •
principal  contributing  sources.  The  calculations did not take into
consideration further industrial  growth in the areas studied.
     While it might be  argued that these three cities are not repre-
sentative of all  U.  S.  industrial  cities,  it seems  almost certain that
similar situations  will  be found in most other industrial cities
                                 63

-------
in which the impact of one or more sources will be such as to contribute
to "hot spots" within that area.
     The results of this study serve to establish a reliable estimate
of the control capability required of control processes and equipment
to deal with specific source problems.  As a limit, flue gas desulfuri-
zation processes and other methods of control must be capable of at
least 90 percent removal.
     This is not to say that all processes must be capable of 90 percent
S02 control.  In many cases, lower levels of control  will be adequate to
meet standards.  In marginal cases, processes such as the dry limestone
injection process which appears to be capable of only 25 percent SO^
control may be adequate to permit compliance.  Each case must be evalu-
ated in terms of its own unique needs and economics.
                               Summary
     What js^ the status of sulfur oxide control technology?  Many factors
enter into the answer.  This paper deals with a few of these factors
that relate to the current definition of the problem and the current
perspective of the need for flue gas desulfurization.  In summary,
these are as follows:
     1.  Low sulfur fuel will be a major tool in achieving ambient
         air quality standards over the next few years.
     2.  Low sulfur fuel, however, falls far short of being a
         sufficient solution for SO  control.

-------
3.  Fuel substitution can and will  be used to a limited extent.
    Major problems exist of a proprietary, legal  and logistical
    nature that will preclude our approaching the full  theoretical
    potential  of fuel substitution.
4.  Fuel cleaning technology is at least ten years away from
    making a major impact on the increased availability of
    clean fuel.
5.  Flue gas cleaning technology is the principal tool  available
    to the Nation for SOz control for at least the next 10 years.
6.  Industrial  combustion is the No.  1 problem source of S02 in
    the United  States needing solution.
7.  Considerable effort must be made  to extend the application
    of processes which have been designed for utility application
    to industrial combustion control.  In addition, increased
    attention must be directed toward developing  processes uniquely
    applicable  to small industrial  combustion, industrial  processes
    and area sources.
                             65

-------
             TOTAL EMISSIONS,
               267 X 106 tons
                  PARTI CU LATE

      23X70*10*I26*}*
       STATIONARY SOURCE EMISSIONS,
               122 X 106 tons
                 PARTICIPATE
                  25 X 106 tons
                      20%
Figure 1. Emission relationships, 1970.

                  66

-------
POWER
PLANTS, 53%
INDUSTRIAL
COMBUSTION, 24%
INDUSTRIAL
PROCESSES, 10%
AREA
SOURCES, 12%
TOTAL MASS
EMISSIONS
                                                               POWER
                                                               PLANTS, 22%
                                                               INDUSTRIAL
                                                               COMBUSTION, 40%
INDUSTRIAL
PROCESSES, 14%
AREA
SOURCES, 24%
DISTRIBUTION IN
AMBIENT AIR
          Figure 2.  Distribution of SOX  emissions in ambient air.
                                    67

-------
                                                  FUEL FOR
                                                  FLUE-GAS
                                                  CLEANING
                                        NATURAL
                                       LOW-SULFUR
                                          FUEL
                                                                                        cc.
                                                                                        UJ
                                                                                1990
Figure 3.  Power demand and utility fossil fuels required to achieve ambient air quality standards.

                                             68

-------
STATUS OF JAPANESE  FLUE GAS  DESULFURIZATION TECHNOLOGY
                              by
                         Jumpei Ando
               Faculty of  Science and Engineering
                       Chuo University
               Kasuga, Bunkyo-ku, Tokyo,  Japan
                               69

-------
              Status of Japanese Flue Gas Desulfurization Technology
                                    Jumpei Ando

Per-capita  energy consumption in Japan is now about one-third that in
the U.S.A.  but consumption per acre of level land is eight times the
U.S. figure resulting in serious environmental problems.  Stringent
regulation  of SO,, emissions has prompted the development of various
desulfurization processes.  More than 100 commercial and prototype
plants for  waste-gas desulfurization are on stream now.  The present
paper will  chiefly describe seven of the processes which seem to be of
more interest for possible application in the U.S.A.
                    1   Outline of SO- control activity

Eecently, about 70$ of energy supply in Japan has depended upon imported
crude oil.  Heavy oil, which is a residue from atmospheric distillation
of crude and contains 2-4$ sulfur, has been a major fuel for power
generation as well as for other industrial activities.  Hydrodesulfurization
of heavy oil has been carried out since 196?•  In 1971i nearly 200 million
barrels of heavy oil, about one-fourth of the total quantity of heavy oil
consumed in Japan, was subjected to hydrodesulfurization giving 287,000 tons
of elemental sulfur as by-product.  Still nearly 3 nillion tons of sulfur
in heavy oil burned produced nearly 6 million tons of S0?, constituting
the chief source of SO,, emissions.  Today's stringent controls on SO,,
emissions have renderea the desulfurized heavy oil, with its 1$ or higher
sulfur content, unsatisfactory for large power plants.

Under such situation, flue-gas desulfurization has assumed greater
importance, which fact has led to construction of more than 100 commercial
and prototype plants.  Most of the plants built so far are of relatively
small capacity, designed to treat waste gas from industrial boilers,
Chemical and smelting plants, etc.  Some of the larger ones are listed
in Table 1.  Major plants under construction or being designed are listed
in Table 2.  Nine maj*or power companies, supplying more than JCffi of the
total electric power needs in Japan, were originally interested in dry
processes but have recently decided to build many plants using wet processes
because wet processes with reheating of the treated gas have proved to
"be less expensive than dry processes.  With some of the wet processes,
moreover, trouble-free continuous operation has been demonstrated and
wastewater has been eliminated.  The total capacity of the desulfurization
plants of the major power companies will increase from 375MW in 1972 to
2,700MW in 1974 and to 4,800MW in 1976.

Most of the desulfurization plants have produced salable by-products such
as sodium sulfite, sulfuric acid, and gypsum, because Japan is subject
to limitations in domestic supply of sulfur and its compounds as well as
in land space available for disposal of useless by-products.  As the
desulfurization projects are making very rapid progress, however, it is
likely that a considerable oversupply will occur in future necessitating
abandonment of a substantial portion of the by-products.

                                  70

-------
                     Table  1    Major waste-gas desulfurization plants in operation

Process developer
Kureha Chemical
Showa Denko
Wellman-MKK
Vellman-SCEC
Mitsubishi-JECCO
Mitsubishi-JECCO
Chiyoda
Chiyoda
Nippon Kokan
Onahama Smelting
Chemico-Mitsui
Sumitomo S.M.
Mitsubishi H.I.
Hitachi Ltd.

Absorbent
NaOH
NaOH
NaOH
NaOH
Ca(OH)2
Ca(OH)2
H-SO., CaCO.-
H-SO,, CaCO-j
24 3
NH , Ca(OH)2
MgO
Ca(OH)2
Carbon
MnOx, HE,
Carbon, CaCO,

Product
Na2SO
Na2SO
H2S04
so2, s
GypBum
Gypsum
Gypsum
Gypsum
Gypsum
H S04
CaSO.,
H2S°4
I "MTT A O/~\
V liil t I f^ &\J .
Gypsum

User
Kureha Chemical
Ajinomoto
Japan Synth. Rub.
Toa Nenryo
Kansai Electric
Onahama Smelting
Fuji Kosan
Mitsubishi Rayon
Nippon Kokan
Onahama Smelting
Mitsui Aluminum
Kansai Electric
Chubu Electric
Tokyo Electric

Plant site
Nishiki
Kawasaki
Chiba
Negishi
Amagasaki
Onahama
Hainan
Otake
Keihin
Onahama
Omuta
Sakai
Yokkaichi
Kashima
Unit capacity
(l.OOOscfm)
176a, 176a
159a
118a
Tj
35
59a
54°
93a
53a
.a
88d
53°
226e
iooa
193a
250a
Date of
completion
1968
1971
1971
1971
1972
1972
1972
1973
1972
1972
1972
1971
1972
1972
a:" Oil-fired boiler



d:  Sintering plant
b:  Glaus furnace      01   Smelting furnace



e:  Coal-fired boiler

-------
Table  2
Process developer
Vellman-MKK
Wellmari-MKX
Wellman-SCEC
Showa Denko
Shell
Mitsubishi-JECCO
Mitsubishi-JECCO
Mitsubishi-JECCO
Kureha-Kavasaki
Rureha-Kawasaki
Babcock-Hitachl
Chiyoda
Chiyoda
  Absorbent
    NaOH
    NaOH
    HaOH
NaOH, CaCO,
    CuO
  Ca(OH)2
  Ca(OH)2
  CaCO,
NaOH, CaC03
NaOH, CaCO'
  CaCO
          ,
      3
     , CaCO,
                   Major flue-gas desulfurization plants under construction
                   or being designed (oil-fired boilers)
 Product
 H2S04
 Gypsum
so2, s
 Gypsum
 Gypsum
 Gypsum
 Gypsum
 Gypsum
 Gypsum
 Gypsum
 Gypsum
                                     User
                        Chubu Electric
                        Japan Synth. Rub.
                        Sumitomo C.C.
                        Showa Denko
                        Showa Y.S.
                        Tohoku Electric
                        Kansai Electric
                        Tokyo Electric
                        Shikoku Electric
                        Tohoku Electric
                        Chugoku Electric
                        Eokuriku Electric
                        Mitsubishi  P.O.
Plant site
Nishinagoya
Yokkaichi
Chiba
Chiba
Yokkaichi
Hachinoe
Kainan
Yokosuka
Shintokufchiffl
Sendai
Mizushima
Shinminato
Yokkaichi
Capacity
(MW)
220
160
125
200
42
125
150, 130
150
a 150
130
100
250
240
Date of
completion
1973
1974
1973
1973
1973
1974
1974
1974
1974
1974
1974
1974
1974

-------
        2   MItsubishi-JECCO lime  (limestone)-gypsum process1*2)


 Developer  Mitsubishi Heavy Industries
            5-1t 2-chome, Marunouchi, Chiyoda-ku, Tokyo

            Japan  Engineering Consulting Co.
            1-4» Ogawamachi, Kanda,  Chiyoda-ku, Tokyo

 Rrocess  description Waste gas is first washed with water for dust removal
 and cooling to about 140°]?.  As the water becomes acidic and dissolves
 metallic components of dust, it is  neutralized with milk of lime to
 precipitate metallic ions, which are filtered off together with the dust.
 The filtrate is used for slaking of lime.  The cooled gas is then sent to
 an absorbing step.  At three plants built recently two plastic-grid packed
 absorbers in series, which are put  together in one tower, are used as
 shown  in Figure  1.                (For new -plants which are being
 designed a one-absorber system will be used.}  Milk of lime is fed to the
 No.  2  absorber.   The gas is introduced into the tfo. 1 absorber and then into
 the ITo.  2 absorber.  The slurry discharged from the No. 2 absorber is
 a  mixture of calcium sulfite and unreacted lime with a small amount of
 gypsum.   The slurry is then led to  the Ho. 1 absorber, where the remaining
 lime is  reacted to  fom calcium sulfite, a portion of the sulfite is
 converted  to bisulfite.  The pH of the slurry discharged from the Ho. 1
 absorber is 4-4»5-  The concentration of the slurries in the absorbers
 Is about 15$.  A  relatively large liquid/gas ratio (20-50 gal/1,OOOscf)
 is used  to prevent  scaling.

 The  pH of the slurry is then adjusted to about 4 to promote oxidation in
 the  following step.  If required, a small amount of sulfuric acid,
 normally less than  one ton per 100  tons of inlet SO-, is added to the
 slurry for the adjustment.  The slurry is then sent to an oxidizing tower
 where  the  sulfite and bisulfite are converted to  gypsum by air oxidation
 using  rotary atomizers invented by  Japan Engineering Consulting Co.(jECCO)
 at a pressure of  50-57psig  and a temperature of 120-180°F.  The atomizer
 is quite  effective  in producing fine bubbles and is free from scaling,
 erosion and corrosion.  The gas leaving the oxidizer contains some SO-,
 and  is returned to  the absorber.  The gypsum is centrifuged.  All of
 the  liquor and wash water are used  for the gas washing and cooling step.
The gypsum grows  into large crystals; its moisture content after ce'ntrifu-
gation is  only 8-10$.  The gypsum thus obtained is of high purity and good
 quality, which make it suitable for use in cement and gypsum board.  The
gas from the No.  2 absorber is passed through a demister, reheated, and
led to a stack.  Wash water of the mist eliminator is also used in the
 system.  Normally no wastewater is  emitted from the system.  More than
    of SOg is recovered.
                                  73

-------
               of d'-velc^-ient  Four  lants are in operation, and five others
TJser
Nippon Kokan
Kansai Electric
Onahama Smelting
Tomakomai Chemical
Kawasaki Steel
Tokyo Electric
Kansai Electric
Tohoku Electric
Kansai Electric
PI -.r.t site
Koyasu
Amagasaki
Onahama
Tomakomai
Chiba
Yokosuka
Hainan
Hachinoe
Hainan
Capacity
fscfn)
37,000
59,000
54,000
35,000
71,000
235,000
235,000
224,000
221,000
Number of
A.b s orb ont ab z o rb o r z
Ca(OH)2
Ca(OH)2
Ca(OE)2
Ca(OH)2
Ca(OH)2
CaCO
Ca(OE)2
Ca(OH)2
Ca(OH)2
2
2
2
2
1
1
1
1
1
         are being constructed or designed as shown ir. the following table:

                                                                    Year of
                                                                    completion

                                                                     1964
                                                                     1972

                                                                     1972

                                                                     1972

                                                                     1973

                                                                     1974

                                                                     1974

                                                                     1974

                                                                     1974


Status _of techr.olcp'  Based on extensive studies with a pilot plant,
Mitsubishi has succeeded in scale prevention.   Scaling can be prevented
by the use of a suitable material, shape, and arrangement of the grid
in the absorber, by the adjustment of the slurry concentration and pH
as well as of the liquid/gas ratio, by the addition of gypsum crystal
seed and thorough mixing of lime and the circulating slurry.

The Amagasaki plant has been in continuous operation since its start in
April 1972 except for the period of shutdown of the power plant.  The
desulfurization plant treats a fraction of flue gas 86,000scfm from a
1JJ6MW boiler'containing about 700ppn S0? to recover about JOJj of the SO-.
The gas velocity in the absorber is about 11 feet/sec.  The pressure
drop in the whole system including the cooler, absorbers and demister is
6 in.HpO.  More than "95/a of calcium sulfite is oxidized into gypsum in
the absorbers due to the low S0« concentration; the oxidizing tower is
almost unnecessary.  The amount of water added to the system is maintained
equal to that removed from the 'system by evaporation in the cooler, by
hydration of gypsum, etc.  Ho water is wasted from the plant.

The Onahama plant treats 54,000scfm of gas from a copper smelter containing
20,000-25,000ppm S02.  More than 99.5$ of the SO- is recovered with a
stoichiometric amount of lime by feeding milk of lime mainly to the No. 2
and partly to the No. 1 absorber.  The SO^ content of the outlet gas is
less than 50ppm.  The plant came on-stream at the end of October 1972
and has been in continuous operation without trouble except for a period
of scheduled shutdown for inspection at the end of November,  ifo scaling
was observed at the inspection.  The gas supplied from tjtie smelter is a

-------
 wet  gas  at  155°F  and results in less evaporation of water in the system.
 Therefore,  the  amount of water fed to the desulfurizaticn plant slightly
 exceeds  that by evaporation, hydration, etc.  A small amount of water
 is wasted after being treated for pollution control.  About 450 tons/day
 gypsum is produced; three oxidizing towers are provided for the oxidation
 of calcium  sulfite  into gypsum; little oxidation occurs in the absorbers
 due  to the  high concentration of SO-.

 A one-absorber  system will be used for the new plants to save invest-
 ment cost.  To  ensure high SO,, recovery    excessive amounts of the
 absorbents, about  105;' of stoichiometric for line and about llO^j for
 limestone,  will be  used.  For pH adjustment prior to oxidation, a
 considerable amount of sulfuric acid will be required to convert the
 excessive absorbent to  gypsum.  Other facilities and treatments are the
 same as  in  the  two-absorber system.

 In the limestone  scrubbing plant in Yokosuka, seawater will be used for
 cooling  flue gas  from an oil-fired boiler.  The seawater is gradually
 concentrated in the gas-cooler and therefore should be wasted after
 "being duly  treated.  Salable gypsum of high purity and good quality will
 "be by-produced  using limestone as absorbent.

 Economics   Investment cost for the Amagasaki plant (59»OQOscfm) was
 $1.46 million including various equipment for automatic control and for
 tests, while that for Tomakomai Chemical (31,OOOscfm) was $0.32 million.
 The  cost    for    larger plants (224,000-235,OOOscfm) is estimated to be
 $2.6-2.9 million  in battery limits.  The desulfurization cost for a
 plant to treat  100,000-150,OOOscfm gas from an oil-fired boiler is
 estimated at 50.67-0.88/bl of oil containing 2.5-3.0^ sulfur, including
 depreciation and  credit on gypsum at $6/t.

 Advantages  High recovery of SO- is attained and good-quality gypsum is
 obtained using  either lime or limestone without scaling problems.  The
 jfotary atomizer is  quite effective for oxidation,  involving no operational
prgblem.  No water  is wasted when gas at temperatures above 250°F is
 treated.  Seawater  can be used for cooling,  although in this case the
used seawater should be discharged.

Disadvantage  Lime is more expensive than limestone.  Although limestone
 is used for the absorbent, an appreciable amount of sulfuric acid is
required to produce gypsum of high purity and good quality.   Oversupply
 of eypsum might occur within several years if too  many gypsum-producing
desulfurization plants are built.
                                       75

-------
      Washing
        and
      cooling
                  Cooling water
                          Lime
0>
Stack  Cooling
  _     water
                                Demlster Blower I
                                      --Q-Ll
                                                  pK adjusting
                                                              <-l| JlCentrlfuge
                                   tentrl-
                                ,,  *
                                 Soot
                                      Lime

                                    Milk of lime
                                                                            NeuSrallzer
                                                                            L
                                    Thick-
                                      ener
                                                        Liquor tnak


            Figure 1  Flow sheet of Mitsubishi-JECCO lime-gypsum process
                             Reed
                            crystal
                            tank

-------
                                                          125)
                             3   Chemico-Mitsui lime process  '  '

 Developer  Chemico, U.S.A.

            Mitsui Aluminum Co. (Miike Power Station),  Orauta

 Constructor  Mitsui Miike Machinery Co. Ltd.
              1-1, 2-chome, Muromachi,  Kihonbashi,  Chuo-ku, Tokyo


 Process description  Chemico scrubbers (two-stage venturi) are used
 for SC>2 and dust removal from flue gas from coal-fired boiler (156MW).
 Carbide sludge (primary calcium sulfite) is used as absorbent.
 Figure 2 presents a description of the total scrubber system.  The
 flue gas  (302,OOOscfm)  after passing  through an electrostatic precipitator
 contains  0.3  grain/scf  of  dust  and 1,300 to  2,200ppni of S02 at 300°?.
 About 75/S of  the  gas is handled by the  scrubber.  Two  scruobers were
 installed but one of them  has been used with the other as a back-up.  The
 gas flows down through  the first venturi section, up through the mist
 eliminator section,  then passes through the  second venturi  and mist eliminates
 sections,  is  reheated and  exhausted to  a stack along with the unscrubbed
 fraction  of the gas.  Milk of lime is mixed  with the discharge from the
 second venturi; the  mixed  slurry is partly recycled to the  second venturi
 and partly fed  into  a delay tank.  The  slurry in the delay  tank is sent
 to  the first-stage venturi.  The discharge from the first venturi,
 consisting mainly of calcium sulfite  with small amounts of  calcium sulfate,
 unreacted calcium hydroxide and-fly ash, is  sent partly to  a delay tank
 and partly to a disposal pond about a nile from the plant.  The decanting
or settling of the solids takes  place  there and the supernatant from the
 pond is recycled  to  the scrubbing system to  prepare milk of lime and also
 to  wash mist  eliminators.   The  outlet gas frcia the scrubber contains
 Otl grain/scf of  dust and  200 to 300ppm of SO-.

 Status of technology The  scrubber is 33 feet in diameter and 66 feet high,
 is  constructed  of stainless  steel, and  is lined with glass  flake reinforced
 polyester material.  The following operation conditions have been'used:
       L/G (venturi + spray) 1st stage, 46 to 59 gal/1,OOOscf

                             2nd stage, 42 to 55 gal/1,OOOscf

       Stoichiometry        100 to 105$ aa  pure Ca(OH)2

       Percentage solids in slurry    3-5$

       Total pressure drop            16 in.E-O

Prior  to  the completion of  the plant, extensive pilot plant tests were
 carried out by Mitsui Aluminum  Co. leading to the establishment of
 operation know-how for  scale prevention.  Precise control of pH to a
 certain narrow range is important for the scale prevention.

 State  of  development  Operation of the  commercial plant started in March
 1972.  After 8 months of satisfactory continuous operation, the plant
vas subjected to  a scheduled shutdown for inspection, wliich revealed
 essentially no scaling.  Operation was resumed soon and has since been

                                       77

-------
carried out smoothly.  The waste disposal pond has a capacity for
holding solids discharged over a period of several years.  However, to
eliminate any possible pollution with eulfite ion in wastewater which could
"be emitted when the facilities are cleaned in a periodical -shutdown of
the plant for inspection, some means of converting calcium sulfite to gypsum
will be adopted within 1973*  Tests have shown that the gypsum can be used
for gypsum board and cement   although it contains a snail amount of fly
ash.

Economics  The commercial plant cost 33«3 million including two scrubbers.
The desulfurization cost is a litule less than 1 mil/Vah.  About 70j£ of
the desulfurization cost is accounted fox by depreciation and interest.

Advantages  Stable operation without scale formation is achieved, removing
both SO- and dust satisfactorily.  Carbide sludge, a cheap source of line,
is used.  The Chemico scrubber is suited for treating a large amount of-
gas.

Disadvantage  Large amounts of slurry and water must be recycled because
of the use of high L/G and low slurry concentration in order to ensure
scale prevention.
                                  78

-------
 Boiler
 flue
VD
           Stack
                 --CD
          Wet carbide pit
N 	 j 	
«D« Fan

n










•*-



A
Reheat D
furnace

eoiste
r
f
                            Makeup
                          i  slurry
Makeup slurry
feed pump
                            let stage
                            recycle*
                 "" "" ""* — — — pump "~ ""

Q~-
                      Aah pond liquor  t
                      return pump
                     ^r
       Dry carbide pit
            slurry
,. —. Makeup slurry
,•__» Bleed slurry
____ Return liquor
                  >LL
                   \'=r^~^r^/
                   Waste disposal pond
                     Note:  Only one scrubber systea presently In operation*
                          • Indicates a closed valve during 2/72 to 9/72.
                         M indicates a closed damper during 2/72 to 9/72.
                         Figure  2    Mttsul-Chemico lime  process

-------
                  Chiyoda dilute  sulfuric  acid process
                                 (Thoroughbred 101 process)
 Developer  Chiyoda Cheaical Engineering &. Construction
            1580 Tsurumi-cho, Tsurumi-ku, Yokohama

 Outline, of the process  Flue gas  is washed with dilute  sulfuric acid
 which contains an iron catalyst and is  saturated with oxygen.  SO-
 is absorbed and converted to sulfuric acid.  Part of the acid is
 reacted with linestone to produce gypsum.  The rest is  diluted with
 gypsum wash water and returned to the absorber.
 Description  A flow sheet  is  shown in Figure 3,
                                    Flue gas is first
 treated by a prescrubber to  eliminate dust and to cool the gas to
 l*fO°F.  The cooled gas  is led into a packed tower absorber containing
 1  inch  Telleretts.  Dilute sulfuric acid  (2 to 5$ H-SO.) which contains
 ferric  ion as a catalyst and is nearly saturated with oxygen, is fed to
 the packed tower.  About 90$ of S02 is absorbed, and partly oxidized
 into  sulfuric acid.

 The product acid is led  to the oxidising  tower into which air is bubbled
 from  the bottom to complete  the oxidation.  Most of the acid at 120-150*F
 nearly  saturated with  oxygen is returned  to the absorber.  Part of the
 acid  is treated with powdered limestone (minus 200 mesh) to produce
 gypsum.  A special type  of crystallizer has been developed to obtain good
 crystalline gypsum 100 to 300 microns in  size.  The gypsum is centrifuged
 from  the mother liquor and washed with water.  The product gypsum is of
 good  quality and salable.

 The mother liquor and  wash water are sent to the scrubber.  The amount
 of input water—wash water and the water to the prescrubber—is kept equal
 to the  amount of water lost by evaporation in the scrubbers and by
 hydration  of gypsum.   No wastewater is emitted.

 State of development   The operation of a pilot plant (6,000scfm) has led
 to the  construction of the following commercial plants:
                                  Commercial plants by Chiyoda process
    User
Plant site

Mizushima
Kainan
Nippon Mining

Fuji Kbsan

Mitsubishi Rayon  Otake
Tohoku Oil       Sendai
Daicel Ltd.      Aboshi
 Source of gas
Glaus furnace
Oil-fired boiler

Oil-fired boiler
Glaus furnace
Oil-fired boiler
Size, scfm

20,600
94,100

52,900
 8,200
59,000
Hokuriku Electric  Shinminato  Oil-fired boiler 442,000
Mitsubishi       Yokkaichi   Oil-fired boiler   413,000
  Petrochem.
Completion
October  1972
October  1972
December 1973
January  1973
October  1973
June     1974
December 1974
                                         80

-------
 Status  of  technology  The iron catalyst is less reactive at. low
 temperature but is as reactive as manganese catalyst at operation
 temperatures above 120eF  (Figure ^0.     I* *s no* poisoned by
 impurities in  the gas, even when flue gas from a coal-fired boiler
 is used.   Catalyst loss is very small (Figure 5).      The towers of
 the commercial plants are provided with rubber or FEP linings.
 Stainless  steel is also usable; the ferric catalyst works also as a
 corrosion  inhibitor.

 A large L/G ratio is required to attain high S0« recovery as shown in
in Figure  6 ;  large pumps and fairly large absorber and oxidizer are
 required as shown in the following table.


                                 Size of towers (feet)


                              Ab so rb_er_	       Oxidizer
      Ca-pacity fscfm)     Diameter  Height    Dianeter Eeight

           117,600           29.7     49-5       13.4    62.7

           294,100           49-5     49-5       21.1    62.7


A double-cylinder type reactor  (Figure  7)   including an oxidizing
 section in the center and a scrubbing section in the outer part has
been developed recently instead of using two towers.  The absorbing
liquor  goes down the scrubbing section,  then goes up in the oxidizing
 section and overflows to the scrubbing section.   The reactor enables
 some savings to be made in floor space and investment cost.

Advantages  The process is simple and the plant is easy to operate.
Even in the event that the gypsum-producing system has to be stopped
for a day or two for repairs, the absorbing system can be operated
continuously.  The concentration of sulfuric acid increases by 1 or
Zfo in this case but S0« recovery is not  decreased.  Catalyst is cheap
and is not poisoned by impurities in the gas.  Salable gypsum of good*
quality is obtained from limestone without scaling problems.

Disadvantage  Large pumps and a fairly large scrubber and oxidizer are
required.  A large L/G is required when  SO- concentration of inlet gas
Is high.
                                  81

-------
                           Cost estimation* ($1 = ?308)

(inlet gas S02 l,000ppm, dust 0.08 grain/scf)
(Outlet gas SO.  lOOppin, dust 0.04 grain/scf)
                                           Capacity
Plant cost (3) (A)
Fixed cost (S/year) (B)
Direct cost (S/year)
Limestone (0.250/lb)
Electricity (0.70/kUh)
Water (80/1,000 gal.)
Fuel oil ($3.18/bl)
Steam (0.120/lb)
Catalyst (60/lb)
Labor (312,000/year/capita)
Kaintenance
Subtotal (C)
Net running cost (3 + C = D)
Overhead (s) (12JS of C)
Sunning cost (D 4- E)
Desulfurization cost f'($/bl)
without by-product $/MWhi
credit v, N '
800IW
(1.51 x lO^scfn1)
17.2 x 106
3.10 x io6

430,000
789,000
25,400
1,035,400
39,700
11,000
144,000
344,000
2,818,500
5,918,500
338,200
6,256,700
0.645
•) 0.992
500MW
(0.95 x lO^scfn)
12.5 x IO6
2.25 x IO6

272,000
529,000
17,000
648,700
25,100
8,000
96,000
250,000
1,845,800
4,095,800
221,500
4,317,300
0.712
1.095
250M&T
(0.475 x 106scfn)
7.0 x 10°
1.26 x IO6

135,600
308,000
8,400
325,600
12,600
4,000
96,000
140,000
1,030,200
2,290,200
123,600
2,413,800
0.797
1.225
         * Estimation made by  Chlyoda  in February 1973
                                    82

-------
                            I

                    Water  dJU Repeater
             Prescrubber
CO
Dust
eliminator
         Waste  gas
                                              Limestone  Water
                                      Dilute
""1
  4

Air
                          O
                                      Water
                       Absorber  Oxidlzer
                                                                Centrifuge
                                                    V
                                                             V/ater
                                                              ^^.u—^V^W ^ -C*^
                 L-o
             Crystallizer
                                                                         Gypsum
                      Figure 3   Flow sheet of Chiyoda Thoroughbred 101 process

-------
 -P
 03
 C
 O
 r4
 •P
 03
 •o
 «H
 K
 O
      80


      60


      40


      20
     Figure
Fe
  •H-f-
              60    80    100  120  140
               Temperature (°F)


              Temperature  and oxidation ratio
             with  catalysts

    1.0
ra
*>
0>
>»
    0.6
           1,000         2,000         3,000

                S(>2 content  (ppm)

            Figure 5   Catalyst requirement

-------
?
o
§
 CM
O
03
   100
    80
    60
    20 .
-X Physical-chemical absorption
       (Chiyoda)

  Physical absorption
                Flue gas containing l,300ppm S02

             H	1	1	1	1	1	1	1	1	H
             200    400     600    800     1,000

             Liquid/gas ratio  ( gal./lfOOOscf )
          Figure 6   Liquid/gas ratio and S0£ removal


                            T Gas outlet
     Liquor
     distri-
     buter
                             Liquor outlet
   Liquor
   outlet
                                             Liquor
                                             inlet
          Figure 7  Double-cylinder type reactor

                              85

-------
                                                1  2)
                 5    Weliman-KSZ sodium process   *
 Developer  Well man-Lord, U.S.A.
            Mitsubishi Chemical Machinery  (MK£)
            6-2,  2-chorne, Marunouchi, Chiyoda-ku, Tokyo

 Process  description  Flue  gas is first washed by a prescrubber installed
 in the lower part of the sieve tray absorbtion tower  (pigure  8).
 The partially cleaned gas  rises in the tower while contacting a
 countercurrent flow of a concentrated sodium sulfite  solution which
 eliminates nore than 9C$ of  the inlet S09 to fora sodium bisulfite.
 Hist droplets are removed  by the elininator and deoister combination
 in the upper part of the tower and the gas is discharged at about 2650?
 after reheating in the oil-fired after-burner.  A sodium bisulfite
 solution is discharged from  the absorption tower and  is stored in a
 surge tank before it is sumped to an evaporator.  In  the evaporator
 the sodium sulfite solution- is heated with steam and  decomposed  into
 S0« gas  and sodium sulfite.
   2                              Na2SO, + S02 -

 The S00  concentration in the gas leaving the evaporator is 90$ after
 the water vapor  is condensed in a cooler.  In this evaporator, sodium
 sulfite  is gradually concentrated and crystallized.   The crystalline
 sodium sulfite is centrifugally separated from the nother liquor,
 dissolved in a condensate  from the cooler, and the solution is recycled
 to "be used as absorbent.   The recovered S02 is used for sulfuric acid
 production.   Tail gas from the acid plant is led into the absorption tower.

 Sodium sulfite is gradually  oxidized into sulfate by  the oxygen in the
 flue gas.   To keep ITa^SO.  concentration to a minimum  and optimum figure,
 a.  small  portion  of the mcfther liquor is purged; this purge liquor is
 used for gas cleaning in the prescrubber.  The bleed  is taken off this
 prescrubber circuit for the  purpose of removing contaminants, which
 otherwise  would  build up in  the system.  The major contaminants are
 sodium sulfate and sodium  polythionate.  The purge otreaa is subjected
 to wastewater treatment that involves the following:  (l) addition of
 HpSO.  to convert NaSO- and ITaHSO, to Ka-SO,;  SO. evolved is sent to
 tfie sulfuric acid plant; (2) alkali is added to^form the hydroxide of
 soluble  metal ions (vanadium, nickel, iron);  this precipitates then as
 hydroxides;  (j)  removal of solids by filtering; (4) neutralization by
 adding H2SO^. The final wastewater is a largely clear concentrated
 solution of  sodium sulfate.  This effluent eventually is sent to the bay.

 State  of development   A commercial plant designed to treat 118,OCOscfm
 of flue  gas  front oil-fired toilers has been in operation since June
 1971 at  Chlba plant,  Japan Synthetic Rubber Co.  A larger plant with a
 capacity of  treating 365tOOOscfm of flue gas  from oil-fired boilers is
 under construction at  Nishinagoya Station, Chubu Electric Power Co.  to
 start  operation  in July 1973.  It has been recently decided to build
 &  237»000sbfa plant (flue gas from oil-fired  boiler)  for Japan Synthetic
 Rubber Co. at Yokkaichi.

Status of technology  In Chiba plant,  Japan Synthetic  Rubber,  flue gag
from two oil-fired boilers  (1JO tona/hr  each)  is  treated by two 16-ft

                                 86

-------
 square sieve tray absorption towers at a rate of 90,000scfa per absorber.
 The plant came on-streara in June 1971 a^-cL has been operated for more than
 8,000 hours   in a year.    The S00 concentration in inlet  gas  normally
 ranges from 1,000 to 2,OOOppm and that in outlet gas from 100 to 200ppm.

 The major problem associated with the process is the necessity  to bleed
 a waste stream from the absorber licuor circuit  to  avoid  build-up of
 contaminants,  primarily sodium sulfate.   The following tabulation
 summarizes the present composition of absorber and  waste  streams:

           Absorber  feed Absorber Cut  Vastewater to   Wastewater after
treatment treatment
Ka2SO- 16-19$ by wt
Ka2S2°5a °
Fa2S04 5-7$
Suspended
solids
pH
COD
Plow rate
2-4$ 0
14-17$ 4-6$
5-7$ 3-6$
1-2$
(10, 000-20, OOOppm)
5-5-5
20, OOOppm
•» MM
0
0
7-16$
2-10ppm
7+0. 1
200ppia
1-1.5 tons/or
                               r

Clear sodium sulfate solution is emitted from Chiba plant.  In new plants
to be built in future, the sodium sulfate solution will be evaporated to
produce solid sodiun sulfate as a by-produc'i; or treated with lime to
precipitate gyp.sua and to recover a sodium hydroxide solution, thus
eliminating the wastewater.

Economics  The investment cost of the Chiba plant was $2,600,000 including
the cost for sulfuric acid plant.  The plant consumes approximately
755,000bbl/year of fuel and produces 13,200 tons/year of aulfuric acid.
The requirements of the desulfurization plant are shown belcw.

          Make-up caustic soda 3-4 Ib/bl oil for 86-93$ recovery of SO,,

          After-burner fuel    10.5 Ib/bl oil to heat to 266°F

          Steam                175 Ib/bl oil
          Cooling water        6.5 tons/hi oil
The investment cost for Nishinagoya plant, Chubu Electric (365,000ccfn)
is $5,200,000 including the sulfuric acid plant with a capacity of 90
tons/day.

Advantages  Stable reliable operation.  High recovery of S02-  The
sulfuric-acid plant is much smaller than usual because the concentrated
SO- recovered is used.

Pi sadvantage  The treatment of sodium sulfate formed by the oxidation is
not simple*
                                    87

-------
                        ----- M
                      H
     Boiler
           S02 gas -to   I
                                                         Stack
                                            gas to
                                              Plant
           ^2^0h plant j,     /\ Absorber
         	1     ]     cL	j>
                                                         Evaporator
                                       Coolin
                                       water
          NaQH
    Soot
en
CO
     Wastewater treatment
           section
Absorbing section    Recovery section
                       Figure  8   Wellman-MKK process

-------
                6  Kureha sodium-limestone process 1«2)


Developer  Kureha Chemical Industry Co.
           1-8, Horidonecho t Nihonbashi, Chuo-ku, Tokyo

State of development  Kureha first developed a sodium scrubbing process
to produce solid sotiiun s\ilfite to be sold to paper aills. In addition to two
176,000scfm plants operated by Kureha since 19°9» two plants have been
licensed, one to Mitsui Toatsu Chemical (ll2,000scfd) which began
operation in September 1971,  and the other to Konan Utility (I23,000scfm)
which started operation in late 1972.  Since the demand for sodiun
sulfite is limited, Kureha has recently developed a sodium- calcium
double-alkali process.  Tests with a small pilot plant led to the
construction of a larger pilot plant (3fCOOscfm) which has been in
operation since July 1972.  The larger pilot plant program is & joint
effort with Kawasaki Heavy Industries.  Two commercial plants will be
completed in 1974 to treat flue gas from oil-fired boilers at power
companies, one at Shinsendai station, Tohoku Electric Power (l50MW),and
the other at Shintokushima station, Shikoku Electric Power (150MW).

Process description  A flow sheet of the sodium-calcium process is shown
In Figure 9.    The scrubbing system consists of a venturi scrubber
vhere water is used to remove particulates and to cool the gas followed
"by a rubber-lined, grid-packed scrubber where SO- is absorbed in a sodiua
sulfite solution.  The water from the dust scrubBer is discharged at
a pH of about 2.5.  The pH of the liquor from the absorber is controlled
to 6.0-6.5.  With an inlet concentration of l,500ppa SO-, 9&fo removal is
achieved; the liquid gas ratio is 7gal/l ,000ft* of gas.  The feed to the
absorber contains 20-25^ sodiua sulfite and has a pH of 7-Q» the calcium
content was reported to be about J>Qppm.  The scrubber discharge contains
about 10$ sodium sulfite, lOye sodium bisulfite, and 2-5$ sodium sulfate.
Limestone pulverized in a wet aill equipped with a cyclone classifier
Is fed continuously along with scrubber liquor into an atmospheric
pressure vessel where sodium bisulfite reacts with limestone to form
calcium and sodium sulfites.
         2KaHSO, + CaCOj— ^ CaSOj'l/fflgO + NagSOj + 1/2H20 + COg

The reaction temperature is somewhat higher than the scrubber temperature
vhich IB about J40°F; residence tine for conversion is about 2 hours.
The slurry from the decomposer is passed through a centrifuge where the
ealciua sulfite crystals are separated from sodium sulfite liquor, which
Is then returned to the scrubber*

                                    89

-------
The calcium sulfite is reacted with air at atmospheric pressure in an
oxidizer developed by Kuraha.  Gypsum is removed from the oxidizer
discharge stream by a centrifuge,  The product is suitable for use in
wallboard and ceaent.

Oxidation of sulfite in the scrubbing and decomposition steps results
in the formation of sodium sulfate which cannot be regenerated by
reaction with limestone.  I;i order to control the eulfate level, a
sidestream fron the scrubber discharge is nixed with calcium sulfite
crystals and sulfuric acid ia added.  The net effect is to convert the
sodium sulfate to calciua sulf&te and produce sodium bisulfite for recycle.
Gypsun is separated by a centrifuge and added to the oxidizer loop.
          B02 + CaS03-l/2E20

Status of pilot Plant operation  The pilot plant (j.OOOscfm) has been
operated continuously since its coaple^ion in July 1972 except for the
scheduled shutdown  for inspection in September and December.  Almost
no scaling was observed.  Operation of the centrifuges has given
effective separation of solid and liquid phases.  Both calcium sulfite
and gypsum discharged from the centrifuge are dump solids which can be
transported by solid-handling equipment, if desired.  The crystals of
gypsum grow to around 100 microns.  The flue gas contains about 6?&
oxygen.  About ?# of the recovered S02 is oxidized to form sodiun
Btilfate.  The sulfuric acid requirement for the decomposition of the
snlfate is 125$ of  the theoretical amount.  Therefore, about 8.756 of
the product gypsum  is derived from sulfuric acid.

Economics  For the  production of a ton of gypsum, 1,200 Ib of limestone,
18 Ib of caustic soda (100#) for make-up, 100 Ib of sulfuric acid (9W,
340kWh of electric  power, 1,460 Ib of steam, and 18 tons of water are
required.  Investment cost is uncertain, but the investment requirement
would be split among the process steps as follows:
      Absorption 30#, Decomposition 50#, Sulfate conversion
      Oxidation
 Advantages  High recovery of S02 is  achieved with- limestone.  No  scaling.
 Sodium eulfate is decomposed to  recover sodium bisulfite and gypsum.
 Good quality of salable gypsum is obtained.  Both  gypsua and calcium
 eulfite discharged from the centrifuge hare less moisture  and are  easy
 to handle.

 Disadvantages  The process is less simple than the lime-gypsum process.
 STSTff* considerable amount of sulfuric acid it disadvantageous
 for plants where the product calcium sulfite or gypsum must be discarded.

                                   90

-------
Fuel
 Air
        X
             Make-up
                NaOH
                            Limestone


                              NaHS03
                                                         HpSOjj.
                                                             ^
                                                                    J   Gypsum
(1)  Reheater     (2)  Absorber     (3)  Mill and  classifier

    Decomposition tank    (5)   Centrifuge    (6)  Sulfate conversion tank

                       (8)  Oxidizer     (9)  Centrifuge
      (?)  Air compressor
              Figure 9   Plow sheet of Kureha sodium-limestone process

-------
                  7   Showa Denko aodiun-limeatone process

 PeTeloper  Showa Denko K.K.
            34» Shiba Miyamotocho, Minato-ku, Tokyo

            Ebara Manufacturing Co. Ltd.
            11-1, Asahiaaehi, Haneda, Ota-ku, Tokyo

 State of development  Showa Denko, Jointly with Ebara, recently constructed
 commercial plants for SO- recovery by sodium scrubbing to produce sodium
 sulfite for paper mills.  The plants include Kawasaki plant (88,000scfm)
 of Showa Denko, Kawasaki plant (I39*000scfm) of Ajinonoto, and Yokohama
 plant (I42,000scfm) of Asia Oil for flue gas from oil-fired boilers, and
 Sodegaura plant (47,000scfa) of Nippon Phosphoric Acid to treat tail gas
 from a sulfuric acid plant.  As demand for sodium sulfite is limited,
 Showa Denko and Ebara have started Joint tests on sodium-calcium process
 to by-produce salable gypsum.  A pilot plant (5»900scfm) has been in
 operation at Kawasaki plant of Showa Denko since 1971*  A commercial
 plant to treat 34°»°00scfa of flue gas is being constructed at Chiba plant,
 Showa Denko, to start operation in June 1973-

 Process description  The Showa Denko-Zbara process features the use of
 a vertical-cone type absorber as shown in Figures  10 and 12.
 A liquid (sodium sulfite solution) is charged from the bottom, blown up
 by the gas to absorb SOO, and flows back to the liquor inlet by gravity.
 Very good contact between gas and liquid particles is attained ensuring
 95-98$ desulfurization at a liquid/gas ratio of 7-14 gal./l,000scf
 (Figure 11).     Pressure drop ranges from 8 to 15 in.H?0.  A flow sheet
 of the sodium-calcium process is shown in Figure  12.    Flue gas from
an .oil-fired boiler containing l,500ppm SO- and about 1 grain/of dust is
 led directly into the scrubber; 95$ of the SO^ and about 60?b of the dust
 is removed by a sodium sulfite solution.  Most of the liquor discharged
 from the scrubber is recycled to the scrubber.  A portion of the liquor
 la led to a reactor and treated with pulverized limestone.

            SHaHSO, + CaCO, = Na2SO, + CaSO, + H20 + C02

 The calcium sulfite is separated from the sodium sulfate solution;  the
 solution is returned to the scrubber.   Calcium sulfite is oxidized in
 an oxidizer to  form gypsum.   As sodium sulfate gradually forma in the
 solution and tends to accumulate,  a portion of the liquor discharged
 from the scrubber is sent to a sulfate conversion step to maintain the
 sulfate concentration at a certain level.  In the conversion step,  the
 sulfate is treated with calcium sulfite and sulfuric acid to produce
 gypsum and sodium bisulfite.

            Na2S04 + 2CaS05 + H2S04 + 2H20 = 2(CaS04«2H20> + 2NaHSO,

 The bisulfite solution is led to the reactor.


                            92

-------
Status of  technology  The pilot plant has 'oeen operated for more than
one year without  serious trouble.  Both line and limestone have been
used for comparison.  Limestone reacts slowly with sodium bisulfite
requiring  a few hours.  By using line the reaction proceeds rapidly.
However, limestone will be used in the commercial plant because it is
anich cheaper than lime, and moreover, larger crystals of gypsum is
obtained with limestone.  An oxidizer developed by Showa Denko and Ebara
will be used in the commercial plant.  The oxidation proceeds a little
«ore slowly but the crystals of gypsum grow  larger than with a rotary
atoaizer developed by JECCO.

Advantages The scrubber is very effective for deeulfurization.  High
recovery of S02 ia achieved consulting limestone.

Disadvantages  The process is less simple than the lime-gypsum process.
The use of a considerable amount of sulfuric acid is a demerit for
plants whose by-product gypsum or calcium sulfite must be discarded.

.Economics  The estimated de sulfur 12 at ion cost for a 120,000scfm plant
for 9556 removal of SO. is shown below.

                      Plant cost    $1.7 million
                      7,000 hours operation in a year
                      By-product gypsum   192,000 tons/year


                   Requirements    (lb -oer bl oil)
KaOH
0.56
E2S°4
5.0
CaO
35.8
Water
560
Steam
0.38
Electricity
lO(kWh)
    Yariable cost           $O.J31/bl oil
    Fixed coat              $0.687/bl oil
    Desulfurization coat    $1.018/bl oil {without credit for gypstua)
                            $0.936/bl oil (with credit for gypsum)
                                 93

-------
              Gas
              1
             10  feet^.
<£>
4=
      Liquid
       Gas     T
Gas-liquid
 separator
                  Absorbing
                    section
         I
    28 feet

         I
         I
         I
                     Liquid
       Figure 10   Vertical-cone  type

            absorber for 60,000acfm
                                            150
                                          3 100
                                          bO
                                          0)
                                          H
                                          -P
                                          O
                                          a
                                          CM
                                          O
                                          Ki
                       50  ..
                                              0
L/G = 14
                           0     1,000   2,000    3,000   4,000

                                  SOg ppm  of  inlet  gas



                              Figure H    Relation  between

                              liquid/gas ratio  (gal./l,000scf)
                              and SO^ removal (pH 6,5}

-------
1C
Ul

                      Gas mixer
                                      Sulfate  treatment
                      Mist
                      eliminator
                                                                       i  Oxldlzer
                                                                                  Centrifuge
                                                                                    Gypsum
                   Figure 12  Showa Denko  sodium-limestone process

-------
                                                              2  4)
               8   NEK ammonia process and ammonia-lime process


 Developer  Nippon Kokan Kabushiki Eaisha
            1-1-3, Otemachi,  Chiyodaku, Tokyo


 Process description  Waste gas at 250°P from an iron ore sintering plant,
 containing 400  to l,000ppm SO. is first led into an electrostatic
 precipitator and then cooled to 140 °F in a cooler with water spray
 (Figure  13).    The  gas is  then led into a screen type scrubber
 (jinkoshi type  scrubber  )    for the absorption of S02 by a liquor
 containing ammonium sulfite.  In the scrubber 16 mesh screens of stainless
 steel  are placed with some inclination in five stages.  On three of the
 screens placed  at the middle  of the scrubber, the ammonium sulfite
 solution flows  slowly forming a liquor film which readily absorbs S0«.
 On  the other two screens placed at the upper part of the scrubber,
 water  flows slowly forming a water film to decrease plume formed by the
 reaction of S0? and ammonia*  About 95$ of S0« is removed when the pH
 of  the circulating liquor is  about 6.  Virtually no ammonia is lost when
 the pH of the liquor  is  6 or  below.

 The outlet liquor containing ammonium bisulfite is sent to an ammonia
 absorber*   Coke oven  gas containing a small amount of ammonia is introduced
 into the absorber.  The  liquor is sprayed to absorb ammonia and to form
 an  ammonium sulfite solution.  A large portion of the solution is returned
 to  the scrubber to absorb SO-.  The rest of the solution is sent to an
 oxidizer where  the sulfite is oxidized into sulfate by air bubbles
 produced by rotary atomizers.  The ammonium sulfate solution is evaporated
 to  produce crystal ammonium  sulfate.

 Nippon Kbkan has  recently developed an ammonia-lime double alkali process
 (Figure 1*0.     The  SO-  absorbing part is the same as in the ammonia
 process except  that no coke oven gas is used.  The liquor from the scrubber
 contains  about  (NSL)  SO,  7-5$,  NH,ES03 7»5$ and (NH.)2SO. 15$.   A portion
 of  the liquor is  sent  td  a reactor end Is reacted with milK of lime (10$
 concentration) under  normal pressure at 210°F.  The ammonia released here
 Is  sent  to the ammonia absorber to be absorbed by the liquor from the
 scrubber.   Calcium sulfate and sulfite are precipitated in the reactor.
 The slurry from the reactor is acidified with sulfuric acid to adjust
 the pH to  4 to promote oxidation.  The slurry is then led into an oxidizer
 equipped with rotary  atomizers to convert calcium sulfite  to   gypsum,
which  is then centrifuged.  Salable gypsum with a good quality is obtained.
The gas from the  oxidizer contains S02 and is sent to the scrubber.

 State  of development  After tests with a pilot plant to treat 17»000scfm
of waste gas  from the iron ore sintering plant,  a prototype plant to treat
88,000scfm gas to produce ammonium sulfate by reaction with coke  oven gas
was courpleted in  early 1972 at the Keihin works  of Nippon Kbkan.


                                   96

-------
 Additional  units  for the ammonia-lime double alkali process were
 completed in November 1972.  The construction and the operation of
 the plants  have been carried out as a research project by Japan Iron &
 Steel Federation.

 Status of technology  The Jinkoshi type scrubber capable of treating
 88,000scfm  gas  is 56 feet high with a cross section of 14 feet x 23 feet.
 The following conditions are used for S0« absorption:


       L/G  12gal/l,000scf       Gas velocity  7 to 13ft/sec.

       Pressure  drop  10 to 12 inches H«0

       Inlet SOg 400  to l.COOppm  Outlet S02  15 to 50ppm

 The outlet  gas  is at 140°F and is not preheated; heavy plume is observed
 from the  stack.   Tests have  indicated that the plume becomes slight when
 the gas is  reheated  to 180°F and nearly invisible at 240°F.

 When coke oven  gas is used as the source of ammonia, H«S in the gas is
 absorbed  to form  thiosulfate.


          2LS + 2SO," -f 2HSO," —^>3S20,   + 3H20


 The thiosulfate is not oxidized into sulfate in the oxidizer.  It is
 decomposed  by addition of sulfuric acid to the liquor discharged from
 the oxidizer.
By the decomposition SO. is released which is sent to the scrubber.
Elemental sulfur -formed by the reaction is removed by filtration.

In the ammonia-lime process, the oxidation of calcium sulfite into sulfate
is hindered under the presence of thiosulfate.  Therefore, it is better
not to use coke oven gas in this case.  To make up a small amount of ammonia
(about jfo} lost from the system, ammonium sulfate is added to the reactor
to react with lime and to generate ammonia.  The pH in the reactor is
maintained at about 11 to ensure gypsum crystal growth in the reactor.
To promote the oxidation of calcium sulfite present in the slurry from
the reactor, the slurry is acidified to pH 4 by addition of sulfuric acid
and then led into the oxidizer.  Gypsum grows into big crystals  (100
to 300 microns) and can easily Tbe centrifuged to a low moisture content
(about 10$).  The liquor from the centrifuge which is acidic is neutralized,
clarified and returned to the system.  No wastewater is emitted.
                                 97

-------
 _Adv£r Adages.  -rhe 3cx-&c-a type ssrabb^r it; effective for SO,. renaval with
 B relatively lev pressure drop.  A lar,^- niuount cf g^s5 up to about
 5GG,GQGscf2i can. be treated in cr.e s crabber.  2y the acnonia process,
 both SO,. In. waste gs.3 and. ar^o.iia in. coke over, gas are utilised to
 produce salable ammonium sulfats.  3y the air.onia-Iiiae process, salable
 gyp sun of good quality is obtained with no scaling problem.  No wastewater
 iB emitted.

                     coke oven, gas is used, hydrogen sulfide in thfc gas
        tates additional facilities.  The screen in the scrubber is
         to corrosion under inadequate operating condition.

              orL  A cost estimation for the anmariiE-line process TO treat
 ilue gas froa. an oil-fired boiler is shown in the following table in
 cc-parison with -;hat for the liae-gyps-on prccesa developed by Nippon
 Kokan which is sinilar to the Mitsubishi- JECCO lime gypsum process.

                     Cost: eatisation (31 = ¥308)
           (SO  in inlet gas l^COpprc, in outlet gas 70ppm)

                                  Agmsr-i a- 1 J_m e p r o c e s_s   Liae-sypsua process
^moxint of gas treated (scfm)       235,000    88? .000     235_T_OOp   882.000
 Investment cost ($1,000;             2,110      4,545       2f27J     5|644
Fixed cost (SljOOO/year)
    Depreciation                        382        822         409     1,055
    Labor (? persons)                    45         45          45        45
    Hepair                               63        136          68       175
    Insurance                             25           J         7
    Hana^ement                           95        553          84       295
Variable cost (Si f 000/year)
    Electric power (€11 mil/teWh)          168        634         1$1       822
    Steam I©Sl.5/ton)                    96        363
    Industrial water («15 nil/ton)          4         14           3        11
    Seawater(03 nil/ton)                                         1         2
    Quick line {©313. I/ton)              168        688
    Slaked lime t«S15. I/ton)                                   2?6       951
    Sulfuric acid (^328. 3/ton)            45        180
   Ammonia (f=>S87.6/ton)                 37        148

   Fuel  C
-------
                       Coke-oven gas
       Gas to stack
           	; Scrubber   i   p—»   j-
             *taJL^^          I  — *      1
                     ,   I  •  ^^w^» - f^MvMH
10
   Gas
        i	
       A
    Cooler
-***


„.*'
„*••


L
1 1
3
w
^

„(


,S
i
-K!

aterl i
'
"i ¥ '
Wi
1
1
1
Cooler i
ID '

>





•
-»



»— -p.








»




L











Te
                    Purified
                    coke-oven gas
Ammonia absorber        f
                        i
                                              Absor-
                                               ber
                   ~U
                                                    o
                                     Filter
                                                                                Tank
                                                                              Evaporator
                                                              Air  Ammonia
                                                                               Ammonium
                                                                               sulfate
                 Figure 13  Nlppn Kokan ammonium sulfate  process

-------
                                                               Recovered water
          Gas to stack
                          Gas to prescrubber
                                  | NH3 absorber
  Prescrubber
o
o
       Thi ckener

              Sludge
Reactor    pH
          adjusting
           tank
                                                                                  Neutra-
                                                                                   lizer
     S02 recovery
      section
              V
NH-j recovery  |
    section   j
NHo regeneration
     section
                   Gypsum
                  production
                    section
 Water
recovery
 section
Figure
                                         Nippon Kokan ammonia-lime process

-------
                          References
1)  H. ¥. Elder, P. T. Princiotta, G. A. Hollinden, and S. J. Gage,
    Sulfur Oxide Control Technology, Visits in Japan—August 1972,
    Interagency Technical Comittee, U.S.A., Oct. 1972

2)  J. Ando, Recent Developments in Desulfurization of Fuel Oil and
    Waste Gas in Japan (1973). prepared for U.S. Environmental
    Protection Agency through Processes Research, Inc., April 1973
    (in English)

3)  Haiendatsuryu no Subete (All About Waste-gas Desulfurization),  Jukogyo
    Shimbunsha, Nov. 1972

4)  M. Yokoi, Ryusan to Kogyo (Sulfuric Acid and Industry), Vol. 26,
    No. 1, 1973

5)  J. Sakanishi, ibid.
                              101

-------
ECONOMICS OF FLUE GAS DESULFURIZATION
                    by

             Gary T. Rochelle
        Control Systems Laboratory
  National Environmental Research Center
      Environmental Protection Agency
      Research Triangle Park, N.  C.
                     103

-------
Summary and Conclusions




     This paper summarizes the results and conclusions of an analysis of the




costs of flue gas desulfurization for fossil fuel boiler plants.  Cost details




of the desulfurization processes have been evaluated and, presented as a con-




sistent set of data from which it is possible to estimate capital and annualized




costs for any given plant conditions.  The following conclusions were derived




from that data.




     1)  Flue gas desulfurization can be applied to 75% of existing fossil fuel




         utility capacity at an annualized cost of 1.5 to 3.0 mills/kwh, which




         is less costly than substitute low-sulfur fuels, typically costing 3.0




         to 6.0 mills/kwh.




     2)  Regenerable processes are generally less costly than throwaway processes




         at waste disposal costs above $3/ton wet sludge.




     3)  Newer processes such as citrate and double alkali may be up to 20% less




         costly than currently available processes such as Wellman-Lord and lime-




         stone scrubbing.




     4)  Annualized costs  for a given process can easily vary from 2 to 6




         mills/kwh with typical variations of plant sizes and load factors.




     5)  Flue gas desulfurization can be significantly cheaper to use with




         medium sulfur coals if only a portion of the flue gas needs to be




         treated in order  to meet emission standards.




     6)  Extrapolation of  this costing technique to industrial boiler conditions




         indicates annualized costs of 60 to 130 C/MMBTU witli tlirowaway processes,




     7)  This cost data base consistently predicts actual first-of-a-kind project




         capital costs within 10 to 20%.





                                     104

-------
 Introduction
     A  thorough evaluation of  technology for control of SCL from power plants
 must include consideration of  numerous processes and approaches as well as
 several criteria of usefulness.  The scope of such an evaluation is illustrated
 in Figure 1.  The  technology to be considered includes not only flue gas clean-
 ing and low sulfur fuels for short term solutions, but also longer term
 alternatives such  as gasification, fluidized bed combustion, or solvent refining
 of coal.  This technology must be evaluated on the basis of the following
 criteria:
     1)  environmental  impact
     2)  stage of  development
     3)  economics
 This paper evaluates flue gas  cleaning processes on the basis of economics*  Its
 scope is indicated by the dotted line in Figure 1.

     For purposes  of evaluation the flue gas cleaning processes considered here
 can be divided into three groups:
     1)  Available throwaway - Limestone scrubbing and lime scrubbing
     2)  Available regenerable - Wellman-Lord, MgO scrubbing, and Cat-Ox
     3)  New/developing
         a)  Throwaway - Double alkali
         b)  Regenerable - Citrate and Stone & Webster/Ionics
 These processes were selected as being representative; they do not include all
 potentially important systems.

     Each of the three groups appears to be identified with one of the evaluation
 criteria listed above.  Throwaway processes have serious problems with environ-
mental impact of waste disposal.  Regenerable processes are often cited as
                                    105

-------
LOW SULFUR
   FUELS
GASIFICATION
 FLUID-BED
COMBUSTION
   SOLVENT
   REFINING
                        OTHER

                     APPROACHES
                                                       TECHNOLOGY
FLUE GAS

CLEANING
                   THROWAWAY
NEW PROCESSES
                                                             REGENERABLE
                          ENVIRONMENTAL
                             IMPACT
DEVELOPMENT
   STATUS

                                             ECONOMICS
                                    Figure 1.  Evaluation of 502 control technology.

-------
having higher costs.  New processes have not yet reached a stage of development

adequate for commercial applications.  To provide a quantitative evaluation for

process selection, these different criteria must be quantified for all of the

process groups.  This paper concentrates upon quantifying the economics of flue

gas cleaning processes, so that they can be compared with each other and with

other approaches to S02 control.  Flue gas cleaning will be evaluated on the

basis of all three criteria -in-«£ paper to be presented at the APCA annual meet-

ing in June 1973.


     In a paper presented in Fall 1972,   we summarized and analyzed the cost

data for five wet scrubbing processes:  limestone scrubbing, lime scrubbing,

Wellman-Lord, MgO scrubbing, and Stone & Webster/Ionics.  That data base has

been expanded to include the double alkali and citrate processes as well as

preliminary information on Cat-Ox.  The capital cost data for Wellman-Lord and

Stone & Webster/Ionics has been modified to reflect the higher costs of purge

handling and S02 reduction being estimated for the NIPSCO Wellman-Lord demonstra-

tion.  Corrections and supplements to the 1972 paper are being prepared and

will be available shortly.  This paper summarizes the results and conclusions

of the revised cost analysis.


     The major sources of cost data are given in Table 1.  The primary sources are

important, detailed estimates.  The secondary sources include project, scoping,

and open literature references of lesser detail or importance.



The Cost Model

     On the basis of the aggregated cost data, CSL has developed a series of

equations and graphs that allow cost estimates to be generated as a function

of process type and conditions of specific installations.  For ease of repre-

sentation, the scrubbing processes are divided into two process areas:
                                    107

-------
                        Table 1.  SOURCES OF COST DATA
Primary


                                           (2)
     Catalytic, Inc. — Limestone scrubbing


                                                    (3)
     TVA — MgO conceptual study; includes limestone


                                                                       (4)

     Davy-Power Gas — Demonstration project costs and proprietary data



     Allied Chemical — Proprietary data on sulfur production
Secondary



     M. W. Kellogg — Cost study of 12 processes



     Stone & Webster/Ionics — Costs of electrolytic regeneration


                                                      (8)
     TVA — Widow's Creek project estimate (limestone)


                                                    (9)
     Monsanto/Illinois Power — Cat-Ox project costs



     Chemico/Boston Edison — MgO project costs



     Bechtel — Limestone scrubbing costs
                                   108

-------
scrubbing and alkali handling.  Scrubbing costs are dependent on the flue gas




rate but not on the amount of sulfur removed.  Alkali handling costs vary with




sulfur rate but are independent of flue gas rate.






     The structure of economics is illustrated in Figure 2.  Capital costs




are made up of four segments:  process direct, other direct, contractor




indirect, and user indirect.  The process direct costs include labor and




material for process equipment.  For scrubbing, process direct costs are repre-




sented as a function of process type, flue gas rate, and retrofit conditions.




Alkali handling process direct costs are a function of process type and the



amount of sulfur removed per hour.  The other direct costs and indirect costs




ate estimated as a percentage of process direct costs.  In this analysis total




capital costs are computed as 170% of process direct costs, unless otherwise




noted.  Capital costs can be expressed as absolute dollars or $/kw of generating



capacity.






     Annualized costs include contributions from utilities and raw materials,




labor, maintenance and capital charges.  Utilities and raw materials cost per




year are functions of gas rate or sulfur rate, process type, and load factor.



Labor is assumed to be $225,000/year regardless of plant size.  Maintenance




cost is estimated as 4.5% of capital cost per year.  Capital charges at 17.5% of




capital cost are meant to include depreciation, return on investment, taxes, and




insurance.  Annualized costs can be expressed as absolute dollars per year or




as cost per unit throughput, such as mills/kwh, C/MMBTU, or $/ton sulfur.




Generally CSL has used mills/kwh.  Variations of annualized cost with load




factor depend on the units used.  At lower load absolute costs per year will




decrease, but relative costs per unit throughput will Increase.



                                109

-------
                          ANNUALIZED
                             COSTS
UTILITIES
AND RAW
MATERIALS
 LABOR
MAINTENANCE
 CAPITAL
 CHARGES
                           CAPITAL
                            COSTS
  PROCESS
  DIRECT
OTHER
DIRECT
CONTRACTOR
  INDIRECT
  USER
INDIRECT
                   Figure 2. Structure of economics.

-------
Estimated Versus Actual Costs




     Using this cost data base and estimating techniques CSL has evaluated




the costs of six actual construction projects.  The results of these evalua-




tions are given in Table 2.  Generally all but 10 to 20% of the actual or




contractor estimated investment costs were predicted from process configurations




and site conditions.  This consistently good prediction of costs was maintained




over a wide range of retrofit conditions, plant size, fuel type, and process




type.  Therefore this technique of cost prediction should be accurate in pre-




dicting variations in process costs and can be relied upon for process comparisons.






     However, the estimated costs are consistently about 15% lower than the




actual project costs.  This error may result from insufficient detail on actual




contractor estimates or from real costs for first-of-a-kind installations.  When




cost detail is lacking, assumptions must be made for the extent of inflation and




other direct and indirect costs.  These assumptions will probably be more con-




servative than the assumptions of a contractor trying to make a profit.  The




costs presented here are meant to be accurate for widespread commercial applica-




tions and do not include first-of-a-kind costs.  Nevertheless, we recognize




that the bias in our estimated capital costs may be partially real,  At the




worst however, the 15% lower capital cost projects would result in annualized




costs that are 5 to 10% low.






     In order to better illustrate project cost estimating, two of the cases




will be discussed in greater detail.  Calculations for the Will County limestone




scrubbing facility are summarized in Table 3.  The power plant is designed




for a new generation of 163 Mw, but its gross capacity is 177 Mw.  We have estimated




the design flue gas rate at 360,000 SCFM or 180 Mw gas equivalent.  The actual



                                      111

-------
                      Table  2.  ESTIMATED  CAPITAL  COSTS OF  DEMONSTRATION PROJECTS  ($MM)
Explained casts

Actual costs

Unexplained costs,
  % of actual costs
Plant characteristics
K)
Will
County
11.3
13.3
15%
Limestone
Retrofit
3% S Coal
177 Mw
Widow's
Creek
27.6
35.0
21%
Limestone
Retrofit
4% S Coal
550 Mw
Sherburne
County
31.0
36.1
14%
Limestone
New
1% S Coal
1360 Mw
LaCygne
29.5
32.5
9%
Limestone
New
5% S Coal
820 Mw
Mystic
5.6
6.2
10%
MgO
Retrofit
2% S Oil
155 Mw
  Mitchell



    7.7

    9.6

    20%
Wellman-Lord
  Retrofit

3% S Coal

115 Mw

-------
         Table  3.   ESTIMATED COSTS  OF WILL COUNTY LIMESTONE DEMONSTRATION


    (Basis:  180 Mw  (360,000  SCFM),  15 tons limestone/hour,  difficult retrofit.)


                    Process  area                            Cost,  $MM


    Scrubber, two modules                                    8.6

    Alkali handling, without  pond, grinding
     for 24 tons limestone/hr                               2^7

    Total                                                   11.3

    Actual Costs (includes considerable overtime)           13.3
        Table 4.  ESTIMATED COSTS OF THE  NIPSCO  WELLMAN-LORD DEMONSTRATION


Process area                     Comments                  Estimated3    Actual  Unexplaine<
                                                          by  model     Estimate
Scrubber
Evaporator
Purge handling

Sulfur plant

Boiler
Total
Single shell with particulate scrubbing
	
Very small, almost pilot scale
Earlier paper
Very small, almost pilot scale
Earlier paper
Not usually required

3.00
1.65
0.65
(0.30)
1.50
(0.95)
0.90
7.7
3.0
1.8
1.3
-
2.6
-
0.90
9.6
—
+ 0.15
+ 0.65
(+ 1.0)
+ 1.10
(+ 1.8)
_
1.9
a
 Includes actual indirect cost rates with predicted direct costs.

                                          113

-------
sulfur design rate is not disclosed, but the limestone rate is 15 tons/hr,



equivalent to about 3 tons/hr of sulfur.  The scrubbers were installed under



severe space limitations and are therefore rated as difficult retrofits.




Normally, the total gas flow could possibly go through a single large scrubber,



but two smaller modules have been used.  Furthermore, the limestone grinding



facilities were designed for almost double capacity (24 tons/hr).  The actual



costs are about $2 MM more than the estimated costs, but include a large  amount



of overtime labor not usually accounted for by the estimating procedure.   A




similar facility designed for 180 Mw, 3 tons S/hr with easy retrofit conditions



and using only one scrubber and matching sized limestone grinding should  cost



only $7.0 MM compared to the estimated cost of $11.3 MM for this facility.






     The estimated costs for the Mitchell Station Wellman-Lord process are




given in Table 4.  The costs are compared on an area by area basis with the



actual contractor estimates.  Costs for the scrubber and evaporator agree very



well.  Costs for purge handling and the sulfur plant reported in the earlier




CSL paper were not well founded and have actually been modified for this  paper




to reflect the much higher costs estimated for the Mitchell Station.  Even as




modified, the contractor costs for these two areas are almost a factor of two



greater than our estimated costs.  However, both the purge handling and sulfur



plant are very small, first-of-a-kind, almost pilot plant facilities.  Since



the estimating technique is intended to be most representative of larger,




widespread applications, the lack of accuracy under these extreme conditions



is not unexpected.  The Mitchell estimate also includes provision for a



package boiler which would normally not be required.  Another factor of great



importance is inflation.  The Mitchell estimate includes actual costs through




1975.  Our time basis is late 1972, so as much as 15% should be added to our



                                   114

-------
estimated costs to account for inflation.  This would put estimated costs
within 5% of actual project estimates.


Comparative Process Costs
     In order to provide a point of process comparison, capital and annualized
costs for a single, base case power plant are given in Table 5.  The range of
annualized costs between processes is quite small, varying from 1.95 to 2.75
mills/kwh.  At the conditions selected, lime and limestone scrubbing are somewhat
less expensive than the regenerable processes.  The new processes, double alkali
and citrate, appear to offer savings up to 20% in annualized costs over available
throwaway and regenerable processes.  The costs for the Cat-Ox process have been
scaled from the actual demonstration project costs and are therefore somewhat
preliminary.  Nevertheless, the estimated annualized cost of Cat-Ox is within
5 to 10% of the other regenerable systems, and actual costs could easily turn
out to be competitive.  At conditions different from the base case, the relative
cost ranking can change.  For example, higher costs of waste disposal would make
throwaway processes more expensive.

     Regenerable Versus Throwaway - The comparative costs of regenerable and
throwaway processes are typified by the annualized costs of Wellmah-Lord and
limestone scrubbing as compared in Table 6.  In the scrubbing area limestone
is more expensive because of the higher capital expense of using limestone
slurry with larger liquid recirculation, more complicated demisting, and
erosion resistant materials.  On the other hand the annualized costs of alkali
handling appear to be greater for Wellman-Lord, because of contributions from
large capital expense and utility requirements.  Therefore the major cost com-
ponents balance out, and overall annualized costs differ by only 10Z.
                                      115

-------
                   Table 5.  COMPARATIVE PROCESS COSTS
(Basis:  500 Mw, 3.5% S coal, retrofit, 60% load, waste at $3/ton wet
         sludge, sulfur credit at $15/ton, particulate removal included.)
      Process
Comparative Process Costs
Capital,      Annualized,
  $/kw         mills/kwh
Throwaway

   Double alkali

   Lime scrubbing

   Limestone scrubbing
   24

   35

   36
1.95

2.40

2.45
Regenerable

   Citrate

   MgO (to S)

   Wellman-Lord (to S)

   Stone & Webster/Ionics (to S)

   Cat-Ox
39
49
50
50
55
1.95
2.40
2.65
2.70
2.75
a
 Capital costs do not include disposal facilities, usually $5-15/kw.
                                   116

-------
      Table 6.  ANNUALIZED COSTS OF LIMESTONE SCRUBBING AND WELLMAN-LORD
Scrubbing

   Operating utilities and labor

   Capital charges and maintenance

   Subtotal
                                                Annualized costs (mills/kwh)
                                                Limestone      Wellman-Lord
                                                scrubbing
0.21

1.33

1.54
0.21

1.09

1.29
Alkali handling

   Operating utilities and raw materials

   Labor

   Capital charges and maintenance

   Waste disposal costs and sulfur credit

   Subtotal
0.23
0.06
Q
0.18a
0.45
0.45
0.06

0.99
- 0.15
0.92
1.35
Total annualized cost
2.46
2.64
 Capital charges are not included for a disposal pond.  Its costs are
 represented by waste disposal costs.
                                      117

-------
     The assumed sulfur credit for Wellman-Lord is $15/ton,  but at that level




it only amounts to 0.15 mills/kwh.  If acid were produced rather than sulfur,




the costs would be reduced about 0.15 mills/kwh in addition to the byproduct




credit which could amount to as much as 0.30 mills/kwh.   Therefore, acid pro-




duction could be as much as 0.30 mills/kwh less costly than sulfur production.






     The key assumption affecting the relative costs of regenerable and throwaway




processes is the cost of waste disposal.  The assumed cost in Table 6 is $3/ton




wet sludge ($1.50/ton dry solids) and corresponds to the typical costs of a




disposal pond.  Figure 3 illustrates the variation of the annualized costs of




lime and limestone scrubbing with waste disposal costs.   The middle range of




regenerable process costs is also shown.  The range of disposal costs at which




regenerable processes are competitive varies from $1 to 3/ton wet sludge.  In




practice, costs for disposing of sludge via dewatering and hauling can run as




high as $7/ton sludge.  Therefore, it is certainly not accurate to say that




throwaway processes are less expensive than regenerable systems, though there




will certainly be localities of low-cost waste disposal where throwaway systems




will be 5 to 20% less expensive.






     New Processes - The double alkali and citrate processes are typical of  the




cost benefit to be achieved by new systems.  The capital cost of double alkali




is compared with lime scrubbing in Table 7.  The double alkali scrubbing system




is much cheaper because it can achieve S09 and particulate removal in a very




simple single stage venturi scrubber with little concern for entrainment and




erosion problems.  Of course the double alkali process requires more equipment




for alkali handling, but the overall capital cost is still 15 to 30% less




expensive than lime scrubbing.  Utility and raw materials costs are the same





                                       118

-------
    3.0
CO
o
o
    2.5
                             LIMESTONE SCRUBBING
   2.0
                                                                 .IME SCRUBBING
                   12345

                       COST OF WASTE DISPOSAL, $/ton WET SLUDGE
                        Figure 3.  Process costs versus waste disposal costs.

-------
          Table 7.  CAPITAL COSTS OF DOUBLE-ALKALI AND LIME SCRUBBING
(Basis:  500 Mw, 3.5% S coal, typical retrofit, includes particulate scrubber.)
Process area
        Capital Costs, $/kw
Double alkali/lime    Lime scrubbing
Scrubbing

Alkali handling (not including pond)

Total
       18.1

        6.8

       24.9
32.0  (25.9)

 2.7
34.7  (28.6)'
"without particulate scrubber.
          Table 8.  ALKALI HANDLING COSTS OF CITRATE AND WELLMAN-LORD
(Basis:  500 Mw, 3.5% S coal, 60% load.)
                                                     Annua1ized costs, mills/kwh
Cost factor
Operating utilities and raw materials
Labor
Capital charges and maintenance
Sulfur credit
Total (alkali handling)
Citrate
0.20
0.06
0.57
-0.15
0.68
Wellman-Lord
0.45
0.06
0.99
-0,15
1.35
                                     120

-------
for both processes, so the difference in annualized costs is a result of capital
cost differences.

     The scrubbing costs of citrate and Wellman-Lord are essentially the same
since both processes use clear solution scrubbing.  The annualized costs of
alkali handling are compared in Table 8.  The citrate process regenerates
scrubbing solution with H2S to sulfur, so no steam is required as with Wellman-
Lord.  Furthermore, the regeneration reactor is quite a bit simpler than the
Wellman-Lord evaporator.  The citrate process requires a unit to produce H_S
from sulfur, but it should be cost equivalent to the Wellman-Lord unit for
production of sulfur from S02.  Therefore, the capital costs, and hence main-
tenance and capital charges, are 40% less for citrate.  The overall costs for
alkali handling are about 50% lower for citrate.

     The reduced costs for new processes may not totally materialize.  The cost
estimates are necessarily based on preliminary information.  Additional processing
equipment may be identified as the processes become better developed.

Cost Variations with Source Parameters
     As just' explained, the total variation of annualized costs with process type
at the base conditions is limited to a factor of about 1.4 from double alkali to
Cat-Ox.  However variation of costs with source parameters such as plant size,
fuel sulfur content, and load factor is much more pronounced.  The variation
of total capital costs (including waste disposal) with plant size is illustrated
in Figure 4.  Costs for the available systems, Wellman-Lord and limestone, vary
a factor of 1.65 from $65-75/kw at 100 Mw to $45/kw at 1000 Mv.  Citrate and
double alkali capital costs are 20 to 50% lover.
                                    121

-------
                                                                  BASIS: 3.5°, S COAL RETROFIT,  1 k* - 2 scfm  	
                                                                           WELLMAN-LORD
NJ
ro
                                        DOUBLE^ALKALI LIME
                                                           200       300


                                                     PLANT SIZE, MW


                                              Figure 4. Total capital costs.

-------
      The variation of  annualized  costs with  plant  size and load factor is




 illustrated in Figure  5.   Over  a  range of  100  to 1000 Mw, 40 to 70% load




 factor,  the annualized costs  for  a  typical retrofit plant with 3.5% S coal




 vary a factor  of  3 from 2  to  6  tnills/kwh.  The sludge waste disposal cost was




 treated in Figure 5 as a materials  operating cost  rather than capital cost.




 Therefore the  limestone process is  relatively  less costly than Wellman-Lord




 at  smaller plant  sizes and lower  load where  capital costs have a greater




 impact than materials  costs.






      Figure 6  illustrates  the variation of annualized costs of limestone




 scrubbing with sulfur  content.  In  most estimates  it is assumed that all of




 the gas  will pass through  the scrubber, in which case the cost varies a factor




 of  1.4 in going from 1 to  5%  S  coal.  The  other curves in Figure 6 represent




 costs if some  flue gas can be bypassed around  the  scrubber while maintaining




 a sulfur emission standard.   Clearly the cost  to treat flue gas from combustion




 of  1% S  coal will be zero  if  the  sulfur emission standard requires 1% S.  Actual




 costs for 3.5% S  coal  with a  2% S standard can be  as much as 50% less than




 costs for treatment of  all the  flue gas.   With a 1% S standard, the cost of




 using 2% coal  with flue gas scrubbing would  only be 1.25 mills/kwh (12.Sc/MMBTU)




 Therefore,  substitution of low  sulfur fuel is economical only if its incremental




 cost  is  less than 12.5C/MMBTU.  With a 1%  S  standard, annualized costs vary a




 factor of 2  from  1.25 mills/kwh at  2% S to 2.5 mills/kwh at 5% S.








Actual Distribution of  Annualized Costs




      It  is somewhat academic  to discuss process costs over a range of source




conditions as  discussed earlier.  In real  life,costs must be considered as applied




 to actual plants.   Figure  7 represents a distribution of calculated annualized



                                    123

-------
K>
jr
         6.0
         5.0
   4.0
     8  3.0
UJ
M
         2.0
                                                      LIMESTONE
                                                      SCRUBBING
                                               LIMESTONE SCRUBBING
         1.0
                 BASIS: 3.5*, S COAL, RETROFIT, WASTE AT S3, TON
                       WET SLUDGE, SULFUR CREDIT AT S15/TON S
                             I
                                         I
I
I    I
           50
                       100               200                    500
                                           PLANT SIZE, MW
                        Figure 5. Annualized costs versus plant size and load factor.
                1000
                             2000

-------
             3.0
Ul
             2.5
         I   2.8
         *i
        o
        o
" ^^
i
               5
              .v
             1.0
             0.5
                                                                        BASIS: LIMESTONE SCRUBBING,
                                                                       500 MW, 60% LOAD, RETROFIT
                                                                       90% REMOVAL
                                                                                     I
                                1.0               2.0              3.0
                                              COAL SULFUR CONTENT, wt %
                                     Figure 6. Annualized costs versus sulfur content.
                                                                           4.0
5.0

-------
    6.0
    5,0
    4.0
o  3.0

a
UJ
ISI
    2.0
    1.0
                   I
I
I
i
I
                  10           20           30          40          50          60           70

                       COAL-FIRED CAPACITY THAT CAN BE RETROFITTED AT COST OR LESS, percent
                                                            80
                                                            90
                    Figure 7.  Distribution of annualized costs over utility population, limestone scrubbing.

-------
costs over a random sample of coal-fired utility plants including 25% of the



plants in the 1969 FPC form 67 survey.  Costs of limestone scrubbing for each



plant were calculated as a function of plant size, load factor, and fuel



sulfur content.  Retrofit factors were estimated on the basis of unit age and



size.  Each system was assumed to treat all of the flue gas for SO  and
                                                                  X


particulate removal.  Waste disposal was represented as an operating cost at



$37ton sludge.  In addition a random variation of plus or minus 15% (over two



standard deviations) was incorporated into the costs to account for variations



in waste disposal costs and retrofit factors that could not be correlated



with plant parameters.





     As shown in Figure 7, 75% of the coal-fired utility capacity in the U. S,



could be retrofitted with limestone scrubbing at a cost of 3.0 mills/kwh or



less, 50% at a cost of 2.0 mills/kwh or less.  Thus representative costs for



widespread application of flue gas cleaning would be 1.5 to 3.0 mills/kwh.



It is also apparent from Figure 9 that 25% of the capacity would cost more



than 3.0 raills/kwh (about 30C/MMBTU) to retrofit with  flue gas scrubbing.



This capacity would probably be most economically controlled by  the use of



low-sulfur fuels, if available at costs less than the  cost of  flue gas



scrubbing.  It would appear from Figure 9 that 15% of  the capacity would cost



more than 6 mills/kwh (60C/MMBTU).  This capacity would certainly consider



clean fuels or another more economic approach to sulfur emission abatement.



Since substitution of low-sulfur fuels is expected to  add an incremental



cost of 30 to 60C/MMBTU, flue gas cleaning would be the better economic



choice for 75 to 85% of current coal-fired capacity.





     As shown in Table 9, costs for typical new plants would be about  1.85



mills/kwh (18.5C/MMBTU) for coal or 1.3 mlllc/lcwh (13C/MMBTU)  for residual



                                      127

-------
             Table 9.   COST OF LIMESTONE SCRUBBING FOR NEW PLANTS
(Basis:   500 Mw,  70% load  factor,  no particulate scrubbing,  70%  indirect
         costs,  and 1972 dollars.)


                                                                      *»
                                               Annual!zed,     Capital,

Coal,
Oil,
Fuel
3.5% S
2.5% S
mllls/kwh
1.85
1,31
$/kw
25
20
 Waste disposal included in operating cost,  but  not  in capital cost.
 Disposal ponds would cost $5-10/kw.
                                      128

-------
                                      26



oil.  Thus  the use of substitute low-sulfur coals or even desulfurized oil



at  incremental costs over 20C/MMBTU would not be economically favored.




Therefore most new fossil-fired plants will probably use flue gas cleaning



until a cheaper alternative  (such as fluidized bed combustion) can be



developed.








Costs with  Industrial Boilers




     Generally the costs presented thus far have related primarily to applica-



tions on utility boilers.  Coal-fired industrial boilers present a different



picture.  A typical coal-fired industrial boiler plant has three boilers with




a total capacity of 500 MMBTU/hr (50 Mw).  The average load factor is probably




about 50%.  Higher excess air (50% compared to 20% for utilities) is used



with the typical stoker-fired boilers, so larger scrubbers must be used.



Capital expense is a major factor that can have more than twice as great an



impact on industrial boiler annualized costs.  Industrial facilities are built




primarily with equity, not 50% debt/50% equity as with utilities; therefore




taxes are substantially higher.  Furthermore, non-regulated industry necessarily




requires higher return on investment, because of the risk involved.  The com-




bined effect of equity financing and higher investment returns is that capital




charges for depreciation, return on investment, and taxes amount to 33Z of the




capital cost per year, not 17.5% as with utilities.  Also, the relative impact




of capital  costs is increased because of the small scale of operation.






     Figure 8 presents the annualized costs of lime scrubbing and double alkali




as applied  to coal-fired industrial boilers.  The capital costs were extra-




polated from utility costs.  Actual capital costs may be lower if shop-fabricated




                                     129

-------
           180
           160
           140
         "120
CO

o
S
o
Nj 100
_J



I


    80







    60
                                                                                  1       T
            40
              30
                    BASIS:  2 Ib S 106 Btu REMOVED, 50*. LOAD,

                    38'o CAPITAL AND MAINTENANCE CHARGES
                          I
                                    I
I
                                                       I
 I       I
                  50
                400
                    100               200


               BOILER PLANT SIZE, 1Q6 Btu/hr FUEL


Figure 8. Annualized costs of throwaway processes on industrial boilers.
600
800   1000

-------
 systems are practical.  Labor and overhead costs were assumed to be constant

 at $40,000/year.  Waste disposal costs were assumed to be $5/ton of wet sludge.

 The annualized costs on a boiler plant of 500 MMBTU/hr are 60 to 80C/MMBTU.

 On a plant of 100 MMBTU/hr the costs are 100 to 130C/MMBTU.  Such prohibitive

 costs indicate that substitute low-sulfur fuels will probably be used if they

 are available.


 References

 1.  Burchard, J.K., Rochelle, G.T. e£ al^. "Some General Economic Considerations
     of Flue Gas Scrubbing for Utilities," Sulfur in Utility Fuels:  The Growing
     Dilemma, pp. 91-124, McGraw-Hill, Inc. (1972).

 2.  Catalytic, A Process Cost Estimate for Limestone Slurry Scrubbing of Flue Gas,
     NTIS No. PB219-016  (1973).

 3.  Tennessee Valley Authority, Sulfur Oxide Removal from Power Plant Stack Gas:
     Magnesia Scrubbing-Regeneration. Contract No. TV-29233A  (draft, 1973).

 4.  Davy - Power Gas, Project Manual for S00 Demonstration Unit, Contract No.
     EPA-68-02-0621  (draft, 197TT.          -             ~ ~~~~

 5.  Allied Chemical, proprietary data, 1972.

 6.  M.W.  Kellogg, Evaluation of S02 Control Processes, NTIS No. PB204-711 (1971).

 7.  Humphries, J. J. e£ jrl., An SO  Removal and Recovery Process, Chemical
     Engineering Progress 67 (5): 65, May, 1971.

 8.  Tennessee Valley Authority, private communication.

 9.  Control Systems Laboratory, EPA, "Briefing Document: Cat-Ox Demonstration"
     (1972).

10.  Quigley, C.P.,  "Magnesium Oxide Scrubbing System at Boston Edison Company's
     Mystic Station," Sulfur in Utility Fuels;  The Growing Dilemma, pp 287-290,
     McGraw-Hill (1972).                                         ~

11.  Sherwin, R.M. £t al., "Economics of Limestone Wet Scrubbing Systems," Proceed-
     ings  £f Second International Lime/Limestone Wet-Scrubbing Symposium. APTD-
     1161, pp. 745-64 (1972).                                       	
                                      131

-------
Notes




1.  The mention of company names or products is not to be considered




    as endorsement or recommendation for use by the U. S. Environmental




    Protection Agency.






2.  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,  NERC/RTP  is providing conversion




    factors for the particular non-metric units used in the  document.




    For this report these factors are:
                British




            1 SCFM (60°F)




            1 short  ton




            1 MMBTU
         Metric




1.61 NM3/hour (0°C)




 .91 metric tons




252,000 kilocalories
                                   132

-------
STATUS OF TECHNOLOGY OF COMMERCIALLY
   OFFERED LIME AND LIMESTONE FLUE
     GAS DESULFURIZATION SYSTEMS
                   by

               I. A. Raben
            Bechtel Corporation
         San Francisco, California
                     133

-------
                              SUMMARY
The  status of technology for commercially offered lime/limestone
SO_  removal processes has been evaluated.   Suppliers have been con-
   L*
tacted to determine the present status of design criteria.  It is im-
portant  to note that most commercial system suppliers will guarantee
80- and 90-percent SO  removals using limestone and lime scrubbing,
                     u
respectively.
Between 1970 and 1973,  liquid-to-gas ratio has increased from 20-30
to 40-80 gallons per thousand cubic feet, and stoichiometry has de-
creased from 1.7 to as low as 1. 2.   The new values for both parameters
tend to make designs more reliable and free of scaling.

There are 21 full-size units being designed or constructed.   The num-
ber of SO.  removal systems (using alkali addition into scrubber) opera-
ting  for any period of time is quite limited.  The Will County Station of
Commonwealth Edison is the only domestic coal-fired unit with scrubber
addition that has  operated for a significant period.  The facility has
shown high SO2 removal but has been plagued with mechanical  problems.
Much has been learned from its operation to advance commercial
technology.

Capital  and  operating costs have been estimated for new units.  The
capital cost of a 500-megawatt unit with the use of limestone is esti-
mated at $45 per kilowatt; a typical operating cost with limestone is
                               134

-------
estimated at 23-25£ per million Btu.  It is not possible to generalize
capital costs for retrofit units because of the special requirements for
each plant.   New SO- removal systems using lime will cost $38 per
kilowatt with typical annual operating costs of 23£ per million Btu.

The two areas  requiring further study are demonstration of long-term
reliability and  sludge disposal management.  Long-term reliability
will be evaluated and hopefully demonstrated when many of the 21 units
come on stream in 1973 and 1974.  To develop ecologically sound sludge
disposal methods,  techniques  must be studied in much greater detail.
                                 135

-------
                       ACKNOWLEDGMENT

The author wishes to express his appreciation for contributions
from the engineers of Bechtel's Scientific Development Air Qual-
ity Group.   The cooperation of the SO- removal system suppliers
is also gratefully acknowledged.
                                136

-------
                            Section 1
                         INTRODUCTION

When the first EPA Symposium on Lime/Lime stone Scrubbing was
held in November 1971 in New  Orleans,  Louisiana, there were only
three wet scrubbing  systems operating.  These units consisted of
alkali injection into the boiler furnace followed by wet scrubbing.
Today there are over 20 commercial systems being designed,  con-
structed,  or operated.  We have gained much knowledge in the past
18 months, but we still have more to learn.

The  lime/limestone  sulfur-dioxide removal process is the  most ad-
vanced system at this time.  The primary reasons for this are:
         The process is more fully characterized than other
         processes.  Approximately 20 pilot plant test pro-
         grams have been conducted over the last 3 years.
         In addition, the EPA Test Facility has been operat-
         ing for over a year, and test results will be reported
         at this meeting.
         The process has relatively low capital and operating
         costs.
         The process can achieve high sulfur-dioxide removal
         efficiency.
                               137

-------
However, these systems are not without problems, which include con-
trol of scaling and erosion/corrosion and solid waste disposal.

This paper describes the commercially offered lime/limestone pro-
cesses for SO^ removal and indicates the type of alkali, capital and
operating costs, types of scrubbers, and design criteria.   It presents
the process chemistry as it relates to system reliability and equip-
ment selection,  and it discusses the commercial systems as to boiler
size,  scrubber type, vendor,  and status.  The paper also evaluates
solid waste disposal.
                               138

-------
                            Section 2
     CHEMISTRY OF  LIME-LIMESTONE SCRUBBING OF
The chemistry of lime-limestone scrubbing is complicated because of
the large number of species present in the system at equilibrium.

2. 1      RAW MATERIALS AND CONSTITUENTS

The three raw materials — gas, fly ash, and limestone — each contain
several constituents that affect the chemical composition of the system.

For the power plant,  the gas supplies  SO ,  SO3> CO , NO, and NO ;
the ash contributes Na,  K, Ca,  Cl, Fe,  Si, and others.  Limestone
gives Ca, Mg,  and other constituents in minor proportions —  Na and K.

2. 2      REACTIONS

The main reactions in the scrubbers are assumed to be:

    •    Absorption of SO
    •    Hydrolysis to form H  SO  acid
    •    Reaction of sulfite ion from H_SO  with calcium ion  from
         CaC03 or  Ca(OH>2              3

These reactions are affected in several ways  by other constituents in
    system.  Detailed studies of the system chemistry have been
                               139

-------
carried out by TVA (Ref. 1) and the Radian Corporation (Ref.  2).
TVA studied the effect of supersaturation, ionic strength, and sul-
fite oxidation.  Radian developed a computer program using 41 spe-
cies and 28 equations to predict equilibrium compositions in the
scrubber circuit.

The main equations in the scrubber can be written (Ref. 3):
    •    H SO j*.HSO"  + H+
             -    +      =
    •    HSO J± H  +  SO
    •    CaCO3(s)^CaCO3(aq)4±Ca+"1"  + CO3  +  H+«± CaHCO*
    •    CaHCO* £ Ca++ + HCO~
    •    Ca++  + SO3  + 0. 5  H2O ;2CaS03  •  0. 5
    •    CaSO, +  1/2 O^^i
              3         2
If the system is assumed to be one of sulfurous acid formation fol-
lowed by the reaction of acid with lime or limestone, then the fol-
lowing effects may influence the overall kinetic rate:
    1.   Diffusion of SO_ to and through the gas film at the
         liquid surface
    2.   Dissolution of SO^
    3.   Hydration of SO- to H2SO3> H*. and HSO"
    4.   Dissociation of HSO, to form SO_
    5.   Diffusion of H-SO. and ions through the liquid film at
         the droplet surface  and into the droplet interior
    6k   Hydration of CaO to Ca(OH>2  when CaO is used
                                140

-------
     7.   Dissolution of Ca(OH)  or CaCO.
                                       3
     8.   Reaction of Ca(OH)  or CaCO, with H  to give C&++
                       ++        =
     9.   Reaction of Ca  with SO,  to precipitate CaSO,
                                 •3                     3

Available data indicate reactions in steps 3, 4,  8, and 9 are rapid.
The controlling mechanisms are therefore either gas diffusion,  liquid
diffusion, CaO hydration, or dissolution of CaCO, or Ca(OH)  .  For
                                               •5          Ci
the most used design case — introduction of CaCO into the scrubber —
gas phase mass transfer of SO  and CaCO^ dissolution are the con-
trolling steps. This case was studied by Boll (Ref. 4) in a three-stage
scrubber, and he found that these two steps were most critical.
2. 3      INFLUENCE OF PROCESS CHEMISTRY ON SO  REMOVAL
         DESIGN                                      L
In gas absorbers, one of the major criteria establishing the size of the
equipment is the  rate of mass transfer from the gas to the scrubbing
medium.  For lime or limestone scrubbing,  gas-phase  absorption or
chemical reaction rate, or both, may be rate-controlling  resistances.
For optimum SO   mass transfer,  it is necessary to maximize the sur-
               L*
face area of the calcium-contributing  reactant and to minimize the re-
sistance of the gas-to-liquid interface — at minimum cost.  Both hy-
drated lime,  Ca(OH)2>  and precipitated CaCO_ [resulting  from the
reaction of Ca(OH)2 with CO. in the flue gas] have very high specific
surface area compared to ground limestone.  Hence, if limestone is
used, it is important to supply a large surface area per volume of
SO, per unit  of time in the scrubber to keep the  solution supplied
with calcium ions.
                               141

-------
Resistance to gas phase SO  mass transfer can be minimized by the
proper choice of scrubber  type and high gas velocity.  Maximizing
reactant surface area can be accomplished by:

    •    Large liquid holdup of slurry in the scrubber
    •    High slurry recirculation rate
    •    High solids content in slurry
    •    Increasing feed stoichiometry
    •    Reducing the particle size of feed

Liquid holdup depends mainly on  scrubber type, in descending order as
follows:  tray-type, packed-bed,  floating-bed, spray tower, and ven-
turi.  Slurry recirculation rates  in the order  of 20 to 80 gallons per
thousand cubic  feet of saturated gas would be  required, depending on
the type of scrubber used.   Solids content of 5 to 15 weight-percent
are normally used in the slurry.   Limestone feed stoichiometry may
vary from  1. 2 to  1.5  (stoichiometry being defined as the molal ratio
of alkali feed to SO. absorbed).

Reducing the particle size  of limestone is an obvious way of increasing
surface area.  Good results have been obtained at a particle size of
80 to 90 percent through 200 mesh. *  It should also  be noted that if
 !U. S. Mesh No.              Typical Median Range
                          90% through basis, microns
      100                            27-66
      200                            14-35
      300                             7-18
                               142

-------
particles are too fine, solids settling may become a problem.  In

addition, the mill capacity will decrease and power consumption in-

crease for finer size reduction at a given limestone hardness. *


High slurry recirculation rate and high solids content in the  slurry will

not only maximize reactant surface area, but will also minimize scal-

ing potential,  if sufficient delay time  (approximately  10 minutes) is

provided in the recirculating tank.
2.4      WATER BALANCE AND AMBIENT QUALITY
         CONSIDERATIONS
It is necessary that wet scrubbing systems be operated in closed loop

to  minimize effects on  adjacent water basins.  In closed loop opera-

tion the amount of make-up water used will be equal to the amount of

water evaporated to saturate the flue gas plus the amount entrained

or combined with solid waste produced.  The remaining  liquor (i.e.,

pond or thickener overflow) is recycled to the process.  From a

process standpoint, closed loop operation requires recycling saturated

solutions.  Process conditions must be carefully controlled to prevent
scaling.
* Lime stone hardness "Work Index" — 6-10.
                               143

-------
                            Section 3
                 COMMERCIAL SYSTEM DESIGN

3.1      PROCESS APPROACH

The development of the lime- lime stone scrubbing system has taken
three process routes:

    •    Introduction of limestone into the scrubber circuit
         (Figure 1)
    •    Introduction of limestone into the boiler to produce
         CaO, followed by scrubbing of the flue gas (Figure Z)
    •    Introduction of lime into the scrubber circuit
         (Figure 3)

The most used process route is introduction of limestone into the
scrubber.  This approach has the advantage  of minimum effect on
the power plant; it can achieve high  SO_  removal with minimum
scaling and plugging.  The disadvantage  of this process  is that
limestone is less reactive than lime; to offset this  limitation, a
higher stoichiometric ratio of limestone to SO   must be used, more
                                            Lt
slurry must be recirculated {higher liquid-to-gas ratio), and a counter-
current scrubber with  several stages is  required.

The second process  approach of introducing  limestone into the boiler
furnace produces a calcined limestone.  Calcium oxide (CaO) enters
                               144

-------
                              GAS TO STACK
 STACK
  GAS
                                                       CfS03+CiS04
                                                         TO WASTE
    Figure 1.  Scrubber Addition of Limestone — SO2 Removal

                              I        •  GAS TO STACK
IOH.EI
C«OGAS



SCIUIIH



PUMP
TANK
••




SITTllft
J
~1
CsCOj*-
                                                               TO WASTE

 Figure 2.  Boiler Injection of Limestone Followed by Wet Scrubbing
       STACK GAS-
                                    •*- 6AS TO STACK
                          SCKUIIEI
                 ۥ0
                                           PUMP
                                           TANK
                                                                ~l
                                                              CiSOs4CiS04
                                                                TO WASTE
       Figure 3.  Scrubber Addition of Lime - SO, Removal
                                145

-------
the scrubber with the flue gas.  Problems with this process include
boiler fouling,  inactivating the lime by everburning,  and increased
scaling in the scrubber at the dry-wet interface.  This approach is
no longer offered by the SO, removal system supplier who originally
                          Ct
offered it, unless requested by the customer.

Scrubbing efficiency can be  increased in the third process approach
by using lime in the scrubber circuit.   This process has two dis-
advantages — the higher cost of lime over limestone and the greater
potential for  scaling under certain process  configurations.

3.2      SCRUBBER DESIGN

In designing a scrubber system, the following main design criteria
must be considered:

    •    Sulfur content of fuel
    •    Ash content of fuel (Is fly ash removed  separately?)
    •    Percent SO  removal required
                   £
    •    Scrubber type
    •    System turndown
    •    Scrubber size and spare capacity
    •    Mist separators
    •    Gas  reheat
    •    Waste disposal
    •    System power losses
    •    Materials  of construction
                               146

-------
3.2.1    Sulfur and Ash Content of Fuel
The sulfur content of the fuel burned in the power plant determines


the required SO  removal to meet the new performance standards


(1. 2 Ib SO  /million Btu for coal and 0. 8 Ib SO  /million Btu for
         Lt                                 w

oil).   Three-percent sulfur  coal requires approximately 80 percent


removal of SO  from flue gas,  whereas 0. 8 percent sulfur coal
             Lt

(depending on heat value) requires 40 percent removal.  Pilot plant


data have shown that SO removal from low sulfur flue gas is less
                       L*

difficult for the same liquid-to-gas ratios (40-80) because the amount


of SO  absorbed per pass is significantly less.  This can be seen by


comparing inlet SO  values  of 2400 ppm (high sulfur coal) to 600 ppm
                  Ct

(low sulfur coal.  The SO  absorption per pass is 1920 ppm for the
                        £»

high sulfur case and only 240 ppm for the low sulfur case.  Control


of scaling and supersaturation  in the scrubbing systems is more


predictable and thereby provides greater system reliability.  In


addition, the lime or limestone requirements are significantly less


for low sulfur fuels.  Waste disposal sites also require less land.





If the fly ash is removed from  the flue gas before SO  removal is


accomplished, the scrubbing system design can be different.   This


is particularly true as to scrubber selection.  A spray column could


be used to obtain 80 to 90 percent SO  removal with lower gas pressure
                                   Lt

drop.  Waste solids disposal is also influenced by this design con-


sideration.
If SO2 and fly ash are removed simultaneously, a venturi followed by


an afterabsorber, two stages of venturi, or turbulent contact absorber
                                147

-------
(multistage) offer good possible choices for high removal.  Waste
solids requirements are higher,  and process chemistry becomes
more complicated.

3.2.2    Scrubber Type

The major criterion for scrubber selection is its capability to remove
both sulfur dioxide and particulates with high efficiency (SO  removal
greater than 80 percent and particulate  removal greater than 99 per-
cent).  Other factors considered are ability to handle slurries without
plugging, cost, control, and pressure drop.

Additional variables are:

    •    Liquid-to-gas ratio
    •    pH control
    •    Stoichiometric ratio of alkali to SO
                                          C*
    •    Percent of solids in slurry
    •    Limestone particle  size
    •    Number of stages

The scrubber types that have been tested to date include the  following:
         Venturi.  The venturi scrubber (Figure 4) has been used
         when both particulate (fly ash) and sulfur dioxide must
         be removed.  The venturi has good capability to remove
         fly ash down to 0. 02 gr/SCF with pressure drops of 10
         to 15 inches HO and liquid-to-gas ratios of 10 to 15 gpm
         per  1000 cu ft gas for typical dust loadings and particle
         size distributions from power plant stack gases.  The
                               148

-------
               TWO-STAGE VENTURI
                    (CHEMICO)
        UMIIOROUE DRIVC
         FOR PIUMI K»
                                    GASINUT
       KCYCU
GASOUtllT
                                                              ONE-STAGE VENTURI
                                                                   (CHEMICOI
      "PIUMIIOS"
                                                         INUTS
                                                     PIUMI 101 SHAFT

                                                    MUiriPU
                                                 rANCENTIAl UMTS'
                                               CAS OUTUT
                                                                                   HOT CASINUT
SPRAY NOZZLE IARR&
 IFOR PIUMB BOH
                                     Zntf. STAG!
                                   •.MIST SEPARATOR
                                                       SPRAY
    •DENTAL KHK."

    "PLUMB 101"

    MIST SEPARATORS
                    VENTURI UNIT WITH AUTOMATIC THROAT AREA CONTROL
                                   
-------
venturi may contain an adjustable throat area that
permits control of pressure drop over a wide range
of flow conditions.  The venturi scrubber is limited
in SO  removal to 40 to 50 percent per stage with
lime-lime stone due to the short liquid residence time.
It therefore requires two stages of Venturis or an
afterabsorber to achieve 85 to 95 percent SO  removal
from high sulfur fuels at total liquid-to-gas ratio of 80.

Turbulent Contact Absorber (TCA).  The TCA (Figure 5)
is a countercurrent multistage scrubber consisting
of screens that both support and restrain the plastic
spheres.  The  spheres move in a turbulent fashion
providing good gas-liquid contact and scale removal.
The number of stages generally are between  two and four
for high SO removal with liquid-to-gae ratio of 40 to
50.  The pressure drop per stage is approximately
2 to 2. 5 inches HO.
                Lf
Marble-Bed Absorber.  The marble-bed absorber
(Figure 6) utilizes a 4-inch bed of packing of glass
spheres (marbles) that are in slight vibratory motion.
A turbulent layer of liquid and gas above the  glass
spheres increases mass transfer and particulate
removal.  This scrubber has been used mainly in
the process where the limestone is added to the furnace
for calcining,  and the flue gases are scrubbed to
remove SO .  Pressure drop is generally 4 to 6 inches
HO.  Liquid-to-gas ratios of 25 to 30 have been used.
Packed-Bed Absorber. The packed-bed absorber must
use open packing to prevent plugging.  Packed towers
have been tested in a pilot plant with high SO  removal
and no significant scaling with low sulfur coal.  Pressure
drops are low — 0. 4 inches HO per foot of packing.
Scale control is extremely important in this type of scrubber,
requiring liquid-to-gas ratios of 30 to 60.

Spray Colurnn. The spray column (Figure 7) is  a counter-
current type scrubber that has a low pressure drop.  The
spray tower requires high liquid-to-gas ratios (L/G=80)
and several stages of sprays to achieve high  SO  removal.
                                             Lt
                      150

-------
                                     RECIRCULATION
                                     OUTLET NOZZLE
Figure 5.   Turbulent Contact Absorber for SO.
             and  Fly Ash Collection  (UOP)
      DEMISTER

 DEMISTER SPRAY

     DRAIN POT

TURBULENT LAYER
    MARBLE BED
UNDER BED SPRAY
                      GAS OUTLET
                                         GAS INLET
        Figure  6.   Marble-Bed Absorber
                        151

-------
GAS OUTLET
HOT GAS
 INLET
                                        DEMISTER
                                        DEMISTER
                                        WASH SPRAY
                                        SPRAY
                                        SPRAY
             Figure 7.  Spray Coltimn
                     152

-------
        It has been tested by TVA,  Peabody Engineering/Detroit
        Edison,  and  Ontario Hydro, The spray column is being
        tested at the EPA Test Facility, Paducah, Kentucky.
        Results  also will be reported in a separate paper today
        (Ret.  5).
    •   Tray Column.  The tray column offers high liquid hold-up
        and high SO  removal at relatively low pressure drop.
        The main disadvantage is the potential for scaling.  High
        liquid-to-gas ratios (L/G = 40) are required.  In addition,
        undersprays are used to wash off soft scale.

    •   Cross Flow Absorber.  The cross flow absorber has a
        short gas path with the scrubber installed in a horizontal
        position.  It  has a  low pressure drop and has been tested
        with  packing or sprays.  It requires  a higher L/G ratio
        to obtain high SO2  removal.
    •   Screen or  Grid Scrubber.  The screen or grid scrubber
        has been recently tested by TVA.  It contains five to
        ten screens (7/8-inch openings).  A liquid-to-gas ratio
        of 50 and a stoichiometric ratio of 1. 5 give SO  removal
        of 75 to  80 percent.  Low pressure drops were observed.


3.2.3   Scrubber Size,  Turndown,  and Spare Capacity
To keep the SO  removal system investment as low as possible,
              2            3
scrubber sizes of 500, 000 ft /min have been developed.  This is

required because the flue gas from a power plant is high — 3, 000

ACFM/MW.  An additional design consideration is scrubber turn-

down,  due to changes in boiler load.  Some scrubbers can be turned

down to 50 percent of design and others require compartmentalization.


In order to insure greater system reliability, spare capacity should

be installed.  Thus, a 500-MW unit could be designed with four

modules (no spare) or five modules to insure backup capacity and

system reliability.  An economic analysis of a five module 500-MW

unit will be presented later in this paper.


                               153

-------
3.2.4    Mist Separators


Mist separators come in various designs,  but are typically of open
construction and low-flow resistance,  similar to cooling tower packing.

The techniques under development to control mist eliminator plugging

employ such principles as:
     VELOCITY
    REDUCTION IN
      TCA
                VS,
                         Mini
                      VELOCITY
                     REDUCTION IN
                      VENTURI
                                         Reduction in upward
                                         gas velocity to 8 feet
                                         per second or less, at
                                         the same time attempt-
                                         ing not to disturb the
                                         uniformity of gas
                                         distribution
WASH
ROW
RATE
n   n  n
      TIME
           vs.
MASH
FLOW
RATE
nnnnn
                          TIME
More frequent back-
washing, "which may
upset water balance
in closed operations
    IU
                  vs.
                      A A A
                         Y/////////////S
                           Y YY
                       More efficient back-
                       washing, for example
                       employment of a tra-
                       veling retractable
                       washer  similar to a
                       conventional soot
                       blower,  in place of
                       stationary sprays,  or
                       washing from both
                       sides
                               154

-------
                 vs.
   THREE - PASS
                    SINGLE - PASS
                       Simplification of the
                       mist separator design
                       itself with possible
                       sacrifice of entrain-
                       ment removal
                       efficiency
tttt
         WATER.
               VS.
                MM
rywvwyyx:
  n  I
                                             Installation of an inter-
                                             mediate wash tray to
                                             lower the slurry solids
                                             entering the separator
DRAIN   FLOW
  j   fttt

                VS.
                             •ROW
                         1
                        DRAIN
                                        Reorientation of the
                                        separator to "cross-
                                        flow" conditions with
                                        the gas flowing  hori-
                                        zontally and the sepa-
                                        rators draining verti-
                                        cally at 90  rather
                                        than countercurrent
                                        at 180°
                vs.
                  '.'.'i
                                    DRAIN
                                             Wet precipitation
                                             with a variety of
                                             configurations
                               155

-------
It seems likely that a satisfactory answer to the problem of mist
separator plugging will soon be found, and that it will consist of
one or more of these measures.

3.2.5    Stack Gas Reheat

As the gas passes through a wet absorber, it will be humidified and
cooled to the adiabatic saturation temperature (normally 120 to 130 F).
Some water droplets may be present since the eliminator does  not
remove all of the mist formed in the scrubber.  If the  scrubber exit
gas is not reheated, it will lose some of its buoyancy as a result
of cooling.  Thus, the effective stack height and the plume dispersion
will be reduced.  This should not be  serious if a high degree of SO
removal is achieved.  However, due to the low efficiency of nitrogen
oxide removal, the ground level concentration of NO  may become
a serious problem, especially for existing plants where the  stack
height is already fixed.  For new plants, this effect must be taken
into consideration in  stack design.

Humidification of the stack gas  is  also objectionable  because condensa-
tion may occur and cause formation of a visible plume.  This acidic
plume may also create a corrosive environment around the  stack area,
although the  stack itself is corrosion resistant.

The obvious  solution is to reheat the scrubber exit gas.  There is general
agreement that the saturated and cooled gas should be  reheated, but
there is no agreement on the degree  of reheating necessary or  desirable.
The consensus seems to be that it is not necessary to reheat it to the
                               156

-------
original temperature (too expensive),  but that it should be heated at
least to a high enough temperature to minimize the water-vapor plume
under most atmospheric conditions.  This, of course, depends on the
local meteorological  conditions,  but 50 F  is the most commonly used
reheat design figure at present.

The  methods of reheating the stack gas may be  divided into direct
heating and indirect heating.

Direct Heating.  The  major advantage of direct heating is operational
reliability, because there is no heat transfer surface on which fouling
can occur.  Other advantages are low investment,  flexibility in degree
of reheat in some  cases,  low pressure drop, and low maintenance.
There  are several methods by which  direct heating can be accomplished.
         Direct-Fired Reh^f^i.   Either natural gas or low- sulfur
         fuel oil may be used,  depending on availability of fuel
         and cost.  Natural gas is the most expensive fuel, but it
         is convenient and clean.  Oil would be less expensive,
         but receiving, storing, and handling would be additional
         problems.  Complete combustion of oil in contact with
         wet gas may also be a problem.  The addition of sulfur
         dioxide and ash to the treated gas  should be considered.
         Flue Gas Bypassing.  Bypassing the scrubber with
         part of the gas stream — followed by mixing this gas with
         the scrubber exit gas — requires minimum investment
         and adds  essentially no operating cost.  However, this
         is possible only when the overall SO2 removal require-
         ment is low  enough to allow such bypass,  i. e. ,  50 to 60-
         percent removal or less.  This may be possible for low-
         sulfur coal-fired boilers when  the scrubber is preceded
         by a precipitator,  but it is impossible for high-sulfur
         coal application.
                                157

-------
         Hot Air Injection.  If the air preheater is designed to
         produce excess hot air over the amount required for normal
         boiler consumption, this hot air may be used to mix with the
         scrubber exit gas to obtain the degree of  reheat required.
         In doing so, the heat efficiency of the boiler will,  of course,
         be lowered somewhat.   Alternately, air can be preheated
         in a separate exchanger (with steam for instance), then
         mixed with the scrubber exit gas.  This alternative gives
         the clean, nonfouling service expected, but it consumes more
         steam, because of sensible heat losses in the air,  compared
         to direct-steam-to-stack-gas exchange.  The added cost might
         be justified on the basis of reliability.
Indirect Heating,  Indirect heating of the stack gas requires an exchanger

to transfer heat from the heating medium to the gas.  The advantage of

this method is usually low operating cost,  especially when heat from the

inlet gas is transferred to the scrubber exit gas (recuperative  heat ex-

change).  The main disadvantages of this method are a higher gas pres-

sure drop, higher investment,  and possible fouling on heat transfer sur-
faces.   If this method is chosen, finned tubes should be avoided, and de-

vices to keep the exchanger surface clean (such as a soot blower) may

be required.   The following are alternative indirect  reheat methods:
         Reheat with Steam. The  exit gas may be heated with
         steam from the turbine cycle in a heat exchanger at the
         scrubber outlet.  This method would  require additional
         coal firing in the boiler to generate the extra  steam and
         modification of the turbine to allow higher than normal
         extraction rates.   In a new plant, a system could be de-
         signed  to provide the steam required.  The additional
         coal firing would still be  more  economical than direct-
         gas or  oil-fired reheater.
         Recuperative Reheat System. A heat exchanger may be
         used for direct transfer of heat from the scrubber inlet
         gas to the exit gas. With this method, heat that would
                               158

-------
         be wasted is recovered.  The disadvantages are the
         large heat exchanger required (because of low-transfer
         coefficient and low-temperature differential), high pres-
         sure drop, and possibility of fouling.   Corrosion would
         also be a problem when the inlet gas is cooled to below
         the acid dewpoint. A cyclic-liquid heat exchange sys-
         tem with heat transfer  from the inlet gas  to water and
         from the water to the scrubber exit gas would permit
         use  of smaller exchangers than those required for  gas-
         gas  exchange,  and the  smaller  surface would reduce
         pressure drop.  Other  disadvantages remain, however.
The choice of a reheating method is not entirely an economic

consideration; the system reliability must be considered as well.


3.2.6    System Power Losses


Wet scrubbing systems have significant energy requirements in

terms  of pumping and fan losses.  When particulatc removal is not

required, the energy requirement is reduced substantially.   Figures

8 and 9 show the  order of magnitude to expect for fan and pumping

losses  from these systems.  To these must be added an allowance
for discharge pumps,  small auxiliaries, lighting,  and instrumentation.


Certain assumptions  are incorporated in these figures,  and the user

is cautioned to consider results obtained from them as preliminary  esti-

mates  only.  The assumptions made in their construction are as follows:


    •    Compressibility  effects for the fan are  negligible.

    •    Fan efficiency is 60 percent.

    •    Motor efficiencies are  93  percent.
                                159

-------
u

§
«/>
O
at
O
13
1
    5 —
    4 —
3—
    1—
                              ' VENTURI FOLLOWED BY SPRAY TOWER
                'TCA

            ' SPRAY TOWER
                I
               10
                             30
 I
40
50
 I
60
                      PRESSURE DROP,INCHES OF WATER
       Figure 8.  Station Electrical Loss as a Function
                 of Draft Requirements
      Pump efficiencies are 85 percent.

      Scrubbing liquor solids are 10 percent,

      Liquid flow is turbulent.  The sum of potential and friction
      losses is 80 velocity-heads.  (This is  typical of 80-
      foot elevation,  plus pipe and fitting losses, with an
      allowance for control-throttling losses. )

      There is single-stage absorption.

      Flue gas generation was assumed  to be 3000 actual
      cubic feet per minute per megawatt.
                             160

-------
   5—1
£  4—
o
O£
O
u.
o
I/)
(/)
2
o
z


i
1
20
1
40
1
60
1
80
1 1
100
1
120
1 1
140
1 1
160
                             l/G RATIO,GAl/1000 ACFM (INLET)




               Figure 9.  Station Electrical Loss as a Function of

                          L/G Ratio and Nozzle Pressure




       3. 2. 7   Materials  of Construction




       Any part of a wet scrubbing system that is in contact with wet SO  gas
                                                                      &

       or acidic scrubbing liquor should be constructed of acid-resistant mate-


       rial.  If solids are  also present in the system, abrasion-resistant mate-


       rial should be used.  The material specified should be able to withstand


       the highest temperatures encountered during normal operation and up-


       set conditions.
                                      161

-------
Stainless  steel 316L is sometimes used for scrubber construction.
However, if chloride is present, stress corrosion will be a problem.
Chloride stress corrosion is  frequently characterized by so massive
a failure that the affected metallic part must be totally replaced.  A
common form of prevention is to provide an upgraded  alloy (e. g. , Alloy
20, Incoloy, or Inconel),  but  an alloy demands a premium cost consistent
with its higher nickel content.

Elastomer-lined (soft rubber or neoprene) carbon steel can be effective
under abrasive conditions and temperatures up to about 175 F.  The
temperature must be well controlled to protect the  lining.  Field rubber
lining costs roughly four dollars per square foot.  A further advantage
of these linings is the lower maintenance costs in descaling.  The
scale-substrate bond is more  readily broken than in the  case  of
stainless  steel or alloy metals.

Rubber-linings applied in the shop have relatively strong vulcanized
bonds because  they are autoclave  cured.   Field  linings employ chemical
curing agents  rather than heat, and their quality is not always compara-
ble.  Ambient temperatures must not fall below 70  F during the curing
period.  Vessel components with short-axis dimensions  exceeding 10
to 12 feet  cannot be shipped by rail, hence, they must  be field-lined.

A serious problem area inside scrubbers is the zone where hot gases
first contact the liquor medium,  because temperature control is  some-
times difficult.  An acid-proof gunite is now  being offered by Pennwalt
for protection of these areas.  The gunite is  bonded directly to the steel
with thickness  from 0. 5 to 1. 5 inches. The gunite will withstand tem-
                                162

-------
peratures up to 750  F,and the applied cost is $4 to $10 per square foot.
Gunite can also be used for duct and stack lining.  In using gunite, the
added weights and stresses must be considered.

Another lining material of promise is glass-flake reinforced-polyester
resin,  such as the Ceilcote 100 and 200 series and Carboglass 1601.
This material is trowel-applied at 40 to 80 mils thickness on carbon
steel and can withstand temperatures up to 200 F under conditions of
continuous operation.  The applied cost is $3 to $4 per square foot.  Its
abrasion resistance improves  when  well wetted, and it can be field-
patched with relative ease, compared to elastomeric linings.

3. 3     WASTE DISPOSAL

The nature of calcium sulfite and  calcium sulfate solids leaving the sys-
tem has created a problem that remains incompletely defined.   Judging
from the experience at Mitsui  Aluminum and at Will County Station
No.  1, there seems  to be some indication that the solids from limestone
systems may not settle as completely in the pond as those from lime
systems.

The limestone solids at Will County have been observed to settle only
to 35 percent.  Under such conditions  they have no bearing strength,
but behave like quicksand.  This not only increases the acreage require-
ments, but it also creates a possible hazard to stray animals and
wildlife.
                                163

-------
One possible solution is to install supplemental mechanical dewatering
equipment.  Another is to use lime in place of limestone,  although at
greater cost.  Also,  the use of coagulants to improve settling is a possi-
bility.  Furthermore, it is not clear whether the lack of consolidation
originates from the use of calcium  carbonate (as opposed to calcium hy-
droxide),  or whether it can be primarily attributed to excessive size
reduction of the limestone feed.  Until this is  established, limestone
systems should have built-in turndown on their size reduction circuits.
The determining factor in size reduction is the classifier circuit, com-
prised of liquid  cyclones.  These return the oversized material to
the ball mill,  but reject the material of requisite fineness so that it
passes on to the process.

It has been found very difficult to reliably simulate the properties of
these solids on a laboratory scale.  The reasons for this  appear re-
lated to ionic strength — i. e. ,  the slow buildup of minor soluble salts
in the circulation slurry that alters a  solubility relationship of its
principal  constituents.  Invariably laboratory  slurries turn out to be
faster settling than encountered at full scale.  Further detailed studies
based on pilot operation are required,

The waste sludge produced in the lime/limestone scrubbing process
requires special study for proper management. The sludge consists
primarily of CaSO3 • 1/2H2O, CaSO4 • 2H2O,  CaCO3 or  Ca(OH), and
fly ash.  The quantities are significant, 500,000 tons/yr,  50-percent
solids for a 500 MW power plant burning 3-percent sulfur  coal.  The
principal  method for most of the 20 full-size installations  is disposal
in a pond  located on the power plant site.  However,  there are many
plants that have insufficient land and therefore must condition the
solids for disposal some distance from the plant.
                                164

-------
Suitable areas that are ecologically sound must be found.  Studies
should be carried out to determine the feasibility of disposing of
these solids in the mines where the coal was produced.  The unit
trains that transport the coal to the power plant could logically carry
the sludge back to the  mines.  The sludge must be dewatered so  that
it can be transferred easily to and from the trains.  Studies  shoxild
bo funded to evaluate this concept.  Other sites that should be studied
are old limestone quarries and salt domes.   These should be accept-
able from an ecology point of view.  The large disposal  sites might
be operated by state agency and a fee be charged for disposal.

Another important disposal technique used where land is not available
at the plant site involves dewatering the sludge and transporting  it by
truck or  barge to a suitable landfill.   Further studies are necessary
to better understand this technique.

Chemical fixation processes are also being developed.   These pro-
cesses generally involve pozzolanic  (cementitious)  chemical reactions
requiring the presence of lime.  Such reactions  lead to the formation
of dry solids with potential landfill and reclamation applications. This
technique requires additional study to define  optimum conditions.

Lirnitr.-d research on the use of fly ash and limestone-modified fly ash
has boon performed.   Additional research is required to explorr llu>
utilization of lirmj/limestone  sludge to perhaps 25 percent of tin- quan-
tity produced.  In Japan, the  sulfite is usually oxidi/.ed to  sulfate and
the  j/ypsum produced is  used  for construction.  This potential markcl
should be further explored in the UnitedStates.
                               165

-------
The Coal Research Bureau of West Virginia University has been per-
forming research under partial support from the Environmental Pro-
tection Agency to develop and evaluate utilization of such a solid-waste
fly-ash mixture.  As a result of this research,  several possible areas
of utilization have emerged.  These include production  of autoclaved
materials such as calcium-silicate brick or block,  aerated or foamed
concrete, and cement materials.

Based on the above comments, it is evident that further study is criti-
cally needed to define a sound plan of solid waste management.
                                166

-------
                             Section 4
             STATUS OF FULL-SIZE INSTALLATIONS

Although process development for SO  removal using lime /limestone
scrubbing continues at  an intensive rate, the U.S. industry has decided
to install full-size units to meet air quality regulations and to accele-
rate technology.  To date, there are 21 commercial-size installations
using lime or limestone scrubbing (see Table 1).  There are also four
scrubber systems for particulate removal  only (not listed in  Table 1)
that have contributed to the state of the art.  Table  1 gives the opera-
tional status and design criteria of 21 installations; 11 units were retro-
fits and 10 installations are planned for new power stations.  The larg-
est facilities are the Mansfield plant of Ohio Edison (two 900 MW -
3 percent S coal),  and the Navajo plant of Salt River Project (three
750 MW - 0. 5 to 0. 8 percent S coal).  These units are designed for alkali
addition into the scrubber circuits.

Other points that should be  emphasized are:

    •   Units Operating —  8
        —   Alkali  furnace injection — 3
        —   Alkali  into scrubber circuit — 5
             Total MW installed - 1900
                                167

-------
    •    Units being engineered and constructed— 12


         —    Alkali furnace injection— 0


         —    Alkali into scrubber circuit— 12


         -    Total MW to be installed - 9600





In 1968 and 1969 at Union Electric's  Meramec Station and Kansas Power


& Light's Lawrence Station, the first full-size units were installed.


Both employed finely pulverized limestone injected into the boiler fur-


nace followed by wet scrubbing utilizing a marble bed scrubber.


Operating experience revealed a number of problems that included


scaling and corrosion.  A typical flow diagram for this process ap-


proach is shown in Figure 10.   The system at Meramec has been aban-


doned, and the Lawrence Unit has been  substantially modified.  A


second unit at Lawrence Station was  installed and started up in


December  1971.  Reported  SO  removal is 70 percent at a  stoichio-
                             (L

metry of 0. 7.





The first full-size limestone scrubbing  system in this country utilizing


alkali addition into the  scrubber liquor circuit was installed as a retro-


fit unit at the Will County Station of Commonwealth Edison.  The unit


involved is a 180-MW cyclone  boiler with a two-stage venturi/absorber


to remove  fly ash, followed by SO .   More has probably been learned


from  this installation about practical operating problems with lime-


stone during its one year of operation than from  any other pilot plant to


date.   The major problems experienced are  maintaining  clean de-


misters, system reliability due to mechanical problems,  and disposal


of waste solids.  SO  removal efficiencies reported were 75 to 85  per-
                    £•

cent.   A typical flow diagram for this process is shown in Figure 11.


Table 2 presents preliminary test data for 1972 (Ref. 7).
                                 168

-------
en
vo
                  Table 1.  OPERATING AND PLANNED  FULL-SIZE  LIME/LIMESTONE DESULFURIZATION

                                             FACILITIES IN THE  UNITED STATES
Utility Company/Plant
1. Union. Electric Co. 
9. Kansas City Power & Light/
La Cygne Station
10. Arizona Public Service Co. /
Cholla Station
11. Duquesne Light Co. (Pittsburgh)/
Phillips Station
12. Detroit Edison Co. f
St. Clair Station No. 6
13. Ohio Edison/Mansfield Station

14. Tennessee Valley Authority/
Widow's Creek Station No. S
15. The Montana Po-wer Co. /
Colstrip Units 1 & 2
16. Northern States Power Co.
(Minnesota)/
Sherburne County Station
No. 1 & Z
17. Northern Indiana Public Service
Co./
Kankakee 14
18. Mohave-Navajo/
Mohave Module (Vertical*)
19. Mohave-Navajo/
Mohave Module (Horizontal?)
20. Salt River Project/
Navajo Station
21. Southern California Edison/
Mohave Station (Horiz/Vert)
Absorbent
CaO

CaO

CaO

CaCO,
3
CaCO
CaO

CaO

Ca
-------
                                        S1ACK
                                               1.0, FAN
                                                   - STACK (.AS RtHIAItfi

                                                        DRAI!>i POTS
                                 RFCYCU
                                                                MAKE UP 1VAUH
                                                                 TO ASH
                                                               DISPOSAL POkD
Figure 10.
Particulate and SC>2  Removal System Employing
Marble  Bed with Limestone Calcining in the  Boiler
(Source:  Combustion Engineering)
                               170

-------
Table 2.  WILL COUNTY UNIT 1 WET SCRUBBER TEST DATA ("A" SCRUBBER, MAY 1972}
Test Number
Date
Load, MW
Gas Flow, 103 CFM
Scrubber System, Pressure
Difference, Inches HO
Dust Inlet, gr/SCFD
Dust Outlet, gr/SCFD
SO_ Inlet, ppm
SO Outlet, ppm
SO Removal Efficiency, %
Absorber Slurry Density, %
Absorber pH
1
5-18
113
335
24.5
—
.0232
1145
67
94
3.4
6.5
2
5-18
113
335
29
.0944
.0079
1140
75
93
5.2
6.3
3
5-19
114
335
21
. 1440
.0073
890
294
67
5.5
7.4
4
5-19
115
340
25
.1470
.0298
930
35
96
5.2
6.3
5
5-20
111
335
24
. 1105
.0261
1130
285
75
2.5
5.7
6
5-20
112
320
25.5
. 1790
.0255
1000
118
88
4.3
5.8
7
5-21
113
315
22.5
—
—
640
18
97
5.0
7.2
8
5-21
115
310
22.0
-
-
910
45
95
-
5.7
9
5-22
110
315
23.2
.3060
.0205
1000
223
81
2.9
5.9
10
5-22
111
335
23.0
.2580
.0334
545
180
67
2.2
5.4
11
5-23
205
16.0
—
—
1200
45
96
-
6. 1
12
5-23
58
215
18.0
-
-
1150
50
96
1.5
6.1

-------
     ""'k^-il
                fU  e
                                                       LIMESTONE
                                               MAKE-UP  |  | IUHKU
                                                WATER
       I.D. IOOSTE«
       y, '*"
          STEAM
           RIHEATER
                'TO SETTUH6
                  POND
Figure  11
   g
Will  County Station Unit No.  1   (Limestone Added
to Scrubber - Venturi plus TCA Absorber)
                    172

-------
The first scrubbing system utiliaing lime introduced into the  scrubber
circuit has been installed at the Phillips Station of Duquesne Power
and Light.  It is presently going through startup and should supply
valuable information when in full operation later in 1973,  This unit
is reported to be designed based on the Mitsui Aluminum unit in Japan,
which uses a calcium hydroxide scrubber  (Ref. 8).   To date the Japanese
system is the most successful operating unit based on a throwaway pro-
cesa.  It has been operating since March 29, 1972 without any  significant
downtime.  SO- removal efficiencies have been reported to be 80 to 90
percent. A typical flow diagram for this process is shown in Figure 12.

The Hawthorne Units 3 and 4 of Kansas City Power & Light are pres-
ently in the first stages of operation.  One uses alkali injection into
the boiler furnace and the  second has been modified to inject lime
into the scrubber.

The Paddyrs  Run Station of Louisville Gas & Electric is the first plant
to test carbide  sludge  [Ca(OH)2].  Test results based on a relatively
short period  of operation indicate high removal of SO
                                                   £t

Kansas City  Power fc Light's  LaCygne plant is presently in startup
and should  be in full operation during the next  60 days.

When most of the 20 units are operating,  significant progress toward
commercial demonstration will be realized.
                              173

-------
      I.D. FAN
FLUE GAS
 BOILER

I

RE
FUII
                                DEMISTERS
                         —1  ; n
                           Ui:  !
                     MAKEUP SLURRY
                         TANK
                                                 r:   :z:._
j MAKEUP SLURRY
   FEED PUMF
                                               BLEED SLURRY
                                              TRANSFER PUMP
               ~i
     DRY CARBIDE PIT
	 RECYCLE SLURRY
    	 MAKEUP SLURRY
	 BLEED SLURRY
      _ RETURN LIQUOR
ASH POND LIQUOR
 RETURN PUMP

        \
                                                                   WASTE DISPOSAL POND
           Figure 12.  Chemico /Mitsui Flue Gas Desulfurization
                        System  (Lime  Added to Scrubber)
                                       174

-------
                            Section 5



    LIME/LIMESTONE COMMERCIAL OPERATING CONDITIONS

        OFFERED BY SO2 REMOVAL SYSTEM SUPPLIERS
Suppliers for SO_ removal systems are growing steadily.  A review
                £

of the market reveals seven  suppliers with full-scale experience.


There are five  other companies that have just recently entered the


market.





The  seven companies with full-scale  experience were contacted to ob-


tain  design information concerning SO  removal guarantees  related to
                                    L*

percent sulfur fuel,  scrubber type, liquid-to-gas ratio, stoichio-


metry, percent solids in  slurry, alkali type,  hold tank residence time,


and any qualifications concerning system design. The results of this


study are presented in Table  3.





It is interesting to note most companies will supply systems using lime


or limestone as the alkali and provide SO   removal guarantees that
                                        £*

vary from 70 to 90 percent,  respectively, or as required  to meet per-


formance standards.   Important operating parameters such as liquid-to-


gas  ratio and stoichiometry have changed in design range.  Liquid-to-


gas  ratio has increased from 20 to as high as  100, the average being 60.


Stoichiometry has decreased from 1.75  to as low as 1.0 based on SO


absorbed, the average being  1.2.  There appears little restriction on


the percent  sulfur in the fuel as  it relates to guarantees or size of units.
                               175

-------
Mc,dui<; designs, as largo as ISO megawatts each,  for units up to 800
megawatts,  are provided by U.S. suppliers.  All process designs are
based on alkali addition to the  scrubber circuit.

When schedule was discussed, most suppliers stated a minimum time
schedule of 24 months from purchase order to  startup.  A preferred
time period would be 30 months, depending on backlog.  The time
required for a bidder to respond to  a functional specification ranges
from 4 to 10 weeks, with 8 weeks being typical.  An additional period
of 8 weeks is required to review proposals, select supplier, and con-
duct preaward negotiations. Based on the  above requirements, the
total time period is approximately 36 months.  In addition, pilot
plant testing to evaluate limestone reactivity is desirable.
                               176

-------
                           Table 3.
  OPERATING  CONDITIONS  OFFERED  BY S02  REMOVAL SYSTEM SUPPLIERS
Supplier
Chemical
Construction Co.
Combustion
Engineering

Babcock & Wilcox


Peabody
Engineering
UOP - Air Correction

Combustion
Equipment Assoc.



Research Cottrell


Scrubber
Venturi 2 stage
spray c olumn
1 or 2 marble beds1*1
(bottom spray only)

-Low pressure drop
quencher plu« tray
absorber
Venturi plus
spray column
Turbulent Contact
Absorber (3 stages)

Venturi (if particulate
included) (2 stage)
spray column
(S02 only)
Multicontact
absorber
Banco systenv'
Alkali
Lime
Limestone
Lime
Limestone
Carbide sludf
Lime
Limestone
Carbide sludj
Lime
Limestone
Lime
Limestone

Lime

Limestone

Limestone

Lime
Liquid to Gas
Ratio, gal/MCF
40-80
80
25-30
25-30
e 25-30
40-50
40-50
e 40-50
(b)
(c)
40
40

60

80


ffl
90lfl
Full Size
Yes




Yea


Yes

Yes


Yes

Yes

Yes


a. Furnace injection of akali not offered
   unless requested.
b. L/C = 20 for venturi,  50-60 for spray
   column with lime.

c. L/C * 80 for »pray columnwth
   limestone.
d. L/C (1 st stage) = 50-60; L/C (2nd
   stage) = 15-30.
t.  Research Cottrell has no restriction on system design or guarantee
   for units using fuel with 2 percent or less sulfur.  For fuel with
   greater than 2 percent sulfur,  they require a 100-MW demonstra-
   tion unit before building a full-size (i.e.,  800 MW( installation.

f,  Bahco is limited to 100,000 CFM, or 35 MW, and offers lime-
   stone systems.

-------
                              Section 6
          CAPITAL AND OPERATING COSTS OF FULL-SIZE

                         SO  REMOVAL SYSTEMS
                           LJ
 In order to develop capital and operating costs,  Bechtel cost estimates


 were used as a basis.  The basis for design of lime/Limestone scrub-


 bing systems was a "typical" new 520-megawatt boiler burning high


 sulfur coal (3 percent S).  SO  removal required is approximately 80
                             £r

 percent.  Particulate removal proceeded the SO  removal.  A spare
                                              LJ

 train was provided for system reliability, and a pond is available for


 disposal of the waste solids.





 It has  generally been the practice to cite capital costs on a "per kw"


 basis, whereas a "per CFM" cost basis is more meaningful across the


 range  of combustion conditions encountered.  Economics of scale pro-


 vide unit savings for scrubbers up to 500,000 CFM.   Above this size,


 duplication of scrubber trains is necessary,  and cost savings cease


 to apply.





A common reason for high "per kw" cost is  the installation of a complete


 scrubber train with isolation dampening so as to make on-stream main-


 tenance.  For a five-train system,  this is equivalent to 25 percent of


 installed cost. This  practice is  not always necessary, being depen-


 dent on utility load balancing  as well as on reliability and proven
                                178

-------
maintenance record for the particular system employed.  Other rea-
sons for "high" costs per kw are due to high contingencies for system
modifications and high escalation rates.

It should be noted that no costs were shown for real estate deeded over
to sludge ponding.  Instead, a flat charge of $2 per ton of wet sludge
(40-percent solids) was allowed.  A slight advantage was given to lime
sludges in that it was assumed the solids would settle to 50-percent
solids.  In addition, it should be emphasized that disposal costs are
highly site-sensitive.  If the sludge must be treated for disposal off
site,  the charge per ton could double to $4. 00 and thus increase opera-
ting costs by 15 percent.

Capital  costs have been divided into direct costs and indirect costs.
Direct costs include yard facilities, raw material handling,  scrubber
system, and sludge handling.   Indirect costs  include field costs, engi-
neering, home office,  escalation, and fee.  Estimated capital cost for
a "typical" new 520-megawatt unit using limestone scrubbing is
$23, 400, 000 or $45 per kw.  The total installed cost is  $ 1 0. 70 per
ACFM.   The estimated cost per kwh is  2.26 mills, based on limestone
at $5. 00 per ton delivered (stoichiometry at 1. 5 and a sludge charge
of $2. 00 per ton). If limestone can be purchased at $3. 00 per ton, the
annual cost  would be 2. 15 mills/kwh.   Based on 10,000 Btu/lb coal,
the cost per million Btu is estimated to be 23£,  If sludge disposal cost  is
$4.00 per ton,  the annual operating cost would be  26£ per million Btu.
Tables 4 and 5 summarize these costs and give the basis for the estimate.

The estimated  capital cost for a "new" 520-megawatt lime scrubbing
system  is $19,600,000 or $38/kw.  This is equivalent to $9/ACFM.
                               179

-------
The estimated cost per kwh is 2. 16 mills based on lime at $20 per ton


delivered.  Sludge disposal costs  are reduced due to less sludge pro-


duced because of lower stoichiometry and better settling of solids to


50 percent by weight.  Annual cost for SO  removal is estimated to be


    per million Btu.  Tables  6 and 7 summarize these costs
Retrofit costs for SO  removal are greatly affected by specific plant
                    C*

conditions.  Therefore,  no effort has been made to present costs for


retrofit units.   More detailed information on cost considerations can


be found in a paper presented by Burchard (Ref. 9).
                              Table 4
           CAPITAL COST - LIMESTONE SCRUBBING
                                                     (a)
Cost Item
Yard facilities
Raw material handling
Scrubber system
Sludge handling (on- site)
Direct costs, subtotal
Field costs, engineering
and home office, escala-
tion, contingency, and fee
Total installed
Thousands of
Dollars
648
2, 347
7, 427
755
11, 177
12,253
23,430
Dollars/kw





45.05
Dollars/ AC FM
Installed





10.70
 a.
520-MW gross output; 3. 1-percent S coal; 81-percent SO2 removal;

particulate matter removed by others; 80-percent load factor; with

ponding; spare train
                                180

-------
                                         Table  5
       ANNUAL  OPERATING COST - LIMESTONE  SCRUBBING
                                                                                  (a)


Coat Item
Basic alkali, tons
Process -water, gal x 10
Power, demand, kw
Power, energy, kwh x 10
Steam, demand, annual Btu x 10
Steam, energy, Btu x 10
Operating labor, 3/shift
Operating labor overhead at 75%
Maintenance at 5% of direct coat
Maintenance overhead at 30%
Capital recovery, 15 yr.
7-1/2% f. 1315 multiplier)
Insurance and property taxes
at 1.9%
Sludge disposal, tons
Total Cost

Total
Requirement
205,000
461,000
7,940
53,400
800,000
615,000
Z5.200







659,000

Range of
Unit Cost
Experience
$ 3.00-9.00
$ 0.02-0. 15
$15.00-40. 00
$ 0.90-5.00
$ 0.08-0.25
$ 0.10-0.55
-







$ 0.75-5.00


Unit Cost,
This Study
$ 5.00
$ 0.03
$35.00
$ 5.00
$ 0.20
$ 0.55
$ 6.25







$2.oo(b)

Annual
Cost,
$Thousand
1,025
9
278
267
160
338
153
118
557
167

3, 081

444
1,318
7,920
Cost
per kwh,
mills
0.293
0.003
0.078
0.076
0.046
0.097
0.045
0.034
0.159
0.048

0.881

0.127
0,377
2.26
a.  520-MWgroes output; 3. 1-percent S coal; 81-percent SO  removal; particulate removal by others;
    80-percent load factor; with ponding; spare train
b.  Waste solids are assumed to be 40 percent  by weight dry solids.  At $4 per ton for sludge disposal,
    annual cost increases to $9,238,000 or 2.64 mills per kwh
                                            181

-------
                              Table 6
              CAPITAL COST - LIME SCRUBBING
                                                  (a)
Cost Item
Yard facilities
Raw material handling
Scrubber system
Sludge handling (on- site)
Direct costs, subtotal
Field costs, engineering and
home office, escalation,
contingency, and fee
Total installed
Thousands of
Dollars
555
592
7.427
755
9,329
10, 262
19,591
Dollar s/kw





37.68
Dollars/ AC FM
Installed





8.96
a.
  520-MW gross output; 3. 1-percent S coal; 81-percent SC>2 removal;
  particulate removal by others; 80-percent load factor; with ponding;
  spare train
                              182

-------
                                          Table  7
          ANNUAL OPERATING  COST - LIME  SCRUBBING
                                                                                (a)
Cost Item
Basic alkali, tons
Process water, gal x 10
Power, demand, kw
Power, energy, kwh x 10
Steam, demand, annual Btu x 10
6
Steam, energy, Btu x 10
Operating labor, 3 /shift
Operating labor overhead @ 75%
Maintenance @ 5% of direct cost
Maintenance overhead @ 30%
Capital recovery, 15 yr, 7-1/2%
(. 1315 multiplier)
Insurance and property taxes
at 1.9%
Sludge disposal, tons
Total Cost
Total
Requirement
94,000
1,754,000
1Z, 109
81,500
800,000
615,000
25,200
-
-
-

—

-
416,000

Range of
Unit Cost
Experience
$15-25
$.02-. 15
$15-40
$.90-5.00
$.08-. 25
$.10-. 55
-
-
-
-

—

—
$.75-5.00

Unit Cost
This Study
$20.00
$00. 02
$35.00
$ 5.00
$00.20
$00.55
$ 6.25
-
-
-

—

—
$2.00(b>

Annual
Cost,
$ Thousand
1,880
35
424
407
160
338
158
118
466
140

2,576

372
832
7.906
Cost
per kwh,
mills
. 537
.010
. 121
. 116
.046
.097
.045
. 034
. 133
.040

.736

. 106
.238
2.260
a, 520-MW gross output: 3. 1-percent S coal; 81-percent SO, removal; particulate removal by others;
   80-percent load factor; with ponding; spare train
b. Waste solids are assumed to be 50-percent by weight dry solids.  At  $4 per ton for sludge disposal,
   annual cost increases to $8,738, 000 or 2.5 mills per kwh
                                             183

-------
                               Section 7

                             CONCLUSIONS
1.     Commercial technology has advanced to the point where
       system suppliers offer guarantees of 80 to 90 percent
       removal for limestone  and lime, respectively.

2.     Process design has become more conservative with higher
       liquid-to-gas and lower stoichiometric ratios.  Both should
       give greater system reliability.

3.     Twenty-one full-size lime/limestone scrubbing units for
       SO   removal are being engineered, constructed,  or operated.

4.     Capital cost for new 500-megawatt SO^-removal limestone
       units burning 3 percent sulfur coal is $45 per kw.   Operating
       cost is  23 to 25 cents per million Btu (including capital).

5.     Capital cost for lime scrubbing units under the same con-
       ditions  as for the previous conclusion is  $38 per kw.
       Operating cost is 23 cents per million Btu.

6.     Two areas requiring further study and evaluation are
       reliability and solid waste disposal.
                               184

-------
                            Section 8
                          REFERENCES

1.    TVA Special Reports on Pilot Plant Operations, Aug 1971,
      Nov 1971
2.    P. Lowell et al, "A  Theoretical Description of the Limestone
      Scrubbing Process,"  (Jun 1971) Vol.  1.  Report No.
      PB 193-029, Clearinghouse for Technical Information,
      Va. ,  2251
3.    A.V. Slack et  al, "Sulfur  Oxide Removal From Waste Gases,"
      Journal of Air Pollution Control Association, Mar 1972
4.    R.A. Boll, "Mathematic Model for SO  Absorption by Lime-
                                          £
      stone Slurry,"  Limestone Scrubbing Symposium,  Perdido
      Bay,  Pensacola, Fla. , Mar 1970
5.    M. Epstein, "Test Results from EPA Lime/Lime stone Scrub-
      bing Test Facility, "  Flue  Gas Desulfurization Symposium,
      May  1973, New Orleans
6.    J. Jones et al,  "Waste Product From Throwaway Flue Gas
      Cleaning Processes,"  Flue Gas Desulfurization Symposium,
      May  1973, New Orleans
7.    D. C. Gifford,  "Will  County,  Unit  1 Limestone Wet Scrubber,"
      Proceedings Electrical World Conference - Sulfur in Utility
      Fuels,  Chicago, 111., Qct 1972
                               185

-------
8.    Interagency Report on Japanese Sulfur Oxide Technology,
      Aug 1972
9.    J. Burchard et al,  "General Economic Considerations
      of Flue Gas Scrubbing  for  Utilities," Proceedings Elec-
      tric World Conference— Sulfur in Utility Fuels,  Chicago,
      111., Oct 1972
                              186

-------
      WASTE PRODUCTS FROM THROWAWAY
       FLUE GAS CLEANING PROCESSES -
ECOLOGICALLY SOUND TREATMENT AND DISPOSAL
                      by

                Julian W. Jones
               Richard D. Stern
         Development Engineering Branch
           Control Systems Laboratory
        Office of Research and Monitoring
      U.S.  Environmental Protection Agency
         Research Triangle Park, N.  C.
                       187

-------
                          ACKNOWLEDGEMENTS






     In presenting this paper, the authors are indebted to numerous




organizations, including Chicago FlyAsh Co., Commonwealth Edison Co.,




Dravo Corporation, EPA National Environmental Research Centers in




Cincinnati, Ohio, and Corvallis, Oregon, EPA Office of Solid Waste




Management Programs, International Utilities Conversion Systems, and




the Tennessee Valley Authority.  The authors are also indebted to




several individuals, but especially to Mr. Frank Princiotta for his




helpful comments and direction, and to Mrs. Charlotte Bercegeay, for




her patience in typing the many "revised editions."

-------
                                NOTES


1.  The mention of company names or products is not to be considered


    as endorsement or recommendation for use by the U. S. Environmental


    Protection Agency.


2.  EPA policy is to express all measurements in Agency documents


    in metric units.  When implementing this practice results in


    difficulty in clarity, NERC-RTP provides conversion factors for


    the particular non-metric units used in the document.  For this


    paper these factors are:
          British


          1 acre


          1 acre-ft


          5/9 (°F-32)


          1 ft


          1 ft2


          1 ft3


          1 ft3/short ton




          lb/in.2


          1 mile2


          1 mile2-ft


          1 ton (short)
Metric
4047 meters'
1233.6192 meters3
0.3048 meter

             2
0.0929 meters


0.0283 meters


0.0312 cubic meters per

  metric ton


70.31 grams per centimeter2

               2
2.59 kilometers


0.7894 square kilometer-meter


0.9072 metric tons
                                 189

-------
                              ABSTRACT






      Extensive application of lime/limestone throwaway processes will



 create  large quantities of wet sludge.  The sludge solids consist



 primarily of flyash or oil soot plus sulfonated and unsulfonated calcium



 salts and contain many impurities  and trace elements associated with



 the raw materials involved.  In addition, neutralization of unmarketable



 abatement sulfuric acid can potentially create large quantities of



 similar sludge.  Based on current and projected utilization of lime/



 limestone throwaway processes, the sludge problem is quantified.



 Results of a preliminary assessment of current technical approaches



 and techniques for treatment and disposal of sludge are discussed,



 including potential for ground and surface water degradation.  An EPA



 program to determine ecologically and economically acceptable methods



 for disposal of lime/limestone sludge is described.  The program includes



 the physical and chemical characterization of sludge materials from:



 representative sorbent/fuel combinations; an evaluation of potential



 water pollution or other problems associated with disposal of both



untreated and treated (subject to dewatering, oxidation, chemical



 fixation, etc.)  sludge;  and a study of the economics of various




disposal/treatment combinations.   Preliminary data and analyses are



presented.
                                190

-------
1.0  INTRODUCTION



     A major problem inherent in any flue gas desulfurization system



is the necessity of disposing of or utilizing large quantities of a



sulfur product.  The sulfur compounds produced are either saleable



(sulfuric acid, sulfur dioxide, or elemental sulfur) or throwaway



(calcium-sulfur compound sludge materials).   The technology of



processes producing abatement sulfuric acid  is currently more advanced



than for those producing the other saleable  products.  To date, however,



most utilities have favored utilization of lime or limestone scrubbing



throwaway processes.



     Although research and development is currently underway in a



number of processes with saleable sulfur products, most will not be



commercially available until the late 1973-1975 period.  Allowing 2



years for retrofit to an existing plant and  5 years for a new installa-




tion including a control system, the earliest application of these



processes could be late 1975 and 1978, respectively.  It is significant



to note that of 23 full-size desulfurization control processes for



power plants in design, construction, or operation, 17 are lime/limestone



scrubbing throwaway systems.  This majority, undoubtedly based on a



higher degree of confidence in understanding, process characterization,



and current commercial availability, and the timing indicated above,



suggests that throwaway processes will comprise a very significant



percentage of flue gas cleaning installations in the near term at least




to 1980. (It should be noted that the double alkali process—soluble



alkali scrubbing, lime/limestone regeneration of soluble alkali—is




                               191

-------
currently regarded by many to be an attractive alternative to lime/




limestone slurry scrubbing.  However, this process results in a very




similar throwaway product.)  The majority of the remaining installa-




tions will very likely be comprised of processes producing sulfuric




acid since these processes appear to be further advanced than those




producing other saleable sulfur products.




     Extensive application of the throwaway processes will create




large quantities of wet sludge with solids consisting primarily of




flyash or oil soot plus CaS03 • 1/2 11,0, CaS04 • 2 IL,0, CaC03, and




possibly some Ca(OH)9 from lime scrubbing processes, and CaS03 • 1/2 H20>



CaS04 • 2 H-,0, and CaCO, from limestone scrubbing processes.  In addition,




neutralization of unmarketable abatement sulfuric acid can potentially




create large quantities of sludge consisting primarily of CaS04 • 2 l^O.




Typically, the water in equilibrium with these sludges will not only




contain varying amounts of these materials as dissolved solids in the




3,000-15,000 ppm range (primarily Ca**, S04=, Mg++, and S03=), but also




many impurities and trace elements found in the applicable raw materials




such as the alkali sorbent, the fuel material, and the process water.




     For a number of years the Environmental Protection Agency




has been engaged in research and developnent concerned with



utilization of lime/limestone-modified pulverized flyash resulting




from lime/limestone based desulfurization processes.  Acknowledging the




potential solid waste and water pollution problems associated with these




processes, the prime objective was that an air pollution problem should




not be transferred to these other areas.  A study of the utilization of




                              192

-------
        '   indicated that only about 7-10 percent of flyash is currently



utilized and that under current technology and economic conditions,



utilization would probably not exceed about 25 percent in the near terra.



Similar limiting conditions are expected to apply to sludge materials.



Additionally, analysis of dry-collected limestone-modified flyash



indicated the presence of potentially toxic elements.




     On the basis of these results, the rapid emergence of lime/limestone



wet scrubbing as the dominant desulfurization process, and a lack of



knowledge regarding potential heavy metal and toxic element involvement



in the scrubber chemistry, EPA reoriented its activities toward ecologi-



cally sound and "safe" treatment/disposal of the waste products from



these throwaway processes.  Techniques for treatment and disposition



may be applicable to sludges produced by neutralization processes.



     This paper quantifies the problem of lime/limestone throwaway




process generated sludge, examines and assesses potential treatment



and disposal alternatives, identifies problem areas, and presents the



EPA program, including results to date.
                               193

-------
2.0  QUANTIFICATION OF THE PROBLEM

     Typical quantities of throwaway sulfur products, potential saleable

products, and coal ash production for a typical 1000 Mw coal-fired boiler

are presented in Table 1.  From this table rough comparisons between the

production rate and storage volumes* for typical throwaway sulfur products

may be compared to typical saleable products and to coal ash, the normal

waste product for a coal-fueled plant.  The following are important

observations which can be made:

     1.  The production rates  (dry basis) of throwaway sulfur product

are approximately 45-80 percent larger (dry) than those of ash normally

produced; this leads to a total (sludge plus ash) throwaway requirement

about 2.5-3.0 times the normal coal ash disposal rate.  The ranges

reflect lime and limestone sludges, respectively.

     2.  Large storage volumes are required for ultimate disposition of

the sludges. For example, for a 1000 Mw coal-fired unit over a 20-year

lifetime, about 860-1100 acres** (1.3-1.7 square miles) of disposal

land would be required for a wet sludge (50 percent solids) ponded to

a 10-foot depth.  The ash alone would require about 250 acres (0.4

square miles) ponded to the same depth.  Thus, the volume required for

the throwaway sulfur product is approximately 275 percent greater than
 *It should be noted that rough estimates for solids packing specific
  volume were used to calculate potential storage volumes required for
  20 years of boiler operation; depending on the process, lime sludges
  can be allowed to settle in a storage pond to a maximum of about 50
  percent solids slurry, or can be dewatered to about 70 percent or
  more solids by various techniques.
**Factors for converting these and other English units to their metric
  equivalents may be found on page iii.
                               194

-------
    Coal  Ash
                                      Table 1.   TYPICAL QUANTITIES OF ASH AND SULFUR
                                 PRODUCTS FROM A 1000 MW COAL-FIRED BOILER CONTROLLED WITH
                              A LIME OR LIMESTONE SCRUBBING FLUE GAS DESULFURIZATION SYSTEM
                                Yearly Production, 1000 MIV
                               	(tons/yr)
                              Assumed Packing Volume
                              (ft3/ton of dry material)
                                 Approximate Volume
                                Required for Storage
                                1000 MW for 20 Years
                               	(acre-feet)	
  288,000  (dry)
                                                                           19
                                                                     (wet;  20% water)
                                        2510








to
Ul


Lime Sludge (dry)
73% CaS03-l/2H70
11% Ca(OH)2
11% CaS04-2H20
5% CaC03

Lime Sludge (wet;50% solids)
Limestone Sludge (dry)
58% CaS03-l/2H20
9% CaS04-2H20
33% CaC03


304,000
47,000
45,000
21,000
417,000
834,000

305,000
45,000
170,000
520,000
22 4210





45 8600
22 5225




   Limestone Sludge  (wet;
     50% solids)
1,040,000
          45
(assumed same as for wet
lime sludge)
10,800
Sulfuric Acid (95% cone)
Sulfur (dry)
243,000
75,600
17.6
23 (80% packing)
1960
800
Assumptions: Coal  - 3.5 percent S content,  12 percent  ash  content.
             Coal  Rate -  2,400,000  tons/yr  for  1000 Mw on  stream  for  6400 hours/yr  (0.75 Ib/Kwh).
             Lime  Sludge  - Based on performance of Chemico Ca(OH)2  scrubbing unit in Japan  (Mitsui Aluminum Co.)
               at  1.28 Ca(OH)2/S02  mole ratio and 85 percent  utilization.
             Limestone Sludge - Based on preliminary EPA/Bechtel/TVA  data from  the  TVA Shawnee Power Plant at
               1.65 CaC03/S02 mole  ratio and 85 percent utilization.
             Removals - 90 percent  of S07 in flue gas  converted to  sulfur ^r—luct,  100 percent of ash  collected.

-------
 for  ash normally produced and the total volume requirement is about
 4  times that for ash alone.  The ranges reflect lime and limestone
 sludges, respectively.
     3.  Potentially large quantities of sulfuric acid can be produced
 by certain flue gas desulfurization processes such as Catalytic
 Oxidation, Wellman-Lord, and magnesium oxide scrubbing.  Approximately
 24 million tons per year of concentrated sulfuric acid can be produced
 per  100,000 Mw of flue gas desulfurization capability.  This is close
 to the total U. S. sulfuric acid production rate, which was 29.3 million
 tons in 1971.  Additionally, the use of a sulfuric acid plant and
 neutralization as a control strategy in the smelting industry
 has  the potential for producing large quantities of sludge.  Assuming
 stoichiomctric quantities, for each ton of sulfuric acid neutralized,
 approximately 1.75 tons of dry sludge would result.
     4.  Elemental sulfur appears the most attractive product in terms
 of production rates and potential storage volume.  Only about 7b,000
 tons per year of sulfur would be produced per 1000 Mw of flue gas
 desulfurization; this leads to a storage area of about 80 acres CO.12
 square miles) over a 20-year lifetime,  assuming a 10-foot depth.
     The numbers of Table 1 become significant when one considers that
 in the foreseeable future, probably through 1980, U.  S. electric
utilities will likely continue the current pattern of ordering wet
 scrubbing systems; a majority of these  orders will probably be for wet
 lime/limestone processes  producing a throwaway sludge.   Forecasts based
on a government interagency Sulfur Oxide Control Technology Assessment
Panel (SOCTAP)  report^  indicate that  for coal:
                               196

-------
     1.  Over 20,000 Mw of generating capacity could be equipped with



S02 scrubbing systems by 1975, but the equipped capacity is estimated



to be closer to 10,000 Mw.




     2.  By 1977, the equipped capacity may be 48,000-80,000 Mw, but




the lower end of the range is considered a more realistic estimate;



i.e., approximately 50,000 Mw.




     3.  By 1980, at least 75 percent of the coal-fired capacity could




be equipped with stack gas scrubbers.  This is equivalent to about




144,000 Mw.  Assuming 75 percent of the total stack gas scrubbing




installations utilize a lime/limestone throwaway process results in



108,000 Mw controlled in this way.




     Known commitments for lime/limestone scrubbing systems indicate




control of approximately 1900 Mw in 1972, 6700 Mw by 1975, and 9000 Mw



by 1977.




     The above data were used to prepare Figure 1 which presents the




growth of coal-fired generating capacity and a comparison with known




committed and forecast utilization of lime/limestone scrubbing processes




for flue gas cleaning.  Although a function of plant size, sulfur




content, and other factors, the total waste sludge generated to 1980




was roughly estimated using the data from Table 1.  The results are




shown in Figure 2.  In addition, an estimate of limestone sludge production




rates and disposal area requirements as a function of plant size and coal




sulfur content was also determined.  These are shown in Figure 3.
                              197

-------
        CAPACITY, KNOWN CONTROL
               COMMITMENTS
        1972     1973    1974     1975     1976     1977    1978     1979     1980

                                      YEAR
Figuie 1. Comparison of estimated electric utility coal generation capacity with current
and forecast control by lime/limestone scrubbing.
                                   193

-------

-------
     1.4
     1.2


 ^S"»
 
 c:

,f   i.o
UJ
     0.8
 a

 I   0.6
     0.4
     0.2
                                 T
                              SULFUR CONTENT OF COAL, \  4.5
                                                                         2.5
                                                                         2.0
                                                                         1.5
                                                                              OJ
                                                                         1.0
                                                                              
-------
     The magnitude of the 1980 estimates of Figure 2 appears significant



from the following comparisons:



     1.  Strip and Surface Mining - The estimated 1800 square mile-feet*



required to be committed for sludge disposal (50 percent solids)  by



1980 is:



         a.  About 89 percent of the total land disturbed by strip



             and surface mining of coal as of January 1965.



         b.  About 3.5 times the square mile-feet which will be



             utilized in strip and contour mining of coal from



             1973 to 1980.  The strip and contour mining utilization



             Ls estimated at 73.5 square-mile feet per year based on



             1967 Bureau of Mines reports.  '    This rate is



             undoubtedly lower than the present and forecast rate,



             but the comparative order of magnitude is obvious.



     2.  Coal and Ash Production - The estimated 112 million tons/year



production rate of wet sludge by 1980 is:



         a.  Approximately 25 percent of the weight of all the coal



             (450 million tons) estimated to be required for electric


                                      f41
             power generation in 1980.



         b.  Approximately 175 percent of the weight of all the wet



             ash (approximately 20 percent moisture) from all the



             coal required for power generation in 1980.  An ash



             content of 12 percent was assumed for the coal.
 *Square  miles  X ft(depth);  e.g.,  180  mi2  X  10  ft  =  1800  mi2-ft,




                               201

-------
 3.0  PRELIMINARY ASSESSMENT OF TREATMENT/DISPOSAL TECHNOLOGY




      Currently,  all of the existing or planned lime/limestone  scrubbing




 facilities for power plants in the United  States  employ  the two major




 ash disposal techniques,  ponding  and landfill,  for lime/limestone waste




 sludge disposal.   Disposal of  the sludge has been chosen because of




 unknown technical  and economic factors  and  the  lack of a readily available




 market for large-scale utilization of  the waste material, although




 research  and development  has shown at  least limited use  is feasible in




 some structural applications.   This  section discusses the results of a




 preliminary  assessment of treatment/disposal technology  made in late




 1972,  to  assist in  formulation  of the  F!PA program.




      3.1   Ponding




      Since ponds are  widely used  by  electric utilities for the disposal




 of  ash, it is  not surprising that  ponding is considered  to be a prime




 means  of  sludge disposal.   In  fact,  ponds are employed for disposal of




 sludge  in  about 60  percent  of  existing or planned U. S.  power plant




 locations  of  lime/limestone scrubbing facilities.  In addition, small




 ponds  are used for partial dewatering and temporary storage of the




 sludge  in  almost all  locations where other methods are being employed.




 In  general, ponds are used where a large area of inexpensive land is




 available near the power plant.   If  sufficient land is available,  the




pond is designed to eventually store wot sludge material  over the



lifetime of the power plant.




     The production rates  (tons/year, dry basis) of sulfur product




waste material are approximately 50 percent larger than the ash normally




                                202

-------
produced.  This leads to a total (sludge plus ash)  waste disposal



requirement about 2.5 times the normal coal ash disposal rate.   In



addition, since some sludge materials retain up to  60 percent moisture



even after an extended period of settling time, pond volumes several



times the size of normal ash ponds are required for sludge disposal.



     This moisture-retaining tendency, or tendency  for the solids not



to compact well upon settling, has been attributed  to the thin  platelet-



like crystal structure of calcium sulfite (CaSOg •  1/2 H-0), which



usually represents the majority of the sulfur in the sludge.  The calcium



sulfate (CaSC>4 • 2 FUO) crystals, on the other hand, are "blocky," and



tend to compact well.  This situation has created interest in means of




oxidizing the sulfite in the pond to sulfate.  Oxidation of the sulfite



could reduce the pond volume requirements, as well  as make the pond



area more easily reclaimable, rather than the possibility of indefinitely



remaining a wet "bog."  Oxidation can also reduce the potential oxygen




demand of the sulfite in solution.  However, oxygen consumption by the



sulfite does not appear to be a major problem because of the slowness



of the reaction under normal pond conditions.



     Oxidation of the sulfite could lead to other problems, however.



Over an ambient temperature range, the calcium sulfate is on the order



of 50 to 100 times more soluble than the sulfite, so the pond water



pollution potential would be greatly increased.  This solubility could



cause problems with reclaimed pond areas through sub-surface dissolution.




An alternative to oxidation might be dewatering using filtration,




centrifugation, etc. prior to ponding.  Solids with approximately






                               203

-------
 30 percent moisture have been  generated  by filtration  in  limestone
 scrubbing pilot  tests  at EPA.   Unfortunately,  it  is not currently
 known whether the  dewatered  material  will  "set-up" permanently;  i.e.,
 will  harden or remain  dewatered after exposure  to rainfall.   (See
 Section  5.2.2.)
      With either a lime  or limestone  scrubbing  system, ponding creates
 the potential  for  water  pollution.  Lime sludge solids consist primarily
 of CaS03  •  1/2 H90,  CaS04  •  2  H-,0, CaCCL,  and possibly some Ca(OH)2,
 and flyash;  limestone  sludge solids consist primarily of  the  same com-
 pounds, excluding  Ca(OH)  .   Both sludges contain  liquor which is, in
 all likelihood,  saturated in calcium  sulfitc and  sulfate.  In fact,
 pilot  and  full-scale operating  data have shown  that it may be super-
 saturated  in calcium sulfate.   Magnesium,  a normal impurity in lime
 and limestone, is  also contained in the  sludge.   Because  of the high
 solubility of  magnesium  sulfite and sulfate, magnesium is a serious
 potential pollutant  in the liquor associated with the sludge.  It is
 anticipated that in most applications, a relatively low limit may have
 to  be put on the magnesium contained  in the lime or limestone, possibly
 rendering dolomitic  limestones unsuitable.
     A variety of additional  chemicals from sources other than lime
 and limestone  are also present  in the sludge.   The chlorides in the
 coal,  for example,  are given  off in the combustion process as HC1.   The
HC1 is removed in the scrubber and, through internal recycle of scrubber
liquor, the concentration of  soluble chlorides builds  up in the liquor.
The concentration levels  reached are dependent on:  the quantity of
chlorine in the coal, the solubilities of the  chlorides formed,  and  the
                               204

-------
quantity of liquor associated with the waste sludge.   The dissolved



solids level in the scrubber liquor is greatly affected by this  chloride



build up, which presents another potential water pollution problem.




     It is generally recognized that flyash contains  trace quantities



of potentially toxic elements, although usually in extremely insoluble



form.  For example, limited analysis of dry-collected limestone-modified



flyash from EPA's dry limestone injection program indicated 50-200 ppm



As, about 200 ppm each of Ba and V, and 200-500 ppm Pb.  It is also



recognized that trace quantities of toxic elements present in coal and



oil are volatilized in the combustion process and are potentially



removed from the flue gas in a wet scrubbing process, subsequently



ending up in the waste sludge.  In addition, lime and limestone can



contain trace impurities which will be present to some extent in the



process streams.  Any trace chemicals in the makeup water will also be



present.  Because of the internal recycle of scrubber slurry, and the



recycle of water from the pond back to the scrubber circuit, it is



possible that trace chemicals from all of the above sources will build



up in concentration with time.  Similar to the chloride problem, the



concentration levels reached are dependent on:  quantities in the above



sources, their solubilities, and the quantity of liquor associated with



the sludge.  Potential exists, therefore, for trace chemical contamina-



tion of both surface and groundwaters unless proper precautions are



taken.




     To prevent unintentional seepage of liquor through the dikes and



floor of the pond into groundwater, it may be necessary to use a




                                205

-------
sealant material.  The pond diking should also he carefully designed




to avoid overflow and run-off of sludge liquor into surface water.




In addition, the system must be operated in a closed- or nearly-closed-




loop mode of operation.  This means that all liquor entering the pond




is recycled to the scrubber circuit; no sludge liquor is released to




any watercourse.




     In the majority of full-scale installations currently in operation,




attempts have been made to operate in a closed-loop mode.  However, in




most of these facilities, seepage, run-off, and other mechanisms could




be postulated which would allow liquor to be released into surface or




groundwatcr, at least periodically.  It should be noted that ash ponds




which have been used for many years have similar potential water pollution




problems,  although there are little available data related to water con-




tamination by ash liquors.  However, soil amendment studies using flyash




have shown trace element ingestion by plant life, suggesting the




possibility of leaching from flyash ponds.




     In summary, although ponding is currently widely used, the following




potential  problems arc evident:




     1.  Volumes several times the size of normal ash ponds,




         because of high moisture content in solids containing a




         significant percentage of calcium sulfite.




     2.  Potential sulfite oxygen demand.




     3.  Potential dissolution of calcium sulfate, with resultant




         water pollution potential.
                               206

-------
     4.  Potential reslurry of dewatered solids when exposed
         to rainfall.
     5.  Impurities and trace chemicals in the sludge, with
         inherent water pollution potential.
     6.  Seepage, run-off, and/or overflow of liquor from the pond.
     3.2  Landfill
     Although landfill operations are included in less than 40 percent
of existing or planned lime/limestone scrubber locations, ash disposal
has involved techniques other than ponding in these locations.
(Actually, in most cases ponding ultimately results in a landfill-type
operation; to reclaim the land when the pond is filled, the "pond" must
be covered with earth, similar to a landfill.)
     Since the sludge material contains large quantities of liquor and
therefore is difficult to handle, dewatering of the material for trans-
portation is likely although not absolutely necessary.  Clarification
alone will probably result in not more than 50 percent solids, so
further dewatering by either filtering or centrifuging will likely be
necessary.  The filtered or centrifuged material can then be transported
to a suitable landfill site.  As mentioned in the previous sub-section,
it is not currently known whether the dewatered material will remain
dewatered after exposure to rainfall.   (See Section 5.2.2.)   If it does
not, a fixation or stabilization technique must be applied.
                                 207

-------
    Chemical fixation processes are currently being developed by many


companies, including IU Conversion Systems, Dravo, Chicago FlyAsh,


Chemfix, and others.  Such processes generally involve pozzolanic*


chemical reactions between flyash and lime.  In most lime processes,


the hydrated lime in the scrubber and the slurry recycle tank is


almost completely converted to calcium carbonate.  Therefore, although


the percentage of unreacted calcium carbonate in a lime process sludge


is less than that in sludge from a limestone process, the sludge


components are identical.  This means that lime must be added to the


sludge to generate the pozzolanic process, possibly along with an


accelerating agent to decrease the curing time.  Reaction of the lime


and flyash leads to the formation of a relatively dry solid in which


the sludge components are physically, and possibly chemically, bound


up.  This is desirable because it decreases the potential dissolution


of these components and thus the possibility of groundwater contamina-


tion through leaching or nearby surface water contamination through


run-off.


    Binding up the sludge components through a chemical fixation process


is currently being applied at the limestone scrubbing facility at


Commonwealth Edison's Will County Station.  It is also planned for the


lime scrubbing facility at Duquesne Light's Phillips Station, to be
*Pozzolan:  Generic name for cementitious material which is based on

 ordinal use of volcanic rock or ash containing silica, alumina, lime,

 P^;,n?-a rf ?r   ,   construction in ancient aqueducts.   Named after
 Pozzuoli, Italy, where it was first found.
                                  208

-------
 started  up in  mid-1973.   It  is  not  currently known whether fixation



 techniques will be  applied to additional projects, although  it is




 known that the techniques are being considered  for a  large new coal-



 fired facility.



      In  summary,  problems similar to those  of ponding exist  with landfill




 operations,  for example:




      1.   Potential  reslurry  of  dewatered solids when  exposed to




          rainfall.



      2.   Dissolution of calcium compounds and trace chemicals in



          the sludge material, with  resultant water pollution




          potential.



 Fixation  processes  may alleviate the problems associated with landfill,




 but these  require further evaluation.




      3.3   Utilization



      As development activity in limestone scrubbing has increased during




 the past  few years, investigations  in the utilization of flyash and



 limestone  waste sludge material have been undertaken. Private companies



 involved  in this area have included  Combustion Engineering,  IU Conversion




 Systems  (formerly G&WH Corson), and  others.  In Japan, oxidation of the



 sulfur products to  gypsum for construction use is underway.  EPA has



 sponsored  work at West Virginia University's Coal Research Bureau(7) and




 at Aerospace Corporation.



     The approach taken in the EPA programs was to determine current and




potential utilization of flyash, assess applicability of limestone waste




sludge material and, based on its properties,  assess its potential for




                                 209

-------
 new  applications.  However, it was found that in  1970, only about




 7  percent  of  all the  flyash produced by coal-burning power plants was




 actually utilized  - primarily for concrete, structural fill, lightweight




 aggregate, raw material for cement, filler for bituminous products, and




 road base  material.   It was estimated that the maximum practical




 potential  usage under current technology and associated market conditions




 would only be about 25 percent.  Thus, approximately 75 percent of the




 flyash produced in the United States was not considered marketable in




 the near term because of:  expected variation in the composition of the




 flyash due to coal composition changes during operation, unfamiliarity




 with this  material, and the general inability to readily compete economi-




 cally with conventionally used materials.  Because of similar limitations




 there is not expected to he widespread utilization of limestone sludge




material in the United States, at least not in the near future.




    It can he concluded, therefore, that although utilization is




desirable  in the long run, near-terra treatment/disposal solutions to




 the sludge problem must occupy a higher priority for EPA research and




development.  However, promising utilization approaches which result in




materials  which can be disposed of as "waste," to be utilized at some




time in the future, are considered applicable to the current problem.




    3,4  Conclusions



    This preliminary assessment of treatment/disposal technology,




made in late 1972,  indicated a considerable number of potential problem




areas, all described earlier in this paper.  Some efforts are being




expended by private industry,  but these seem to be directed toward



                                210

-------
solving the problems of a specific application,  and are in many




instances quite limited in scope.  Because of the wide variety of




scrubber systems, sorbent/fuel combinations, power plant land use




considerations, etc., there is quite an apparent need for a program




with a broad, national scope which will encompass as many situations




as possible.  This broad-based program would take into account currently




available information, generate new information, and disseminate findings




to all electric utilities, private industry, and the general public.  It




also would have the advantage of reducing the cost of specific application-



oriented development by, among other things, avoiding duplication of




effort.




     EPA has initiated a program with the objective of meeting as many




of these requirements as practicable.  Due to fiscal and other constraints,




the program, although broad in scope, includes neither the full range of




sorbent/fuel combinations, nor their associated treatment disposal tech-




niques.  In addition, demonstration of the most promising treatment/




disposal technology is yet to be undertaken.  However, the EPA program



described in Section 4.0 is considered a major first step.
                                   211

-------
 4.0  THE EPA PROGRAM
      An EPA contract,  "Wet  Collected  Limestone-Modified  Fly Ash
 Characterization  and Evaluation of Potentially Toxic Hazards" was
 formalized  late in  1972  with  the Aerospace Corporation.  The contract
 provides for a detailed  characterization of wet collected limestone-
 modified flyash and  an evaluation  of  the potential toxic hazards posed
 in processes that may be performed  in subsequent handling, disposal,
 or utilization of the sludge.  The  need for this study was based on
 the results  of previous  programs conducted by LPA which  indicated the
 following:
      1.   Analyses performed on dry  collected limestone-modified
          flyash identified  the presence of heavy metal trace
          elements that could  pose a toxic hazard, depending upon
          their chemical/physical state.
      2.   The wet collected  limestone-modified ash may contain even
          greater quantities of the heavy metal elements as a con-
          sequence of the wet  scrubbing of the ash and flue gases,
          and poses a potential hazard  in its disposal or utilization.
      3.   Water-soluble elements pose  an additional health hazard in
          the disposal of waste waters  because of the large quantities
          of  water required  for the process scrubbing.
      However, when procurement for the current contract was initiated,
potential utility sources to obtain representative sample types were
limited;  disposal was essentially limited to ponding; and commercial
acceptability of throwaway processes  and the coresponding quantity of
material requiring disposal was unknown.  These factors led to a
                                   212

-------
program of limited scope with prime ecological  emphasis  on  toxicity.



Since that time, additional utility sources  using different sorbent/



fuel combinations and applying other treatment/disposal  techniques



have become operative.  These additional combinations  with  their



different elemental compositions and treatment/disposal  techniques



needed to be taken into consideration.  Although toxicity is still



important, toxic element concentrations are  expected to  represent only



a very small percentage of the total quantities to be disposed of.



Based on the above and the current and projected magnitude of the



sludge problem, EPA now plans an expanded program to allow a more



complete assessment of ecological acceptability, technical state of the



art, and economics for the various treatment/disposal techniques.



    The objective of the expanded program is to determine ecologically



and economically acceptable methods for treatment/disposal of lime/




limestone sludge and to provide pertinent input for the establishment



of realistic water and solid waste regulations and standards.  Sample



materials, representative of as many situations of lime/limestone wet



scrubbing process applications as practicable, will be obtained.  In



addition, test, operational and economics data from a wide variety of



sources will be taken into consideration.



    The basic elements of the program are as follows:



    1.  An inventory of sludge components, including chemical



        analysis of various types of sludge and the raw materials



        from which they are formed  (lime or limestone, coal or oil,
                                 213

-------
     process  water).   Sorbent/fuel  combinations to be studied




     are  limestone/Eastern and Western  coals, lime/Eastern coal,




     and  double  alkali/Eastern or Western coal.




 2.   An evaluation of  the potential water pollution and solid




     waste problems associated with disposal of the sludge,




     including consideration of existing or proposed water




     effluent, water quality and solid waste standards or guide-




     lines.   The  information would  also assist in the evaluation




     of potential treatment/disposal techniques described below.




3.   An evaluation of treatment/disposal techniques with




     emphasis on ponding and "fixed" and "unfixed" landfill




     (and related land use applications).  Physical analyses



     and tests of various sludges will be conducted,  including




     determination of the effects of dewatering,  oxidation,




     chemical fixation, aging,  etc., on stability, compactibility,




     leachability of solubles,  potential pond seepage, potential




     run-off problems,  and other disposal considerations.   The




     economics of various treatment/disposal combinations  will




     also be studied.




4.  A recommendation  of the best available technology for sludge




    treatment/disposal based on the evaluation described  above.




    Pertinent input for the establishment of realistic water




     and solid waste regulations and standards may also be provided.
                               214

-------
      The  approach taken by the EPA in undertaking this program is to
 utilize EPA funded or co-funded scrubber sources, the cooperation of
 several utilities who have installed scrubber systems and (as much
 as possible)  the knowledge and technical expertise of EPA water and
 solid waste organizations.   EPA personnel conducted a survey of
 potential electric utility participants for  the planned program.
 Factors in the  study included facility availability, additive/fuel
 combination,  facility size,  treatment techniques  (dewatering, oxidation,
 chemical  fixation, etc.). and final sludge disposition (ponding,
 landfill,  other).  Table 2 presents the results in order of facility
 availability  to support the  EPA program.
      Upon completion of the  survey, contacts were made with several of
 the utilities to determine their willingness to provide sludge and raw
 material  samples, as well as information concerning their sludge treat-
 ment  and  disposal activities.  In addition, discussions were held with
 several scrubber system companies, waste treatment/utilization/disposal
 companies,  and EPA water and solid waste personnel regarding review and
 integration of activities and information.
     Aerospace Corporation efforts have been limited to the scope of
 the current EPA contract.  However, these efforts have been conducted
 and coordinated to the maximum extent possible with EPA efforts toward
meeting the objectives of the planned program expansion.  Surveys of
current treatment,  disposal, and utilization technology, as well as
chemical analyses and physical testing of samples from the TVA Shawnee
facility,  have been conducted.  Progress of Aerospace and EPA efforts
are discussed in Section 5.0.
                                   215

-------
Table 2.  POTENTIAL SLUDGE TREATMENT/DISPOSAL PROGRAM UTILITY PARTICIPANTS
                   (X = Current; P = Possible Additions)
Facility
(Availability
Status)
TVA-Shawnee
(Current]
City of Key
West-Stock
Island
(Current!
Commonwealth
Edison Co. -Will
County
(Current)
Southern
California
Edison-Mohave
Kansas City
Power 5 Light-
Havv'thorn
(Current)
Kansas Power §
Light-
Lawrence
(Current)
Louisville Gas
& Electric-
Paddy's Run
(Current)
Sorbeut
Fuel
Limestone now, .^^^
lime later .^^^
*s^^ Eastern
^^^ coai
Limestone .->**
j^""^^
(coral marl) ^^
^*^^ Residual
^^^ oil
Limestone ^^^^
.s^ Eastern
.^ coal
Limestone ^^'
5 lime ^*-^^
^^-^^'^ Western
^^^ coal
Boiler *^
injected s/~^
limestone s^
S^ Coal
.S (possible E&W
^ blend)
Boiler ,^"
injected ^^^^
1 imes tone ^^~^
.^*^ Eastern
^^"^^ coal
Carbide ^^**
sludge ^^"'^
(Ca(OHK}^^
^s- — Eastern
<-^"^ coal
Scale
Proto-
type
Full
Full
Pilot
Full
Full
Full
Dewatering Technique
Ciuri-
f icr
X

X

X

X
Filter
X

V
\


X
Centri-
fuge
X


P



Dryer


P




Pond

X
X
(Well
points)

X
(Well
points)
X

Final Disposition
Ponding
X


X
X
X

Landfill

X
(Unfixed)
X
(Sealed)
(Fixed)



X
Other.








-------
                    Table 2 (Continued).  POTENTIAL SLUDGE TREATMENT/DISPOSAL PROGRAM UTILITY PARTICIPANTS

                                           CX = Current; P = Possible Additions)
Facility
(Availability
   Status)
Sorbent
                 Fuel
Scale
Clari-
  fier
                     Dewatering Technique
                                                            Filter
Centri-
  fuge
                                                                Dryer
                                             Pond
                                                    Final Disposition
                                                                                               Ponding
                                                          Landfill
                                                                              Other
Northern
States Power-
Black Dog
 (Current)
Limestone
                Western
                coal
                           Pilot
Kansas City
Power §
Light-
LaCygne

 (Current)
Limestone
                Eastern
                coal
                           Full
Arizona
Public
Service-
Choi la
(Approx
mid-1973)
Limestone
                          Full
               Western
               coal
                                                        X
                                                     (Solar
                                                     evap)
Duquesne
Light-
Phillips
(Approx
Bid-1973)
Lime
                          Full
               Eastern
               coal
                                               X
                                            Curing)
                                                            X
                                                         (Fixed)
Detroit
Edison-
St. Clair
(Late 1973)
Limestone
                          Full
               Eastern
               coal
                                                                 (Unfixed)

-------
 5.0  PROGRAM PROGRESS


      5.1  Industry and  Intra-EPA Coordination


      The information highlights discussed here are based on an initial


 survey conducted  as part  of  the EPA  program  via  informal  meetings


 and/or recent publications.   The information generally  falls  into the


 categories  of general,  technical,  and  economic.


          5.1.1  General - Lea.chate pollution of  groundwater or water


 courses appears to  be the primary  concern for  sludge disposal.  Results


 of  various  studies  have shown a cause  for this concern.   For  example,


 greenhouse  studies  have shown that application of  selected samples of


 flyash  to soil increases the  availability of boron, molybdenum, zinc,


 phosphorus,  and potassium to  plants  by supplying soluble  forms of these

                                       ( Q~\
 elements  and/or modifying the soil pH.  }  These results  add  credibility


 to  the  implication  that these elements and others  can leach from flyash


 even  though  the ash  consists  primarily of glassy silicates considered


 relatively  insoluble  in aqueous  environments.  While leaching in this


 particular  application  is undoubtedly  a function of surface area


 exposed to permeating moisture,  it provides  additional evidence that


 water quality may be  sacrificed by unsuitable disposal of flyash/


 sulfur  sludges.  Further,  at  least one state has preliminarily disapproved


disposal of a particular sulfur sludge on the basis that potentially toxic


materials in the sludge were  in excess of potable water standards.


     Coordination with EPA water personnel indicated that because of a


 lack of detailed information, there are presently no Federal water



                                   218

-------
effluent, water quality, or groundwater standards* which take into

account power plants with on-site desulfurization systems.   As a

result, State and local regulatory agencies are applying available

water standards  dealing with receiving stream quality and  potable

water.  The Federal Water Pollution Control Act Amendments  of 1972

require the setting of standards applicable to effluents and ground-

water, and also establish a goal of zero pollution discharge by 1985.

     Based on the above, a study on recycle, reuse, or treatment of

effluent waters related to flue gas lime/limestone wet scrubbing

processes is under consideration by EPA water personnel.  This study

would be closely coordinated with the program discussed in this paper

and represents additional broadening of the overall EPA program.  The

study, directed toward determining the implications of open-loop or

partially open-loop systems, would involve analyzing effluent liquor

downstream of the scrubber system.  In addition, the study would

include the potential for treatment of the water after scrubbing and/or

dewatering.  Noxious effects as well as toxic hazards would be con-

sidered, with technology and economics determined  for all cases.

     Coordination with EPA solid waste personnel also indicated that

there are no specific Federal standards* related to power plant sludge

disposal.  However, considering  the current  lack  of information concerning

sludge properties,  disposal of sludge  in sanitary  landfills  is not

expected to be permitted.
*Current and proposed State and Federal guidelines, standards, etc.,
 are being compiled for correlation with EPA program results.

                                 219

-------
      For these and other reasons,  there is  considerable interest  in



 treatment of the sludge to combine primary  and trace  elements  into



 new crystalline phases  and also to reduce the leaching  rate.   One



 such treatment technique under development  results  in the production



 of synthetic aggregate,  which  would be  used immediately or  stored for



 subsequent use.   This lime/flyash/sulfur sludge aggregate would be



 used in load-bearing applications  in place  of natural aggregate with



 either conventional cement binder  material  or with  lime/flyash/sulfur



 sludge binder material.   Synthetic aggregate  production was advanced



 as a means  of raw material (limestone aggregate) conservation  as  well



 as ecologically  sound disposition  of sludge.   This  would free  limestone



 supplies  for scrubbing and other uses.   Without  this  approach, local



 limestone availability in sufficient quantity to supply sorbent materials



 for large scale  wet scrubbing  installations has been  questioned.




      On-site flyash availability in  sufficient  quantity for some  treat-



 ment  formulations may be  questionable.   In  other words,  even with



 all  the plant  flyash available, some utilities may  need  additional



 flyash or a  substitute, depending  on final  disposition  of the  treated



 sludge.   For utilities which currently market  their flyash, the problem



 is even greater.   In fact, one such utility is currently looking  for a



 flyash substitute for their sludge disposal process.  Although optimum



 formulations are not presently known, applications  such  as structural



or non-structural landfill, base course materials, and  synthetic



 aggregate are expected to require different mix proportions.
                                  220

-------
     Additional information related to sludge  disposal  has been generated



through EPA efforts in mine drainage pollution control  and solid waste



disposal.  EPA mine drainage activities have resulted  in  numerous



reports dealing with sludge produced by neutralization of acid mine



drainage.  This type of sludge was one of those tested at Transpo  '72.



(See Section 5.1.2, below.)  The reports cover areas  such as  in-situ



sludge precipitation, sludge supernatant treatment,  thickening and



dewatering, use of latex as a soil sealant,  and technical and economic



feasibility of bulk transport.



     Although primarily concerned with municipal sludge,  many EPA  solid



waste activities relate to the CPA sludge program under discussion.



Examples include the following:



     1.  Methods of removing pollutants from leachate water.



     2.  Determination of organic and metal characteristics  of




         leachate from selected landfills.



     3.  Evaluation of landfill liners.



     4.  Development of mathematical models to determine effects



         of landfill leachate on groundwaters.



     5.  Leachate pollutant attenuation in soils.



     6.  Moisture movement in landfill cover material.



     The EPA reports and other results of continuing coordination will



be taken into account under the current program.



         5.1.2  Technical - A wide divergence in physical properties



between untreated sludges has been reported.  Settling rates varied



considerably and thickener design parameters have shown large spreads




                                 221

-------
                                                       fg-j
 from  source to  source and even within the same source.v J  Although


 the large variations observed in sludge behavior are currently not


 completely explained, it is known that sulfite/sulfate ratio, percent


 flyash, and other measures of chemical composition have a definite


 effect.  However, the operating conditions which result in the difference


 in sludge properties (or its chemical composition] are not as well


 defined.


      As mentioned earlier, there is considerable interest in treatment


 of sludges in pozzolanic fixation processes using lime.  The production


 of the sulfur sludge pozzolan is based on:


      1.  The reaction of lime with soluble sulfates originating in


         either the flyash or the sludge water to form calcium sulfate


         and tie up water.


      2.  The reaction of lime, sulfate, iron oxide, and/or alumina


         to form complex crystalline sulfoferrites or sulfoaluminates


         such as ettringite (A1203 • 3 CaS04 - 31 H20).


      3.  The reaction of lime with the glassy silica of the flyash


         resulting in the well-known pozzolanic reaction proceeding


         slowly to form the calcium silicate, tobermorite.


     These reactions reportedly result in significant changes in physical


 and chemical properties of the sludge.  A reduction in permeability, an impor-


 tant property directly related to ecological disposal, has been reported.


This  is based on falling head permeability tests indicating a decrease


by about 2 orders of magnitude between raw sludge and treated sludge
                                  222

-------
             (10,11)
after a week.         It has also been reported that the trace  element


concentration in leachate from a sulfur sludge pozzolan produced from


acid mine waste sludge was also reduced by about 2 orders of magnitude.


Considering these results, the availability of soluble contaminants


to groundwaters could be reduced on the order of 10,000 times.


     It is postulated that the expansive nature of ettringite crystalli-


zation during hardening seals the mass and prevents shrinkage cracks,


thereby reducing permeability.  Thus, dimensional stability appears  to


be another important property of the sulfur sludge pozzolan especially


when compared to untreated sludge.


     There is also some evidence that inclusion of sulfur sludges


improves the properties of conventional (limc/flyash/aggregate) pozzolans.


The mix containing sulfur sludge is reported to develop superior early


strength; greater final strengths appear possible.  For example, com-


pressive strengths of up to 350, 750, and 1100 psi have been reported


for sulfur sludge pozzolan samples cured at 100°F for 2, 7, and 28 days,


respectively. *-  J  With unspecified formulations and cure conditions,


unconfined compressive strengths of 800 and 1600 psi after  14 days and


1550 and 2700 psi after 28 days were also reported/11-'  It is postulated


that these properties are also the result of the ettringite crystalli-


zation.

                          (12)
     Transpo '72 results/    on the other hand, indicated paving


material compressive strengths of about 100 and 225 psi for an acid


mine drainage sludge pozzolan and an S02 scrubber sludge pozzolan,


respectively, after 28 days at 73°F.  However, this is a single test



                                    223

-------
 result  and time  did not permit optimization of the  formulation.   In



 addition,  it  is  reported  that inadequate compaction, sealing and



 curing  time due  to inclement weather and other factors, may have com-



 promised performance.  Cube compressive strengths of synthetic



 aggregate  made from acid  mine drainage sludge and SC^ scrubber sludge



 and  cured  at  73°F were approximately 3500 and 3300  psi, respectively,




 after 28 days; and approximately 6500 and 6000 psi, respectively,



 after 90 days.   Synthetic aggregate used directly as a crushed aggre-



 gate in conventional binder material and as an aggregate in the



 sludge/flyash/lime paving mixture was reported to perform adequately,




 with no evidence of failure.



         5.1.3   Economics - The cost of local (within 20 miles) disposal



 of wet  (50  percent solids) lime/limestone sludge, including an additive



 for pozzolanic fixation,  has been estimated by one  source to be as low



 as $2.50 and by  another to be as high as $10/ton.   '  However, the



 bases for these  estimates are not completely known  and may be signifi-



 cantly different from each other.  The cost of the  additive (primarily



 lime) has been estimated  at from less than $18 up to $20/ton;  a 3



 percent by weight addition for sludge treatment would result in a cost



 of $0.50 to $0.60/ton of wet (50 percent solids)  sludge.



     The cost of aggregate formation from the sludge has been estimated



from about $5 to $8/ton.   This compares to a cost of from $1.50 to $8/ton



for naturally occurring aggregate;  this cost depends on local supply,



transportation,  etc.   A typical  average cost is  reportedly about $2.50/ton,
                                 224

-------
     The choice between disposal by landfill and synthetic aggregate



production depends on many factors, such as natural aggregate availa-



bility and costs, transportation costs, land availability, and plant



size.  Because of the many variations in these factors from one power



plant to another, it is difficult to select a single overriding con-



sideration.  However, it has been estimated that the break-even point



for synthetic aggregate production (versus disposal cost) is approxi-



mately 200,000 tons/year wet sludge production.  Below this sludge



production rate, the economics reportedly favor direct disposal.  The



bases for this break-even point is unknown.  However, the plant size




for 200,000 tons/year of wet sludge can be estimated from Figure 3,



Section 2.0.



     5.2  EPA/Aerospace Corporation Contract



     Initial results of analytical and physical property tests on



limestone scrubbing sludge and process materials from the TVA Shawnee



power plant burning Eastern coal are discussed below.  It should be



noted that these results are of a very preliminary nature and are based



on a single sample from one source.  As such, these results may or may



not be representative of those from other samples  from the same source



or from different sources.




         5.2.1  Chemical Analysis  - Emission spectrographic analysis



of clarifier underflow liquor and  solids  (separated by centrifugation)



and of the limestone indicated:




     1.  Liquor - No toxic components; dissolved solids content



         representative of equilibrium calcium sulfate concentration.





                                   225

-------
      2.   Solids  -  Presence of  several heavy metals,  including

          lead, gallium,  chromium  and titanium.   In this specific

          case, based  on  the  above results of  analysis of  liquor

          which had  been  in equilibrium with the  solids, these

          elements  do  not  appear to present a  toxic hazard.  However,

          the  specific compounds and their solubilities under other

          conditions are  presently unknown.

      3.   Limestone  -  Presence  of  strontium in addition to carbon,

          sodium, magnesium,  aluminum, silicon, potassium  and

          calcium.   All other elements found are  constituents of

          tramp clay minerals in the limestone.   No other  mineral

          phases  were  found.  The  limestone showed few major impurities

          that could cause potentially toxic effects.

      Ion  microprobc mass analyzer (IMMA) results have shown good agree-

ment  with the emission spectrographic analysis.  More precise analyses

of solids by spark  source mass spectrometry (SSMS) and liquors by atomic

absorption  (AA)  are in progress.  Available physiological concentration

effects data and toxic and hazardous element  standards are being com-

piled for correlation with analyses of untreated and treated sludge

materials and run-off, overflow,  and leachate liquors.

     Tests to determine whether potentially toxic components may

sublime from a pond surface  subjected to drying by solar heating indi-

cated that this was not a problem up to 650°F.

          5.2.2  Physical Properties Testing - Testing of unconditioned

(raw) sludge from the clarifier underflow produced the following

preliminary results:
                                  226

-------
1.  Drainability and Water Retention  - In  a  falling head



    column,  a naturally settled sludge of  approximately



    50 percent solids showed steady state  drainage of  7.5 X  10



    cm/sec.   After air drying to less  than 15-20  percent



    moisture, the column was found to  be virtually impervious



    until several days later.  At that time, initial drainage



    was noted, but at a rate about half that indicated above.



    At intermediate column dryness, initial  rate  and delay  in



    drainage were intermediate.  In each case the final drainage



    rate was approximately equivalent  to the steady  state rate



    with the 50 percent solids sludge.  In the experiments  with



    sludge dried above 50 percent solids,  the column appeared  to



    regain sufficient water to return to the initial water  content



    (about 50 percent) and permeability.  The effects of extended




    time, and other dewatcring and compaction techniques  on this



    sludge behavior are not currently known.



2.  Corrosion - Sample  specimens  of 1100 aluminum and 1010 mild



    steel have been exposed to sludge solids and supernatant



    liquor.  The aluminum  specimens  have shown no degradation.



    The effect on steel  specimens differs between that portion



    suspended within the solids and that immersed within the



    liquor.  After  1 month, the area in contact with the solids



    appeared dull black,and a weight loss of about 1 percent



    was measured.   After 2 months, the area within the solids



    was heavily encrusted with corrosion products and bound




                               227

-------
          sludge particles and a net weight gain of more than
          10 percent was measured.  The area in the liquor dis-
          played rust  (Fe(OH)3) which was non-adhering and
          easily rinsed from the specimen.
      3.   Viscosity - The viscosity of sludge samples with solids
          contents varying from 50 to 60 percent was measured with
          a viscometer which uses a rotating cylindrical cup immersed
          in the fluid being measured.  Peak viscosities of 120 poises
          for 60 percent solids and 20 poises for 50 percent solids
          were measured.  Thixotropic behavior was exhibited by all
          samples.
      4.   Bulk Density - Bulk density was determined for sludge with
          water content from zero to 100 percent.  Normally settled
          and dried sludge bulk density was 1.2 g/cm3 (75 Ib/ft3).
          With increasing water content, the bulk density increased to
          a maximum of about 1.7 g/cm3 (106 lb/ft3) for a water content
          of about 30 percent.
      5.   Shrinkage - Sludge cast into a known volume and allowed to
          air dry exhibited linear shrinkage of about 4 percent.
          Shrinkage of this magnitude could produce cracks in dried
          ponds, providing paths for subsequent leaching.
      5.3  Concluding Remarks
      The  authors were concerned with presenting results obtained so
early in  the EPA program.   However, the overriding consideration in
presenting these results was to transfer information in this
important area as  soon as  possible.
                                  228

-------
     Additional information and details pertaining to industrial



activities and EPA program progress related to sludge treatment



and disposal can be found in other papers presented at this symposium,
                                 229

-------
6.0  PRELIMINARY CONCLUSIONS

     In addition to the obvious conclusion that much extensive


research and development remains to be done,  the following preliminary

conclusions, based on initial progress, are drawn.

     6.1  Untreated Sludge

     1.   Based on reported sludge variability and divergence of

         properties between and within sources,  there is  a

         definite need to thoroughly characterize sludge  materials

         before a disposal/utilization scheme can be determined.

     2.   Sublimation of potentially toxic  components docs not

         appear to be a problem at upper extremes in ambient

         temperature.

     3.   Air-dried,  naturally settled sludge  exhibits  a tendency  to

         regain moisture back to its equilibrium  moisture content and


         initial  permeability.   Effects of other  treatments  and extended

         time arc currently unknown.


     4.  Mild steel  does  not  appear  to  be  a satisfactory  material for

        use in storage  or  handling  of  untreated  sludge.

     5.  Sludge pumpability appears  to  be  significantly reduced as

        solids content  increases  from  50  to  60 percent solids.
                                                                 i
     6.  Air-dried sludge exhibits a tendency to  shrink enough  to

        produce  cracks which could  serve  as  paths for leaching to

        occur.

     6.2  Treated Sludge


     1.  The  treatment of sludge in a pozzolanic fixation process

        .appears to significantly reduce the permeability of the

                                  230

-------
        material; the  leachability of trace elements from within



        the material is not clearly established.



     2.  The extent  to  which lime, flyash, and sulfur sludge



        proportions influence permeability, leachability, and



        strength development is not clearly established.



     3,  The inclusion  of  sulfur sludge  in conventional  (lime/



        flyash/aggregate) pozzolans, for load-bearing applications,



        produces a  material which appears to develop superior early



        strength and  (possibly) greater final strength.



     4.  Pozzolanic  mixtures of  lime, flyash, aggregate  and  sulfur



        sludge  may  be  useful for road basecourse  construction;



        they  do not appear useful for wearing surfaces.



     5.  Synthetic aggregate produced from  lime,  flyash,  and sulfur



        sludge  may  be  suitable  for  basecourse construction  either



        directly as crushed aggregate with  conventional binder



        materials or  as  aggregate with  a  lime/flyash/sulfur sludge



        binder  material.  Substitution  of  synthetic aggregate for



        all  cr  part of the natural  aggregate required  in a  construc-



         tion  mixture  could significantly  increase the  use of the



        waste sludge  input.



     6.   Optimum formulations for  the various disposal/utilization



         techniques  are not  clearly  established.




     6.3  General




     The reader is  reminded of  the nature of the results on  which



these preliminary conclusions are based.  Much more data and information



must be obtained and interpreted before firm conclusions can be reached.




                                 231

-------
The attainment of these firm conclusions in the shortest possible



time will require a high degree of cooperation between industry and



governmental agencies to assess, develop, and apply the sludge treat-



ment, disposal, and utilization technologies.
                                   232

-------
                             References



1.  The Aerospace Corporation,  "Technical and Economic  Factors



    Associated with Flyash Utilization," NTIS No.  PB 209-280, final



    report on EPA Contract F04701-70-C-0059.



2.  Brackett, C. E., "Production and Utilization of Ash in the



    United States," presented at the Third International Ash Utilization



    Symposium, Pittsburgh, Pennsylvania, March 13-14, 1973.



3.  Sulfur Oxide Control Technology Assessment Panel (SOCTAP),  "Final



    Report on Projected Utilization of Stack Gas Cleaning Systems by



    Steam-Electric Plants," submitted to the Federal Interagency



    Committee, Evaluation of State Air Implementation Plans, April 15, 1973.



4.  Princiotta, F. T., Kaplan,  N., "Control of Sulfur Oxide Pollutants



    from Power Plants," presented at EASCON  '72, Washington, D. C.,



    October 18, 1972.




5.  U. S. Department of the Interior, "Surface Mining and Our Environment -



    A Special Report to the Nation," 1967.




6.  Young, W. T., "Thickness of Bituminous Coal and Lignite Seams Mined



    in 1965," Bureau of Mines 1C 8345, August 1967.



7.  Condry, L. Z., Muter, R. B., and Lawrence, W. F., "Potential



    Utilization of Solid Waste from Lime/Limestone Wet Scrubbing of



    Flue Gases," presented at the Second International Lime/Limestone



    Wet Scrubbing Symposium, New Orleans, Louisiana, November 8-12, 1971.



8.  Martens, D. C., Plank, C. 0., "Basic Soil Benefits from Ash




    Utilization," presented at the Third International Ash Utilization



    Symposium, Pittsburgh, Pennsylvania, March 13-14, 1973.




                                    233

-------
 9.  Selmeczi, J. G., Knight, R. G., "Properties of Power Plant Waste




     Sludges," presented at Third International Ash Utilization




     Symposium, Pittsburgh, Pennsylvania, March 13-14, 1973.




10.  Minnick, L. John, "Multiple By-Product Utilization," presented




     at the Third International Ash Utilization Symposium, Pittsburgh,




     Pennsylvania, March 13-14, 1973.




11.  Bitler, J. A., Minnick, L. John, "Lime-Sulfur Dioxide Scrubbing



     System and Technology for Utilization of Underflow Sludge,"




     Industrial Waste, March/April 1973.




12.  Brink, R. I!., "Use of Waste Sult'ate on Transpo '72 Parking Lot,"




     presented at Third International Ash Utilization Symposium,




     Pittsburgh, Pennsylvania, March 13-14, 1975.




13.  Gifford,  D. C., "Will County Unit  1  Limestone K'et Scrubber,"




     presented at AIChE 65th Annual  Meeting, New York City,  N. Y.,




     November 28, 1972.
                                   234

-------
      TEST RESULTS FROM THE EPA
LIME/LIMESTONE SCRUBBING TEST FACILITY
                    by

         M. Epstein,  C.  C.  Leivo,
        C. H.  Rowland, S. C. Wang
            Bechtel Corporation
         San Francisco,  California
                     235

-------
                 ACKNOWLEDGEMENT

The authors wish to acknowledge the valuable contribution of
the Bechtel, TVA and EPA on-site personnel at the Shawnee
Test Facility.
                            236

-------
                             Section 1
                         INTRODUCTION

In June 1968, a three phase program was initiated whose aim was the
testing of a large, versatile prototype system to fully characterize
wet lime and limestone scrubbing for removal of sulfur dioxide and
particulates from boiler  flue gas.  The Office of Research and Mon-
itoring (OR&M) of the Environmental Protection Agency (EPA) is
sponsoring this program,  with Bechtel Corporation of San Francisco
as the major contractor and test director, and the Tennessee Valley
Authority (TVA) as the constructor and facility operator.

Phase I of the test program consisted of preliminary engineering,
equipment evaluation and site selection.   Phase II involved the detailed
design and construction of the facility and the development of the test
plan and mathematical models for predicting system performance.
Phase III,  the testing portion of the program, began in March 1972.

The test facility consists of three parallel scrubber systems each
capable of treating approximately 30, 000 acfm (10 Mw equivalent)
of flue gas, which are integrated  into the flue gas ductwork of an
existing  coal-fired boiler at the TVA Shawnee Power Station, Padu-
cah, Kentucky.
                                237

-------
This paper will cover,  primarily, the test results for wet limestone
scrubbing from September  1972 to April 1973 at the test facility.
The operability and reliability of the facility during  the limestone
testing will be covered in a second paper at this symposium (Ref.  1).
The results of air-water and sodium carbonate testing from May 1972
to August 1972 have been presented in Ref.  2.
                               238

-------
                             Section 2

                  TEST PROGRAM OBJECTIVES


The overall objectives of this program are to evaluate the perfor-

mance, reliability and economics of closed-loop limestone and lime

wet scrubbing processes.  The following are specific goals of the

program:
    •    Investigate and solve operating and design problems,
         such as  scaling, demister  plugging, corrosion and
         erosion  (see Ref. 1).

    •    Generate test data to characterize scrubber and sys-
         tem performances as a function of the important
         process variables.

    •    Study various sludge disposal methods.

    •    Develop mathematical  models to allow economic
         scale-up of attractive operating configurations to full-
         size  scrubber  facilities and to estimate capital and
         operating costs for the  scaled-up system designs.
    •    Determine operating conditions  for optimum SO2
         and particulate removal, consistent with  operating
         cost considerations.

    •    Perform long-term reliability testing.
                               239

-------
                             Section 3
                         TEST FACILITY

The test facility has been described in detail in Ref. 3.  It consists of
three parallel scrubber systems,  each with its own slurry handling
system.  Scrubbers are of prototype size,  each capable of treating
approximately 30,000 acfm of flue gas from the TVA Shawnee coal-fired
boiler No.  10.  Therefore,  each circuit is  handling the equivalent
of approximately 10 Mw of power  plant generation capacity.  The
equipment  selected was sized for  minimum cost consistent with the
ability to extrapolate results to commercial scale.  The 30,000 acfm
scrubber train was judged to meet these requirements.  Boiler No. 10
burns a high-sulfur bituminous coal leading to SO2 concentrations of
2300-3300  ppm and inlet grain  loadings of about 2 to 6 grains/scf in
the flue gas.

Based on their ability to handle slurries without plugging or excessive
pressure drop and on their reasonable cost and maintenance, the  fol-
lowing scrubbers were selected for testing:

    (1)   Venturi followed  by a spray-tower (after
         absorber)
    (2)   Turbulent contact absorber (TCA),
    (3)   Marble-bed absorber (Hydro-Filter)
                              240

-------
Figures 1, 2, and 3 depict, approximately to scale, the three scrub-
ber types  along with the demisters (and demister wash sprays) se-
lected for de-entraining liquor in the gas streams.  Note that, for
the TCA scrubber (Figure 2), a Koch wash tray has been installed
between the uppermost stage and the demister, in order to provide
a liquid barrier to prevent the bulk of the entrained slurry from im-
pinging directly onto the demister.

Typical system configurations for the venturi, TCA and Hydro-Filter
systems are shown in  Figures 4,  5, and 6, respectively.  Such pro-
cess details as flue gas slurry saturation sprays and demister or
Koch tray wash sprays are not shown.
                               241

-------
                         FIGURE I
            SCHEMATIC OF VENTURI SCRUBBER
                  AND AFTER-SCRUBBER
       CHEVRON DEMISTER
     AFTER-SCRUBBER
INLET LIQUOR

         THROAT


  ADJUSTABLE PLUG


 VENTURI SCRUBBER
                                GAS OUT
                                                DEMISTER WASH
INLET LIQUOR
     EFFLUENT LIQUOR
      APPROX. SCALE
                            EFFLUENT LIQUOR
                           242

-------
                          FIGURE Z
         SCHEMATIC OF THREE-STAGE TCA SCRUBBER
                 WITHOUT TRAP-OUT TRAY
                           GAS OUT
        CHEVRON DEMISTER
INLET KOCH TRAY
      LIQUOR
                                       KOCH TRAY
                   EFFLUENT KOCH
                 TRAY WASH LIQUOR
               STEAM WASH
      RETAINING GRIDS
                IN
                           0   o
                           °°o°o
                           o o  ow
    o
o o °o
                             00
                                      INLET LIQUOR
                                       MOBILE PACKING SPHERES
                                            I-
                                                5'
                       H
                                            APPROX. SCALE
                        EFFLUENT LIQUOR
                            243

-------
                       FIGURE 3

      SCHEMATIC  OF HYDRO-FILTER SCRUBBER
DEMISTER WASH

  INLET LIQUOR





  INLET LIQUOR


       GAS IN
                       GAS OUT
CHEVRON DEMISTERS






TURBULENT LAYER




GLASS SPHERES





     EFFLUENT LIQUOR





      y
   i	\
  APPROX. SCALE
                    EFFLUENT LIQUOR
                         244

-------
N>
J=
w*
                 FLUE
                          VEWW1

                         SCRUBBER
                             ®
                                                                                    FIGURE  4

                                                                                 EPA TEST FACILITY


                                                                        TYPICAL  PMKfSS  FLOW  DIAGRAM


                                                                               FOR VENTURI SYSTEM
ft




i
»}»»

A r^ A
/* f\ f\
f+ ^ ^
WWW

\ /





"AFTER
SCRUBBER
                                                                                      1.0. FAN
                                                                                                                                         SETTLING PONO
                  O   Gas Composition

                  ®   Paniculate Cwnposition & Loading

                  ©   Slurrv or Soliffs Composition
— —   to Stream

—   Liquor Stream

-------
                                                                           FIGURE 5
                                                                         EPA TEST FACILITY

                                                              TYPICAL   PROCESS   FLOW  DIAGRAM

                                                                         FOR TCA SYSTEM
KJ
JT
en
                                                                                                                                         SEHLINGPOND
          O  Gas Composition
          ®  Particulate Composition & Loading
          (•)  Slurry or Solids Composition
	   Gas Stream
—   Liquor Stream

-------
KJ
                                                                         FIGURE  6
                                                                      EFA TEST FACILITY

                                                            TYPICAL  PIOCESS  FlOW  DIACfAM

                                                                FOt  HTMO-FILT1I SfSTEl
                                                                                                                                 SETTLING POND
         O   G»sComp«iUon
         ®   Particulatc Cocnpositttm 4 Loading
         ©   Slurry or Solids Composition
—> _  Gas Stream
__  Liquor Sirnm

-------
                             Section 4
                         TEST PROGRAM

 The following contains a brief description of the test program.  A
 more detailed description has been presented in Refs. 2 and 3.  In
 Table  1, a description of the reports which are presently scheduled
 for general distribution is presented,

 4. 1      TEST PERIODS AND TEST PROGRAM SCHEDULE

 The following sequential test blocks have been defined for the test
 program.

     (1)  Air-Water Testing
     (2)  Sodium Carbonate Testing
     (3)  Limestone Wet Scrubbing Testing
     (4)  Lime Wet Scrubbing Testing

The  test program schedule is presented in Figure  7. As  can be
seen, the air-water and sodium carbonate tests have been completed.
As of mid April 1973,  limestone wet scrubbing short-term factorial
tests were  approximately 90% complete and longer-term reliability
verification tests were approximately 25% complete.
                               248

-------
                                                     Table  1
                                  TOPICAL AND FINAL REPORT DESCRIPTION
K>
£
            Report Title

     1.  Air-water, Sodium
        Carbonate and Open-
        Loop Limestone Test
        Results
2.  Limestone Wet-Scrubbing
   Test Results
     3.  Lime Wet-Scrubbing
        Test Results/Limestone
        Reliability Test Results
        Status
    4. Final Report
             Information to be Included

Summary of operational problems and resolutions,
planned and actual test designs, results of air-water
and NazCOj testing, utilization of data for model de-
velopment, results of open-loop limestone testing
with interpretation of data.

Summary of operating problems and resolutions
associated with reliability verification testing,
planned and actual test design, results of  closed-
loop factorial tests, interpretation of data, status of
process model development and selection of param-
eters for limestone long-term reliability testing.

Summary of operational problems and resolutions
associated with lime reliability verification testing,
planned and actual test designs, results of factorial
lime testing, status of process model development,
interpretation of data and status of limestone relia-
bility testing.

Summary of total test program with particular em-
phasis on lime and limestone reliability test results,
mathematical models, scale-up design and economic
studies.
                                                                                      Estimated General
                                                                                       Publication^ Date

                                                                                           June, 1973
September, 1973
                                                                                       February,  1974
                                                                                           July,  1974

-------
                                        FIGURE 7

                                        PHASE  III

                                 SHAWNEE TEST SCHEDULE
TEST PROGRAM FUNCTIONS
SYSTEM CHECK-OUT
AIR-WATER & SODIUM CARBONATE TESTING
LIMESTONE WET-SCRUBBING TESTING:
Short-Term Factorial Tests
Reliability Verification Tests
Short -Term Factorial Tests
Reliability Tests
LIME WET-SCRUBBING TESTING:
Short -Term Factorial Tests
Reliability Verification Tests
Reliability Tests
ENGINEERING & COST ESTIMATE STUDIES
1972
MAMJ JASOND
123456789 10
•^•MMM
OPEN-LOOP TESTING-*
BOILER OUTAGE
SYSTEM MODIFICATK
1973
JFMAMJ JASOND
11 12 13 14 15 16 17 18 19 20 21 22
i 	 1 j — *-CLOSED-LOOP TESTING
_J |
i j
/ -
DNS
«••••
1974
J F M A M J
23 24 25 26 27 28

••


KJ
U1
O

-------
4.1.1   Air-Water Testing





These experiments, which use air to simulate flue gas and water to


simulate alkali slurry, are designed to determine pressure drop model


coefficients and observe fluid hydrodynamics (e.g.  Hydro-Filter tur-


bulent layer) for all three scrubbers in clean systems.





4.1.2   Sodium Carbonate  Testing





Two series of sodium carbonate tests have been designed.  The first,


or high concentration series, utilizes concentrated (~1 wt. % sodium


ion) water solutions of sodium carbonate to absorb SO_ from flue gas
                                                    c*

and from  a synthetic flue gas composed of air and SO_.  These tests
                                                   L*

are designed to determine uncertain model coefficients for  the case


where gas-side mass transfer is rate controlling.  The second,  or  low


concentration series, uses  dilute (~0. 1Z5 wt. %  sodium ion) sodium


carbonate solutions to absorb SO  from flue  gas  and synthetic flue gas.
                                £

For this series,  gas-side mass transfer is not rate controlling and


liquid-side mass transfer uncertain coefficients can be calculated


using relationships for gas-side coefficients  developed from the high


concentration tests.  These runs also help ascertain the absorption


capability of liquors  associated with some variations of the Double


Alkali scrubbing process (see Ref. 4) over a range of operating


conditions.
                                251

-------
4.1.3    Limestone Wet-Scrubbing Testing


The primary objectives of these test sequences for closed liquor
    'i;
loop  operation, are:
    (1)  To characterize,  as completely as practicable, the
         effect of important independent variables on partic-
         ulate removal and SO;? removal.

    (2)  To identify and  resolve operating problems, such as
         scaling and demister plugging.

    (3)  To identify areas or regions for reliable operation
         of the three scrubber systems,  consistent with reason-
         able SC>2 removal, and to choose attractive (economic)
         operating  configurations  from within these regions.

    (4)  To determine long-term  operating  reliability with
         attractive  configurations for one or more of the scrub-
         ber systems and to  develop more definitive process
         economics data and scale-up factors.
In order to accomplish the first objective,  it will be necessary to

make a large number of "short-term factorial tests" for each scrubber

system. Based on tests performed to date, it appears that meaning-

ful data on SO  and particulate removal can be obtained during tests
             LJ
lasting  from 4 hours to one day.


In order to accomplish the second and third objectives, a relatively

small number of longer-term (2+ week) "reliability  verification tests"

will be  made on each scrubber system.   These longer-term tests

will also be useful to:
 u
 Closed vs. open liquor loop operation will be discussed subsequently.
                                252

-------
        Obtain more reliable material balances.
        Quantify any variations in SO  and particulate removal
        and system slurry compositions with time.
The fourth objective will be accomplished by running "reliability tests, "

lasting from 4 to 1 0 months,  on attractive operating configurations for

one or more of the scrubber  systems.  During these tests, the systems

will be carefully monitored for potential long-term reliability problems,

such as erosion and corrosion of system components.  The ability to

effectively operate such systems under varying load and SO  inlet

conditions will also be studied during this test period.


Early during the short-term  factorial test period (see Figure 7 ) it

became apparent that it was not feasible to  operate the test facility
                                                            o..
                                                            *i*
in a totally closed liquor loop,  without facility modifications.    A

closed-loop test is a test wherein the raw water input to the system

is nearly equal to the water normally exiting the system from the

humidified flue gas and the waste  sludge transferred to the pond.  In

an open loop system,  raw water input is significantly greater than

the water outflow in the exit  gas and sludge.  Therefore,  process

liquor must be discharged  from the system to maintain an overall
  These modifications were completed during a five-week boiler-outage
  in February and March,  1973 (see Figure  7 ).  The major modifica-
  tions included:  eliminating pump sealing water on the Allen-Sherman-
  Hoff pumps by changing from Hydroseals to Centriseals; humidifying
  the hot inlet flue gases with slurry instead of with raw water; and
  washing demisters with process liquor diluted with raw water instead
  of with raw water only.  Other major modifications to the systems during
  the boiler outage, not necessarily effecting the water balance, are dis-
  cussed in Ref.  1.
                                 253

-------
water balance.  In a commercial system such discharge may not be
acceptable due to potential water pollution problems.  Also, during
open-loop operation reliability may be unintentionally enhanced since
the additional raw water added tends to desaturate liquors returning
to the scrubber, thereby tending to reduce scaling and plugging.  It
is expected, however,  that important SO removal data has been ob-
tained during the  short-term open-loop factorial testing, since other
pilot-scale testing has  indicated that SO  removal is relatively in-
dependent of liquor composition for limestone scrubbing systems.

To date, therefore,  the order of limestone testing has been (see
Figure 7 ):
     (1)  Open-loop short-term factorial testing.
     (2)  Closed-loop long-term reliability verification testing.

4.1.4   Lime Wet Scrubbing Testing

This  test series,  which involves introduction of hydrated lime  (calcium
hydroxide) directly in the scrubber circuit, will  resemble the lime-
stone wet scrubbing test program.  The major difference will be the
absence of any open-loop tests.  Again,  tests will be divided into
three general categories (see Figure  7 ):  short-term factorial tests,
longer-term reliability verification tests and long-term reliability
tests.
                               254

-------
4. 2.     LIMESTONE WET SCRUBBING TEST DESIGNS

4.2.1    Short-Term Factorial Testing

The test sequences for the short-term factorial experiments are all
full or partial factorial designs (with centerpoint replicates included)
based upon the chosen independent variables, their levels and the
restraints of time as outlined in  Figure  7.  The choice  of the inde-
pendent variables and their levels was based upon pilot plant test
results, the restraints  of the systems and results from mathemati-
cal models which relate the dependent and independent variables.

Table 2  shows the independent variables and  levels which are to  be
investigated during limestone factorial testing for the three scrubber
systems.  These variables and levels are always tentative, since the
experimental test program is continuously reviewed and up-dated, as
information is generated.   Intermediate variable  levels for center-
point replicate runs (e.g., stoichiometric ratio =1.5) have not been
included in Table 2.

The results, to date, of short-term (open-loop) factorial testing for
the three scrubber systems are  presented in Section 5.

4.2.2    Reliability Verification  Testing

Assuming that each reliability verification test will last about 2  weeks,
and assuming 1/3 downtime for each system   ( for inspections,  clean-
ings,  etc.), only about six tests  can be made for  each  scrubber  system,
                                255

-------
                                                                                              TABLE  2

                                                       INDEPENDENT VARIABLES AND LEVELS FOR LIMESTONE FACTORIAL EXPERIMENTS
Venturi System
Variable
Scrubber configuration
Gag Flow Rate
Liquor Flow Rate to
Venturi and to Spray Tower*
Stoichiometric Ratio ~
Percent Solids
Recirculated
Effluent Hold Tank
Residence Time
Venturi Pressure Drop
Number of Spray Headers
in Spray Tower
Levels
1 Venturi alone
2 Spray tower alone
3 Venturi and Spray tower
1 IS.OOOacfm
Z 30,000
1 300 gpm
2 600
1 1.00 mole/mole
2 1.75
1 6%
2 IS
1 2. 3 min
2 4.6
3 40
* 60
1 6 in. HO
29
* 12
1 2 headers
2 4
TCA System
Variable
Number of Stages
Gas Flow Rate
Liquor Flow Rate
Stoichiometric Ratio
Percent Solids
Recirculated
Effluent Hold Tank
Residence Time

Levels
1 1 stage
Z 3 stages
1 IS.OOOacfm
2 30,000
1 600 gpm
2 1200
1 1.00 mole /mole
2 1.75
1 67.
2 15
1 4. 6 min
2 30

Hydro-Filter System
Variable
Marble- Bed Height
Gas Flow Rate
Total Liquor Flow
Rate
Stoichiometric Ratio
Percent Solids
Recirculated

Levels
1 ? inches
Z 5
1 20, 000 acfm
2 30,000
1 400 gpm
2 BOO
1 1.00 mole/male
2 1.75
1 6%
2 15

10
Wl
          *  After Mid-May 1973, the maximum liquor flow to the  spray tower will be 1200 gpm.
         **  Stoichiometric ratio is defined as moles CaCO^/mole SO  inlet.

-------
given the restraints of time outlined in Figure 7.  Obviously,  not all
variables which are assumed to affect system reliability can be com-
prehensively studied within the six- run limitation.*

Earlier open-loop factorial runs were all made at high stoichiometries
(greater than 1.75 moles CaCO,/mole SO_ abosrbed) and had scrubber
                              •J          Lt ••- ™ -•"-•--—  •
inlet liquor pH's ranging from 6. 0 to 6. 3.  It is planned that a majority
of the reliability verification tests will be made at reduced  scrubber
inlet liquor pH's (5. 7-5. 9), reduced stoichiometries (1. 1- 1. 5 moles
CaCO /mole SG>2 absorbed) and,  of course,  slightly reduced SO-,
removals.   These "reduced-pH" runs should result in improved over-
all system reliability, because of larger oxidation rates at  the lower
pH's and, of course, increased limestone utilizations  (less CaCO3
added/mole SC^ absorbed).
The higher oxidation rates at reduced pH's will result in improved clari-
fier operation and in a larger percentage of "seed" CaSO4 crystals within
the process slurry**. An increase in the percentage of CaSO4 crystals
will probably result in reliable  operation at lower percent solids recir-
culated and/or in smaller effluent hold tank sizes (residence times),
which is economically desirable.  The increase in limestone utilization
results,  of course, in a reduction in waste mass  solids handling and in
limestone requirements.  These improvements may be offset,  however,
by the  need for larger scrubbers  in order to obtain the specified removals.
 *
    These tests will be supplemented with reliability verification tests
    with the EPA pilot TCA scrubbers at Research Triangle Park,
    N.C. (see Section 4. 4),
    Oxidation of sulfite to  sulfate in the E. H. T. can also be increased
    by air- sparging (Ref.._5).
                                257

-------
 On Tables 3,  4,  and 5,  the proposed limestone reliability verification
 test plans for the venturi,  TCA,  and Hydro-Filter systems are shown.
 For the venturi system,  the  effect of percent solids  and gas rate on
 reliability, with two different demister types, will be  tested.  For the
 TCA system, the effect on reliability of percent solids,  residence time,
 air sparging, and gas rates are tested.  For the Hydro-Filter system,
 the effect on reliability of percent  solids,  gas rate,  and scrubber inlet
 pH (at 10% solids) are  tested.   Solids  separation tests for  the clarifier,
 filter, and centrifuge will be made on the  three systems throughout the
 entire test period.

 The runs listed in Tables 3,  4, and 5 have been listed in the expected
 order in which they will be made.  An attempt has been made to start
 the testing on each system at conditions which are expected to give
 high probability for reliable operation (e.g.  high L/G, high effluent
 residence time).

 As of late April 19*73, the initial  runs on Tables 3, 4,  and 5 for the
 three scrubber  systems have been in progress for about 20 days each
 (see Section 7).

 4.3      ANALYTICAL SCHEDULE

 Samples of slurry, flue gas and limestone are taken during each test
 run for chemical analyses, particulate size  sampling and limestone
 reactivity tests.  Batch samples  of coal are taken periodically for
 chemical analysis.  Locations of sample points are shown on Fig-
ures  4,  5,  and 6.   In addition,  clarifier settling tests,  filter leaf
tests, and filter and centrifuge "operational tests" are also being
                                258

-------
                            Table 3

   PROPOSED TEST PLAN FOR RELIABILITY VERIFICATION
                TESTS WITH VENTURI SYSTEM
Venturi and Spray Tower
Venturi pressure drop = 9 in. H~O
Liquor rate to venturi = 600 gpm
Residence time =20 min

Stoichiometric ratio = 1. 3-1. 5 moles CaCO3/moles SO2 absorbed (expected)
Inlet liquor pH = 5. 7-5. 9

       Percent                                            Expected
Run     Solids        Spray Tower    Gas Rate, Demister   Percent
 No.  Recirculated  Liquor Rate, gpm   acfm      Type   SO2 Removal
1
2
3
4
5
6
15
8
15
8
(a)
(a)
600
600
600
600
1200
1200
20,000
20,000
20,000
20,000
20,000
30,000
A
A
B
B
(a)
(a)
60-70(b)
60-70
60-70
60-70
70-80
65-75
(a)  To be determined from runs 1-4.

(b)  Expected SO2 removal at high-pH (6. 0-6. 3) is from 70-80%.
                              259

-------
                           Table 4
PROPOSED TEST  PLAN  FOR RELIABILITY VERIFICATION
                 TESTS WITH TCA SYSTEM
Three stages of TCA spheres,  5" per stage
Liquor rate =1200 gpm
Stoichiometric ratio = 1.3-1.5 moles CaCO3/moles SO2 absorbed (expected)
Inlet liquor pH = 5. 7 - 5. 9
Expected SO2  removal = 80 - 90%.  For high stoichiometry (> 1. 5) and high
    pH (6. 0 - 6. 3) expected SO2 removal is from 90 - 95%.
Run
No.
1
2
3
4
5
6
7(b)
Percent Solids
Recirculated
15
8
(a)
15
8
15
15
Residence
Time, min.
20 (EHT)
20 (EHT)
20 (EHT)
5-6 (EHT)
5-6 (EHT)
5-6 (EHT)
4. 6 (RT)
Air-
Sparging
No
No
Yes
Yes
Yes
Yes
No
Gas
Rate, acfm
20, 000
20,000
20, 000
20, 000
20,000
30, 000
20,000
(a)  To be determined from results of runs  1 and 2.
(b)  Will be run if time permits.
Note;   "EHT" stands for effluent hold tank and "RT"  stands for
       recirculation tank.
                             260

-------
                          Table 5

         PROPOSED TEST PLAN FOR RELIABILITY
               VERIFICATION TESTS WITH
                 HYDRO-FILTER SYSTEM
Height of marble-bed = 3-1/2 inches

Liquor rate = 800 gpm (total)

Residence time =  30 min

Stoichiometric ratio = adjusted to give desired pH or SO2 removal

   (expected between 1.3 and 1. 6 moles CaCO3/moles SO2 absorbed)
Run   Percent Solids     Approximate       Gas         Expected
No.   Recirculated     pH to Scrubber   Rate, acfm  SOg Removal, %

  1         10              5.7          20,000         70
  2         10              6.2          20,000         80
  3         15              5,7          20,000         70
  4          6              5.7          20,000         70
  5         15              5.7          30,000         70
  6          6              5.7          30,000         70
                            261

-------
conducted periodically.  A  typical analytical schedule for a limestone



\vet-scrubbing test and a summary of the analytical methods for deter-



mining important species in slurry,  coal and alkali are presented in



Ref. 2.








4.4      EPA PILOT PLANT SUPPORT








In support of the Shawnee prototype testing activities, two smaller



scrubbing systems (300  acfm each),  which are capable of operating



over a wide range of operating conditions, have been installed at



EPA's Research Triangle Park,  N.C.  facility.  The small pilot scale



units are capable of simulating the TCA  scrubber circuit and have



already generated large quantities of closed-loop data on certain TCA



configurations.
                                262

-------
                              Section 5
      SHORT-TERM FACTORIAL LIMESTONE TEST RESULTS

 5. 1      SO  REMOVAL RESULTS

 In this section, the significant SO  removal results from the open-
 loop limestone factorial test sequences are presented graphically.
 The complete listing  of all pertinent data, in tabular form,  for all
 of the open-loop runs made prior to the boiler outage (see  Figure 7),
 will be presented in the first interim report (see Table 1).

 The SO  removals during open-loop testing have all been corrected
 for the dilution effect of water vapor and  reheater gas pickup  by the
 flue gas.  The SO  removals have also been corrected for  DuPont
 SO- analyzer calibration errors associated with unstable composi-
 tion of SO- calibration gas (from September 15 to  October  13,  1972)
 and the deterioration of the DuPont analyzer optical filters  (from
 June 23 to December 1,  1972)."

 After mid-November, 1972, when a 60 wt. % limestone slurry addition
        ##
 system  was installed for the three scrubber systems, some  problems
 developed with the calibrations of the limestone additive magnetic
 flowmeters  at the reduced slurry flow rates.  All  of the stoichiometric
 *These correction factors were furnished by the DuPont Co.  The
  corrections were small for SO2 removals  greater than 70%.
**Previously, a 1* wt. % limestone addition  system was in use.
                                 263

-------
ratios after mid-November were higher than their "nominal"  (see Table 2)
values and in excess of 1. 5 moles CaCO3 added/mole SO2 inlet.  For these
high values of stoichiometric ratio, and at the prevailing high-inlet
SO_ concentrations (2300-3300 ppm), the SO. removals were  not
   Ci                                       t*
significantly affected by changes in stoichiometric ratio  (see Section
6. 1 and 6. 2).

Prior to system modifications made during the five week boiler outage,
reasonable material balance closures for calcium and  sulfur could
only be obtained with the TCA system.  During this period of opera-
tion, the venturi and Hydro-Filter  systems still had the  clarifiers
and process water hold tanks included in the  main slurry loops, while
the TCA system had been modified so that the main slurry stream
circulated between the hold tank and scrubber,  with a bleed stream
from the main slurry stream routed to the solids separation area
                                      *
(clarifier, vacuum filter or centrifuge) .  The poor material  balances
for the venturi and Hydro-Filter systems were attributable to solids
build-ups (or depletions) within the clarifiers.  For the TCA scrubber,
the calcium and sulfur leaving the  system could be obtained from the
measured flow rate and  solids analysis of the "bleed stream"  to the
solids separation area, and the clarifier could be excluded from the
material balance enclosure.  Typical system configurations for the
open-loop limestone tests have  been presented in Ref.  2, along with
the material balance results for six TCA runs.  The closures were
all within the limits  of the estimated experimental accuracies during
those periods of time.
*
   During the five-week boiler outage, the flow configurations for the
   venturi and Hydro-Filter were converted to ones  similar to that
   of the TCA system.  These modified configurations have been shown
   in Figures 4,  5 and 6.
                               264

-------
Although satisfactory material balance closures were not obtained
during the open-loop factorial testing for the venturi and Hydro-Filter
systems, confidence in the generated data is based on the following:

     •   "Wet" chemical analyses for SO- in the inlet and
         exit gas streams repeatedly corroborated
         DuPont SO  analyzer measurements.
     •   Sulfur removals in longer-term reliability veri-
         fication runs, with excellent material balance
         closures for calcium and sulfur (see Section 7),
         have been in close agreement with open-loop
         replicate'- runs for the venturi,  Hydro-Filter and
         TCA systems.

It is recognized that SO   removal is affected by SO. inlet gas con-
centration and scrubber inlet liquor temperature (Ref. 6).   Car«,
therefore, has been exercised in  segregating these non-controlled
independent variables in the presentation of the data.

5. 1. 1   Venturi System

In Figure 8, the effect of gas  and  liquor flow rates on SO, removal
                                                       &
for the venturi scrubber with  9 inches of pressure drop is shown.
Only  data taken after December 1, 1972 has been plotted, since there
is some  evidence that SO- removal corrections (for the deteriora-
                        Lt
tion of the DuPont optical filters)  were not accurate at low SO,
                                                           tt
removals.
j_
   Replicate runs are made with identical values for all controlled
   independent variables.
                               265

-------
                             FIGURE 8
           PRELIMINARY RESULTS FOR SCL REMOVAL IN THE
        CHEMICO VENTURI WITH A NINE INCH PRESSURE DROP.
    60
   50 ••
   40  .
i
Ul
otf
 CM
O  30 +
Ul
u
   20 -f
             LIQUOR FLOW RATE = 600 gpm
             LIQUOR FLOW RATE = 300 gpm
         STOICHIOMETRIC RATIO > 2 moles CaCOa/mole SO2 absorbed
   10 -f  SOj INLET CONCENTRATION = 2200-3000 pom
         PERCENT SOLIDS = 5-8%
         HOLD TANK RESIDENCE TIME = 33-70 min.
         SCRUBBER OUTLET LIQUOR TEMP.= 107-120 °F
          15,000        20,000         25,000       30,000
                    GAS FLOW RATE, acfm @ 330°F
                        266

-------
Figure 9 illustrates the effect of gas and liquor flow rates on SO  re-


moval in the four-header  spray tower.   The outlet liquor temperatures


(shown in the figure) varied considerably from run to run.  The effect


of outlet liquor temperature on SO  removal was generally consistent,


with higher SO  removal at lower temperature.  A curve representing
              c*

a median liquor temperature of approximately 100 F has been drawn


for each liquor flow rate.






Figure 10 is a cross-plot of Figure  9, showing the effect of liquid-to-


gas ratio and gas velocity on SO- removal at a scrubber outlet liquor
                               Ct

temperature of about 100 F.  The SO_ removals are outside the range
                                    £*
                                                                 •tf
of interest for commercially acceptable gas velocities (>7 ft/sec). '


The results from Figure 10 appear  to agree reasonably well with the


spray tower data taken by the Hydro-Electric Power Commission of


Ontario (Ref.  7), after correcting for the effects of inlet gas SO»


concentration (see  Section 5. 1.2).




                                 ;'»;'*

A spray  tower limestone depletion   run 463-1A was made to deter-


mine the  effect of stoichiometric ratio and  inlet scrubber liquor pH


on SO_ removal. Results from this  run  are presented in Table «>  and
      £

in Figure 11.  SO^ removals for this run were low because,  at that
                 L*

time, the liquid-to-gas ratio could  not be maintained greater than
   A modification to increase the maximum liquor rate from 600 to

   1200 gpm for the spray tower is scheduled for completion by end

   of May,  1973.  This will allow for a liquid-to-gas ratio of 53 gal/

   mcf at a gas velocity of 7. 5  ft/ sec, which should produce an SO2

   removal of about 65%.


 ##
   A limestone depletion run is a run in which no limestone make-up

   is added during the test period.
                                267

-------
                                FIGURE 9
                   EFFECT OF GAS AND LIQUOR FLOW RATES
             ON S02 REMOVAL IN THE FOUR-HEADER SPRAY TOWER
  70 -•
1(78-92)
  60 ••
$50
£
        (71-103)
        320 gpm
        2-3% SOLIDS
  30 ••
O LIQUOR RATE = 450 gpm
^ LIQUOR RATE = 300 gpm
STOICHIOMETRIC RATIO = 1.5-3.0
PERCENT SOLIDS = 5-9%
RESIDENCE TIME = 40-106 mfn.
SO2 INLET CONCENTRATION
    = 2,500-3,400 ppm
             (98-111)
             (102-114)
             (109-114)
             (79-116)
                                       (97-113)
          LIMESTONE
          DEPLETION
          RUN 463-1A
          HIGH STOICH.
          RATIO
                                                      (113-120)
                          (111-118)
        NUMBERS IN PARENTHESES REPRESENT
          LIQUOR OUTLET TEMPERATURES( °F).
                                                       J80-101)
                                                      (116-123)0-—
  20
         10,000
                     20,000
              GASRATE,acfm@330°F
                         30,000
                            268

-------
                           FIGURE  10
         EFFECT OF LIQUID-TO-GAS RATIO AND GAS VELOCITY
         ON S02 REMOVAL IN THE FOUR-HEADER SPRAY TOWER
70 ••
60 ••
40 -•
30 ••
20
SO2 INLET CONCENTRATION = 2,500-3,400 ppm
STOICHIOMETRIC RATIO = 1.5-3.0
PERCENT SOLIDS = 5-9%
HOLD TANK RESIDENCE TIME = 40-106 min.
SCRUBBER OUTLET LIQUOR TEMP.» 100 °F
                        3.7 ft/sec
                                           2.5 ft/sec
   10
     20
  30        40        50
LIQUID-TO-GAS RATIO, gal/mcf
60
70
                          269

-------
              TABLE 6
SPRAY TOWKH LIMES CONE
      IUIN WITH FOUIl UF.ADKRS
           RUN NO. 4l.f-IA
f
PATE
1/21/71







IN
TIME (5"
noiw w
MOO 7.
1200
lino
1400
1500 6.
li.OO
1700
1806
LET .Sl.UHKV
-IRS
, ,i f) 1 1
. •
70 7. 30



n i.. =>o
(.. 10
6.25
6.20
1900 6.40 6.30





1/24/73





2000
2100
2200
2300 . <:.
2400
0100
0200
0300 7.
0400
0500
0600
(>.2O
(,.20
u.OO
9=, f.. 10
6. 10
1'., 2 »
6.30
93 6, 10
6. 30
6. 10
6.20
O700 8.92 6.15



0800
0900
1000
6. 10
(,.25
!,. JO
1100 9.00 6.10


1200
1300
i,. 50
6. 10
mif.T ^
<:Sc. REM°
ppm
1125 71.
1094 70.
1094 60.
106) 58.
303! 53.
1031 49,
1011 48.
1000 47.
2969 47.
3000 46.
1125 43.
3250 44.
3187 44.
1063 44.
1063 44.
3063 44.
3000 41.
3063 42.
3061 41.
3125 42.
3125 42.
3156 42.
3156 41.
3031 42.
2875 41.
2813 41.
2813 42.
SrOIClHOMETRJC RATIO,
VAL. 10 PAL MOI.RSC.^O,
MOI.K SOj ARSOHBRD
0
0 1H
0 19
0 13
0 10
0 H. 6
0 7.4
0 6.5
0 S. H
» S.I
5 4.H
5 4. S
0 4. 1
5 3.'*
5 l.i.
0 1.4
5 1.2
5 1. 1
0 2.9
0 2.H
0 2.7
5 2.6
0 2. 5
0 2.4
0 2.3
5 2.2
5 2. 1
1400-V
• j System down due to high fan vibration.
210oJ

2200
6. 15
2300 5.95 6.20

1/Z5/7J

2400
0100
0200
6.20
6. 10
t.05
0300 6.60 6.00



0400
0500
0600
6, 10
6.00
6.00
0700 7.95 6.20



0800
0900
1000
6.20
6,00
6.0;
1100 6.62 6.10




1200
1100
1400
1500
6. 10
6.00
2812 51.
2812 47.
2750 43.
2812 44.
2562 41,
5 2. 1
0 2. 1
5 2,0
5 2.0
5 .0
2812 47.4 .9
2750 46.0 .8
2812 47.0 .8
2750 44.5 ,8
2750 44.
2813 45.
5 .7
5 .7
2813 43.0 .6
2813 43.
2500 44.
5 .6
5 .6
2438 43.0 .0
1375 43.0 .5
2375 43.0 .5
5.90
1600 5. B5 5.90


1700
180(1
5.95
5.70
1900 7.50 5.80



2000
2100
2200
5.70
5. 65
5. 60
2300 6.22 5.70

1/26/73

2400
0100
02 00
5.70
5. bO
5.60
0300 6, 7H 5. 50



0400
0500
0400
5.40
S.45
5.35
0700 6.99 5.45



osao
0900
1000
5.40
5.20
5.20
1100 h. 10 S.40

1200
5. 30
2375 40.0 .5
2312 36.0 .5
2000 36.0 .4
2312 36.0 .4
218K 35.0 .4
1812 32.5 .4
1750 30.0 ,4
2000 33.5 .4
1918 31.5 .4
1875 32.5 .3
1688 11.5 .3
1781 31.5 .3
1 906 11.5 . ,
1906 10.0 .3
1813 29.5 .1
1906 29.0 .}
2000 2H.O .)
2344 2f,.0 .3
2344 24.0 .2
2500 23,0 .2
2500 21.0 .2
2500 21.0 .2
          270

-------
                          FIGURE 11
                   EFFECTOFINLETLIQUORpH
       ON S02 REMOVAL IN THE FOUR-HEADER  SPRAY TOWER
              (LIMESTONE DEPLETION RUN NO. 463-1A)
   60
  50  +
§  40  ••
O
*-
Ul
y  30  ••
   20  ••
GAS RATE = 20,000 ocfm @ 330 °F
LIQUOR RATE = 450 gpm
SO2 INLET CONCENTRATION =1,750-3,200 ppm
PERCENT SOLIDS = 6-9%
SCRUBBER OUTLET LIQUOR TEMP. = 97-113 °F
HOLD TANK RESIDENCE TIME = 56 min.
                   HIGH STOICH. RATIO
                     (GREATER THAN APPROX.
                     1.4 MOLES CaCO«/MOLE
                     SO  ABSORBED)
   10  '   I  I   I  I   I  1  |  I  I   1  I   I  I   I  I  I   I  I   |
     5.0          5.5          6.0          6.5          7.0
                     SCRUBBER INLET LIQUOR pH
                            271

-------
approximately 30 gal/mcf.


                                  j-
In Table 6,  the stoichiometric ratio  has been calculated from the

estimation of the original Ib-moles of CaCO  in the system and of

the SO  absorbed.  A comparison between the stoichiometric ratios

calculated in this manner with those obtained from the solids analysis

could not be made, unfortunately, because of uncertain solids analy-

tical results during this period.



As can be seen from Table 6, the SO  removal and inlet liquor pH
                                   L*
remained at steady values of 40-44% and 6. 0-6. 3, respectively,  for

a long period  of time  (about 35 hours)  before the removal and pH

began to drop. The stoichiometric ratios for this period of time were

greater than 1.4 moles CaCO- added /mole  SO  absorbed.  The SO
                            3               £•                  {*
removal for this high stoichiometry region has been included as a

data point in Figure 9.



Figure 11 shows the effect of inlet liquor pH on SO  removal for Run

463-1A as the limestone in the system was depleted.   Similar  effects

of pH upon SO. removal have been reported in Refs.  8 and 9.
   The stoichiometric ratio (moles CaCOj/moles SO2 absorbed) of
   the scrubber inlet liquor changes with time as the SO? is absorbed
   (i. e., one mole of CaSOx is formed and one mole of CO2 is
   evolved for every mole of SO, absorbed).
                               272

-------
5.1.2    TCA System

                      i't
The results of the EPA  TCA limestone runs are summarized in Fig-
ures 12 through 15.


Figure  12 shows the effect of height of spheres (5 and 10 inches;/stage)
and gas rate on SO_  removal in the TCA with six grids and three stages,
                  b
The effect of spheres vs. no spheres in the six-grid TCA on SO~ re-
moval is illustrated  in Figure 13.


The liquor and  gas rate effects on SO  removal in the four-grid three-
stage TCA are  presented in Figure 14.   Figure 15 is a cross-plot of
Figure  14, showing the effect of liquid-to-gas ratio and gas velocity
on SO   removal.
     tL

The variation in SO  removal for 5 inches of spheres per stage in the
six-grid, three-stage  TCA (shown as open circles in Figures 12 and
13) is attributed, mainly, to differing average  values of SO- inlet
concentrations. In Table 7, the SO   removals and operating conditions
                                  £
for these runs  have been compared..


R.H.  Borgwardt (Ref. 9) of EPA has reported  that,  for his pilot scale
TCA (see Section 4.4), the percent SO- removal is inversely pro-
                                    iL
portional to the one-tenth power of inlet SO? concentration.  The
   From November 4, 1972 to January 15,  1973,  TVA has conducted
   a special series of tests with the TCA scrubber to provide process
   and equipment scale-up and design information for the 550 Mw coal-
   fired TVA Widows Creek Unit 8 retrofit limestone scrubbing system.
   The results shown on Figures 12 through 15 do not include the TVA
   tests.
                               273

-------
                                    FIGURE 12
                    EFFECT OF HEIGHT OF SPHERES AND GAS RATE
                 ON S02 REMOVAL IN THE SIX-GRID THREE-STAGE TCA
  100
   95 -
         A (7.1-7.9 In.H.O)
   90 ••   J-
m^

Q  85 4
0(4.4-5.1 ln,H90)
   80  -
   75 ••
   70  ••
                    O"


                     .J
                                 (8.2-9.9 In. H20)
                                  (5.9-7.7 in.H20)
                                                               1(12-15 !n.H2O)
                          LIQUOR RATE = 1 , 170-1 ,220 gpm
                          SO2 INLET CONCENTRATION = 2,400-3,300 ppm
                          STOICHIOMETRIC RATIO = 1.2-2.0
                          PERCENT SOLIDS = 6.5-10.5%
                          HOLD TANK RESIDENCE  TIME = 4.6 min.
                          SCRUBBER OUTLET LIQUOR TEMP. = 111-125 °F

                               HEIGHT OF SPHERES
                                         5 INCHES/STAGE
                                         10 INCHES/STAGE
   65
                          NUMBERS IN PARENTHESES REPRESENT TOTAL
                              PRESSURE DROPS (EXCLUDING DEMISTER).
     1	1	1	1	1	1	1	1	1	1	1	1	1	1	
        15,000
                   20,000                25,000
                       GAS RATE, ocfm@280°F
30,000
                                  274

-------
  100
   95  -•
   90 -•
   85 ••
I
cTso
LU
u
   75 -•
   70 --
   65 --
   60
                                  FIGURE  13
                    EFFECT OF SPHERES VS NO SPHERES AND
                GAS RATE ON S02 REMOVAL IN THE SIX-GRID TCA
1 1 1 1 1
/ 102-1 18 °F
T /
i
X
    (5.5in.H2O)
            (6.2fn.H2O)
L) (2.0in.H20)
                                 M (2.5in.H2O)
                                                   |(3.6?n.H0O)-
         LIQUOR RATE =1,190-1,210 gpm                            .
         SO2 INLET CONCENTRATION = 1,700-2,950 ppm
         STOICHIOMETRIC RATIO ^ l .2-2.0
         PERCENT SOLIDS = 7.5-11.5%
         HOLD TANK RESIDENCE TIME = 4.6 min.
         SCRUBBER OUTLET LIQUOR TEMP. = 110-120 °F (EXCEPT AS NOTED)
               HEIGHT OF SPHERES
           O 5 INCHES/STAGE (3 STAGES)
               NO SPHERES
         NUMBERS IN PARENTHESES REPRESENT TOTAL PRESSURE DROPS
               (EXCLUDING DEMISTER & KOCH TRAY).
                              •H	1	1	1	1	1-
	1	1	1	1	1
       20,000
 GAS FLOW RATE,acfm @ 280 °F
      275
         15,000
                                                  25,000

-------
                                FIGURE 14
                      EFFECT OF LIQUOR AND GAS RATE
           ON S02 REMOVAL IN THE FOUR-GRID THREE-STAGE TCA
  100
                 (3.6 In.
   95 ••
   90 -•
   35 . .LIQUOR RATE=900 gpm'
   80  ••
I/I

1
£  70
   65 -
   60  -•
   55  -•
   50
                                                  T    I
       LIQUOR RATE=1200 gpm
                                    (5.5 ln.H20)	n(7.
                                     (3.8ln.H20)_
LIQUOR RATE=600 gpm
                                                         (4.4in.H2O)
                  SO2 INLET CONCENTRATION = 1,800-2,500 ppm
                  STOICHIOMETRIC RATIO = 1.5-3.0
                  PERCENT SOLIDS = 6-11%
                  HOLD TANK RESIDENCE TIME = 18-35 mtn.
                  SCRUBBER OUTLET LIQUOR TEMP.  = 111-123 °F
                  HEIGHT OF SPHERES - 5 INCHES/STAGE
                                                          «
               NUMBERS IN PARENTHESES REPRESENT TOTAL PRESSURE
                      DROPS (EXCLUDING DEMISTER & KOCH TRAY).
  H	1	1	1	1	1	1	1	1	1	1	1	1	1	
            15,000
                        20,000
                   GAS RATE,acfm @ 280 °F
25,000
                                276

-------
  100
   95 -•
   90 -
   85  -•
0^75  +

=  70-1-
   65  -•
   60 -
   55  ••
   50
                          FIGURE 15
         EFFECT OF LIQUID -TO-GAS RATIO AND GAS VELOCITY
         ON S02 REMOVAL IN THE FOUR-GRID THREE-STAGE TCA
         9.8 ft/sec
                                                sec
SO2 INLET CONCENTRATION - 1,800-2,500 ppm _
STOICHIOMETRIC RATIO » 1.5-3.0
PERCENT SOLIDS = 6-11%
HOLD TANK RESIDENCE TIME = 18-35 mln.
SCRUBBER OUTLET LIQUOR TEMP. - 111-123 °F
HEIGHT OF SPHERES = 5 INCHES/STAGE
     20      30      40       50      60       70
                      LIQUID-TO-GAS RATIO, gal/mcf
                              -4-
                               80
90
                            277

-------
                            Table 7
       EFFECT OF INLET SC>2 CONCENTRATION ON
         REMOVAL IN A SIX-GRID THREE-STAGE TCA
Run No.
SO2 Removal, %
Inlet SOg Cone. , ppm
Gas Rate, acfm @ 280°F
Liquor Rate, gpm
Stoichiometric Ratio
Scrubber Outlet Liquor
Temperature, °F
Percent Solids Recirc.
Hold Tank Res. Time, min.
Ht. of Spheres/Stage, in.
Pressure Drop, in. ^O
409-2A
&
414- 2A
90+3
2800-3250
20, 100
1, 190
1.4-1.6
112-122
7-11
4.6
5
5.9-7.7
416-2A
95+1
1750-2200
20,000
1,195
>1.5
111-118
8-9
4.6
5
5. 8-6. 6
410-2A
b
411-2A
87+3
2500-3150
15, 100
1, 180
1.2-2.0
111-120
7-8. 5
4.6
5
4.4-5. 1
415-2A
95+2*
2250-2750
15,250
1, 200
>1.5
102-118*
7-11
4.6
5
5. 0-6. 0
*High removal may also be due to lower outlet liquor temperature.
                              278

-------
difference of 5% in SO  removal between runs 409-2A and 416-2A,



for average SO_ concentration differences of 3000 and 2000 ppm,



respectively,  is in agreement with EPA's pilot results.  The 8%



difference in the SO   removal between runs 410-2A and 415-2A is



attributable both to the differences in inlet SO_ concentrations and



the differences in  the scrubber outlet liquor temperatures.






5.1.3    Hydro-Filter System






Figure  16 summarizes the effect of gas and liquor flow rates on SO_
                                                                L*


removal in the Hydro-Filter with five inches of marble bed height.



Figure  17 is a cross-plot of Figure 16,  showing the effect of liquid-



to-gas ratio and gas velocity.






The extended  dash-lines showing the ranges of SO_ removal in Figures


16 and 17 indicate where the ranges of SO_ removal would have been
                                        £4

if corrections had not been made (for the deterioration of the DuPont



analyzer optical filters) for these runs.  As mentioned previously,



there is some doubt about the accuracy of these corrections at low



SO  removals.






5.2     ANALYTICAL RESULTS






A comparison between measured and predicted liquid and solids analy-



tical data for the venturi and TCA systems during open-loop testing is



presented in Section 6.2.  Analytical data for the closed-loop testing



is presented in Section  7. 1.
                                279

-------
                               FIGURE 16
                 EFFECT OF GAS AND LIQUOR FLOW RATES ON
     S02REMOVAL IN THE HYDRO-FILTER WITH FIVE INCHES OF MARBLES
2  80 +
 04
8
I-

£  60 ••
 40 ••
 20
                                 TOTAL LIQUOR RATE = 400 gpm
                                 TOTAL LIQUOR RATE = 600 gpm
                                 TOTAL LIQUOR RATE = 800 gpm

                            SO2 INLET CONCENTRATION = 2,400-3,200 ppm
                            STOICHIOMETRIC RATIO = 1.1-1.9
                            PERCENT SOLIDS = 5-7%
                            HOLD TANK RESIDENCE TIME = 50 mtn.
100  +                       SCRUBBER OUTLET LIQUOR TEMP. = 115-125 °F
                  -11 1n.H0O)
*
                                                         (10-12 in.H2O)
               ri                                       T
               U(B-9?n.H20)                            i

                                                        U(9-10Fn.H2O)

                    NUMBERS IN PARENTHESES REPRESENT HYDRO-FILTER PRESSURE
                        DROPS (EXCLUDING DEMISTER) IN A SCALE-FREE BED.

           •I	1	1	1	1	1	1	1	1	1	1	»	1	h—
            20,000               25,000               30,000
                       GAS FLOW RATE,acfm @ 330 °F
                             280

-------
                              FIGURE 17
          EFFECT OF LIQIUD-TO-GAS RATIO AND GAS VELOCITY ON
     S02 REMOVAL IN THE HYDRO-FILTER WITH FIVE INCHES OF MARBLES
  100
   80  ••
0  60
LUI
 CM
-40 +
   20  ••
     10
                 7.7 ft/sec
             SO2 INLET CONCENTRATION = 2,400-3,200 ppm
             STOICHIOMETRIC RATIO = 1.1-1.9
             PERCENT SOLIDS = 5-7%
             HOLD TANK RESIDENCE TIME = 50 mm.
             SCRUBBER OUTLET LIQUOR TEMP. = 115-125 °F
20
 30        40        50
LIQUID-TO-GAS RATIO, gal/mcf
60
70
                               281

-------
5.2.1    Liquid Data

                                                           *
Table 8 shows the average scrubber inlet liquor compositions  for the
open-loop factorial runs.  During the period of open-loop testing,
there did not appear to be a continual build-up of magnesium,  sodium
or chloride ions within the liquor.  The large concentrations of chlor-
ide ions  are attributable to chlorides present in the coal which were
converted to HC1 and absorbed from the flue gas in the scrubber.  A.
Saleem of Ontario Hydro (Ref. 11) has reported similar chloride
concentrations during limestone wet scrubbing tests with flue  gas
from a coal-fired boiler.

Table 8 shows that the venturi and Hydro-Filter systems had lower
overall dissolved solids  than the  TCA system.  This was expected,
since the quantity of input raw water for these systems was greater
than for the TCA system (the  TCA system was more "closed-loop").
Actually, the concentration of sulfate within the TCA liquor is close
to the predicted "saturation" level for sulfates.
   The liquid analytical data are tested for consistency by inputting
   the measured compositions and pH's into a modified Radian
   Equilibrium Computer Program (Ref.  10),  which then predicts
   the ionic imbalance.  For the data shown in Table 8, the ionic
   imbalances were all less than  10%.
                               282

-------
                        Table 8
AVERAGE LIQUOR COMPOSITIONS AT THE SHAWNEE TEST
            FACILITY DURING OCTOBER, 1972

C!** •pii'KI") f* i*
System
Species Concentration, mg/1 (ppm)

S
-------
5.2.2    Solids Data

Analyses of the Fredonia  Valley limestone used at the Shawnee facil-
ity showed an average composition of 90% CaCO3, 5% MgCO3 and 5%
inerts.   Dry sieve analyses showed approximately 90% of the ground
limestone passing through 325 mesh.   A MikroPul* particle  size analysis
showed approximately 7% of the ground limestone less than 3 microns,
30% less than 6 microns and 85% less than 20 microns.

The coal burned in boiler No.  10 during these limestone tests is Old
Ben 24 and contains approximately 18% ash,  10% total moisture, 3.2%
sulfur and 0. 3% chloride.   The analyses of ash from boiler No.  10
showed about 50% SiO2, 18% Al2O3, 16% Fe2O3, 7% CaO,  1. 3% MgO,
1. 3% SO3,  2. 3% K2O,  1% Na2O and 3. 2% ignition loss.
The composition of solids in the slurry is determined by the moles
CaCC>3 added per mole SC>2  absorbed,  the overall percent oxidation of
sulfite to sulfate within the  system and the percent  of fly ash.  The mole
percent oxidations averaged approximately 30% during the  open-loop
factorial testing and the fly  ash comprised from 30 to 50 wt. %  of the
solids for the three scrubber systems.

5. 3     PARTICULATE REMOVAL RESULTS

Particulate removal results for the three scrubbers are presented in
Tables 9,  10, and 11.  Only those data which were  taken at close-to-
isokinetic sampling conditions have been included in the tables. All
 A division of United States Filter Corporation
                                  284

-------
                                               Table 9

                PARTICULATE REMOVAL IN VENTURI AND SPRAY TOWER SCRUBBER
                                     DURING OPEN-LOOP TESTS



K)
00
U1










Run No.
415-1A
414-1D
414-1D

414-1C
417-1A
414 -IE
418-1C
*
453-1B
*
454- IB
: #
456-1 A
Date
11-09-72
11-12-72
11-14-72

11-15-72
12-22-72
12-25-72
12-27-72

12-31-72

1-04-73

1-05-73
Gas
Rate,
acfm @ 330°F
30,000
30,000
29,900

29,900
30,000
30,000
14, 900

14,900

14,900

14,900
Liquor Rate,
gpm
Vehturi 1 Spray Tower
305 0
305 0
305 0

305 0
605 0
300 0
600 0

12 460

12 450

12 450
Pressure Drop,
in. HO
Venturi
9.0
9.0
9.0

6.4
9.5
12.0
12. 5

2.5

0.75

0. 70
Spray Tower
1. 1
1. 0
1.0

1.0
1.0
1. 0
0.2

0. 25

0. 25

0. 25
Grain Loading,
g/scf
Inlet
4. 38
2. 1
3. 32

3.40
3. 38
4. 17
6.39

2.6

4. 62

3. 38
Outlet
0.012
0. 010
0. 013

0.02
0. 012
0.009
0. 114

0. 004**

0.07

0.056
Percent
Removal
99. 7
99. 5
99. 6

99.4
99.6
99.8
98.2

99.8

98. 5

98. 3
 Spray tower only.
"Data point questionable.

-------
                                                   Table 10
                       PARTICULATE REMOVAL IN TCA SCRUBBER WITH NO SPHERES
                                         DURING OPEN-LOOP TESTS
      Run No.
             Date
                Gas
               Rate,
            acfm @ 330°F
             Liquor
             Rate,
             gpm
           Pressure
         Drop, in. H2O
                                                                  Grain Loading, g/scf
                                                                          Inlet
                                                                               Outlet
                                   Percent
                                   Removal
KJ
00
WC-5
WC-5A
WC-5A
WC-11
WC-12
12-21-72
 1-06-73
 1-09-73
 1-12-73
 1-14-73
19,200
19,300
19,300
19,400
19,300
730
730
730
745
375
1. 5
1. 5
1.5
1.5
1.0
                                                               *
1. 70
4.16
1. 32
3.29
3.65
0. 004
0. 029
0.019
0. 017
0. 022
99.8
99. 3
98. 6
99. 5
99. 4
       These listed values are the assumed pressure drops across the scrubber.  Increases in total pressure
       drop for these runs were most likely due to pluggage of the inlet gas duct during testing.

-------
                                                     Table 11

              PARTICULATE REMOVAL IN HYDRO-FILTER SCRUBBER DURING OPEN-LOOP TESTS
Run No.
427-3A
427-3A
426-3B
427-3C
427-3B
428-3A
428-3A
428-3A
438-3A
440-3A
440-3A
Date
11-13-72
11-16-72
11-28-72
12-02-72
12-24-72
12-28-72
12-29-72
12-30-72
1-07-73
1-11-73
1-13-73
Gas
Rate,
acfm @ 330°F
20,000
20,000
20,000
20,000
20,000
20,000
20,000
20,000
19,900
12,500
12,500
Liquor
Rate,
gpm
810
810
810
800
805
810
810
810
400
600
600
Pressure
Drop, in. H^O
12. 0
12.0
10.0
12. 5
11. 0
11.5
11. 5
11. 5
7.0
6.8
6.8.
Grain Loading, g/scf
Inlet
2.6
3. 32
4.43
4. 24
2.19
3. 78
4. 12
3.63
4.20
3. 82
3.59
Outlet
0.030
0. 035
0.032
0.033
0.027
0. 025
0.016
0. 035
0.020
0. 042
0. 066
Percent
Removal
98. 8
98. 9
99. 3
99.2
98, 8
99. 3
99.6
99.0
99.5
98.9
98.2
hJ
00

-------
of the outlet particulate data have been corrected for soot-

contamination from the gas rcheaters.  The soot amounted to less

than 30% of the total mass of the outlet particulates.



During the open-loop factorial testing there were solids accumulations
                                                        *.tf
                                                        'i1*
(depositions) in the demisters for much of the test period.    These

solids accumulations may explain some of the very low measured out-

let grain loadings in Tables 9, 10, and 11,  especially for the TCA

(multi-grid tower) at 1. 5  in. HO of pressure drop.
                              Ci
 During open-loop testing, the demisters were all washed from above
 with raw water (see Figures 1,  2, and 3 in Ref.  2).  During the boiler
 outage,  provisions  were made for washing the demisters from below,
 with a mixture of clarified liquor and raw water,  and for the installa-
 tion of a Koch tray  in the TCA scrubber (see Figure  2).
                               288

-------
                             Section 6
          MODELING OF OPEN-LOOP LIMESTONE DATA

 Results of the open-loop factorial limestone test program are pre-
 sented in Section 6. 1 as equations relating sulfur removal to mea-
 sured parameters for the three scrubber systems.  These "linear"
 equations, which were produced by  a statistical analysis of the data,
 are the simplest available that can adequately depict the data  within
 the range of variables measured.

 In Section 6.2, a theoretical approach is employed, for the spray
 tower, TCA and Hydro-Filter,  to again relate SO  removal to the
                                                c*
 measured parameters.  A general closed-form  equation is developed,
 which is not incompatible with boundary constraints,  and which should
 permit reasonable extrapolations.   Those variables   which were
 found to be significant in Section 6. I  were introduced into the general
 closed-form equations.

 In Section 6. 3, complex mathematical models are discussed for com-
puting SO2 removal  and slurry compositions for the scrubber  systems.
 These models are, generally, systems of partial differential  equa-
tions, which are  solved with numerical computer methods.

Other models for predicting pressure drops, particulate removal and
gas and liquid.side mass transfer coefficients, will be presented in
the first interim report (Table 1).

                               289

-------
6. 1      STATISTICAL MODELS FOR SO2 REMOVAL
Presented in this section are the results of a statistical analysis of

the data from the open-loop limestone factorial runs.  The objective

of the linear equations presented  in this section is to identify the im-

portant independent variables affecting SO2 removal.


Variables may not  appear in the linear equations for a number of
reasons.  Some of  these are:
         The variable did not significantly affect
         removal over the range tested,  i.e. the vari-
         able was not statistically significant in improv-
         ing the fit of the equations over the range tested.

         The effect of the variable was masked by a simul-
         taneous variation of more  significant variables.

         The variable was substantially constant for the data
         set being analyzed.

         The variable was "non-controlled" (e.g.,  inlet gas
         SO2 concentration,  liquor temperature) and may
         not have varied in a manner condusive  to determina-
         tion of its effect on SO  removal.
6. 1. 1    Venturi Scrubber


A stepwise regression analysis of the 10 venturi runs  (with no slurry
in after-scrubber) made in December, 1972, resulted in the following
equation:
                                290

-------
               % Removal = 9.4 + 0. 04 L + 0. 9 Ap            H)


The range of variables covered by these runs,  and therefore the reg-

ion of application, is:
        Gas Flow Rate:  15, 000-30, 000 acfm
        Liquor Flow Rate (L):  300-600 gpm
        Pressure Drop (Ap):  6—12 in. I^O
        Inlet SO2 Concentration:  2, 400-2, 800 ppm
        Stoichiometric Ratio:  1.0—3.0** moles CaCC>3/mole SO2 inlet
        Percent Solids Recirculated:  6-7%
        Hold Tank Residence Time:  33-70 min.
        Scrubber Outlet Liquor Temperature:  112—117 F
Equation 1  accounts for 89% of the total variation of the data.  Gas

rate,  Stoichiometric ratio, hold tank residence time, percent  solids,

inlet SO  concentration and liquor temperature were not statistically
       ti»
significant  (over the above ranges) in improving the fit of the equation.


6.1.2   Spray Tower


The following equation was fit to data from the 15 open-loop test runs

made with the four-header spray tower with no liquid to the venturi

(see  Figures 9 and 10):
   The effect of pressure drop on SO2 removal was only observed be-
   low 9 in. H2O.   Changes in pressure drop above this value (e.g.,
   9-12 in. H2O) had little effect on SO2 removal.

   High stoichiometries O1.75 moles CaCOs/mole SO2 inlet), after
   mid-November 1972, were the result of calibration problems with
   the limestone additive flowmeters (see Section 5. 1).
                               291

-------
                    % Removal = 16 + 0.9 L/C                 (2)

 The range of variables covered by these runs, and therefore the reg-
 ion of application, is:
         Gas Flow Rate:  10, 000-30, 000 acfm
         Gas Velocity:  2.5-7.5 ft/sec
         Liquid-to-gas Ratio (L/C):  13-61 gal/mcf
         Inlet SO2 Concentration: 2, 700—3, 300 ppm
         Stoichiometric Ratio:   1. 0-3. 0 moles CaCO3/moles SO2 inlet
         Percent Solids Recirculated:  2-8%
         Hold Tank Residence Time: 40-106 min.
         Scrubber Outlet Liquor Temperature:  98—128°F
Equation 2 accounts for 95% of the total variation of the data.  Inlet
SO- concentration, Stoichiometric ratio,  percent solids recirculated,
hold tank residence time and liquor temperature were not statistically
significant (over the above ranges) in improving the fit of the equation.
At constant L/C, a gas velocity of 7. 5  ft/sec  gave slightly more
removal (~3%) than 5 ft/sec.   This velocity effect was not observed
below about 5 ft/sec.   Percent solids were actually between 6—8%
for most runs,  with only one run at 2%.

6. 1. 3   TCA Scrubber

The following equation was fit to the 31 EPA TCA runs (see Figures
12 through 15):

     % Removal = 47 + 0. 034 L +  1.4 PSR 4- 0. 5 Hp - 0. 006 ppm  (3)
                                292

-------
The equation accounts for 85% of the total variation of the data.  The
range of variables covered by these runs, and therefore the region of
application, is:
         Gas Flow Rate:  15,000-27, 500 acfrn
         Gas Velocity (V): 6-11 ft/sec
         Liquor Flow Rate (L):  600-1,200 gpm
         Percent Solids Recirculated (PSR):  6-14%
         Inlet SO2 Concentration (ppm):  1, 800-3, 200 ppm
         Stoichiometric Ratio:  1. 0—3. 0 moles CaCO3/mole SO2 inlet
         Hold Tank Residence Time:  3. 8-35 min.
         Scrubber Outlet Liquor Temperature:  105—122°F
         Number of Grids: 4, 6
         Total Height of Packing (Hp): 0-30 in.
Percent removal decreases with increasing inlet SO_ concentration
                                                 £*
(~6% per 1000 ppm).  Gas velocity, Stoichiometric ratio, hold tank
residence time, scrubber outlet liquor temperature,  and the number
of grids were not statistically significant (over the above ranges) in
improving the fit of the equation.  Although not an independent varia-
ble, the pressure drop in the scrubber was also examined and was
found to not have a  significant effect on removal.  For example,
runs giving  92% removal have been made at pressure drops of 4, 6
and 9 in. H2O.

The 17  EPA six-grid TCA runs were analyzed as a group.  The follow-
ing equation was fit to these runs (see Figures 12 and 13):

       % Removal = 67 + 0. 02 L + 1. 0 V + 0. 44 Hp -  0, 006 ppm  (4)

The equation accounts for 78% of the total variation of the data.   The
analysis is restricted to the previously mentioned range of variables,
                               293

-------
 with the exception of percent solids recirculated,  which only varied
 from 7—10%.  Note the gas velocity term for this group, which was
 not  statistically significant for the entire set of runs  (see Eq. 3).
 Again,  stoichiometric ratio,  hold tank residence time, and scrubber
 outlet liquor temperature did not significantly affect SO  removal
 over the above ranges.

 The 14 EPA four-grid three-stage TCA runs were also analyzed as
 a group.  The following equation was fit to these runs (see Figures
 14 and  15):

          % Removal = 53 + 0. 04 L + 1.4 PSR - 0.007 ppm       (5)

 The equation accounts for 96% of the total variation of the data.   These
 test runs were made with 5 inches  of spheres per stage, for a total of
 15 inches.   The range of variables is otherwise the same as that for
 the 31 test run group.   Gas velocity, stoichiometric ratio, hold  tank
 residence time, and scrubber outlet liquor temperature did not  sig-
nificantly affect SO_ removal over  the above ranges.
                  £+

 The following equation was fit to the 16 runs made without spheres in
the five-grid TCA tower (TVA tests):
               % Removal = 90 + 0.034 L-0.46 T_             (6)
                                                 Lj
The range of variables covered by these runs, and therefore the reg-
ion of application, is:
                                294

-------
        Gas Flow Rate:  19, 000-30, 000 acfm
        Gas Velocity:  7.5-12 ft/sec
        Liquor Flow Rate (L):  375—1, 070 gpm
        Inlet SO2 Concentration:  2, 200—3, 200 ppm
        Stoichiometric Ratio: 1. 0—3. 0 moles  CaCC^/moles SC>2 inlet
        Percent Solids Recirculated:  14%
        Hold Tank Residence Time:  5—15 min.
        Scrubber Inlet Liquor Temperature (T  ): 63—110  F
        Scrubber Outlet Liquor Temperature: ^9—115 F
        Pressure Drop:  1—7 in. HO
Equation 6 accounts for 65% of the total variation of the data.  Stoi-
chiometric ratio, inlet SO- concentration and hold tank residence
time were not statistically significant (over the above ranges) in im-
proving the fit of the equation.
6.1.4    Hydro-Filter Scrubber

                                                     *
A stepwise regression analysis of 27 Hydro-Filter runs  resulted in
the following equation (see Figures 16 and 17):


                 % Removal = 17. 9 + 0.1  L - 2. 0 V            (7)


The range of variables covered by these runs, and therefore the reg-
ion of application, is:
   Eleven runs made during October, 1972, were excluded from the
   analysis due to doubtful low values of SO2 removal obtained dur-
   ing this period.   Recent closed-loop data has validated this
   exclusion.
                                295

-------
         Gas  Flow Rate:  12,000-30,000 acfm
         Gas  Velocity (V):  3-8 ft. /sec.
         Liquor Flow Rate (L): 200-800 gpm
         Inlet SOZ Concentration: 2, 000-3, 500 ppm
         Stoichiometric Ratio:  1.5—3. 0 moles CaCO3/mole SO2 inlet
         Percent Solids Recirculated:  6—12%
         Hold Tank Residence Time: 50 min.
         Scrubber Outlet  Liquor Temperature:  85—125 F
         Height of Marbles:  3-5 in.
Equation 7 accounts for 94% of the total variation of the data.  Inlet

SO- concentration, Stoichiometric ratio,  percent solids, liquor tern.

perature, and height of marbles were not statistically significant

(over the above ranges) in improving the  fit of the equation.
6. 2      CLOSED-FORM CORRELATIONS FOR PREDICTING SO-
         REMOVAL
Analysis of the closed-loop limestone data, using the Radian Equilib-

rium Computer Program (Ref. 10), has shown that the equilibrium
                                          Jf
mole fraction of SO_ over the bulk liquid,^- , is negligible with res-

pect to the SO- mole fraction within the gas, for the  spray tower,  TCA
             *             *
and Hydro-Filter Scrubbers .  For this condition, the following rela-

tionship represents SO- absorption:
   Due to low liquor residence times,  the amount of limestone dis-
   solved within the venturi scrubber is relatively small.  Hence, y-
   can be significant.  Also, U/f can be significant for the spray tower,
   TCA and Hydro-Filter scrubbers operating at low stoichiometries
   (<1.5 moles CaCOs/ mole SO2 absorbed) and, consequently, low
   inlet liquor pH's  «6.0).

                                296

-------
where
             - gas -side mass transfer coefficient
             = gas -liquid interfacial area per scrubber volume
             = axial distance
             -gas rate per cross. sectional area
             - Henry's law constant
             - liquid- side mass transfer coefficient

Also, scrubber computer models utilizing previously fitted gas-side
mass transfer coefficients (see Ref.  2) have shown that liquid-side
resistance controls (i.e. , -f?t/tfZ
-------
          T~ = liquor temperature,  F.
The liquid-side coefficient is expected, therefore,  to be a function of
gas and liquor flow rates, scrubber configuration,  amount of dissolved

reactant, inte

temperature.
                                                          -i-
reactant,  interfacial concentration of dissolved SO?  (H  SO )   and
                                                 £*   C*  j
The form of Eq.  9 has been fitted (by multiple regression) to the open-

loop limestone data  for the spray tower, TCA and Hydro-Filter scrub-

bers.  The  significance of the independent variables in the fitted equa-

tions was demonstrated by the statistical analysis (see previous

section).


The open-loop limestone  data was all obtained at relatively high  stoi-

chiometries (greater than 1.5 moles CaCO  /moles SO  absorbed) and

consequently, high scrubber inlet liquor pH's (from 6. 0 to 6. 3).  Within

this regime of operation, stoichiometric ratio showed no  apparent  ef-

fect upon SO- removal (see previous section).  The effect of stoichio-
            £+
metric ratio  for the scrubber systems  will, hopefully, be obtained

during the closed-loop testing now in progress.  Other variables

which showed negligible effects upon SO- removal during  the  open-

loop testing, such as percent solids recirculated,  may also have more

significant effects at reduced stoichiometries during closed-loop

testing.
*                                                  \j
   As the concentration of SO, in the gas increases, XA increases and
  ^L decreases.  This is an explanation for the empirical fact that
   as the SOg inlet gas concentration increases, for limestone wet-
   scrubbing systems,  the SO£ removal decreases.
                                298

-------
The effect of inlet gas SC>2 concentration (a non- controlled independent
variable) upon SC>2 removal has only been included in the fitted equa-
tion for the  four-grid three-stage TCA, although it is presumed that
a similar effect exists for the other scrubbers.  Also, the effect of
inlet scrubber liquor temperature (a non-controlled independent vari-
able), which was determined to be significant from the TVA TCA runs
(see Eq. 6), has not presently been included in the closed-form equa-
tions.  The  effects of inlet gas concentration and temperature will be
included in the final forms of all the correlations,  once the analyses
of other  pilot  data and the Shawnee closed-loop data has been completed.

6.2.1    Spray Tower

The following equation was fit to 15 open-loop  limestone test runs
made with the 4-header spray-tower (see Figures  9 and 10 and
Eq. 2):
where:
             = liquid to gas ratio in scrubber, gal/mcf
Equation 1 1 accounts for 94% of the variation in the data (correlation
coefficient of 0. 97) with a  standard error of estimate of 3. 7% SO,
                                                              &
removal.
                               299

-------
 6.2.2   TCA
 The following equation was fit to 11 open-loop limestone test runs

 made with the four-grid,  three- stage TCA scrubber (see Figures
                    *
 14 and 15 and Eq. 5):
                                            »'-a*/  ,    \a^1
                                            L   ("K*    J
where


         L   = liquor flow rate per cross -sectional area,  gpm/ft


                 _ concentrati°n *n inlet gas, mole fraction
Equation 12 accounts for 99% of the variation in the data with a stan-


dard error of estimate of 1. 3% SO- removal. As previously men-
                                 fit
tioned,  it is assumed that the measured effect of ffSO*.  for the TCA

scrubber will be similar for the other systems.




6.2.3    Hydro-Filter




The following equation was fit to 27 open-loop limestone test runs

made with the Hydro-Filter scrubber (see Figures 16 and 17 and Eq.


7):**
   Two runs at relatively high weight percent solids and one "lime-

   stone depletion" run were eliminated from this analysis.


*#
   Eleven runs made during October,  1972, were excluded from the

   analysis due to anomalously low values of SO£ removal obtained

   during this period.  Recent closed-loop data has affirmed this

   exclusion.



                               300

-------
Equation 13 accounts for 95% of the variation in the data with a stan-
dard error of estimate of 4. 1% SO, removal.
                                 Lt
6. 3      COMPUTER MODELS FOR PREDICTING SO, REMOVAL
         AND SLURRY COMPOSITIONS
6. 3. 1    Scrubber Models

In Ref.  12, mathematical models were presented for predicting SO-
removal in the venturi, TCA and Hydro-Filter scrubbers.  The models
are, generally, sets of partial differential equations which describe
SO. absorption into the process liquor (in accordance with the two-
film theory), 'reaction between the absorbed SO, (H SO  ) and the
species in the liquor and the dissolution of solids (e.g., CaCO.) with-
in the process  liquor.   The assumption has been made,  for these
systems, that the liquor is at all times in equilibrium with an inter -
facial vapor pressure  of 0. 1 atrn  of CO,, i. e., the rate of absorp-
tion of CO, from the flue gas is large. The thermodynamic equilibria
for the models are obtained from the Radian Computer Program
(Ref. 10).

Computer models have been written for the three scrubber systems,
which numerically solve the systems of differential equations.  It has
been planned to fit the gas and liquid-side mass transfer coefficients
to the high and low-concentration soda-ash data (see Section 4. 1.2)
and then fit the solids  dissolution rate constants to the limestone data.
The fitting of the gas.side coefficients for all three scrubbers has
been presented in Ref. 2.   To date, only the liquid-side coefficient
                                 301

-------
for the venturi scrubber has been fit to the low concentration soda-ash
data.  The results of this fit will be presented in the first interim re-
port (Table 1), along with the correlations for the gas-side coefficients.

As discussed in the previous  section, the equilibrium mole fraction
of SO? over the bulk liquid is essentially zero  for the spray tower,
TCA and Hydro-Filter scrubbers, for the open-loop  (high-stoichio-
metry) data.  For this regime, therefore,  the models describing SO_
absorption  for the three scrubbers can be greatly simplified (see
Eq. 8).  For the venturi scrubber,  however, the residence time of
the liquor is low (<0.1 sec.), the dissolution of limestone within
the scrubber is small and,  consequently, the equilibrium mole frac-
tion of SO, over the bulk liquid not  zero everywhere, for the ranges
         £•
of variables tested. Results from the venturi computer  model,  using
the previously fitted gas and liquid-side mass transfer coefficients,
have shown that an assumption of zero dissolution of solids will give
a reasonable fit to the open-loop limestone data.   These results will
be presented in detail in the first interim report (Table  1).

6.3.2   Simulation Model

The simulation model  is a computer model which determines the slur-
ry compositions of the waste streams and the  scrubber inlet and out-
let streams for the three scrubber  systems.   The major assumption
in the model is equilibrium between the liquid and solids in the slurry
   This implies that the kinetic rate of dissolution of limestone within
   the scrubbers is high.
                                302

-------
leaving the effluent hold tank, at a specified equilibrium partial pres-

sure of CO  .   The equilibrium relationships between the liquid and

solid species are obtained from the Radian Equilibrium Program.


The simulation model takes as imput all of the  independent variables

(e.g. , stoichiometric ratio),  the percent sulfite oxidation, the percent

ash in the solids and  the concentration of chloride and magnesium in

the process liquor.     If a scrubber model (either simplified  closed-

form or computer model) is used,  the  simulation model will (itera-

tively) predict SO- removal,  as well as the slurry compositions.  If
                 L4
no  scrubber model is used, then SO removal must be input into the

simulation program,  along with the independent variables.


Two results from the simulation model will be  presented here.  The

first simulation, for  venturi runs 419-1A and 421-1A, is shown pic-

torially in Figure 18  and incorporates  the venturi  scrubber  computer

model (which assumes zero dissolution of solids).  The predicted

removal of 45%  is close to the average measured values of 42 ± 5%

(see Figure 8 and Eq. 1).  The second simulation, for TCA run

412-2A, is shown pictorially  in Figure  19,  and does not incorporate

a scrubber model (the measured removal of 96% was input to the
   The specified CO2 partial pressure was chosen to match the mea-
   sured E.H. T. outlet liquor pH's and compositions.  Predictions
   with the Radian program indicate relatively constant CO? equilib-
   rium partial pressures from 0. 05 to 0. 1 atm.

*** *£
""  Ultimately, models will be  developed for predicting sulfite oxi-
   dation, ash in the solids and the concentrations of chloride and
   magnesium in the liquor.
                                303

-------
                                                                                FIGURE  IS
                                                  BECHTEL LIMESTONE / LIME WET- SCRUBBING SIMULATION PROGRAM
JOjttMOVAL-   45%
 SIMULATION OF VENTUtl SOtUlK* SYSTEM KUNS
    419-lA/ttt-IA ON DECEMKR 23-25, 1*72.
 KAL SCtUUH MOOCl USED IN SIMULATION.
Scnibbw Inlet Uquor lot. •  600 PJM ( 1 4,9
Go. lot.- I5,000ocf«.( 1.600t>-w>le/W>@ 3»°F
Inlet SO 2 Concentration .  2, 900 par*
Stolchloinerrlc Ratio «  1 .5
L/C-   Mool/mcF
Percent Solid* Reclrcuwled -  6 %
Percent Soli* (Xtetwraed -  14 %
Percent Flyori) In Soil* -  40%
Percent Sulflfe Oxidation -20%
Concentration of Chloridd *  14 pnt-moli/IOOO liten
Scrubov T«np«aiuni < II7°F
MAX* ASSUMPTIONS:
o Effluent Held Tonfc Slurry in evitltbrtim
  wlnSO.05ott.COj.
e G»SO3 & CaSO4 oancenrrationi In liquor
  drmm leaving E.H.T. or» 1 J. 1 time* rtvt
  evulllbrtum lalwrotian levelt. nwpectiwly.
e Seiubber Sluny (In contact with flue-gal)
  In equilibrium with O.I arm CO., i.e.
  rot* of frontier of CC>2 it l«nj*.
• Zero ditnlution of iclidt In icnibbor
  fall diuolution of »li-«ole/nt C02^^


X

V r-
SPt
Total
Total
Total
Toto
Total
Toto
H2O
6.0 Ib-fwU/nr CoCO
0 Ib-mole/hr MgCC
^^-»H«3.4

,
Flow Rate, Ib-mole/hr
5reC>" UquW | Solid
Totol Co 9.7 109.3
Total MO 0 0
Ton! COj 0.6 78.0
TotolSOj 2.2 27.1
Total SO. 6.6 4.2
Total 0 4.3 0
MjO 16,900
1
„.„ Flow Rote. *t-«ole/nr
Clt- Liquid 1 Solid
Co 10.0 109.1
MQ 0 0
CO2 0.9 78.0
SO2 0.7 26.9
SO3 6.6 4.2
Cl 4.3 0
16,900

3 1 >• 1 . 4lb-mol.Ar CO j
EFFLUENT




WASTE DISPOSAL Wmre Stream

\
«dES ^%
Total Co 6.0
Total Mg 0
Total CO2 4.3
Total SOj 1.4
Total SO] 0.3
Total Cl 0.1
H2O 370
1 Mot«-uo Water
140pm

-------
  SOj REMOVAL -   96 %
    SIMULATION Of TO SCtUUER SYSTEM RUN
      4J2-2A ON OCTOBEt 29-31, I972t
    REAL SCRUBBER MODEL NOT USED IN SIMULATION
      (SO. REMOVAL INFUT TO PROGRAM.)
o
VI
 0.4fc-«oMirSO2
 300 Ib-mle/hr CO,""* "I
9.0tt>-i»VWS02    J
300 fe-molo/W C02 ~
                                FIGURE »
   BECHTa LIMESTONE , LIME WET-SCRUBBING SIMULATION PROGRAM

              Scrubber Inlet Liquor Rote »1. ISOgpm (32,700 Ib-mole/M
              Ga Rota = 27,SOOaefm ( 3.000 Ib-mole/W) @  290°F
              Inlet SO2 Concentration =  3,000p(xn
              Stoichiometric Katie - 1.25
              L/G =   55 gol/mef
              Percent Solidt Reciroulated - 8.5%
              P*rc««> Soli* Oivctwx^w) =  18%
              Percent Flyah in Solldi « 30%
              Percent Sulfite Oxidation =  30 %
              Concentration of Chloride! * 42.5gm-fiiole/10QO liten
              SaubUr Tempemlure =  120°F
                            I
                                                    MAJOR ASSUMPTIONS-.
                                                    e Effluent Hold Tank Slurry in equilibrium
                                                      wtthO.OSatmCO,.
                                                    e CbSOj & CoSO4 eonc*n*raHoni In liqwx
                                                      itream leaving E.M.T. are I & I  fimei the
                                                      •avrllbrtum tatuntlon leveli, reipecflvely.
                                                    e Scrubber Slurry On contact with flue-gal)
                                                      in equilibrium •!* 0.1 atm CO   i.e.
                                                      rate of tranler of CO2 li lanje.
                                                pH-e.3
                                                                 SPECIES
                                                                          Flow Rare, Ib-molt/V
                                                                Total Co
                                                                Total Mg
                                                                Total CO 2
                                                                Total SO2
                                                                Total SO3
                                                                Total Cl
                                                                H20
                          Liquid   |   Solid
                           25.4
                            0
                            1.6
                            1.2
                           tl.S
                           25.0
                          32,700
321.«
 0
 72.2
177.9
 71.4
 0
11.3 lb-mol.Ar CoCO j <
   0                 *
     Ib-mole/hr CO j
     Nor Computed
                                            pH = Not Computed
SPECIES
Total Co
Total Mg
Total CO2
Total SO2
Total SO3
Total Cl
HjO
Flow Rote,tt>-mole/Kr
Liquid | Solid



Not Computed



                                                                          EFFLUENT
                                                                         HOLD TANK
                                                                                                                                                                     Waita Stream
                                                                                                                                                                        1990"
SPECIES
Total Co
Total Mg
Total CO2
Total SO2
Total SOj
Total O
H2O
Flow RoM,
Ib-moleAr
11.3
0
2.6
6.0
2.6
0.4
4tt>







                                                                                          Make-up Water

-------
program).   The agreement between the predicted and measured scrub-
ber inlet slurry compositions for this TCA simulation is shown in
Table 12.
                               306

-------
                           Table 12

   COMPARISON OF MEASURED AND PREDICTED SLURRY
 COMPOSITIONS AT SCRUBBER INLET FOR TCA RUN 412-2A
                Gas Rate = 27,500 acfm

                Liquor Rate = 1, 170 gpm

                L/C = 5? gal/mcf

                Pressure Drop = 14 in.

                Three stages,  5 inches/stage
Species
pH
SO/
CO/
so/
Ca + +
Mg+ +
cr
Species Concentrations, gm mole/1
Liquid
Measured 1
5.9
1.8
1.2
24
35
5.5
43
Predicted
6.3
2. 1
2.7
19
43
0*
43*
000 liters
Solid
Measured
—
21

0
220
86
51
2

0
1

1 Predicted
—
300
120
120
550
0
—
:': Input to computer model.
                             307

-------
                             Section 7



           RELIABILITY VERIFICATION TEST RESULTS





As mentioned previously,  the objects of the reliability verification


tests are to (1) identify areas or regions for reliable operation con-


sistent with reasonable SO  removal,  (2) choose attractive operating


configurations from within these regions, (3) obtain more reliable


material balances, and (4) quantify any variations in SO  and parti-
                                                      L*

culate removal and system slurry compositions with time.
As discussed in Section 4. 2. 2,  the initial tests are to be run at re-


duced scrubber inlet liquor pH's (5.7—5.9), to increase system relia-


bility and limestone utilization.  A modest reduction in SO_ removal
                                                         Ct

(from high pH performance) is the price of the increased reliability


and limestone utilization.





Presently, initial runs are in progress on all three  scrubbers (see


run Nos.  1 on Tables 3, 4 and 5).  The Hydro-Filter (Run 501-3A)


was  sUrted.up on March 14,  the TCA (Run 501-2A) on March 22 and


the venturi-spray tower (Run 501-1A) on April 9,  1973.





The performance data for  these three reliability verification runs


will be presented in Section 7. 1 and the results  of material balances


for sulfur and calcium  (which were satisfactory) in Section 7. 2.
                               308

-------
 7. 1     PERFORMANCE DATA

 Data for the first 400-500 hours of operation on the initial runs is
 summarized in Figures 20,  21 and 22.  The upper section of each fig-
 ure shows the operating periods (blank space indicates shut-down),
 and such critical variables as SC>2 removal, liquor pH and stoichio-
 metric ratio.  The middle plot gives some analyses of solids in the
 scrubber inlet liquor.  The lower plot gives concentrations of some
 dissolved species in the scrubber inlet liquor.

 Also shown on Figures 20,  21 and 22,  are the depletion (line-out)
 periods for the tests.   Fresh limestone slurries  (no CaSO  "seeding")
 were introduced into the effluent hold tanks  and SO-  absorption was
                                                 ft
 used to reduce the slurry pH until the desired level of SO-  removal
 was attained.  This  level was approximately 10%  below that attain-
 able in high-pH open-loop operation (see Section 4.2.2).

 Before beginning limestone addition, the systems were inspected to
 be sure that they were free of scaling or erosion  that might have
 occurred in the high-pH period of the  line-out. Periodic (approxi-
 mately weekly) inspection shut-downs are scheduled  in order to moni-
tor scaling and erosion in sensitive areas of the systems.

Operability  and reliability of the three scrubber systems during  the
initial runs  is discussed in another paper presented at this  symposium
(Ref. 1).
                               309

-------
                                       Ol£
                       •HOMUHtl-IA
  jpnif
  nW.\
          r*»
, i  8
10 'I  .?  !L g fr

  8 »  > 7, §
      n
  . .  -+>
  3  O


        1
ft)
ro
o
o
-a
a>
3
LQ

s-
sr
o



n>
-5


3


Ol

O
    3' '- " I 1  go
        n t— is*

        fi §
        ^-» 3 3 IX*

        ? I" a.P

        5  3
        i
                           • TOfAL WHtHVIO lOHt» «- MAGNESIUM («• ** )

                           O CAtCWM fU ** )    fl iOOK»*CN«*l
                             cmMMia-l
                                      O o«iON»n ico,, • I
                                              CALEMOAI MY

-------
                                                                                   Ll£
z  ™  -g  -if
?   i  §   s
o   -  -<  x
.0'  .o'
 g   S
                             I
    txi
    Q   — •   -i
             I   I
                     o  5   8
                     -s1  <   a
                     1  s   i
     3   3'
             8

            •7
              i
             tn
             o
 -n
 _>.

«§
 O

 ~
 i

 m
 Q-
  n>
  O
  -s
   r+
   C
   -s
   -5




   01
   O
              T  —•   O"  	
              i5   °-   o-  3_

             CQ   n   5   «
              ffl   O       "*
             ^  3   J  on
             7^  £^-   —  /^i
              ^-ro  =*  M
              °-  §  "°   *

              P  i   "   f

              n   -    N>  -
                      SO  r*0
                       I   O
                  °  K)

                  I  ,  
-------
                                                        ZL£
  ;  I S  O -g  S
  *  -, <  ii  ?  g"

  «?    5" 8 2* *

  1  *  I  O
        =•53  o
        n  c  c  "
        g" B: f 3;

        *  j? *  *

        D I i- O
        5  0. "* M
        I <; JF 
 O-
 01
 o
-s
c


CJ1
o
  t\3
         O  I  II  O
         O  —•     I
1
z
s
ff
2/40
z Q »
2s
5 ^ 10
n - 0
-X A
\ / \
\ / x — ^--
                         ill  .
                         e y o  8
                         1 ^ »

                         ill
Clli CONC(

INSCIUMf

SOtlOS,


5 B
                                     \
                                  1/73   t   V73  I   J^   I  I/SI
                                                                TEST tlMt, kun

                                                               ^4  I  V17  I   3XII

                                                                O.L1NOAI UV
                                       * TOTAL OtISQLVCD SO4.HM  *  UAGNiSHJM 4M«  )

                                       O CAIOUM Ki ** t     ^  IOWUM 0* * 1

                                       D SUUATI 80 'I      •  JUIFITI (SOj- )

                                       A OlOtHK (Cl- 1      O  CAOOHAH fCOj - I
                                                                                         CT
                            ?   »
                                                                                                           ,.i
                                                                                                           ion 5
                                                                          8

-------
    T  •-
        3C  2

            <
    Q   TO  Q
   fi   •   A

    S   s'  3"
    •   _  o>
   -S   i'  y
    r  *  f
   2   "
   S S
                        ©
        3

        9
-s


PO
 o
 o

 rt
 CD
 o.
 o
-a
 n>

 2
 a.
 Cli
 01
 o
 -I
 o
 -s



 cn
 o
 ro
 3=>
             ?£  if
                O  5   5"
                o  _   —
                         n
                         S

                -
             cj
             S t
                     -n°  3
                     ^
                                                        TOIM 611501VID HXIW  *  «»O«SIU*I 
-------
    :i ™ p C C  O
    3  x 8  £|   1

    JL H #  "  i  f
    3T  30 O  O

    fill
    28  i ^ -"
    3  «
    «  -I °°
    f j  3
    I   .
      •  "
      "  s
       s
       3
     1  r
         ©
                 a i« «
                           |>
                           i
                                           TOTAl, EXCLUDING KMISTM t KCWTH r*A*




                                         -DCMISTEI & ICOCM Iwv
(£1



 CD


 ro
           o  r

           M* -tl
           O  C
      s~  o
      -» M
                                                     •\
                                                      ^OUTIEI fL>ll
 O
 o
 3




 O.
i^sl  -»
o  8  '„  §
§  -  -  i
rll
 o
T3
 CU

 f+
 —i.

 3
 CL

 fa
 a>
 -s
 3



 CJ1
  I
 r-o
                                                           HIT TIME, ham

                                                        4/13  I   4/l<

                                                           CAUNDAI OAY
                              •  TOTAL DISSOIVIO SOLICH • MAONI1HIM (Ma " 1


                              O  CALCIUM {fa** )      6 1QOHJM [N= ' )


                              D  SOLFATIBO^-)      • SUlFlnOO,')
                                            CWOtlOE Kl - )
                                                          O  CAUONATE (CO3 p I
                                                                                                                 '•""i
                                                              &


                                                          .   t"0.
                                             4/13   li   4/IJ   1   4/14   I   «/13   I   '''^

                                                                      CALENQU OAV

-------
                                                               SLE
   *5 J?1  r
   r y |- a  f  «

   f §" "^ >  1  sS
   f S  « 5.  8  I
   ' =* -    f  «
   « 3  f     a  |

   '= i S     I  ©
      ®        ^  W
      S.           IS
 ro
 ro
 ro
 O
•a
 (T>
 -s
o
-s
Q.


O
ff
-5

-S
c


S
?f  51  ?
n  c   c   »

§" £  f  ^
i  e  I  *

o I  s-  o
? 1 i
-    2
          P  I
                                                                      HiT TIMi, k..

                                                                      3/19   I   V30

                                                                      CALErNDAI DAY
a.
00
00
                                        *  TOIAI. OIHOLVtO ICXHO


                                        O  CALCIUW fCa ** )
                                           CHICHIW (Ct - )
                                                             SODIUM [Na " )
                                                          O  OUKDNATE (CO., • 1
                                                                                                                  1,000  O
                                         »     «
                                               3/U  1    V\7  1   J/18
                                                                                        A~.

                                                                                          '
                                                                                              i/n   i   J/23   i
                                                                      CALENDAR DAY

-------

„
1 * f i f §* if
« -1 ^ " * ? ° •
y *> g « ? 3
i 8 Q. « S » j "
x °- *^ — " &j a" i * «


3- 8 *« 1 8 §
T =1 ^ f o
<* 1 ^ ? ?
= i* f® , »
r f8 js-
3 o | ^
-n r
-J j t
3 ii -
ro " "
no
D r: }C p '••
--- s' i a s g j | ,..
Q r s: f j- Ml,
O  o~
3 » 9 1 ° 6 •". ,„
r-+ • — ^ 3 <" - ? ^ '
n g- 5- o
:± £T £ ° w
C S i. I' ?
n> 2 ^ i I
Q. "~* ii "" • • /
^~* s v[ 1 M Isi"
g-g-o- azaJO
"^ n ^ ~~ ' ^ 3 5 °

o> 3tK^ § s ' n
(-h ' 0 S ° ? -'
N 3 "n S • ' to
,3 f 2" ill •
^ J > I ^ = •
Q.
& rt. '
I- ' '
Q5 1 1 '
-« If8 .
Q. s «
3 lh »
4 Jig.
in sss
— " 5- »
(D "M
-s

.-s
14,000
tn
o ta,M
i
s 0
3 > '•"
i. I '•""
^
oo 1 ••""
ca :
. I i.«
3
1 •
3
a MO
£
1
i *°°
•e

»

IOC
o
9L£
fND (UN KI-M ! ! ItGlN RUN Ml -31
^^^^ SYSIU* DWINtD TO tINt-OUT IMSKCTIOM
V\
\ A >*
v V/~. \ /^/
^\ \ \ i/
L V 1
/— IO?M. (XCLUDINO 01 Mint
sc — ^ ^ n^
A^,(0
y~ Of MIITEt
,-. INLEKUn ^"CXl,
K *~i~~^ '~~~$
i — DCWNCOMH OUTLET ILAII 1
^— Will OVIMLOW CXJTLET (LAI) ^


-
C JdO 3W 300 HO MO 140 180 400 4M t*0 460 *
Ttil TIME, hwn
cAiiMOAi a*v

^ - ^ -
# "*

"
/ I I*

^^" "^J
!

-^ ^^" *^^
v
" \ '_
i — TOTAV SUlFUi [SOjl ^J .f^
^— CALCIUM (C*O) y— iUUHE t5OJ / n
3- ^.j-i^^ f d J^ r-t*^~* — •»--.
OUOnO)BOnoMOHOMO*00*?04«9«40««
nsr TIM*. hM,
I J/73 1 J/I4 1 VW 1 Wt 1 V» 1 V30 1 J/ll 1 4/1 1 */5 1 4/1
CAUNMI OAV
* TOTAL Dt»OLVlD VTHIOS * MAGNCSHJM (*»»*")
O CAlCW* (C« ** ) n WO*** (Si. » )
O SULFAH &Qt ' 1 • iULFin tJOj" 1
A CHLOtIM (O - I O OMONAn {CO} * )
•
"
-
*]
A

O *'
0 0"
D-






. •-
^

0 *

i °, i , , , , i i , i . . i n


-1 o
» "
-B - |
-73 „



«S
-:«
1 1?
JlO z

j* "J 2
j! -°
i(j
1: ??

!:! Hi
J"^ ?
S H
.« 1 1 1
D

Z
„ i i
i i
D ! £
^
:E i
0 I I

' I ?
•
5 r- ^ ^
1 y i
, 1*1
0 ~"
I J
5 S n
XM1
" 'H
» "6
9



14,000
11,006

W'°" s
Ii
0
4,flOO 5
Z
is
I, MO c
5
0 7
s
*» B
» ?
i
400 ^
JOO

»0

T»
0
1/3)     1     1/J*     1

-------
                                                                     L\.£
   "   —-  *  ~  t
   V  70  O   &  Q   a
   2.  a  o. eg  s   "

   11?Ii  S
   •5   2  0.1  8  §
           r

           0?
           g
                 tw
Id
 I
 O
 o
 3
s  i  1   s
f  5:  £  3-

 O
 T3
 0>

 pu
 3
 10

 o.
 Q)
 C+
 O>
  O
  o
  -s
  c
  00
  CO
  D3

 O •*   — •
                      ~  «.OCD



                      S  1,000
                                             *  TOTAL DIUOIVCO KM.IOS  *  MAONHH** PT* BOj' )

                                             A  CMOUDt CO - )       O  CAttOUTE (CO, * 1
                                                                                                          4/11   I   VI*

-------
An overall summary of the initial run data appears in Table 13, which

presents average values for some  significant parameters (from

Figures 20,  21  and 22).  The listed pH values may be in error.




A summary of the liquid analytical data is presented in Table 14.

Most dissolved  species appear to have approached steady state con-
centrations during the period of operation.  However, magnesium

ion (Mg   ) concentration exhibited a steady

and TCA  systems (see  Figures 20 and 21),
ion (Mg  ) concentration exhibited a steady increase in the venturi
It is of interest to compare the liquid analytical data for the closed-

loop and open-loop runs (see  Tables 8 and  14).  For the TCA system,

sulfate concentrations for both periods of operation were close to the

"saturation" levels.  The chloride concentrations, however, differed

significantly.  For the venturi and Hydro-Filter systems, the  sulfate

concentrations during the initial closed-loop runs was well above that

measured in the open-loop tests. As expected,  in all three systems,

the level of total dissolved solids during the closed-loop runs was

far greater than that obtained during open-loop testing.




Lack of confidence in the pH meters led to  a decision to control SO,

removal in the initial tests rather than to directly control pH at the

desired 5. 7 to 5.9 region.  Results of open-loop limestone depletion

runs were used to estimate SO  removals consistent with the desired
    * *                        *
pH.    In general,  these SO?  removal levels were selected at about
*
   There have been some problems with pH meters at the facility.

##
   See Figure 11 and Table 6.
                               318

-------
                          Table  13
           AVERAGE CONDITIONS FOR INITIAL
            RELIABILITY VERIFICATION RUNS
Parameters
Operating Time, hrs
Gas Velocity, ft/sec
L/C, gal/mcf
Pressure Drop, in. H.O
Percent Solids Recirc.
Percent SO_ Removal
Stoichiometric Ratio
Limestone Utilization
Inlet Liquor pH
Percent Oxidation
Dissolved Solids, pprn
Hydro -Filter
Run 501. 3A
520
5
53
9
11
65-70
1.25
80%
5,8
30
8000
TCA
Run 501-2A
550
7.8
80
6
15
80-85
1.20
83%
5.8
20-30
7500
Venturi
Run 501. 1A
410
5
40*»
10.5**
15
70.75
1.5
67%
5.8-5.9
15
7000
L/G's of 40 for spray tower and 40 for venturi.
9 inches across venturi and 1. 5 inches across spray tower.
                            319

-------
                         Table 14

      AVERAGE  LIQUOR COMPOSITIONS FOR INITIAL
            RELIABILITY VERIFICATION RUNS
Scrubber
System
Liquor
so3= co/
Venturi 200 200
TCA 150 150
Hydro-
Filter 300 150
Species
so/
1500
1800
1800
Concentrations, mg/1 (ppm)
c."
2000
2000
2000
Mg+*
250
300
200
Na+
50
50
50
Cl"
3000
3000
3500
Total
7200
7400
8000
This species increased gradually throughout the time period.  The
values shown on this table are the maximum values, at the end of
the plotted periods.
                           320

-------
10% below that attainable in open-loop operations at a pH above 6. 0
(high pH).  Control of SO  removal was established by varying the
rate of limestone addition.

7. 1. 1    Venturi Run 501-1A (see  Figure 20)

Open-loop testing at high-pH indicated approximate SO,  removals to
                                               *
be 42% in the venturi and 57% in the spray tower,  which is equiva-
lent to an overall removal,  for the combined system, of 75%. Thus,
in order to achieve the desired low-pH operation , a target of 65%
removal is indicated.

From April 14 to April 21,  removal was controlled at about 74%
with an average stoichiometric ration of 1.5 (moles CaCO, per mole
SO- absorbed) and an average oxidation of 15%.   This implied higher
pH than desired,  although the scrubber inlet liquor pH (which is sub-
ject to question) was apparently 5. 8 — 6. 0 during this period.

From April 22 to April 27,  SO, removal was controlled  at about  70%.
In this period the stoichiometric ratio rose from 1. 3 to  1.8.  This
increase in stoichiometry, while maintaining the same SO, removal,
^
   This removal was estimated from Eq. 11 and "corrected" for the
   change in inlet SO  concentrations (see Eq. 12 and Section 5. 1. 2).
                    &
** The high-pH removal at the selected venturi run conditions was
   originally estimated at 85% for the closed-loop operation.  The
   early part of Run 501-1A is thus at a pH and SO, removal some-
   what higher than desired.
                               321

-------
may indicate "degradation" in the system (e.g. , drop in limestone
reactivity,  erosion of spray nozzles).  Oxidation and apparent inlet
liquor pH remained at 15% and 5.7 — 6.0 respectively.

Toward the end of the operating period depicted  on Figure 20, the
SO_ removal dropped below  70% and was restored to a value  slightly
above  70%.  On April 28 (low removal) the  stoichiometric ratio was
about  1. 5 and then rose to above 2. 0 by April 30.  The apparent
scrubber inlet liquor pH remained at 5. 9 — 6. 0 in this period.

7.1.2    TCA Run 501-2A (see Figure 21)

The estimated high-pH value of SO. removal for the TCA system,
                                                         ,tf
operating at the test conditions of Run 501-2A, is about 95%'  (see Eqs.
5 and  12).  Therefore, the controlled SO, removal for Run 501-2A
                                       £
was chosen at  85%.  After an initial operating period (from March 24
to March 31) in which there  were  relatively large fluctuations in SO,
removal, a relatively steady period of about 5 days ensued (from April
1 to April 6),  where the  SO, removal varied from 80 to 85% and the
                          £»
scrubber inlet liquor pH from about 5. 7 to  5. 9.  The stoichiometric
ratio during this period was  about 1.15  (which corresponds to a
limestone utilization of about 87%) and the percent oxidation was
about  30%.
*
   A removal of about 96% was obtained in the line-out (depletion)
   period for Run 501-2A. This confirms the estimate from the open-
   loop data.
                                322

-------
Towards the latter part of the plotted operating periods (from April
12 to April  15 and from April 18 to April 20), while the SO removal
                                                        L^
was still controlled between 80 and 85%, there appeared to be an in-
crease in the stoichiometry to an average value of 1. 4 (limestone
utilization of 71%), which again  could indicate "degradation" in the
system (e.g. , drop in limestone reactivity, pluggage of spray noz-
zles).  The  percent oxidation during these periods dropped slightly
to about an  average of 20% and the inlet pH ranged between 5. 7 and
6.0.

The system was shut down a number of times due to  solids pluggage
of the inlet  duct in the vicinity of the humidification section, and re-
sulting increase in total system pressure drop (see Ref.  1).

7.1.3    Hydro-Filter Run 501-3A fe 3B (see Figure  22)

The predicted high-pH value of SO_ removal for the Hydro-Filter
system,  operating under the test conditions of Hun 501-3A, is about
80% (see Eqs. 7 and 13).  Therefore, the controlled SO7 removal
                                                     C»
target for Run 501-3A was 70%.  During most of the  operating period
for Run 501-3A, the SO^ removal was controlled between 65 and 70%,
the average stoichiometric ratio was about 1. 3 and the average per-
cent oxidation and inlet liquor pH were about 30%  and 5. 8, respec-
tively.  As mentioned previously, the measured pH's, during this
operational  period, were in doubt.
#
   During a brief period of high stoichiometric ratio (about 1.5) and
   inlet liquor pH (about 6. 1) at about 110 hrs.  in Figure 22, the SO2
   removal rose to about 80%,   This  substantiates the predicted high
   pH removal.
                                323

-------
 After the system was drained to remove debris (marbles) on March
 28 and March 29, another depletion (or line-out) period was conducted
 for Run 501-3B.  From April 3 to April 13,  the SO- removal was held
 between 65 and 70%, and the average stoichiometric ratio was about
 1.4.   The stoichiometry,  during this period  of the run, appeared to
 gradually increase, form an initial average value of about 1. 3 (from
 April 3 to April 6) to a final average  of about 1. 5.  The percent oxi-
 dation remained relatively steady  during this period (at about 30%)
 as well as the  inlet liquor pH (at about 5.8).   The increase in stoichio-
 metry, for the  same SO_ removal, could again indicate some "degra-
 dation" within  the system.

 7.2     MATERIAL BALANCES

 As mentioned previously (see Section 5. 1), during open-loop testing,
 good  material balances for calcium and sulfur could only be obtained
 with the TCA.   The poor material  balances for the venturi and Hydro-
 Filter systems were attributable to solids build-ups (or depletions)
 in the clarifiers, which could not be excluded from the material bal-
                 *
 ance  enclosures.   During the five-week boiler outage, the venturi
 and Hydro-Filter flow configurations  were modified to ones similar
 to that of the TCA system (see Figures 4, 5  and  6). It was expected,
 therefore, that good material balances for calcium and sulfur would
 be obtained on  all three scrubber systems, based on the measured
*
   For the TCA,  the main slurry stream circulated between the
   hold tank and scrubber, with a "bleed stream" from the main
   slurry stream routed to the solids separation area.
                               324

-------
flow rate and solids compositions of the bleed streams to the solids
separation area and the measured limestone addition rates and SO.
removals.

7.2.1   Venturi Run No. 501-1A

Table 15 gives the results of a material balance for calcium and sul-
fur for venturi Run 501-1A, during a continuous 142 hour operating
period from April 14 to April 19, 1973 (see Figure 20).

The results of the balance showed that the measured sulfur discharged
(4.38 Ib-moles/hr) is 3. 1% less than the measured SCX absorbed
(4.52 Ib-moles/hr) and that the measured calcium added (6. 10 Ib-
moles/hr) is 6.0% less than the measured calcium discharged (6.49
Ib-moles/hr).  Both closures are satisfactory,  in spite of  some dif-
ficulties experienced in measuring the limestone  feed rate during the
                       *
initial reliability tests.

The ionic balances for the solids analyses, from  which the calcium
and  sulfur discharge rates were calculated,  averaged less than +3%
(more cations  than anions) for the bleed stream shown in Table  15.

Note that for both sulfur and calcium the measured inlet and outlet
rates do not necessarily balance during each individual computational
period in Table 15.  This is due to the unsteady conditions which
prevail (e.g.  changing percent solids) and the resultant accumulation
   This measurement problem will be alleviated after May 4, when
   replacement flowmeter elements (for smaller flow ranges) are in-
   stalled in all three limestone feed system magnetic flowmeters.
                               325

-------
                                                                                     FABLE  15
                                                                 MATERIAL BALANCES FOR VENTUKI RUN NO. 501-1A


Date




Time


Length
of
Period.
hours

Gi«
Flow
Rate.
acfm
@ 330 OF
Inlet
S02
Cone. ,
ppm


S02
Removal.
%

Bleed Stream
Flow
Rate.
gpm

4/H 0200-0800 6 19.750 3.000 73 12.5
4/14 0800-2400 16 19.850 2.950 72 14.8
4/15 0000-2400 24 19,750 1.050 74 It. 6
4JJ6 0000-0800 8 19.90(1 1.150 76. 16.6
4/16-17 OBOO-OSOO 24 20. 000 3.05.0 73 IS. E
4/17 0800-2400 16 19.900 2.850 75 15.5
4/18 0000-2400 24 ZO.OOO 2.900 73 15. S
4/19 OOOD-2400 24 20,000 2.950 72 15.8
Totals 142
Solids
In
Bleed.
•urt. %
13.7
14.0
14.6
14.9
15, S
It. 0
15.7
16.0

SOX in
Solids
»sSO3,
«.!-•%
29-3
24.9
24.0
27.4
24.4
32.4
29.3
25.3

Ca in
Solids
as CaO.
wt.%
_
*'
Flow
Rate,

gpm
26-0 1.2J
27.6 1.47
27.7 1.53
29.0 1.25
27.3 1.45
29. S 1.41
27.5 1.40
26.6 1.59

Solid s
Content,

wt. 1,
Sulfur Balance
SQt
Absorbed,

Ib-moles
56.7 Z7
59. 3 70
57.4 111
53, 0 40
59,! HI
59. S 71
58.1 106
57.5 106
642
SOX in
Solids
Ditch. ,
Ib-xnolea
Calcium
Ca in
L-S
Feed,
Ib- moles
21 31
56 101
96 150
37 41
9-9 149
&9 98
iia uo
106 156
<>22 &66
Balance
Ca in
Solid I
Ditch. ,
Ib-molet
26
89
158
54
155
U6
159
160
922
CJ
KJ
01
                 Average rates. Ib-moles/hr:
                     SO2 absorbed  = 642/142 ~- 4.52
                     SOX discharged = 622/142 = 4. 38
                     Ca added      = 866^142 = 6. 10
                     C» discharged  = 922/142 = 6.49
Average stoichiometric ratio, moles Ca added/mole SOj absorbed:
 Based on limes, one added and SOj absorbed - S66/642 = I. 35
 Based on solids analysis                  = 922/622 = 1.48

-------
(or depletion) of the species in. the system.  However, over a long
period of time (e. g. ^150 hours) the accumulation term becomes
negligible as compared to the total input or output for the entire
computational period.

The average stoichiometric ratio in Table 15  of 1.48 moles Ca/mole
SO9 absorbed based on solids analysis is probably more accurate
   £t
than the value of 1. 35  based on the measured  limestone addition rate
and SO, absorption, because of uncertainties  in the  limestone slurry
       C*
feed rate.

7.2.2    TCA Run No. 501-2A

Table  16 gives the  results of material balance calculations for TCA
Run 501-2A, covering a period of 150 hrs.  uninterrupted operation
from March 30 to April  6, 1973  (see Figure 21).

The results of the balance showed that the sulfur discharged {4, 34
Ib-moles/hr) ie 7% less than the SO. absorbed (4.67 Ib-moles/hr),
while the calcium added (4.45  Ib-moles/hr) is 11% less  than that
discharged  (4.99 Ib-moles Air).  The closures are considered  to be
quite acceptable.

jn Table 16, the sulfur input in each individual computational period
is generally greater than ftie output,  and the reverse is  true for
calcium.  The ionic imbalances for the solids analyses during these
periods were mostly positive (more cations than anions) and averaged
abovit +5%.  In other words,  the reported sulfur content in the bleed
solids might have been too low, or the calcium content too high, or
                                327

-------
                                                                       TABLE  16
                                                   MATERIAL BALANCES FOR TCA RUN NO. S01-ZA





3/30-31
3/31
4/1
4/Z
4^3
•4/4
4/5-6
Total*





2300-0800
0800-2400
0000-2400
0000-2400
0000-2400
0000-2400
OOOD-0500

Length
of
Period.
hours

9
Ih
24
14
24
24
2<>
150
Gas
Flo*
Rate.
acfm
@3oo°r
20,200
20,000
20,000
20,000
20, 000
20,000
20. 000

Inlet
S02
Cone. ,
ppm

2410
2460
2520
2540
2700
2680
2580


S02
Removal,
1,

84
87
S2
85
82
82
84

Bleed Stream
Flow
Rate.
ipm

14. 1
!+. 5
1 J. 7
13,9
14.9
14. 0
13.7

Solids
in
Bleed,
wt. %
14. 3
14. 1
15.3
15.0
15.5
14.2
14.6

SO, in
Solids
as SOj.
wt. %
27.8
24.3
28.9
27.8
33.2
34.6
31. 5

Ca in
Solids
as CaO,
wt. %
25.2
22. 6
Z4.4
24.0
it,. 6
24.2
2!. 5

Limestone Feed
Flo*'
Rate.

gpm
1. 03
1. 10
1. 17
1.25
1. 07
0. 96
0. «

Solids
Content,

wt. 7,
S8.4
i9.2
S3. 5
S6.3
fco.o
59.6
60. 3

Sulfur Balance "x
S02
Absorbed.

Ib-moles
43
74
107
112
115
114
13*
701
SOX in
Solids
Disch. ,
Ib-moles
35
54
100
96
127
113
126
451
Calcium
Ca in
L-S
Feed.
Ib-moles
3"!
75
103
119
112
100
119
fc(,7
Balance
Ca in
Solids
Disch. .
Ib-moles
45
72
121
118
141.
113
134
749
Average rates. Ib-moles/hr;

       SOz absorbed     -   701/150   =   4.67
       SOX discharged   -   651/150   =   4.34
       Ca added         -   f.C7/i^O      -i.45
       Ca discharged       749/150   =   4.99
Average stoiehiometric ratio,  molea Ca added/mole SO2 absorbed;
       Based on limestone added and $02 absorbed
       Based on solids analysis
667/701  =  0.95
749/651  -  1.15

-------
both.   If this factor is taken into account, either or both of the sul-
fur and calcium balances would  be better than those reported.

Again, due to uncertainties in limestone addition measurement, the
average stoichiometric ratio of 1. 15 moles Ca/mole SO  absorbed
based on the solids analysis is a more reliable number than the value
of 0. 95 based on the measured limestone addition rate and SO-
                                                           Cf
absorption.

7.2.3   Hydro-Filter Run No. 501-3A

Table  17 gives the results  of material balance calculations for Hydro-
Filter Run No. 501-3A,  covering a period of 150 operating hours from
March 16 to March 22, 1973  (see Figure 22).

For sulfur,  the average discharge rate (4. 11 Ib-moles/hr) is only 3%
less than  the SO,, absorption (input) rate (4.24 Ib-moles/hr). For
calcium,  the rate of addition {4.49 Ib-moles/hr) is 13% less than
the discharge rate (5. 16 Ib-moles/hr).   The balance is satisfactory,
considering the uncertainties in the limestone slurry addition rate
during the period.

The ionic imbalances for the solids analyses, from which the calcium
and sulfur discharge  rates were calculated, averaged less than +2%
(more cations than anions) for the bleed stream shown in Table 17.

Again, the average stoichiometric ratio in Table 17 of 1.26  moles Ca/
mole  SO, absorbed based on solids analysis is probably more accurate
than the value of 1. 06 based on  the measured limestone addition rate
and SO. absorption.

                               329

-------
                                                                                     TABLE  17

                                                             MATERIAL BALANCES FOR HYDRO'-FILTER RUN NO.  501-3A



Date


3/16
3/17
3/17
3/17
3/16
3/18
3/18
3/19
3/19
3/19
3/20
3/20
J/ZO
J/21
3/2'l
3/Z1
3/22
3/22
3/22
Total a



Time


1600-2400
0000-0800
0800-1600
1600-2400
0000-0900
1100-1600
1600-2400
0000-0800
0800-1600
1600-2400
0000-0800
0800-1600
1600-2400
0000-0800
0800-1600
1600-2400
0000-0800
0800-1600
1600.2400


Length
of
Period,
hours

8
8
8
8
9
5*
8
8
8
8
8
8
8
8
8
8
8
8
8
150

Gas
Flow
Rate,
acfm
@310°F
20,400
20.000
20,300
20.500
20,300
20,300
20,200
20.700
20,350
20.300
20,350
ZO, 600
20,200
20,250
20.150
20, 150
20,150
20,450
20,100


Inlet
S02
Cone. ,
ppm

2950
3050
3100
3150
3320
3050
2800
2900
2800
2750
2800
3020
2950
Z850
2850
2730
2750
2900
2870



S02
Removal,
1,

66
70
72
68
66
70
73
68
77
71
67
68
69
70
64
63
69
67
65

Bleed Stream

Flow
Rate,
gpm

24.7
22.5
26.3
25.4
22. 3
22.6
23.7
21.6
21. 6
25.0
22.1
19.9
19.0
19.2
17.3
16.5
18.2
19.6
19.6


Solids
in
Bleed,
wt. %
9.9
9.2
9.0
9.0
9.8
10.8
10.4
10.2
10.7
10.7
10.9
11.5
11.7
11.3
10.9
10.7
10.9
10.8
10.2


SOX in
Solids
as S03,
wt. %
25.4
21.6
23.0
24.4
27.9
30.0
30.4
31.6
26.0
28.7
28.7
30.3
33.5
26.4
27.5
27. 1
27.5
28. 3
27. 3


Ca in
Solids
as CaO,
wt. %
23. 7
22.8
23. 7
25.8
28.7
26.2
25. 6
24.9
25.6
24,3
23.3
24.5
24.9
22.7
23.3
22.5
22.8
22.5
23.7

Limestone Feed

Flow
Rate,

gpm
0.98
1. 12
1.42
1.60
1.75
1.68*
1. 47
1.93
1.87
1.35
1.52
1.45
1. 19
0.80
0.84
0.97
0.98
0.96
1.00


Solids
Content,

wt. %
35.0
57.6
59.8
50.6
49.9
51.3
53. 1
33.5
55.6
45.4
41.7
43.9
57.0
62.8
60.7
61.3
59.9
60.7
62. 1

Sulfur Balance

so2
Absorbed,

Ib- moles
33. 1
35.6
37.7
36.6
41.7
22.6
34. 4
34.0
36.5
31.0
31.8
35.2
34.2
33.6
30.6
28.9
32.2
33. 1
31.5
f,3f.

SOX in
Solids
Disch. ,
Ib-moles
33. !
23.7
28.8
29.5
36.5
24.5
40.0
37. 1
32.1
41. 1
37. 1
37.3
40.2
30.8
27.8
25.6
29.3
32. 1
29. 1
fiUi
Calcium

Ca in
L-S
Feed,
Ib-moles
16.0
36.8
49.4
43.2
52. 1
40. 5
42.5
30.0
58.2
31.4
31,3
32. 1
38.4
30.1
30. 0
35. Z
34. 1
34. 3
37.0
673
Balan.c

Ca in
.Solids
Disth. ,
Ib-molcs
44. 1
35.7
42.4
44.6
53.6
30.6
48. 1
41.8
45.2
49.7
43.0
43. 1
42.6
37.8
33,7
30.4
34.6
36.4
36.1
774
OJ
to
O
ivera^e rates, Ib-moles/hr
SO2 absorbed =
SOX discharged -
Ca added
Ca discharged =

636/150 =
616/150 =
673/150 =
774/150 =

4.24
4. 11
4.49
5. 16
                                                                       Average stoichiometric ratio, moles Ca added/mole SOg absorbed :

                                                                                Based on limestone added and SO^ absorbed  =     673/636

                                                                                Based on solids analysis                     =     774/M6
1.06

1.26
        ' The scrubber was shut down on 3/18,  0900-1100 hours (no flue gas addition).  However, the limestone addition was continued during this 2-hour period* and the length of limestone

         addition was 7 hours instead of 5 hours.

-------
                            Section 8


                          REFERENCES


 1.    H.  W. Elder, et al. ,  "Operability and Reliability of the EPA
       Lime/Limestone Scrubbing Test Facility, " presented at Flue
       Gas Desulfurization Symposium, New Orleans, Louisiana,
       May 14-17,  1973

 2.    F.  T. Princiotta and M.  Epstein,  "Operating Experience with a
       Prototype Lime-Limestone Scrubbing Test Facility, " presented
       at the Sixty-Fifth Annual Meeting of the A. I. Ch. E. ,  New York City,
       November 26-30, 1972

 3.    M. Epstein, et  al. , "Test Program for the EPA Alkali Scrubbing
       Test Facility at the Shawnee Power Plant, " presented  at Second
       International Lime/Limestone Wet Scrubbing Symposium, New
       Orleans, Louisiana, November 8-12, 1971

 4.    F.  T. Princiotta and N.  Kaplan, "Control of Sulfur Oxide Pollution
       from Power Plants, " presented at EASCON, Washington, D. C. ,
       October 18,  1972

 5.    R.  H. Borgwardt.  Limestone Scrubbing at EPA Pilot Plant,  Progress
       Report No.  3. EPA Report, October 1972

 6.    J.  M. Potts, et al. , "Removal of  Sulfur Dioxide from Stack Gases
       by  Scrubbing with Limestone Slurry:  Small Scale Studies at TVA, "
       presented at Second International Lime /Limestone Wet Scrubbing
       Symposium, New Orleans,  Louisiana, November 8-12, 1971

 7.    A.  Saleem,  et al. ,  "Sulphur Dioxide Removal by Limestone Slurry
       in a Spray Tower,  ibid.

 8.    A.  D. Little, Inc.,  Evaluation of Problems Related to Scaling in
       Limestone Wet Scrubbing,  EPA Report, April  1973

 9.    R.  H. Borgwardt,  Limestone Scrubbing at EPA Pilot Plant,  Progress
       Report No.  6,  EPA Report, January 1973

10.    Radian Corporation, A Theoretical Description of the  Limestone-
       Injection Wet Scrubbing Process,  NAPCA Report, June 9, 1970

jl.    A.  Saleem,  J. Air Pollution Control Assoc., Vol. 22, No. 3,
       March 1972

                                   331

-------
12.    M. Eptstein, et al., "Mathematical Models for Pressure Drop,
      Particulate Removal and SOg Removal in Venturi, TCA and
      Hydro-Filter Scrubbers, " presented at Second International
      Lime/Lime s tone Wet Scrubbing Symposium, New Orleans,
      Louisiana, November 8-12, 1971
                                 332

-------
   OPERABILITY AND  RELIABILITY
OF EPA LIME/LIMESTONE SCRUBBING
          TEST FACILITY
                by

            H. W.  Elder
     Steering Committee Member
     Tennessee Valley Authority
             L. Sybert
       Test Program Director
         Bechtel Corporation
           J. E. Williams
   Onsite Technical Representative
   Environmental Protection Agency
            P. E.  Stone
       Test Facility Supervisor
     Tennessee Valley Authority
                  333

-------
OPERABILITY AND RELIABILITY OF THE EPA LIME/LIMESTONE SCRUBBING TEST FACILITY
                                  ABSTRACT


            Results of research and development programs to study lime/
  limestone wet scrubbing have shown that effective sulfur dioxide removal
  is feasible.  One of the major questions remaining is the long-term reli-
  ability of the process when applied under actual conditions on a power
  generating unit.  The objective of the EPA-Bechtel-TVA test program at
  Shawnee is to define a system which operates both effectively and reliably.
  The test facility design considerations to permit investigations of opera-
  bility are covered in this presentation along with the experience during
  about 1 year of operation.  Preliminary conclusions which might affect
  selection of components for full-scale systems are presented.
                                Introduction
            As  part  of an  extensive research  and  development program by the
  Environmental Protection Agency  (EPA),  an experimental test facility was
  designed and  constructed to  study and evaluate  the feasibility and economics
  of closed-loop lime/limestone wet-scrubbing processes and develop mathe-
  matical  models to  allow  effective and economic  scale-up of practical
  operating configurations  to  full-size scrubber  facilities.  Interest in
  alkali scrubbing has grown because of its basic  simplicity and comparatively
  low capital cost.  The drawbacks of alkali  scrubbing are the need to dis-
  pose of  large quantities  of  solid wastes and the tendency of accumulation
  of solids to  plug  the system.

            The emphasis on development of technology for sulfur dioxide
  control  during a relatively  short period of time resulted in pilot-plant
  studies of limestone  and  lime scrubbing by  several organizations.  These
  programs  led  to the  conclusion that lime/limestone scrubbing is based on
  feasible  technology.  However, the process  alternatives are not well de-
  fined and the long-term reliability of  the  configurations tested to date is
  still questionable.   One  of  the main objectives  of the prototype-scale test
  program  is to provide a better understanding of  the technology and to study
  the factors which  affect  reliability.   The  program is funded and directed
  by EPA; Bechtel Corporation  prepared the detailed design of the test fa-
  cility, and developed, the test program; and TVA  constructed and operates
  the test  facility.
                                    334

-------
          Three parallel scrubbing systems (10-raw each) were installed
at TVA's Shavnee station near Paducah, Kentucky.   The system was described
at the limestone symposium in New Orleans during October 1971.   The facility
has been operated now for approximately 1 year.   The process results are
covered in a paper presented by Dr. M. Epstein (Bechtel Corporation) at
this meeting.  The intent of this paper is to review the operability and
reliability of the test equipment during the early phases of the program.
The design considerations in providing methods for evaluation of reliability
are reviewed followed by a description of operating experience.
                          Design Considerations
          The test facility is not sufficiently large to have a significant
effect on stack gas effluent quality at the station where it is installed—
its only function is to provide information on lime/limestone scrubbing
technology.  Therefore, it was designed to be a versatile, relatively sophis-
ticated system with emphasis on data generation and collection.
Unit Size

          Most of the test facilities used to study sulfur oxide and particu-
late matter removal are small ones capable of treating only up to a few
thousand cftn of flue gas.  Because of the differences in both the performance
characteristics of the equipment and scrubbing efficiencies of small-scale
test facilities and commercial-size units, and the associated uncertainty
and difficulty involved in using small-scale data in full-scale unit design,
the three scrubbers at the Shawnee test facility are of prototype size,
capable of treating up to 30,000 acfm of flue gas (320°F, 1^.3 psia).

          The size selected will allow extrapolation by a factor of 10 (an
acceptable scale-up) to commercial-size units.  The systea is designed so
that the three scrubbers can be operated simultaneously in order to acquire
information on different types in the shortest period.  In retrospect,
because of the multitude of problems associated even with routine operation,
maintenance, and evaluation of process data, it might have been more prudent
to concentrate on a single, versatile, system.
Flexibility

          The selection of the scrubbers was based on the evaluation of
the performance characteristics of the various designs in terms of their
overall effectiveness as S02 absorbers and particulate collectors under the
operating conditions expected for the process in commercial applications.
The selected types of scrubbers are:

                                  335

-------
     1.  Variable throat venturi scrubber followed by an after-scrubber
         absorption section (spray tower or Pall-ring packed bed).

     2.  Turbulent contact absorbers (TCA).

     3.  Hydro-Filter marble-bed absorber.

It is beyond the scope of this paper to discuss the types of scrubbers
considered and their advantages and disadvantages.

          The following describes some of the flexibilities that have been
designed into the test facility:

     •    The three scrubbers,  including the flue gas inlet and exhaust
         facilities, slurry handling equipment, and clarifiers are in-
         stalled in parallel.   Only the alkali addition system and the
         final solids dewatering and disposal facilities are common.

     •    Scrubber internals can be changed.  For example, the packed-bed
         section of the venturi after-scrubber can be replaced by a
         four-header spray tower (the four headers  can be operated in
         any combination).   The TCA can be operated as a one-, two-,  or
         three-bed unit, with  a variety of liquor flow piping arrangement.

     •    Arrangements are being made to provide parallel limestone and
         hydrate addition systems.

     •    The scrubbers can accommodate different types of mist eliminators.

     •    Each scrubber system  has its own oil-fired reheater for inde-
         pendent evaluation of exhaust gas characteristics with up to
         125°F (above scrubber exhaust temperature) reheat capability.

     •    A heat exchanger was  provided to evaluate  the effect of cooling
         of the slurry feed on scrubber performance (only one scrubber
         at a time).

     •    Various solid disposal configurations can be evaluated.   They
         include:  clarifier/pond,  clarifier/centrifuge/pond, and
         clarifier/rotary vacuum filter/pond.

     •    To evaluate the degree of supersaturation in the circulating
         slurry system and the associated possible  scaling problems,  the
         residence time of the circulating slurry solutions in the scrub-
         ber effluent hold tanks can be varied between k and 60 minutes
         (not necessarily in all three scrubber systems).
                                  336

-------
         To determine uncertain model coefficients for gas side con-
         trolling mass transfer and obtain design information for
         sodium carbonate aqueous scrubbing, a common S02 additive
         system was provided.  This system can also be used to compen-
         sate for rapid and significant drop in S02 concentration in
         the flue gas feed.

         The ducts, piping, and pumps in each scrubber system are
         designed to allow the required variation of the turndown ratio
         and the ratio of liquid to gas flow rates.

         Variable speed pumps have been used for flow control wherever
         possible in order to avoid potential solids buildup and/or
         rapid erosion of throttling mechanisms (rubber pinch valves
         or metal control valves).
Instrumentation

          The control of the test facility operation is carried out from a
central control room.  Important process control variables are continuously
recorded or indicated on graphic panelboards by remote conventional
(pressure, temperature level, and water, liquor, and air flows) and special-
ized instrumentation (radiation-type densitometers, magnetic-type flow meters).
The SOg concentration of both the flue gas feed and scrubbed gas streams to
and from the scrubbers can be automatically analyzed by UV spectrophotometer;
also C02, 1^0, 02, and H2 can be continuously monitored by gas chromatographs.
The analytical results of these instruments are also remotely recorded in the
central control room.


          The number of local and board-mounted instruments is about 1500.
Some of the more  important control loops are listed below:

     •   The flue gas flow to each scrubber is  controlled by a flow-
         indicating control loop which senses the flow via a venturi flow
         element and sets the position of the damper at the inlet of the
         induced-draft fan.

     •   The S02 injection system to the flue gas feed is on a control
         loop set by the Du Pont UV analyzer using a valve in the S02
         line as the final control element.

     •   The reheater outlet temperature is controlled by a temperature-
         indicating control loop which senses the exit temperature and
         adjusts the fuel oil and combustion air rates to the burners.

     •   The reheater and the induced fan are interlocked such that the
         fan has to be operational in order to fire the reheater.

                                   337

-------
     •   The internals of the venturi after-scrubber and the TCA and
         Hydro-Filter scrubbers are lined with neoprene rubber.   To
         protect the rubber lining, a temperature sensor is provided
         at the bottom of each scrubber which trips the ID fan and
         reheater systems and admits water to the emergency sprays in
         the scrubber gas inlet duct.  (The temperature trip is set at
         190°F.)

     •   The induced draft fans are protected by automatic shutdown
         control loops in case of high bearing vibration and temperature.

     •   Each scrubber is protected by a vacuum relief system which pro-
         tects the induced draft fan against high vacuum (-3^ in. water gage).

     •   The amount of,limestone and water to be added to the limestone
         slurry tank is controlled by a cascade of a level indicator con-
         troller, to a weight of limestone, to volume of water ratio
         controller.  This system maintains the desired limestone slurry
         concentration at any set level in the mixing tank.  The limestone
         slurry is then pumped by remotely controlled pumps to feed the
         scrubber systems separately.

     •   The scrubbing slurry to each system is tied to a flow-indicating
         control loop.  The loop uses a magnetic flow meter as a sensing
         device and a variable speed pump as the final element to maintain
         the desired flow.

     •   The levels in the effluent hold tanks are controlled by the amount
         of slurry bled to disposal.  The levels are sensed by diaphragm-
         type elements.

     •   The clarifier underflow is set by a density-indicating control
         loop.  The underflow density is sensed by a gamma radiation type
         element and the flow rate is set by a signal to the variable speed
         pump.

     •   Alarms for high and low values of different variables are also
         provided where necessary.

         An electronic digital data acquisition system is utilized to record
automatically and continuously over 150 selected operating data points.  In
addition, about 125 channels are provided for alarm condition printouts.  This
system is hard wired for data output in engineering units directly on magnetic
tape which is transmitted from the field to Bechtel Corporation in San Francisco
for data evaluation along with manually recorded pertinent information.

         An X-ray Fluorescence Spectrometer is used to determine the following
species in liquid samples:

                            Ca**, S, 1C*", Cl"

                                  338

-------
Solid samples can also be analyzed after dissolution in an appropriate
solvent (e.g. HCl).  All functions of the X-ray unit (such as sample
presentation, goniometer angle, slit, collimator) are controlled by a
Nova 1200 minicomputer.  Sample identification is fed to the system via
an ASR 35 teletype.  All the data are reduced by the computer and printed
out on a line printer in addition to being recorded on a magnetic tape.


Provision for Inspection and Cleaning

          An equipment inspection program for the test facility is scheduled
to observe the performance and condition of the equipment after periodic
intervals of operation.  The results of the inspection are correlated to
process conditions (temperature, pH, chemical composition, particle size,
distribution of limestone slurry feed) and the time period for which the
equipment was exposed to the above conditions.  Trends based on quantitative
data and observations are being analyzed and evaluated, and corrective actions
will be taken to resolve the problems.

          Two general categories are covered by the program:

     •   Evaluation of the corrosive tendencies of the various process
         fluids using corrosion coupons installed in preselected lo-
         cations in the process equipment.

     •   Observation and evaluation of the equipment performance with
         regard to erosion, solids deposition, scaling, etc.

          Corrosion test racks are installed in the flue gas inle*b ducts,
scrubber exhaust ducts downstream of the reheaters, at different locations
in each scrubber, in the effluent hold tanks, recirculating tanks, and clari-
fiers; twenty coupons are mounted on each test rack.1  The selection of the
coupons and the location of the racks were based on experience and anticipated
locations of corrosive conditions.

          The equipment performance checks are made periodically at various
points in the system for:

                    Erosion
                    Deposition of solids and slurry sludge
                    Corrosion (this is separate from the evaluation
                     of the corrosion coupons)
                    Formation of chemical scale
                    General conditions
1 Carbon steel (A-283), Cor-Ten B (A-588), l&l S.S., 3l6L S.S., 18-18-2 S.S.,
  Alloy 20 Cb-3, lHO S.S. , W»6 S.S. , E-Brite 26-1 S.S. , Hastelloy B, Hastelloy
  C-276, Incoloy 800, Incoloy 825, Inconel 625, Monel 400, 70-30 Cu-Ni,
  Aluminum 3003, Flakeline 200 (Coated M.S. ), fiberglass-reinforced plastic
  (Bondstrand), Transite.  In addition, specimens of fiberglass-reinforced
  Furan, resin, natural rubber, neoprene and butyl rubber, and five stressed
  alloy specimens were installed at selected locations.

                                 339

-------
The location of the inspection points and the frequency of inspection are
based on the severity of the process conditions (temperature, pH, etc.),
hydraulic characteristics of the system, equipment vendor recommendations,
and experience by others.  Wherever possible, photographic records of the
condition of the equipment are established and used in the evaluation of
the system.

          The observation windows provided on the venturi, TCA, and Hydro-
Filter scrubbers have only limited use in limestone service because the slurry
interferes with visibility.

          Access doors on major equipment and Victualic couplings on the
slurry piping system permit quick cleaning and disassembly.


Materials of Construction

          The selection of material for construction was based on anticipated
process conditions and design and economic considerations.  For example:
Rubber lining of the scrubbers was selected for erosion and corrosion protec-
tion; the stainless steel construction of the venturi scrubber was necessitated
by high throat velocities.

          The following list gives an overall picture of the materials of
construction:

     •   The inlet flue gas ducts are of carbon steel and are insulated.

     •   The scrubber exhaust gas ducts are of 31&L stainless steel.

     •   The induced-draft fan internals are of 3l6l> stainless steel.

     •   The venturi scrubber internals are of 316L stainless steel.
         The after scrubber shell is of carbon steel coated with I/1*-inch
         neoprene lining.

     •   The TCA and the Hydro-Filter scrubber internals are of carbon
         steel lined with neoprene.

     •   The chevron demisters are of stainless steel.

     •   The direct-fired reheater shells are of carbon steel lined
         with refractory.

     •   The tanks are of carbon steel with internals lined with Flake-
         line (polyester glass) or with neoprene rubber.   The tank
         agitators are rubber coated.
                                  340

-------
     •   The limestone slurry tank is of 3l6*L stainless  steel.

     •   The carbon steel clarifiers are lined with Slakeline and
         the rakes are stainless steel.

     •   All of the slurry piping above 2-1/2-inch diameter is  of
         carbon steel with internal neoprene rubber lining.  The lines
         are not heat traced or insulated.

     •   All of the slurry piping smaller than 2-1/2-inch diameter
         is of stainless steel.  The lines are not heat  traced  or
         insulated.

     •   The water lines are of carbon steel and are heat traced and
         insulated.

     •   The slurry pumps have rubber-lined casings and  impellers.

     •   The slurry piping is valved with rubber-lined plug valves or
         rubber pinch valves.

     •   The entire SOg injection system is of 3l6L stainless  steel.

     •   The instrument probes (temperature for example) in the
         scrubber and duct internals are of Jl6L stainless steel.

     •   The solid bowl centrifuge and the rotary drum vacuum filter
         are constructed of J16L stainless steel.

     •   Slurry piping to the pH and conductivity meters is made of FVC.
                         Operating
          Firm conclusions regarding the effect of operating variables on
reliability can be made only after relatively long-term tests.  Operation
to date has been mainly restricted to short duration testing of performance
variables.  However, observations of trends and short-term effects are perti-
nent.  Also, the experience with instrumentation, solids handling, and
materials evaluation should be of interest.


Effect of Operating Variables on Reliability

          Although the effects of all variables may not be evident from the
experience to date, the factors are listed below together with applicable
observations from the test program and from supporting studies.


                                 341

-------
          Open Versus Closed-Loop Operation.  One of the objectives of
 development work on lime and limestone scrubbing is to establish conditions
 for operation without liquid discharge.  The criteria for quality of liquid
 effluent streams are not well developed; therefore, zero discharge may not
 be required but certainly is a safe level.  One of the main factors which
 affects reliability of lime/limestone scrubbing technology is the low solu-
 bility of the calcium absorbents and reaction products.  If the scrubbing
 slurry were discarded after a single pass through the scrubber, a minimum
 amount of precipitation would occur.  When the slurry is recycled, dissolved
 solids build up in the liquid phase and more precipitation occurs in the
 scrubber.  The precipitation of calcium sulfite and calcium sulfate leads
 to scaling in the scrubber if provisions are not made for controlling the
 precipitation.

          The test facility design included pumps with water seals for bear-
 ing protection, water quench sprays for gas cooling, water sprays for mist
 eliminator wash, and dilute slurry feed.  The water required for operation
 exceeded the makeup requirement so that the systems operated for about 6
 months with partially open liquor loops during limestone scrubbing tests.
 This was not considered to be a serious problem because the results of
 factorial testing of performance variables are not likely to be significantly
 influenced by the water balance.  However, not a great deal was learned about
 the effect of scaling potential on reliability during this period.  Essen-.
 tially no scaling occurred.

          During February and early March of this year, the systems were
 modified to permit closed liquor loop operation.  The absorbent feed system
 was changed to allow feeding of slurry with up to 60$ solids.  The water
 seals on the pumps were converted to mechanical seals supplemented with air
 purge.  The quench spray system was modified to use slurry and the mist elim-
 inator wash system was converted to use clarified water.  Required revisions
 to flow measurement and control devices were made.

          Since the modifications were completed in mid-March the systems
 have been operated with liquid discharge approaching the quantity of moisture
 that would reside with settled sludge.  No serious scaling has been observed
 during the relatively brief period of operation.  One serious encounter with
 scale during a special test is discussed in the section below related to
 stoichiometry.

          Gas Velocity.   The effect of gas velocity on entrainment is a
major reliability factor.  The need for high solids concentration in the
recirculated slurry aggravates mist recovery because the extent of entrain-
raent influences the amount of solids that impacts on eliminator surfaces
 and must be removed by washing.
                                  342

-------
          Most of the tests in the Hydro-Filter scrubber have been carried
out at about 5 ft/sec; a few were made at about T-5 ft/sec.   Because of the
few comparative data points the correlation between velocity and pressure
drop increase in the mist eliminator is not clear.  However, the rate of
solids accumulation was higher at the higher velocity.  During the tests at
the lower velocity, manual cleaning of the mist eliminator was required
about,once per month.  Only limited wash with fresh water was used during
this period.  An improved wash configuration has been installed to reduce
the cleaning requirement.

          The TCA scrubber has been operated during most of the test periods
at three different velocity levels, 5.9, 7.8, and 9.8 ft/sec; a few tests
were run at 11 ft/sec.  This system has been operated less than the Hydro-
Filter scrubber and no conclusions can be reached regarding effect of gas
rate on carryover.

          During a special test to simulate the TVA pilot-plant configura-
tion, slurry entrainment was excessive at 12 ft/sec and the gas rate had to
be reduced to 8 ft/sec for the run.  The higher value had been acceptable in
the pilot-plant (l-mw size) work; this indicates that gas-liquor distribution
may not have been as good in the larger scrubber.

          Most of the testing in the venturi-spray tower system has been with
slurry fed only to the venturi; in this mode of operation, the spray tower
serves as a large disengaging chamber and carryover to the mist eliminator
was not a problem.  When slurry was introduced in the spray tower, the gas
velocities were generally low, 2.5 to 5.0 ft/sec, although a few tests were
run at the maximum velocity of 7.5 ft/sec.  Mist eliminator deposits were
minor in all tests.

          Until the recent modifications were made, the mist eliminators
were washed intermittently with limited amounts of fresh water on the top side
only.  Piping was changed so that wash liquor can be applied to both sides
which should improve the flushing action: a mixture of makeup water and clari-
fied recycle liquor (about half and half) is now being used.  There has been
no indication of significant scaling in the demister since the changes were
made.  However, use of recycle liquor for wash has caused sulfate scaling in
pilot-plant studies and this could be a potential problem during extended
tests.

          Liquid Rate.  The primary effect of liquor rate on operability is
to control scaling by providing sufficient volume to accommodate the "make"
of reaction products.  Of equal importance is the flushing action in the
scrubber to avoid silting of suspended solids.  The latter requirement is
met incidentally because the trend in development of lime/limestone scrubbing
technology has been toward higher and higher liquor rates.  At the test fa-
cility, operation has generally been at maximum liquor rate—800 gal/min
(L/G of 27-5*0 in the Hydro-Filter, 1200 gal/min  (L/G of Uo-8o) for the TCA,
and ^50 gal/min (L/G of 20-^0) for the spray tower.  The spray tower system '
is being modified to increase the liquor rate to 1200 gal/min.
                                   343

-------
          Except for minor deposits in the Hydro-Filter bed, silting has not
 occurred.   Scaling has not been a problem because of the open-loop operation
 during most of the program to date.  The minor amounts of scale formed during
 the brief period of closed-loop operation is encouraging.

          Solids Concentration.  The suspended solids concentration in the
 recirculated slurry is important to the process because sufficient unreacted
 absorbent to provide the required dissolved calcium is needed and recycled
 calcium salts provide seed sites for precipitation; the effect of fly ash has
 not been established.  On the other hand, erosion increases as the suspended
 solid concentration goes up; therefore, an optimum level should be established
 and, although this has not been done, it remains  one of the objectives of
 the program.  The composition of recirculated solids is probably also important,
 but the degree of control is questionable.  If the sulfate content could be
 increased,  the total solids requirement might be reduced.  Tests of methods
 to promote  oxidation are planned.

          Percent solids recirculated has been varied from 5 to 12; a few
 tests were  made with 15$ solids.  The most serious effect of suspended solids
 has been erosion of spray nozzles.  The Hydro-Filter nozzles are constructed
 from stainless steel and have polyurethane internal liners and swirl vanes.
 The liners  failed after about 1800 hours of use and have been replaced with
 an improved design.  Open-type spiral nozzles constructed from stainless steel
 were chosen for the spray tower and these are also showing signs of wear; the
 same type except with stellite tips have been ordered for replacements.  The
 large (300  gpm) open-type nozzles used in the TCA scrubber have given good
 performance.

          Because of frequent changes and short periods of operation at each
 solids level, the effect of solids concentration on erosion rates was not
 determined.  This information will be obtained during the longer term tests.

          Stoichiometry.  Much of the pilot-plant work on limestone scrubbing
 has been done with a limestone feed rate equivalent to 1.2-1.5 times the
 amount required to remove all the sulfur dioxide from the inlet gas.  During
 a special test carried out in the TCA system to simulate the TVA pilot plant,
 the intended feed stoichiometry was 1.5.  Problems with flow control resulted
 in an actual stoichiometry of about 2.7.  The scrubber walls and grids were
heavily scaled during the 3-week, closed-loop test.  It has since been deter-
mined in pilot-plant studies that the excess limestone was the primary cause
 of scaling.

          EPA is conducting a test program in a small pilot plant (500 cfm)
to support the Shawnee program.  This program, under the direction of R. H.
Borgwardt, has been highly productive.   In one series of tests the pilot
plant was configured to simulate the TVA grid-packed tower and several poten-
tial causes of scaling were studied including limestone type, limestone
particle size, inlet gas temperature, presence of fly ash, and stoichiometry.


                                  344

-------
The factors were systematically tested and a-11  except stoichiometry had no
effect on scaling in the ranges studied.   With the same stone ground  to the
same particle size, scale formed rapidly (less than 40 hours) at a stoichi-
ometry of 2.5; the scale was similar to that found at Shawnee (primarily
CaSOa).  The scale growth rate was estimated to be 60 mg/m2.   At a stoichi-
ometry of 1.25* no scale formed during Uoo hours of operation even when the
hot gas (285°F) entered the scrubber without a precooling step.   The  effect
of high stoichiometry was also confirmed in the TVA pilot plant.

          From the results of the EPA pilot-plant studies, it appears that
calcium sulfite scaling occurs when the precipitation of calcium sulfite in
the tower exceeds a critical rate that is influenced by pH (rate of dissolu-
tion of limestone).  It was estimated that the critical pH for the system
tested is 6.2 - 0.1.  When this value is exceeded, scaling is likely  to
occur.  The effect of pH  (stoichiometry) on the dissolved calcium sulfite
in the scrubber discharge is shown below.
          Average Liquid-Phase Composition of Scrubber Slurrya

                     at 1.25 and 2.|? Stoichiometry

                       (L/G = 37.4, no fly ash)
                                       Stoichiometry
                   Scrubber effluent   1.25x    2.5*
                       so3, ng/i       3,?4o   1,580
                       S0a               774     424
                       C02               240     292
                       Ca                771*     670
                       pH                5.8     6.2
          a From "Limestone Scrubbing of S02 at EPA Pilot
            Plant," Progress Report 8, March 1973, p. H*


The difference in  sulfite level represents the increased precipitation  (and
scaling) at the higher stoichiometry.

          The test program planned for Shawnee specified a stoichiometry
range of 1.25 to 1.75»  Actual values were normally  higher than 1.5 as  a
result of problems with the feed  system.  Based on the results of the pilot-
plant tests, the stoichiometry will be lowered in order  to reduce the risk
of scale formation.  Minimum levels will be established  by desired sulfur
dioxide removal efficiency.  A potential problem  of  operation at relatively
low pH is loss of  limestone reactivity, probably  because of calcium sulfite
precipitation on the particles.   This phenomenon  has been observed in the
TVA pilot-plant program.
                                  345

-------
          Limestone Particle Size,  The effect of stoichiometry on pH points
out the importance of any factors which will influence dissolution rate of
the absorbent.  Limestone particle size is one of these factors and although
no systematic study has "been carried out in the test program, differences
in sulfur dioxide removal have been attributed to particle size of the stone.
An empirical test to relate particle size, feed stoichiometry, and unreacted
stone in recycled slurry to dissolution rate is needed for application of
results to different situations.

          Reheat.  The scrubbed gas is saturated when it leaves the scrubber
and entrainment plus any condensation which might result from cooling would
be detrimental to downstream equipment.  Therefore, some level of reheat
is required.  Reheat for each system is provided by combustion of oil in
direct-fired burners installed in a combustion chamber in the exhaust duct.
Operation of these units has been unacceptable.  The cooling effect of the
cool, oxygen-deficient, saturated scrubber exhaust plus poor atomization of
oil as a result of operation of a single nozzle over a wide range of flow
rates have caused incomplete combustion.  In addition to interference of
soot with particulate measurements, the poor combustion resulted in deposits
of unburned oil and soot in the ductwork above the reheater.  These accumula-
tions of combustibles ignited on two occasions and could have caused injury
to personnel or damage to equipment.  Fortunately the fires were extinguished
without serious effects.  After the second incident, the local plant manage-
ment advised that the test facility should not operate until the hazard was
eliminated.

          Both parts of the problem were investigated simultaneously during
a period while the test facility was not operating because of a scheduled
boiler outage.  Stainless steel sleeves (4o-in diameter by U-ft high) were
installed to isolate the combustion mixture from the scrubber exhaust until
combustion of the oil could be essentially completedj the effective combus-
tion volume is about 50 ft3.  Mechanical atomizing nozzles were installed
instead of the turbulent mixing type provided initially with the combustion
system.  The types now in use are designed for a narrow range of oil flow
rates and will have to be changed when the reheat requirement changes.  How-
ever, nozzle replacement is a simple job.

          The changes have been effective.  Essentially no soot is visible
in the stack gas and particulate samples have shown no evidence of carbon
from the reheaters.   Plans for installation of an external combustion system
have been deferred.
                                    346

-------
Instrumentation

          S0g Analyzers.  Experience with the six Du Pont Model ^00 UV S02
analyzers at the test facility has generally "been good.   Most problems have
been associated with the sample handling system rather than with the instru-
ments.  Initially the sample handling system was particularly vulnerable to
condensation, dust, oil, corrosion, or combinations of these factors.   This
led primarily to line leakage, line plugging, plugging of the filters, or
coating of the lens.  All of these effects resulted in erroneous S02 read-
ings.  It was also found during early attempts to calibrate using standard
reference gas, that the certified values are given for a specified tempera-
ture.  Use of the reference gas at other than the specified temperatures can
also lead to erroneous results.


          In order to eliminate the problem areas, the sample handling system
was modified in November 19T2 as follows:

     1.  All heat sinks and sharp bends in the sample lines were eliminated.
         A new 5/8-inch diameter Dekeron sample line was installed to replace
         the original 1/Wnch stainless steel line.   Heat tracing was in-
         stalled the full length of the sample line.

     2.  Stainless steel shields furnished by Du Pont were installed around
         the probe filters.  The original ceramic probe filters were also
         replaced by type Jl6 stainless steel probe filters recently developed
         by Du Pont.

     3.  An automatic zero and air blow-back system was installed on the S02
         analyzers in the inlet gas duct as had originally "been installed on
         those in the outlet gas ducts.

     k.  All stainless steel lines and fittings were replaced with Dekeron
         or teflon wherever  possible.

     5.  Calibration methods were changed to use a stainless steel wire mesh
         reference filter rather than use of standard reference gas bottles.

          One additional problem associated with the instrument rather than
the sample handling system was also encountered.  The interference filter in
the optic section of all six analyzers was found to deteriorate with time.
All of these filters, which screen out all except the desired light wave
lengths, were replaced.  The failure and subsequent deterioration was attri-
buted by Du Pont to have been caused by exposure to freezing conditions prior
to installation.  The freezing resulted in minute cracks which then deterio-
rated with time as theorized by Du Pont.

          Following the modifications made to the sample handling system
and the replacement of the interference filters,'operation for the past 5 to 6
months has been essentially trouble-free with only a minimum of preventive
maintenance time.
                                    347

-------
          Magnetic Flow Meters.  Operating experience with the Poxboro mag-
 netic  flow meters at the test facility has generally "been good.  The main
 problem has been in obtaining accurate flow measurements at very low flow
 rates  with meters designed to measure flow over too wide a range.  This is
 a problem inherent in the nature of a test facility where testing over a
 wide range of variables is desired.  To assure accuracy, Foxboro recommends
 a minimum linear velocity of 5 ft/sec through the flow element.  Periodic
 cleaning of the electrodes and calibration checks are also required to
 compensate for drifting of the meters.  Routine cleaning and maintenance
 checks are now made on a monthly basis.

          Control Valves.  Operating experience with control valves at the
 test facility is similar to experience with flow meters.  Operation has been
 generally good when control valves are used in reasonable design flow ranges.
 However, when excessive throttling is required to obtain very low flow rates,
 severe erosion in a relatively short period of time results from the increased
 velocity through the throttled valve.  This has been observed with both
 stainless steel plug valves and rubber pinch valves used as control valves.

          pH Meters.  Operating experience with pH meters at the test fa-
 cility has thus far been limited to in-line flow-type meters.  No significant
 scaling of electrodes has been noted to date with limestone.  However, fre-
 quent  calibration checks using a buffer solution are required to maintain the
 desired accuracy.  The current frequency in use at the test facility is twice
 per week on a routine basis or more frequent if required.  Because of the
 desirability to be able to control pH to within t 0.1 pH unit, future test
 program plans include evaluation of another type of pH meter*

          Density Meters.  Operating experience with density meters at the
 test facility includes both the Ohmart radiation-type meter and the bubble-
 type meter.  Both systems require further study and modification to achieve
 adequate reliability in their respective control service.

          Data Acquisition System.  Operating experience with the EMC data
 acquisition system at the test facility has been improved during the past 6
months.  Early in the test program much difficulty was experienced in recover-
 ing the data from the tapes recorded automatically onsite.  Extraneous
 characters were being recorded on the tapes and were interfering with the
 data recovery.  Changes were made to reduce the industrial noise of the system
 and a  special computer program was also written to help recover the data from
the tapes.   The system is also sensitive to dust, and particularly to coal
dust.   Since the system is neither enclosed nor located within a pressurized
area,  periodic cleaning of the recorder on a weekly basis was initiated.
Subsequent to these changes, operation of the data acquisition system has
been very good.
                                  348

-------
          Based on the operating experience at the test facility, two areas
 appear to be critical*  Some of the field effect transistor (FET) circuits
 appear to be extremely sensitive and should probably be redesigned if time
 permits.  Compatibility of the tape recorder with the equipment used to
 recover the data from the tapes also appears to be very sensitive to dust,
 noise, alignment, etc.  However, neither problem area appears to be insur-
 mountable with proper care or design of the equipment.

          X-Ray Computer System.  Operating experience with the Siemens
 X-ray computer system at the test facility has been similar to that on the
 data acquisition system.  Some mistakes made earlier in the program on the
 data acquisition system were avoided on the X-ray computer system.  Both the
 X-ray unit and the computer were enclosed in a pressurized air conditioned
 room.  This has minimized the problems of recovering the data from the tapes.
 Also when problems do occur in removing the data from the tapes, the special
 computer program prepared for use with the data acquisition system is
 available.

          Operation of the X-ray unit has been very satisfactory.  Most prob-
 lems have been associated with either the interface between the X-ray and the
 computer or in the peripheral hardware equipment.  Once these problems were
 corrected and a weekly cleaning schedule was established, operation of all
 associated equipment has improved significantly.


 Waste Solids Handling

          One of the most important considerations in use of lime/limestone
 scrubbing technology is disposal of waste solids.  A section of this meeting
 is scheduled for discussion of the work related to handling and disposal of
 sludge.  The test facility is equipped to study alternate methods for sepa-
 ration of solid and liquid phases in the scrubber discharge stream.

          Pond.  A three-section settling pond was constructed in an area
 that had previously been an ash storage pond.  The dikes are made from fly
 ash and are covered with local clay.  A small starter pond was used during
 the early open-loop tests and has been filled and retired.  The pond arrange-
ment which will be used for the remainder of the closed-loop program is
 discharge of sludge into a large settling area and return of the supernate
 through a smaller "polishing" pond.  The slurry to the pond can come directly
 from the scrubber circuit, from the thickener underflow, or from the filter
and centrifuge as reslurried cake.  The ponds are being equipped with instru-
mentation for evaluation of seepage.  During the early stage of use before a
significant area of the pond is covered with settled solids, seepage has been
excessive.   Small-scale permeability tests are being made in an effort to
 determine the ability of settled sludge to seal the clay lining.  An im-
pervious membrane may be required to reduce seepage to an acceptable level.
                                   349

-------
          Clarifier.  Each scrubbing system is equipped with a separate
 clarifier;  the venturi and Hydro-Filter  systems have  20-ft  diameter  units  and
 the TCA system, because of a higher recirculation rate, has a JO-ft diameter
 thickener.  The performance of the larger unit has been satisfactory but
 the solids  carryover in the overflow has been a problem with the smaller
 units.  The settling characteristics of the reaction products (particularly
 calcium sulfite) are poor and insufficient time is available during the
 compression or hindered settling period to produce a clear overflow.   Vari-
ation in the feed rate may contribute to poor performance.   It might  be
necessary to operate with a turbid overflow.   This probably would not be
serious if the solids in the overflow concentration could be controlled.

          The concentration of solids in the underflow approaches the ex-
 pected final settled density of sludge (approximately
          The program for improved clarification and thickening includes:

     •   Maintain the feed flow to the clarifiers as steady as possible.

     •   Optimize the oxidation of calcium sulfite to calcium sulfate.

     •   Investigate the effect of the particle-size distribution of the
         ground limestone on both the SQ^ absorption efficiency and
         settling characteristics and adjust limestone grinding accordingly.

     •   Maintain the percent solids in the clarifier underflow under
         density control and let the sludge level vary in the tank.  The
         set point of the density controller could then be adjusted
         (i.e., the underflow rate increased) to prevent sludge overflow.

     •   Operate the clarifier in series with the vacuum filter or centri-
         fuge to attain minimum solids concentration in the recycle liquor
         and maximum final solids concentration in the solids disposal
         stream.

     •   Use of flocculents.

          Filter.  Initial tests of a rotary vacuum filter to separate the
solid and liquid phases in the clarifier underflow were not successful.  The
cake is thixotropic and, although it appeared dry and firm under vacuum,
internal water was released and the cake became fluid when the vacuum was
released.  The wet, sticky cake would not separate from the filter cloth.
Further dewatering was restricted by formation of cracks in the cake which
allowed bypassing of air.
                                   350

-------
          The vendor has recommended installation of compression rolls  and
air blast discharge to reduce the final moisture to approximately ^0$.
These changes are being made*

          Centrifuge,  Short-term, exploratory tests of the centrifuge  were
carried out in late April.  Both clarifier underflow and a sidestream from
the scrubber recirculation loop were tested.  The results of these cursory
tests are shown below.
                      Summary of Centrifuge Tests
         Machine    Centrifuge     Peed    Wt $      Wt #      Wt %
  Test    speed,       feed       rates,  solids   moisture  solids in
 series    rpm        source       gpm    in feed   in cake   centrate

    I     2000    HP clarifier    11-22   15-22a    ^3-^7     0.2-0.6
                    bottoms
II     2000    HP clarifier    10-22
                 bottoms
                                                              0.5-0.5
  III     2000    HP clarifier    10-22   19-29*    39-^2     0.1-0.5
                    bottoms

   IV     2500    HP clarifier     9-33   19-2?a    36-^0     0.1-1.1
                    bottoms

    V     2500    Scrubber bleed  11-35   10-1^     37-^1     0.1-0.6
                    (clarifier
                    bypassed)
 a Increase the values by about 3 for pump seal water correction.
 * Test Series II was a replicate of Test Series I.
           It appears  that the  centrifuge  is  effective in reducing the
 solids content well below the  level attained by settling and that the  cen-
 trate clarity is  satisfactory.   The centrifuge  will be incorporated into
 the loop of one of the scrubbing systems  for long-term tests.
                                    351

-------
Materials Evaluation

          Selection of construction materials for components of scrubbing
systems  is an  important economic consideration.  One objective of the test
program  is to  evaluate corrosion/erosion rates for alternate materials.
The evaluation involves comparison of several different materials in
installed components as well as exposure of test coupons at appropriate
locations.

          Equipment Inspection.  A thorough inspection of al.1 system com-
ponents  was conducted during the extended boiler outage.  The systems had
been operated  during limestone scrubbing tests that totaled about 1800 hours
for each train.

          The mild steel gas ducts between the boiler and scrubber inlet
transitions had localized deposits of loose fly ash.  The surfaces had a
thin coating of iron oxide scale except at uninsulated flanged connections
where moderate pitting had occurred; these flanges are now insulated.

          The rubber lining in the scrubbers was in excellent condition;
no erosion or deterioration was noted.  The rubber linings in pumps, piping,
and process water tanks were also in excellent condition.  Slight wear was
noted on some of the rubber-coated agitator blades.

          The only damage noted in the Flakeline lining in effluent hold
tanks and clarifier tanks were several hairline cracks which did not appear
to penetrate the entire thickness.  The cracks were most prevalent at the
junctions between the baffles and tank walls.

          The most severe corrosion was found on type J>l6 stainless steel
surfaces particularly on mist eliminator blades in the TCA system.  The
general  attack was in the form of pitting and some pits were as large as
1/16-inch diameter and 50 to 55 mils deep.

          The only significant erosion of equipment noted was on pump sleeves
and at intersections of wires of support grids in the TCA scrubber.  Weight
loss of mobile bed packing material has also been detected.  The polypropylene
(and polyethylene) spheres in the TCA scrubber have worn noticeably.  Some
are so thin that they have lost mechanical strength and the walls have col-
lapsed; a random sample of these showed about 60$ weight loss.  The bulk of
the spheres were still intact but had an average weight loss of 20$.  The
glass marbles in the Hydro-Filter have lost about 6% of their initial weight.
                                   352

-------
          Test Coupons*  Test coupons of several different materials of con-
struction were exposed for 70 days or more in various slurry and gas
environments.  Stressed and welded metallic materials were also tested.

          Corrosion of Hastelloy C-276 was negligible to 5 mils per year;
this alloy showed no evidence of localized attack in any test location.
Next in resistance were alloys Inconel 625, Incoloy 825, Carpenter 20Cb-3,
and Type Jl^L stainless steel with corrosion rates for each material rang-
ing from negligible to 5* 7, I1*, and 15 mils per year, respectively.  These
alloys had very few minute pits and/or crevice corrosion.  One specimen of
Type 3l6 stainless steel was grooved and the weld of another specimen was
attacked.  Type J516L is the fifth alloy in resistance and the least expensive
of this group.

          Three nonferrous alloys, Cupro-Nickel 70-30, Monel UOO, and
Hastelloy B each had minimum rates of < 1 mil and maximum rates of ^9, 57,
and 100 mils per year, respectively, with 1 or 2 specimens pitted.  In three
tests of Monel and in one test of Cupro-Nickel 70-30, the welds were inferior
to the parent metal.

          A group of five alloys, Type kk6 stainless steel, E-Brite 26-1,
Incoloy 800, USS lS-lB-2, and Type 301* stainless steel, had rates that ranged
from negligible to a "greater than" (>) value which indicates that the speci-
men was completely destroyed at one or more test locations.  The values for
failures ranged from > l4o mils per year for Type kk6 to > 200 for both
USS 18-18-2 and Type 30UL stainless steels.  These five alloys were highly
susceptible to localized corrosion.

          Another group of alloys, Type 1*10 stainless steel, aluminum 3003,
mild steel A-283, and Cor-Ten B, had minimum rates of < 1 mil per year and
maximum rate of >250 for Type UlO to > lUoo for mild steel and Cor-Ten B.
Pitting and crevice corrosion occurred on the four alloys.

          In general the stressed specimens (five alloys only) were not
corroded at rates higher than their counterpart disk-type specimens.

          Specimens of Bondstrand ^000, Flakeline 200, and Transite were
tested at 21 locations.  Bondstrand showed good resistance in 12 tests and
poor in 9 tests.  The evaluations for Flakeline and Transite were:  good,
2 and l^j fair, l4 and 2; and poor, 5 and 5, respectively.  Only 6 specimens
each of Qua-Corr plastic and of butyl, natural, and neoprene rubbers were
tested.  The results were 5 good and 1 poor for Qua-Corr and 6 good for each
of the rubbers.
                                  353

-------
              Recommendations Based on Experience to Date


          Some of the preliminary conclusions discussed in this section may
change as more information is obtained during long-term tests.  However,
several large-scale installations are in the design stage and some of the
trends which have been noted during operation of the test facility are worth
considering.


Procea s jgentrol

          During the open-loop testing,  runs at a given set of conditions
were normally short (2 days or less).   Process control during these tests was
accomplished by setting the limestone feed rate at the value required for the
desired stoichiometry based on S02 concentration in the inlet gas; slurry pH
and removal efficiency were, therefore,  dependent variables.  This approach
met the objectives of the test program.

          When closed-loop testing was begun in March, the program emphasis
was placed on long-term reliability tests.  Because of the effect of pH on
scale formation in the tests discussed under stoichiometry, it was decided to
operate with pH control; the limestone feed rate is manually adjusted to main-
tain a nearly constant value of pH.  System response has been good in the
relatively low pH range  (5,8-6.0) studied.  It appears that if reliable pH
meter operation can be established, automatic control of limestone feed rate
to maintain pH at values below 6.0 would be feasible.


Gas Cooling

          If a heat sensitive lining is used  in the scrubber for  erosion/
corrosion protection, gas cooling ahead of the scrubber is required.  Efforts
to control solids deposits at the point of liquor or slurry addition have been
only marginally effective when sprays are used.  The venturi scrubber configu-
ration provides a clean  separation between the wet and dry areas  and solids
have not accumulated  at  the transition point.

          When gas  cooling by humidification  ahead of the absorber  is required,
a venturi scrubber  may be the most effective  device for addition  of the  liquor.


Mist Eliminator

          Most of  the testing to date has been with excessive  fresh water
addition for washing  the mist eliminators.  This  operation  is  not consistent
with the long-term reliability  evaluation.  The  systems recently  have been
                                   354

-------
modified to permit use of recycled water for wash of the mist  eliminators.
The liquor has varied from a ratio 1 part fresh water to 3  parts  clarified
liquor to a half and half mixture.  In the Hydro-Filter and venturi  scrubbers
the full underside is washed intermittently at a rate of about 1  gpro/ft2  on
a cycle that has averaged 1 minute on and 3 minutes off.  In the  TCA,  the
wash is added on a valve tray beneath the mist eliminator and  entrained liquor
from the tray flushes the underside of the eliminator.

          No significant scaling has been detected although some  solids de-
posits have occurred.  It appears that precipitation on mist eliminator
surfaces can be controlled by dilution of clarified liquor  with available
makeup water.
Reheaters

          Use of direct-fired, in-line, oil burners for reheat can lead to
incomplete combustion and accumulation of oil-saturated soot.   The problem
apparently has been solved at the test facility but turndown is complicated
by the need for nozzle substitution.

          Use of an external combustion system and admixture of completed
burned combustion products with the scrubber exhaust gas would be a more
satisfactory system for direct-fired reheat.
Clarifiers

          The small (20-ft diameter) Clarifiers appear to be underdesigned
to provide both a dense underflow (kQ% solids) and a clear overflow.  The
settling rates and settled density of the sludge are poorer than the design
basis and preliminary results indicated the volume of the Clarifiers should
be increased; the additional volume required has not been established.

          The Clarifiers appear to be underdesigned to provide both a dense
underflow (kofi solids) and a clear overflow.  The settling rates and settled
density of the sludge are poorer than the design basis and preliminary re-
sults indicated the volume of the Clarifiers should be doubled.
SOo Analyzer

          The continuous, on-line S0a analyzers  (Du Pont Model 400 UV) are
accurate and reliable provided that high-quality gas samples are delivered
to the units.  A nonplugging probe is needed and heated, well-insulated sample
lines (Dekeron) are required to prevent condensation and plugging.  Precise
calibration techniques are necessary.


Materials of Construction

          Rubber lining of scrubbers, tanks, pumps, and piping has been
entirely satisfactory.  Use of less expensive polyester-glass lining may be
acceptable, but further exposure is required before a firm conclusion can be
reached*  Erosion of mobile-bed spheres has been excessive; improved materials

                                   355

-------
are being tested.  Type Jl^L stainless steel has been badly pitted in
partially wetted areas, particularly the surfaces of mist eliminators.
Nonmetallic materials will be tested in these areas.   Test coupons of
highly resistant alloys (HastelUoy, Carpenter 20) have shown no signs of
attack, but these are relatively expensive.
                                  356

-------
               SCRUBBING EXPERIMENTS
        AT THE MOHAVE  GENERATING STATION
                          by

 Alexander Weir, Jr., Principal Scientist for Air Quality
Lawrence T. Papay, Director of Research and Development
          Southern California Edison Company
               2244 Walnut Grove Avenue
                 Rosemead,  California
                           357

-------
                            PREFACE

The pilot testing program was set up to make comparative tests
of different scrubber types and reagents to determine SO^ and
particulate removal capabilities.  Reliability testing of mech-
anical equipment and hardware was not a part of the program.
The performance results in the succeeding discussions therefore
do not imply that reliable operation of these levels can be
attained.  The follow-on module program described will hopefully
give a better measure of the total performance.  It should be
pointed out though that with the press of time a large step is
being taken in proceeding from the pilot to the full size module,
viz., good engineering practice would dictate an intermediate
prototype first.
                               358

-------
This paper presents work performed at the Mohave Generating
Station during 1971 and 1972.  Eight pilot plant scrubbers and
four different reagents - soda ash, limestone, lime and ammonia -
were studied.  We intend to present some of the highlights of
this work as well as the conclusions which allowed us to select
two of the scrubber types for construction of two 450,000 scfm
scrubbers.  The work reported here was performed by a number of
contractors and paid for by an even larger number of organizations

Figure 1  (attached) presents some of the organizations which
participated in this work.

Figure 2 presents a further description of the scrubbers.

The scrubbers were all essentially cylindrical in shape and,
as may be seen, most of the scrubbers tested were between two
and three feet in diameter,  with the exceptions being the six
foot diameter spray drier and the 19-inch diameter TCA.  The
scrubber lengths ranged from eight feet to 27 feet.  However,
although the scrubbers were somewhat comparable in physical size,
their maximum gas handling capacity varied considerably ranging
from 1100 scfm in the WPS to 3500 scfm in the SCE scrubber.  It
should be noted that the flow rates listed were not necessarily
the optimum flow rate for SC^ or particulate removal but the
maximum (except for the SCE scrubber) under which the scrubber
would operate.  The capacity of a scrubber, rated in scfm per
megawatt varies from station to station, depending on the type
of fuel, the excess air used in combustion, the amount of leak-
age through air preheaters and other factors.  At Mohave we have
determined that 2800 scfm is approximately equal to one megawatt
downstream of the electrostatic precipitators resulting in tested
scrubber capacities from 0.4 to \\ megawatts.  Actually, the
capacity of the scrubbers is a function of the actual linear
velocity through the scrubbers.  While the temperature of the
gas entering the scrubber varies considerably, the exit gas
temperature remains constant, ranging from 120° to 130° F regard-
less of the outside air temperature.  The gas at the exit is
saturated with water.  By using the exit gas conditions, the
cross-sectional area of the scrubber, and the maximum scfm
capacity of the scrubber we are able to calculate the "super-
ficial" (i.e. disregarding scrubber packing) linear velocity
through the scrubber and thus determine a relative "size factor"
for extrapolation to larger scrubbers.  In Figure 2 the scrubbers
are listed in order of increasing linear velocity.

Finally, the last column in Figure 2 indicates the reagents
tested with the scrubbers.  Most of the experiments with AIS
scrubber were performed with soda ash but a few experiments
                               359

-------
                                            Figure 1
                           MOHAVE  SCRUBBER  PROGRAM
              SCRUBBER
                              MANUFACTURED BY   TESTING PERFORMED BY  WORK SPONSORED BY
U)
CT>
O
     WPS HYDRO  PRECIPITROL
     CVS CHEMICO VENTURI
US   PEABODY-LURGI IMPINGE-
     MENT SCRUBBER
                             DESEVERSKY ELECTRO-   SCE-TRUESDAIL LABS
                             NATOM  CORP.
CHEMICAL CONSTRUC-
TION COMPANY

PEABODY ENGINEERING
CO.
     PPA POLYPROPYLENE  ABSORBER   HEIL WITH FLUOR
                                 PACKING
     TCA  TURBULENT CONTACT
          ABSORBER

     SCE  EDISON SCRUBBER

     AIS   ATOMICS INTERNATIONAL
          AQUEOUS CARBONATE
          SCRUBBER

     RHS  ROTATING  HORIZONTAL
          SCRUBBER
                             UNIVERSAL OIL
                             PRODUCTS

                             STEARNS-ROGER, INC.

                             BOWEN ENGINEERING
                             CO.
                             HAZEN RESEARCH, INC.
SCE-BECHTEL CORP.


SCE-BECHTEL CORP.


SCE-BECHTEL CORP.


SCE-BECHTEL


SCE-TRUESDAIL LABS

SCE-ATOMICS INT'L.
                    SCE, U.S. LIME.
                    HAZEN RESEARCH AND
                    NLA
                                        SCE
NAVAJO AND MOHAVE
PARTICIPANTS

NAVAJO AND MOHAVE
PARTICIPANTS

NAVAJO AND MOHAVE
PARTICIPANTS

NAVAJO AND MOHAVE
PARTICIPANTS

SCE

16 UTILITIES WHO ARE
MEMBERS OF WEST
ASSOCIATES

NATIONAL LIME ASS'N
AND SCE

-------
                                        Figure 2

       MOHAVE PILOT PLANT SCRUBBER TEST  CONDITIONS
U>
        SCRUBBER

AIS  CONVENTIONAL
     SPRAYDRIER
     FOLLOWED BY CYCLONE
     SEPARATOR

RHS  HORIZONTAL LIME KILN
     CONTAINING BALLS OR
     CHAINS

CVS  SINGLE STAGE
     VENTURI WITH FIXED
     ANNULAR THROAT
     FOLLOWED BY CENTRI-
     FUGAL SEPARATING
     CHAMBER

WPS  VERTICAL ABSORBER
     FOLLOWED BY WET ELEC-
     TROSTATIC PRECIPITATOR

LIS  VARIABLE THROAT
     VENTURI FOLLOWED BY
     THREE-STAGE IMPINGE-
     MENT  TRAY VERTICAL
     ABSORBER

PPA  VERTICAL ABSORBER
     PACKED  WITH 7 1/2 FT.
     OF PLASTIC "EGG-CRATE"
     PACKING

TCA  VERTICAL TURBULENT
     CONTACT ABSORBER
     CONTAINING THREE
     STAGES OF "PING PONG
     BALLS"

SCE  FOUR-STAGE
     HORIZONTAL CROSS-
     FLOW  SCUBBER
                                 HEIGHT
                                  OR
                                (LENGTH)


                                  12
                                  10
                                  8
                                 15 1/2
                                  14
                                 12 1/2
                                  27
DIAMETER
 (INCHES)
   I.D.

   72
  36
 26 3/4
 27 1/4
  30
  24
   19
 MAX. GAS
FLOW RATE
  TESTED
(SCFM 60°F)

   1375
   1300
   1037
   1100
   1332
   1850
   1230
 LINEAR
VELOCITY
AT EXIT

  0.9
  3.4



  5.0
  5.1
  5.1
  11.0
  11.7
                                  15
  22
  3500
  24.9
   TEST
 REAGENTS

 SODA ASH
   (LIME)
   LIME
SODA ASH
   LIME
LIMESTONE
 AMMONIA
SODA ASH
   LIME
LIMESTONE
SODA ASH
   LIME
LIMESTONE
SODA ASH
   LIME
LIMESTONE
,   LIME   v
(LIMESTONE)

-------
were performed with lime.  On the other hand, most of the
experiments with the SCE scrubber were performed with lime,
with only a few experiments being performed with limestone.
Ammonia, as gaseous ammonia and as ammonium hydroxide, was used
only with the WPS.

Some of the variables studied in addition to the scrubber and
reagent type included reagent composition (including pH and
percent solids) and, in the case of lime and limestone slurries,
the effect of soluble sodium salts such as occur in cooling tower
blowdown water.  The gas flow rate was an important parameter.
As previously mentioned, there appeared to be an optimum gas flow
rate for each of the scrubbers tested.  Some of the scrubbers
were operable over wider ranges than the other due to their con-
struction.  For example, with the TCA scrubber at too low a gas
velocity the ping pong balls stayed immobile on the lower support
screen while at a higher velocity they congregate at the top of
the stage, i.e., the bottom of the next higher stage.  The liquid
flow rate and the L/G ratio (GPM/1000 scfm) were also important
variables.

The number of stages was varied with the TCA scrubber and the SCE
scrubber and we were quite pleased to find that our theoretical
predictions on the effect of staging were confirmed experimentally,

The gas pressure drop was measured under a variety of conditions
and this, of course, determined the fan requirements for larger
size scrubbers.  We operated with both FD and ID fans and per-
haps we can summarize our findings in this area by indicating
that both of our 450,000 scfm test modules will have FD fans even
though the power required because of the higher gas temperature
will be greater.

We also studied sludge separation with centrifuges, rotary filters
and thickeners, the degree of oxidation from sulfite to sulfate
and the properties of the slurry and the sludge itself.

It is not possible in the time allotted to present much more than
a cursory overview of the effect of the different variables or
to itemize all of the operating problems encountered with the
various scrubbers.   Primarily, the latter resulted from plugging
in scrubbers,  demisters, reheaters, centrifuges, and piping.


Effect of Inlet SC>2 Concentration and L/G

The Mohave Generating Station  is located some 90 miles south of
Las Vegas and burns low sulfur coal transported by a 285-mile
slurry pipeline from the Black Mesa Mine.  The average sulfur
                              362

-------
content of the coal burned over the last two years was about 0.3870
sulfur which results in less than 200 ppm of S02 in the exhaust
gas.

After we initiated our experiments, it became evident that the
S02 inlet concentration varied with the station operating condi-
tions.  In order to have consistent data we had to maintain the
inlet S0« concentration at a constant value.  Accordingly, we
arbitrarily spiked the inlet gas to 400 ppm of SOo.  This corres-
ponds to about 0.83% sulfur coal, the highest sulfur content coal
predicted for Mohave.  In the two 450,000 scfm scrubbers that are
currently under construction we will have to burn about five tons
per day of sulfur per each scrubber in order to increase the con-
centration to 400 ppm for test purposes.

Most of the data presented in the remainder of this paper is
based on an inlet gas concentration of 400 ppm of S02-

We did perform some experiments, however, with five scrubbers on
the effects of inlet S02 concentration.  The data from one of
these scrubbers is presented in Figure 3.

Prior to performing these experiments, our calculations, based
on a rather simplistic theory, indicated that we should obtain
straight, parallel lines at lower SOo concentrations, the degree
to which they can be considered parallel is debatable.  The shape
of the curve near the origin varies with different scrubbers,
being convex in some cases rather than concave as shown here.

Figure 3 does show, however, the effect of L/G on S02 exit con-
centration, i.e. increasing L/G increases the S02 removal.
However, the CVS scrubber exhibited a minimum in the curve.
That is, after a certain L/G was reached an increase in L/G
resulted in an increase in S02 exit concentrations.

Often considerations of SCH removal specify scrubber performance
in terms of percent sulfur removal without regard to the inlet
S02 concentration.  When the data in Figure 3 is presented in
terms of percent SOo removed, as is shown in Figure 4, it is
obvious that the inlet S02 concentration is an important factor
in characterizing scrubber performance.  While this data is for
one scrubber at one gas flow rate, we have observed similar
trends at other flow rates with four other scrubbers.  Apparently
this is due to the solubility of the reagent, the gas flow rate
and the gas-liquid contacting mechanism.

In summary, Figure 4 indicates that for a given L/G ratio and gas
flow rate, a decrease in the inlet S02 concentration results in
an increase in the S02 removal efficiency, until some maximum
                               363

-------
              Figure 3

  EFFECT  OF L/G ON OUTLET SO2
CONCENTRATION AT DIFFERENT INLET
  CONCENTRATIONS SCE SCRUBBER
            -2500 SCFM
  1OOO
                 1000 PPM
                   INLET
                          1600 PPM
                          / INLET
                         600 PPM
                           INLET
         200 PPM
           INLET
                          400 PPM
                           INLET
               L/G
               364
                    GPM
                  1OOO SCFM

-------
                FIGURE 4
     EFFECT OF INLET SC>2 CONCENTRATION
          ON PERCENT SO2 REMOVAL
100
                        SCE SCRUBBER
                          2500 SCFM
       200
400
600
1000
1600
          INLET SO2 CONCENTRATION (PPM)
                  365

-------
 removed  efficiency  (around 400 ppm of  862)  is reached.  At lower
 values of  SOo  inlet  concentration, it  becomes increasingly dif-
 ficult to  achieve high percentage 862  removal.  Thus blanket
 edicts to  remove a certain percentage  of  the sulfur in  the fuel
 pose  a problem for the scrubber designer  as well as the utility
 user.


 Reagent  Comparison

 Figure 5 presents a  comparison of three different reagents in  the
 same  scrubber, the packed grid tower  (PPA). As may be seen, with
 this  scrubber  soda ash is a much more  efficient scrubbing reagent
 than  limestone while lime is intermediate between the two.  Also
 shown on Figure 5 are a few data points obtained with ammonia.
 Similar  data for soda ash, lime and limestone obtained with the
 TCA scrubber are presented in Figure 6 along with the same ammonia
 scrubbing  data obtained with the WPS.  The  data of Figures 5 and 6
 indicate that  the combination of ammonia used with the WPS scrubber
 is superior in S02 removal to soda ash, lime, or limestone used
 in either  the PPA or TCA scrubbers.  However, the experiments  with
 ammonia  were discontinued because of plugging difficulties in  the
 scrubbing  section of the WPS when operating on boiler flue gas.
 In addition, significant amounts of unreacted ammonia passed
 through  the scrubber which creates an  additional air pollution
 problem  by substituting NHg for S02-

 The problems associated with NH^: its  cost, the solubility of
 the sulfite-sulfate in water, disposal, the lack of a regeneration
 process, and the possibility of substituting ammonia for S02,
 contributed to the decision not to perform  any further tests with
 NH3, even  though it appears to be the most  reactive reagent.

 While Figure 5 indicated that soda ash was  the most effective
 scrubbing  reagent at any given L/G ratio with the PPA, the data
 of Figure  6, with the TCA scrubber operating at its optimum gas
 flow rate  of 1020 scfra, present a different story.  With this
 scrubber,  the data presented indicate  that  lime is a superior
 reagent  to soda ash over the operable  L/G range, and that lime-
 stone is superior to both lime and soda ash at L/Gs greater than
 58.  A similar anomaly also exists with the LIS scrubber as shown
 in Figure  7.  With this scrubber, while soda ash is superior to
 lime, limestone is superior to soda ash at  L/G ratios above 47.
 The superiority of limestone to lime in this scrubber was ex-
 plained by the plugging with lime slurry on the underside of the
 impingement trays.  The reason for the superiority of limestone
 to soda ash is difficult to explain.  You will note, however,
 that the S02 levels in the region of interest in Figure 7 are
below 6 ppm S02.   The same type of instruments (Dynasciences)
                               366

-------
                 Figure 5
         REAGANT COMPARISON
   400
       T
T
                    T
  T
 T
a.
o.
<
cc
LLJ
O

O
o

 CM
o
CO

K-
III
   200
Q  100
80
40
    20
    10

    8
                 LIMESTONE
                 PPA-139O SCFM
  LIME
 PPA-139O SCFM
      ©

   AMMONIA
   WPA-1OOOSCFM
          SODA ASH
        PPA-139O SCFM
            I
          10
           20
    30
                 L/G
 40

GPM
50
                     1000 SCFM
60
70
                    367

-------
                  Figure 6
          REAGANT COMPARION
 6OO
 4OO
  2OO
Q.
Q_
<
DC
LU
o
    LIMESTONE
    TCA 1O2O SCFM
  1OO

   80
   6O
O  4O
LU
_l


O
   2O
   1O
                 SODA ASH
                 TCA-1O2O SCFM
-  AMMONIA
0 0   WPS-
   1OOO SCFM
                        LIME
                       TCA-IO2OSCFM
             I
         I
                      I
I
1
        10
    2O   30


        L/G
                      4O   5O

                      GPM
    6O
7O  8O
                   1000 SCFM

                   368

-------
  600
  400
                 Figure 7
          REAGANT COMPARISON
              LIS-980SCFM
  200
Q.
Q.



O
UJ
O
O

CM
100-
                L/G
                   1000 SCFM

                   369

-------
were used  in  testing the PPA, TCA, LIS,  CVS,  SCE and AIS scrubbers
and in all four  instruments  the presence of NOX influenced the
wet electrochemical cello  Specifically  NOX interference in the
instrument resulted in readings of SOo 8-10 ppm higher than actual.
This correction  had to be  subtracted from  the instrument reading.

It is possible that the superiority of limestone to soda ash in
these two  scrubbers is possibly due to measurement error.  However,
the superiority  of lime over soda ash in the  TCA scrubber is more
difficult  to  explain and neither our test  contractor nor their
consultants was  able to explain this anomaly.

With the few  experiments with lime performed  in the AIS scrubber,
soda ash was  more reactive,  while in the SCE  scrubber lime was
a more reactive  reagent than limestone.  Comparative data with
the three  reagents utilizing the Chemico Venturi scrubber is
presented  in  Figure 8.  With this scrubber the exit SOo levels
are high enough  (168 ppm with limestone, 120  ppm with lime, and
55 ppm with soda ash, based  on 400 ppm inlet  concentration) to
eliminate  possible instrument measuring  errors.  This data indi-
cates, as  expected, that soda ash is superior to lime which in
turn is superior to limestone.  In this  scrubber, operating at
its optimum gas  flow rate of 950 scfm, the S02 removal efficiency
(based on  400 ppm inlet S02  concentration) at the optimum L/G
ratio of 43,  was about 86% with soda ash compared to 58% with
limestone.

Because of the solubility of sodium salts, a  sodium system offers
many advantages  due to reduced probability of scrubber plugging.
However, the  cost, without regeneration, is high.  Approximately
$500,000 dollars was spent trying to regenerate the sodium scrub-
bing solution on a pilot plant scale without  success.  Laboratory
experiments indicated that we could regenerate a sulfite-bisulfite
solution with lime, but not with limestone.   However, on the pilot
plant scale using actual stack gases, we found considerable oxida-
tion of the sulfite to sulfate and we were not able to regenerate
the sulfate.  In addition, the regeneration pilot plant design
was such that the calcium and sodium systems  become intermingled
and we had plugging of scrubbers and centrifuges with the calcium
salts.  We hope  that the other papers in this symposium on double
alkali processes will be able to report more  success with their
experiments than we are able to report.


Scrubber Comparison

Comparative data on five scrubbers operating  with soda ash is
presented  in  Figure 9.  This indicates that the AIS scrubber
requires a much  lower L/G ratio to reach an exit gas concentra-
tion of 40 ppm (90% removal) than the TCA, PPA, LIS or CVS scrubber
                               370

-------
500
400
               Figure 8
       REAGANT COMPARISON
           CVS-950 SCFM
       10
             L/G
                1000 SCFM
               371

-------
             Figure 9
SCRUBBER COMPARISON-SODA ASH
                 CVS
               88O SCFM
                  AIS
                    SCFM
 /1300
                          TCA
                        1O1O SCFM
                      PPA
                    144O SCFM
  -  9OO SCFM
     1O  2O
30


L/G
 4O   5O

   GPM
10OO SCFM
6O   70   8O
              372

-------
as well as indicating that the CVS scrubber, as tested, was unable
to achieve an exit S(>2 level of 40 ppm even with soda ash.

Similar data with limestone is presented in Figure 10.  As indi-
cated, the LIS scrubber (a Lurgi Venturi followed by a three-
tray tower) was able to achieve the lowest S02 exit concentra-
tions (around 7 ppm 802) while the CVS scrubber minimum SO? con-
centration (ca. 170 ppm) was considerably above the desired 40
ppm.

Data on six scrubbers with lime is presented in Figure 11.  This
data indicates that the RHS scrubber requires the lowest L/G ratio
to achieve the desired 40 ppm S02 outlet concentration.  The
mechanism of operation of this scrubber makes it difficult to
really determine the real ratio of liquid to gas because of the
"random" method of contacting.  The L/G presented in Figure 11
for the RHS is based on the lime feed rate, the actual amount
of liquid contacted per unit of gas flow is undoubtedly higher.

The L/G ratio used for the SCE scrubber was the actual amount
of liquid contacted with the gas, but since this scrubber is a
four-stage device, this amount of liquid is contacted with the
gas four different times.  Thus, the L/G as plotted is correct
for comparison of mass transfer performance, but the amount of
liquid pumped should be multiplied by four to compute the horse-
power required for liquid pumping to be on a comparable basis
with the other scrubbers tested.  The total horsepower required
is composed not only of the liquid pumping horsepower  (and in the
case of the RHS, the power required to rotate the scrubber) but
also that of mixers, thickeners, instruments, etc.  The major
contributor in most of the scrubbers tested, the fan power re-
quired, is a function of the pressure drop through the system.


Reagent Utilization

The composition of the scrubbing reagent influences the degree
of reagent utilization as well as scrubber performance.  At Mohave,
cooling tower blowdown water is to be used as makeup water for
the scrubbers, therefore experiments with added soluble sodium
salts were performed.  This was found to increase the scrubber
performance with limestone slurries.  Calculations indicated
that with a 5% NaCl - 5% Na2S04 solution saturated  with CaC03,
the sorption capacity on a once-through basis was almost five
times greater than with a solution saturated with CaC03 alone.
These calculations also indicated that the contribution of sodium
sulfate is more effective than that of sodium chloride.  Experi-
ments with sodium salts added to a soluble lime slurry did not
seem to improve the performance of the SCE scrubber.
                               373

-------
                Figure 10

 SCRUBBER COMPARISON - LIMESTON
  4OO
CL
Q.

-------
                 Figure 11

     SCRUBBER COMPARISON - LIME
  4OO
  28O
  2OO
a.
o.


O 12°
z
UJ
o

§
o


ON
CO


UJ
80




6O
40



32


28
   2O



   16




   12

                      CVS
                    1OOO SCFM
      II

      II
        -RHS

        975 SCFM
      I





     \  \
           0
              PPA

             139O SCFM
                            TCA

                          1O7O SCFM
      SCE

    3OOO SCFM

    - 1
                 LIS

            © 131O SCFM
          10
           2O   3O
                 L/G
                 375
                        40
                    GPM
SO
6O
7O
                    1OOO SCFM

-------
Increasing the suspended solids  (i.e. calcium sulfate, calcium
sulfite and unreacted calcium carbonate) concentration from 5%
to 15% and reducing the calcium  carbonate concentration from
1.5 to 0.5% would increase the limestone utilization rate since
less unreacted limestone would be lost with the filter cake.

Thus, in the first case
               limestone utilization - 1 - 1.5 - 70%
While in the second case
               limestone utilization - 1 - 0.5 = 96.6%

Experimental values in the first case were around 66%, while
in the second case were around 95% as indicated in Figure 12.
As indicated, experimental determination of the degree of reagent
utilization is difficult.  In this four-day test, five different
methods were used, four of them based on time-consuming wet
chemical analyses of the cake and the slurry, the fifth based
on the S02 removal from the gas.  As may be seen, this latter
method indicated a limestone utilization rate of 117%.

It should be pointed out that, while increasing the solids content
in a limestone scrubbing slurry increases the degree of reagent
utilization, it also increases erosion of nozzles and pumps as
well as increasing the probability of plugging.  It was for this
reason that our 450,000 scfm Vertical Module is being designed
to use a 5% rather than a 157o solids slurry.

Experiments with the SCE scrubber using a soluble lime solution
were made to determine the degree of reagent utilization.  In one
75 -hour test, this was determined to be 93.3% while in a second
159-hour test, the degree of reagent utilization was 92.8%.


Particulate Removal

An objective of the Mohave Test Program equal in importance to S02
removal has been that of particulate removal downstream of the
98% efficient electrostatic precipi tator s .  The dust loading down-
stream of these precipi ta tor s , based on two years of operating
experience and the manufacturer's guaranteed performance with all
sections in service, is 0.05 grains /scf.  Optimum precipitator
performance has typically met the manufacturer's expected per-
formance with all sections in service of a dust loading of 0.03
grains /scf.  Most of the data presented in Figure 13 was based
on inlet grain loadings ranging from 0.01 to 0.02 gr/scf.  How-
ever, with some precipitator sections out of service, due to
discharge wire failures, the precipitator manufacturer would
                               376

-------
                          FIGURE 12
           DETERMINATION OF LIMESTONE UTILIZATION
                   BY DIFFERENT METHODS
10
            ttt>

-------
 expect grain loadings as high as 0.095 gr/scf and we have occasion-
 ally recorded grain loadings this high.   Figure 13 presents a
 composite of the particulate removal data obtained downstream of
 these precipitators with five different scrubbers.  There are
 many variables which affect particulate removal performance,  for
 example increasing the gas flow rate (and hence the pressure  drop)
 seems to increase the particulate removal performance on the  CVC,
 PPA,  TCA and LIS scrubbers and in the SCE scrubber at gas flow
 rates above  1750 scfm.   Increasing the liquid flow rate on the SCE
 scrubber is  beneficial,  at 2500 scfm,  the particulate removal
 efficiency is increased from 75% to 90% by increasing the liquid
 flow rate from 50 to 66 GPM.   At a given liquid and gas flow rate,
 we have obtained good data correlations  (with TCA, PPA,  CVS,  and
 SCE  scrubbers)  on a log-log plot of % particulate removal versus
 inlet grain  loading.   A straight line results with increased
 particulate  removal at higher inlet grain loadings.

 This  increased efficiency at  higher inlet grain loadings is un-
 doubtedly due to the fact that at higher dust loadings  the stack
 gas  is composed of larger diameter particles  due to  precipitator
 sections being out of service.

 Particle sizes  in the stack gas at Mohave are normally  very small.
 About 90 cumulative wt  %  of the particles are less than 4 microns
 in diameter  with 70 wt  %  less than 1 micron in diameter,  40 wt %
 less  than 0.5 micron in diameter and about 15 wt % less  than  0.3
 micron in diameter.   With the SCE scrubber, our test data indi-
 cated about  75% removal of 0.3 micron  diameter particles,  87%
 removal of 0.5  micron particles,  96% removal  of 1.0  micron dia-
 meter particles,  and 97%  removal of 1.5  micron diameter  particles.


 Pressure Drop

 One important characteristic  of a scrubber which contributes  a
 great deal to both  the  capital  cost and  the operating cost is  the
 pressure drop through the scrubber.  The  pressure  drop is,  of
 course,  primarily a function  of the mechanical design of the
 scrubber.  Plugging of  the scrubber or demister and  operation
 near  the flooding point  (particularly with the TCA scrubber and
 the LIS  scrubber; caused  extremely  high  pressure drops.  Partial
 plugging occurred with all of  the  scrubbers tested,  particularly
 with  hydrated lime  slurries,  except with  the  SCE scrubber.  This
 scrubber also had the lowest  pressure drop  (ranging  from 
-------
                  FIGURE 13
             MOHAVE PILOT PLANTS
                  PARTICULATE
                 REMOVAL DATA
20
        500
 1000     1500     2000

GAS FLOW RATE-SCFM
       379
2500
                                              3000

-------
It was found that the contribution of skin friction and wake
contribution of the stage separations was about 7  (% £ v2) while
that of  the liquid spray was about 0.4  (% fv2)(L/G) so that pres-
sure drop could be correlated by the following equation:

               P -   [7 + 0.4 (L/C)]    (% e v2)

Comparison of pressure drop through the various scrubbers is
presented in Figure 14.


Operating Problems

Insufficient time prevents a detailed discussion of all of the
operating problems encountered in the pilot plant program.  As
previously mentioned, tests with the WPS were discontinued pri-
marily due to plugging, although unreacted ammonia passing through
the device also was a contributing factor in this decision.  There
was an accumulation of solid material at the wet-dry interface
with the RHS device, the lime utilization rate was high, and
particulate removal performance was marginal.

The planned test program of the AIS was completed but the mode
of operation was found to be different than that originally
proposed.  Initially it was planned to operate the spray drier
with a thick slurry feed so that the spray droplets would be
"BB sized11 particles when dried.  It was postulated that this size
particle could easily be collected in the downstream cyclone
separator.  However, when the spray drier was operated in the
slurry (as contrasted to solution) mode (about 30 wt % Na2COo)
it was found that the soda ash utilization rate was about 25%,
making the reagent cost prohibitive.  When the spray drier was
operated with lower concentration solutions, the droplets formed,
and hence the dried particles, were too small to be collected
efficiently in the mechanical separator    In many of the experi-
ments, the dust loading leaving the system was greater than the
inlet dust loading.

Bechtel's test report on the LIS device stated "The Lurgi im-
pingement scrubber scaled up so badly several times in the lime
slurry tests that the desired gas flow rate could not be maintained."
Various parts had to be wire brushed, scraped, or acid cleaned.
The vanes above the venturi also had to be acid cleaned and the
vane demister plugged.  In the limestone tests, appreciable build-
up occurred on the underside of the impingement trays as well as
a considerable accumulation of solids in the bottom of the liquid
downcomer trays.  Flooding difficulties caused cancellation of
the tests at the higher liquid and gas rates.
                               380

-------
oo
                         Figure 14

                SCRUBBER PRESSURE  DROP

                   AP      GAS FLOW RATE       L/G
  SCRUBBER    INCHES OF H?0      SCFM      GPM/100Q SCFM


    SCE             10           3000            17
    PPA            2.5            1390            54
    TCA            6.2            1020            67
    CVS            9.0            950            36
    US            11.5            1300            35

-------
 The  PPA liquid  inlet  spray header plugged  several times as did
 the  coarse  screen and strainer  intended to protect this header.
 A thick scale was also deposited on  the egg crate packing.  With
 regard  to the demister,  the Bechtel  test report stated "The PPA
 absorber demister was completely unsatisfactory for hydrated lime
 slurry  operation."  After a ten-day  test with limestone, it was
 found that  a hole had been eroded in the fiberglass wall from the
 force of the slurry spray.

 During  initial  limestone tests  with  the CVS there was an apprecia-
 ble  buildup of  solids (about one inch thick) in the duct from the
 venturi to  the  demister.  After a ten-day run, the thickness of
 these solids was between 2 and  3 inches.  A slight buildup of
 solids  occurred in the separator.  With the lime tests, only a
 relatively  thin scale was deposited  in the throat, but thick
 (1%  inch) scale was deposited on the vessel walls and the line
 to the  entrainment separator.   This  scale also plugged the pressure
 measurement lines, thus explaining some of the erratic BP measure-
 ments obtained.  The  major problem with this scrubber was its
 relatively  low  S02 removal efficiency.

 The  major difficulty  with the TCA scrubber concerned erosion of
 the  "ping pong" balls.  Initial measurements after about 200 hours
 of operation indicated a wear rate of about 0.5 percent per day.
 For  the ten-day test  with limestone, new balls were used and the
 loss in weight  over 10.8 days was 4.457o loss (0.246 grams/ball)
 or 0.43% per day loss.  The problem  with ball erosion, in addi-
 tion to the cost of ball replacement every three or four months,
 is that when the balls wear through, they fill with slurry and
 the  balls remain immobile.  Eventually, the slurry solidifies
 and  if  enough balls are immobile the pressure drop increases
 greatly.  This  spring, balls constructed of a new flexible soft
 material instead of the previous hard material were tested.
 However, in several days of testing  so many of these balls split
 in half that they were all replaced with the old style balls.
 There were  other difficulties with the TCA scrubber; spray nozzle
 erosion, demister plugging, and some buildup of solids in low
 velocity areas.

 As previously indicated, there were no difficulties with the SCE
 scrubber itself.  There were problems with pump seals, as there
 were with all of the  scrubbers, but  the design of this scrubber
 is such that one pump  can be taken out of service and repaired
without  degradation of the exit SC^  concentration.

As previously indicated, difficulty was experienced with I.D.
 fans due to uneven buildup of material on the rotors.  More
problems were experienced with  centrifuges than rotary filters.


                              382

-------
Reagent Cost

The annual cost of the reagents is a significant contributor to
operating costs.  Since each mole of S02 required to be removed
requires one mole of lime (CaO) or limestone (CaCOo) or soda ash
(Na^COo), the theoretical minimum requirements (100% purity of
reagents, 100% utilization) to remove 32 Ibs of sulfur or 64 Ibs
of S02 are 56 Ibs of CaO or 100 Ibs of CaC03 or 106 Ibs of Na2C03«
Based on supplier quotations for delivery to the Mohave Generating
Station, 90 miles south of Las Vegas, Nevada, and including a
delivery charge of $5/ton, the information presented in Figure 15
was prepared.  The important fact to note is that if limestone is
used in the "non-optimal utilization" system, lime costs on an
annual basis are lower.  On the other hand, if an "optimal utiliza-
tion" limestone system is used, maintenance costs could be higher
due to the higher solids concentration in the slurry system.
Furthermore, this system has a greater potential for plugging and
particulate carryover.


Reagent and Scrubber Selection for Full Scale Testing

It should be emphasized at this point that the evaluation made
by Southern California Edison of various scrubbers and reagents
applies only to the Mohave Generating Station.  Different condi-
tions at other generating stations might lead to conclusions dif-
ferent than the ones reached here.

The criteria on reagent cost and operation have been summarized
on a comparative basis in Figure 16.  The numerical ratings
range from 0 to 3, with 3 representing outstanding performance,
2 average, 1 poor and 0 assigned where the minimum required per-
formance was not demonstrated.

Figure 16 compares potential scrubbing reagents in terms of
performance, cost and control factors.  Based on the pilot plant
data, the best reagent system is soluble lime.  Except for the
potential plugging problems with certain types of scrubbers, this
reagent has exhibited overall good characteristics.  Without
regeneration, the cost and waste disposal problems associated
with soda ash are prohibitive.  Limestone, the other possible
reagent choice, suffers in regard to removal capability and
system control.  The addition of sodium and/or solids to lime-
stone salts improves performance but increases the waste disposal
problem.  Ammonia is not recommended at all due to the potential
presence of ammonia in the stack gas.

Based on the experimental data, operating problems and the SO2
and particulate removal needs at Mohave and Navajo Projects, the
                               383

-------
                                FIGURE 15
    %  UTILIZATION
                               REAGENT COSTS

                                          LIMESTONE +
                              LIMESTONE  (NA & SOLIDS)  LIME  SODA ASH
   66
 95
 98
99
    REAGENT REQUIRED
     (LB/LB OF SO2)
  1.56/1
1.56/1
1.88/1     1.66/1
u>
00
    REAGENT COST
     ($/TON TO MOHAVE )
  12.50
12.50
  21
50
    ANNUAL COST
$1,725,378      $1,198,684 $1,090,500 $4,896,160

-------
             Figure 16
COMPARISON OF SCRUBBING REAGENTS



CRITERIA
5O2 REMOVAL
UTILIZATION
U)
s EROSION & WEAR
MATERIAL COST
DISPOSAL
LIQUOR CONTROL
PLUGGING & SCALING
TOTAL


SODA
ASH
3
3

3
1
0
2
3
15


SOLUBLE
LIME
2.5
3

3
3
3
3
2
19.5


0.5%
LIME
2.5
2

2.5
3
3
2
1
16


LIME-
STONE
1
2

1
2
1
1
2
12

LIME-
STONE
+ SODIUM
2
2

1
2
1
1
2
11
LIME-
STONE
+ SODIUM
+ SOLIDS
2
3

0
3
1
1
1
11



AMMONIA
3
1

3
1
0
2
2
12

-------
 following conclusions were  reached:

      1.    It  is  essential that a  full  scale prototype  scrubbing
           system be  constructed and operated reliably  before
           selection  and  installation of equipment  to treat all
           of  the stack gas  at either Mohave or Navajo.

      2.    Construction of two different test module systems would
           give a higher  probability that a workable system for the
           entire plant could be developed in a reasonable time
           frame.

      3.    The scrubber/reagent combinations which  seemed to offer
           the greatest chance of  success were the  TCA  scrubber
           using  limestone and the SCE  scrubber using lime.

      4.    Flexibility should be included in the systems design
           including  provisions to incorporate packing  other than
           ping pong  balls in the  TCA scrubber and  the  ability
           to  test this scrubber with lime and the  SCE  scrubber
           with limestone.

Although we believe  the Mohave Generating Station  today is one
of the "cleanest" coal burning stations in the United  States,
with  S02 emission levels ranging  from  200 to 250 ppm and dust
loadings from 0.01 to 0.03  grains/scf, Clark County, Nevada
District Board of Health believed that an improvement was neces-
sary.  Clark County  adopted a regulation limiting  S0«  emissions
to 0.15 Ib S02/million Btu  after  the plant was in  operation.
In January of this year, the Clark County authorities  granted
variances  for Mohave which  require compliance with this S02 emis-
sion  limit.  This is intended to provide the time  required to
carry out  the required prototype module test program.  This program
includes installation and testing of two different scrubber module
prototypes each having a capacity equal to one-fifth the stack
gas flow for each of the two Mohave boilers.  Based on module
test  results and operating  experience, the type and specific
design for the full  scale scrubber systems would then be selected.

In accordance with this compliance plan and variance, we started
construction February 15, 1973, of two 450,000 scfm test modules
at Mohave.  The  time allowed for design and construction of these
modules is extremely short, particularly since we believe that a
450,000 scfm scrubber is larger than any operating today in the
United States.  We are required to start checkout of the Horizontal
Module by  November 1, 1973  and commence operation by December 1,
1973 with  checkout of the Vertical Module to start January 1,
1974  and operations  commencing March 1, 1974.  However, we are
proceeding with this "forced draft" schedule in order  to have
                              386

-------
time for an abbreviated test program before selecting the final
module design.  At the conclusion of our test period, we will
construct at least ten more modules of similar size at Mohave.

These test modules are based on the results of our pilot plant
tests but represent a considerable increase in size.  Normally,
good chemical engineering practice would be to extrapolate pilot
plant data by a factor of 10 when installing a new chemical pro-
cess.  However, with the Horizontal Module, which is based on the
SCE scrubber, we also have a new unconventional piece of equipment,
the scrubber itself.  Considering the difficulties encountered
with other large scale scrubbing systems, it required considerable
management fortitude to start construction of a new device 150
times larger than was tested on a pilot plant scale.  The relative
scale up is shown in Figure 17 which is an artist's rendering of
the full size scrubber compared to the initial pilot plant.

The Vertical Module is based on the TCA scrubber.  Construction of
a 450,000 scfm TCA scrubber represents a size factor increase of
450 over the 1,000 scfm pilot plant scrubber data.  While TCA
scrubbers have been constructed in larger sizes at other locations,
operating difficulties informally reported to us by other utilities
did nothing to allay our fears.  Therefore, provisions are being
incorporated in the design of this scrubber to allow us to replace
the balls and screens, if necessary, with the PPA type egg crate
packing or other packing material or possibly convert the device
to a simple spray tower.  An artist's rendering of the Vertical
Module is presented in Figure 18, compared to the initial pilot
plant size.  Unfortunately, the man in Figure 18 is smaller than
the man in Figure 17 so that a visual size comparison of the two
scrubbers is not available.

Some comparative information being incorporated in the test modules
design is presented in Figure 19.  It is perhaps of interest that
an 18 MW transformer is being installed to provide power for these
two 160 MW modules.  However, the 6000 hp F.D. fan motor for the
Vertical Module will obtain power directly from a 13.8 kV source.

Acknowledgement

As indicated previously, testing of the TCA, PPA, US and CVS
scrubbers was paid for by the Navajo Project Participants  (Salt
River Project - Operating Agent, Arizona Public Service Company,
City of Los Angeles Department of Water and Power, Nevada Power
Company, Tucson Gas and Electric Company  and the United States
Bureau of Reclamation) and the Mohave Project Participants
(Southern California Edison Company - Operating Agent, City of
Los Angles Deoartraent of Water and Power, Salt River Project,
aSd Nevada Power Smpany).  Southern California Edison served
as Program Manlger for these tests while Bechtel Corporation
served as the test contractor.


                               387

-------
                           SIZE COMPARISON BETWEEN PILOT
                        AND FULL-SCALE HORIZONTAL SCRUBBERS
oo
oo
     MliT FACILITY
FULL-SIZE HORIZONTAL TEST MODULE
                                                                               Figure 17

-------
CO
ID
                     SIZE COMPARISON BETWEEN THE VERTICAL
                TEST MODULE, PILOT-PLANT AND EXISTING FACILITIES
     VERTICAL      PILOT
     iODULE      FACILITY
    MOHAVE
GENERATING STATION

-------
u>
t£
o
SO2 REMOVAL WITH LIME SLURRY

SO2 REMOVAL W/LIMESTONE SLURRY

THICKENER TANK

SCRUBBER DIMENSIONS

GAS VELOCITY IN SCRUBBER AT 128° F

GAS REHEAT TEMPERATURE RISE

SYSTEM  PRESSURE DROP
     NOMINAL
     MAXIMUM

FAN HORSEPOWER
     NOMINAL
     MAXIMUM

SLURRY  PUMPS
     NOMINAL
     MAXIMUM

LIQUID FLOW RATE
     NOMINAL
     MAXIMUM

US RATIO IGPM/1000 SCFM)
     NOMINAL
     MAXIMUM

ELECTRIC POWER
     NOMINAL
     MAXIMUM
                                             FIGURE 19
                                       450.000 SCFM MODULES
                                     MOHAVE GENERATING STATION
                                          HORIZONTAL  MODULE
                                          400 PPM TO  40 PPM

                                          250 PPM TO  40 PPM

                                          60 FT. DIA. x 16 FT. HIGH

                                          15 FT x 30 FT. x 60 FT LONG

                                          21.6 FT/SEC.

                                          80° F
6 INCHES H20
10 INCHES HO
                                          1200 HP
                                          1750 HP
                                          8 - 300 HP
                                          12-300 HP
                                          12.000GPM FOR EACH OF 4 STAGES
                                          24,000 GPM FOR EACH OF 4 STAGES
                                          20
                                          40
                                          3.0MW
                                          5.0MW
                                 VERTICAL MODULE
                                 250 PPM TO 40 PPM

                                 400 PPM TO 40 PPM

                                 50 FT DIA. BY 8 FT. HIGH

                                 18 FT x  40 FT. x 90 FT HIGH

                                 12.6 FT/SEC.

                                 80° F
31  INCHES H2O
40 INCHES H2O
                                 4850 HP
                                 6000 HP
                                 4 - 500 HP
                                 6 - 500 HP
                                 37,350 GPM
                                 37.350 GPM
                                 83
                                 83
                                 8.8 MW
                                 10 MW

-------
Testing of the AIS scrubber was paid for by the following members
of WEST Associates:  Southern California Edison Company,  Arizona
Public Service Company, City of Los Angeles Department of Water
and Power, City of Colorado Springs Department of Public  Utilities,
Colorado-Ute Electric Association, Inc., El Paso Electric Company,
Idaho Power Company, Montana Power Company, Nevada Power  Company,
Pacific Power and Light Company, Public Service Company of Colorado,
Public Service Company of New Mexico, Salt River Agricultural
Improvement and Power District, San Diego Gas and Electric Company,
Tucson Gas and Electric Company, and Utah Power and Light Company.
Southern California Edison Company served as Program Manager for
these tests while the Atomics International Division of North
American Rockwell served as the test contractor.

Funding for the RHS scrubber tests was primarily provided by the
National Lime Association although Southern California Edison
Company provided some financial support.  Both organizations pro-
vided test personnel.

With regard to the WPS, EPA provided advice on instrumentation
and test procedures, but the funding was provided solely  by
Southern California Edison.  The test program was conducted by
Southern California Edison with assistance from Truesdail Labora-
tories.

Funding for the SCE scrubber test program described here  was pro-
vided solely by Southern California Edison Company.  The  test
program was conducted by Southern California Edison and Truesdail
Laboratories.

Funding for the two 450,000 scfm test modules is being provided
by the Navajo Project Participants (represented by Mr. Tom Morong,
Chief Engineer and Assistant General Manager of the Salt River
Project) and the Mohave Project Participants (represented by Mr.
Jack B. Moore, Vice President-Advanced Engineering of the Southern
California Edison Company).  Program Management of this Test
Modules Program is the responsibility of Southern California
Edison Company.  Stearns-Roger, Inc. is responsible for the
design and procurement of equipment for the Horizontal Module,
with the Bechtel Power Corporation serving as the contractor at
the site.  The Bechtel Power Corporation is also responsible for
the overall design and installation of the Vertical Module with
Universal Oil Products providing the scrubber and process design
under subcontract to Bechtel Power Corporation,

Finally, we feel somewhat embarassed about taking credit for the
work of so many individuals and would like to publicly acknowledge
that the individuals listed below made many identifiable contri-
butions to the experiments reported here.

     Atomics International Division of North American Rockwell
          Dr. Dennis Gehri
          Mr. Donald Gylfe

                               391

-------
Bechtel Power Corporation
     Mr. Paul"Grimm
     Mr. Robert Keen
     Dr. Wen Kuo
     Mr. Fred Miller
     Mr. Angelo Sassi
     Dr. J. L. Shapiro

National Lime Association
     Mr. Clifford Lewis

Radian Corporation
     Dr0 Phil Lowell
     Dr. Delbert Otmers

Southern California Edison Company
     Mr. S. T. Carlisle
     Mr. E. J. Fletcher
     Mr. John M. Johnson
     Dr. Dale G. Jones
     Dr. E. A. Manker
     Mr. W. Carl Martin
     Mr. Richard B. Rolfe

Steams-Roger, Inc.
     Dr. Robert M. Christiansen
     Mr. Keith S. Campbell
     Mr, John D. Ferrell
     Mr. Dave Naulty
     Dr. J. Louis York
Truesdail Laboratories
     Mr. Harold Decker
     Mr. Harold A. Kerry
     Dr. Olgart Klejnot
     Dr. Marty Prieto
     Mr. Eli San Jose

U. S. Lime Division of Flintkote Corporation
     Mr. Dan Walker
                           392

-------
 A REVIEW OF BABCOCK & WILCOX
AIR POLLUTION  CONTROL SYSTEMS
      FOR UTILITY BOILERS
               by

          J. F. Stewart
 Fossil Power Generation Division
     Power Generation Group
         Barberton, Ohio
                393

-------
 A REVIEW  OF BABCOCK &  WILCOX AIR  POLLUTION CONTROL
 SYSTEMS FOR UTILITY BOILERS	

 J. F. Stewart, Fossil Power Generation Division
 Power Generation Group, Barberton, Ohio
 Presented to
 Environmental Protection Agency
 Flue Gas Desulfurization Symposium
 New Orleans, Louisiana
 May 14-17,1973
 INTRODUCTION

 The Federal Air Quality Act of 1967 was set up in
 an attempt to deal with air pollution problems in
 this country on a regional basis. Over 90 air quality
 control regions were originally designated to be the
 basis for regional administration and enforcement
 of the Act. There presently exist a total of 247 air
 quality control regions through out the nation. The
 Act also  called  for the Department of  Health,
 Education, and  Welfare (HEW) to  develop and
 publish air pollution  criteria  that  indicate the
 extent to which pollutants are detrimental to the
 health and property of people and how emissions
 can be limited and controlled.

 In an effort to speed up progress  toward cleaner
 air, Congress passed the Clean Air Amendments of
 1970 that set  up  the  Environmental Protection
 Agency (EPA), directed the EPA  to set national
 primary  and secondary ground level air quality
 standards,  and  set   a   timetable  for   their
 implementation. The initial set of standards was
 issued April 30,  1971, and covered six common
 pollutants, including  particulate  matter,  sulfur
 dioxide and nitrogen oxides. Table I lists ground
 level standards that would  primarily affect utility
 boilers. Primary  standards are intended to protect
 the  public health  with an adequate margin  of
 safety, and the secondary standards are intended to
 protect the public  welfare,  and  consider such
 factors as soiling, corrosion and vegetation damage.

 Each state was responsible, by January 31, 1972,
 to develop and submit an implementation plan for
approval by the EPA. The state plans are required
to  include emission control  limits designed  to
achieve and maintain these national ambient air
quality standards. The States have until July, 1975
 to achieve air quality equal to or better than the
 national primary standards. Secondary  standards
 are to be met 27 months later, depending on the
 availability of adequate control technology, land
 use and transportation control.

 The ammended  Clean  Air Act of 1970 also has
 resulted in the establishment of emission standards
 for new stationary sources. Limits for particulate
 matter, sulfur dioxide  (SO2), and nitrogen oxides
 (NOx) related to fossil fuel-fired steam generators
 of more than 250 million BTU/hour fuel  imput are
 shown on Table II.

 This  paper  will  deal with   the  research and
 development   programs,    and   demonstration
 scrubber systems that Babcock & Wilcox has been
 developing during recent years in  an attempt to
 provide the electric utilities with solutions to their
 present and future air pollution control problems.

 PARTICULATE  MATTER

 Particulate emissions from  most  coal-fired and
 some oil-fired  utility boilers have been controlled
 for  many  years  with   mechanical  collectors,
 electrostatic precipitators  or  a combination  of
 both. In order to meet the stringent particulate
 collection efficiency required by the EPA for new
 sources, the engineer today is faced with a difficult
 problem.  Further,   many  states  are  adopting
 particulate emission limits for existing units which
 are more stringent that the EPA standards for new
sources. For example, the state of New Mexico has
recently passed an Air Quality regulation that will
limit  particulate  stack emission  levels  to  .05
Ib/MKB input by 1975.
                                               394

-------
Engineers today can choose from four basic types
of  collection devices when selecting participate
removal equipment. These devices include:
    1.  Cyclone-type   mechanical  collectors and
       classifying hoppers.
    2.  Electrostatic precipitators.
    3.  Wet impingement-type scrubbers.
    4.  Bag filter houses.

    1.  Mechanical Collectors

       Mechanical   collectors   and   classifying
       hopper devices are typically low collection
       efficiency  components  when  handling
       flyash from a  coal-fired boiler. Collection
       efficiencies  for these  devices will  range
       from  20 to 75%, considerably below the
       requirements  being  set by the EPA and
       state authorities.

   2.  Electrostatic Precipitators

       The electrostatic precipitator  has  become
       the principal gas cleaning device for boilers
       where fine particles cannot be collected in
       a mechanical device. Properly designed and
       arranged, precipitators are able to perform
       at high collection efficiencies  over a wide
       range of particle sizes. However, as these
       systems age, electrodes corrode and break,
       hoppers   bridge,   deposits    form   on
       insulators,  and more  frequent flashover
       occurs with  a resulting  increase  in  the
       emission rate. To compensate for  these
       malfunctions,  and they most certainly do
       occur, the designer  must allow for some
       redundancy  and   conservatism   in   the
       precipitator design so that design efficiency
       can be maintained without reducing boiler
       load  or  incurring  an   outage.   It  is
       conceivable that a precipitator  designed for
       92  to   96%   collection  efficiency   can
       continue to perform near design efficiency
       with some degree of malfunction; however,
       the margin  of safety  becomes very slim
       when  a unit is designed  for  greater than
       99% collection efficiency. These  are  the
       minimal  efficiencies that are  going  to be
       required  in order to meet the  0.1 Ib/MKB
       set  forth by the EPA. A typical  10,000
       Btu/lb fuel with 20% ash will require the
       precipitator to  perform  at  an  average
       efficiency of 99.6% in  order to meet this
       requirement.   It  would appear  that  the
prudent designer may have to provide spare
precipitator  sections  which   could  be
isolated  from  the operating precipitator
modules with the boiler in service to permit
routine maintenance, hopper cleaning, etc.
if  the required emission level is to be met
on a day in day out basis. One alternate to
providing a modular precipitator  design
would  be  for the   engineer  to  select
conservative precipitator designs  that  are
less subject to discharge electrode failure.

Precipitator  efficiency  is  controlled  by
many factors which include dust size and
loading,  gas  temperature, sulfur  oxides
concentrations,  moisture  content,   ash
chemical composition, treatment time and
gas velocity. The shift of many power plant
operators to low  sulfur western fuels is
currently  having a marked affect on  the
operation of  their  existing precipitators,
which in many cases, were designed  for
high  sulfur  eastern fuels. The  immediate
result can be a marked increase in stack
emission due  to many  factors  not all of
which are  entirely  related to  the  sulfur
content of  the fuel. In  some cases,  the
higher moisture content of the western fuel
can  have a compensating effect  on  the
reduction of the  sulfur of the  fuel with
respect to precipitator performance.

Many   operators   are   redesigning   old
precipitators, adding on new sections, or
installing completely new precipitators to
perform on  low sulfur fuels based on little
operating experience or  resistivity data
applicable to their specific gas conditions.
Also, little is known  today to what  extent
coal  composition  can  be varied without
affecting precipitator efficiency. It  would
appear that considerable effort  is required
to establish  what effect parameters such as
coal and ash composition chemistry, mode
of burning, flue gas moisture, SO2  and SO3
concentration,  temperature, and velocity
have on precipitator performance. This will
be especially important in designing high
efficiency precipitators for low sulfur fuels
that  are expected  to  have  a wide variation
in  coal composition over the plant life. It is
expected that these plants will be supplying
a large portion of our future fossil energy
requirements.
                                               395

-------
3.  Wet Impingement-Type Scrubbers
   Wet impingement-type scrubbers have been
   solving a multitude of varied problems for
   the  chemical,  pulp and paper  and steel
   industries for many years. Within the past
   four years, a number of utilities have begun
   to  install  prototype  and  demonstration
   scrubbers  for  the purpose of  removing
   particulate and/or SO2 from boiler flue gas.
   Particulate removal in these scrubbers is
   accomplished in a  number of ways.  Some
   designs rely on quenching the flue gases to
   the adiabatic saturation temperature with
   wetting  and/or   agglomeration  of   the
   particles  in  a low velocity duct.  Their
   removal    from    the   gas   stream   is
   accomplished by  gravitational forces and
   en trainmen t separators.

   Other devices impact  the  quenched gases
   on a wetted packing, such as marbles, balls,
   or bubble caps, and remove  particulate by
   the process of inertia! impaction. Particles
   are removed from  the gas  stream in this
   process because the particles are  unable to
   follow the gas  stream around the packing,
   resulting in the particle impacting against
   the  packing. Collection efficiencies  for
   devices of this type are dependent on the
   particle   size distribution   of   the  dust
   entering the scrubber. Most scrubbers have
   good   collection   efficiencies   on   large
   particles greater than one micron; however,
   the  collection  efficiency  for submicron
   particles  can decrease  rapidly unless the
   particles  are  accelerated  sufficiently   to
   cause impaction  on the packing surface.
   The performance  of wetted packing can be
   affected  significantly unless the gas and
   liquor distribution  remains  uniform over
   the bed surface.

   Removal  of  submicron particles  can be
   accomplished  in  a wet scrubber  if the
   particles  are accelerated sufficiently and
   then permitted  to collide  with or  on a
   droplet surface. This can be accomplished
   in  a  high  energy  venturi  scrubber. The
   smaller  the  size  of the  particle  to  be
   removed,  the  higher  the   velocity  and
   energy required. Most of the energy losses
   in  a venturi result from accelerating the
   scrubbing liquid. In a venturi scrubber, the
 probability of a particle  colliding  with a
 water  droplet  is   greatly  increased   by
 maximizing the number of water droplets
 in   the   throat   area.   This   can   be
 accomplished to a degree by first atomizing
 the   liquid;  however,  more  complete
 atomization  of  the liquid droplet  can be
 produced by the shearing action of the gas
 stream. The accelerated particles impact on
 the fine liquid droplets which subsequently
 collide  with each  other and agglomerate.
 The gas stream is then  decelerated  and the
 water droplets with their captured particles
 are removed from the gas stream by gravity
 or inertia! separation.

 Venturi scrubbers have been used for many
 years to scrub fine fumes such as the  salt
 cake  generated  in Kraft  recovery  boilers
 where 40 to 50% of the particles  are  less
 than  one  micron  in size. More recently,
 high  energy venturi scrubbers have been
 employed  to scrub  the iron  oxide fume
 emitted from  Open  Hearth  and  Basic
 Oxygen  Furnaces.  In this  application,
 where 90% of the particles  are less than  one
 micron, the energy requirements amount to
 a 50  to 60 in. wg gas side pressure loss to
 obtain virtually a clear stack.

 What energy requirements are required in
 the  case  of  boiler  fly-ash?  It  is  first
 necessary  to  define  the  particle   size
 distribution of the fly-ash to be collected.
 The classical method for obtaining  particle
 size  distribution for fly-ash  has been to
 obtain a fly-ash  sample according to ASME
 PTC-27 and determine  specific gravity  and
 particle size distribution by Banco Analyses
 (ASME PTC-28). Figure 1 is a plot showing
 typical  fly-ash particle size distribution for
 large  pulverized coal-fired  or cyclone-fired
 boilers. Banco data is usually not reported
 below a particle diameter of  2 microns
 because   the   smallest   size   fraction
determined  with this analysis is 1.7 to 2
microns. It can readily be seen from Figure
 1  that  a  significant  percentage  of  the
fly-ash  (6  to 7%)  exists  in  the fraction
below two microns, which  may or may  not
follow the same distribution slope as  the
larger material fraction. The fraction below
                                           396

-------
2 microns is the most difficult to collect,
regardless of the type of collection device
employed.

A  venturi scrubber functions much like a
sieve; that is, it has a cut-off point, which is
a  function  of  gas  velocity  and  recycle
liquor   rate.   These   two   parameters
determine the  venturi pressure loss which
for a given collection efficiency must be
made to vary  inversely with  dust particle
size.

The  theoretical  venturi  grade  efficiency
curves shown in Figure 2 illustrate how the
cut-off point for  a venturi can be shifted
with  energy level.  These  are  calculated
curves and assume ideal conditions that do
not occur  in an  actual operating  venturi.
The  collection efficiencies  indicated  are
significantly greater  than those expected
for an operating unit and can be effected
by venturi  design. These  curves illustrate
how  the particle size  cut-off point  in a
venturi  can be  shifted  with changes  in
venturi energy level.

It  therefore   is  quite  apparent  that
submicron  particle size distribution for a
fly-ash  must  be  determined before  the
required venturi energy levels for a given
collection   efficiency  can  be calculated.
Submicron  fractions  have in the past been
examined   using  light  scattering  devices,
ultracentrifuge  techniques,   transmission
electron  microscope,  and  the  scanning
electron   microscope.   These  methods,
however,    suffer   from    the   same
shortcoming,  i.e.,  not  reproducing  the
particle  size distribution  as it exists in the
gas  stream. It  was expected  that  dust
loading  and particle size  distribution  can
vary considerably, according to application;
therefore,   some  tool was  needed  that
would measure the actual dust loading and
particle size distribution as exist in the gas
stream of an intended installation. Such a
technique  would eliminate the need for
expensive and time  consuming  field  pilot
plant tests, which  only  indirectly give
particle size distribution and are subject to
error in extrapolation to  commercial sizes.
To  perform  these  measurements, B&W
utilizes  a  commercial  cascade  impactor
modified to include a cyclone separator in
series with from  one  to  seven impactor
stages.1  These components are assembled
in a probe that can be inserted in a duct for
dust sampling. An iso-kinetic gas sample is.
drawn through the sample  probe, which*
acts like seven Venturis in series.  Figure 3
shows the sampling train employed.

Each impactor stage  has an  orifice  and
collection  cup as illustrated  in  Figure 4.
The  orifice  diameter and  the  distance
between the orifice and cup determine the
particulate collection characteristics of the
stage.    Some   typical   particle    size
distributions for  fly-ash determined  with
this device are shown in Figure  5 along
with  a  typical Bahco analysis for fly-ash.
Using  theoretical  analysis  coupled  with
experimental  results,  the  Research   and
Development Center of Babcock & Wilcox
at  Alliance has developed a mathematical
model   to   predict   venturi   scrubber
performance. With this model,1  we are able
to  construct theoretical grade  efficiency
curves,  shown typically in  Figure 2 for
various venturi design configurations, and
apply   particle   size   distribution   data
determined with  the  cascade impactor to
predict theoretical venturi performance for
different energy levels.

Wet scrubbers may not necessarily be the
final answer  to  every  dust  collection
problem especially in water-scarce areas or
where    visible    vapor    plumes    are
objectionable. Large quantities of water are
evaporated in cooling the flue gas stream to
its  adiabatic saturation temperature.  This
water quantity can be as much as 900 gpm
for an 850 MW scrubber unit. Another loss
is the dilution water required to remove the
ash as a slurry from  the system. Some of
the dilution water  can be recovered  with
suitable   thickening   and    dewatering
equipment;   however,  some  degree  of
blowdown will  be  required  for   these
systems due to dissolved  solids  buildup.
The extent of the blowdown will depend
on  the  chemistry of the fly-ash  and the
make-up water supply to the scrubber.
                                          397

-------
 Bag Filter Houses
 In   applications  where  extremely  low
 discharge emissions are required, the energy
 requirements   for  a  wet  scrubber may
 become excessive when compared to that
 required to obtain the  same performance
 with a bag filter house.

 Highly efficient filter houses are commonly
 used to  collect dusts from cement plants,
 fertilizer plants, metallurgical furnaces, and
 other  applications  where collection  of
 submicron  material   is  required.  The
 successful application of a full-scale filter
 bag house to a 320 MW oil-fired boiler  for
 the  control of visible stack emissions was
 demonstrated  in 1968.3  A number of pilot
 filter houses have been  installed in recent
 years on pulverized  coal-fired  boilers  to
 determine pressure drop requirements,  air
 to cloth ratios, and operating performance
 as  well  as determining  the economics of
 applying these systems to coal-fired units.
 The performance of a large full-scale filter
 house  applied  to a coal-fired  unit has not
 yet  been  demonstrated.  Application  of
 filter  houses   to  coal-fired  units in the
 future  will probably  be  warranted  for
 special  situations where  extremely high
 collection  efficiencies  are required and
 after  taking   into   consideration  site
 location, fuel source and water availability.

 When the engineer today starts to  consider
 the   type   of  particulate   collection
 equipment that will best fit his present and
 future needs, there will be factors  which in
 the  past did  not enter into  his  decision
 making. Preference may be given  by  some
 to  wet  collection  systems  since  many
 designs  may be augmented in the future
 with  various   basic  materials,   such  as
 limestone for   removal  of sulfur  dioxide.
Others  may  have  a  ready  market for
dry-collected  fly-ash  and  will   find   a
 combination  precipitator  for removal  of
 the bulk of the ash and wet scrubbers for
removal  of  submicron  dusts  and/or SO2
removal the best overall solution. Some will
find that application of a wet scrubber to
their particular  fuel will result  in serious
scaling  of the slurry  lines  due   to the
chemistry  of   the  ash.  When  more
        experience  is gained, it may be possible to
        predict  what ashes are  likely to  cause
        scaling problems in scrubbers,

 SULFUR DIOXIDE REMOVAL SYSTEMS

 No processes have received more attention by the
 different engineering disciplines  in  recent  years
 than  those that are being developed for removal of
 sulfur dioxide from boiler  flue gas. In the mid to
 late  sixties,  the  primary  incentive  for  their
 development  was  the  attractive  price for sulfur
 which peaked out at $40 per long ton in 1968. The
 price for sulfur in recent years has steadily dropped
 to  levels that have forced the closing of a number
 of  Frasch process mines in the Gulf Coast region.
 Recent prices for Canadian sulfur, most of which
 comes as a by-product of natural gas production,
 have  been quoted as low as $9.50 per long ton
 delivered to  some midwest markets.  There will
 always be a price available for recovered sulfur and
 sulfur products, but it would appear that this price
 will be adversely influenced in the future as more
 crude oil stocks are  desulverized  and as  sulfur is
 recovered from flue gas sources.

 NON  RECOVERY SYSTEMS

 The  development  of flue  gas   desulfurization
 systems  by B&W  has been  concentrated  on both
 recovery and  non-recovery systems. A  number of
 bases  can be  employed in  a scrubber system  for
 removal  of  SO:  as a waste  product. These  bases
 include  lime,  high  calcium  limestone,  sodium
 carbonate,  sodium  hydroxide and ammonia.  A
 development program was initiated by B&W five
 years  ago to determine the performance for various
 basic   materials  in  a number  of wet scrubber
 devices. Out of these investigations emerged the
 B&W  limestone wet scrubbing  system.  Bases such
 as sodium carbonate, sodium hydroxide, ammonia
 and lime were found to give better SO2  removal
 performance than  limestone,  but these  systems
 have high raw material costs and many of the bases
 result in  sulfur products that have a high solubility
 in water and would be difficult and expensive to
 dispose of.

The chemical  costs of lime could be reduced  by
injecting  pulverized  limestone into  the boiler
furnace  to  accomplish  its  decomposition  to
calcium oxide. There are  several disadvantages to
                                        398

-------
 this method of operation:
    1.  Injected pulverized stone increases the dust
        loading and duty on the scrubber system.
    2.  Universal  application   of   the  injection
        system was not deemed possible due  to
        possible   pluggage   in   reheater   and
        economizer    sections    and    slagging
        conditions that could occur due  to boiler
        design or the type of fuel utilized.
    3.  An injection system could not be applied
        to oil-fired units due to the tendency for
        limestone to deposit on the furnace walls
        or convection  surfaces resulting in serious
        changes  to  the furnace, superheater, and
        reheater heat absorption.
    4.  The problems  associated with circulating a
        lime slurry are well documented in calcium
        base pulping. The problem  of maintaining
        system  chemistry   to  prevent scaling  of
        piping and  other hardware in a lime system
        was  considered  more  critical  than  the
        control required for the limestone system.

 Concurrent with determining that limestone was an
 acceptable  base  for an  SOj  removal system, a
 program   was   initiated  for   evaluating  and
 determining the scrubber that had the potential for
 at least 80% sulfur dioxide removal.

 This  degree  of  SO2  removal could   not  be
 accomplished in a  single  or double stage venturi
 with limestone.  A  counter current  tray absorber
 did look promising,  however, when  considering
 both   performance  and  probable  operating
 problems.

 Again, the  engineers at  B&W's  Research and
 Development  Center  at  Alliance   attacked  the
 absorber  problem  with  a fundamental approach.
 They  felt it would be extremely  dangerous  to
. scale-up pilot plant test results to a 125 MW size
 absorber  unless  the absorption mechanism  with
 limestone was understood. A mathematical model4
 was  developed  to determine  what  effect  the
 significant variables have on scrubber performance.
 The model was  later confirmed in  the laboratory
 pilot plant shown in Figure 6. This model considers
 the normal  operating variables such  as flue gas
 flow,  recycle liquor  rates, slurry  concentration,
 reaction  rates and diffusion   constants  for  the
 chemical   species   involved.   In   addition,  the
 comparative reactivity of various  limestones was
 determined so the  prediction  of SOa absorption
 could be adjusted accordingly.
Other  factors  that influenced  the  decision  to
proceed   on  development  of   limestone  wet
scrubbing as a  first  generation  system for SO3
removal were the low cost and high abundance of
high calcium  limestones in  most areas  of the
United  States. The reaction  products from this
system, calcium sulfite and calcium sulfate, have
low  water solubilities which reduce the potential
for this system to create a water pollution problem
from disposal of spent react ants.

In  order to  evaluate the  suitability of  various
limestones for use in the limestone system, several
methods for measuring  limestone reactivity were
developed. One method involves chemical titration
of  a slurry sample prepared  from a pulverized
sample of the limestone. The quantity of titrant is
plotted as a function of time, while simultaneously
taking into account the  change occurring in stone
fineness   during  the   titration.  This  result  is
compared with the titration rate for the standard
stone sample utilized  for the pilot plant and model
test work. This test is used primarily for screening
purposes to determine those materials that should
be further screened in the small pilot plant.

This method  of laboratory pilot plant testing of
stones for use in limestone systems provides  for
excellent  control  over  all  test  conditions.  In
addition, the testing methods are  not subject to the
many  uncontrolled  variables that  occur  when
conducting tests with costly field pilot plants.

Confirming tests  of  limestone performance have
been run under closed cycle conditions which are
very close to those the stone will experience at the
final installation. Closed  cycle  testing has been
conducted in a larger laboratory pilot plant that
includes a  furnace that can  burn 500 Ib/hr of
pulverized coal, a steam generating bank to cool
the   combusion  gases,  a  particulate  venturi
scrubber, a B&W counter  current tray absorber,
steam coil reheater and ID fan.

The  slurry  portion   of the  system  includes  a
limestone preparation and recirculation system,  a
thickener and a vacuum belt filter. It is possible to
operate  this  system  with maximum  recovery of
water to determine the effects of dissolved solids
buildup on scaling and SO} absorption for various
limestones   and   fuels.  This   pilot  plant  is
instrumented with controls,  sampling equipment
and  is  capable  of  continuous round-the-clock
operation.  Figure  7  is a schematic showing  the
closed  cycle  pilot plant which  is located at  our
 Research Center.
                                                 399

-------
 Waste Disposal

 Spent slurry and fly-ash disposal requirements for a
 coal-fired boiler  with a limestone wet scrubbing
 system will be about double that normally handled
 for the boiler alone. This will be a severe burden at
 many locations and could require this material be
 dewatered and hauled  away for  disposal. All the
 problems associated with sludge disposal are not
 fully known. The possibility of utilizing the waste
 stream  form this process as a useful or valuable
 product  is  considered  highly  remote  in  this
 country.

 B&W's  efforts at waste disposal with this process
 have been directed toward conversion of the waste
 stream  to a form that will facilitate its disposal,
 minimize  its effects  on  the environment  and
 reduce, for the customer, the  quantity of sludge
 for disposal. Research efforts currently in progress
 have not progressed  sufficiently  to  permit  a
 meaningful report at this time.

 Demonstration Systems

 The first full scale demonstration  of the B&W
 limestone wet scrubbing system is located at the
 Will County  Station of  Commonwealth Company.
 This system  is retrofitted on a 163 MW net B&W
 radiant cyclone - fired boiler that went in service in
 1955. Two venturi • absorber scrubber modules
 treat the entire  flue gas flow leaving the unit,
 770,000 ACFM.

 The  expected   performance  for  this  system
 corresponds to an outlet dust emission of 0.0124
 grains per DSCF and  a  sulfur dioxide removal
 efficiency 81.6 percent. The system  and its major
 components are shown schematically in figure 8.

 Each scrubber module consists of a variable throat
 area  venturi  that  removes  the  particulate  and
 provides an initial stage of sulfur dioxide removal,
 followed by  a B&W countercurrent tray absorber.
 Venturi  and  absorber  sprays both  drain  into
 separate recirculation tanks that provide for delay
 time  to  complete chemical  reactions prior  to
 recirculating the slurry back to  the module. Three
 100 percent  capacity venturi recirculation pumps,
 each  rated  at 7300 gpm,  and  four-60  percent
 capacity absorber recirculation  pumps, each rated
 5250 gpm were provided for the two modules.

 Flue gases leaving each absorber pass through a
bare tube steam coil reheater and then  to a 2250
hp ID booster fan. Both booster fans discharge into
the existing ID fan inlets.
 The  major  control  functions for  the system,
 limestone  feed rate,  venturi  spray  liquor rate,
 venturi  AP,  slurry  solids  concentration  and
 limestone  milling  system  operation are  all
 controlled automatically from the scrubber control
 panel located in the existing boiler control  room,
 Figure  9 shows the instrumentation and control
 diagram for the scrubber portion of the system at
 Will County.

 Limestone is  received  by river barge  and stacked
 with the existing coal handling equipment at Will
 County. Stone is transferred by conveyor to two
 storage  silos.  Each   silo  discharges on  to  a
 gravemitric feeder that supplies one  of two 100
 percent  capacity wet  ball mill and classification
 systems. Each mill has a capacity to grind 12 tons
 per hour of limestone to a fineness of 95 percent
 minus  325  mesh. Twenty  percent  solids feed
 limestone slurry leaving the milling system is  stored
 in a slurry storage tank from where it is transferred
 to the scrubber modules.

 Spent slurry from the scrubber is pumped to a 65
 foot diameter thickener. Clarified recycle  water
 discharges  to  a 5.5  acre pond and returned  to the
 cycle with the reclaim pumping system. Thickened
 slurry  underflow is pumped to a loading station
 where fly-ash  and other  dry  additives will be
 blended  with  the thickener underflow, to modify
 the sludge  sufficiently to produce a stable land fill
 material.

 Detail  engineering  for the Will County project
 commenced in  September,  1970.  One  scrubber
 module was placed in service on February  23,1972
 with the second module going in service on April 7,
 1972.
A second demonstration limestone  wet scrubbing
system is being  supplied for a new  electric power
generating station located near La Cygne, Kansas.
The 820 MW net station is a joint project of Kansas
City Power &  Light Company and Kansas Gas and
Electric Company. The 6,500,000 PPH Universal
Pressure  Cyclone boiler will be fueled by  a low
grade  bituminous  coal  obtained  from  nearby
surface mines.

 The air  quality system, shown schematically  in
 Figure 10,   consists  of  seven venturi-absorber
 scrubber modules designed to handle the  entire
 flue gas flow  of 2,370,000 ACFM. The system is
 designed for  98.75%  particulate removal   which
 corresponds to  an outlet particulate  emission of
 0.10 lb. per million BTU fuel input. Sulfur dioxide
 removal  for   the  system  is  designed   for  80%
 efficiency.
                                              400

-------
 Each scrubber module consists of a variable throat
 area venturi followed by  a  B&W  countercurrent
 tray absorber. Each module  has a  100% capacity
 venturi recirculation pump,  each  rated at 7750
 gpm and  supplied with 350  hp drives. A total of
 seven,  100%  capacity,  10,300  gpm  absorber
 recirculation pumps, were supplied, each with 400
 hp drives.

 Flue  gases  leaving the seven  modules, are  first
 reheated 25 F with a bare-tube steam coil reheater
 and discharged by six induced draft fans, 7000 hp
 each, through a 700 ft. stack.

 Limestone  consumed  by the  scrubber  system
 comes from local quarries and is delivered by truck
 and   conveyed  with  the plants  coal  handling
 equipment to two limestone storage silos. Two full
 size  110-ton per  hour wet  ball mills grind the
 limestone to a fineness of 95% minus 200 mesh.
 Pulverized limestone slurry is then stored  in  two
 200,000 gallon capacity slurry storage tanks.

 Spent   slurry,   containing    fly   ash    and
 calcium-sulfur-reaction products, is pumped to a
 160 acre settling pond, with clarified recycle water
 returned to the system with two-100% capacity
 pond return pumps.

 This generating unit is currently in  the start phase
 with  a scheduled commercial  operating date of
 May 1,1973.

 SULFUR  RECOVERY SYSTEMS

 It is too early to determine the total magnitude of
 sludge  disposal  problems  associated with  the
 nonrecovery   sulfur   removal   systems.   Some
 operators   may  find  no economic  means  for
 disposal of waste products and will direct  their
 attention  to processes that minimize this problem.
 One sulfur recovery process that can be applied as
 a retrofit  to existing units is a  wet  MgO  system
 B&W has been developing for  the past six years.

 B&W MgO System

 Scrubbing flue  gas with MgO and  recovering the
 sulfur  values is  not a new system. For over  20
years, many calcium sulfite pulping processes have
been  converted  to an advanced pulping-recovery
process developed  and patented jointly by Howard
Smith Paper Mills Ltd., Weyerhaeuser Company,
and Babcock &  Wilcox. Over twenty installations
of this type both in the United States and abroad
have been installed at both new and existing pulp
mills  that utilize  this process to  recover sulfur
dioxide from the flue gas leaving chemical recovery
boilers.  Figure 11  shows a flow schematic for a
typical Magnesium Bi-Sulfite Pulping and Recovery
process.   The  scrubbing   liquor,  a  mixture  -of
magnesium  sulfite  and bisulfite, is utilized as* a
cooking  liquor during the pulping process. Weak
liquor from the digester is concentrated to 50 to
55%  solids in a multiple effect evaporator  and
burned  in  a  B&W  recovery furnace. Dissolved
lignins from the pulping process supply  the fuel
that maintains the combustion process. Magnesium
sulfites and sulfates are thermally decomposed to
sulfur dioxide and magnesium oxide. Magnesium
oxide  is removed  from the flue gas stream with
mechanical  collectors, washed to  remove soluble
impurities, and slaked to magnesium hydroxide in
hot water.  The slaked magnesium  hydroxide is
then   added  to  the sulfite-bisulfite  scrubbing
solution  to remove the sulfur values from the flue
gas stream in a venturi  or contact tower before
exhausted the gases to the atmosphere.

The application of  this  process to  utility boiler
stack gas application results in a number of process
changes.  Most of these are associated with the
regeneration  portion of  the  cycle  due  to  the
thermal    requirements    for     drying    and
decomposition which must be supplied from fossil
fuels, since the recovered magnesium salts have no
heating  values.  A  schematic of  this  process is
shown in Figure  12.

The first step  in the process involves quenching of
the hot flue gases and removal of particulate from
the flue gas stream. Particulate removal can also be
accomplished  in a  high  efficiency  electrostatic
precipitator.    Sulfur   dioxide    removal   is
accomplished  in an  absorber of  the same design
utilized for limestone scrubbing. Magnesium sulfite
hexahydrate slurry is removed as a blowdown from
the  process,  concentrated  and   passed  to  a
dewatering step  where partial dehydration of the
crystals   takes  place.   Surface   moisture  and
additional waters of  hydration are removed in a
drier. The dried crystals are decomposed in an  oil
or gas-fired furnace that discharges a sulfur-rich gas
suitable as a feed to a contact acid plant or to an
elemental sulfur conversion plant, depending  on
the desired product. Magnesium  oxide  formed
during the decomposition  step is  recovered with
mechanical collectors, slaked and recycled to  the
scrubber.
                                              401

-------
 B&W-Esso Dry Sorbent Flue Gas Desulfurization Process
SUMMARY
One  of  the  more  promising  dry  flue  gas
desulfurization  processes  is  the  system  being
developed jointly by B&W and Esso Research and
Engineering Company  (ERE) with the support of
17 electric utility  companies in the United States
and Canada, Dry sorbent systems are attractive for
many reasons. They avoid plume problems created
by wet scrubbing and do not have some of  the
water  disposal  problems  associated  with wet
scrubbers.  Dry  systems  which  operate in  the
temperature range  of  600 to 700  P places  the
sulfur dioxide removal  equipment ahead of the air
heater in a  boiler cycle. This should result  in
reduced maintenance and  improved performance
for the air heater.

In  August,  1967,  B&W  and  Esso Research and
Engineering jointly  began  the  study of  a dry
sorbent flue  gas  desulfurization  process. These
studies showed that the  development of a good
sorbent material  that  would  absorb and desorb
sulfur dioxide in the temperature range of 600 to
700 F was feasible. Utility support was solicited
and obtained in 1969, and a three-phase research
and development program established. The final
phase  of  this  program  will involve  the design,
installation, operation  and testing of  a  150 MW
demonstration  system.  The  B&W-Esso  process
utilizes an electrostatic precipitator for controlling
particulate emissions  and a  dry  sorbent  for
controlling sulfur  dioxide  and  nitrogren  oxide
emissions.

It provides advantages over  existing  systems in
that the process reduces nitrogen oxide and sulfur
oxide emissions simultaneously and also minimizes
the  problem  of  disposing of  spent material.
Product gas from  the  process  is suitable for  the
production of  saleable  sulfur  or  sulfuric  acid.
Figure 13  represents an 800 MW coal-fired utility
boiler with the B&W-Esso system.
The  programs  described  here  are  part  of a
continuing research and development effort that
will provide the power  industry with some of the
answers to their air pollution control problems.

Coal remains as our most abundant fossil fuel, with
proven reserves estimated at 830 billion tons, most
of  which cannot be utilized  for future  power
production without some degree of sulfur removal.

REFERENCES

1.  Downs,  W.,  Strom, S.S., New Particle  Size
    Measuring  Probe  •   Application  to Aerosol
    Collector and Emissions Evaluations,  ASME
    Paper No. 71 - WA/PTC-7., 1971.

2.  Boll, R. H.,  "Particle Collection and Pressure
    Drop   in   Venturi  Scrubbers",  I&EC
    Fundamentals, Vol  12, pp  40-49,  February
    1973
3.  Bagwell, F.A., Cox, L.F., Pirsh, E.A., Design
    and Operating Experience With a Filterhouse
    Installed On an Oil-Fired Boiler, Air Pollution
    Control Association, St. Paul, Minnesota, 1968.

4.  Boll, R.H.,  A  Mathematical Model of SO2
    Absorption   by  Limestone  Slurry,   First
    International  Lime/Limestone  Wet Scrubbing
    Symposium,  Pensacola,  Florida,  March  16,
    1970.

5.  Downs, W., Kubasco, A.J., Magnesia Base  Wet
    Scrubbing  of Pulverized Coal Generated Flue
    Gas • Pilot Demonstration, Project Sponsored
    by   NAPCA,   Contract    CPA-22-69-162,
    September 28, 1970.
                                              102

-------
                        TABLE I
              National Primary and Secondary

               Ambient Air Quality Standards
      Pollutant
                               Ground Level Concentration
                                     Not To Exceed
                               Primary
             Secondary
Particulate Matter (a)

Sulfur Oxides (b)

Nitrogen Oxides {b)
 75

 80

100
 60

 60

100
(a)  Annual geometric mean
(b)  Annual arithmetic mean
                      403

-------
                          TABLE 2


           EPA Standards of Performance for New

            Fossil Fuel-Fired Steam Generators
Fuel
Emission Limits Not To Exceed
Parti culate
Ib/MKB
Coal . 1 (a)
Oil {a)
Gas (a)
SO2 NOX
Ib/MKB Ib/MKB
1.2 .7
.8 .3
.2
(a)  Visible emissions will be limited to not greater than 20%
    opacity.
                       404

-------
Particle diameter,
  DP (microns)
   100
    10
                                        I
      01
5  10            50

 Cumulative mass, % less than DP
90
99
                             Figure 1
            Typical fly ash particle size distribution
             for large PC and Cyclone fired boilers
                         Banco-Analysis
                               405

-------
  Theoretical
grade efficiency,
   100
    10
                    Press, drop
                      in wg
                        6
                        10
                        20
                        40
     .01
                          .1

              Particle diameter, DP, (microns)


                   Figure 2

Theoretical venturi grade efficiency curves
i.o
                                    406

-------
                    Temperature
10
TO
o
    Cascade

    impactor
0)



w ra

'o *-•
                                                     Drying
                                                    Column
                                                         cu
                                                         -*—'
                                                         03

                                                         E
                                                         o
                                                         4-J
                                                         o
                      Vacuum

                       pump
                             Figure  3

                   Cascade impactor sample train
                                407

-------
o

1 1
1
1 1
inch

                       Enlarged section A-A
                         Section A-A
                                               Collection
                                                  cup

                                               Spring


                                               Jet spindle

                                               Gasket
        Figure 4
   Cascade impactor

           408

-------
Particle diameter,
  DP (microns)
     100
      10
      0.1
            Boiler A
            Boiler B
            Boiler C
        0.1    1       10        50        90     99

                Cumulative mass, % less than DP
                      Figure 5
     Cumulative particle size distribution
                         409

-------
     Figure 6
Laboratory pilot plant

         410

-------
            Figure 7
Limestone wet scrubbing pilot plant

-------
                                                                  Limestone
                                                                   bunker
                                          To sluGge
                                        treatment plant
                    Figure 8
Will   County limestone wet scrubbing system

-------
                                                      Venturi -.Absorber Modules
< Slurry
Di
867 GPM
                                           11.612GPM
                                                Figure 9
                           WILL COUNTY INSTRUMENTATION AND CONTROL DIAGRAM

-------
                Figure in
La Cygne Limestone wet scrubbing system

-------
CWS
                                            STEAM FOR
                                         PROCESS & POWER


                                           RECOVERY  FURNACE
                                                                   MECH. DUST
                                                                    COLLECTOR
                                                     I      I
                                                  DIRECT-CONTACT
                                                   EVAPORATOR
                                                                             MAKE-UP
                                                                             (M9(OH)2
              COOKING
                ACID
              STORAGE
                           MULTIPLE-EFFECT
                            EVAPORATORS
                          SULFUR BURNER -
                           GAS COOLER
STRONG
  RED
LIQUOR
STORAGE
 WEAK
  RED
 LIQUOR
STORAGE
                                                Figure 11
                     Flow diagram of magnesium  base pulping and recovery

-------
                                               SO2 GAS TO
                                               ACID PLANT
                                     ABSORBER
                                      RECIRC.
                                       PUMP
                         VENTURI
                         RECIRC.
                          PUMP
F. O. FAN
TO SETTLING POND
             MAKE UP WATER &
             ADDATIVE FOR pH CONTROL
                                                  RECYCLE &
                                                  MAKE-UP WATER
                                        Figure 12
                              Magnesia scrubbing system
BOILER


H2O MAKE-UP

-------
               Electrostatic
               precipitator
                                                                         Stack
                                                                 Primary air heater
                                                                    (tubular)
                                                                  Air in
Tempering air duct
Primary air outlet
                           Figure 13
               800  MW B&W-ESSO system

-------
ONE  YEAR'S PERFORMANCE AND OPERABILITY
  OF  THE CHEMICO/MITSUI CARBIDE SLUDGE
  (LIME) ADDITIVE  SO2 SCRUBBING SYSTEM

                     at

                OHMUTA #1

            (156 MW-Coal Fired)
                    by

      Jun Sakanishi,  General Manager
            Miike Power Station
           Mitsui Aluminum Co.
            Ohmuta City, Japan
   Robert H.  Quig, P. E., Vice President
         Pollution Control Company
     Chemical Construction Corporation
              One Penn Plaza
           New York, New York
                      419

-------
                            FOREWORD
It is a great honor and pleasure to introduce and participate at this sym-
posium with Jun Sakanishi, General Manager of the Miike Power Station
(referred to by Chemico as Ohmuta No. 1, which is part of the Mitsui
Aluminum Company Ltd., Japan.  Sakanishi San has provided the super-
visory leadership necessary  to assure the success of this full-scale
multi-stage SO2 Scrubbing System which was retrofitted onto his  156 MW
coal-fired boiler.

Sakanishi San's participation has spanned the total history of the project
ranging from the basic design and pilot testing phases,  through engineer-
ing, construction, and start-up; to over one year's operation.  It was
Sakanishi San who trained his existing power plant operators to thoroughly
understand all functions of the system in terms of what had to be done
and, most important -- why it had to be done.   The interaction of design
engineering, comprehensive  training and skillful operation  has paid off--
the first commercial-sized,  successfully operating, highly efficient,
totally reliable SO  Removal Scrubbing System.

While it is Chemico's pleasure to assume credit for the basic design en-
gineering and the associated  process know-how, major acknowledgments
must be extended to the outstanding engineering and construction staffs
of  the Mitsui Miike and  Mitsui Aluminum companies who completed the
project in the unbelievably short period of nine months --within budget.
Similar work in the U.S.A.,  taking into consideration problems of work
practices and productivity in the construction area as well as the many
factors which can adversely affect engineering and procurement schedules,
would require a much longer period of time.

The System has been  totally reliable and available for all power generation
requirements of the Mitsui Aluminum Plant --a facility which is totally
dependent upon a low  cost, reliable power supply.  Much credit for this
reliability must be attributed to the exceptional quality of Japanese equip-
ment manufacturers.  The supply of conventional  system components such
as fans,  pumps, controls, etc.,  and the associated equipment servicing
by the Japanese,  has  proved  to be everything which they claimed and more.
This is something which American vendor counterparts should note that
they need to seriously improve upon.

The success of this installation is indicative of the effectiveness which
can be achieved,  if required,  on American  power plants because of the
following reasons:
                                  420

-------
(a)    The inlet SO2 concentrations at the Ohmuta Scrubbing System
      are quite similar to most of the SC>2  concentration ranges
      measured by Chemico during different pilot testing operations
      at 17 different power plant sites throughout the U.S.A. over
      the past five years. Only Western coal-fired operations have
      been different in that these inlet 803 concentrations are gen-
      erally lower than that experienced at Ohmuta.

(b)    Carbide sludge (calcium hydroxide) used at Ohmuta and vari-
      ous grades of lime which are expected to be proposed for most
      applications in the  U.S.A., have sufficiently common proper-
      ties for SO2  scrubbing purposes to correlate their mutual
      successes as additives. Simultaneous SO2 absorption pilot
      testing, using both types of materials under similar conditions,
      has demonstrated these correlations. This is not the case for
      considerations involving limestone.

(c)    The Ohmuta Plant is essentially a base loaded operation but
      has been subjected to sufficient load  swings and 803  concen-
      tration variations to demonstrate that the scrubbing system
      can readily handle  various turndown conditions.

      Since American applications envision the use of 150 MW
      modular trains, which is the approximate size of Ohmuta,
      there are no further scale-up factors involved.

(d)    Fly ash removal at Ohmuta is accomplished by depending upon
      a relatively efficient,  previously existing precipitator.  The
      first stage scrubber/absorption section also acts as  a polish-
      ing function to remove residual ash in the flue gas and has
      functioned to provide a system outlet loading as low as
      0.001 grains/SCF.

      The decision to incorporate  a precipitator on American in-
      stallations versus solely using a wet scrubber for dust col-
      lection, will be a function of client considerations for site
      conditions and the  results of associated cost /benefit  studies.
      SO.  absorption is common to both dust collection techniques.
        £t

(e)    The Ohmuta €3803/804 effluent bleed disposal system is
      designed to operate, and has operated, as a closed recycle
      loop between the scrubbers and the disposal ponds.  The re-
      cycle liquor has been totally saturated with sulfate for ex-
      tended periods of time (months) without significant scale,
      build-up or other deposition occurring in the scrubbers.
                             421

-------
            This success of precluding build-up has been essentially
            attributed to pH control throughout the liquid system com-
            bined with strategic utilization of fresh make-up water which
            is always required in any scrubber system to compensate
            for conventional stack evaporative losses.

            Some "spokesmen" have attempted to  speculate that the pond
            is purposely "blown down"  or "dumped" into an adjacent bay.
            This is not so.  There were periods during the year, however,
            when extensive rainfall from typhoons created an emergency
            pond overflow condition.  This problem  has been resolved by
            increasing the pond area and building  up of higher dike re-
            tainer walls.

            Pond management of this  effluent material is  achievable at
            conditions which are no more or no less complex than that
            experienced on existing plant sites in  this country today.
It is clear that the removal of SC^ for power plant stacks by the use of
any method,  is going to significantly increase steam /electric generation
production costs in this country.  The rationale for or against this  SO%
control has obviously become a matter of multi-opinions on many issues
ranging from questionable need for such SC>2 compliance in the first place,
cost benefits associated with the various methods, and who should pay for
what and how.  These issues are subjects, the rhetoric of  which, Chemico
defers to others.  We do say, however, that the SC>2 can be chemically
absorbed from the flue gas at high efficiency in a reliable manner,  and
have demonstrated as such at Ohmuta.  Thus, Chemico respectfully sub-
mits to this symposium that the criteria,  established by an adhoc panel
of the National Academy of Engineers concerning demonstrated SC>2 tech-
nology for over one year,  has been achieved.

There  is no technical reason why SC>2  removal via the demonstrated
Ohmuta scrubbing technique cannot be achieved in a manner which would
not adversely affect availability and reliability of the steam /electric
generation facilities.
                                         Robert H.  Quig,  P.E.
                                         Vice President
                                         Chemical Construction Corporation
                                         Air Pollution Control Company
                                         New York, New York
                                         April 26, 1973
RHQ:rg

                                   422

-------
                                         HEMICO
                            FLY ASH - SO2  SCRUBBING  SYSTEM
                                     MITSUI  ALUMINUM CO
                                   OHMUTA  PLANT  (156  MW)


                             SCHEMATIC   PROCESS   FLOWSHEET
NJ
OJ
1— _
1
EXISTING

FACILITIES

1—
*
RETRO- FITTED^
FACJLITIES

•
            BOIlER
PRECIP  i.D   STACK BOOSTER
      FAN       ~~
DELAY
TANK
                                        TWO STAGE
                                     VENTURI SCRUBBER
                                                            ONE MILE - C.SOj /C. 50./FLY ASH
                                                                    CARBIDE SLUDGE
                                                                 TANK
E_XISTIN&^ ASH
   POMD

-------
1.      Background of pro.ject of SQg control plant

       Mitsui Aluminum Co.,  Ltd. was established in 1968 with the
       objectives of 1) to support the local coal industry whose economic
       base was being weakened due to the drastic change in structure
       of the fuel source supply and 2) to  develop an Aluminum smelting
       plant which had been planned by Mitsui Group for many years.

       We believe that a basic requirement for the Aluminum Smelting
       Industry is to secure a cheap and stable supply of electricity.
       Miike Mining Station of Mitsui Mining Co., Ltd. located in Ohmuta
       City, is enjoying the largest quantity of coal production (6 million t/y)
       in Japan and produces high quality coal for the Steel  Industry.  In
       the process of this production, however,  a large quantity of slime
       coal which is unsalable waste coal is also produced.   Under the sit-
       uation, the basic requirement for Aluminum Smelting Industry has
       been resdlved by utilizing the low-grade slime coal.

       Although the coal produced by Miike Mining Station (so-called  "Miike
       coal") is the best in quality in Japan, it has two unfavorable character-
       istics.  First, it is easy to cause clinker problems and second, the
       coal has a high sulfur content. At the time of commencing the project
       of the power plant (1967-1968), therefore,  our primary efforts in boiler
       design were "how to avoid clinker trouble" and this was successful.
       As to the high sulfur content, nobody anticipated at that time the
       severe pollution  control codes which are prevailing today and there
       were no SO^ standards to be applied by codes.  Under the situation,
       we thought that it might be acceptable enough to build a 130 meters
       high stack as a measure of SO2 control which was also a recommendation
       of the Japanese MITI (Ministry of International Trade and Industry).

       The rapid progress of the Japanese economy has made it necessary to
       quickly strengthen pollution controls.  Asa result, power companies
       and others have been obliged to change their fuel sources.  The com-
       bination use of low-sulfur oil and high stacks was one of the most pre-
       vailing counter-measures at that time. In order to meet the initial
       objectives of establishing the company in the first place, Mitsui
       Aluminum Company was determined to continue the use of coal as a
       primary fuel source and to solve the SO2 problem through the operation
       of coal fired boilers withSC'2 control

       The power plant  was completed in March 1971.  At the start  of opera-
       tions coal which  was to have a heat value of 9,000 Btu/Lb and 2.1 %
       of sulfur content as  had been originally planned,  was changed to coal

-------
with a heat value of 10, 300 Btu/Lb and 1.7% of sulfur content.  One
year later, in March 1972,  the SO,, control system was completed
and placed in service.  Since then, the power plant with the SO2
scrubbing system has been operated  smoothly without any major
trouble,  satisfying  not only the national government control code,
but also the agreements with the local municipal administration.

Attached TABLE I shows the history of the power plant and the SO2
control plant at Mitsui Aluminum Co.

The work which Chemico had done to date best suited our criteria.
Mitsui Miike Machinery executed a license agreement with Chemico
and then we began to design the Ohmuta system.

Reasons for selection of Chemico/Mitsui Miike process are as
follow s:

First:   -  We believe that Chemico is one of the most  experienced
           companies in the field of chemical process engineering
           and  construction  in the world.  They are especially well
           experienced in applications of the large-sized venturi
           scrubber which is a critical part of a totally engineered
           scrubbing system.  Accordingly, easy size-up of the
           venturi absorption tower and reliable process analysis
           could be expected with Chemico.

Second:  -  The process including the venturi-absorption tower is
           relatively very simple when compared to other SO2 control
           technologies.  This advantage means relative easier
           operation and maintenance of the plant which requires a
           minimum training of operators.

Third : -  The use of venturi vessel scrubber designs  allowed that
           continuous long-period operation and high efficiency
           would be assured.  In addition,  the Chemico type SOg
           control plant allowed for a shorter construction schedule
           which could meet our target schedule of completion  of
           the plant construction by March, 1972 to satisfy the time
           schedule of MITI's air pollution code.

Fourth: -  The estimate of construction cost was comparatively low,
           relative to other technologies.

Fifth : -   Mitsui Miike Machinery  has had extensive experience
                             125

-------
           in the supply of chemical machinery. Its factory is
           located near the site so that the sufficient arrangement
           could be expected for our various technical requirements.

I wish to note that,  in order to assure the success in operation of the
first commercial SO2 control plant applied to a large-scale thermal
power station based on the wet Calcium base in cooperation with
Chemico/Mitsui Miike,  we installed a pilot plant witti a gas flow of
1/200 of the commercial plant at the site of the power plant when we
placed an order for the  commercial plant.  The  operation of this pilot
plant in technical cooperation with Chemico provided much knowledge and
experience in construction and operation of the commercial plant.
                            426

-------
2.     Brief history of selection of Chemico/Mitsui Miike process.

       In Japan, the dry type SO, control processes had been under develop-
       ment, sponsored by MITI and power companies. Dry type SO2 control
       processes were omitted from our selection because there still seemed
       to be many difficulties to be resolved in both the technical and econo-
       mical aspects.  In addition, these dry processes were considered in-
       applicable to our coal fired power plant which exhausts so much fly
       ash.

       Since we had to achieve reliable operation of the SC>2 control plant within
       the  limited time of construction,  the following criteria were established
       for  process selection after our various investigation and examinations.

       First: -    Economical requirement.
              Relatively low capital and  operating costs are critical in order
              to maintain power generation cost as low as possible. To
              meet this requirement the wet  type scrubbing process using
              a Calcium basis is more advantageous.

       Second:-  Simplicity of the process.
              Only relative simple developments or improvements to the
              process could be tolerated in order to achieve reliable oper-
              ation easily.  This requirement is very important since it
              will result in lower maintanance cost and easier operation of
              the plant.

       Third: -    Requirement for the absorbent.
              There is a plant of The Electro Chemical Industrial Co., Ltd.
              (Denki Kagaku.Kogyo Co., Ltd.) located in Ohmuta City
              where a lot of wet carbide sludge was disposed and dry carbide
              sludge is also being produced.  As these absorbents were
              available at low prices, we planned to adopt a Calcium additive
              SO2 scrubber system  even if we had a disadvantage of a possible
              problem of build up of scale -  a major concern of all at the time.

       Fourth: -   By-product from  waste  disposal.
              By-products such as sulfuric acid, elemental sulfur and sodium
              sulfite were not preferable under the market conditions at
              that time and the local conditions at the  site.  In case of Calcium
              base being adopted, either throw-away system or Gypsum
              process without the secondary pollution is suitable to our
              requirements.
                                    127

-------
Fifth:-     Requirement for space available.
       Since very small space was available,  the system which had
       the least number of scrubber-absorption towers required
       was one of our basic requirements.  Therefore, we required
       that the process designer should be well experienced.

Sixth:-     High efficiency.
       It was imperative that  high removal efficiencies of dust and
       SO_ at low stoichiometric requirement should be achieved.

Seventh:-   Assurance of stable  and continuous long period operation.
       It was also required that the process should not be too much sen-
       sitive in operation,  but rather  flexible to meet the change in
       load factor of boiler operation and the  fluctuation of sulfur
       content in coal.
                             428

-------
3.     Outline of SO 2 control plant and basic specification.

       The outline of this Power Plant and the SO- control plant is shown
       in the attached TABLE II.

       First:-     History of SO2 scrubbing with many different types of
              calcium additive processes has  shown that build-up of
              scale is one of the most difficult problems. However, we
              made the decision to employ this process despite the opinion
              mentioned above being widely accepted.

              The basic resolution was to establish optimum operating con-
              ditions which would minimize or eliminate scale by first using
              a  pilot plant.  Furthermore,  we planned to  install spare cap-
              acity for respective parts of the system including an extra
              scrubber vessel.  This stand-by scrubber unit and other
              spare equipment was found unnecessary later when the internal
              inspection of the scrubber system was carried out in November,
              1972 after the system had been in  service for  six months.

       Second:-    The project of by-product Gypsum production was postponed
              in order to first achieve successful performance of SO- control
              project.  We selected the throw-away system which enabled us to
              utilize the existing ash pond for disposal of slurry in a form of
              Calcium Sulfite (approx.  80 wt.%) and Calcium Sulfate,  (Now
              we are going to start Gypsum production.)

       Third:-    By the request of The Electro Chemical Industrial Co.,
              Ltd.  who supplied carbide sludge to the  SO2 control plant
              where carbide  sludge is used as absorbent, we installed a
              carbide sludge receiving system to handle either wet carbide
              or dry carbide whichever supplied.

       Fourth:-    In order to avoid problems  in boiler operation due to rapid
              pressure change in the duct when an unexpected trouble with a
              booster fan occured, we developed a specially designed duct
              which can handle backward flow of gas without any damper.

       Fifth:-     In determining the gas flow, we settled K value with a
              figure of 10 anticipating that it would be  strengthened to 5 in the
              future and we decided to install two units of scrubbers having a
              capacity of 75% each of total gas flow.   Two  reasons for the
              reserve capacity were a possible build-up of  scale and the anti-
              cipated strengthening of K value in future when two scrubbers will
              handle total gas flow.  The operation of one scrubber is enough
              to meet the present local code.
                                   429

-------
We proceed with the project with the basic specifications mentioned
now and we succeeded in completion of construction of the SO,
control plant within a short period of nine months in a  limitea area
of approx.  2, 000  square meters (21, 500 square feet).  In the mean-
time, the preventive measures against the disposal of  waste water
was also taken.  Flow sheet and arrangement are shown in the at-
tached drawings 1 and 2 respectively.
                            430

-------
4.     Outline of operation status.

       On March 29, 1972, the operation of the SO2 control plant was started
       and an excellent record of continuous operation for 202 days until
       October 17, 1972 was established.

       Further continuous operation was assured and no necessity of major
       modification was confirmed when the internal inspection of the plant
       was carried out for the first time during the annual maintenance shut-
       down of the power  plant from October 17 to November 10, 1972.  The
       plant  has been operated further quite satisfactorily since the plant
       operation was re-started on November 11,  1972  except a shut-down of
       the power plant for 4 days in January this year  due to mechanical
       trouble in boiler.

       Following is outline of operation status.

       First:-     Continuous operation for a long period.
              Fortunately, the operation was carried out without shut-down
              although there were several troubles, all of which  were just
              minor and resolved without difficulty while the system remained
              in service.
              Please refer to TABLE III for details.

       Second;-   Internal inspection during the annual maintenance shut-down
              of the power plant.  Every detail of internal parts was inspected
              twice,  namely during the annual maintenance shut-down of the
              power plant in October/November last year and at the time of
              boiler shut-down in January this year, and no major defect was
              found.
              Please refer to TABLE IV.

       Third:-    One of the  major items for daily maintenance  work.  PH
              values is being controlled strictly by means of sampling of
              recycle liquor at each 1st and 2nd stage every hour,   measur-
              ing pHvalue with a portable pH meter and then adjusting the
              quantity of make-up slurry at each 1st and 2nd stage using control
              valves so as to maintain pH value at a pH with the  tolerance of
              plus and minus 0.2 within the range of preferable pH value.

       Fourth:-   Basic  requirement for operation.
              Since it was the first commercial SO2 control plant in the
              world and any failure should be avoided, the operation was
              carried out at lower side of pH value within the range of
              preferable value where build-up of scale was not noticed
              during pilot operations.  The SO2 removal efficiency was
                                   431

-------
       sacrificed to some extent at the beginning. We are now trying
       to raise up the pH value gradually maintaining the smooth
       operation of the plant.

Fifth;-     Ash disposal pond construction.
       Please see the attached drawing 3, Ash Pond Layout. Originally
       the ash pond was  constructed for ash disposal,  however, it
       is used now for settlement of waste slurry from the throw-away
       system.  Supernatant is returned to the plant for re-use.  The
       ash pond occupies approx.  a million square feet and is  located
       at an area of approx. a mile away from the power plant. The
       wall of south-western side is lined with a polyvinyl film to seal
       the leak water in  order to avoid a secondary pollution problem.
       Perhaps because  of the heavy rains and mine water which flow
       into the pond,  there  is no accumulation of sulfuric acid, and
       we have had no trouble at all maintaining the plant operation
       during this one year.  The returned water is re-used for spray
       water, level control and make-up of Carbide slurry.

Fortunately,  operation is being carried out quite satisfactorily. The
requirements for a large quantity of recycle water had been regarded
as one of the  disadvantages  of Chemico process at the time of process
evaluation. Actually, however,  this disadvantage was proved later
to be one of the biggest advantages because of low solid concentration,
high SCL removal efficiency with lower pH value resulting in prevention
from build-up of scale.   The data on operation experience are shown
in the attached TABLE V.
                              432

-------
E conomic Features

Plant cost is about one billion Japanese Yen which is about 3 million
dollars and this amount works out at about 6, 500 Japanese Yen which
is about 25 dollars per KW excluding Gypsum plant.

Actual operating costs greatly underran the originally  planned budget
of 0.3 Japanese Yen/KWH which is about 0,1 cent/KWh for the past
year of plant operation.  We can definitely say that we have been able
to achieve the goals in our economic evaluation which had been planned
as one of the most critical requirements at the time of commencing
our project.  Our analysis indicates that firstly,  70% of the operation
cost is spent for interest, which means that the lower  plant cost will
greatly contribute to the lower operation and secondly,  the low cost
of absorbent by use of waste Carbide sludge shows  clearly an advantage
of wet type Calcium basis SC<2 control process. In addition, the mea-
sures taken to protect the plant from the possible build-up of scale
have been also very useful to reduce the maintenance cost.

The use of the  scrubber system has allowed the increased use of lower
cost slime coal in our boiler operation.  We have increased the quantity
of slime coal from 30% (on dry basis) to more than 80%.  This lower
cost coal contributed  in avoiding an increase of the power generation
cost by offsetting the  increased cost due to construction and operation
of the SO2 control plant.

The ash disposal pond is presently being filled with the waste slurry
consisting of mainly Calcium Sulfite. We are planning to start Gypsum
production within this year so that we can expect further reduction of
the operating cost by  sales of this by-product in place  of waste  dis-
posal. There is a difinite market for gypsum in Japan.

Conclusion
It is almost one year since the plant was completed. In spite of being
one of the first large-sized commercial plants in the world,  it has
shown an excellent result beyond our expectation with great satisfaction.
Moreover, to our surprise, so many people concerned not only in
Japan, but also in the United States and many overseas countries are
very interested  in our success.

It is our intention that to further try to attain higher performance and
more economical operation of the plant.  I believe that our strong
intent to maintain coal firing power station brought us the great success
in the  wet calcium base SOg plant.
                            433

-------
                  TABLE   I
        "Brief history of projects of Miike Power Station
         and SO3 Control Plant"	
Year

1967

1968

1969



1970
1971
1972
   Miike Power Station

Project was started.

(Nov.) MITI approved project.

Construction was  started.
    Control Plant
(Mar.) Test operation was
completed and the power plant
operation was started.

(Fuel coal:  10,250 Btu/Lb
            S:  1.7%)
(Feb. ) K value:  29.5

Investigation was started.

Approach to domestic and
foreign process owners
was started.
Research test was carried
out at Mitsui Engyo Co. ,  Ltd.
(Mitsui Salt Ind. Co., Ltd.)
using IHI-TCA test plant.

(Feb.) K value:  17.5

(Nov.) First survey team
was sent to overseas
countries.

(April) Second survey team
was sent.

(June)  Purchase order was
placed to Chemico/Mitsui
Miike Machinery Co., Ltd.

(July) Test using a pilot
plant was started.
(1,500 SCFM)

(Jan.)  K value:  11.7

(Mar. 29) Operation of SO2
control plant was started.

(Apr.) SO2 control plant
passed acceptance test by MITI.

-------
Year
1972
      Miike Power Station
SO2 Control Plant
(June) Installation of an additional    (May) Performance test
pulverized coal dryer was completed, was carried out.
             (Fuel Coal: 9, 550 Btu/Lb
                        S:  1.8%)
             (Oct.) Power plant was shut down
             due to the second annual maintenance
             and inspection.
1973
(Jan. ) Power plant was shut down
due to boiler tube leakage (for
4 days).
             (Feb.) (Fuel Coal:
                   9,900 Btu/Lb
                   S:  2.2%)
(Sept. ) Plant performance
was checked under various
operating conditions.

(Oct.) Inspection was made at
the first time to check
internals  of equipment of SO2
control  plant during the
power plant shut down.

(Oct.) Test operation of
Gypsum plant was completed.

(Jan.) K value:  9.34

(Jan. ) Inspection was made
again during shut down  of the
power plant.
                                  435

-------
                   TABLE II
               "Outline of Power Plant and
                   SO» Control Plant"
Electric power plant

(1)     Generator:

(2)     Tubine:

(3)     Boiler:



(4)     Combustion system:


(5)     Coal treatment:



(6)     Electrostatci Precipitator;


(7)     Stack:


SO,, control plant
 '"""""Si """" ~ ~  L J_J- J-TLLJ- ^™^^^
(1)     Process:
(2)
Capacity:
(3)     Performance guarantee:
(4)     Equipment details:
                             174, 000 KVA, 15, 000 V,  60 HZ

                             156, 250 KW 3, 600 rpm

                             490 T/Hr, Manufactured by
                             F.W. - I.H.I.,  Single drum natural
                             circulation type.

                             Pulverized coal firing.  Front
                             burners arrangement type.

                             Two slime coal  dryers, having
                             a capacity of 24 T/Hr and 45
                             T/Hr. each.

                             Dust removal efficiency:  98.7%
                             Outlet dust loading: 0.25  Gr/SCFD

                             425 ft high,  concrete outer shell.
Chemico/Mitsui Miike Machinery
process, wet type and Calcium base.

2 units of gas flow of 241, OOOSCFM
each.

One unit only is operated for the time
being to handle 75% (241, 000 SCFM)
of total gas flow (319, 000 SCFM)

Two units will be operated simultan-
eously when required to handle to
total gas flow in future.

SO 2 removal efficiency:  90% or mere
Dust   "         "     :  90% or more
Stoichiometric  requirement of
absorbent              120% or less
Continuous reliable operation.
Boost-up Fans   : 1, 000 KW x 2 units

Recycle Pumps: approx. 200 KW  x
                6 units.
                             436

-------
                                    Scrubbers        : 241, 000 SCFM x
                                                       2 units.
                                    Reheat Furnace   : 1 unit incl. Fan etc,
                                    SO _ Analyzers    : 2 units.
(5)     Disposal system:             Throw-away system.
                                    Waste slurry is transferred to ash
                                    pond and supernatant is returned
                                    from ash pond  to SO- control plant.
(6)     Gypsum plant:                Under planning.
                              437

-------
                           TABLE III
               "Report on minor troubles occured in
               operation of SO  Control Plant"	
1.
Modification for improvement
of carbide sludge receiving
system:
2.
Improvement of ash pond water
return pump:
3.
4.
Improvement of flushing
schedule of washing spray at
mist eliminator:

Change of make-up point of
Carbide slurry:
1)   Protector was installed to
    protect Agitator from damage
    caused by various foreign
    matters contained in wet
    carbide sludge.

2)   Dry carbide sludge contains
    approx. 3% of carbon particles
    and other foreign matters.
    Cyclone separator was installed
    to avoid possible troubles of
    piling-up in a slurry tank and
    errosion caused by carbon
    particles and foreign matters.

    Wooden chips flowed from
    timber yard into ash pond by
    heavy rain of typhoon in June,
    1972 and plugged suction side
    of ash pond water return pump.
    Use of substituted water caused
    increase of pressure drop of
    gas at 1st stage mist eliminator
    due to insufficient pressure of
    water for mist eliminator
    washing and may have caused
    build-up of scale.  Improvement
    was made and strainer of
    pump was installed.

    Flusing schedule was improved.
    Feed point at 2nd recycle line
    was improved to avoid build-up
    of scale.
                                   438

-------
5.     Improvement of connecting
       tubes for langential nozzles of
       recycle liquor of scrubber:
6.     Improvement of rubber lining
       arrangement after butterfly
       valve:
1)   Scrubbing liquor ran out due
    to loose fitting of chemical
    tubes. The tubes were changed
    to rubber tubes.

1)   Rubber lining of pipe inside was
    damaged at a part after butterfly.
    valve due to turbulence.  The
    rouble was  resolved by instal-
    lation of orifice.
                                    439

-------
                            TABLE  IV

                    "Status of equipment and internal
                    parts of SO2 control Plant inspected
                    during annual maintenance shut-down
                    of the power plant"
       Carbide Sludge
       receiving system:

       Scrubber:
       Recycle liquor piping;
       (rubber lining)
       Mist eliminator:
5.     Piping to disposal pond:
6.     Recycle water pipeline:
Piling-up of carbon particles was
found in slurry tank.

No defect was found at glass flakes
lining.  No build-up of scale was
also found.

No defect was found in general.
Thin deposit of scale  (1/25" to  1/12"
thick) was found but the plant was
operated without trouble.
Slight damage of rubber  lining  was
found at a part of piping  after
butterfly valve.
Partial build-up of scale was found
at the parts of:

1)   area affected by trouble with
    ash pond water return pump at
    the time of typhoon in June, 1972.

2)  outer area water coverage of
    spraying water was  not sufficient-

No trouble is anticipated in future
operation because of minor modifi-
cation in mist eliminator sprays.

No erosion,  corrosion and build-up
of scale were  found.

Muddy scale was found, but removed
completely without trouble.
                                    440

-------
7.     Pumps:

8.     Ducts:


9.     Delay tank:
       No trouble was found.

       Slight damage was found at expansion
       .pint.

       Piling up of carbon particles contained
       in slurry was found.
       Exhaust gas containing a very small
       quantity of 803 was connected with
       a stack thru,  vent gas piping,
441

-------
                          TABLE  V
                          "Operation Data"
                                               Most of figures were
                                               obtained in operation
                                               during January/February
                                               1973.
1.      Characteristics of coal.
              Heat value:                   9.800 Btu/Lb
              Moisture:                    6%
              Ash content:                  2.91%
              Volatile matter:              35.44%
              Fixed carbon:                 36.86%
              Moisture content:             1.59%
              Sulfur:                       2.2%
              Ash fusibility:                S.P. 2330<>F
2.      Characteristics of carbide sludge:
                  Moisture      Consumption         Purity
       Wet carbide:  55%     approx. 80 T/D       85% as Ca(OH>2
       Dry            6%        "    30 T/D       90%
       Total:          -              110 T/D (wet base)
Ca(OH)2
wt.%
j:84.3
87.8
CaSOg
1/2H00

1.8
3.9
CaSO4
2H2O

1.0
0.3
CaCOg

6.9
4.9
Si02

1.7
2.7
A1203

0.7
0.7
Fe2
wt
°3
.%
1.1
0.
7
200 mesh
thru.
%
99.72
87.12
 Dry
3.      Main operation data (as of October, 1972)
       Output:                     156, 000 KW
       Coal:                         9,700 Btu/Lb   S:1.9%
       Gas load to handle:          75%
                                   442

-------
             SO 2 concentration            Boiler outlet        1, 900 ppm
                                          Scrubber outlet     180 ppm
                                          Stack               610 ppm

             Gas temperature             Boiler outlet        315°F
                                          Scrubber outlet     130°F
                                          Stack               212QF
                                          Reheated gas        172op

             K value :                     7.84

             Ca(OH)24                    Stoichiometric requirement:
                                              105 to 110%

4.     Return liquor from ash pond.

             pH    8.2

                      ppm  Mg++70 ppm  Cl~ 728 ppm   NO2-0.28ppm
                 — 1,012 "   Fe++0.45 "   Mn++-
             SS    5.2   "   COD 6.4 "

5.     Slurry analysis.

             Ca(OH)2  CaSOq   CaSO4  CaCQ3   SiO2  A^Oa  Fe2O

    Make-up  87.3      0.5      0.7    2.3     1.9      0.24   0.8

    Bleed      6.0     69.5      12.2    1.02    5.8      2.6    0.3

-------
                          TABLE  VI

                      Explanation of "K value"
                      which is used as the SC^
                      emission standard in Japan
K value
       Constant ranging from 5.26 to 29. 5 according to the area stipulated
       by the Air Pollution Control Law of Japan and is shown in the following
       equation concerning SCX allowable emissions.
                       Q  =  K  x 10-3 x He2
       Where     Q:
Quantity of SO,, allowable emissions from
the stack.                *(NM3/HR)
                            *  Cubic meters at 0°C and 1 atm.
                  He:       Effective height of stack         (Meter)
                            calculated by the following Bosanquet's formula.
                            He  =  Ho  + 0.65  (Hm  +  Ht)
                  Ho:       Actual height of stack          (Meter)

                  Hm:      Equivalent height of stack      (Meter)
                            effected by discharged flue gas
                            momentum.
                  Ht:       Equivalent height of stack      (Meter)
                            effected by discharged  flue gas
                            temperature.
                  Both Hm and Ht are calculated by the following formulas
                  Wm  -  4-77           x /Q-1  vg
                  Hm  '  1*0.43  V_        V^
                                  Vg
                  Ht =  6.37 gxQl - t    (loge  j2  + _2_    2)
                                V3T!                J


                  J  =  V      (0.43 PI   .  0 28  VgTl
                                                  g*t

-------
Where            Qi:       Quantity of wet flue gas discharged from
                                                   (M3/SEC)
                  Vg:      Discharge velocity of the above gas    (M/SEC)
                  V:       Wind velocity                        (M/SEC)
                  TI:       Temperature at which density of discharged
                           flue gas becomes equal to that  of atmosphere (°K)
                  At:       Difference between temperature of
                           discharged flue gas and T^      (°C)
                  G:       Temperature gradient          (°C /M)
                  g:       Gravity constant      9.81     (M/SEC2)
                  T! and G are usually  assumed to be 288 K (15 5C) and
                  0.0033 respectively.
In case of Ohmuta city, (where Miike Power Station is located.)
                  K  =  9.34	  usually employed
                  K  =  5.26	   employed in an emergency when
                                           judged by the local government.
For the Ohmuta area, the K-value has been progressively decreased, leading
to more stringent emission limitations with time,  such as
                           1968:   K  =  29.50
                           1970:   K  =  17.50
                           1972    K  =  11.70
                           1973    K  =   9.34
                                  445

-------
•t
4S
01
                                                                          1st st-ze
                                                                          venturl
                                                                        recycle
                  Vet carbide pit
                                                                                       Aflh pond llouor
                                                                                       return pucp
                Dry carbide ptt
                                  Recycle slurry
                         ... — Makeup slurry
                         __.__ Bleed slurry
                         .„	Return liquor

Note:  Only one scrubber systea presently In operation.
       •  indicates a closed valve during f/72 to r>/72.
      M indicates a closed damper during f/72 to 9/72.
Waite disposal  pond
                                               Process Flow Sheet

-------
-tr
                                                             *  «V

                                           *«ey jf\<\^   V*
                                           ^f^y^\\^   *

-------
                              T=«b 1173
Ash

-------
-
t
                                        #1  Ohmuta  scrubber

-------
.-•
Ul
O
                                        #2   Ohmuta pond system

-------
               PANEL DISCUSSION:

       SIGNIFICANCE OF OPERATION TO DATE
OF 156-MW CHEMICO/MITSUI LIME SCRUBBING SYSTEM
                         451

-------
                               PANELISTS

P. Wechselblatt - Chemical Construction Corporation
J. Craig - Southern Services
H.W. Elder - TVA, Muscle Shoals
F.T. Princiotta - EPA
                           Pond Saturation
     P. Wechselblatt— In the questions that were raised  perhaps  the
one that everyone was most interested  in was the degree of  saturation
of the pond.  And I think that in order to view this,  it  is necessary
to understand exactly what is happening in the system. A great amount
of liquid is being recycled (perhaps 90 gal/1000 cfm in total), and there
is a huge storage of liquid.  The amount of liquid going  to the pond  and
coming back from the pond is very low.  In fact, if the liquid coming back
from the pond were boiler feed water,  it would change the degree  of
supersaturation by only 4 percent in the scrubber.  Therefore,
except for the fact that Bob Quig took a walk on it—and  I  am sorry there
was no picture of just a hard hat where he sank—the pond was completely
irrelevant to the degree of saturation or supersaturation in the scrubber.
And when we argue about the pond points, I would suggest  that if you
looked at the points in the scrubber recycle loop you would find  invariably
that the liquid is saturated or supersaturated; therefore,  the only line  in
question on the job is the return line from the pond to the scrubber.   That
line, of course, could scale if you had supersaturation,  but all  other lines
in the system are operating in the supersaturated mode.
                                 452

-------
                       Lime/Limestone Buffering
     The second thing that I would comment on  is  the  buffering action of
limestone and the buffering action of lime. There really  is  very  little
black magic in this.  The numbers are reproducible.   Everyone has  done a
great deal of work in the last 5 or 6 years.   When I  was at the  startup
of the plant in Omuta and helped with the performance tests,  we  altered
the carbide sludge to the system because our guarantee to  Mitsui Aluminum
was a more severe guarantee than their requirement from the State, from
the municipality.  So it was necessary to run  the system at times  for
guarantee purposes for the period for the guaranteed  tests at a  stoichi-
ometry where we could obtain better than 90 percent S02 removal.  Under
most of the yearly conditions, since the code  requirement  is not that
stringent, the plant operates between 80 and 85  percent removal  with a
lower utilization of carbide sludge.  The system was  remarkably  responsive
to alterations in the carbide sludge feed.  For example, if you wanted
to raise  the pH of the system from  7 to 8.5 (which might have a
stoichiometric change of perhaps 0.9 to 1.2), it was possible to watch
the pH change with the change in lime feed.  The response time  of the
system is excellent.  Now you try to balance the response time  of the
lime system against the buffering action of the limestone system.   I
really do not want to get into a hassle about it, but this rapid
response  really is a tremendous aid.  Particularly when
putting the system on automatic control (which would necessitate sulfur
dioxide measurements, gas volume measurements, and things of that sort),
you want  the system to respond very quickly.  This system responds almost
instantaneously.  And I might say that with a change in feed you can
                                   453

-------
watch the S02 out of the stack increase as you change the hydroxide  feed.
It is almost instantaneous.   You can watch it take place.
                              Mist Eliminators
        Another comment that was made this morning had to do with  mist
eleminators.  Chemico operates mist eliminators very successfully.   We
have had problems on other jobs, but we have never had a  problem with
mist eliminators.  We operate them on a rotating cycle and at a higher
velocity than anyone talked  about this morning.  In fact, if you take a
look at the design, you will  find that usually about 20 percent (15  to
20 percent) of the cross-sectional area is taken up by supporting
structure.  Therefore, the mist eliminators in general operate at  about
20 percent higher superficial velocity than anyone calculates when they
calculate a cross-sectional  area of the vessel.
                   Japanese Versus American Results
     I was impressed when I  saw the plant in operation.  I have seen
plants of various sorts in this country, and I would just like to  say
this, and I feel very strongly about it:  it is kind of a challenge.  I
am disheartened when I see a system designed in this country that  was
erected by the Japanese in 9 months.  We started at the same time
in this country, doing the same design work, and the first unit we will
have operational in the United States will be in another 2 to 3 months—
almost a 30-month schedule compared with their 9-month schedule.  They
take  our design sheets and process flow sheets, which they pay for;
they build a plant; and they start operating a plant.  They have no
particular problem, and if they can do it there, we can do it here.  And
I think it is a challenge to the industry.  There is no reason why the
Japanese can do it better than we can.
                                454

-------
     H. W. Elder - Okay, the subject is of significance and I  think  we
will all have to agree that in Japan the significance is unquestionable.
It has been an excellent installation.  The Japanese have done one
fantastic job of installing equipment and making it run.  And  it is  a
real credit to the initiative and fortitude of the Japanese that they
had this large measure of success.  We have a problem now in translating
that technology to application in this country, and this is in no way
taking a shot at what has been done in Japan.
                        Transfer of Technology
     I think there are some questions that must be raised about transfer
of technology from one application to another.  I hate to beat this  horse
some more, but this closed-loop business is important, I think; and  Pete
(Wechselblatt), I am not sure I agree with your total conclusions on that
point.  Number 1, it is too bad we have to talk about closed-loop systems.
We need some guidelines on quality of water that can be discharged from
any plant, not only limestone scrubbing power plants, but any  situation.
And that would give us some guidelines then as to how much blowdown  we
could take and how much fresh water we can add.  Therein lies  the
important point, because if you can add no fresh water, I can  practically
guarantee that you will have scaling.  Based on TVA's pilot plant operation
and my conversations with experienced persons, I believe that
use of saturated and supersaturated solutions for washing mist
eliminators (or for points where dilution is required) is absolutely
essential if you are going to avoid scaling.  So I think that  you have
to use at least some makeup water to make the system run.  So  then it
is just a question of the amount.  For your situation or for the Mitsui

                                  455

-------
situation, particularly with the low solids concentration  in  the
recirculated stream, the sulfate concentration can be rather  critical.
I think that the accepted (well, it is pretty well agreed)  way  to
control sulfate scaling is by recirculation of calcium sulfate  to  provide
precipitation sites.  And the lower the solids concentration, of course,
the lower the sulfate concentration.  So if it becomes saturated with
sulfate, I would expect scaling to occur.  The low solids  in  the loop
coupled with the clarifier liquor returned from the pond with a high
sulfate concentration could be dangerous it seems to me.  A couple of
other points of uncertainty, and again I am just raising questions, the
fact that the waste material settles in a fashion that is  a little bit
unusual based on our experience is a little perplexing.  I  think that if
the sulfite/sulfate ratio in the reacted solids is similar to that which we
have experienced, you would expect it to settle very poorly.  So this says
to me that maybe there is something different about the carbide sludge as
compared with freshly calcined lime—differences in particle  size  or
differences in surface characteristics, for example.  But  if  the sulfite/
sulfate ratio is the same and if the particle size is the  same  as  freshly
calcined lime, you would expect the reacted product to behave the  same
way.  As far as flyash removal from the system, I am not sure how  important
this is.  If I were personally putting in an order for a system for control
of S02 from a new power plant, I would certainly want to take the  flyash
out as well as avoid the expenditure for a precipitator.  What  effect
the dust has in admixture with the calcium reaction products  is really
not clear, but  it is worth  raising the question.   It affects both  the
chemistry of the liquor and, of course, it affects the erosion  or

                                456

-------
mechanical considerations in the scrubber system itself.   So  there  are
really four main points that I am raising relating to the transfer  of
technology:  the amount of fresh water and how it can be  added;  the
question of low solids; the effect of sulfate on precipitation;  and the
effect of flyash in the recirculating slurry loop.  As I  say, I  am  not
throwing rocks.  I am just raising questions.
      J.  Craig—I  will  keep my discussion  pretty  brief.   I feel  there will
 be a number of questions  from the audience  and  probably  some discussion
 on the  panel.   Apparently, the installation  is  meeting Mitsui's design
 criteria and  operating goals.   Mr.  Sakanishi  (J.  Sakanishi,  Mitsui
 Aluminum Co.)  and all  the employees at his  station  are to be commended
 on a well  run  and well  maintained operation.  Now,  as to what I saw there:
 I saw a  station  operating base-loaded, with  apparently very  little variation
 in the  flue gas  flow rate.   We observed operating logs for 8 days; based
 upon that,  there was probably plus  or minus  2 MW variation from around the
 156-MW  capacity of the unit.  In comparing  that variation to our stations,
 where we can  go from probably 40 percent to 110 percent  of the  flue gas
 flow rate in  a matter of  24 hours,  it leaves some question in my mind
 whether you can  just extrapolate Mitsui's operation to a typical plant in
 our system.   We  also observed that sulfur content in the coal was  pretty
 constant at Mitsui;  this  is because of where they get their  coal and how
 it is processed.   Again,  if you look at our sulfur content,  it  could vary
 as much  as  2  to 4 percent sulfur in a relatively short period of time.   I
 believe  both  of these favorable operating conditions allow Mitsui  to run a
 chemical plant the way a  chemical plant should  be run—either at or near
 steady  state  conditions.   I am not fully convinced myself that  we  can run
                                  457

-------
 our  typical  units in this same kind of mode.  I, too, walked on the pond,
 but  that  is  a little bit misleading—I am not a Sumo wrestler.  The pond
 has  a crusted surface approximately 6 inches thick.  It held me up, but
 Mr.  Sakanishi warned me not to go too far out.  We penetrated the crust
 very easily  with a stick; below the crust was a very pasty material.  I
 might indicate that the top layer was thixotropic in nature.  As a con-
 sumer, we would like to buy a system that includes a solution to the prob-
 lem  of waste disposal.  Mitsui apparently has solved their problem for a
 short operating time; they are going to a gypsum final  product.  Again,
 I am not sure that is the best solution for our needs.   In addition,
 there is sonic question regarding the water balance.  Members of the group
 walked around the pond and could not find any visible overflows from the
 pond.  That was on one given day and I cannot say that that was the way
 it was for a year.  We were not able to resolve the question of open-
 loop versus closed-loop operation.

                          Areas of Commonality
     F.  T. Princiotta —  First of all, I think we should discuss areas
 of commonality and areas of difference, with perhaps the first statement
 being that I do not think any one plant in the United States, for that mat-
 ter, is  completely representative of all  other coal-fired plants; it is
 an impossibility.   We tried to do it at Shawnee, I might add, and it is
just impossible.   Everybody has a different sulfur content in the coal,
a different ash content, and different sludge disposal  logistics.  So I
think it is unreasonable to expect one particular plant, particularly in
                                  458

-------
a foreign country, to answer all  questons regarding this  technology.
However, I think anyone who has studied the system carefully  has  seen  that
there are really amazing parallels to many U.  S.  applications,  many
areas of commonality.  In the first place, it  is  a lime-based American
system utilizing American-based technology. And, as everyone knows,
much potential exists for these systems in the United States  due  primarily
to the low cost of the input alkali.  In my opinion (based on discussions
I have had with Combustion Engineering people, Chemico's  pilot plant
experience, and the English experience 40 years ago), there is no reason
to believe, well there is no hard evidence, that there is any difference
in the characteristics of carbide sludge as opposed to calcium hydroxide.
The major difference that I see is that it is  a pure form of calcium
hydroxide.  Other areas of commonality include unit size, percentage  of
sulfur in the coal, and use of a precipitator.  The 156-Mw unit can  be
easily scaled up without any increase in size  of the basic module which
can handle this amount of flue gas.  The system uses 2 percent sulfur
coal, whereas many coals in the United States  may have higher sulfur con-
tent.  Yet, if you averaged out the sulfur content of all the coals  in
the United States, I suspect it would be close to 2 percent.  Admittedly,
it would be nicer if that plant happened to have 3,4, or 5 percent sul-
fur coal; but it is relatively close to many U. S. coals. Regarding  the
precipitator upstream of the scrubber, as you well know from several  dis-
cussions so far, many utilities plan to employ a precipitator.  The  Navajo/
Mohave experiments will have a precipitator upstream of the scrubber.  And
it is not necessarily any dramatic, in my opinion at least, cost advantage
                                  459

-------
 to  go  In with an all-scrubber system as opposed to a precipitator/
 scrubber.  I  think Gary Rochelle  (G. T. Rochelle, EPA) treated this in
 one of his earlier papers.  So again, it is not representative to all
 cases,  but a reasonable case.

                            Areas of Difference
     In my mind, the major areas of difference that are really signi-
 ficant were brought out by John Craig; they are in the areas of the sys-
 tem being base-loaded with relatively constant coal.  Again, of course,
 there are many systems in the United States that approximate this condi-
 tion but there are also many plants, particularly the older units, that
 do  not.  They do have wide swing loads, as we heard for the Will County
 unit.  This is an area of difference and could affect the control of the
 system.  I think we should also mention that within several months,
 Pete Wechselblatt referred to it informally, there will be a Duquesne
 Light scrubber on the line.  This will be very similar to the Japanese
 Chemlco scrubber and will, I think, answer quite a few questions about
 relating this technology to the U. S. situation.  I believe that much
 of  the success did relate to the use of Japanese equipment and Japanese
 construction personnel.  I think there is no question (based on Shawnee
 experience, and experience of chemical plants, refineries, and other chem-
 ical operations) that the reliability and quality control of American
equipment leaves a lot to be desired.  There Is no question about that
 in anyone's mind, really.  Regarding the blowdown situation, my own belief
 is that It is not very relevent because the model proposed for sulfate
                                  460

-------
scaling, which is what is at issue here,  indicates  that  the degree of super-
saturation of calcium sulfate is the significant parameter.  Now this has
not been proven, but it is the most widely held theory and  I think that
Pete Wechselblatt is right in the calculation that  there is only a minimum
of 4 percent difference in this degree of supersaturation in the scrubber
itself.  So I am not worried about whether the pond leaches or not.  Of
course, it presents a water pollution problem that  is  certainly serious
in the U. S.  But, as far as operability  is concerned, that may not  be
significant.  I think I should briefly mention the  settling differences.
Based on the results presented yesterday  in the sludge discussions,  I
think it should be mentioned that probably what we  are seeing  here is
not any difference in sludge.  What we are seeing is drying of the top
layer of sludge.  Based on everything that I have seen and heard,  I
would suspect that, in times of rain, this would probably end  up reslurry-
ing again and forming the boggy type sludge materials  that have plagued
other installations, like Will County.  So in my mind,  there  is nothing
magical about the sludge.  And I think we heard earlier  that,  given  the
right weather contitions, people have walked on Will County Sludge.

                                Open Discussion
     We are now open for discussions and anyone who wants to challenge
anyone else on the panel, go  right ahead.
     P. Wechselblatt -~ I am not picking points.  I recall one fact which
has not been stated and that  is that the sludge going to the pond  contains
approximately 15 percent calcium sulfate and, therefore, 85 percent calcium
sulfite.  And the significance of that is that calcium sulfate must be
                                   461

-------
coming out of solution because 15 percent is too high a number for 863
collection.  So there is calcium sulfate coming out of solution and this
is a demonstration that the solution is in fact over saturation.
     H. W. Elder — Frank {Princiotta),  I would like to question  one
of your points.  You say that the average sulfur in U. S.  coals is about
2 percent, but the system does not operate on a national average sulfur
level, you know.  It operates on what you have to burn.  I think you will
have to agree that there is a significant difference in removing lOOOppm
and SOOOppm of flue gas.  I think that the level certainly must influence
the liquid-to-gas ratios and other important design considerations. So
differences between the levels operated at Mitsui and those that we face
in the TVA system, for instance, are important differences.
     F. T. Princiotta  — Yes, and I agree.  But 2 percent, in my  opinion,
is not very far from many high sulfur coal problems, the average probably
being around 3 percent.
     I. S. Shah   (Combustion Equipment Associates, formerly with  Chemico)
— I was part of the successful  installation at Mitsui/Miike.  Last year
we were talking about whether to use lime or limestone, and I think there
is only one system in the world  that is operating successfully without mech-
anical problems of any kind and  that happens to be Mitsui  Aluminum. We
have a lot of installations with limestone, regardless of equipment prob-
lems or process problems, that are not still operating successfully.   So
now here we see a successful operating unit.  Instead of asking questions
about why it works, let us start asking how it works.  And let us  learn
something from that and make the technology better, rather than fighting
                                    462

-------
with "My lime system works and my limestone system does  not work."   The
second question regards the saturation solubility.  Everybody  has talked
about solubility of sulfur/sulfite, in the presence of all the compo-
nents.  Everybody is theorizing.  It should be this because that is  what
I have.   TVA got some that they say is the sulfite/sulfate ratio.  Maybe
they have a unique oxygen, unique coal, unique flyash.   Maybe, then, no-
body else has it.  We talked about the saturation solubility within  the
scrubber just like Pete (Wechselblatt) said.  It is 90;  it must be almost
beyond supersaturation because the amount of bleed that  you are taking out
from the scrubber loop is practically, I would say, within 5-10 percent
of the total recycled liquor.  Now we talked about the  ponds.   The next
installation that is going to come on line and which is  under  design, I
think, is TVA's Widow's Creek.
                    Open - versus Closed-Loop Operation
     Can I ask you a question, Bill (Elder)?  Have you  designed that pond
for no leaching, no overflow, or complete recycle loop?  And  if not, why?
     H.W. Elder — Because we cannot control scaling if we operate  in a
completely closed loop.
     I.S. Shah — So you said open-loop is required for reliability  of
operation.
     H.W. Elder — Well, you will hear more about this  during  the Widow's
Creek discussion, but basically we are. .  . .
     I.S. Shah — I was asking a question.  Is open-loop necessary for
reliable operation of the S02 systems or not?  In my opinion  it may  be
necessary.
                                     463

-------
      H.W.  Elder —  I was  trying  to  answer the question.  The system does
 have a  horizontal mist  eliminator and that mist eliminator is washed
 with once-through water.  The scrubber  loop itself is closed-loop except,
 of course, for  the  makeup required  for  evaporator of that that goes with
 the solids.   But the design of the  mist eliminator loop itself is open-
 loop.   We  would like to learn how to put that back into the scrubber sys-
 tem.  Based  on  some of  the progress we  have made at Shawnee, that may
 happen.  But right  now, it is designed  for the mist eliminator to have a
 blowdown from it.   You  are right, we probably do not know exactly what
 the sulfite/sulfate levels are.  But we know what happens when it gets too
 high:  we  have  grown some mighty nice gypsum crystals in the mist elimina-
 tor by  trying to use saturated (or  at least high sulfate level) liquor for
 wash.   So  we  know the effect even if we do not know the reasons.
     Alex  Weir, Jr. (Southern California Edison) — We have been able to
 operate, in essence, a closed-loop  system for about 3 months, and have
 washed with  lime slurry mist eliminators without any really serious prob-
 lems.
     Bob Sherwin (Bechtel  Corporation) — There is one point that is not
 clear to me.  I would just like to ask what is being done with the precipi-
 tator ash  at the Mitsui aluminum plant.   Does this go out to the lime pond?
     F.T.  Princiotta — I  do not believe it does, but why not ask Pete (Wech-
 selblatt)?
     P.  Wechselblatt — Well,  I do not know where it goes.   Mr.  Sakanishi
would.  But it does not go the lime pond because Mitsui  is  trying to
 keep the calcium sulfite as  pure as possible for gypsum.
                                   464

-------
     A. Weir — Okay, it does not get mixed.
     P. Wechselblatt — No, it does not.
     Joe Selemczi (Dravo Corporation) — I also had the pleasure of the
congenial hospitality of Mr. Sakanishi.  There are two points I would like
to contribute to this discussion.  One is that when we are talking about
sulfur content of coal, we have to look at the Btu value,  which was somewhat
lower for Japanese coals than for the Eastern coals in the United States.
So that coal would calculate out to something like 2.5 to 2.7 percent sul-
fur based on Eastern coals.  There is also a lot of varied thinking about
the sludge and why somebody can walk on the water of Gethsemane there.
Actually, what Mitsui has going into the pond is a lean slurry of calcium
sulfite sludge containing about 3 to 4 percent solids.  You can compare
this with a horizontal, long settling tank.  The heaviest fractions settle
out first and build up close to the discharge points.  That is what you
can walk on; nobody walked out to the section of the pond that is farther
away from the discharge point.  Now I can say this with confidence because
we have examined the sludge from Japan as well as many dozens of other
sludges.  When you operate the thickener with 3-4-5 percent solids content
in the sludge, you are in the free settling zone.  Correspondingly, your
discharge will behave entirely differently in your pond.  When you have a
20 percent or higher solids content thickener underflow, settling 1s going
to be hindered.  The discharge 1s not going to settle preferentially.  You
are not going to have the larger particles settle out close to the discharg
point and the fine particles travel maybe a few hundred feet or maybe a few
thousand feet If you happen to have that kind of distance.  These are the
                                    465

-------
comments I wanted to make.
     Ab Saleem (Peabody Engineering) — My comments are related to the L/G's
which are used and the S02 removal that can be obtained.  I would like to
submit to the panelists and the audience:  if you sprayed all that water
to an open shell with no internals you will get removal of 90 percent or
more.  And to me the 300-mm system pressure drop is excessive and unnecessary.
     F.T. Princiotta — Sounds like you may be selling a spray unit.  I
might add that, although it was not brought out, there is some particulate
removal requirement in the venturi scrubbers.  As I recall, it is approx-
imately 0.007 gr/scf of particulate outlet of that system.  As far as I
know there is nothing in the United States that can match that right now.
So keep in mind pressure drop does buy us some particulate removal.
     J. Ando (Chuo University) — Mr. Sakanishi wants me to explain some-
thing for him.  There is a variation of the operating load.  As I said,
the operating load is fairly uniform for the Mitsui aluminum plant.  But
he has the experience that on a few occasions the operating load could
amount to nearly 50 percent of the usual operating load, but he could not
follow up the variation.  It is not very often, so they operated very care-
fully and they have done it without any problem.
     J. Craig — Let me go a little bit further then.  It is not often
and for what time period did this occur?
     J. Ando — (after conferring with Mr. Sakanishi) - Several times 1n
1 year.
                                    466

-------
     J. Craig — But not on a daily  basis?
     J. Ando — Just for one day he  dropped,  and  then backed up.
     J. Craig — For one day?
     Question — I would like to ask Pete  (Wechselblatt)  a question.  Pete,
how critical is water makeup at various locations in the  scrubber  and mist
eliminator?  And if it is extremely  critical, what would  happen  if that
loop were closed and how, when you talk about the pond's  being supersatu-
rated, does it take a little bit of  water  to  drop the supersaturation in
order for it not to scale?
     f. Wechselblatt — Our newest jobs in this country are designed to
utilize fresh water makeup; that is, the liquid lost by evaporation. This
water is used preferentially for pump seals,  for fan spray wash  water
(if in fact there is a wet fan), and for mist eliminator  wash water.  Of
course there are differences.  I am  sure different equipment  suppliers and
different designers have different rates at which they  wash mist eliminators,
for example.  But the rate set at Mitsui is such that the normal evapora-
tion loss of the system (and therefore that concomitant makeup)  is satis-
factory to use all the fresh water makeup  to  wash the mist eliminator and
allow you to have approximately 50-75 gal/min left over for fan  spray wash
water and 5 gal/min per pump seal.  So in our view, the quantity of evap-
oration down to about 50 percent load is satisfactory to  add the fresh
water to where we think it is most important.
     Paul Cho  (EPA Region V) — I would like  to answer one of the questions
that John  (Craig) brought up.  That is, your worry about  some of the plants
which  are  operated  under  a  fluctuating  load  condition,   I think that is
not a technical problem in itself.  And secondly, people  are concerned
                                     467

-------
 about  open-loop or closed-loop.  I think this is a little bit confusing
 here without defining what percentage of makeup water we are talking
 about:  if operating loss itself were defined, you could see an absolute
 closed-loop system.  And I further suggest that a power plant in this
 country has cooling water operations and there is some border proton water
 available.  I was wondering if anyone has looked into this area as makeup
 water.
     J. Craig — Well, the discussion really is the applicability of Mitsui
 and what Mitsui demonstrates to the U.S.  Regarding the ability to bypass
 gas, are you saying you bypass a certain percentage of untreated flue gas
 to the stack?  If so, will  the regulations in Japan allow you to do this?
     P. Cho •— No, I think that the bypass I am talking about is that
 instead of using one scrubber, probably you can use two.  This is perfect
 application in terms of your fluctuating load condition.  If you have two
 modules and bypass you can vary the stoichiometric ratio as well as the
 amount of flue gas introduced into the scrubber.
     J. Craig — Well, we are talking about probabilities versus what we
 saw at Mitsui and the probability is an expensive gamble.
     F.T.  Princiotta — I think it is a good point though.  To some extent,
 John (Craig), would you not agree, that costs can help you buy reliability;
 having extra scrubbers, I do not think one can argue, does improve reli-
 ability.
     H.W.  Elder— I think the important point here on turndown though,
 Frank (Princiotta), is that with variation in gas flow rate or variation
 in sulfur level, the trick is to absorb a certain amount of sulfur in a
given volume of liquor and to precipitate part of it and accommodate some
as soluble solids until it gets out of the scrubber.  And if the ratio of
                                     468

-------
absorbed sulfur per unit volume changes  for any  reason,  it can upset
this balance between solubility in the scrubber  and,  therefore, pre-
cipitation in the scrubber.   And that is the key point,  really.
     A.V. Slack (TVA, Muscle Shoals)  —  There have  been  a great many
postulations here.  I wish we had a little more  in  the way of hard data.
Pete (Wechselblatt), as I understand  it, you consider that it does not
make any difference, as far as operation of the  scrubber is  concerned,
whether one operates with saturated liquor return or not.  Is that a
correct interpretation of your remarks?
     P. Wechselblatt — Yes.
     A.V. Slack — Well, this does not agree at  all with not only data
at TVA and EPA, but it does not agree with data  in  Europe.   I have seen
a plant there in which the lime system is being  operated closed-loop  with
saturated return liquor, and there was scaling.   And I  cannot  see much
difference in data, really, between your operation and  the  other.  So I
think this is a postulation on your part.  Now if you have  pilot plant
data in which you have actually operated a saturated return  liquor,  I would
like to see it.  I think, though, that at Mitsui the data that Gerry
McGlamery (TVA, Muscle Shoals) showed indicates  or proves that the  return
liquor is not saturated and until you do saturate it you cannot say  that
you would not have trouble either in Japan or in this country under  such
an operating mode.
     P. Wechselblatt — In answer to that, I would point out the very small
quantity of liquid going to and from the pond.   If the liquid coming back
from the pond is, for example, at 50 percent of saturation,  this system
is a more closed system than if you operated a pilot plant with a thickener
                                    469

-------
and underflow at 35 percent solids.
     Pat Rapier (Burns & Rowe) — In continuance of the question that came
up regarding the 300-mm pressure drop as being excessive, I have some ex-
perience.  It was many years ago, but it is directly related to this.  If
you take out the mist eliminators; if you take out all the impingement
baffles in the collector, whatever it is; if you use, as a substitute,
a 316 stainless cyclonic collector; if you precede the collector with a
316 stainless fan and keep the tip speed of the fan blades at over 12,000
ft/min ( I believe that will correspond to a 4- to 6-inch pressure gain
through the fan);  and then if you put the slurry in through the fan it-
self and let the fan do the contacting, you create an extremely high
collision rate between the particles in the air stream and with the water,
which indeed the fan itself can accelerate and atomize.  We tried this
using a 30,000 ft^/min unit, which is much smaller than we are talking
about here, but the same type of operation could be done on the larger units.
It would scrub  out about 100 percent of the dust particles that were available
down to the micron range, and down probably about as low as 0.25 micron.
The pH of the liq'uor was not regulated because in those days we were not
interested in trying to remove sulfur, but it was removing sulfur dioxide.
My recollection is that the pH of that liquor was coming out at about 5.5.
The circulation rate going in for 30,000 ft3/min scrubbing was 72 gal, a
large portion of which was recirculated.  We were furnishing a moderate
amount of evaporation water for makeup* but the main portion was recircu-
lated.
     P.  Wechselblatt —• Commercially, a device such as that might be repre-
                                    470

-------
sentative of a Typson disintegrator, which I believe is  comercially  avail-
able for particulate removal.
     Del Ottmers (Radian) — I have a question to be directed  to Mr  Wech-
selblatt.  As I understand it, you have a scrubber with  a hold tank.  You
are operating with two feeds,  two scrubbers with a total L/G of 90.   I
was interested in the supersaturation and how you control this.  In  that
regard, I wondered about the volume of the hold tank that you are using
and the size of the stream of slurry that you are taking to your pond.
     P. Wechselblatt — Well,  the size of the hold tank  provides approx-
imately 30-min residence time, and the bleed from the system to the  pond
I believe is 3 percent of the liquid rate.  But again, it is the kind of
number I would rather give you after the meeting.  It can be made available
to you; I just do not recall the number.
     R.H. -Quig (Chemical) — Bill (Elder), I direct a question to you.
You and I have had a running commentary from time to time on this recycle
loop.  I just want to put a couple of things in perspective.  Do you essen-
tially feel that this problem of closed-loop saturation falls into two areas
of concern?  One is that you are concerned that, if you have a saturated
return system, you will plug and scale the scrubber vessels regardless of
the different geometries or shapes or what they are - spray towers, ping
pong balls, venturi, whatever.  Is that a major point of concern?  The
second point is, and I agree with you for sure on this  one, the concern
for overall pond management and the impact of water quality standards.
But could you first comment on your concern for plugging and scaling as
related to this problem of saturation.  How fast do you plug and scale
your pilot units when you truly have saturated liquor 1n terms of sulfate
                                    471

-------
coming back?
     H.W. Elder — Well, you know, I can not say flatly that if it is
saturated it will scale.  What I do say is that if you can not add any
fresh water in the system, it would certainly scale.   Now it becomes
relative then, and the less saturated the recycle water the more fresh
water you can add.  And that has to be in the right direction.  Now the
other point, how long does the pilot plant operate before it scales up?
You know, we have scaled it up at 4 hours and we have used lime.  And
we have run for 1000 hours with limestone without scaling it up.
     R.H. Quig — Saturated?
     H.W. Elder — Yes, well saturated to the extent  that the purge from
the system is a slurry with 40 percent solids, 60 percent water.  That
goes out and does not come back.  That simulates what would happen if you
put it in a storage pond.  But under those conditions, yes, we have oper-
ated without scaling for a continuous run of 1000 hours.   We, I think, have
a pretty good understanding of the system, the chemistry, and the tech-
niques with limestone.  I do not understand the lime  system that well  and
perhaps if I did I would be less concerned about some of  these points  that
I have raised.  Again, it gets back to the reactivity of  the absorbent.
I talked about that yesterday: the difference in stoichiometry and its
effect on scaling in the scrubber itself.  Now what we are saying really
is that the reactivity changed.  More calcium was available to react with
the $02 that was absorbed.  Lime is a more reactive material than limestone.
Therefore, the amount of calcium available with the lime  system is higher
than with limestone.  I would expect the tendency to  precipitate sulfite
                                   472

-------
in the system 1s greater with lime.   As I  say,  there  are  some questions
1n my own mind about the difference  between lime and  limestone chemistry
that sort of confound our conversation.
     R.H. Quig — Well, yes, perhaps so.   But I would like  to refer to
this chart that you fellows prepared here.  It has several  points  of data
you have grabbed, based on information from visitors  to the Omuta  system.
And that chart (which we do not really agree with, but just let  us assume
at the moment that it is correct) shows a saturation  condition of  several
weeks.  We did not plug and scale, and I know the unit was  operating at
full load with the impact of the changing S02 concentrations due to the
different coals.  Now your chart says in effect that the  scrubber  system
operated saturated for about 6.5 weeks, if we look at the chart, under a
saturated or supersaturated condition.  And then your chart says that,  for
the rest of the year, it operated at less than that.   So  I  submit  that if
your data are correct  ( and it is not because you only have partial  infor-
mation due to returning visitors) it operated for 6 weeks  totally saturated.
I think that in itself has significance.  We happen to know that it opera-
ted much longer than that.  But giving you credit for you assumptions, 1t
still operated.
     H.W. Elder — Well, of course, you know the obvious  question. If 1t
operated a few weeks, why not a few months, or a few years?  We would have
had a lot more confidence in the performance had it operated 6 months
rather than 6 weeks with saturated conditions.
     R.H. Quig — Well, I would just like to conclude our position on that
issue:  we have gone through ranges of saturation and we find no impact
there.
     F.T. Princiotta — I think we a running out of time,  and I am sure
                                    473

-------
people are anxious for a break, so we would like to conclude discussions
on the panel.  I would like to thank all the panel members.
                                    474

-------
              THE
      TVA WIDOWS  CREEK
LIMESTONE  SCRUBBING FACILITY

             PART I
     FULL SCALE FACILITY
               by

         B .  G . McKinney
      Power Research Staff
   Tennessee Valley  Authority
     Chattanooga, Tennessee
          A. F. Little
 Division of  Chemical  Development
   Tennessee Valley  Authority
     Muscle Shoals, Alabama
          J. A. Hudson
  Division of Engineering Design
   Tennessee Valley  Authority
      Knoxville, Tennessee
                  475

-------
                                   THE

                            TVA WIDOWS CREEK

                      LIMESTONE SCRUBBING FACILITY

                                 PART I

                           FULL SCALE FACILITY

                                   BY

                             B.  G.  McKinney
                         Power Research Staff
                       Tennessee Valley Authority
                        Chattanooga,  Tennessee

                             A.  F. Little
                   Division of Chemical Development
                       Tennessee Valley Authority
                         Muscle Shoals,  Alabama

                             J.  A. Hudson
                    Division of Engineering Design
                       Tennessee Valley Authority
                         Knoocville, Tennessee
                               ABSTRACT


          The design of a limestone slurry scrubbing system for TVA's Widows
Creek Unit 8 (550-MWj plant located in Northeast Alabama) is discussed.  Basic
design premises on which the design is based are presented.  Descriptions of
major components of the scrubbing facility are described.

          Estimated capital and operating costs are contained in this paper.
The estimated capital cost of this installation is $^2,000,000 not including
a portion of the solids disposal costs.  The estimated capital cost is $1*3,636,000
if the full initial scrubber effluent ponding costs are Included.  The estimated
operating cost is approximately 2.9 mills per killowatt hour generated.
                                  476

-------
                                  THE

                           TVA WIDOWS CPEEK

                     LIMESTONE SCRUBBING FACILITY

                                PART I

                          FULL SCALE FACILITY

                                  By

                            B. G. McKinney
                         Power Research Staff
                      Tennessee Valley Authority
                        Chattanooga, Tennessee

                             A. F. Little
                  Division of Chemical Development
                      Tennessee Valley Authority
                       Muscle Shoals, Alabama

                            J. A. Hudson
                   Division of Engineering Design
                      Tennessee Valley Authority
                        Knoxville, Tennessee


          In mid-1970 TVA made a decision to install a full-scale
demonstration limestone scrubbing SOg removal system on generating Unit 8
at Widows Creek Steam Plant (in Northeast Alabama, near Chattanooga,
Tennessee).  The primary objective is to work out design and operating
problems that affect both S02 removal efficiency and process reliability,
with emphasis on the latter.

          Since late 1970 considerable small scale, pilot plant and
prototype developmental work has been done on the limestone scrubbing
process in an attemptto establish design parameters for the full scale
demonstration unit.  The results of the developmental work are described
in a separate paper.  A draft environmental statement covering the full-
scale demonstration S02 removal system and waste disposal pond was prepared
and issued on June 30, 1972.  Following receipt and resolution of various
state and federal agencies' conanents, the final environmental statement
was issued January 15, 1973?
                                  477

-------
          Following finalization of the environmental statement, the project
authorization was submitted to and approved by TVA's Board of Directors in
February, 1973-  Since engineering had commenced prior to the project author-
ization  it enabled construction to begin in March, 1973-  The engineering
for the project is about 50 percent complete.  Equipment bids have been
received and contracts awarded on most of the long delivery equipment items.
Specifications and requisitions for other material and equipment are at
various stages of completion.

          The mechanical completion date for the facility based on the present
project schedule is mid-1975•  Pre-operational testing is scheduled to be con-
ducted in the third quarter of 1975.

          An attempt is made in this paper to discuss the design of the
facility and the estimated capital and operating costs.  The design areas
of the facility to be discussed are:  basic design premises, limestone
handling and grinding system, scrubber system, and solids disposal system.

          Figure 1 is a general plot plan of the Widows Creek Steam Plant
showing the general location of the scrubber facility, limestone storage,
handling and grinding facilities, and the solids disposal area with respect
to existing facilities.
Basic Design Premises

          The following design premises were established for the scrubbing
facility during the planning stages of the project:

      1,  Coal analysis (as fired basis)

          a.  Ash content, 25$

          b.  Sulfur content, b.yf>

          c.  Moisture, 10$

          d.  Heating value, 10,000 Btu/lb

      2.  Capacity

          a.  Maximum power generation rate for Unit No. 8, 550-MW

          b.  Stack gas rate at capacity,  1,600,000 acfm at 280°F
               (5,325,000 Ib/hr)



   Based on a total of 33$ excess air including air heater leakage.
                                  478

-------
                                      NEW ASH DISPOSAL POND
              EXISTING ASH DISPOSAL
              AREA FOR UNITS T & 8
    3OO KV SWYO
                                                    POND RECYCLE WATEfr
                                                     PUMPING STATION
    POWERHOUSE
    UNITS 7*9
POWERHOUSE
UNITS 1-6
                                                                                           RELOCATED
                                                                                           WIDOWS CREEK
                                                                                           CHANNEL
                                                                                       UNIT 8  SCRUBBING
                                                                                       SYSTEM SOLIDS
                                                                                       DISPOSAL  POND
UNIT 8 SCRUBBING SYSTEM

   RIVER WATER PUMPING STATION

       LIMESTONE GRINDING AREA
                                              LIMESTONE DEAD STORAGE AREA
                                  GUNTERSVILLE    RESERVOIR
                                               FIGURE  I
                                       GENERAL   PLOT  PLAN

-------
       3.   Sulfur dioxide removal

           a.  Percent removal, 80

           b.  Inlet concentration, 3^40 ppm (wet basis); 37to ppm (dry basis)

           c.  Outlet concentration, 650 ppm (wet basis); 750 ppm (dry basis)

       k.   Particulate removal

           a.  Inlet particulate loading,  5.6 gr/scf (dry); 5.1 gr/scf (wet;
              3.6 gr/acf (280°F)

           b.  Particulate level at scrubber exit,2 0.020 gr/acf (125°F saturated);
              0.022 gr/scf (wet); 0.026 gr/scf (dry)

       5.   Stack gas reheat temperature, 175°F (50°F rise)


Limestone  Handling and Grinding

           A wet grinding system for limestone was chosen over dry grinding
because it is less expensive and does not produce a dust problem.  Even if
dry grinding were used, the ground limestone probably would be slurried
before feeding to the scrubber.

          A schematic drawing of the limestone handling and grinding system
is shown in Figure 2.   The facility is designed for receiving limestone by
both rail and truck from the quarry.   The limestone is conveyed from an
unloading hopper to either the live storage silo or the dead storage area.
Material will be reclaimed from the dead storage area as required to maintain
an adequate level in the live storage silo.

          Limestone is conveyed from the live storage silo to a wet ball
mill where it is ground from the purchased size (3A by 0 in.) to the
desired size.  The resulting slurry is pumped from the ball mill through
a classifier where the oversized particles are separated and recycled to
the ball mill.  The product slurry (Uo# solids) from the classifier goes
to a surge tank from which it is pumped to the scrubbing area limestone
slurry storage tank.
   Based on **• conventional ASME sampling technique.
                                   480

-------
GROUND
LEVEL?
 BELT
FEEDER
      UNLOADING
      a RECLAIM
        HOPPER
                       TO FUTURE
                       SURGE HOPPER
                                                         FROM POND WATER
                                                        ^RECYCLE PUMPS
                                                            CYCLONE
                                                           CLASSIFIERS,
00
TRANSFER
  POINT
                          BELT
                       CONVEYORS
               LIVE
             STORAGE
               SILO
             'BELT
           CONVEYOR
     DEAD
   STORAGE
                                           BUCKET
                                          ELEVATOR
                                CONVEYOR
                                                             MILL SLURRY SUMP
                                                               TANK a PUMPS
                      VIBRATING
                       FEEDERS
                                                                       L.S. SLURRY SURGE
                                                                       TANK a TRANSFER
                                                                           PUMPS
                                                                 TO L.S. SLURRY STORAGE
                                                                  TANK (SCRUBBER AREA)
                                        FIGURE 2

                   LIMESTONE HANDLING AND GRINDING SYSTEM

-------
          The grinding system is designed to produce a ground particle
size of 90 percent minus 200 mesh at the design rate of 50 tons per hour
dry limestone.  The 50 tons per hour limestone rate corresponds to a
limestone feed stoichiometry of 1.5 at full load and design conditions.
Pilot plant work indicates that little or no improvement in SO^ removal is
obtained by using limestone quantities above 1.5 of stoichiometric amounts.

          The unloading and conveying facilities up to the live storage
silo are designed for a 900 tons per hour limestone rate.  The live storage
silo has an effective capacity of 6400 tons.  The unloading and storage
facilities up to and including the live storage silo are designed to
accomodate a future limestone scrubbing facility on Widows Creek No. 7
generating unit which has a maximum rated capacity of 575-MW.  The dead
storage area is designed to accomodate 120,000 tons of limestone corres-
ponding to approximately 50 days capacity at maximum limestone  requirements
for Units 7 and 8.  The live storage silo will have about six days' capacity
initially and about three days if used to supply the maximum limestone
requirements for both Units 7 and 8.  The design of the unloading and
conveying systems will permit the use of only day-shift personnel for the
limestone unloading and reclaim operations.

          The belt conveyor and bucket elevator conveying limestone from
the live storage silo to the limestone surge hopper are designed for 200
tons per hour to accomodate both the Unit 8 scrubbing facility and a future
Unit 7 scrubbing facility.  The limestone surge hopper has a 50 ton limestone
capacity and is designed only for Unit 8 as are all the limestone facilities
downstream of it.

          A weigh feeder is used to feed limestone from the surge hopper to
the ball mill at controlled rates.  The grinding system is a closed circuit
wet ball mill system with cyclone classifiers.  The ball mill has a 11'-0"
inside diameter shell 20'-0" long and is equipped with a 1250 horsepower
motor.  The ball mill and cyclone classifiers are equipped with rubber liners.
A rubber liner was selected for the ball mill rather than a steel liner
because it provides a longer wear life, less maintenance, and a lover noise
level.

          The mill discharge slurry (about 65 percent solids) goes into the
primary compartment of a two compartment sump tank from which it Is pumped
to the cyclonic classifiers.  Sufficient pond water is added to the primary
sump compartment BO that the overflow or product slurry from the classifiers is
a 1*0 percent solids slurry.  The overflow from the  cyclonic classifiers goes
to the secondary compartment of the mill slurry sump.  The product slurry
overflows from the secondary compartment of the mill slurry sump tank into
a limestone product slurry surge tank which has about five minutes capacity.
                                   482

-------
At reduced rates slurry underflows from the secondary compartment into the
primary compartment of the mill slurry sump tank for recirculation to the
classifiers to maintain a constant feed rate to the classifiers.   Three 50
percent capacity cyclonic classifiers are provided with one being an
installed spare.

          Underflow from the classifiers containing oversized particles is
fed back to the ball mill for regrinding.  Sufficient pond water is also
added to the ball mill to maintain a 65 percent solids slurry in the mill
discharge.

          The product slurry (lv&f> solids) is pumped from the product slurry
surge tank to a limestone slurry storage tank located adjacent to the
scrubber facility about 300 yards from the grinding facilities.  The slurry
storage tank has a capacity of about 181,000 gallons corresponding to about
eight hours storage capacity based on the «M>vimny limestone slurry requirement.


Scrubber System Design

          A simplified flow diagram of the scrubber system is depicted in
Figure 3>  The scrubbing system consists of four identical trains each
handling 25 percent of the flue gas from  Unit 8.  The flow diagram shows
the major components typical for each train and those components common to
the four trains.

          The flue gas after exiting the existing electrostatic precipitators
passes through a fan, venturi, liquid-gas separation chamber, absorber,
entrainment separator, and a reheater for each train into an existing stack.
Flue gas bypass ducts are provided around each scrubber train to prevent an
undue amount of boiler downtime because of scrubbing system malfunctions,
particularly during the initial shakedown operation of the system.  A plenum
connecting the inlet ducts to the four fans is provided to permit operation
of up to T5# load with one of the scrubbing trains out of service.

          Limestone slurry, pond water recycle, and make-up river water  are
added to an absorber circulation tank.  Slurry is pumped from the absorber
circulation tank into the absorber and drains back into the tank.  The
recycle pond water is regulated to maintain the solids concentration  level,
about 10# by weight, in the absorber circulation tank.  The absorber circul-
ation tank overflows into a venturi circulation tank from which  slurry is
pumped to the venturi and drains  from the separation chamber  back  into the
circulation tank.  The solids concentration in the venturi  circulation tank
is dependent upon the inlet flue  gas flyash loading but will  normally run
about 15 percent.
                                    483

-------
CO
•p
RE HEATER-^ r-
""""* I FROM STEAM / 1
[ B.cao tf t

\_ I
n TO ASH ,*_! 4-
UbrUoAL rTJNU
, , . — •• 	 >,
•^

1 VENTURI \
-TO STACK PLENUM f /
v.
_ FROM ESP X 	 \ 	 '
^J •*• j^ J
^3— .
,-. FROM B.C a D TRAIN VENTURI
^ VENTURI CIRC. TANKS ^| n lAfcutt
^v^ TO SETTLING
'"^ POND \y-
ENTRAPMENT
SEPARATOR
r
TkT.itirA-/r
J \
1
{
r
i *•
i
i
\


* fc
TRAINS < )—
B.caD ^J
1
ABSORBER
L
i
ir^ '
tir
1
1
1
^
CIRC. ABSORBER CIRC.
PUMPS TANK a PUMPS
i
i ,
SLURRY PUMP
SEAL WATER
, HEADER
/ ^t ^.
™ I'D
rRAINS<*— 1

r
i
TO P-«
TRAINS/ f—
B.C8D J ^__
__i— ~s RIVFR
— «T>*Jl!l£i 	 j
WATER PUMPS
FROM POND WATER A-N
RECYCLE PUMPS ^
FROM LS. SLURRY /->
TRANSFER PUMPS ^
1 i


L S. SLURRY
STORAGE TANK
                 EFFLUENT SLURRY

                SURGE TANK 8 PUMPS
FEED PUMPS
                                      FIGURE 3
                          SCRUBBER SYSTEM  FLOW  DIAGRAM

-------
          The venturi circulation tanks on each of the four trains overflows
into a common effluent slurry surge tank from which the effluent is pumped
to a settling pond.  Supernatant pond water is recycled to the absorber
circulation tanks and to the limestone grinding system.

          Four flue gas fans are provided, one per train, to supply the
static pressure for the system losses.  The design capacity for each of the
fans is 1»OO,000 acfm (280°F) with a test block capacity of U60,000 acfm
(280°F).  The fans will be equipped with induction motors and variable speed
fluid drives to permit flue gas flow regulation.  The static pressure will
be sufficient to convert the Unit 8 furnace from a forced draft furnace to
a balanced draft furnace.  The fans are a double width, double inlet and
radial tip blade design with an A-2U2 steel casing and impeller with high
strength alloy wear plates.  After investigating and evaluating the advantages
and disadvantages of locating the fans upstream of the Venturis versus
downstream of the reheaters, it was decided to locate them upstream of the
Venturis.  It was concluded that the dry fly ash erosion problems would be
less severe than corrosion and solids build-up problems vith the fans located
downstream of the reheaters.

          The venturi is used to cool, saturate, and remove fly ash from  the
flue gas.  It is designed to operate with a differential pressure of five
inches of water at full load conditions; however, higher differential pressures
may be used at reduced loads.  The venturi is a rectangular throat design
equipped  with a motor operated variable throat mechanism.  The venturi
housing material is 3l6L stainless steel and the convergent section is lined
with castable silicon carbide (2 inches thick).  The throat damper blades
are lined with fired alumina brick.  The venturi throat is 23*-0" wide
(inside), and has an effective throat depth of 28" (2-lU" sections).  The
expected SOg removal in the venturi is about 10 percent.

          Constant speed elastomer lined centrifugal slurry pumps will be
used to pump slurry from the venturi circulation tank to the venturi.  A
total of six venturi circulation pumps are provided, one operating pump per
train and a common spare for each two trains.  The pumps are designed to supply
a liquid circulation rate to the Venturis corresponding to an L/G of 10 gal/
Mcf of saturated  flue gas at the design flue gas rate.  At reduced flue gas
rates the L/G will be higher.

          The venturi circulation tanks have a capacity of ^0,000 gallons
and are l8'-0" in diameter and 25'-0" high.  The retention time in the tanks
is about 12 minutes based on the venturi circulation rate.  The tanks will be
provided with agitators to maintain solids suspension.
                                   485

-------
           The absorber is  a grid type  absorption  tower.   It  is designed
 for a superficial  gas  velocity of approximately 12  feet per  second at the
 design gas rate.   Its  dimensions are 30  feet vide,  l6 feet deep, and $k
 feet high.   The absorber has  five grids, 65-70  percent open  area, spaced
 either four or five  feet apart.   A final grid selection has  not been made,
 but preliminary plans  are  for them to  be of the floor grating type and made
 of 316 stainless steel (top and bottom grids) and fiberglass reinforced
 polyester  (intermediate grids).   The expected SOg removal efficiency for the
 multi-grid absorber  is about  70 percent  and the pressure  drop is expected
 to be about 2 inches of water with clean grids.   The  corresponding overall
 expected S02 removal including the venturi  is about 80 percent.  The slurry
 is distributed in  the  absorbers  using  316 stainless steel distribution
 headers equipped with  low pressure drop  (5  psi) spray nozzles.  The use of
 turning vanes  in the venturi-absorber  sump  to give  better gas distribution
 in the absorber is being studied.

           The  absorber circulation pumps  will be  elastomer lined centrifugal
 slurry pumps designed  to supply a maximum slurry  rate to  the absorbers
 corresponding  to an L/G of  60  gal/Mcf  at  the design gas rate.  A total of
 10 pumps are provided, two  operating pumps  per  train  and  a common spare for
 each  two trains.    The  pumps are  equipped  with variable speed drives for
 varying  the slurry rate to the  absorbers.

          The  absorber circulation tanks  have a capacity  of  1^8,000 gallons
 and are 33  feet in diameter and  25  feet high.  The  retention time in the
 tanks based on the maximum  absorber circulation rate  is seven minutes.
 Agitators are  provided in the  tanks to maintain solids suspension.

          The  entrainment separator is a  chevron  vane type (k pass) located
 vertically  in  a horizontal  shell.   The vanes will be  316  stainless steel,
 12 inches deep and spaced on 1^ inch centers.  The  superficial gas velocity
 in the entrainment separator is  approximately nine  feet per  second at the
design gas  rate.   The  face  of  the  eliminator blades are washed continuously
with  once through river water  at a rate of  1 gpm/ft.   The wash water drains
 from  the entrainment separators  to  the ash  disposal pond.  The environmental
statement may  be referred to for the water  quality  considerations for the
ash pond overflow which will consist of ash sluice water  from all Widows
Creek generating units and the entrainment  separator  wash water from the
Unit 8 scrubbing facility.
                                   486

-------
          The venturi liquid-gas separation chambers, absorber shells,
venturi circulation tanks, and absorber circulation tanks will be
constructed of Corten A steel and coated or lined inside.  Although a
final coating or lining selection has not been made, the primary con-
siderations are for a poXyurethane elastomer base coating (50 mils)
for the separation chamber and absorber shell, and flakeglass reinforced
polyester resin (60-8© mils) for the tanks.

          Indirect tubular exchangers (one pass) are used for reheating
the flue gas from the saturation temperature of about 125°F to 175°P
(50°F reheat) to desaturate and provide buoyancy for the gas.  Steam at
approximately 500 psig and 650°F from Unit 8 is used as the heating medium.
The material of construction to be used for the tubes has not been defined.
Corrosion specimen are being evaluated in pilot plant tests to determine
the material of construction for the reheater tubes.

          Slurry piping four inches and larger in diameter will be soft
rubber-lined carbon steel.  Smaller slurry piping will be FVC-coated
aluminum or 3l6L stainless steel.  A decision has not yet been made on
the type of slurry block valves but they will likely be either an
elastomer lined plug type valve or a knife gate type valve.

          Steam operated soot blowers will be provided at the locations
in which solids deposition is expected  to occur based on research and
development experience.  The locations where soot blowers are to be
provided include the inlet ducts to the venturi, the elbow between the
absorber  and the entrainment separator, and the reheater section.

          A perspective drawing of the scrubber area is given in Figure k
which shows the general arrangement of the equipment.  The controls for
the system are located in the existing Unit 8 control room to minimize
operating personnel requirements.  Adequate controls and instrumentation
are being provided to minimize operating personnel, to monitor and control
the system performance, and to provide the necessary data for studying system
variables.

          Provisions have been made in the design of the scrubber system to
make the system as flexible as practical.  The absorber is being designed
so that it can be operated as a multi-grid, turbulent contact or spray
tower absorber.  The base case for design is the multi-grid  absorber which
has been discussed above.
                                    487

-------
                                    RE HEATER

                                    fNTRAlNMENT SEPARATOR
                       ABSORBER CIRCULATION
                          TANKS
              -ABSORBER CIRCULlTIOtl
                 PUMPS
                                                       f UJE G*S DUCT
                                                         POWER HOUSE
                                               EKISTING ESP'S
               FIGURE 4
SCRUBBER AREA  PERSPECTIVE VIEW

-------
          Provisions axe made in the absorber design BO that it can readily
be converted to a mobile bed type absorber should it prove desirable in the
future to obtain  improved S02 removal.  The variable speed absorber circu-
lation pumps permit the use of lower L/G's required for mobile bed absorbers..
Also the flue gas fans are designed with sufficient static pressure head
to accomodate the higher pressure drops associated vith mobile bed scrubbers.

          Another back-up which is being investigated is the changing of
the limestone scrubbing process to a double-alkali process.  Investigations
have been conducted which indicate that the conversion to a double-alkali
process is feasible.  Pilot plant tests are planned in late fiscal year
1973 to further define a double-alkali process for the Widows Creek Unit 8
scrubber facility.  Hopefully pilot plant testing will show that a double-
alkali system is a possible and practical back-up process for the limestone
scrubbing process being installed.
Solids Disposal System Design

          Ponding was chosen as the most feasible method of disposing of
the waste by-product solids from the Widows Creek scrubbing facility.  The
by-product solids consisting of flyash, reaction products (hydrates of
calcium sulfite and sulfate) and unreacted limestone are pumped to the
pond as a 15-16 percent solids slurry.  The supernatant liquor from the
pond is recycled to the limestone grinding area and the scrubber area.

          Thickeners could be used to concentrate the -urge slurry and
reduce the pumping rates to and from the ponds.  However, the savings in
pumping costs do not Justify the additional capital investment for
thickeners.  Also there is doubt as to how effective thickeners would be
in concentrating the purge slurry because of the very poor settling
characteristics of the reactant products in the slurry.

          The poor settling characteristics of the solids in the bleed
slurry to the pond result in large pond volume requirements for solids
disposal.  Based on the TVA pilot plant data, the weight percent water in
the settled solids after settling for extensive periods (up to 2to days)
ranges from 57 to 66.  For planning the solids disposal pond requirements
for the scrubbing facility 60 percent water in the settled solids was
assumed.  This corresponds to a pond volume requirement of about 1.^ cubic
yards per thousand pounds of solids pumped to the disposal pond.  Based
on the expected average coal analyses with regard to sulfur ezid ash content
the solids disposal rate in terms of settled solids corresponds to about
150 cubic yards per hour at full load conditions.
                                   489

-------
           The  initial  scrubber  effluent  solids disposal pond vill have an
 effective  volumetric capacity of U.5 MM  cubic yards.  The area  of the pond
 including  dikes  is about  100 acres and the dikes average about  30 feet in
 height.  The life of the  pond is estimated to be about seven years based
 on projected capacity  factors.   The pond dikes are designed so  that they
 can be raised  10 feet  to  increase the pond capacity to 5.8 MM cubic yards
 (increase  of 1.3 MM cubic yards).  The total estimated pond volume required
 over the remaining life of the  planl^ based on the above settled solids
 density data and the projected  capacity  factors, is 9«3 MM cubic yards  and
 thus  an additional storage volume of approximately 3.5 MM cubic yards would
 be required for  which  no  provisions are  made.  Due to the developmental
 nature of  the project  and the possibility of developing means of enhancing
 the settling characteristics of the solids, it was not deemed justifiable
 to provide for the total  estimated pond  requirements for the remaining life of
 the Unit 8 generating  unit,  TVA is continuing research and development work
 on means to increase the  compaction of the settled solids to reduce the pond
 volume requirements for solids  disposal.


 Estimated  Capital Costs

           In Table I,  the estimated capital costs  for the wet  limestone
 scrubbing  facility at TVA's Widows Creek  Unit 8 are summarized.  These costs
 are based  on cost estimates prepared by  TVA, which are based on detailed
 layouts and firm prices for much of the  major equipment.  The estimate is
 based on construction  beginning in March, 1973, and with completion in
 August, 1975.  The estimated direct cost for the scrubber facility is
 $22,36*0,000 not  including a portion of the solids disposal costs.  If the
 additional cost  of solids disposal is included, the direct capital cost is
 $23,61*0,000.  The total field construction costs are $31.400,000 and
 $33,018,000 respectively.  The  total project costs are $1*2,000,000 and $^3>
 636,000 respectively.  The cost  estimates do not include the cost of land
which is associated with  the pond or the other parts of the scrubber facility.


Estimated  Operating Cost

           The estimated operating costs  are summarized in Table II.  The
estimated  total  operating cost  is approximately 2.9 mills per kilowatt
hour generated.  For an investor-owned utility, the capital charges portion
of the operating cost would be  higher for the same total investment due to
differences in cost of money, income taxes, and so forth.  The  estimated
total operating  cost for  an investor-owned utility, based on the same para-
meters (capital  investment, operating life, capacity factor, etc.) except for
capital charges, is approximately 3.5 mills per kilowatt hour generated based
on capital charges of 15 percent of the  total investment per year.
                                   490

-------
TABLE I                   WidovB Creek Unit 8

                    Limestone Wet Scrubber Facility

                     Capital Cost Estimate Summary


Item                                                          Estimated Cost, M$

Grading, landscaping, yard drainage, surfacing                         ITT
Roads, sidewalks, bridges                                              608
Power house modifications                                               35
Electrical equipment building                                          100
Ductwork                                                              2000
Fans                                                                   960
Reheaters and soot blowers (includes steam & condensate piping)       1525
Railroad facilities                                                    320
Ball mill building                                                     200
Limestone grinding facilities                                          255
Limestone conveying facilities                                         919
Limestone storage facilities                                           562
Mobile equipment for limestone handling                                155
Scrubber area foundations                                              350
Scrubber area steel structures                                        1110
Scrubbers                                                             1280
Pumps                                                                  T65
Tanks (including linings and agitators)                                390
Entrainment separators                                                 150
Piping                                                                2^T1
Elevator                                                                80
Painting                                                                30
Instruments and controls                                               550
Electrical work                                                       2821
Electrical transmission plant                                         1526
Cranes and hoists                                                       92
Solids disposal area                                                   T6T
Construction facilities                                               2162

          Total direct cost subtotal                                22,360

Field general expense                                                 26TO
Allowance for shakedown modifications                                 2000
Contingencies

          Total field construction subtotal
                                   491

-------
 Table   I,  Continued
 Item                                                          Estimated Coat, M$

 Indirect  Costs
   Engineering design
   Managers  office  -  Office  of Engineering Design & Construction
   Office  of Power  (research,  development, & coordination)
   Initial limestone  supply  and preoperational testing
   Employee  compensation  benefits
   Administrative and general  expenses
   Interest  during  construction
   Other

          Total indirect cost subtotal                              10,600
Total project cost excluding supplemental pond costs                42,000


Additional supplemental pond cost allocations

  Direct costs
  Field general expenses
  Contingencies
  Indirect costs

          Subtotal                                                   1,636
Total project cost including supplemental pond costs                43,636
Notes

1.  The total project cost with additional pond cost allocations from
    separate authorization contains a total of about $2.8 million
    ($2.2 million direct cost) for scrubber effluent solids disposal
    pond.  The pond will have an effective capacity of 4.5 million
    cubic yards which will last about seven years at projected Unit 8
    load factors based on present expected settling characteristics.
    The pond dikes are designed so that they can be raised in elevation
    to obtain an additional storage capacity of 1.3 million cubic yards
    at an estimated cost of about $1.6 million ($1.2 million direct
    costs).

2.  The estimate includes about $2.1 million ($1.2 million direct
    costs) for solids disposal over and above the pond costs.

3-  The estimate includes about $7.0 million ($4.0 million direct
    costs) for limestone handling, storage and grinding facilities.
                                   492

-------
TABLE II
                    Estimated Annual Operating Cost

                        TVA Widows Creek Unit 8

                      Limestone Scrubbing System
                                                 Estimated Annual Cost,
   Item

Direct costs

  Raw material - limestone^/

  Conversion costs                .  .
     Operating labor & supervision.!/
     Utilities
     Analyses     .
     Maintenance^/
       Subtotal conversion costs

  Subtotal direct costs
                                      Limestone  ,
                                      Processings
                                          299
Scrubber
  Area
                                                        821
   105
   986
    30
  1390
  2511

  3332
 Solids
Disposal
                                                                      27
                                                                       5
                                                                      10
    172

    172
 Total
                         821
             159
            1033
              50
2$

3803
Indirect costs

  Capital charges^/
                                          TOO
  3170
    770
3/
4
  Overhead
     Plant, 20# conversion cost            60
     Administrative, 1056 operating labor	3
       Subtotal indirect costs            763
Total annual operating cost

Operating cost, mills/kWh generated

Notes

I/ Based on capacity factor of 65 percent (3135 x 10° kWh/year generated).
2/ Limestone handling, grinding and storage facilities.
   273,600 tons limestone at $3 per ton delivered.
   Operating labor and supervision at $6 per hour.
   Annual maintenance costs are based on 4.0, 6.0, and 3*0 percent of total
   field construction of $5.6 million, $23.1 million, and $U.3 million,
   respectively,for limestone, scrubber and solids disposal areas.
6/ Annual capital charges based on 10 percent of total investment (25-year
   life)  except capital charges on $2.8 million for pond costs in the solids
   disposal area are based on 20 percent (7-year life).
                                   493

-------
          It should be noted the effect which capital investment has on
the operating costs.  The capital investment is approximately 50$ of the
operating costs.  The second largest contribution to the operating cost
is maintenance and the third largest contribution is utilities.

          Also, the capacity factor (65 percent) used in this estimate
may be high for a unit approximately 10 years old.  A decrease in the
capacity factor would increase the operating cost substantially.
                                   494

-------
                           THE
                   TVA WIDOWS CREEK
             LIMESTONE SCRUBBING FACILITY

                         PART II
   PILOT-PLANT AND  PROTOTYPE OPERATING EXPERIENCE
                            by

J.  J.  Schultz, T. M. Kelso,  J.  L. Graham, and J.  K. Metcalfe
              Division of Chemical Development
                 Tennessee  Valley Authority
                  Muscle Shoals, Alabama
                       N. D.  Moore
                   Power Research Staff
                 Tennessee  Valley Authority
                  Chattanooga, Tennessee
                               495

-------
                                  THE

                            TVA WIDOWS CREEK

                      LIMESTONE SCRUBBING FACILITY

                                 PART II

             PILOT-PLANT AMD PROTOTYPE OPERATING EXPERIENCE

                                   By

      J. J. Schultz, T.  M. Kelso, J. L.  Graham, and J. K. Metcalfe
                    Division of Chemical Development
                       Tennessee Valley Authority
                         Muscle Shoals,  Alabama

                                  And

                              N.  D.  Moore
                          Power Research Staff
                       Tennessee Valley Authority
                         Chattanooga, Tennessee


                                ABSTRACT
          The Tennessee Valley Authority (TVA) has "been engaged in a
limestone - wet-scrubbing pilot-plant test program for the past 2 years.
The purpose of the program is to provide information for the design and
operation of a demonstration-scale (550 mw) scrubber system now under
construction at TVA's Widows Creek power station in northeast Alabama.

          The pilot—plant testing program consisted of two phases.  The
objective of the first phase (l year) was to select a scrubber type for
the Widows Creek project.   Four different types of scrubber systems
were evaluated.   As a result of this evaluation, a venturi (for particu-
late removal) followed by a multigrid scrubber was selected as a basis
for the design of the Widows Creek system.

          The objective of the second phase was to closely duplicate
the Widows Creek design in the pilot plant and evaluate the following.

          • Process equipment and construction materials.

          • Mist eliminator system.

          • Scrubber operation during simulated changes in the
            operation of the boiler.

          • The effect of the scrubber grid configuration and the
            temperature of the scrubbing slurry on S02 removal
            efficiency.
                                     496

-------
          The following conclusions can te drawn from the pilot—plant
studies.

          • Limestone wet scrubbing provides an effective method for
            removing S02 from boiler flue gas.

          • An S02 removal efficiency of JCffr can "be expected using
            the multigrid type of scrubber with a scrubbing liquor
            recirculation rate (L/G) of 50 to 60 gallons per 1000
            cubic feet of gas and a Ca:S02 mole ratio of 1.5.  The
            efficiency can he increased to a"bout 9df> if a two-stage
            mobile—"bed scrubber is used.

          • Techniques for adequate long-term disposal of the spent
            solids from the scrubbing slurry (including fly ash)
            have not "been developed.  Initially, ponding of the spent
            solids will "be used for the Widows Creek system.

          • The long-term operating reliability of the limestone - wet-
            scrubbing system will be largely determined by proper design
            and selection of suitable construction materials to cope
            with the erosive and plugging characteristics inherent in
            slurry scrubbing.
                                     497

-------
                                  THE

                            TVA WIDOWS CREEK

                      LIMESTONE SCRUBBING FACILITY

                                 PART II

             PILOT-PLANT AND PROTOTYPE OPERATING EXPERIENCE

                                   By

      J. J. Schultz, T.  M, Kelso, J. L. Graham, and J. K. Met calfe
                    Division of Chemical Development
                       Tennessee Valley Authority
                         Muscle Shoals, Alabama

                                  And

                              N. D. Moore
                          Power Research Staff
                       Tennessee Valley Authority
                         Chattanooga, Tennessee
Introduction

          TVA has been engaged in a limestone — wet—scrubbing pilot—plant
test program for the past 2 years.  The program is designed to provide
information for the design and operation of the full-scale (550 mw) lime-
stone wet scrubber which is being installed on unit 8 at the Widows Creek
generating station.  The information obtained from the pilots-plant program
includes the following.

          • Selection of scrubber type.

          • Selection of optimum operating conditions.

          • Evaluation of scrubber performance during simulated changes
            in boiler operation;

          • Identification of suitable process equipment and construction
            materials.

          The program was performed in pilot plants located at TVA's
Colbert Power Plant near the fertilizer research and development
facilities at Muscle Shoals, Alabama.

          The first pilot plant, a temporary unit located near the unit
No. 3 boiler, was operated for 1 year before it was dismantled to make
room for the installation of an'electrostatic precipitator.  The evalua-
tion of four basic scrubber types was done in this pilot plant.  The

                                   498

-------
second one is a permanent unit located near the unit 5 boiler,  designed
to closely duplicate the system planned for Widows Creek.   It has been
operated for 1 year and is still being operated.   Except for changes
in the scrubber type, the same general mode of operation was used in
T>oth pilot plants; a discussion of the work performed in both follows.
General Description of Process

          Flue gas for the pilot plant is obtained from a pulverized
coal-fired boiler in which coal containing about 4$ sulfur and 15$ ash
is burned.  The boiler is normally operated with 20$ excess air.   The
flue gas contains about 6$ oxygen (including air heater leakage) and
2800 ppm S02.  It can be withdrawn from the boiler outlet duct on
either side of an electrostatic precipitator; fly ash loadings are
about 4 gr/scf or 0.1 gr/scf.  The temperature of the gas entering the
pilot plant is about 300° F.

          In the pilot plant the flue gas is scrubbed with a limestone
slurry containing 15$ undissolved solids (combined limestone reaction
products and fly ash).  This level of solids is optimum for maintaining
the stability of the slurry pH and S02 removal efficiency.  Spent
scrubbing slurry is continuously bled to a series of settling tanks
where the solids settle.  The settled solids contain from 55 to 65$
water "by weight.  The supernatant liquor from the settling tanks is
returned to the scrubbing loop to maintain the solids content at 15$.
The liquor system is operated on a closed—loop basis; no liquor is
overflowed to waste.  This mode of operation is planned for the
Widows Creek system.  The only liquor lost from the system that must
be replaced with fresh makeup water is (l) the water required to
humidify the hot (300°F) flue gas and (2) the free and chemically
combined water in the settled solids.  This loss is equivalent to about
1 gpm of makeup water in the 1-^aegawatt pilot plant.  With this mode of
operation, 'the liquor phase of the slurry system contains about 10g/l
total dissolved solids and is supersaturated with calcium and sulfur
salts.

          Pulverized limestone (normally 75$ minus 200 mesh) is mixed
with water (part of the makeup) to give a 60$ solids slurry.  Slurry
of this concentration has excellent storage characteristics and requires
only mild agitation to maintain homogeneity.  The fresh slurry is fed
to the scrubbing slurry retention tank at the required rate (usually
1. 5 moles of Ca per mole of S02 in the inlet flue gas).
                                      499

-------
 vtork. Performed in the  First  Pilot
 Plant (Temporary Unit  at  Boiler No. 3)

           The  following types  of scrubbers were tested in the unit 3
 pilot plant.

           • Ventri-Rod1 followed by a packed—bed Crossflow scrubber.

           • Ventri—Rod/open  spray tower.

           • Mobile-bed scrubber (TCA)2.

           • Ventri-Rod followed by a multigrid scrubber.

           The  objective of the evaluation was to determine which system
 tfould give the "best combination of S02 removal efficiency and operational
 reliability.   Pilot—plant operating experience and data for the first
 three scrubber types were presented last year at the Second International
 Lime/Limestone Wet Scrubbing Symposium ("Scrubber-Type Comparison/1 by
 T. M.  Kelso, P.  C. Williamson, and J. J.  Schultz).

           The  present paper will cover the testing and development of
 the Ventri—Rod multigrid scrubber system which ultimately led to its
 selection  for  the commercial-size installation at Widows Creek.

           All  of the scrubbers tested gave acceptable S02 removal
 efficiencies at  approximately the same liquid to gas ratio (L/G,
 gallons/1000 ft3).  The TCA and multigrid scrubbers could be operated
 at higher  gas  velocities than the Crossflow or spray tower (12 ft/a vs.
 8 ft/s).   These  data are shown in Table I.

           The  Crossflow scrubber was abandoned for slurry scrubbing
because of rapid plugging of the packed bed.  The Ventri—Rod spray
 tower, although  nonplugging, did not give acceptable S02 removal
 efficiency at  allowable pressure drop (less than 15 in H20) if the gas
velocity exceeded 5 ft/s.   The lack of sufficient space for fitting to
the Widows Creek unit precluded the use of a' low-velocity scrubber.

           The  TCA and multigrid scrubbers were nonplugging and reliable.
The TCA scrubber (containing two or three 1—foot beds of spheres) had
excellent  gas—liquid contacting characteristics and gave the best S02
removal efficiency (90$).   The major disadvantage of the TCA scrubber
was erosion of the packing spheres (polypropylene).   A 2.% weight loss
of the spheres occurred during 1000 hours of operation and failures
due to puncturing began to occur.   Universal Oil Products (manufacturer
of the TCA scrubber) has a program underway to develop a more erosion-
resistant  sphere material; thermoplastic rubber has shown promise.
  A type of venturi scrubber manufactured by Environeering, Inc.
2 Turbulent Contact Absorber manufactured by Universal Oil Products
  Company.                           500

-------
                                                        TABLE I
in
O
Typical Operating Characteristics

Scrubber system
Ventri-Kod/Crossflow
Ventri— Rod/ spray tower
Three stage mobile bed (TCA)
Ventri-Rod/multigrid
of Scrubber Systems Tested
L/G,
gal/1000 ft3
55
65
55
50
in TVA Pilot
AP,
in HP0
10
15
10
8
Plant
Velocity,
ft/s
5
5
12
12
     S02
   removal
efficiency,

     80


     75

     90


     75

-------
          The multigrid  scrubber, while not as effective in removing
 S02  as the  TCA, has the  advantage of having few  internal parts.  The
 pressure  drop across the multigrid scrubber (containing five 65$ open
 grids spaced on 5—foot intervals) is only about  2 inches of HeO;
 consequently the particulate  (fly ash) removal efficiency  is poor.  To
 remove the  fly ash, a venturi  (Ventri—Rod) was installed upstream of the
 multigrid scrubber.  The Ventri—Rod, operated at an L/G of 10  and a
 pressure  drop of 5 inches of H20, was effective  in removing the fly
 ash  from  the flue gas and also providing a sharp wet-dry junction in
 the  inlet gas duct.
Selection of Scrubber for Widows Creek

          As a result of this evaluation, the venturi/multigrid scrubber
system was selected as the basis for the design of the Widows Creek
system.  A Ventri—Rod type of venturi was used in the pilot plant; however,
a standard venturi with similar operating characteristics may be used in
the Widows Creek installation.  The expected S02 removal efficiency using
the multigrid scrubber is about 75$ (TOO ppm S02 in the gas exhausted to
the atmosphere).  If higher S02 removal efficiency is required, the
scrubber can be converted to a two- or three—stage TCA type.  A new pilot
plant was "built to continue the evaluation of this system.

          It should be noted that a venturi or some other type of scrubber
effective for particulate removal is necessary at Widows Creek because
the present electrostatic precipitator is not very efficient.  The ash
collected in the precipitator is slurried to the ash disposal pond and
that collected in the scrubber goes to the scrubber solids disposal system.
Work Performed in the Second Pilot
Plant (Permanent Unit at Boiler No. 5)

          A perspective view and flow diagram of the unit 5 pilot plant
are shown in Figures 1 and 2.  The gas capacity is equivalent to 1
megawatt (2^00 acfm at 120°F), and the design closely duplicates that
planned for the Widows Creek installation.  The objective of the new
pilot—plant test program was to evaluate the following:

          • Mist eliminator design and operation.

          • The effect of grid configuration and slurry temperature
            on S02 removal efficiency.

          • Scrubber performance during simulated changes in power
            plant operation.

          • Process equipment and construction materials.

                                    502

-------
                                  FIGURE  1




Limestone -- Wet—Scrubbing Pilot Plant  at  Colbert  Power Plant (Permanent Unit)
                                   503

-------
Ul
o
                                        WATER
               FLUE
               GAS
                    I
                    I
         SLURRY
      DISTRIBUTOR


      MULTIGRID
      SCRUBBER


FORCED DRAFT FANS
                       VENTRI-ROD
                         ELEMENT
             PULVERIZED
             LIMESTONE
   SLURRY FROM F-4
 «
                     F-2
               LIMESTONE SLURRY
                  FEED TANK
   F-3
 SLURRY
RETENTION
   TANK
                                                   MIST ELIMINATOR
                                                     TO SEWER
               MAKE-UP WATER
                                                                                SCRUBBED
                                                                                   GAS
                                             DIRECT
                                             FIRED
                                            REHEATER
                                 F-4
                               SLURRY
                              RETENTION
                                 TANK
                              (OPTIONAL)
SETTLING
  TANKS
                                               FIGURE 2
                    Flow Diagram;  Limestone - "We-t>-Scrubbing Pilots-Plant (Permanent Unit)

-------
          A discussion of these items follows.

          Mist Eliminator Design;  Removal of entrained scrubbing slurry,
and consequently particulate, from the gas leaving the scrubber was a
major problem identified during the unit 3 pilot-plant program.  Commer-
cially available mist elimination systems installed in the vertical
scrubber housings were inadequate.  The most severe problem with this
type of system was plugging of the mist eliminator element by accumulation
of solids from the entrained slurry.  The plugging occurred "because there
was not a sufficient amount of fresh water available to continuously wash
the mist eliminator when the scrubber slurry system was operated on a
closed-loop basis.  A total of only 1 gpm of fresh water was required
for makeup in the 1-megawatt pilot plant, whereas most vendors recommend
a continuous wash of about 2 gpm per square foot of mist eliminator face
area.  This is equivalent to about 8 gpm in the 1-megawatt pilot plant,
or 8 times the required makeup rate.  In some tests a blend of 3 parts
clarified liquor from the slurry settling tanks and 1 part fresh water
was used to wash the mist eliminator intermittently (twice an hour) at
the vendor's recommended rate of 2 gpm per ft2.  This method was not
satisfactory because of scaling (calcium sulfite and sulfate) caused
by the supersaturated clarified liquor (Figure 3).

          An effective mist elimination system was ultimately developed
and is now in use.  It consists of a horizontal housing connected to the
outlet of the vertical scrubber tower.  The superficial gas velocity
through the horizontal housing is 9 ft/s.  A V-pass chevron element is
installed in a vertical position in the housing.  The element  (Type 316
stainless steel blades spaced on 1—1/2—in centers) is 12 inches deep and
has a face area of k ft2.   It is washed continously with fresh water
which is discarded to the fly ash disposal pond separate from the lime-
stone slurry disposal system.  Fly ash pond effluent water, used in a
few tests for washing, was as effective as fresh water.  The particulate
loading of the gases exhausted to the atmosphere averaged 0.04 gr/scf.

          Tests were made to determine the minimum effective wash rate
with this configuration.  A wash rate of 0.25 gpm/ft2  (l gpm total) was
effective, but was marginal in regard to uniform coverage of the chevron
element.   At this rate (1 gpm total) the wash water was drained into
the slurry system as makeup.  Maximum effectiveness was obtained with a
wash rate of 1 gpm/ft2 (4 gpm total).  Since this was approximately 4
times the makeup required for the slurry system, the excess water was
routed to the fly ash disposal pond rather than to the limestone slurry
disposal system.  The composition of the discarded wash water was
monitored in the pilot plant at various wash rates to determine the
environmental impact caused by this mode of operation  (Table II).  Based
on these data a wash rate of 1 gpm/ft2 is planned for the Widows Creek
installation.  The excess wash water not required for makeup will be
discarded to the ash disposal pond.
                                     505

-------
-
-
                                                                       ,HARD SCALE
                                                                          SOFT  SOLIDS
                                                FIGURE 3



                               Mist Eliminator After 270 Hours of Operation



    (Washed Intermittently from Both Sides wifrh a Blend of Clarified Settling Tank Liquor and Fresh Water)

-------
                                                        TABLE n

                      Composition of Effluent Water from Once-Through Mist Eliminator Wash System*

                                          Limestone - Wet-Scrubbing Pilot Plant
                                                  Composition,  mg/1 at the following wash rates
en
o
••J
Wash rate, ^pm/ft2
Component
Total dissolved solids
Aluminum
Ammonia nitrogen
Arsenic
Barium
Beryllium
Cadmium
Calcium
Chloride
Chromium.
Copper
Cyanide
Iron
Lead
Magnesium
Manganese
Mercury
Nickel
Phosphate
Potassium
Selenium
Silver
Sodium
Sulfatf
Zinc
PH
1.0
1000
<0.2
0.21
0.002
<0.1
<0.01
0.0042
220
2k
<0.05
0.02
<0.01
5.8
0.033
6.5
0.16
<0.0002
<0.05
0.11
2.2
0.012
<0.01
8.1
TOO
0.0?
3.1
0.50
1900
<0.2
0.25
0.002
<0.1
<0.01
0.0013
440
hO
<0.05
0.02
<0.01
0.07
0.011
8.2
0.14
<0.0002
<0.05
0.03
3-0
0.024
<0.01
8.8
1000
0.02
7.2
0.25
2200
<0.2
0.34
0.010
<0,1
<0. 01
0.0031
430
120
<0.05
0.03
<0.01
5.5
0.016
18
0.37
<0. 0002
<0.05
0.03
2.6
—
<0.01
11
1200
0.14
2.7

0.25b
1900
0.5
1.9
0.001
<0.1
<0.01
0.0015
370
70

-------
           Grid Configuration;   The S02 removal efficiency obtained in the
 rrrultigrid scrubber is  dependent upon the gas-liquid contact  imparted by
 the grids.   Tests were made to determine the effect on S02 removal effi-
 ciency and pressure drop caused by varying the number and open area of
 the grids.   When the scrubber  was  operated with 5 wire mesh—type grids
 (75% open area,  1/8-inch—diameter  wire with 7/8—inch—square  openings)
 at  an L/G of 50,  the S02 removal efficiency and pressure drop averaged
 70# and 1. 7 inches of  H20, respectively.   Only 2 to 3 percentage points
 increase in  S02 removal efficiency was obtained when the number of
 grids was increased to 10.   The pressure drop increased only slightly
 (less than 0. 5 inch H20).   The relationship of the number of grids to
 S02 removal efficiency is shown in Figure k.

          Another series  of tests  was made to determine the  effect of
 less  grid open area and consequently more contact  surface on  the pressure
 drop  and S02 removal efficiency.   Combinations of five  and six stainless
 steel wire mesh grids  containing 40 and 6Cff> open area were tested.   The
 hOf> open grids were made  of 0.075-inch-diameter wire with 1/8-inch-square
 openings.  The 6Of> open grids  were made of 0.09*1—inch-diameter wire with
 1/2—  by 5/l6-inch openings.  The wire  used in these grids was smaller
 and more closely  spaced than that  previously used to provide  more surface
 (target)  area for gas  and slurry contact.   In some tests,  additional
 grid  surface area was  obtained (with respect  to the downward  flow of
 slurry) "by installing  a kCff> open grid  on a 65-degree angle "between two
 horizontal grids.   A summary of the grid combinations tested  and
 operating data obtained are shown  in Table III.

          There was  no significant improvement (compared with previous
 tests with 75% open area grids)  in the S02 removal efficiency or
pressure  drop  across the  scrubber  when the open area was reduced to
 60$.   The removal efficiency and pressure drop increased to 78$ and
3.7 inches of  H20,  respectively, when  the bottom four 6of> open grids
were replaced with 40# open area grids.   The  S02 removal efficiency
was further  increased  to Q2f> by increasing the L/G from 50 to 60.   The
 addition of  a  sixth grid  (^C#  open,  slanted between the bottom two
grids) did not cause any significant change.   Grids with kQJ>  open area,
although better for  S02 removal, are not  practical because of the
tendency  of the grid to collect  trash  that  would normally pass through
a grid having 6Cff>  or more open area.

          On the basis of this evaluation,  five  grids containing
approximately 65%  open area will be used  in the  Widows  Creek  installa-
tion.  The construction of the grids may be similar to  floor  grating
 (5/8- by 3-3/l|-inch-rectangular openings) made of  reinforced plastic or
Type 316  stainless steel.  The pressure drop  and S02 removal  efficiency
obtained in the pilot plant with this  type  of grid construction was
essentially the same as that obtained  with  the standard wire mesh grids.
                                      508

-------
fc
Ul
o
u.
CO jjr
^^^
fT -J
0 §
^1. '^^^
o
llJ 2
O UJ
< o:
z ^
o iii O
*° O ^
Ul ?
CL
Ul
CO
Ul
££
0

«t

3
2

1

n c
1 1 I 1 I '
rv/^

: xr :

>^ S02 REMOVAL EFFICIENCY
.S WAS 70% WITH FIVE GRIDS
S 1 1 1 1 f
      67          8         9         IO
      NUMBER  OF  GRIDS (75 7o OPEN AREA)
                     FIGURE 4
Effect of the Mumber of Grids on SO^ Removal Efficiency
         Limestone - Wet-Scrubbing Pilot Plant
fl

-------
                                                 TABLE III
(Jl

o
Evaluation of Grids in the
Multigrid Scrubber*
Limestone — Wet-Scrubbing Pilot Plant
Grids

Distribution13
Test No.
1
2

2

k


5


Total
5
5

5

6


6


No.
5
1
^
1

1
k
1
l
h
1
Open area, $
60
60
to
60

60
to
to (Slanted)
60
to
to (Slanted)
L/G
50

50

60


50


60

SO g removal
APj in H^O efficiency, #
1-7 70

3-7 78

k.Q 82


4.1 77


4.U 84
     * All tests made at  12  ft/s  superficial gas velocity.
     ° Grid distribution  listed in order  starting from top  of scrubber.

-------
          Slurry Temperature;  During the course of a year's operation
of the pilot plant, the S02 removal efficiency varied as much as 10$
(Figure 5).  This change in efficiency is largely attributed to the
change in temperature of the scrubbing slurry (about 25°F, from 100°F
in winter to 125°F in summer) caused by ambient temperature fluctuations.
The lower vapor pressure of S02 over the scrubbing slurry at the lower
temperatures presumably gave better S02 removal during cool weather.

          The feasibility of removing heat from the scrubbing slurry
to obtain satisfactory S02 removal efficiency was tested in the
pilot plant.  A Teflon-tube heat exchanger manufactured by DuPont was
submerged in the multigrid slurry retention tank and river water (57°F)
passed through the tubes at a rate of 10 gpm.   The heat exchanger was
effective in reducing the temperature of the slurry from the normal
120° to 125°F and maintaining it at 100° F.  About 5056 of the heat load
for the cooling was accounted for in the cooling water discharged from
the heat exchanger; the remainder was assumed to be lost by convection.
An overall heat transfer coefficient (U) of 25 was obtained.  This method
of increasing the S02 removal efficiency is not practical because of
the large heat transfer area and high water rate required, which would
be even higher in a larger unit because a smaller percentage of the
cooling would be accomplished by convection.  Also, the temperature of
the cooling water obtained from the river in summer would be about 80°F,
reducing the AT significantly.   An equivalent increase (L/G of about 20)
in the pumping rate of uncooled slurry to the scrubber would increase
the S02 removal efficiency about as much as cooling the slurry to 100° F.

          Power Plant Simulation Tests;  A series of tests was made in
the pilot plant to determine whether the limestone — wet-scrubbing process
could te operated and controlled over a range of simulated power plant
operation conditions.   The flow rate and S02 concentration of the gas,
the recirculation rate of the scrubbing slurry, and the Ca:S02 mole
ratio were varied to simulate scheduled and unscheduled changes in boiler
operation.   A diagram of the pilot-plant configuration used for this
series of* tests is shown in Figure 6.

          Six tests were run consecutively, all on a closed-loop "basis.
The retention time and the solids content of the scrubbing slurry were
9 minutes (based on a pumping rate of 105 gpm to the multigrid scrubber)
and ll$, respectively.  The S02 removal efficiency ranged from ¥$ to
82#; 44$ when the Ca:S02 mole ratio was reduced to 0.8, and Q2f> when
the gas velocity through the multigrid scrubber was reduced to 8 ft/s
and the Ca:S02 mole ratio and L/G were 1.5 and 77 > respectively.
Operation of the pilot plant was routine and no difficulty was
encountered in maintaining control of the process.  The S02 removal
                                   511

-------
     IOO
      90
   o
   z
   y  eo
   o

   ti-
   ll.
   UJ
in
—«  «s

KJ  UJ
   (VI

   O
   o  70
      60
     50
        10
                                     L/G = 5O AT A GAS  VELOCITY OF  8 FT. /SEC
                                       JL
15
20
25
30
35
                              SLURRY  RATE TO SCRUBBER,GPM/FT.2

                                 OF SCRUBBER CROSS SECTION


                                          FIGURE 5


                        Effect of Slurry Temperature on SOp Removal Efficiency
4C

-------
                                                                                              SCRUBBED
                                                                                                CAS
                                  WATCH
                                                            MtST ELIMINATOR
                                    SLUKRY
                                  DISTRIBUTOR"
                                    MUtTl«fll|>   ~
                                    SCRUBBER\
                                    (BORIDSt   1--
Ul
«>
CO
                          FORCED DRAFT FANS
(FROM CYLINDERS)
                            VENTRI-ROO
                              ELEMENT
               PULVERIZED
               UME3TOHC
   SLURRY FROM  F-4
                                                              'TO SEWER
                                                              MAKE-UP MMTER
                         r-z
                  UMESTOHE SLURRY
                     FCfO TAHK
                                       F-3                F-4         SETTLING
                                  SLURRY  RETENTION   SLURRY RETENTION   TANKS
                                       TANK                TANK
                                       C9MIMUTC RETENTION TIME 1

                                                FIGURE 6
                                                                                        DIRECT FIRED
                                                                                         REHEATER
  Flow Diagram:  Limestone — Wet-Scrubbing Pilot—Plant  Configuration Used During Power Plant Simulation Tests

-------
efficiency fluctuated with changes in L/G and in the gas flow (simulating
a 15 and 35$ change in load).  The pH of the scrubbing slurry and S02
removal efficiency declined rapidly when the flow of fresh limestone was
stopped to simulate a pump failure.  However, operation returned to
normal within 4 hours after the flow of fresh limestone was restored.  A
summary of the test conditions, S02 removal efficiency, and pH of the
scrubbing slurry for each test is shown in Figures 7 through 12.

          Slurry Settling and Compaction;  The slurry disposal pond
planned for the Widows Creek installation represents a significant
portion of the project cost.  The initial size and estimated life
planned for the pond depend upon the settling and compaction characteristics
of the spent slurry.   In the pilot-plant tests, initial settling occurred
rapidly (within 2k hours) to produce a clear liquor, but little further
settling took place even over extended periods of time.  A summary of the
long—term settling data is shown in Table IV.  The poor settling and
compaction are attributed to the presence of small flat calcium sulfite
crystals which normally constitute about 90$ of the calcium—sulfur
reaction products.   Only about 10$ of the reacted sulfur forms gypsum,
which would be a more desirable form "because the crystals are large and
settle well.

          Attempts were made to convert (oxidize) the calcium sulfite
to calcium sulfate (gypsum) by treating the spent slurry with excess
oxygen (10 times stoichiometric).   These tests were unsuccessful.   In
other tests oxidation of the slurry was promoted when the flue gas was
mixed with ambient air to give a 1:1 volume ratio "before the scrubber.
It would not "be practical to do this in a commercial—size unit because
of the large volume of air required, but the method was used to determine
the effect of increased oxidation on the settling and compaction charac-
teristics of the slurry.   The flue gas—air blend contained iy/o oxygen
and about 1100 ppm S02.   In one short test oxidation was increased to
the extent that 90$ of the reacted sulfur was in the form, of gypsum.  No
quantitative settling data were obtained during this test, but compaction
of the solids appeared much improved over that of less oxidized slurry.
In a longer test (9 days) only 53$ of the reacted sulfur formed gypsum.
There was no significant improvement in the short— and long—term settling
characteristics of this slurry (53$ oxidized).   These data are shown in
Figure 13 and Table V.

          Optimum Operating Conditions:  Various modes of operation
and test conditions were evaluated during the pilot—plant program to
determine the best combination for S02 removal and ease and reliability .
of operation.   The following conditions are considered optimum and are
being used for the planning and design of the Widows Creek system.

          • Pressure drop across the venturi (or Ventri—Rod)
            for fly ash removal	  5 in H20

          • L/G to venturi (slurry) 	  10 gal/1000 ft3

                                    514

-------
  O >
  S O
  UJ Z
  a: uj
  Sfc
    UJ
tn
in
       80
75
    I
    a.
     BOILER  SIMULATION TEST NO. I
     UNSCHEDULED  SHORT TERM DECREASE OF LOAD
                    BASE CONDITIONS
                                              GAS  FLOW
                                         r   REDUCED 15%
                                                                  -H*
                           BASE  CONDITIONS
              GAS  FLOW. 2100ACFM AT  I20°F
        Ca/SO2 MOLE  RATIO:
              L/G  ro
              L/G TO MULTIGRID;50
         L/G TO  V-R;I2
         L/G TO  M-G:59
SLURRY TO  SCRUBBER
                                     F-3 SLURRY FROM SCRUBBER
                            23456
                                             HOURS
                                            FIGURE 7
                        The Effect of a 15% Reduction in the Gas Flow on the S0g
                                                                               8
                            Removal Efficiency and pH of the Scrubbing Slurry

-------
 o
'£
 uj
     BOILER  SIMULATION TEST  N0.2
     UNSCHEDULED  SHORT TERM  INCREASE  IN  LOAD
75
   -BASE CONDITIONS
                          GAS FLOW INCREASED  15%
BASE  CONDITIONS -
TO
 6
- GAS FLOW:2100 ACFM
—Co/S02'. 15
. L/G TO V-RMO
  L/G TO M-G.'SO
                                 Ca/S02:1.3
                          F-4  SLURRY TO SCRUBBER
                          F-3  SLURRY FROM SCRUBBER
                                          I
                      23456

                                       HOURS

                                      FIGURE 8

                The Effect of a iyf> Increase in the Gas Flow on the S0g Bemoval

                         Efficiency and pH of the Scrubbing Slurry
                                                                            8

-------
         o >
         SE "
         ij Z
         
-------
                 BOILER  SIMULATION  TEST N0.4
00
S02 REMOVAL
EFFICIENCY, %




Q.

75
70


6

5

SCHEDULED SHORT TERM DECREASE IN LOAD
(FLOWS WERE ADJUSTED TO MAINTAIN A Ca/SOo MOLE RATIO OF 1.5 AND
AN L/G OF 10 AND 50 TO THE V-R AND M-G) 	
^S**-r^ ,_ r— X'^ BASE ~"
GAS FLOW REDUCED 35% , CONDITIONS
	 ^r *H ^1
^ BASE CONDITIONS
F-4 SLURRY TO SCRUBBER
— —

1 ii i i i 1 i 1
123456789 10
HOURS
FIGURE 10
                              The_Effe_ct__of a 35% Decrease in the Gas Flow on the S0g




                                 Removal Efficiency and pH of the Scrubbing Slurry

-------
§".
O  >
2  u
(O
 en

 to
   -
   u.
   U.
   UJ
80



70



60



50



40
           BOILER SIMULATION  TEST NO.5

           UNSCHEDULED AND  SCHEDULED  INCREASES  IN  SO2  LOAD  TO SCRUBBER
-BASE CONDITIONS


          4000 PPM  SO2 Ca/S02'-l.5
                    -7IOO PPM  S02, Ca/S02: O.8

              •BASE CONDITIONS
                          F-4 SLURRY TO SCRUBBER
                         F-3 SLURRY FROM SCRUBBER
                                                                              BASE  CONDITIONS
                                                                       2100 ACFM
                                                                       2800 PPM S02

                                                                       Ca/S02:i.50

                                                                       L/G TO V-R:IO
                                                                       L/G TO M-G:50
                        8      12     16     20      24

                                                 HOURS

                                            FIGURE 11
                                                            28
                                                             32
                                                 36
40
44
                     The Effect of Q02 Concentration and Ca:SOg Mole Ratio on the


                           Removal Efficiency and pH of the Scrubbing Slurry

-------
     o >
     2 £
     UJ Z
     0* UJ
U1
KJ
O
          85
          SO
          75
          7O
       x
       ex
  BOILER SIMULATION TEST  NO.6
  SCHEDULED DECREASES IN LOAD (OVERNIGHT) USING A CONSTANT
  PUMPING RATE TO  SCRUBBER

                                             GAS FLOW REDUCED 35%
                                             L/G TO V-RM5
                                             L/G TO M-G=77
                        BASE CONDITIONS
DAYTIME
GAS FLOW.-2IOOACFM
L/G TO V-R:IO
L/G TO M-G'50
                                                          OVERNIGHT
                                    Ca/S02 MOLE  RATIO: 1.5
                                  F-4 SLURRY TO SCRUBBER
                                 F-3 SLURRY  FROM SCRUBBER
                      16
                  32
                48
  64
HOURS
SO
96
112
128
                                           FIGURE 12

           The Effect of Reducing the Gas Flow 35% and Maintaining a Constant Pumping Rate

            to the Scrubber on the S03 Removal Efficiency and pH of the Scrubbing Slurry

-------
                                                                  TABLE IV

                                                   Long-Term Settling and Compaction Data

                                                   Limestone - Wet-Scrubbing Pilot Plant



                                                                                                                           Percent
                                                                                                                      water in settled
        Limestone  Percent solids	Volume percent after	solids after
         source  d  in slurry  at   	gO  days	  	 60 days        160 days        240 days        300 day?              Days
  Test  and grind  start of test   Solids Liquid  Solids  Liquid  Solids  Liquid  Solids  Liquid  Solids  Liquid  |o55160  240300

   lb       A           13.7         33      67      32      68      -       -       32      68      32      68    67  66   -    66   66
    b
in
to
2
c
Ub
B
C
D
12.9
ik.O
15.0
31
-
29
69
-
U
30
2k
—
70
76
—
27
23
—
73
77
— —
27
23
29
73
77
71
66
-
60
66
57
60
63 -
57
— —
63
57
60
   a
     Includes fly ash removed from the flue gas.
     Flue gas contained a fly asli loading of 2 to 5  grains per  standard cubic foot.
   c Flue gas contained a fly ash loading of 0.1 to  0.5  grain per standard cubic foot.
   d A = The Stoneroan, Inc. (Tiftonia), Chattanooga, Tennessee,  Jk% minus  200 mesh.
     B = The Stoneman, Inc. (Tiftonia), Chattanooga, Tennessee,  8856 minus  325 mesh.
     C = Longview Lime Company, Birmingham, Alabama, 75$ minus'  200 mesh.
     D = Fredonia Valley Ojiarries, Inc., Fredonia, Kentucky,  795& minus  200 mesh.

-------
NJ
KJ
te iv/v/
uj
l-
u.
**
oc
o
•=>
3 75
||SO
CO tig.
i_ i^
ujw
o
£ 25
0.
Ld

O
> 0
1 I I

i I 1 1 1

TESTS MADE USING A 1.50 Co/S02 MOLE RATIO,
LONGVJEW LIMESTONE (75% -200 MESH), AND
15% UNDISSOLVED SOLIDS

—
o— ^^-ltl--^.^><
r"
IN THE SCRUBBING SLURRY
Jto>*
^j^^ 	 -f^" ^^
^N> 	 °^
O USING NORMAL FLUE GAS,
260O PPM S02 AND 5%
10% OXIDATION

_
• USING 5O/50 BLEND OF
AMBIENT AIR, 1120 PPM
53% OXIDATION
1 1 1

1 2 3
02


—
FLUE GAS AND
S02 AND 13% 02

: 1 I II

456789
                                      DAYS  OF PILOT PLANT OPERATION

                                                FIGURE 13


                    Effect of Increased Oxidation on Settling of Solids  from Scrubbing Slurry

-------
                                                      TABLE V
Effect of Oxidation on the Settling and Compaction Characteristics
of Spent Scrubbing Slurry — Limestone - Wet-Scrubbing Pilot Plant
Ul
NJ
u>
Test
1
(I0ff> oxidation)
2
(53% oxidation)
Percent solids
in spent slurry
at start of test
15.0
16.7
Percent water
in settled
Volume percent after solids after
kO days 120 days Days
Solids Liquid Solids Liquic- 40 120
23.5 76.5 23.1 76.9 57 57
27.0 73.0 26.0 74.0 53 53
NOTE:  The percent oxidation is defined as the percent of the total reacted sulfur in the slurry existing as calcium
       sulfate (gypsum).

-------
          • Gas velocity through multigrid scrubber
            (Superficial, based on outlet conditions)

          • L/G to multigrid scrubber (slurry)

          • Number and spacing of grids in multigrid
            scrubber

          • Open area of grids

          • Solids content of scrubbing slurry (total
            undissolved, including fly ash)

          • Retention time in delay tank of slurry fed
            to multigrid scrubber (based on pumping rate)

          • Retention time in delay tank of slurry fed
            to venturi

          • Ca:S02 mole ratio (based on inlet S02
            concentration)-

          • Limestone particle size

          • Solids content of fresh limestone slurry

          • Mist eliminator wash rate (fresh water on a
            once—through basis)
12 ft/s

50-60 gal/1000 ft3


5 on 5— ft centers

60-65*


1936


7 minutes


Not critical
1.5
    -£00 mesh
  gpm/ft2
Operation of Prototype Scrubber

          Tests were made in one of the three EPA prototype (10 mw) scrubbers
(designed "by Bechtel Corporation) located at TW's Shawnee Power Plant near
Paducah, Kentucky.  Over a period of 2 months, tests were made which were
designed to simulate previous operation in the 1-megawatt pilot plant at
Colbert and to provide scale-up and design information for the Widows Creek
installation.   The TCA scrubber was used; the packing spheres were removed
and 5 grids were installed at approximately 1»—foot intervals.   The slurry
system was modified to closely duplicate the closed—loop mode of operation
used in the pilot plant.   A diagram of the prototype plant is shown in
Figure 14.  Initial operation was suspended after kQ hours of operation
because the chevron mist eliminator, located in the vertical scrubber
housing, became plugged with solids from entrained scrubbing slurry.  This
problem was anticipated based on pilot—plant experience.

          The design or the prototype system precluded the installation
of a horizontal mist eliminator housing at the outlet of the scrubber
to duplicate the system developed in the pilot plant.  Instead, a multiple
venturi tray device (FlexiTray, manufactured by Koch Engineering Co., Inc.)
was installed in the vertical scrubber tower between the slurry distributor
and mist eliminator.   The FlexiTray was designed to provide an interface
between the closed—loop scrubbing slurry system and a once—through wash
                                     524

-------
10
ui
                                                DIRECT FJREO REHEATER
                                                                SCRUBBED GAS
                                                CHEVRON MIST ELIMINATOR
            FREOON1A
            LIMESTONE
                          MULT 16 RID
                          SCRUBBER
                        (MODIFIED TCA)
                                                                                                UNDERFLOW TO
                                                                                               DISPOSAL POND
                                       SLURRY
                                      RETENTION
                                        TANK
CLARIRED
  LIQUOR
HOLD TANK
THICKENER
                                                        FIGURE
                          Flow Diagram - Modified TCA Prototype Scrubber at Shawnee Paver Plant

-------
system for the mist eliminator.  Fresh water, fed across the top of tne
FlexiTray at a rate of about 60 gpm, was effective in keeping the top
of the tray and the mist eliminator clean; the water was discarded.
Blasting the "bottom of the tray with steam (150 psig) once per shift
for a"bout 1 minute was effective in preventing accumulation of solids
in this region.  A diagram of the prototype scrubber and FlexiTray
installation is shown in Figure 15.

          When the scrubber was operated at a gas velocity of 12 ft/s
(design rate for Widows Creek at full load), carryover of scrubbing slurry
into the open-loop FlexiTray wash system was excessive (30 gpm).  Continuous
operation could not 'be maintained at the high gas flow rate because (l) the
maximum slurry purge rate to the disposal pond during closed—loop operation
is only about 1 gpm per megawatt (10 gpm total) and (2) slurry lost to
the FlexiTray wash system was routed to the fly ash disposal pond and could
not be returned to the closed-loop slurry system.  Excessive carryover of
scrubbing slurry did not occur in the pilot plant when the scrubber was
operated at a gas velocity of 12 to 13 ft/s.  These data are shown in
Figure 16.  The reason for the excessive entrainment of slurry in the
prototype scrubber is not known but may "be caused by the smaller ratio
of scrubber wall area to gas volume as the size of the scrubber increases.
The relationship between velocity and entrainment for the larger Widows
Creek scru'bber is not predictable.

          Carryover of scrubbing slurry into the FlexiTray wash system
was negligible when the gas velocity was reduced to 8 ft/s (two-thirds
of Widows Creek design).   This velocity was selected for a long—term
test designed to compare the operating characteristics of the prototype
system with the Colbert pilot plant.  Scrubbing slurry containing 15$
undissolved solids was fed to the multigrid scrubber at an L/G of 50.
The hot flue gas was humidified with slurry fed to the inlet duct at a
rate of JO to 50 gpm.  The retention time of the scrubbing slurry in the
delay tank was 7 minutes based on the pumping rate to the multigrid.
The tap of the FlexiTray was irrigated with fresh water on a once—through
basis at a rate of 45 to 60 gpm.  The 'bottom of the tray was blasted with
steam (150 psig) once per shift.

          Crushed limestone (1-1/2—inch) was supplied by Predonia Valley
Quarries, Inc., Fredonia, Kentucky, and pulverized by TVA at the Shawnee
site.   The limestone is relatively soft and easily pulverized to produce
a large fraction (95$) of minus 325-mesh material, of which about 50$ -is
minus 8 microns.  The pulverized limestone was fed as a 60$ slurry into the
scrubber system at a rate intended to maintain a Ca:S02 mole ratio of
1. 5 based on the inlet S02 concentration; however, chemical analysis of
the solids from the spent scrubbing slurry taken near the end of the
test indicated a Ca:S02 mole ratio equivalent to 3.0.  The magnetic
flowmeter used to measure the fresh limestone feed rate was calibrated
at the beginning and end of the test, and the indicated flow rate at the
end of the test was only about one—half of the actual rate.
                                      526

-------
                                          DIRECT FIRED REHEATER
ui
K>
                  FLEXITRAY
            FLEXITRAY WASH
            WATER TO SEWER


              MULT 16 RID
              SCRUBBER
FREOON.A    --^   DISPOSAL POND
                                SLURRY
                               RETENTION
                                 TANK
CLARIRED
  LIQUOR
HOLD TANK
                                                                     THICKENER
                                             FIGURE 15


                    Flow Diagram — Modified TCA Prototype Scrubber with FlexiTray

-------
Q_
O
or
o
en
o
UJ
UJ
9.0

7.0


5.0

4.0

3.0


2.0
 1.0


0.5
                           PILOT PLANT
                               I-MW
       PROTOTYPE
          PLANT
         10-MW
     01	
       0
        5          10         15
        VELOCITY, FT./SECOND


             FIGURE 16

Effect of Gas Velocity on Entrainment

        of Scrubbing Slurry
                                              20
                         528

-------
          The S02 removal efficiency gradually increased from about
70$ in the beginning of the test to 80$ at the end.   The pressure drop
across the system (including the wet-dry junction, scrubber, FlexiTray,
and mist eliminator) increased from 3.5 to 6.0 inches of H20 during
the course of the test (19 days).  The undissolved and dissolved solids
contents of the scrubbing slurry -were steady at about 14 and 0.5$>
respectively.  The pH of the slurry fed to the scrubber ranged from
5.8 to 6.0.  The trends in S02 removal efficiency, pH and solids content
of the slurry, and pressure drop are shown in Figure 17.

          From 10 to 2056 (5-10 gpm) of the FlexiTray wash water fell
through the tray into the multigrid scrubber.  The fall-through water
exceeded the closed—loop water "balance based on a water content of 50$
in the discarded spent slurry.  The extra water (5-10 gpm) was removed
from the system with the dilute thickener underflow (about 25$ solids)
which was discarded to the disposal pond.  No clarified liquor was
returned to the scrubbing system from the disposal pond.

          Operation of the scrubbing system was routine and control of
the system was maintained without difficulty.  However, the performance
of the thickener was marginal, yielding a more dilute underflow than
expected.  The dilute underflow provided a convenient means of purging
the excess water from the system caused by leaking of the FlexiTray
wash water into the slurry system.

          After the long-term test was completed, and before the plant
was shut down, a series of short>-term screening tests was made designed
to determine the effects of L/G, gas velocity, and Ca:S02 mole ratio on
the S02 removal efficiency.  These data are shown in Table VT.  In all
tests, the S02 removal efficiency increased with increased L/G and with
increased gas velocity (over the range of 8 to 10.4 ft/s).  The effect
of the Ca:S02 mole ratio on the S02 removal efficiency  could not be
identified.  This is attributed to the high mole ratio  due to the error
in the flowmeter.  These data agree closely with data obtained in the
pilot plant while operating under similar conditions (except for the
high Ca:S02 mole ratio).

          Operation of the prototype plant was discontinued after
completing the screening tests.  The plant had Tseen operated continuously
for 27 days.  Following the shutdown the plant was inspected.  Heavy
deposits of scale (mostly calcium sulfite and sulfate)  had accumulated
on the grids and walls of the scrubber.  A discussion of the inspection
follows.

          Wet-Dry Junction;  The flue gas entered the scrubber through
a UO-inch-square horizontal duct.  Scrubbing slurry (about 30-50 gpm)
was sprayed into the duct (90 degrees to the gas flow)  h feet from the
entrance to the scrubber.  The slurry spray was effective in reducing
the temperature of the flue gas from 300°F to about 125° to 150° F before
entering the rubber-lined scrubber, thus affording protection to the
liner.  About 70$ of the cross-section axea of the duct was plugged with
an accumulation of fly ash and solids from the scrubbing slurry.  The
accumulation extended about 2 feet upstream from the point of slurry
                                   529

-------
U Ul

e 5

 «u-
o u.
to w
90




80




TO




60
                         i	1	1	1	i
             I     234     5     6    789     IO    II    12    13    14    15    16    17     18    13

                                                           DAYS



                                                     FIGURE I?




               Prototype Multigrid Scrubber—Trends  in S0g Removal, pH and Solids Content  of  the




                                Scrubbing  Slurry, and Pressure Drop Across System

-------
                                                                            TABLE VI
Screening Tests, Prototype Limestone - Wet-Scrubbing Pilot Plant at Snavnee Stean Plant
Flue gas
(Maltigrid Scrubber
Veloc ity
Flov rate, through AP
Test Duration, acfm at scrubber, Ca:SOj> nole ratio sc:
So. hr 120 °F ft/s





Ul
LO


WC-4A
HC-5A
WC-6A
WC-7A
WC-BA
WC-9A
HC-10
WC-U
WC-12
8
4J2
8
3
8
4
4
24
64
a Combined pressure
14,720
14,720
14,720
14,720
19,170
19,170
19,170
14,720
14,720
drop across
8.0
8.0
8.0
8.0
10.4
10.4
10.4
8,0
8.0
wet-dry
Indicated
1.5
1-5
1-5
1-5
1.5
1.5
1.5
1.2
1.2
Junction and
o i o*.+<*4 >w m
Actual i:
3-0
3.0 1
3-0
3-0
3-0
3-0
3-0.
2.4
2.4
grid section

across
rubber, Ito
a Hj£a i
3.8
.4-3.5
3-0
2.7
5-6
6-3
6-5
4.4
3-9
of scrubber

vith Five Grids)
PlexiTrav
Once-through wash vater
tray, From tr&yc
SPP
60
60
60
60
60
60
60
60
60
.
gpn Solids, $>
48 0.10
56 0.10
48 0.10
46
85 3-6
85
86
50 0.7
45 1.1

Fall through, £ P
12 2.5
4 2.4
12 2.4
14 2.4
3.4
3-5
3-4
10 2.6
15 2-6

Scrubbing slurry
L/C
70
50
35
25
25
35
50
50
25

5-9
5-9
5-9
6.0
5-8
5.8
5-9
5-9
5-9

Temp,
•F
99
-
-
99
97
94
105
103
110

Solids,
14
14
14
15
14
14
14
13
14

soa
removal,
88
75
76
72
79
86
90
86
75

  Conbined pressure drop across FlexiTray and chevron mist eliminator.



DOTE:  Slurry retention time v&3 7.2 minutes, based on lU,720 aefn at 120*F and an L/G of 50.

-------
 injection.  The soot blower  (operated once per shift with air at 90 psig)
 was not effective in keeping this area clean.  No accumulation of solids
 occurred  downstream from the point of slurry injection.

          Grids:  The scrubber contained five Type Jl6L stainless steel
 grid assemblies spaced at approximately 4—foot intervals.  The open area of
 the wire  mesh  (1—1/8—inch—square openings with 0. 148-inch—diameter wire) was
 75%f but  the effective open area of the grid assembly was only about 50$
because of the grid support and retaining structures.

          A coating of hard scale formed on all of the grids and about 40
to 50$ of the open area was obstructed.   The scale deposits were identified
 as predominately calcium sulfite, although small amounts of calcium sulfate
 and calcium carbonate were present.

          The bottom grid was scaled the most.   The scale was about 1/8-inch
thick on  the grid wires and retaining assemblies.  Nodules of scale (about
3/4—inch  diameter) had formed on the top and bottom sides of the grid where
the wires intersected.   The scale formation was most severe in the center
section of the grid; an area of about 1 square foot was completely plugged.

          The deposit on the intermediate grids was uniform and about 1/16-inch
thick.   The top grid was partially plugged with a mixture of scale and soft
solids from the scrubbing slurry.

          Scrubber Walls:  The rubber-lined walls of the scrubber were coated
with hard scale ranging from 1/8— to 1—inch thick.  Petrographic examination
indicated alternate layers of calcium sulfite and calcium sulfate of varying
thickness.  Approximately equal portions of calcium sulfite and calcium
sulfate were present according to chemical analysis.

          In addition to the general scaling, an accumulation of solids
resembling a stalactite occurred on one wall between the top two grids.  The
formation, about 12 inches thick and J5 feet long, was a soft mixture of scale
and solids from the scrubbing slurry.
                                 •
          FlexiTray and Mist Eliminator;  Solids from entrained scrubbing
slurry had accumulated on the bottom of the FlexiTray to a thickness of
2 to 4 inches, but the openings in the tray were not plugged.

          The chevron mist eliminator was generally clean except for the
portion located directly above the FlexiTray overflow and drain trough.
The pressure drop across the FlexiTray and mist eliminator assembly
increased from 2.J to 2.6 inches during the period of operation (27 days).

Pilot—Plant Simulation of Prototype Operation

          Scaling of the scrubber did not occur during prolonged operation
(8000 hours) of the Colbert pilot plant.  The reason for the heavy accumu-
lation of scale in the prototype scrubber is. unknown, but it may have been
caused by the following:

          • Inadequate humidification of the hot flue gas "before it
            entered the scrubber.
                                   532

-------
          • Poor distribution of the gas and scrubbing liquor in the
            scrubber.

          • Pine particle size of the Fredonia limestone.

          • High Ca:S02 mole ratio (1.5 to J. 0) caused "by inaccurate
            metering of the fresh limestone feed.

          A test was made in the pilot plant to determine if the
conditions used in the prototype plant would also cause scaling in the
pilot plant.

          The pilot plant was arranged to closely duplicate the equipment
configuration and operating conditions used in the prototype.   The velocity
of the gas through the scrubber was maintained at 8 ft/a.   Scrubbing slurry
containing 15$ solids was fed to the scrubber at an L/G of 50.   The reten-
tion time of the scrubbing slurry was 7 minutes based on the pumping rate
to the scrubber.  Finely ground Fredonia limestone was obtained from  the
prototype grinding and storage facility.  The Ca:S02 mole ratio was main-
tained at 2.5 to duplicate the average limestone feed rate used during
the test in the prototype system.  Makeup water was added to simulate the
partial open—loop operation of the prototype system caused by the dilute
thickener underflow.

          Operation was routine.   The S02 removal efficiency averaged
68$ compared to 75$ in the prototype.  A comparison of the trends in
the S02 removal efficiency and pH and solids content of the scrubbing
slurry for the prototype and pilot plants is shown in Figure 18.

          The test was terminated after 1J days of continuous operation.
The grids and walls of the scrubber were coated with heavy deposits of
scale.  The extent and composition of the scale deposits in the pilot-
plant scrubber were very similar to the deposits found in the prototype
A comparison of the chemical composition of the scrubbing slurry and
scale deposits from the pilot-plant and prototype operations is shown
in Table VTI.

          Based on these tests and similar tests conducted by EPA in
a small pilot plant at Research Triangle Park, North Carolina,  it has
been concluded that the scaling of the prototype and pilot-plant scrubbers
was primarily caused by the combination of (l) high Ca:S02 mole ratio and
(2) the fine particle size of the soft Fredonia limestone.

Process Equipment and Construction Materials

          One of the primary objectives of the pilot-plant program was
to evaluate process equipment and construction materials.   This information
is required for the design of the Widows Creek system.
                                   533

-------
                                                       PROTOTYPE (O-O-O)
en
U)
4=
    gjf
    Ui Ul
     o u.
     in u
     IT
     o:
     CO
     £D
     O
     (0
80


70


60



50


6.4

6.2

6.0

5.8
        CO
        UJ
        (A
        O
 14 -
 13-
 12 -
                                                FIGURE 18


                 Comparison of Trends in 50^ Removal Efficiency and pH and Solids Content


                         of the Scrubbing Slurry in the Prototype and Pilot Plant

-------
                                                                        TABLE VII
Comparison of Chemical Analysis of Limestone - Wet-Scrubbing Slurry_
and Scale Deposits from the Prototype Multigrid Scrubber
at Shavnee Steam Plant and Pilot Plant at Colbert Steam
HLant





Scrubbing slurry
Filtrate, gm/1




Ul
U)
en


Plant
Prototype
Pilot plant



Duration,
hr
500
310



Limestone
sourcea
Fredonia
Fredonia




Ca
27.4
30-7


Cakej % by vt
Sulfite
MB s
2.0 3-8
0.9 4.2



Total
S Ca
5-5 0.8
4.9 1-1



Total
S
0.4
0-5


Total
dissolved
solids
5-4
7.1




Ca
23.2
21.5


Chenical scale, % by wt
Grid
Sulfite
Mg S
0.4 10.4
0.1 14.5


Walls
Total
S
16.4
17-1



Ca
25-5
19-7



Mg
0.1
0.4


Sulfite
S
9-3
8-9


Total
S
19.2
11.3


 Limestone supplied by Fredonia Valley Quarries,  Inc.,  Fredonia, Kentucky,  and ground by 3VA (50# -8 microns) in prototype dry ball mill at the Shawnee
 Steam Plant.
 Scrubbing slurry contained 15$ undissolved solids by weight.

IJOTE:  Both plants vere operated using a Ca:SOa mole ratio of 2.5-5-0 based on the S02 concentration of the inlet flue gas (normally 2800 ppn).  The
       SQS removal efficiency varied from 65 to 755&-

-------
          In as many cases as possible, commercially available equip-
ment was used and observed over extended periods of time.  Pumps,
valves, piping, and process instrumentation were evaluated in this
manner.  Most of the construction materials and coatings were evaluated
in a special test device designed to simultaneously expose several
different materials to the same chemical and erosive environment.

          A summary of the conclusions drawn from this evaluation
follows.

          Pumps;  Soft rubber—lined (natural or neoprenej centrifugal
pumps were normally used to circulate the scrubbing slurry.  The rubber
housing liner and impeller coating showed very little wear after 9000
hours of operation.  Severe erosion of the housing and impeller occurred
within 500 hours when an unlined Type JOU stainless steel pump was used.

          Flow Control and Isolation Valves:  Plug, pneumatic pinch, and
"butterfly valves were evaluated for controlling the flow of scrubbing
slurry.  The plug and pinch valves were lined with soft rubber.  The
body and disk of the "butterfly valve was coated with abrasions-resistant
polyethylene.

          The 3/l6-inch-thick rubber coating on the "body and plug of the
plug valve eroded through during 1200 hours of operation.  The pneumatic
pinch valve was not suitable "because of severe vibration caused "by pulsa-
tion of the rubber sleeve.   The butterfly valve provided an effective
means of trimming the flow of slurry,  although positive shutoff was not
possible.  The polyethylene lining was in good condition after 1000 hours
of service.

          Knife—type gate valves (unlined stainless steel construction)
were effective for isolation and on-off service.

          Piping;   Rubber hoses were used for most of the slurry piping
in the pilot plant.   The hoses had soft rubber (neoprene) liners and
gave excellent service.   The velocity of the slurry through the hoses
and piping normally did not exceed 8 ft/s.   In addition to the rubber-
lined hose,  special test sections of pipe of the following construction
were tested:

          • Mild steel

          • Type 3l6L stainless steel

          • Glass-reinforced polyester

          • Mild steel lined with polypropylene

          • Mild steel lined with urethane

          • Mild steel lined and coated with polyethylene

                                    536

-------
          Unlined mild steel pipe is not suitable because of severe
erosion and corrosion.  Some types of urethane—lined steel pipe failed
because of blistering of the liner.   All other materials tested gave
good service.  Selection of any one of the types of pipe construction
tested in the pilot plant (except mild steel and some urethane linings)
for a commercial unit would depend largely upon cost.

          Instrumentation;  Several instruments were used for process
control.  A discussion of the performance of the instruments follows.

          • SOP Determination-—An ultraviolet analyzer (DuPont photometric)
            was used to monitor the concentration of S02 in the flue gas
            before and after the scrubber.  This unit has given good service
            and required only a minimum of maintenance.   Leaky valves in
            the automatic sample selection system caused the initial instal-
            lation to be unreliable.  Replacement of the automatic sample
            selection valves with manual valves cured the problem.

          • pH Measurement'—The pH of the slurry in the multigrid retention
            tank was measured continuously with a system manufactured by
            Universal Interloc, Inc., consisting of a waterproof sensor
            and amplifier assembly  submerged  directly in the  slurry reten-
            tion tank.  The pH measurement is indicated and recorded by
            an instrument located in the control room.  This system has
            worked very well and agrees closely with laboratory pH measurements,
            The unit has required no maintenance,  -nere has been no indication
            of electrode scaling or erosion over 6 months of continuous expo-
            sure to the scrubbing slurry.

          • Slurry Density—The solids content (density) of the slurry was
            monitored continuously with a differential pressure (bubble
            tube) device.   Two tubes (l-inch pipe) are submerged directly
            in the slurry retention tank at a fixed pressure differential.
            A constant air purge is supplied to each tube.  The differential
            in pressure of the air purge is proportional to the density of
            the slurry.   This device has been very reliable and useful as
            a trend indicator.   The absolute solids content of the slurry
            is determined gravimetrically in the pilot-plant laboratory.

          • Measurement of Slurry Flow—The flow rates of the fresh
            limestone slurry (6(# solids) and the scrubbing slurry
            solids) are measured with magnetic flowmeters manufactured
            by the Foxboro Company.   The meters are lined with an abrasion-
            resistant material (Adiprene—L) similar to soft rubber.  The
            readout system is all electronic, consisting of a field-mounted
            signal converter and a recorder located in the control room.
            The performance of these meters has been very good and only
            a minimum of maintenance has been required over a period of
            18 months.
                                    537

-------
          Metals  and Coatings;  Several metals and coatings were
 evaluated for possible use in the Widows Creek installation.  Types 316
 and 3l6L stainless steel showed good resistance to corrosion and erosion
 in the scrubber (grids) and piping system.  Severe pitting and general
 corrosion occurred when stainless steel was used as a tube material for
 an indirect reheater in the scrubber exhaust gas duct.  Reheat tubes made
 of other materials (inconel 625, Incoloy 825, Carpenter 20 Cb-3, and
 Hastelloy C-276)  are currently being evaluated.  Mild steel, Cor-Ten and
 Type 304 stainless steel are generally not suitable for use in the slurry
 or gas systems because of severe pitting and general corrosion.

          Soft rubber (natural or neoprene) showed good resistance to
 erosion; however,  blistering and separation from the base metal (mild
 steel) occurred frequently.   Excellent bonding and resistance to erosion
was obtained with a urethane elastomer (Urecal 2003) manufactured "by the
Urecal Corporation.
          Limestone wet scrubbing provides an effective method for
removing S02 from boiler flue gases.  An S02 removal efficiency of 70$
can be expected with a multigrid type of scrubber at an L/G of 50 to 60.
The efficiency can be increased to about 9<$ if a two-stage mobile- bed
(TCA) scrubber is used.

          Long— term operating reliability is largely determined by proper
design and selection of suitable construction materials to cope with the
erosive and plugging characteristics inherent in slurry scrubbing.
                                   538

-------
    STATUS  OF C-E's AIR QUALITY
          CONTROL SYSTEMS
                  by

      M.  R. Gogineni,  Supervisor
Chemical Process Design and Development
        J.  R. Martin, Supervisor
   Systems and Equipment Development
   Kreisinger Development Laboratory
   P. G. Maurin, Assistant to Manager
      Air  Quality Control Systems
        C-E  Combustion Division
      Combustion Engineering,  Inc.
          Windsor, Connecticut
                     539

-------
    STATUS   OF  C-E's   AIR  QUALITY  CONTROL   SYSTEMS
                 INTRODUCTION
   The need for removing S02 from stack gases of oil
 and coal-fired combustion equipment has been and is
 being emphasized by the stringent emission limitations
 established by the Environmental Protection Agency
 and other government agencies. The EPA standards
 are 1.2 lb/106 Btu for coal and 0.8 lb/106 Btu for oil.
 More stringent  requirements in certain districts  have
 been passed.  An  example is  Clark County,  Nevada
 which limits S02 emissions  for a  1,500,000 kw steam
 generating station  to 0.15 lb/106  Btu or ]^th of the
 EPA requirement.  In  order  to meet  these  require-
 ments, a very high percentage of 862 produced by the
 combustion of most oil and coal fuels must be removed
 from stack gases.
   There are several ways to classify processes for the
 removal of S02 from stack gases: wet or dry, recovery
 or non-recovery, and absorption, adsorption,  or cata-
 lytic  oxidation/1'  In  reviewing these  processes and
 many  others, Combustion  Engineering decided  that
 wet  lime/limestone scrubbing  without  recovery  of
 sulfur was most worthy of development.
  The development of C-E's air quality control system
 started in  1964 with the construction of a small pilot
 facility in our laboratories. A second pilot application
 on a Detroit Edison Co. unit in 1966 and 1967 followed.
 Table  I  lists the  full-scale  installations'2' that have
 been sold to date by C-E. In a move to further accele-
 rate development, a large prototype unit of 12,500-cfm
 capacity was constructed in  our laboratory and began
 operations in early 1970.<3)
  This paper describes the C-E AQCS for  the removal
 of particulate matter  and SOa from stack  gases  of
 steam  generators emphasizing  C-E's experience with
 the full-size field units.
                                       SYSTEM DESCRIPTION
                             In the tail-end C-E Air  Quality  Control System
                           (Fig. 1), a slurry of pulverized limestone or slaked lime
                           enters directly into the reaction  tank (located at the
                           bottom of the scrubber). Recirculation pumps convey
                           the scrubbing slurry from the reaction tank to underbed
                           spray nozzles. The incoming  gas, laden with dust and
                           S02, contacts the  sprayed slurry  and continues to the
                           bed. The removal of SOz and particulate matter occurs
                           in the bed. The scrubbing cycle continues with the
                           reacted materials draining to the reaction tank which
                           is designed to provide for completion of chemical re-
                           actions and precipitation of solids. From the reaction
                           tank, pumps recirculate the scrubbing solution.
                             A bleed line provides the necessary solids removal to
                           a  clarifier or pond. Here, solids  settle  and clarified
                           water is available  for  recirculation. The cleansed flue
                           gas passes through a mist eliminator for removal of the
                           remaining water and is then reheated for induced-draft
                           fan protection and plume control.
                             The furnace injection process (Fig. 2) involves the
                           injection of an  additive which  contains a  high  per-
                           centage of calcium, such as limestone, into the furnace
                           of a steam  generating unit.  The  pulverized additive
                           calcines in the furnace  and reacts with the combustion
                           gases, removing 20 to 30%  of the sulfur oxides, in-
                           cluding all the 80s.  The flue gas  enters the scrubber
                           and  the  process proceeds as  described above. Due to
                           additive  distribution and boiler deposit problems, the
                           furnace injection system is not now being offered.
                             The basic design variations of tail-end systems result
                           from additive selection — limestone or limj — and are
                          additive preparation and system control. Although both
                          additives enter the system as a slurry, the  limestone
                           (usually received in the — Yi to —2-inch size range) re-
        Utility
Union Electric

Kansas Power & Light

Kansas Power & Light

Kansas City Power & Light

Kansas City Power & Light

Northern States Power
Northern States Power
Northern States Power

Louisville Gas & Electric

Southwest Public Service
C-E CONTRACTS FOR
      Unit
 Meramec 2

 Lawrence 4

 Lawrence 5

 Hawthorn 3

 Hawthorn 4

 Sherburne 1
 Sherburne 2
 Black Dog

 Paddy's Run 6

 Nichols Station B
                                                  TABLE I
                                                AIR QUALITY CONTROL SYSTEMS
 Size, Mw
   140

   125

   430

   100

   100

   690
   690
 Prototype
ISOOOcfm
    65

   350
     Fuel
Coal

Coal/Gas

Coal/Oil/Gas

Coal/Gas

Coal/Gas

Coal, Low Sulfur
Coal, Low Sulfur
Coal, Low Sulfur

Coal/Gas

Coal/Gas
      System
Furnace Injection,
Limestone
Furnace Injection,
Limestone
Furnace Injection, •
Limestone
Furnace Injection,
Limestone
Converted Tail-End,
Limestone
Tail-End, Limestone
Tail-End, Limestone
Tail-End. Limestone

Tail-End, Lime
(Ca(OH), Sludge)
Tail-End, Lime
                                                    540

-------
         mcir MS
    10,
9*««*«T9»-
_m»viet
wjrr*
_jro OTH«
KHirS
s«ri~\
r
.A
...-•" h
,,**'

"- CJICJOK
* 1
a.
y
r^M
V
«V<
\-JTIM»««
— MIXCft
\-5MI«r PUMP
cr/ov nwr
*— «« now MCTE«

                                                                                          *»    V*
                                                                                                  ni*r
                                                                                  umtKrum nuf
  SCDUCKHS
                                         Fig. 1: C-f foil-end AOCS
                                   RUCTION TANK
                                      III
           Fig. 2: C-E furnace injection AQCS

Quires size reduction and water addition. A wet milling
sYstem is used for this. Lime requires a slaking opera-
tion to  form a calcium hydroxide (Ca(OH)2) slurry.
A direct supply of solid "Ca(OH)2 is slurried to facilitate
"lira-system transport and control.
  For a CaO/Ca(OH)2 system, the additive enters the
faction tank as  required to maintain the spray slurry
Pi I at a level high  enough to maintain SO2  removal
efficiency, yet low enough to avoid  CaS03 scaling.  For
a  calcium  carbonate  system,  fuel  sulfur  content
Determines additive feed rate.

                SYSTEM DESIGN"*
System Chemistry
  The possible  reactions taking  place  in  the  wet
'''ne/limestonc SOa removal system have been studied.
'•*ur conclusions relative to the controlling reactions are
''ttswl on consideration of theoretical equations in light
°r operating experience in both  field  and lalwratory
systems. Because there are considerable differences in
the operating conditions required to provide adequate
SOj removal in the absence of scale or deposit forma-
tion when utilizing lime or limestone as additives, the
chemistry  of the systems are treated separately. The
essential reactions governing these systems are:
Calcium hydroxide or lime system reactions
  CaO +  H20>=tCa(OH)2                      (1)
  Ca(OH)2 + S02 & CaS03 + H20             (2)
  CaSOs + SO2 4- H20 * Ca(HS03)2            (3)
  Ca(HS03)2 + Ca(OH)2 * 2CaS03 + 2H20     (4)
  2CaS03 + 02-»2CaSO<                      (5)

Calcium carbonate or limestone system reactions
  CaCOs + C02 -r- H20 ^ Ca(HC03)2           (6;
  2S02 + Ca(HC03)2 ^ Ca(HS03)2 + 2C02     (7)
  CaS03 + S02 + H20 ^ Ca(HS03)2            (8)
  2CaS03 + 02 ^ 2CaSO4                      (9)
  Ca(HS03)z + 2CaC03 n Ca(HC03)2 + 2CaS03 (10)
Following  the  initial steps of hydration (Eq. 1) and
formation  of calcium sulfite (Eq. 2), removal of SOa in
the lime or calcium hydroxide system depends upon
the formation of calcium bisulfite by reaction of sus-
pended calcium sulfite with  sulfur  dioxide and water
(Eq. 3).
  The control of sulfite scaling requires that a minimum
amount of free hydroxide ion be recirculaled  to the
scrubber, therefore, fresh additive (lime or slaked limf)
is added in a reaction tank  external to  the scrubber
where calcium sulfite is formed (Eq. i). An amount of
calcium sulfile equivalent to  the 862 removed (or lh«
                                                541

-------
 fresh Ca(OH)z added) is conveyed from the system to
 a pond or vacuum filter and the remainder recycled to
 continue the removal process.
   The principal absorption reactions for the calcium
 carbonate system are shown in Eqs. 6, 7 and 8. Sulfur
 dioxide reacts with the relatively soluble bicarbonate
 to form  calcium bisulfite.  In addition, solid calcium
 sulfite recycled from the reaction tank reacts  with 862
 to form bisulfite.
   The reactions in  which sulfite is oxidized to sulfate
 (Eqs. 5 and 9) and soluble bisulfite is converted to in-
 soluble calcium sulfite (Eq. 10) account for the  waste
 products  as well as the regeneration of the solid calcium
 sulfite reactant that is recirculated to the scrubber. The
 ratio of calcium sulfite to calcium sulfate found in the
 air quality control system solid waste depends upon the
 extent to which these reactions go to completion.

 Scru66cr Design
   The primary function of the scrubber is to transfer
 SOz from the flue gas into the liquid. The SOz remains
 in the liquid or is converted partially to solid sulfur
 compounds in the scrubber. Hence, knowledge of vapor-
 liquid mass  transfer  rates is  important  in  scrubber
 design. The marble bed scrubber, which has a turbulent
 layer, acts like an absorption tray.
   The vapor-liquid equilibrium line and efficiency are
 needed for tray design. Typical operating and equilib-
 rium lines are shown in Fig. 3. The operating line is the
 material balance line and has a negative slope of L/G
 (liquid to gas ratio).
                            x0     x
               MOL£ FRACTION OF S02 IN UQUID, X
           Fig. 3i Vapor-liquid matt trontfw

  The inlet gas composition YI is known, and the SOz
concentration in the  incoming liquid Xlt  which  is
usually zero, is also  known. Hence,  point  1,  which
 represents the inlet condition is known. The operating
 line can be drawn through Xj Yi with a negative slope
 of L/G.
   The  intersection of the operating and equilibrium
 lines (X, Ye) represents the  scrubber outlet gas and
 liquid conditions for a theoretical stage  which repre-
 sents 100%  overall tray efficiency. The overall tray
 efficiency E is defined as:
                    E
Y.-Y,
Y,-Y.
(ID
 and is usually less than 100%. It can be seen both from
 Fig. 3 and Eq. 11 that the actual outlet liquid and gas
 compositions (Xo, Y0) can be predicted using the tray
 efficiency and the point of intersection of the operating
 and the equilibrium lines.
   Laboratory tests  show that the tray efficiency of the
 marble bed scrubber is 90 to 95%. This indicates that
 the marble bed scrubber  is a good liquid-gas contactor.
   Once the tray efficiency is  known the number of
 marble beds needed to obtain a required SOz removal
 can be determined from  the operating and the equilib-
 rium lines.
   Knowing the total  alkalinity of the liquid  available
 in the scrubber,  which is a function of the additive
 dissolution rate for slurry systems, the equilibrium line
 is plotted using a computer program.<5> For a specific
 L/G, the equilibrium line shown  in  Fig. 3 moves to
 the left or right depending on  the alkalinity of liquid
 available  in the scrubber and, hence, affects the SOj
 removal  efficiency.  The SOz  removal efficiency  is
 defined as:
                                                                                Y,-
                                                                                                        (12)
If the  additive dissolution  rate  is  high enough to
maximize the available alkalinity  in the liquid in the
scrubber, the equilibrium  line will move to the far
right to a point where Y. can become zero. This repre-
sents  the most  favorable  condition for SOa  transfer
from  gas to liquid. Comparison  of  Eqs. 11 and 12
shows that  the  SOj removal efficiency will approach
the tray efficiency as Y. approaches zero.
Reaction Tanks
  The function of the reaction tanks is to provide:
  a. Dissolution of the  additive in order to convert
     the highly soluble bisulfite present in the liquid
     leaving the scrubber to relatively insoluble sulfite.
  b. Precipitation of calcium sulfate which is formed
     in  the system (scrubber and/or reaction tanks)
     due to the oxidation of sulfite.
  c. Precipitation of calcium sulfite.
  Additive  dissolution rates vary considerably  with
the type, origin, preparation, and concentration of the
additive. At C-E, a prototype  scrubber system,  pilot
                                                    542

-------
plant scrubber system, continuous flow  stirred tank
reactors, and batch reactors have been used to deter-
mine the dissolution rates for individual additives used.
  The following rate expression for the  precipitation
of calcium sulfate has been developed:'6'
                R - -KZ(C-C.)2             (13)
  The rate of desupersaturation, R, is proportional to
the gypsum seed crystal concentration, Z. The differ-
ence between the actual concentration,  C,  and the
equilibrium  concentration, C«  of SO^  or Ca++ is the
driving force terra. Although  it is more accurate to
express the -driving force in terms of the  activity and
the solubility products, the driving force in Eq. 13 is
given in concentrations for  convenience,  by including
the factor for converting activities into concentrations
in the rate constant K.  Using the rate constants de-
termined  by experimentation,'7' the  rate  expression
given by Eq. 13 is employed in the design of reaction
tanks to ensure that calcium sulfate supersaturation is
eliminated.
  Laboratory studies are in progress to determine the
calcium sulfite precipitation kinetics and the oxidation
kinetics of sulfite to sulfate.

      DEPOSIT AND SCALE FORMATION
  Calcium sulfite and calcium sulfate scaling can  be a
problem for lime/limestone  wet  scrubber systems.
Scaling occurs when the solutions  are supersaturated
to a point where heterogeneous crystallization takes
place resulting  from  nucleation. The ratios of the
products of the activities (A) of Ca++ and SO4 or SO3
to their solubility product constants (KSp) as a measure
of the degree of supersaturation are:
                ~AQa-HAgQ- "1  <1 Subsaturation
                 ^	I  =1 Saturation
                   Sp(CaSOa)J >1 Supersaturation
   SP(CaS04)
  Laboratory experiments  have shown  that  hetero-
geneous crystallization is not significant until the ratio
of the activity product to the solubility  product con-
stant reaches about 1.5 for calcium sulfate and about
7 for calcium sulfite.
  Heterogeneous crystallization is  minimized by pro-
viding seed  crystals for  homogeneous crystallization
and  by designing the reaction tanks so that the liquid
leaving them is  close to  saturation  and not highly
supersaturated. This requires knowledge  of precipita-
tion  kinetics of calcium sulfate and sulfite.

Calcium Sulfite Deposition
  Calcium sulfite (CaSOa • HH20) is formed in the
scrubber  under  those  conditions  that   favor  sulfite
formation. These conditions  are apparent when one
considers  the sulftte-bisulfite equilibrium and compares
the relative  solubilities of the corresponding calcium
salts. As seen in  Fig. 4, extremely soluble  bisulfite  in
                                                                 MOLE FRACTION SULFUROUS ACIO-OISUl.FITE-SUI.riTe
                                                                                   pH
                                                          Fig. 4: Mole fraction sulfurou* ocid-bitulfite-tulfiit vs. pH
solution changes to relatively insoluble sulfite when the
solution pH shifts from 4 to 10. When SOz is absorbed,
the scrubber solution  is usually  between pH 4 and 6
and, therefore, the predominant  species is bisulfite.  If
the pH of the scrubber solution containing bisulfite is
suddenly raised either in localized areas or in a reaction
tank, crystallization of calcium sulfite will occur.
  Experimental work  with lime  scrubbing has shown
that sulfite  scaling occurs in the scrubber bed when
free  hydroxide is introduced according to the above
explanation. By proper control of the pH of the spray
slurry (less  than  10)  entering the  scrubber,  calcium
sulfite scaling will not occur in the scrubber.*8*
  In the calcium carbonate system, the buffering action
of the carbonate-bicarbonate couple (Eq. 6) maintains
a system  pH between 5 and 7, thus sulfite scaling is
not encountered.

Calcium Sulfate Deposition
  The solubility  of calcium sulfate is only slightly in-
creased with increasing pH, and calcium sulfate scaling
is  related to the tendency  of this  material  to form
extensively  stable  supersaturated   solutions.  While
chemical theory predicts that a given ionizable species
will not remain in'solution when the solubility product
of its component  ions has  been exceeded,   calcium
sulfate may be held in solution to an extent twice thot
predicted  before  crystallization of calcium  sulfate will
occur (gypsum, CaSO« • 2HaO).
  The significance  of this phenomena to  scrubber
operation  is that SOa removal can be accomplished
while scrubbing with  solutions containing more than
the theoretical  calcium and sulfate ion  concentrations,
but less than some experimentally determined level at
which precipitation will occur  within the  scrubber
proper.
  Crystallization  from supersaturated  solutions car
occur by  two processes, formation of new crystals or
nucleation and growth of existing crystals.
                                                     543

-------
   The internal  surfaces of the scrubber can provide
 nucleation sites, thus resulting in scale formation. For
 many crystal systems, growth will occur without sig-
 nificant nucleation if sufficient seed crystals are pro-
 vided. Work by other investigators*61 has shown that
 supersaturated calcium sulfate solutions can be effec-
 tively  desupersaturated by  circulation  of  1 to  5%
 gypsum seed crystals.
   By  employing  this  technique,  operation  free  of
 calcium sulfate scaling has been demonstrated in both
 laboratory and  field installations. This  seeding tech-
 nique is the key to "closed loop" operation in which the
 only liquid leaving the system is  by evaporation or
 combined with the solid by-product of the scrubbing
 system.
   The potential for sulfate scaling is generally more
 prevalent in the calcium carbonate system than in the
 calcium  hydroxide system.  The  calcium  hydroxide
 system reaches steady state below the value of calcium
 sulfate saturation.

 Control Techniques™
   The controls for the C-E AQCS depend upon the
 additive  type. The limestone tail-end system can  be
 controlled by the limestone  feed to the system. The
 limestone feed is determined by the 862 in the flue gas
 or sulfur in the fuel. The  solids concentration in the
 slurry is maintained to prevent calcium sulfate scaling.
   The tail-end lime system  is controlled by  the pH
 of the spray slurry and the total solids in the system.

 Slurry Circulation
   Slurries containing 2 to 10 wt. % solids are circulated
 in the lime/limestone wet scrubber systems in order to
control  scaling in the  system and to improve SOa
removal and additive utilization. The circulation of the
high solids slurries in the system could  cause serious
erosion problems in pumps, piping, nozzles, etc.
   AQCS solids are different from  other solids. Little
technology is available  to  design pumps, piping, and
nozzles that can  handle AQCS solids, therefore,  C-E
undertook extensive testing of these components.
   Several pumps of different materials  have been in-
stalled on C-E systems both in the field and in the
laboratory. The information obtained from the evalu-
ation of these pumps is being used in current system
  Several different materials have  been evaluated to
determine piping material that can withstand erosion
and  corrosion. The piping is designed  so that slurry
velocities in the pipes are high enough to  prevent
settling.
  Severe nozzle plugging problems have occurred when
commercially available nozzles were used. Therefore,
C-E has  developed a nozzle that does not plug, erode,
or corrode in the scrubber environment.
Solid-Liquid Separation
  Solid-liquid separators such  as  clarifier-thickeners
and vacuum filters separate the solids from the liquid.
The solids are disposed of and the liquid is returned to
the system.  This equipment is presently being designed
by vendors  to C-E's specifications.

Gas-Liquid Separation
  The gas leaving the bed carries water droplets which
contain solid particles and dissolved solids. Mist elimi-
nators are used to remove water droplets from the gas.
Proper design of these sections is essential to prevent
both plugging with  solids and  re-en train men t of the
liquid collected on the  surfaces.
  Gas velocity and the distance between the bed and
the  mist  eliminators are important operational and
design variables. Extensive test work  has been  con-
ducted in the laboratory and in the field to evaluate
these variables. Results of these studies are being used
in the design of mist eliminators for C-E systems.

Reheating
  The purpose of reheating the gas leaving the scrubber
bed  is to protect the  I.D. fan and to reduce plume
formation.  The amount of  reheat required depends
upon several factors such as atmospheric conditions
and  stack height.

Additive Preparation and feeding
  The limestone is ground to a small size (about 80%
thru  200  mesh). The  slurry from  the  limestone mill
scalpers (about 60 to 70% by weight) is stored in the
limestone slurry storage tanks. The slurry is then trans-
ferred to tanks, where it is diluted with  make-up water
to reduce the solids  content  to  10  to 15% by weight.
This dilute slurry is added to the reaction tanks for use
with the slurry spray to the scrubber bed.
  For a  lime system,  slaking, storage, dilution, and
transport  steps are involved. Use  of Ca(OH)2 rather
than CaO eliminates the slaking operation.
  The additive handling system is designed so that the
solids do not settle in pipes and tanks.

                 EPA CONTRACT
  C-E has recently completed  a research  contract for
the Environmental Protection Agency (EPA) to op-
timize lime/limestone wet scrubbing processes for SOj
and participate removal in a marble bed scrubber.
  Three types of tests were conducted  on the 12,000-
CFM C-E marble bed scrubber. Soluble system  tests
using once-through sodium carbonate scrubbing  solu-
tions were  conducted  using a  single  marble bed  to
determine the  vapor-liquid  mass transfer character-
istics. Limestone furnace injection tests were conducted
using boiler calcined limestone and flyash mixture to
determine the system performance and the solid-liquid
mass transfer rates. Limestone tail-end system  tests
                                                    544

-------
were conducted  using  limestone  slurry to determine
whether single marble bed results can be extrapolated
to predict the performance of a two-bed scrubber and
to determine the solid-liquid mass transfer rates.
  The results of the once-through soluble system tests
using a Na2COs scrubbing solution show that:
  1. The marble bed  scrubber  is a  good liquid-gas
    contacting device  with an overall tray efficiency
    of 90 to 95%.
  2. The SOz removal  in  the  marble bed scrubber is
    limited by the vapor-liquid equilibrium.
  3. Liquid to gas (L/G) ratio and the scrubber liquor
    alkali composition  strongly influence 862 removal.
  The C-E scrubber at EPA's Shawnee test  facility
was originally supplied with  one marble bed.  The
current commercial  systems offered by  C-E for high
sulfur fuel applications have scrubbers with two marble
beds.  Therefore,  C-E  recommended  that a  second
•narble bed  be  installed in the  scrubber  so  that a
meaningful evaluation  of C-E's  current  commercial
offering  could  be accomplished.  Due to the lack of
available funds  and the  scheduling problems,  EPA
decided against installing the second bed. As a reason-
able alternative,  both C-E and EPA agreed that tests
would be run on the C-E AQCS prototype to determine
whether the performance of a two-bed scrublx-r could
be predicted from the single marble  bed perform;,.,,',-
of the EPA test facility at Shawnee.
  A series of tests were run on the C-E \QCS prolutyp,.
using limestone slurry with one and two marble U-ds
Test results show  that the performance  (SO2 removal
efficiency and elimination of scaling) of the scruhbrr
with two marble beds can be predicted by exIrapolaliriL'
the single bed test  results. The S02 removal clfinem-ics
of the lower and the upper beds appear to l«: the same
based on the SOo concentrations entering the respective
bed.

Kansas  Power and Light — Lawrence 4
  This  125-Mw coal  and gas  fired  C-E boiler  was
retrofitted with a furnace injection AQCS in 1968. The
AQCS consisted of  two wet  scrubbers  and  related
equipment (Fig. 5). During 1971, it was recognized Unit
this system could  not continue  to operate as an open
system (i.e., liquid blowdown).  While liquid was never
discharged to  a natural body  of water,  the sludjje
disposal  pond constructed at the plant site in 1%8
never reached saturation because of the small quantity
of scrubber effluent flow into this pond during the throe
years of operation and because of intermittent operat ion
of the AQCS on Unit 4.
                                                            STACK
 xESTONE SUPPLY
   COM. SUPPLY
                                                                                           SETTLING POND
                               Fig. 5: Flow diagram of AQCS for Unit 4 at KP&L

                                                  545

-------
  Utilizing  the results  of tests  conducted  at  our
Kreisinger  Development Laboratory during 1971, it
was  decided that the system could be  modified to
ensure scale-free closed-loop  operation.  In order to
verify the design modifications, a  series of tests were
conducted  in  February  and  March of  1972 which
demonstrated scale-free operation for periods extending
to two weeks.
  During these tests, all process streams were measured
for chemical constituents in both the solid and liquid
phases. In addition, the  flue gas entering and leaving
the AQCS  was  measured for  both gaseous and solid
constituents using EPA  approved methods. Table II
gives a typical set of data obtained in this test series.
                      TABLE II
              TYPICAL OPERATING DATA
              KANSAS POWER & LIGHT 4
                   February, 1972
                     5.0  gr/DSCF
                     0.06 gr/DSCF
                    1800 ppm
                     550 ppm
                     9.2
Inlet dust
Outlet dust
Inlet SO,
Outlet SOj
PH

% Solids
CaO
SO3
SO4
Solid analysis
CaS04 • 2H>0
CaSO, •
Flyash
CaO
                   Underbid Slurry
                        7.3
                       960 ppm
                       250 ppm
                      2,150 ppm

                        20%
                        15%
                        60%
                         5%
Pot Effluent
     7.3
   1.250
   1.300
   2,450
Since these results agreed with the data obtained in our
laboratory with respect to seed crystal concentration
required to prevent CaSC>4  scaling, and SOa removal
was maintained at 70 percent for the single marble bed
scrubber, it was decided to  complete  the  modification
in order to operate  the system at the 7 to 10% solids
level at full load.
  The modifications to Unit 4 were made during the
spring and summer of  last year. Subsequent operation
of the system in the fall of last year revealed a problem
with  maintaining mist eliminator performance while
operating with high solids  (7 to 10%). An extensive test
program has been undertaken  to solve this problem.

Kansas Power and Light — Lawrence 5
  Unit 5 at the Lawrence  Station of Kansas Power and
Light, a 430-Mw C-E coal, gas, or oil fired unit was
first placed in service in the  spring of 1971. The AQCS
supplied for this unit was based on the same design as
Unit  4. The system consisted of  six scrubbers,  each
handling  approximately  165,000 CFM  of flue gas at
125 F. The additive, calcium carbonate, was pulverized
and injected into the boiler, calcined, and then conveyed
to the scrubber for use in the SO2 absorption process.
  The AQCS was not placed in service until November
1971 since the unit was operated on gas until that time.
The system was operated for three weeks without any
serious problems until the sludge pond became satu-
rated with calcium sulfate, at which time scaling in'lhe
scrubber was observed.  Since  tests were  already  in
progress on Unit 4 to enable operation of the AQCS as
a closed system without scale formation, the AQCS on
Unit  5 was operated on an intermittent basis  during
the winter season when the unit fired coal.
  After the successful completion of the closed loop
modifications on Unit 4, plans were made to "update"
Unit  5. Figure  2  shows the modifications that were
made. These modifications included installation of a
reaction  tank where crystallization, dissolution, and
precipitation reactions take place; mixing devices for
the reaction tank; fiberglass and/or rubber lined piping
for the high solids recycle system; new underbed slurry
pumps; improved controls that can withstand the wet
scrubbing environment;  redesigned Blurry nozzles  to
decrease chemical and mechanical wear  problems; and
a new demist system. Further, it was decided from test
data  obtained during the winter  of 1971-72 that we
should reduce the gas throughput of the marble bed
scrubbers in order to minimize carryover from the bed
due  to  localized high velocities,  and incorporate the
concept of  spare scrubbing  capacity by  adding two
additional scrubbers to this unit.
  The modifications were completed late last fall and
the AQCS was returned to service for the winter period
of coal operation.
  The unit has operated approximately 1800  hours
(since January). The results duplicate closely the tests
run last year on Unit 4.
  Presently, an extensive series  of performance tests
are being conducted on this unit  by C-E to determine
full-load capability of the system.

Kansas City Power and Light — Hawthorn it
  At the Hawthorn Station of the Kansas City Power
and  Light, a furnace injection  system was retrofitted
on a 100-Mw low-load-factor unit. The system consists
of two scrubber  units and related  equipment, and
includes a water seal by-pass (Fig. 6). The system was
started up in September 1972.
  Preliminary operation of the system indicated that a
serious maldistribution of additive existed between the
two scrubbers. After several weeks of testing, it was
also determined that the  back-pass of this boiler was
not  capable of  handling  the additional dust loading
created by the furnace injection of limestone.
  Therefore, in late 1972, it was decided to modify the
system  to  a  tail-end limestone, single  marble  bed
                                                    546

-------
                                                                       	MODIFICATIONS
                                                                             LATE
                                                                                           *-TO POND
                                                                                             FROM POND
                         Fig. 6:  Proeew flow schematic KP&L Hawthorn Station Units 3 and 4
system with the pulverized limestone being blown into
the  scrubber inlet  duct (Fig. 6). Modifications have
been completed,  but the system has not been tested
to date.
  A test series was  run during March to determine the
Performance of this system on low sulfur, high alkali
Wyoming coal. The limestone feed  \vas maintained at
Zero for these tests because there was significant amount
of alkali in the coal. Some  coal samples  indicate as
much  as 200%  stoic-biometry of alkali  (if all were
available for reaction with the sulfur in the coal). The
results  shown  in  Table  III were obtained  during this

                     TABLE III
                  OPERATING DATA
            KANSAS CITY POWER &  LIGHT
         HAWTHORN 4,  WYOMING COAL TEST
         Inlet SOj:             700 - 300 ppm
         Outlet SO]:            300 • 40  ppm
         % Solids              3.0 - 6.0
         Underbed slurry  pH     5.7 • 7.1

             UNDERBED SLURRY LIQUID
                 CHEMICAL ANALYSIS
         CaO                  850 ppm
         SOi                  150 ppm
         S0<                  220 • 2520 ppm

lcst program. The S02 removal varied from 60 to 90
-------
                                                                                            SLUDGE
                              Fig. 7:  Proeew flow schematic IG4E Paddy'* Run No. 6
  This  system represents the culmination of a three-
year  effort  initiated in  1970. Pilot plant  tests were
conducted at  C-E's Kreiainger Development Labora-
tory in 1971 to determine if an AQCS could be designed
that  would  operate with the Ca(OH>2 sludge that
Louisville Gas & Electric had available. The Ca(OH)2
sludge is a  waste product from  the manufacture  of
acetylene.
  In  June 1971, a 100-hour  continuous test  on the
12,000-CFM prototype scrubber system at KDL was
completed and, as a result,  the  system  design was
finalized and fabrication of the system began. Later in
1971, it became apparent because of  more stringent
water pollution regulations that the system would have
to be operated as a closed system. Therefore, additional
testing in the  laboratory was done in February  1972.
The findings of these tests  showed that the  closed
system was accomplished by obtaining essentially zero
oxidation and  providing sufficient  seed  crystals  to
precipitate the calcium sulfite without forming scale.
  This system is presently in the process of starting up.
At this  time, there has been approximately 300 hours
of operation with one scrubber. The chemistry of the
system has undergone preliminary checkout and con-
firms the findings of our laboratory work.
  The scrubbers have been operating exceptionally
well. S02  removal has been 85% or higher depending
on the L/G ratio and additive feed rate being main-
tained. It appears  that  the guarantee level of 80%
sulfur dioxide removal can be  obtained with a total
L/G of 40.
  The Ca(OH)2 sludge feed rate is controlled auto-
matically by the reaction tank pH. This control system
has been in service continuously since start-up without
a problem.
  The vacuum filter system has also operated without
problems.  The filter cake has varied from 30 to 50%
solids,  the lower solids being quite muddy  and the
high solids being handled easily.
  The data in Table  IV are for  a  150-hour run con-
ducted during the last week in April.

Northern Stales Power — Sherburne County 1
  The systems for Northern  States Power,  Sherburne
County Station have  been  designed to clean the  flue
gas from two  690-Mw  C-E coal-fired units. These units
                                                   548

-------
                     TABLE IV
             LOUISVILLE GAS & ELECTRIC
                  OPERATING DATA
      Inlet SOi                 2400 • 2800 ppm
      Outlet SOj                300 -  450 ppm
        Undarbed Slurry
      pH         9.2
      CaO      850 ppm
      SOT      20° PPm
      SOJ      800 ppm
      % Solids    9.5
Pot Effluent
   5.2
 1.100 ppm
 1,300 ppm
  850 ppm
   9
      Solids
      CaS04-2HjO   1-2%
      CaSOj • ViHjO   78%
      Flyash         20%


will fire a low sulfur Western coal. Typical analysis of
this coal is given in Table V. The addition of limestone
to the system  will  he varied to maintain  the  sulfur
dioxide removal to satisfy air quality requirements and
balance the chemistry in the scrubbers. Since significant
quantities of alkali exist in the flyash of this Western
coal, the limestone addition will serve as a supplement.
Tests run at  our facilities indicate  that  for certain
                                             TABLE V
                                     NORTHERN STATES ROW Eh
                                      TYPICAL COAL ANALYSIS
Btu/lb
Moisture
Sulfur
Ash

MgO
CaO
SO.
                                               COAL
ASH
8,130
23.5%
 0.8%
 9.0%

 5.9%
21.9%
 1.4%
                        sulfur concentrations in the coal, all the alkali required
                        can be supplied by the flyash.
                          Another  interesting aspect  of the  design  of  these
                        systems is related to the appreciable quantity of mag-
                        nesium in the flyash of the coal to be  fired. Tin; sulfur
                        salts of magnesium are extremely soluble,  thcn-fori'.
                        the rate  of dissolution of magnesium must he deter-
                        mined in order to predict  the  total dissolved solids in
                        the systems at steady state. The prediction of total
                        dissolved solids was  required  to determine possible
                        effects on the system chemistry  (i.e., ionic strength of
                                                                       CUSTOMER
                                                                       DISPOSAL
                            Fig. 8: Proposed process flow schematic NSP Black Dog Unit


                                                 549

-------
tbe process solutions) and to provide information to the
Water Resources Commission of Minnesota. A develop-
ment program conducted  at our laboratory  predicted
a total dissolved solids level of 15,000 ppm, dictating
closed loop operation. Closed loop system  operation
has been incorporated by  providing adequate  seed
crystals and a reaction tank external to the .scrubber
beds to carry out crystallization and precipitation.

Northern States Power — Black Dog Prototype
  AB part of the project to develop the tail-end lime-
stone  scrubbing systems  for  the  Sherburne County
Station of Northern States Power, a prototype facility
has been constructed at NSP's Black Dog Station. The
purpose of this facility is to verify the design for
Sherburne  County, test various system components,
and provide a  training  facility  for NSP  operating
personnel.
  This facility is a scaled down duplicate of the Sher-
burne County design except that no additive grinding
equipment was incorporated. The scrubber module has
a capacity of 12,000 ACFM at 130  F. All other related
equipment is included in the facility as shown in Fig. 8.
  Tbe  Black  Dog system began  air/water testing
during the first two weeks of March 1973. On Marehll,
it was placed in service with a stream of flue gas being
diverted from the Black  Dog unit precipitator  inlet
                                             duct.  Since that date, approximately 1,000 hoars of
                                             operation have been logged.
                                               The test series at the Black Dog facility includes the
                                             evaluation of the following parameters: (a) gas velocity,
                                             (b) additive feed rate, (c) percent solids being circu-
                                             lated, and (d) L/G ratio variation. The steady-state
                                             concentration  of soluble cations (i.e., magnesium) is
                                             being studied  because of their relatively high concen-
                                             tration in Western coal.
                                               The initial tests at Black Dog agree with our labora-
                                             tory pilot plant data  as shown in Table VI. Current
                                             plans are to complete operation of this facility by late
                                             August of this year.

                                                                 TABLE VI
                                                           OPERATING DATA FROM
                                                        LABORATORY PILOT PLANT TEST
                                                        FOR NORTHERN STATES POWER
                                                      Inlet SOi               BOO ppm
                                                      Outlet SOi              200 ppm
 I-
 V)
 o
 o
 u>
     100
80-
      60-
 0-
 o
 <   40-
 O
 u.
 O   20-
 3*
        0-



—


—
< < * j
!.
,f, 	 "
£ ?
\ *
1 "]>
1 *" f V
Ai.*.. >•
5-
S 5. *».
* i / t

v <
f > \
\Jf /^
> i ; j<
, <
< ^ " <
^ :
                                                      L/G
                                                      Underbed slurry pH
                                                      Pot effluent pH

                                                      Underbed slurry
                                                      Li Quid analysis
                                                                    SOT
                                                                    SOT
                      20
                      6.5-7.1
                      5.8 - 6.3

                      600 ppm
                      900 ppm
                      150 ppm
                    4.000 ppm

                   MILLS
                    PER
                                                                     $/HR    KWHR  
-------
                  ECONOMICS
  Table VII shows the assumptions for a 600-Mw unit
and Fig. 9 shows the estimated operating costs.
                     TABLE VII
          ASSUMPTIONS FOR DETERMINING
         COSTS FOR A TYPICAL 600-Mw UNIT
Heat rate
Load factor
Fuel ash
  Sulfur
  HHV
Ash drop out
  To precipitator
Electrostatic precipitator
AQCS tail end
         9500 Btu/Kwhr
              70%
              12%
Additive limestone
Electric power

Water

Disposal

Maintenance
Labor
Capital
          1 1,700 Btu/lb
              15%
              85%
          90% effective
  SO, Emission Rate 1.2 lb/10« Btu
      95% pure • 130% stolen.
        $3.SO/ton delivered
       $l,00/ton wet grinding
         2.5% of 600 Mw
           $.007/Kwhr
          1.75 gpm/Mw
          $.13/1000 gal.
             $4/ton
          50% dry solids
        4.3% of capital cost
4 men/shift - 4 shifts - $15.000/yr/man
         17.5% of $34/Kw
                 REFERENCES
1. MAUBIN, P. G. and JONAKIN, J., "Removing Sulfur
   Oxides from Stacks," Chemical  Engineering, April
   27,1970.
2.  JONAKIN, J. and MARTIN, J. R., "Applications of the
   C-E Air Pollution Control System" Second Inter-
   national Lime/Limestone Wet Scrubbing Sympo-
   sium, New Orleans, La., November 8-12, 1971.

3.  PLUMLEY,  A.  L. and GOGINEM, M,  R.,  "Research
   and Development in Wet Scrubber Systems," Second
   International Lime/Limestone Wet Scrubbing Sym-
   posium, New Orleans, La., November 8-12,1971.

4.  GOGINENI, M. R., TAYLOR, W. C, PLUMLEY, A* L.,
   and JONAKIN, J., "Wet Scrubbing of Sulfur Oxides
   from Flue  Gases,"  American Chemical Society  Na-
   tional Meeting,  New York, New York, August 27-
   September 1, 1972.

5,  Radian Corporation, "A  Theoretical  Description of
   the Limestone Injection — Wet Scrubbing Process,"
   A Report to NAPCA, HEW contract No. CPA-22-
   69-138, June 1970.
6.  LESSINO, R., "The Development of a Process of  Flue
   Gas  Washing  Without Effluent"  Journal of  the
   Society  of Chemical  Industry, Transactions  and
   Communication, 57, pp. 373-388, November 1938.
7.  RADEH, P.  C., "Bench Scale  Studies of CaSO* De-
   supersaiuration Kinetics," C-E Internal Reports.
8.  MAUBIN, P. G.t "The Combustion Engineering Air
   Pollution Control System,1" Instrument Society of
   America, 15th Annual Power  Division  Symposium,
   Dallas, Texas, May 22-24, 1972.
                                                 551

-------
  MAGNESIA SCRUBBING
           by

   Gerald G.  McGlamery
      Design Branch
Tennessee Valley Authority
  Muscle Shaols, Alabama
             553

-------
                          MAGNESIA SCRUBBING  _!/

                   Gerald  G. McGlamery, Design  Branch
           Tennessee Valley Authority, Muscle Shoals, Alabama
          Under  the  sponsorship of the Environmental Protection Agency  (EPA),

 the Division of  Chemical Development of the Tennessee Valley Authority  (TVA)

 has just  completed an  intensive design and cost study of magnesia scrubbing-

 regeneration processes  for sulfur oxides removal  from power plant stack gas.

 This  investigation, which has been carried out over the past tvo years, covers

 scrubbing systems using both aqueous slurries and solutions.


          When published shortly, the study report will be the fourth in a series

 of conceptual design and cost studies which TVA has prepared for EPA since 1967*

 The first two studies were concerned with the use of lime or limestone  as

 absorbents, which convert the gaseous sulfur oxides to solid compounds  (calcium

 sulfite and calcium sulfate) that are discarded.  These were called "throw-

 away" processes.  The third study included processes using aqueous ammonia

 solutions as the scrubbing medium and recovering  the sulfur oxides as ammonium

 sulfites which were converted to sulfate and used as an intermediate in the

production of fertilizer products.  This was the  first recovery  system examined

 in which materials could be produced for sale to  offset, at least partially,

the cost  of operation.
 I/   Paper presented at EPA Flue Gas Desulfurization Symposium, New Orleans,

Louisiana, May 1^-17, 1973-
                                  554

-------
          Processes recovering the sulfur oxides in a useful form are
potentially superior to the throwaway type because they do not generate
solid waste disposal problems and also offer the possibility that sales
revenue will reduce the cost of sulfur oxide removal.  It should be kept
in mind, however,  that recovery processes are generally more complex, may
be more expensive to install, and require a definite commitment to sell
the products produced.

          In recent years, numerous processes have been proposed for sulfur
oxide recovery and some of these are currently undergoing demonstration on
large scale (100 MW or larger).  The purpose of the EPA-TVA conceptual design
series is to subject the more promising of these to a detailed study in which
the best design is developed from available data; capital and operating costs
are estimated on a uniform basis; a market survey is made to estimate sales
revenue; total cash flow is related to economic promise; and needed research
and development are identified.

          Scrubbing with magnesium oxide slurry to form magnesium sulfite,
followed by decomposition to produce concentrated SOg (MgO is recycled) is
one of the more promising processes for sulfur dioxide removal.  Slurries
of magnesia are good absorbents; however, the most outstanding assets of
the concept are (l) the ease of separation of the sulfite salts formed from
the scrubber liquor, (2) the ability to regenerate and recycle the absorbent,
magnesium oxide, (3) the avoidance of a solids disposal problem, and (k) the
capability of separating, both financially and operationally, the power unit
scrubbing system from the chemical manufacturing and marketing function.  At
the same time, the process does require extra expense for (l) drying and
calcining the intermediate MgSOg and MgSOj^ formed, and (2) the apparent need
                                  555

-------
for two scrubbing stages on coal-fired units to avoid mixing fly ash with the



undiBsolved absorbent.  As with all aqueous scrubbing processes, stack  gas



reheating, if required, would also add expense.  The potential of the process,



however, is outstanding enough to merit demonstration on a 155-MW, oil-fired



power unit of Boston Edison.  This system, jointly funded by EPA and a large



group of chemical companies and utilities, started up in early 1972.





          In the regeneration of the absorbent, sulfur dioxide is released



at concentrations practical for conversion to sulfuric acid, liquified



sulfur dioxide, or elemental sulfur.  With the limited market for liquified



sulfur dioxide and the higher cost of conversion to sulfur, the product



receiving primary attention in the report is sulfuric acid.  Commercial



grades of acid including 98$ concentration and oleum are easily produced in



the process.





          Around the world, development work on magnesia scrubbing for



power plant stack gas has followed at least three major technological routes.



The Russians, Japanese, and Americans have concentrated on the use of mag-



nesium sulflte-magnesium oxide slurries having a basic pH; whereas a German



company, Grillo, has researched the use of an absorbent activator, manganese



dioxide, with the scrubbing slurry.  In addition, using technology associated



with sulfite pulping practice, at least one American company has also investigated



the use of magnesium sulfites in acidic solution so that simultaneous particulate



and 802 removal can be accomplished with a single scrubber in coal-fired unit



applications.





          Each of these three scrubbing schemes are given detailed review in



the present study and are described as follows:
                                   556

-------
          Scheme A • magnesia slurry variation—Wet scrubbing with magnesium
oxide-magnesium sulfite-water slurry to absorb S0g and form undissolved
MgSOyfiHgO plus some MgSOg.SHgO.  The MgSO^.fiBUO is thermally converted to
trihydrate and dried to form anhydrous MgSCU.  This material along with any
sulfate formed by oxidation is calcined with coke to generate MgO for recycle
and S02 for production of HgSO^ by the contact process.  A flow diagram of
Scheme A is shown in Figure 1.

          Scheme B - MgO-MnOg slurry variation—Wet scrubbing with magnesium
oxide-magnesium sulfite slurry containing a scrubbing reaction activator,
manganese dioxide.  The. sulfites, sulfates, and unreacted manganese  dioxide
are dried and calcined to regenerate the absorbent and activator with the
S0g rich gas being processed to HpSO, .

          Scheme C - clear liquor variation—Wet scrubbing of stack  gas to
remove particulates and absorb SOg simultaneously with an acidic solution
of magnesium sulfites, followed by separation of Insoluble fly ash and
liquor and addition of MgO to the liquor to precipitate MgSOo.&gO.   The
crystals of sulfite are then converted to trihydrate, dried, and calcined.
MgO Is recycled and SOg processed to acid.

          The above variations can be applied to multiple power units more
economically than with Individual power plants by taking advantage of a
concept called "central processing",  fy processing (calcining and acid
production) the dried sulfite material from several scrubbing operation*
in a single, large plant, a more efficient operation (higher annual  operating
time) can be derived and economy of scale can be achieved.  This concept is
not a technological variation, but deserves separate consideration (Scheme D)
to evaluate the economic merit of the idea which of course, can be used in •
other type absorbent processes as veil.
                                  557

-------
    Stack
     gag —
                  Water
Particulate
  Removal
  System
tn
vi
CO
                Thickener
                   I
                Ash to pond
        Recycled
          Pond
         Water
 To Stack

H
                                                Sulfur Dioxide
                                                  Absorber
                                                        Slurry
                                 Devatering
                                   System
                                              Liquor
                                                              Water MgO

                                                             11
                            Makeup System
                                   Dryer
                     Air

                     Fuel
                                                             MgSO,
                                                                               Recycle MgO
                                               SulfurIc Acid
                                               Plant and/or
                                               Liquified S02
                                                and/or
                                                Sulfur
                                                                               Calciner
                                                                                                 -Fuel
                             Figure 1  Flov Diagram - Magnesia Slurry Scrubbing-Begeneration

-------
          Some earlier preliminary investigations (1970) indicated the

possible use of a variation in magnesia scrubbing for N0x control.  Recent

vork (1972), however, indicates that no more than 10$ removal can be expected;

therefore, magnesia scrubbing should not be counted as a means of meeting new

Federal NO  emission standards.
          A


Study Assumptions

          Recovery process economics depend on several factors including

power plant size, type of fuel burned, sulfur content of fuel, operating

factor, plant location, unit efficiency, and unit status (new vs. existing).

For detailed design and cost estimating purposes, it was necessary to assume

a combination of conditions as a base case for both oil- and coal-fired units.

In the economic evaluation, the effect of variation in the major  parameters

was determined.  The basic conditions assumed are as follows:

          fewer unit size, MW                      500
          Sulfur content of coal, %                  3.5
          Sulfur content of oil, %                   2.5
          Ash content of coal, %                    12
          Heating value of coal, Btu/lb          12,000
          Heating value of oil, Btu/lb           18,500
          Boiler excess air and leakage, %
             coal-fired unit                        33
             oil-fired unit                         15
          Degree of dust removal, %                 99
          Degree of S0p removal, $
             slurry Schemes A,B,D                   90
             solution Scheme C                      77
          Boiler type                               Horizontal,  frontal-fired
          Plant location                            Midwest
          Capactiy factor, % of nameplate
             rating
             first to 10th year                     80
             llth to 15th  year                     57
             l6th to 20th year                      UO
             21st to 30th year                      IT
             Avg over life of unit                  1*8.5
          Air preheater exhaust temp,  °F           310°F
          Stack gas reheat temp,  °F                 175°F
          Unit heat rate, Btu/kWh                9,000
          Product storage, days                     30

                                   559

-------
Process Equipment

          The scrubbers, ductwork, and fans are the most expensive items
in a magnesia scrubbing process because they must handle the full flow of
gas (over 1,000,000 tons/day for a 1000-MW boiler).  The slurry or solution
processing, drying, calcining, and sulfuric acid units handle a lower through-
put of material depending on S content of fuel.  Stagewise scrubbing (see
Figure 2) will be necessary when using slurry scrubbing on coal-fired units to
keep the  majority of the fly ash from entering the drying and calcining opera-
tions.  For particulate removal, a venturi type device using clarified, circu-
lated water is chosen although  electrostatic precipitators and bag filters
could be used*  Electrostatic precipitators have shown variations in outlet-
loadings due to operating characteristics and time, and bag filters are
more expensive.

          In scrubbing sulfur oxides, slurry systems can utilize venturi,
mobile bed (plastic sphere type) or spray units.  For solution scrubbing
service (Scheme C}, plate and packed scrubbers might be added to the list
of acceptable devices, but consideration must be given to residual fly ash
carryover causing plugging.  In any case, corrosion and erosion protection
should be provided by linings such as rubber or polyester-fiberglass resins.

          At this time, mist elimination performance in slurry scrubbing
service is of concern with a variety of designs and materials of construction
currently in use or under study.  In this design, the chevron vane type device
constructed of a corrosion-erosion resistant material or coated is used.
                                  560

-------
               Figure 2
  PERSPECTIVE VIEW OF A MAGNESIA
 SCRUBBING SYSTEM FOR SO2  REMOVAL
ON A COAL-FIRED 500-MW POWER UNIT-
    TWO-STAGE VENTUR1 CONCEPT

-------
          Reheating can be accomplished by indirect steam heat exchange on



new units for which design provisions have been made in the steam cycle;



however, existing units are not likely to have excess steam available; there-



fore, direct fuel oil reheat is preferred.  Neither of these methods is the



most economical choice available, but the reliability of the indirect liquid-gas



heat exchange method considered in previous studies has become suspect.





          Solids separation in the magnesia  slurry processing area probably



can be accomplished best by first thickening the 10$ solids slurry to kOfi



and then centrifuging to a cake containing less  than 1556 free water.  Since



good test data are not available, separation by filtration cannot be ruled



out; however, cakes containing less than 15% water may be more difficult to



obtain.





          Although rotary type devices are being utilized for drying and



calcining in the Boston Edison demonstration project, discussions with



vendors have Indicated that possible greater efficiency and lower cost



could be obtained with fluid bed units.  In the absence of test data, some



doubt remains; however, fluid bed systems appear to be the better choice.





          The "dry" gas cleanup system for calciner off-gas and the sulfuric



acid plant utilize relatively well established technology; therefore, few



unforeseen problems should arise in these areas.  If desirable, the magnesia



process could easily be added to existing acid units which currently burn



elemental sulfur.
                                   562

-------
Investment Requirements
          Summarized investment under various combinations of conditions are
given for Scheme A and limestone-vet scrubbing in Table 1.  Depending on sulfur
in the fuel, fuel type, plant size, and status, the magnesia Scheme A  Invest-
ment varies from k-2Tf> higher than limestone-vet scrubbing; however, salable
product is produced rather than a waste material.  Scheme C has the lowest
investment requirement ($36.2/kW for base case) for coal-fired power unit
scrubbing systems; however, SOg removal for Scheme C may not be sufficient in
all cases to meet Federal emission standards for new units.  In addition, the
data supporting clear liquor scrubbing are limited; therefore, the scheme  should
not be considered as the leading process.

          A detailed, direct investment breakdown of the 500-MW base case
covering Scheme A is shown in Table 2.

Operating Costs
          A summary of average annual and unit operating costs at TOGO hrs/yr
operation under regulated economics are given in Table 3 for both magnesia
Scheme A and limestone scrubbing (low and high cost).  A detailed operating
cost breakdown for the Scheme A base case (500-MW, 3.5$ S  in coal, new unit)
is shown in Table h.

Evaluation Considerations
          Evaluation of recovery processes brings in factors such as product
marketability and price, return on investment and taxes, and project financial
promise, all of which make the analysis more difficult than for throwaway
processes.  It would be desirable, of course, that recovery methods show promise
of a net profit, but this  is not essential because recovery should be preferable
to throwaway, even at a net loss, aa long as the loss is  lower than for throw-
     systems.  The cost of limestone-wet scrubbing was used as the criterion  for
                                  563

-------
     Table 1

              Capital Requirements for Magnesia Scheme- A
                      And Limestone-Wet Scrubbing


     Conditions                     Capital. $/kW of power generating capacity

                                      Magnesia Scheme A        Limestone-Vet
                                       vet scrubbing             scrubbing

Base case - coal-fired units

 (500-MW, new power unit, 3«5$ S in
  coal,reheat to 175°F)                      ^3.5                  35.2

Exceptions to base case (coal-fired units)

  Existing power unit                        ^9.3                  39.9
  2% sulfur                                  37.6                  32.3
  5% sulfur                                  1*8.5                  37.8
  200-Mtf                                     58.4                  46.0
  1000-Mtf                                    33-1                  27.4
Base case - oil-fired units

 (500-MVf, new power unit, 2.5# S in
  oil, reheat to 175*F)                      2k.9                  21.4

Exceptions to base case (oil-fired units)

  Existing unit                              27.8                  24.8
  1% sulfur                                  19.8                  19.0
  4# sulfur                                  29.1                  23.U
  200-MW                                     33.4                  28.5
  1000-MW                                    18.8                  16.6
                                  561

-------
                                         Table 2   Process Equipment and Installation
                                    Analysis-Direct Cost for Scheme Aa (Thousands of Dollars)
Particulate
scrubbing
Equipment
Material 828
Labor 2*0
Piping & Insulation
Material 327
Labor 177
Ductwork, dampers, &
Insulation
Material 752
Labor Inc.
Concrete-foundations
Material 105
Labor Inc.
Structural
u, Material 135
% Labor 180
Electrical
Material 2l6
Labor Inc.
Instruments
Material 111
Labor 57
Faint
Material 66
Labor Inc.
S02
scrubbing

997
256

236
101


878
Inc.

115
Inc.

1*5
190

3*0
Inc.

135
75

60
Inc.
Slurry
processing Drying

*24
75

52
20


-
-

30
Inc.

26
31*

73
Inc.
(Additional
29
10

12
Inc.

*90
176

2
3


20
Inc.

38
Inc.

8
11

43
Inc.
Calcining
X
665b
215

2
3


44
Inc.

46
Inc.

1*
18

39
Inc.
HgO
slurrying

115
27

15
k


-
-

10
Inc.

3
4

32
Inc.
Instruments)
11
k

k
Inc.
32
11

5
Inc.
39
13

2
Inc.
New
HgSOij.
production

925
Inc.

410
Inc.


741
Inc.

188
Inc.

99
Inc.

207
Inc.

185
Inc.

66
Inc.
ngSO^ Optional by- Fuel oil
storage pass duct storage Total

163
Inc.

.
-


-
—

18
Inc.

3
k

14
Inc.

Inc.
Inc.

1
Inc.

122
22

Inc.
Inc.


454 Inc.
Inc. Inc.

21
Inc.

2
Inc.

10
Inc.

1
Inc.

Inc.
- Inc.

4729
1011

104*
308


2889
Inc.

571
Inc.

435
441

974
Inc.
_• _
5*3
170

216
Inc.
^l^^^BBI*^
Sub-total
  Direct costs  319*       3528
785
810
26k
2821
203
178   13,331
ft  Hev plant, coal-fired, 500-Mtf, 3.556 sulfur In coal, l,0to,000 scfln stack gas, 378 tpd B^SO^.  Inc. = included.

b  Includes most instrumentation.

-------
                                        Table 3   Average Annual and Unit Operating Costa
in
o*
o»
Conditions





Magnesia Scheme A
Average
Annual
Cost $


Unit
Operating
i t
Cost $/ton coal


Limestone-Wet
Low Limestone Cost,
On-Site
Average
Annual
Cost $
Solids Disposal
Unit Operating
Cost $/ton
Coal
Scrubbing

High Limestone Cost.0
Off -Site
Average
Annual
Cost $
Solids Disposal
Unit Operating
Cost $/ton
Coal
Base case - Coal-fired units
(500-MW, new power
S In coal, reheat

unit, 3.5^
to 175°F)
7,01*8,900
Exceptions to base case (coal-fired
Existing Power unit 7,762,500
5.0# Sulfur 8,066,600
200 m
1000 MW
3,870,700
10,635, too


5-37
units)
5.79
6.15
7.21
4.19


5,376,300

5,927,900
5,894,000
2,869,200
8,230,900


4.10 7

4.42 8
4.49 8
5.35 3
3*24 12


,621,500

,253,700
,861,700
,633,1*00
,883,100


5.81

6.15
6.75
6.77
5.08
                         JL
Base case - Oil-fired units
 (500-MW, new power unit, 2.5#
  S in oil, reheat to 175°P)
                       4,159,800
$/bbl oil
                                             0.83
                 3,343,600
$/bbl oil
  0.66
                                                                                                 JL.
4,112,500
                  $/bbl oil
0.82
  Exceptions to base case (oil-fired units)
    Existing Power Unit
    lf.0# Sulfur
    200 MW
    1000 MW
     7000 hrs/year operation

     Limestone  at  $2.05/ton and variable  on-site disposal costs  for calcium solids,  ranges  from $2.85/ton to $1.33/ton
4,51*8,800
4,973,500
2,305,600
6,317,100
0.88
0.99
1.12
0.65
3,755,100
3>747,300
1,836,700
5,160,400
0.73
0.74
0.89
0-53
4,566,000
5,046,700
2,046,900
6,848,000
0.89
1.00
0.99
0.70
    Limestone  at  $6.00/ton and $6.00/ton for disposal  of calcium solids

-------
               Table k   t Regulated Company Economics—Total Venture Average Annual
               Manufacturing Costs for 98% H2SO4 from Scrubbed Power Plant Stack Gasa
                            Scheme A-Magnesia Slurry Scrubbing-Regeneration
(500-mw new coal-fired power unit, 3.5% S
Annual quantity 	
Direct Costs
Delivered raw material
Lime
Magnesium oxide (98%)
Coke
Catalyst
Subtotal raw material
Conversion costs
Operating labor and
supervision
Utilities
Fuel oil
Steam
Heat credit
Process water
Electricity
Maintenance


134.1 tons
1,086 tons
763 tons
1,800 liters



39,200 man-hr

5,356,000 gal
440,000 M Ib
20,300 MM Btu
2,207,500 M gal
66,760,000 kwh

i/i fuel; 1 1 0,400 tons/yr 1 00% H2 SO4 )
Total annual
Unit cost, $ cost, $


16.00/ton
102.40/ ton
23.50/ton
1.51/liter



6.00/ man-hr

0.09/gal
0.55/M lbb
-0.40/MM Btu
0.03/M galb
0.006/kwhb

Labor and material, .06 x 21 ,732,000
Analyses
Subtotal conversion costs
Subtotal direct costs
Indirect Costs
Average capital charges at 14.9%
of initial fixed investment
Overhead
Plant, 20% of conversion costs


















2,100
111,200
17,900
2,700
133,900


235,200

482,000
242,000
(8,100)
66,200
400,600

1,303,900
85,000
2,806,800
2,940,700


3,238,100

561,400
Cost/ton
of acid, $


.019
1.007
.162
.024
1.212


2.130

4.366
2.192
(.073)
.600
3.629

11.811
.770
25.425
26.637


29.331

5.085
Administrative, research, and service,
1 1% of conversion costs
Subtotal indirect costs



Total annual manufacturing costs





forH2S04


Cost/ton
of coal
burned, $
5.371
308,700
4,108,200
Total
annual
cost, $
7,048,900
2.797
37.213

Cost/ton
of acid t$
63.850
aBasis:
  Remaining life of power plant, 30 yi.
  Coal burned, 1,312,500 tons/yr-9,000 Btu/kwh.
  Stack gas rehca t to 17 5° F.
  Power unit on-strcam time, 7,000 hr/yr; acid plant on-stream time, 7,000 hr/yr.
  Midwest plant location-1972 costs.
  Capital investment, $21,732,000; working capital, $505,600.
''Cost of utility supplied from power plant at full value.
                                                       567

-------
 comparison.  Both high cost  ($6/ton limestone, $6/ton solids disposal cost) and
 low  cost  ($2.05/ton limestone, variable cost on-site solids disposal) limestone
 systems were estimated using the same basis as used for the magnesia schemes.

          The basis on which the recovery process is financed is a major
 consideration in evaluating economic promise and acceptability.  If a power
 company finances the entire project, it can be assumed that the investment
would become part of the rate base on which the company is allowed to earn
what the regulatory authority regards as a reasonable return on investment.
If sulfur oxide removal, either by throwaway or recovery method, were to
increase operating cost, then the price of power to consumers presumably
could be raised to offset the extra cost.  Under such basis, sulfur oxide
removal (even by recovery) could be considered as necessary for production of
power just as is the boiler operation, dust removal, or cooling water system,
and the costs, therefore, passed on to the consumer.  It is true that rate
increases are often strongly contested and delayed, and that the full adjust-
ment may not always be allowed, so that the power company has the incentive
to avoid extra investment and expense.  In general, however, the power com-
pany has a more or less assured profit.  For this reason, there is little
risk and capital can be attracted at regulated rates of return.

          Since power companies generally are not familiar with chemical
production and marketing, there would be some advantage if a chemical company
built and operated the recovery process for a fee and marketed the products.
For a private, nonregulated company to enter into such an activity, however,
the project would have to be promising enough to attract the necessary
capital from investors.   It is difficult to say how much promise is needed
because this varies with the situation.  Generally, it is considered that
the projected cash flow (depreciation plus profit after taxes) should pay out
                                   568

-------
the original investment in less than five years or on another basis which takes
into account the time value of money, the interest rate of return after income
taxes should be about 15$.  For the relatively high investment required by sulfur
oxide recovery processes, this is a major hurdle.

          A characteristic of the central process concept is that features
of both regulated and nonregulated economics can be utilized to advantage.
If by cooperative arrangement, magnesia scrubbing and drying operations are
placed under power industry (regulated) economics, and regeneration and acid
manufacture are covered under nonregulated economics, the power company can
avoid any responsibility for acid production and marketing, and the chemical
company can reduce its capital responsibility to levels more likely to achieve
successful profitability.  By charging the power company for regenerated MgO
and also by selling sulfuric acid, enough revenue might be obtained for a
chemical company to Justify manufacturing acid from magnesium sulflte rather
than by the more conventional purchase of elemental sulfur.  The revenue
obtained for regenerated MgO would, of Bourse, depend on the resultant cost
to the power company for magnesia scrubbing as opposed to limestone-wet
scrubbing or other feasible alternatives.

Comparison and Profitability
          An important consideration for comparison under regulated
economics, and in profitability analysis under nonregulated economics, is
net sales revenue for the sulfuric acid.  A market review for the acid
resulted in the following conclusions:
1.  The growth rate of sulfuric acid production is about U-6# per year
    generally paralleling that of the phosphate fertilizer industry.
    Capacity from magnesia scrubbing-regeneration will most likely enter

                                   569

-------
    the market at a moderate to slow pace.



2.  The best end-use market appears to be the phosphate fertilizer industry



    as an acidulant for phosphate rock.  The product is used in many other



    applications, however, any of which merit consideration.



3'  The most promising locations for magnesia scrubbing-regeneration



    systems appear to be on waterways serving the areas where sulfuric



    acid is now heavily marketed.  Areas on the Ohio and Mississippi



    Rivers, and along the Gulf and East Coasts are prime spots.



k.  Sales price will be based on competition in each individual area



    plus flexibility of demand.  In those areas where by-product acid



    or low cost sulfur are available, competiton will be greatest.



5.  Expected net sales revenue after shipping and sales expense are



    deducted could average about $8.00/ton of 100$ acid for single-site



    systems and $12.00/ton for large central processing units.  In the



    better locations, these net backs and maybe more should be attainable



    through the 1970's.



6.  Long term marketing contracts appear to be practical since the



    likelihood of escalating sales revenue due to prolonged product



    shortages is not expected.






          Comparison under regulated economics--A comparison of Scheme A



with both high and low cost limestone scrubbing is given in Table 5-  The



values shown are the cumulative net annual costs over the life of the power



unit; thus, the values represent the total bill including return on invest-



ment and income taxes for particulate and SOg control over the power plant



life.
                                   570

-------
                        5.  Cost of Magnesia Scheme A vs. Limestoixe-Wet Scrubbing under Regulated Economies
             Conditions
                   Cumulative net annual costs, $ millions
                                                   Scheme A~                   Limestone-Wet  Scrubbing
                                                   __________   Low limestone process  coat"    High limestone process costc
Base  Case  -  coal-fired units
   (500-Mtf, new unit,  3.5$ S
   in coal,  ^8.556 avg. capacity
   factor  over 30 yrs, reheat  to 175°F)

  Exceptions to base  case  (coal-fired units)
   Existing unit (27  yr life)
   5% sulfur
   200-MW
   1000-MW
162.1
160.2
180.5
 91.1
237-9
136.2
133.4
148.5
 72.7
208.3
170.6
162.3
194.3
 82.6
283.2
Base Case - oil-fired units
   (50O-MW, new unit, 2.5# S in oil,
   48.5# avg. capacity factor over 30
   yrs, reheat to 175°F)

   Exceptions to base case (oil-fired units)
   Existing unit (27 yr life)
   4# sulfur
   200-MSf
   1000-MW
 95.9
 8k. 2
 93-7
111.0
 54.4
ilfl.6
 84.1
 93-8
 46.3
129.5
 94.2
 92.3
112.1
 47-7
154.3
a  Net sales revenue assumed at $8.00/ton of acid.

k  Limestone cost - $2.05/ton; on-site pond disposal of solids.

   Limestone cost - $6.00/ton; off-site solids disposal cost, $6.00/ton.

-------
          The  only magnesia  case with costs lower than the low cost limestone



system is a 1000-MW, coal-fired unit using Scheme C, the least developed



variation.  Units smaller than 3°0-MW would most likely use limestone scrubbing



if funding were under regulated economics.  Because the incremental cost of



producing additional acid exceeds $8.00/ton (net sales revenue for acid),



increasing on-stream time and higher sulfur content of fuel do not improve



the magnesia process economics.





          Profitability under non-regulated economics—Based on projected



revenue from acid sales alone, all magnesia cases examined have negative



interest rates of return and no payout.  If additional revenue in the form



of a fee equivalent to the cost of limestone-wet scrubbing or other compet-



itive  502 control method is charged by the chemical company for sulfur oxide



abatement, profitability can be derived.  Shown in Table 6 are payout periods



in years and interest rates of return in % for Scheme A assuming revenue from



both a fee and acid sales.






          As would be expected, the results depend on the size of the fee



charged; for a fee equivalent to a high cost limestone scrubbing process,



desirable profitability could be achieved in some cases and for a smaller



fee equivalent to a low cost limestone process, low profitability would



result in all cases.  Funding under this concept will probably be limited.






          Profitability of cooperative central process ventures—With the



separation of investment and operating responsibility and the advantage of



economy of scale for large central acid complexes, cooperative ventures



(Scheme D) between power companies and chemical companies are the best



route to financial funding of magnesia systems.  Given in Table 7 are the






                                   572

-------
Ul
-»j
w
   Conditions*

   Base Case - coal-fired units
      (500-MV,  new unit,  3.536 S in coal
      48.5# avg. capacity factor over
      30 yrs,  reheat to  175*F)

   Exceptions  to base  case (coal fired)
      Existing unit (27  yr life)
      5% sulfur
      200-MW
      XOOO-Mf
Base Case - oil-fired units
  (500-MW, new unit, 2.5* in oil,
   48.5# avg. capacity factor over
   30 yrs, reheat to 175°F)

Exceptions to base case (oil fired)
   Existing unit (27 yr life)
   Itjt sulfur
   200-MW
   1000-MW
for Pollution Abatement
Low
Payout
7.6
coal
iver
Ired)
7.7
7.7
8.3
7-1
7.2
1,
ver
red)
7.0
7-5
7.6
6.8
equivalent payment1*
Interest rate
, yrs of return, $
8.8
8.4
8.5
7.4
10.0
9.8
10.0
9.0
8.8
10.9
High equivi
Payout, yrs
5-6
5.7
5.4
6.7
4.8
6.1
6.0
5.8
7.1
5.3
                                                                                                   Interest rate
                                                                                                   of return,  j>

                                                                                                      14.9
                                                                                                      13.6
                                                                                                      15.7
                                                                                                      11.0
                                                                                                      18.1
13.0
                                                                                                         12.7
                                                                                                         14.0
                                                                                                         9.9
                                                                                                         15.8
      Bet sales revenue assumed at $8.00/ton of acid.
   b
      Equivalent to limestone-wet scrubbing cost assuming low limestone price, on-site pond disposal of solids.
      Equivalent to limestone-wet scrubbing cost assuming high limestone price, off-site disposal of solids.

-------
Table ?.
            Profitability of Central Regeneration-Acid Manufacturing Unit Under
            Cooperative Economics.*  Magnesium Sulfite Supplied from Combinations
            of New 200, 500, or 1000-MW Units Burning Coal with 3.556 Sulfur.
            Regulated Magnesia Scrubbing Costs Equalized to High and Cow Projected
            Limestone-Wet Scrubbing Process Costs.
                             Payout, years
                                     Interest rate of return, %
    Case
Units and size

200-MW equiv.
5 x 200-MW equiv.
10 x 200-MW equiv.
15 x 200-MW equiv.
Recycle MgO
at $25/ton

   None
   6.6
   5.2
   4.6
Recycle MgO
at $55/ton

    8.4
    3.4
    2.7
Recycle MgO1
at $25/ton

   Neg.
   8.2
  14.0
  17.2
Recycle Mgu
at &55/ton

     3-3
    26.6
    35-5
    1*0.6
500-MW equiv.
2 x 500-MW equiv.
4 x 500-MW equiv.
6 x 500-MW equiv.
Recycle MgO1
at $15/ton

   None
   9-9
   7.7
   6.5
Recycle MgOc
at $55/ton

    5-1
    3-5
    2.8
    2.4
Recycle MgO
at $15/ton

   Neg.
   0.3
   5.1
   8.7
Recycle MgOc
at $55/ton,

    14.4
    25-3
    34.1
    39-7
1000-MW equiv.
2 x 1000-MW equiv.
3 x 1000-MW equiv.
Recycle MgO
at $10/ton

   None
   9.9
   8.3
Recycle MgO
at $55/ton

    3.6
    2.9
    2.5
Recycle MgO
at $10/ton

   Neg.
   0.1
   3-5
Recycle MgOc
at $55/ton  .

    24.4
    33.0
    38.5
a  Nonregulated portion of system with 10 yr life; acid revenue - $12/ton.
b
   Equivalent to limestone-wet scrubbing costs assuming low limestone price, on-site
   pond disposal of solids.

0  Equivalent to limestone-wet scrubbing costs assuming high limestone price, off-site
   disposal of solids.
                                             574

-------
payouts and interest rates of return for Scheme 0 systems assuming revenue
from both acid and recycle MgO sales.  The price of recycle MgO must be such
that magnesia scrubbing cost does not exceed that of competitive limestone
scrubbing for the same power unit.  For a 500-MW, coal-fired unit, only
about $15-$20/ton could be paid for recycle MgO before exceeding the low
cost limestone system; however, approximately $55/ton could be paid if
competition came from a high cost limestone system.

          The results in Table 7 indicate that the smaller the power unit
supplying MgSO^ and the larger the acid complex, the better the profitability
which could be achieved.  A 3000-MW equivalent acid plant supplied by fifteen
200-Mtf units would show excellent profit making potential - 17.2$ interest
rate of return with $25/ton for recycle MgO and $12/ton for sulfuric acid or
1»0.6f> return for $55/ton recycle MgO and $12/ton acid.

Conclusions
          The more important conclusions derived from, this study can be
summarized as follows:
1.  Sulfur dioxide absorption by magnesia slurry scrubbing is effective
    and the major portions of the process as conceptualized utilize
    proven technology.
2.  Magnesia scrubbing, like limestone scrubbing, is not an effective
    means of NOX removal from power plant stack gas.
3.  Magnesia slurry scrubbing-regeneration has been tested in laboratory
    and pilot plant stages and at least one large scale demonstration is
    underway.
k.  Although limited experience is available to guarantee performance,
    equipment for commercial systems can be obtained at this time from
    vendors and fabricators.
                                  575

-------
5.  For moat U. S. fossil -fueled  power plants, achievable net sales revenue



    for recovered 9&f> sulfuric acid will probably average only $8-12/ton



    over the next decade or so; however, there will be applications where



    better netbacks are obtainable.  Competition will continue from other



    sources of by-product sulfuric acid and virgin acid made from low cost



    sulfur.



6.  Primary economic factors are investment, product volume (depending on



    power unit size and sulfur content of fuel), net sales revenue (from



    all sources), competitive cost of alternatives and basis of financing.



    Raw material, labor, shipping costs, on-stream time, and plant age



    are significant, but not nearly as Important as the primary factors.



7*  Under total regulated financing, magnesia systems can compete with



    limestone scrubbing on larger (toO-MW or greater) power units.  The



    limestone process would be favored in rural areas (low cost limestone



    and space for solids disposal) whereas the magnesia scrubbing-regeneration



    process would appear more desirable in crowded metropolitan areas.



8.  Total nonregulated industry financing and operation appear unlikely;



    however, with a large fee for pollution abatement and large size units,



    such funding can be considered.



9*  A cooperative venture between several power companies and a chemical



    company with each supplying capital for and operating their portion



    of the process, appears to be a good way to fund a magnesia system.



    It will be necessary for the regeneration-acid plant to charge a



    service fee for MgO processed from MgSCU in order to obtain sufficient



    revenue for desirable profitability.
                                  576

-------
10.  Thus far, Interest in the magnesia scrubbing-regeneration process has



     centered more on replacement of sulfur as raw material in existing



     sulfuric acid plants with existing markets rather than for added



     capacity to meet increasing acid markets.



11.  There are a limited number of locations that can support a central



     process installation.  The Midwest, along the Ohio and Mississippi



     Rivers, and the Gulf and East Coasts are prime targets.



12.  For short range shipping distances (O-50-miles), the cost of shipping



     MgSOo and MgO between sites in a small part of the total process cost



     and will not greatly influence process application; however, as



     distances exceed 100 miles, shipping cost becomes much more significant.





Additional research and development of the process should be performed



primarily on the demonstration level to determine effect of process factors



such as contamination build-up over long periods of MgO recycle, corrosion-



erosion of construction materials, scaling difficulties, and adaptability



to power plant operation.  Some work on oxidation, crystal growth,



effectiveness of additives such as manganese dioxide and the manufacture



of sulfur in the calciner should be performed on the bench or pilot levels.
                                  577

-------
       OPERATIONAL PERFORMANCE OF THE
    CHEMICO BASIC MAGNESIUM OXIDE SYSTEM
        AT THE BOSTON EDISON COMPANY
                     PART I
                       by

                George R. Koehler
         Chemical Construction Corporation
               New York, New York
This paper is Part I of a two part paper and  supplements
Part II of this paper being  presented by Mr.  C. P. Quigley
of Boston Edison Company.
                         579

-------
The work upon which this publication is based was




performed pursuant to Contract No. CPA 70-114




with the Environmental Protection Agency.
                      580

-------
                 OPERATIONAL PEFORMANCE OF THE
              CHEMICO/BASIC MAGNESIUM OXIDE SYSTEM
                 AT THE BOSTON EDISON COMPANY
              	PART I	
                          G. R. Koehler

 INTRODUCTION
   The necessity for control of sulfur oxides emissions to the atmosphere
 has become increasingly apparent.  The deleterious effect of this compound
 on human health and the environment has been documented.  The growing de-
 mand for energy compounds the problem.  Total SO2 emissions in 1965 were
 25 million tons per year.  It is anticipated that within the next decade, if no
 control is imposed,      these emissions will rise to 75 million tons per year
 in the United States.  Of this amount, two-thirds will originate from the  pro-
 ducts of combustion of fossil fuel fired power generating stations.   Of the avail-
 able technology, one of the  most promising seems to be flue gas scrubbing at the
 generating facility itself.
   In its search for  suitable technology to control sulfur oxides emissions
 CHEMICO developed a process scheme which uses  the well  established techno-
 logy of acid gas absorption  by alkaline scrubbing media with precipitation of the
 sulfite formed as a solid product. Further process steps include dewatering of
 the solid product and eventual reclamation of the alkali with recovery of the SO2.
   In June 1970, the United  States Environmental Protection Agency and  the
 Boston Edison Company agreed to provide the funds for a large prototype Sulfur
 Dioxide Recovery Plant using the Chemico/Basic MgO Sulfur Recovery Process.
 These integrated facilities were comprised of an SO2 absorption plant at Boston
 Edison's Mystic Station and a regeneration facility  at Essex Chemical's  Rumford,
 B.I. Sulfuric Acid Plant.  Construction at both plants was  completed by April,
 1972.
   Capital funds for the absorption system were provided by the Boston  Edison
 Company while E.P. A. furnished funds for installation of the magnesium oxide
 regeneration system and appropriate acid plant modifications as well as  opera-
ting funds for the absorption/regeneration systems.
                                   581

-------
   The SO^ absorption plant is designed to remove 90% of the sulfur dioxide
 formed when burning fuel oil containing two and one half percent sulfur.  The
 regeneration facility is designed to recover the alkali in a form suitable  for re-
 cycling to the absorption facility and send the recovered SC^to  a  sulfuric acid
 plant for the production  of 98% sulfuric acid.
   Operations to date have demonstrated each of the process steps.  SO  re-
                                                                    £
 moval efficiencies of 90% or better have been consistently obtained despite the
 use of a lower sulfur fuel oil than originally expected.   Several hundred tons
 of commercial grade sulfuric acid have been produced and marketed in the con-
 ventional manner.  Operations have, however, been intermittent due to the nec-
 essity to make numerous relatively  minor changes and adjustments to  the system.
 Yet to be demonstrated is the long term operability of the process and the ability
 to use Magnesia which has been recycled many times.  Present efforts are being
 concentrated on entering into long term operations in  order  to ascertain  system
 availability  and reliability.
   Problems associated  with process consideration are addressed in this Part
 I.  Mr. C. P. Quigley of Boston Edison will describe mechanical problems en-
 countered, corrosion—errosion experience and related equipment malfunctions
 in the second part of this paper.
GENERAL PROCESS DESCRIPTION
   Chemico/Basic's Magnesium Oxide System for the recovery of sulfur dio-
 xide from power plant flue gases is shown schematically in Figure  1.
   This process, which utilizes the sulfur dioxide absorption characteristics
 of an aqueous slurry containing magnesium oxide,  magnesium sulfite and
magnesium  sulfate is comprised  of five primary operations:
                 Absorption System
                 Centrifuge System
                 Dryer  System
                 Magnesium Oxide System
                 Calcination System
                               582

-------
Absorption System
   The flue gas containing sulfur oxides enters a venturi absorber (Figure
2) and contacts the absorbing media which is an aqueous slurry magnesium
oxide,  magnesium sulfite and magnesium sulfate.  The process of SO2 re-
moval  is explained by conventional mass transfer principles. The venturi
absorber can be shown by analogy to correspond to a co-current packed vessel.
   In the scrubber, the liquid slurry is injected and flows on surfaces over
which passes an accelerating gas stream.  The high velocity gas  passing over
the liquid causes wave motion on those surfaces.   The waves increase in ampli-
tude finally dispersing as fine droplets in the gas stream.  Thus, the whole mass
of liquid  is dispersed in the  form of atomized droplets.
   In the process described  in this paper, the dispersed droplets have a median
size of 400 micron and the surface area available for mass transfer averages
13 ft 2per ft3 of gas.  This absorption surface is dispersed thru and flows with
the gas stream eliminating the problems of plugging associated with conventional
packed towers (the surface area per unit volume is approximately equivalent to
dumped 3 inch rachig rings). Due to system dynamics this surface area relation
is relatively invarient over wide turn-down ratios in the  venturi scrubber and
can be used as an approximation over the power plant boiler's operating range
of 40 to 155 MW.
   Similarly, system efficiency can be predicted with a fair degree of accuracy
by substituting a Sherwood Number of 2 (SH =2) in conventional mass transfer
relations.  Using this estimate a maximum efficiency of 96% can be predicted
for the system, deviations from that removal efficiency being caused by equili-
brium partial pressures  of SO^ over the  droplet surfaces greater than zero in
the dynamic system.
   To date, the absorption system has operated satisfactorily. Sulfur dioxide
removal efficiencies of 90% or  greater have been consistently obtained with
SO2  concentrations in the outlet flue gas being generally 90 ppm  or less measured
by both a DuPont 460 process analyzer and wet chemical methods.  No scaling
or pluggage in the venturi absorber has been encountered despite almost conti-

                                  583

-------
 nual operation of this equipment in a recycle mode during the duration of the
 operations at the Mystic Station.
 Centrifuge Systern^
   A bleed from the absorption system (Figure 3) enters the centrifuge where
 the solids in the slurry are separated and the centrate is returned to the  absorp-
 tion system.  This bleed stream is controlled to maintain a constant solids con-
 tent in the recycle slurry and to remove product magnesium sulfite  and any un-
 reacted magnesium oxide and precipitated magnesium sulfate.  The system is
 operated so that the absorbed SO  is removed as an equivalent amount of magne-
 sium sulfur compounds. A Bird solid bowl centrifuge is used to dewater the
 solids preparatory to drying. Except for a few isolated incidents of shear pin
 breakage, this item of equipment has operated satisfactorily, yielding 50% or
 greater removal of the  solids in the slurry bleed stream.
 Dryer System
   The wet centrifuge cake  containing magnesium sulfite, magnesium oxide
 and magnesium sulfate  plus carbon and other solids removed in the  venturi
 absorber-centrifuge system is passed to a rotary counter-current dryer to re-
 move both unbound water and water of crystallization.
   The dryer with its associated feed conveyor and product conveying system
 has been the most frequent  cause of the  operational difficulties which have caused
 shut-down of the complete system.  The problems encountered have ranged from
 an initial complete carry-over of dryer  product into the off-gas due to excessive
 dryer gas velocity to an inability to convey the dryer product resulting from
auxilary   equipment breakdown.  In order of occurrence these problems and their
 subsequent solutions have been:
       1.   Total carryover of dryer material in the exit gas stream.
           The dryer was designed to process a material of much
           greater crystal size than encountered in actual operation.
           High draft resulted in a gas  velocity which exceeded the
           pneumatic transport velocity of the dried material.   Dryer
           action placed all of the dryer inventory in the air stream.
           All lifter flights were modified or removed in order to re-
           duce internal dusting in the dryer.  Normal operating draft
                               584

-------
was reduced from a design of 1" to a range of 0.1" to
0.01" (this caused subsequent problems in draft con-
trol) .
Adherance of centrifuge cake to dryer walls causing
plugging of the dryer.  This was caused by a change
in the consistency of the centrifuge  cake from a dry,
fine  sand-like material to a wet, fluid mud-like material.
Increasing exit gas temperatures to increase feed end
surface temperatures resulted  in granulation of the feed
(this problem of adherance was a recurring one and was
associated with other properties of the centrifuge cake as
described in later sections).
Granulation of the dried material resulted in a  nonuniform
product.  Many of the granules were several inches in dia-
meter and jammed product materials handling equipment.
The cross screw conveyor and  weigh belt at the top of the
product silo were bypassed  and a lump crusher was installed
at the boot of the bucket elevator to reduce the size of any
large granules to 1" or less.
Disintegration of the hydrated crystal during drying caused
excessive production of "fines" (Figure 4).  Heavy loading
of fines in the dryer off gas caused excessive carryover of
particulate matter from the cyclone emission control equip-
ment installed with the dryer.  Dryer off-gas was ducted to
introduce it to the venturi absorber after  initial clean up in
the  cyclones.  The venturi absorber, operating at a five inch
pressure  drop across  it's throat provides satisfactory parti-
culate emission control of the dust  carryover from the dryer
Further complications were anticipated as the dryer off-gas
was expected to furnish 15°F of reheat to the saturated flue gas
after sulfur oxide removal.  With the dryer gas now diverted to
the inlet of the  absorber during normal operations, it now enters
the stack at saturation temperature. To date no raining or pre-
                         585

-------
     cipitation from the stack has been noted despite exit
     gas velocities  of nearly 80 ft.per second. A careful
     watch is being kept over this part of the operation and
     initial design of a reheat system has been completed
     should it be needed.
5.   Dust  accumulation on a baffle installed over the spill
     back  bin of the dryer built up to restrict  the flow of gas
     from the dryer.  Initially, this material was manually
     pushed from the  shelf.   Recently the feed screw conveyor
     has been lengthened to introduce the centrifuge cake further
     into the dryer and a dust collection  system was installed to
     transport dust  accumulated in the spill back bin to the product
     silo,  allowing removal of the baffle.
6.   Addition of cyclone underflow to the centrifuge cake fed to the
     dryer caused setting of the centrifuge cake accompanied by
     fouling of the screw and overloading of the dryer feed conveyor
     motor. Cyclone  underflow was diverted  from the feed screw con-
     veyor; recently a dust collection system  was installed to remove
     cyclone underflow to the product silo.
7.   Operation with recycled MgO caused a reoccurrence
     of dryer feed sticking on the dryer walls which eventually
     pluged the dryer. It is felt that this was caused in part by a
     higher percentage of unreacted MgO in the centrifuge cake due
     to inadequate slaking of the recycled MgO produced in the early
     reclamation attempts.  Continued efforts to improve the quality
     of the recycled MgO in the calcination facility, installation of
     "knockers11 on  the dryer feed end, provision for heating the
     MgO  slurry tank  to improve  the slaking of the recycled  MgO
     (Table 1) appears to have eliminated this  problem.
                        586

-------
 Magnesium Oxidejystem^
    The anhydrous magnesium sulfite and magnesium sulfate produced in the
 dryer  is conveyed to a storage silo before transportation by truck to the re-
 covery acid plant.  The same transportation facilities are used to return re-
 generated magnesia to the  magnesium oxide silo at the power plant. No pro-
 blems  of pluggage of these silos has been encountered to date.  The measured
 rate of rehydration of the magnesium sulfite on standing in the atmosphere is
 relatively low, Figure 5.  The "lump" crusher installed upstream of the silo
 allowed the passage of very large pieces (1-1/2" x 2IC).  These pieces proved
 difficult to handle in the pneumatic  trucks and Rumford's conveying system.
 A scalping screen was installed in the truck loading chute and oversize material
 is  conveyed to the load-out belt by additional conveying  equipment.
    Recycled alkali absorbent and make up magnesium oxide are  fed with water
 to  an agitated tank where a slurry is prepared for introduction to the recycle
 system manually on pH control. Numerous pluggage  problems were traced to
 tramp  material entering  at this point.including granules  of magnesium sulfite
 not completely removed from the truck in the transport operation. Initially,
 system protection was provided by  interposing a vibratory screen at the dis-
charge  of the weigh belt feed.  Eventually the necessity to slake the recycled
 MgO at higher temperatures in order to increase its reactivity required the
 addition of an MgO tank equipped with seal legs to prevent steaming in the weigh
 house.  This device now  serves as  a "tramp" materials  trap.
 Calcination System
    The dry product transported from the power plant  is received, weighed and
 pneumatically conveyed to a storage silo. It is fed to a direct fired rotary cal-
 ciner at a metered rate,  and calcined to generate  sulfur dioxide gas while re-
 generating magnesium oxide.  Coke can be added to provide a reducing atmos-
 phere to reduce the residual magnesium sulfate to MgO and sulfur dioxide.  The
 hot flue gas containing sulfur dioxide and MgO dust enters a hot  cyclone where
 essentially all the MgO dust is returned to the calciner.  The flue gas then
 inters  a venturi scrubber for final  MgO dust cleaning.  At the same time the
     is cooled and adiabatically saturated.

                                   587

-------
   At the Essex Chemicals installation, the resultant product from the
recovery of the sulfur dioxide is 98% sulfuric acid.  The saturated flue
gas is cooled in a direct contact cooler to meet the requirements of the
acid plant water balance. The cleaned, cooled flue gas then directly enters
the drying tower of an existing 50T/D acid plant.
   The regenerated magnesia is cooled, conveyed to the magnesia storage
silo and recycled back to the power plant site for reuse.
   As would be expected in a plant of this complexity numerous operating
problems occurred.  These problems and 'their solutions were:
       1.   Excessive leakage at the seals of the rotary calciner
           preventing the attainment of the required neutral or
           reducing atmosphere.  All seals had to be remachined
           to provide minimum clearances.  Careful attention was
           required to  see  that all view and instrument ports are
           closed during operation.  Finally a new seal was designed
           and installed to  completely correct the problem.
       2.   Granular  nature of dryer product caused material handling
           problems which required the installation of heavy duty
           belts on weigh feeders to prevent ripping of the belts.
       3.   Balancing the draft requirements of the calcination facility
           and acid plant caused "puffing" at feed hood and weigh house.
           Modification of the gas exit transition of the calciner to provide
           smooth gas  flow was one of the steps taken to reduce pressure
           loss.
       4.   Gritty nature of calciner product caused unreactivity in reuse
           at Boston Edison.  Installation of pulverizing equipment to
           provide a grind  of 100% thru 100 mesh and 70% thru 325 mesh proved
           satisfactory.
                                 588

-------
        5.   Tramp materials in calciner product caused
            breakage of pulverizer. Installation of magnetic
            separator and vibrator screen at screw conveyor
            discharge eliminated this.
        6.   Severe dusting in calciner obscured flame, tripping
            flame safety controls and interrupting operations.
            Installation of secondary "flame scanner" and exten-
            sion tubes partly alleviated problem.  Eventual  employ-
            ment of full time operator on the firing platform during
            operation reduced the number of shutdowns.
        7.   Inability to use coke.  Initial  charge of coke received was
            off specification, and contained 47% ash.   Fear of contami-
            nation of  the calcined product and the acid plant precluded
            its use.  For several months the calcining facility was
            operated  without the introduction of additional carbon and
            attempts  were made to adjust the fuel ratio to compensate;
            this resulted in a harder "burn" of the product magnesium
            oxide.  Finally a coke of less than 10% ash was substituted
            and coupled with the improved operations detailed above
            produced  dramatic results.  A comparison of the two cases
            is given in Table 2.
DESCRIPTION OF INSTALLED FACILITIES
   Boston Edison's Mystic Station located in Everett, Mass, is presently
comprised of six oil-fired power generating units having a total rate capacity
of 619 MW.  All units are equipped with electrostatic  precipitators which are
presently de-energized as the station burns fuel oil exclusively instead of
coal.  An additional unit which will be nominally rated at 600 MW is now under
construction.
   The Chemico/Basic Magnesium Oxide System is operated on Unit No. 6 of
this installation.  This unit has a rated capacity of 150 MW and was placed  in
commercial operation in May 1961.  The boiler is a Combustion Engineering
controlled circulation unit with a rating of 935,000 pounds per hour continuous
                              589

-------
 capacity at 2,150 psi max.  The unit burns 9700 gallons per hour
 of No. 6 fuel oil at full rated capacity.  Design criteria for the scrubber
 installation are given in Table 3.
 PROCESS  CHEMISTRY
   Magnesium oxide is introduced into a mixing tank with water where con-
 version to the hydroxide commences.  This mixture is added to the recycle
 stream of  the venturi absorber where it is contacted with the entering flue
 gas containing sulfur oxides.  The reaction between these two components
 produces principally magnesium sulfite, some  of which is  oxidized to magnesium
 sulfate.  Conventional representation  of the chemical reactions is shown in
 Figure 6.
   A better representation of the postulated reaction mechanism shown in
 Figure 7 which indicates that the formation of the magnesium sulfite proceeds
 thru the reaction of magnesium ion in solution and sulfite ion in solution.   The
 product of this reaction is a salt of low solubility (MgSOQ solubility 15 gm/liter
                                                      o
 versus MgSO^ solubility of nearly 1800gm/liter).  As mentioned in the previous
 section a bleed stream is sent  to a centrifuge where the crystalline magnesium
 sulfite is removed  in a quantity sufficient to balance the incoming sulfur oxides,
 while the centrate (a slurry containing magnesium sulfite nuclei) is returned to
 the absorption unit.
   Upon initiation of operations from a cold start, a major change in the con-
 sistency of the centrifuge cake takes place  after a period varying from 8 to 16
hours.  The cake changes from a dry, fine sand-like consistency to a thin mud.
Quantitatively, during the initial period  the centrifuge cake may contain as much
as 65% crystals of  size 40 to 80 micron  while in operations after the first several
hours the particle size of the centrifuge cake is less than 15 micron.
   It was also determined that  the state  of hydration of the  crystalline magnesium
 sulfite varied between these two conditions. The  larger particle size material
was associated with a higher percentage of MgSOg. 6H-O while the centrifuge
 cake containing more of the finer crystals contained approximately equal amounts
of MgSO3. 6H2O and MgSO,. 3H2O.   The distribution of this latter case is  shown
 in Table 4.
                               590

-------
   Initially, the slurry had been held at 120°p for several days with no
flue gas scrubbing.  The complete absence of MgSO3.6H2O in the slurry
confirms that at temperatures above U0°p only MgSO3-3H2O is stable,
and the hexahydrate formed has a metastable existance.
   The proposed mechanism for this phenonema is shown in Figurte 7.
The existance of the (Mg. 6H2O)   coordinate ion is known and has been
identified by infrared spectroscopy.  Apparently this material dehydrates in
solution at elevated temperatures to form the trihydrate.  A proposed alter-
nate path shows the formation of the coordinate ion 3.3H2O.
   In the system as currently operated, the compounds crystallize as dis-
tinct entities.  The magnesium sulfite hexahydrate as a rhombic crystal and the
trihydrate of the same salt crystallizes as a trigonal pyramid.
   Operating difficulties have been associated with this transition of the  centri-
fuge cake.  However, adherance of dryer feed to the dryer walls has also
occurred while drying a centrifuge cake containing 40 to 80 micron size  crystals.
Dryer heat load is not adversely;affected by the transition during start-up as the
MgSOg.3H2O contains half as much water of crystallization as the hexahydrate
and therefore an equivalent amount of unbound moisture can be removed in the
dryer without affecting process economics.
   A comparison between the average stream analysis obtained during operations
with virgin magnesia and recycled regenerated MgO produced prior to sustained
coke introduction is given in Table 5.  Recent operations  incorporating all the
improvements outlined thus far have shown a return to a stream analysis
similar to that  for virgin MgO; most importantly, a reduction in unreacted MgO
in the centrifuge cake to 1. 4% instead of the higher average value indicative of
hard burned magnesia.
FUTURE INSTALLATIONS
   With limited operation of the plant and lack of data on the effect of repeated
recycle of the absorbing alkali  it is difficult to determine what other  modifica-
tions might be incorporated in a new plant.  Based on the initial operating ex-
perience at Boston Edison's Mystic unit it appears that future installations
should  have additional "lump breakers" and screening equipment located at
                           591

-------
 at strategic points in the plant. These should be installed at the MgO tank in
 the absorber recycle stream at the dryer product discharge and the magnesium
 sulfite storage silo discharge.
     The type of centrifuge cake encountered up to the present time could be
 handled readily in a dryer of different design. A dryer incorporating internal
 chains, external rappers,  lowered gas velocity and provisions to reduce the
 internal dust recycle would be suitable.
     The calcining facility should incorporate pulverizing equipment of more ro-
 bust design in order to ensure more continuous operation with a grind suitable
 for reuse in the absorption system.
 POTOMAC ELECTRIC POWER CO. - DICKERSON #3
     The previously mentioned provisions have been incorporated into-the
 pollution control  facility currently being constructed for Potomac Electric Power
 Co.  This plant is a coal fired unit and the venturi absorption system has  been
 designed to handle  the flue gas equivalent to 100 MW from the boiler (Figure 9).
     Flue gas containing SO2 and fly ash passes into the first stage of a two stage
 venturi scrubber where fly ash is  removed  using recirculated water as the scrub-
 bing media. A bleed stream from the scrubber is thickened  to concentrate the
 fly ash as a slurry underflow which is pumped to a disposal area.  Overflow from
 the thickener is returned to the scrubber circuit for reuse.  A bypass  flue gas  duct
 has been installed around the precipitator to the scrubber absorber.   This will
 allow the introduction of clean flue gas containing little fly ash into the first stage.
Dampers have been provided in these ducts which will allow the scrubber  absorber
 to handle either flue gas partially  cleaned in the precipitators or flue  gas  coming
directly from the air preheaters.   Thus the operation will provide the means of
evaluating wide ranges of particulate emissions on the magnesia slurry system for
the design of new plants and retrofit systems.
    Flue gas leaving the particulate removal system enters the second stage
of the two stage venturi where it is contacted with the aqueous slurry  containing
magnesium oxide, magnesium sulfite and magnesium sulfate  and from that point
follows the process description of an oil fired magnesium oxide system.
                               592

-------
   In addition to the incorporation of additional size reduction equipment in
this operation the principal difference is in the design of the dryer for the
system. The dryer supplied is a co-current unit, equipped with flights  of
internal chains, some external rappers, toothed lifter flights and a diameter
sufficient to reduce the gas velocity to one half that obtained at the Mystic
Station.
   While it is anticipated that any additional advantages incorporated at
Boston Edison will be added to  the Potomac Electric  facility it appears  at this
time that they will be minor ones.
                                593

-------
                       Test Coupon Location
FIGURE  1

-------
                                  ANNULUS
                                  SPRAYS
                                                                                   CONE
                                                                                   WASH
           TANGENTIAL
              NASH
          CLEAN GAS
           OUTLET
    TO STACK
INTERMITTENT
   MIST
 ELIMINATOR  L—  X
   SPRAYS      "
                                         --NORMAL LIQUOR LEVEL
                                                           PUMP
                                                          SUCTION
                    Figure 2.  Sectional view of absorber at Boston Edison Company. Boston, Mass.

                                                     595

-------
                                                                                            KOUCEO- DRAFT FAN
         CONVEYOR
tn
(0
0
                                                                                                                                                                     ELEVATOR

                                                                                                                                                                         CONVEYOR I

                                                                                                                                                                            f EffiH FEEDER
     FMi CAL.CIIIWG
         SYSIE»
                                   FROM DISCHARGE EXISTING MUCED4RAFT FAN
•OTHER
LIQUOR
PUB'S
                                                                                         CEXTRIFUGE STSTEN
                                                                                                                                    DRYER SYSTEM M»j + IfSCM + «fO)
                                                           Figure 3. Process flow<9uoram toi magnesia slurry SOj recovery system
                                                                    at Boston Edison Co., Boston. Mass.

-------
           L  JrystalE -  plus'200 rsesh  'i-accion
                       IOOX
Aiii.yc'rons  I.eSC. - plus 200 :nesh  Tractiou dried
                  IOOX

                    Fiqure 4

                      597

-------
tn
(£>
CO
      g
      u

      S.

      of
                                    8     10    12    14     16     18     20     22     24    26    28    30
                       Figure 5. Percent water absorbed by anhydrous MgSOs on exposure to air.

-------
                      CHEMISTRY  OF MAGNESIA SLURRY S02
Ul
to
to
                            RECOVERY PROCESS




ABSORPTION



     MAIN REACTIONS



           MgO + SO   +  3  HO  •*  MgSO  • 3 H 0
                   —       +•          «j      f,

           MgO + S02  +  6  H20  -*




     SIDE REACTIONS
S09 + H~(
H £
2 + MgO
-T + 7 H (
3 -* Mg(HS03)2
-> 2 MgS03 + H2
) -> MgSO. - 7
               MgO +



                     + 1/2 02 + 7 H20  -v  MgS04
                                                         Figure 6

-------
                 POTOMAC ELECTRIC POWER CO. PROTOTYPE

             PRECIPITATOR/SCRUBBER -  ABSORBER

          MgO  ADDITIVE SYSTEM  FOR S02 RECOVERY
                   SCHEMATIC PROCESS FLOW SHEET
                             BY-PASS
                                                          SCRUBBER/ABSORBER
                                                               STACK REHEAT
                                                               ALTERNATIVES
                                                              • FUEL BURNERS
                                                              • STEAM COtLS
                    ELECTRO-STATIC
                     PRECIPITATOR
       TO DRY ASH HANDLING SYSTEM
                                                                    DUST COLLECTOR
                                                    CENTRIFUGE
                          MgO I  i
                          SILO '  *
 RECYCLED
POND WATER
                                  MOTHER
                                  LIQUOR
                                   TANK
                     SLURRY
                     TANK
TRANSFER
 TANK
                                                                               CRYSTAL
                                                                                DRYER
          MgO FROM ACID PLANT
                                                                                             MgSO3 TO ACID PLANT

-------
                        TABLE 1
Effect of slaking temperature on recycle MgO utilization
Slurry Tank
Temperature

   58° F

  108°

  180°
MgO Hulk
Density
      Ib/ft3
32.3  "
                   Unreacted MgO In
                   Centrifuge Cake

                        17*
                         1.
                       TABLE 2
Effect of coke addition  (4- day continuous operation with coke
                         feed interrupted for 12 hours.)
With coke addition

Without coke addition
      Product Bulk
      Density

       2^.7 lb/ft3

       68.2 lb/ft3
                                                SQ2 Rating
                              21.9
                               601

-------
                                Table 3
                     UNIT NO. 6 S00 RECOVERY SYSTEM
  OPERATING CRITERIA
            GAS FLOW, ACFM                             425,000




            GAS TEMPERATURE,  °F                             300




            DRY GAS,  LB/MIN                              21,369




S           WATER VAPOR,  LB/MIN                           1,620




            S02,  LB/MIN                                   63.2




            FLY ASH,  LB/MIN                                0.91




            S02,  PPM  (DRY BASIS)                          1,410




            FLY ASH,  GR/SCFD  (DRY BASIS)                 0.0228




            FUEL  OIL, CONSUMPTION,  GAL/HR                9,700




            FUEL  OIL  SULFUR CONTENT,  %                     2.5

-------
                        Table 4  OBSERVATION OF CRYSTAL SPECIES DISTRIBUTION
  Sample
                  % Solids
  07202-05-AT08
  07202-09-AT08    6.52
§ 07202-11-AT08
LO
  07202-14-AT08
07212-02-AT08
0712-06-AT 08
+ 200 Mesh
  {Vol.  %)
4.39
6.52
8.03
8.82
6.25
6.3
0
0.65
1.45
1.05
0.65
0.55
  Composition {%)
      - 200 Mesh
•MgS03        MgS04
Tri    Hex
100
50
51
56
43.3
49.5

50
34.
24.
40.
36.


6
3
9
5


4.
6.
4.
6.


4
4
0
2
20

12
16.
14.
18.



2
0
0
% Moisture

Centrifuge   Dryer
Cake        Product
                                                                                     8
                                                                                       24

                                                                                       13

-------
        Table 5   STREAM ANALYSIS
     S02 Abatement  System,  Mystic Station #6
Slurry
  Solids
  MgO
  MgSO:,
  MgSOjJ
  Filtrate
Centrifuge Cake
  Solids
  MgO
  MgSOo
  MgSOif
  Centra te
  Solids

Dryer Product
  Solids
  MgO
                                 Average Compositions
                          Fresh MgO  Feed      Recycle MgO Feed
10;8
 6.2
55.51
85*0
 1.36
57.85
 Mr. 12:
87.2
 2.6
8^,63
11.9
 6.83
59.61
                                                      10.8
83.0.
58.03
 2.59
                        . 67
95.66
  Acid Insoluble
  .17
 7.68
 0.55

-------
            OPERATIONAL PERFORMANCE OF THE
            CHEMICO MAGNESIUM OXIDE SYSTEM
             AT THE BOSTON EDISON COMPANY
                         PART II
                            by

                   Christopher P. Quigley
                   Boston Edison Company
                   Boston, Massachusetts
This paper if Part II of a two part paper and supplements
Part I of this paper being  presented by Mr. G. Koehler
of Chemical Construction Corporation.
                             605

-------
                        OPERATIONAL PERFORMANCE OF THE
                        CHEMICO MAGNESIUM OXIDE SYSTEM
                         AT THE BOSTON EDISON COMPANY
                        	PART II	

                                 C. P. Quigley


 INTRODUCTION

          The magnesia scrubbing system on Boston Edison Company's Mystic

 Unit #6 has now been in operation for over a year.  From the utility company's

 point  of view we consider the results obtained to date to be a combination of

 successes and failures.

          Success has been achieved in that the process has been demonstrated

 chemically.  The scrubber can consistently remove high levels of SO2 from the

 flue gas.  The sulfites produced have been successfully calcined at the acid

 plant.  The SO2 liberated in the calciner has been used to make 98% commercial

 grade acid.  The regenerated magnesium oxide has been reused successfully in

 the scrubber for short periods of time.  Further, the scrubber itself has re-

 mained free of the scaling and plugging problems that have so seriously hindered

 the development of calcium scrubbing systems in this country.

          In spite of these successes in demonstrating the chemistry of the

 process on this large scale, the equipment and the process have failed to dem-

 onstrate a reasonable availability commensurate with power plant equipment needs.

 In this first year, total scrubber operating time has not exceeded 1200 hours.

Availability has been a very low 177. of the Unit #6 operating hours.  The longest

 sustained run has been four and one half days.  On regenerated magnesium oxide,

 the longest run has been limited to two days.


OPERATING EXPERIENCE

Process Chemistry

          Of the problems encountered in the past year, the most serious have been

 1) related to the crystal size produced in the scrubber, 2) the inability to date  to

                                     606

-------
 line out the calciner operation so as to produce an acceptable absorbent  in



 terms of sulfa.ce content and magnesium oxide reactivity and 3) the  scaling of



 dryer internals currently occurring when scrubbing with recycled magnesium oxide.



 These problems are addressed in detail in a supplement to this paper  presented at



 this symposium by Mr. George Koehler of the Chemical Construction Corporation.



          Beyond these process problems, other problems have occurred although



 they must be considered relatively minor and more readily correctable in  nature.



 The following outlines some of the experience with equipment components during



 this initial running period.



 Corrosion/Eroaion Experience




          The equipment components and piping of the scrubbing system are fabricated



 mainly of carbon steel.  Therefore, considerable attention was paid to corrosion



 control.  The internals of the scrubber and the flue gas ductwork have been  coctftd



 with corrosion resistant materials.  pH of the slurry is maintained normally above



 7.0 with only a few short duration tests being conducted at about 6.0 during the



 year.  A few excursions to 4.5- 5.0 pH have occurred for very short periods.






          The internals of the scrubber and the flue gas ductwork  leaving the



 scrubber have been coated with a spray-applied polyester lining. This  lining  has



 stood up well.  Erosion has been noted only at the leading edge of  the venturi




 annulus.  This annulus is subject to direct impingement by slurry crystals  on  its



 relatively sharp-edged top surface.  Some metal wasteage has occurred where  this




 coating has been eroded.  A cladding of the top surfaces of the annulus  ring with



 stainless steel should correct this problem.




          The flue gas ductwork to the scrubber has been protected  with  a I*1 lining



of a corrosion Inhibiting gunlte.  During this first year, no corrosion  of  this



ductwork has bean experienced although the uncoated fan casings of  the new  scrubber




booster fans were corroding severely in the same environment.  Subsequent guniting



of these fan casings has eliminated this problem.



                                       607

-------
          Viton expansion joints were installed in the flue gas ductwork as a
 corrosion protection measure and have stood up well to date.
          Severe corrosion/erosion of cast iron pump impellers occurred in the
 slurry recirculating pumps.  These impellers required replacement within the
 first six months of intermittent operation.  New 316  stainless steel impellers
 have been installed and show no corrosive attack at this time.
          Ring header piping at the top of the absorber feeds slurry to several
 tangential wash nozzles above the venturi throat.  Pip.ing in this header has
 experienced heavy wall thinning at points where slurry sharply changes flow
 direction, such as at reducers and at tees.  Several leaks have resulted requiring
 maintenance welding.  No metallurical evaluation has yet been made to determine  the
 cause but erosion due to slurry impingement is believed to be responsible.
 Scaling and Plugging Experience
          A small ring of scale has deposited at the interface between the enter*
 ing flue gas and the slurry liquor at the top of the scrubber.  This scale deposit
 has been too small to affect the flows of slurry or flue gas and does not appear to
 increase in size.  No maintenance work for removal has been required.  The remainder
 of the scrubber, a rather open vessel and the demisters have shown no evidence of
 scale or deposit buildups.  The pressure drop across the demister has not increased
 since initial operation.
          Scale like deposits have been experienced in dead end legs of magnesium
 oxide recirculating piping.  This occurrence resulted when steam piping used for
 freeze protection was activated this winter.  It is believed that over heating of
 this static slurry accelerated hydration and resulted in the buildup of hard scale*
 like deposits plugging the line.  Provisions for rodding at pipe ends and provisions
 for bleed piping to prevent static slurry buildups have been installed.
          Slurry settling to the bottom of inactive vertical discharge lines at  the
 slurry recirculating pumps resulted in plugging of these lines.  The material  de-
posited and built up at the closed pump discharge valves.  Small bleed piping  was
                                      608

-------
installed above each valve to appoint  of  lower hydraulic pressure at Che common
suction header to the pumps.   The problem has been alleviated somewhat by providing
a continuous flow path for the settling slurry.   In addition, when the system is
shutdown, one of the recirculating pumps  is  operated  continuously to move the
slurry in the recirculating lines.
          The tendency of the centrifuged wet cake to scale has been the process
phenomena that  has inhibited continuous  operation of the  scrubber system.  Hard
scale deposits in the centrifuge, the  ribbon conveyor feeding the dryer and the
dryer itself have plagued this installation.
          Hardened scale deposits left in the centrifuge after  shutdown have
caused repeated shear pin failures and centrifuge motor overloads on startups.
Provisions to backwash the centrifuge  were installed  and the centrifuge is washed
at each shutdown.  The problem is still present  but appears now to be manageable.
          Both rubber housings and steel housings have been tried with relatively
poor results in an effort to cope with scale deposits in the dryer feed conveyor.
This conveyor continues to be a high maintenance item and  a source of many  system
shutdowns.
          The scaling in the rotary dryer of course  has been  the greatest problem
of all and until resolved, a austained run of this  plant cannot be achieved.
Winter Conditions
          Winter operation pointed out deficencies unique  to  cold weather operation.
Freezing and failure of steam traps on steam tracing lines,  freezing of control air
lines because of inadequate air  drying equipment and discomfort for operator and
maintenance personnel required to work for  long periods on equipment outdoors
had to be faced.
Miscellaneous Problems
          Flue gas  dampers are installed to isolate  the scrubber from the gas
path to  the stack.  These dampers are  driven by pneumatic power drives.  Several
occasions of  sticking dampers have been experienced  and on three occasions,
                                      609

-------
 misoperation of these dampers resulted in an unsafe condition in the boiler
 requiring emergency  tripout of the unit.  New more powerful damper drives are
 about to be  installed,  in the meantime, operating procedures are in effect to
 assure proper damper operation so as to prevent the cutoff of a flue gas path
 to the stack by the  failure of dampers to operate.
 Spillage and Housekeeping
          The conveying and handling of the magnesium sulfite and magnesium oxide
 material, much  of which is of dust size, the removal of scale buildups in the
 dryer,  the occasional spillage of -slurry through the scrubber overflow line, and
 the bleeding of plugged piping have all contributed to a continuing housekeeping
 problem and  an  excessive maintenance cost relative to cleanups.
 Operation Staffing
          Operation at Mystic has been staffed by two Edison operators on a 24
 hour  basis.   Chemico has supplemented this force with one man coverage around the
 clock for operational guidance and test purposes.
          Operation at the calciner facility is covered by two Essex Chemical
 operators and by one Chemico operator per shift.
 Operating Costs
          Edison operating costs for the first year of the test and demonstration
 program at Mystic Station are now projected to be $525,000.  This projection is
 based on accrued costs for the first eight months and estimated costs for the
 remainder of  the year.
          Of  these costs,  $235,000 are for Edison operating labor, $240,000 are
 for Edison and  invoice maintenance labor and materials and $50,000 are for metered
 services to  the scrubber system such as water, power and fuel.
          These costs are  heavily unbalanced because of the low operating hours
and the many problems encountered during this first year.
          A  substantial part of the operating labor and maintenance invoice costs
are related to cleanup work associated with the dryer scaling and material spillage
                                      610

-------
problems.   These costs  would  be  substantially reduced if the scaling problem were
solved and the system was  operable  for  reasonable periods of time.
          Maintenance costs, although  only  27, of plant investment, must be considered
excessive  in view of the few  operating  hours.  At this  time, we  do not have sufficient
operating  experience to accurately  project annual maintenance costs.
          The cost for  services  would increase four  fold for the fully operable plant
but must be considered  high based on  actual operating time.  This is due to the use
of power and fuel through  much of the plant standby  period this  past year.
Magnesium Oxide Requirements
          MgO makeup requirements can not  yet be ascertained.  A sustained run of
both plants for at least a few months will be required  to get meaningful data in
this area.
Operating  Constraints On Boiler
          There have been no  constraints placed  on  the  operation of  this unit by
the addition of the scrubber  system.  Whether the  scrubber is  in service or not,
the unit can continue to operate.  Through use of •  bypass damper  system, the
scrubber can be put into and  taken  out  of service without  influencing  boiler
operation.  When the scrubber is operating, high  sulfur residual oil is burned
in the boiler.  By simple valving,  the  boiler is  changed over  to a  low sulfur oil
supply before the scrubber is taken off the  line.
Modifications Dictated  By Operating Experience
          The following features in hindsight would be  of  value if available  on
this prototype:
          1.  Tank for draining scrubber system,  tank to be equipped to assure
              slurry will stay in suspension.
          2.  Clean outs at pipe bends  and dead  ends so as to  facilitate  redding lines.
          3.  Provisions for  continuous flow in  slurry lines which are valved closed.
          4.  Careful design  of piping  to eliminate high velocity impingements  of
              slurry at tees, ells and  reducers.  Careful selection of piping mater-
              ials to provide greater erosion resistance when required.
          5.  Strategic location of crushing systems to handle magnesium sulfite
              slags.
                                     611

-------
           6.   Strategic location of screening and/or crushing equipment to
               handle contamination of .feeds in transport, etc. that have con-
               tributed to plugging problems.

 LARGE SCALE SCRUBBING STUDY
          A recent estimate of the cost to install and operate scrubbers at
 Mystic Station has been made (Figs. 1 & 2).  Under this scheme, four units
 totaling  1050  MW would be equipped vith magnesium oxide scrubbing systems and
 would burn 2.5% S residual fuel oil.  A separate estimate for calciner operating
 costs was made (Fig. 3).  The overall economics of scrubbing and calcining as
 compared  to the present requirement of burning 0.571 maximum sulfur content fuel
 oil  (Figs. 4 & 5) clearly shows a large economic advantage to scrubbing.
          Scrubbing costs equate to $1.00 per barrel of fuel oil burned ($0.16 per
 million Btu).  Present cost differential between low and high sulfur oil is
 twice this amount and savings in excess of $7,000,000 per year are potentially
 available to our customers who now pay these costs through a fuel adjustment
 clause.

 OVERVIEW
          At this writing, we are still attempting to get this process on-stream
 for  a continuous operating period.  A continuous run is required to successfully
 level out the  calciner operation.  A continuous run is required to develop cost
 data for the process.  A continuous run is required to demonstrate that the process
 has  the reliability necessary for adaptation to power plant equipment.
          We are confident that the dryer scaling problem will eventually be
 overcome.  It  should be susceptible to solution by equipment modification or by
 equipment replacement.  We are hopeful that a solution to this scaling problem
will then result in continuous operation of both plants so that refinement of
 operation, in  particular operation of the calciner, can then be accomplished.
Many of the problems now encountered would be eliminated as they relate directly
                                      612

-------
to our inability to keep the equipment running.




          Further, we are confident that this process represents  some  of  the



most promising technology today and if successfully developed can assist  many



utilities who are in a position to combine their operations with  a chemical



operation for disposal of the S02 by-product.
                                       613

-------
                                   FIGURE 1
Scrubbing Systems
  Unit
	     MW
4      150
5      150
6      150
7      600
                             Preliminary Estimate
                      Capital Investment Requirements for
                       MgO Scrubbing of Stack Gas SC>2 at
                      	Mystic Station	
                                       Capital Investment
                                              $  2,000,000
                                                2,000,000
                                                2,000,000
                                                5.000,000
MgSOj Ccntrifuging, Drying

MgO Storage

MgS03 Storage

Media Preparation

Initial M&O Charge
                                                            $11,000,000

                                                               3,300,000

                                                                400,000

                                                                700,000

                                                                100,000

                                                                300,000
Other Captlal Costs

Services to Battery Limits

Foundations

Site Work

Edison Direct Engineering


Total Capital (1972 $)

Less Existing Unit #6 Base Unit

Total Additional Captial

Escalation to 1976

Contingency

Total Additional Investment
                                         $   500,000

                                           1,000,000

                                             250,000

                                             250,000
                                                               2,000,000

                                                             $17,800,000

                                                              (1,000,000)

                                                              15,800,000

                                                               4,200,000

                                                               1.000,000

                                                             $21,000,000
                                      614

-------
                                   FIGURE  2

                     Annual Owning and Operating  Costs  for
                           MgO Scrubbing Systems  at
                     	Mystic Station
Basts:  Mystic Units M,  f>5,  #6,  #7 = 1,050 MW

Fuel Oil Usage o 7,300,000 bbls/year @ 2.57. sulfur

     S02 removal equivalent to 0.3"{, sulfur fuel oil

Cciptlal Investment « $21,000,000


Fixed Charges                                                  $3,100,000

Maintenance                                                       800,000

Labor                                                             200,000

Supervision                                                       100,000


Utilities

Electric Power                                                    800,000

Fuel Oil                                                          400,000

Water                                                             100,000

Mg,0 Make-Up                                                       250,000

        Total Annual Owning and Operating Cost                 $5,750,000
                                       615

-------
                                   FIGURE 3

                            Annual Operating Costs
                               Calcining Plant
Capital Investment:       $4,500,000


Amortization

Interest

Maintenance

Insurance and Taxes

Labor + Supervision
      + Overhead

General Administration


Utilities

Electric Power

Voter (Process, Cooling)

Fuel Oil

Chemicals

                  Subtotal

Management Fee

                  Total Annual Operating Cost
$  450,000

   200,000

   200,000

   150,000

   200,000


    70,000




    50,000

     5,000

   300,000

    25.000

$1,650,000

   250.000

$1,900,000
                                      S16

-------
                           FIGURE 4

                       Overall Economics
                MgO Processing of Flue Gas  SO2
                	Annual Operating Costs
MgO Scrubbing                                $5,750,000

Calcining                                     1,900,000

Transportation                                   80,000

Gross Scrubbing Costs                         7,730,000

Sulfur Credits                                  400.000

Net Scrubbing Costs                          $7,330,000
Cost/bbl fuel oil                               $1.00
                           FIGURE 5

                  Estimate of Annual Savings
                 Scrubbers at Mystic Station
Annual Fuel Useage (bbls.)                     7.300.000

Annual Cost - LS oil ($4.70/bbl.)           $  34,300,000

Annual Cost - HS oil ($2.70/bbi.)           -19.700.000

Gross Fuel Savings/yr.                       14,600,000

Annual Scrubber Costs                      __- 7,300.000

Net Annual Scrubber Savings                   7,300,000
                              617

-------
DESIGN  AND INSTALLATION OF A PROTOTYPE
    MAGNESIA SCRUBBING INSTALLATION
                     by

      B . M. Anz,  C .  C. Thompson,  Jr.,
             and J. T. Pinkston
     United Engineers & Constructors Inc.
               1401 Arch Street
         Philadelphia,  Pennsylvania
                     619

-------
                                INTRODUCTION








          The design and installation of this prototype magnesia scrubbing




installation represent an action of the Philadelphia Electric Company to




cope with the problem of stack gas emissions at its Eddystone Generating




Station.  Although the topic of major interest here today is the removal




of S02 from flue gas, at Eddystone there was a problem of removal of particu-




lates as well.  Because of this and other complications, Philadelphia




Electric chose to follow the procedure described in this paper rather than




to purchase any package system being offered for S0~ removal.








          The design which will be described here is the result of engineering




work by Philadelphia Electric and their Architect-Engineer, United Engineers




and Constructors Inc.  It had the following three main goals:




          1.  Build a prototype S02 removal system with a capacity equivalent




              to a 120 megawatt unit.




          2.  Add equipment which would achieve reduction in the emission of




              particulates.




          3.  Minimize the possibility that the operation of this prototype




              S02 scrubber would affect the availability of Eddystone No. 1




              to generate power.




This design was based on solid engineering principles which were applied to



known chemical process work involving magnesium oxide.








          All of this makes it necessary to describe some of the character-




istics of Eddystone No. 1, the station to which this magnesia scrubbing




                                     620

-------
system is being retrofitted.   Eddystone No.  2, which is  similar  to Eddystone




No. 1 in general size and some other characteristics,  will  also  have  scrubbing




units retrofitted to it after the successful operation of this prototype.




Both stations burn coal averaging about 2.5% sulfur, 8 to 12% ash and 13,500




Btu/lb. heating value.








          Eddystone No. 1, which went into service in November  1959,  has a




nameplate rating of 325 Mw, but a generating capacity of 360 Mw.  Steam con-




ditions are 5000 psi and 1150°F, with two reheats at 1050°F. This pressure




and temperature are the highest used in any power station in the U.S., and,




as far as we know, in the world.  It is a matter of considerable difficulty




to start up or to shut down this station.  Accordingly, it was  essential that




special care be taken to prevent any upset condition in the scrubber from




reflecting on the generating unit itself.  A lot of flexibility was engineered




into this scrubbing unit.








          The space available for the scrubbing system was severely limited.




It was a factor the engineering team always had to keep in mind, and had an




effect on the number and sizes of the parallel scrubbing trains.  When




complete, this overall installation will have essentially  three  scrubbing




trains for each 360 MW boiler unit through which  the  stack gases' from




Eddystone No. 1 and Eddystone No.  2 can be passed.  Elaborate ducting will




permit various combinations of these scrubbing lines  to be used.  Although




particulate scrubbers, ducts  and valving will be  installed for  all three




trains for Unit No.  1, only one complete scrubbing  line, i.e.,  both  particulate




and S02 scrubbing, equivalent to 120 MW will be installed  in this,first phase.



                                     621

-------
          The Eddystone station is exactly in line with the main runways used




by jet aircraft at the Philadelphia Airport, and only a few miles away.




Obviously no unusual plume could be tolerated.  Even if we could have convinced




the environmentalists that this steam is harmless, the safety consideration




would require that a plume be avoided.








          The best approaches to papers for oral presentation with slides,




and papers in written form, are not entirely consistent.  In order for slides




to be read at all by people in the back of the room and to be understood in




the short length of time they can be shown, they must be kept very simple.




As an aid to describing this process to you, several ultra-simple slides have




been prepared.  They are included in the written text, because they may help




clarify some points, but mostly because the figures had to be prepared anyway.




Also, in the written there is provided a process flowsheet with a material




balance.  It was not originally prepared for this type of use, and there really




is given much too much information for the space available.  Still, with a low




power magnifying glass,  you can read it.









          Finally, let me repeat that this is not a stripped down plant.




In all aspects,  the engineering reliability was placed ahead of cost.  There




are several items of equipment which we might not put in a second generation




plant.  There is more than liberable interstage storage.








          We hope to put th±s prototype scrubber into operation during the




fall of 1973.
                                    622

-------
                             PROCESS DESCRIPTION









Humidiflcation and Particulate Removal









          Eddystone No.  1 is equipped  with mechanical collectors  and  electro-




static precipitators upstream of the induced draft fans.   The collection




efficiency is about 93%.  This is no longer good enough and the plant design




was required to provide for particulate removal as well as SO- removal.









          Obviously, one approach to additional particulate removal was  the




installation of wet scrubbers.  An early design question was whether  to  install




a separate wet scrubber from the S0« scrubbing system or to combine the  func-




tions.  Clearly the scrubbing device which we have in mind for S02 removal




would remove enough particulate matter to bring the station into compliance




with regulations.









          The decision was  reached to provide a separate venturi-type wet




scrubber  using only water  for the removal of particulates  and  for  the humidi-




fication  of  the stack gas.   There are several reasons  for  this decision.  First,




it was  expected that the humidification of  the stack gas prior to  bringing  it




into contact with any chemical  scrubbing medium might  assist  in  avoiding




plugging  problems.  Second, it  has been established  that some components  of




fly ash such as vanadium and iron compounds can catalyze  the  oxidation  of




magnesium sulfite (MgSO,)  to magnesium sulfate  (MgSO.).  The  presence of  some




sulfate in  the scrubbing system is, of course,  unavoidable because some of  the
                                      623

-------
S00 is oxidized to SO- before it reaches the scrubber.  Nevertheless, it
  4-                  -}


appeared desirable to hold down the quantity of this complicating sulfate as



much as possible.  Third, there is no easy way to eliminate fly ash from the



circulating scrubbing slurry.  A blowdown stream might be taken, but this



would cause a large loss of magnesium and would add to the solids disposal



problem.  Finally, there is the matter of general system reliability.  In the



present design, if difficulties in the operation of the S0_ scrubber are en-



countered, particulate removal can be continued by simply by-passing the SCL



scrubber.







          A price was paid for this decision.  Not only is there the capital



cost of the first scrubber in which only water is used as a scrubbing medium,



but there is a serious operating cost.  The consumption of electrical power



for the entire scrubbing system was almost doubled by the use of this first



stage wet scrubber.







S02 Scrubbing Step







          A fundamental feature of the process being installed at Eddystone



is that the chemical compound which reacts in the scrubber to bind the SO. is



not MgO, or Mg(OH)2, but MgSC>3.  MgStX reacts with H.O and SO. to produce



Mg(HS03)2.  As you know, MgSC>3 (more properly MgSO^.GHJ) in this system) is



not very soluble in H20.  Accordingly, it is actually used as an approximately



10% slurry of crystals in a dilute solution of Mg(HSO,)_, with a pH of about 6,
                                     624

-------
          On the other hand, Mg(HSO,)_ is  very  soluble.  Thus,  the fundamental




chemical reaction which occurs in the scrubber  tower results  Ln the  conversion




of a slightly soluble material to a very soluble material.  This approach offers



a good possibility of avoiding the problem of plugging in  the scrubber  tower.








          A price may be paid for this feature, because the equilibrium vapor




pressure of S02 over a slurry of MgSO« in a slightly acid  dilute solution of




MgCHSO )  is higher than that for a slurry of MgO or Mg(OH>2 in H-O. Accord-




ingly, it is possible that removal of S02 from  stack gases will not  be  quite




as effective in the Eddystone plant as if an alkaline slurry were used.



Precise equilibrium values in the concentration range of interest are not



available to us.  From the curve relating scrubbing efficiency  to the




acidity of the scrubbing slurry in a particular experimental unit under




consistent operating conditions, it can be seen that the difference between




S02 removal at a pH of 6 and at pH 8 is not very great.








          The equipment in which S02 is removed from the stack gases is a




commercial device manufactured by Environeering, Inc. known as  a Ventrl-Rod




unit.  A simplified sketch of this contactor is shown.  Essentially there are



two scrubbing stages.  In each a slurry of magnesium sulfite is sprayed upward,




co-current with the stack gas, through openings between cylindrical rods.  A



venturi-like effect similar to a fountain is obtained as the streams flow




between these rods, assisting in the contacting between the stack gas and the



slurry of magnesium sulfite.
                                     625

-------
          Mist is eliminated by the use of backwash sprays of water onto




louvers in  the top section of the scrubber.









          This scrubber has an overall height of fifty-three feet.  It is




rectangular in shape, fourteen feet six inches by twenty-five feet.









          Magnesium sulfite slurry is circulated to it at a rate of 13,400




gpm, or 133,800 Ib/min.  The humidified gas rate is 268,000 actual cubic feet




per minute or 17,300 Ib/min. giving an L/G of 50.  This high slurry-to-gas




ratio is designed to minimize concentration changes across the SO- scrubbing




system proper, and to insure a rather constant acidity of the scrubbing




medium.
          Neutralization
          In the scrubber surge tank the following reaction occurs:




             Mg(HS03)2 + Mg(OH)2 + 10H20 - >2MgS03.6H20




Here we have the first formation of a slightly soluble material.  By carrying




out this reaction in a large agitated tank - its capacity is 60,000 gallons -




having a hold up time of four minutes, any plugging problems should be held




to a minimum.  The temperature in this tank where the neutralization occurs




is close to 129 F, a condition under which the hexahydrate is formed.  At




higher temperatures the trihydrate can form.   Its crystals tend to be smaller




than those of the hexahydrate, and might cause more handling problems.
                                     626

-------
          The pH of the slurry in this surge tank  ts  controlled  at 6 by



regulating the rate of addition of Mg(OH)   slurry.
MgSO- Recovery
          The recovery of the MgSO,,.6HJ) formed in scrubber surge tank is




begun by diverting a drag stream from the main scrubber circulation loop.



It flows first to a thickener.  Here, again, is a piece of equipment which




may not be absolutely necessary.  The conclusion to include it was based on




the desirability for more surge capacity in the recovery portion of the plant




and the prospect for improved centrifuge operation.








          The thickener has a diameter of forty feet, and a straight side




height of twelve feet.  It receives 1975 Ib./min of slurry containing  165




Ib./min MgS03,6H20.  The underflow from the thickener going to the centrifuge



contains about 25% MgS03.6H20.








          In the stainless steel solid bowl centrifuge, a cake of




MgS03.6H20 crystals wet with a solution of MgSO, is recovered.  These wet




crystals of MgS03.6H20 are discharged from the centrifuge, through a vertical




chute into a screw feeder which provides a seal and a continous flow of wet




solids, into a rotary kiln type dryer.  Combustion gases from an oil burner,




tempered by a side stream of stack gas from the first induced draft fans,




flows co-currently with the crystals.  Anhydrous MgSO» is discharged from



the dryer and conveyed to the storage silos.








          Some MgSO^ will also, of course, be present.



                                     627

-------
 Regeneration of MgO









          For this prototype facility, the MgSO_ will be trucked to a




 sulfuric acid plant about twenty miles from Eddystone.  There it will be




 heated in an oil-fired fluidized bed reactor to decomposition.  The MgO




 formed goes overhead with the SO^ and the combustion gases, from which it




 will be separated by cyclones.  The stack gas is cleaned up by a combination




 venturi and a packed scrubber tower before being admitted to the sulfuric




 acid plant itself.









          The selection of an optimum temperature in this fluidized bed




 regenerator represents a compromise.  On the one hand it is desired to reduce




 the MgSO, to MgSO  and to insure the substantially complete decomposition of




 MgSO- to MgO and SO , both of which are favored by higher temperatures.  On




 the other hand, excessive temperatures will produce "dead burned" MgO which




 is not very reactive chemically and is thus not really effective for further




 S0? removal.  The final operating temperature for this stage has not been




 selected, but will probably lie in the range of 1650° - 1750°F.  Fluidized




bed reactors are particularly well suited to precise temperature control.  They




 also allow for precise control of oxygen in the reactor which eliminates the




 necessity of adding oxygen scavengers such as carbon to the bed so the




MgSO,  can be decomposed.









          The regenerated MgO will be trucked back to Eddystone for re-use.




 For a prototype operation, this sort of thing can be tolerated.  For a full
                                    628

-------
scale operation at a major generating station,  however,  this  trucking will




not be practical.  The sulfuric acid plant should be located  at  the station




site.









Stack Gas Reheat
          When the stack gas emerges from the S02 scrubber, its temperature




is about 130 F and it is substantially saturated with water vapor.  To avoid




a billowy white plume from the stack as this mixture emerges and water con-




denses, some kind of reheat is necessary.  In this retrofit unit at Eddystone,




the practical way to accomplish this was to install an oil burner mounted




directly in the flue gas duct leaving the S02 scrubber.








          Close to one million Btu/hr. will be required for every one  F of




reheat for the total output of Eddystone No. 1 (all three final scrubbing




lines).  The fuel cost alone will be close to $5000 per year per °F of




reheat.  Obviously, the amount of reheat used will be held to a minimum,




depending on local conditions.
                                     629

-------
                                  ECONOMICS








Capital Cost








          The best estimate we can make now is that the total capital cost



for this version of a magnesium oxide scrubber is in the range of $55 - $65




per kilowatt.  This is for a new unit.  For a retrofit, you should add about



$15 per kilowatt.








          Of this basic cost, $45 - $50/KW is for the scrubbing unit itself.




About $10 - $15 is for the facility to convert MgSO« to MgO for recycle and




to carry the liberated S0« on to concentrated sulfuric acid.  Our thought is



that the sulfuric acid plant should be located adjacent to the power station.




The hauling of MgSO- and MgO to and from a distant regeneration facility and




sulfuric acid plant can become very expensive.








          This operation of a sulfuric acid plant in conjunction with a power




station is not something that will bring universal happiness.  To the power




station staff it will be an utter nuisance.  Its presence may become an up-




setting factor in the local sulfuric acid market.  This acid which is produced



has got to go somewhere, at whatever price can be obtained for .it.  Moreover,




the acid plant which will ultimately be built at Eddystone will be distinctly




below the best economic size, since it will have a daily capacity of only




about 350 tons.  At present, in the Delaware Valley, we do not see any over-



riding problem in this regard.  This brings us on to the matter of operating



costs.
                                    630

-------
Operating Cost








          We estimate that the overall operating cost of the final  facility




at Eddystone, including the scrubbing unit itself, the MgO regenerator and'-




the sulfuric acid plant, will lie between 1.3 and 1.4 rails per kilowatt hour.




The factors going into this estimate are as follows:








          Fuel cost                               $0.70/MM Btu




          Power cost                                 7 mils/KWH




          Fixed charges                             15% of capital/yr.




          Maintenance                                3% of capital/yr.








          It was further assumed that there will be a loss of MgO of 5% per




cycle.  This is probably a very conservative estimate, but in the absence of




operating experience, it appeared best to include it.








          The operation of this scrubbing facility as described earlier, with




a wet scrubber ahead of the S02 scrubbing unit, will consume approximately 3%




of the power generated by the station.  If the wet scrubber were eliminated,




power consumption would drop to about 1.75%.
                                     631

-------
       C.EACTIOAJ  1AJ 5£!2U&BEIZ
O-i
OJ
NJ

-------
                           PH
MA60E61UM SYSTEM
iwu 	
99 	
98 	
% 	
V
V
3
vi
5
kJ
V3
2
t£l
VJ
VP
95 	
^ 	
^ 	

92. 	
91 	
90 	
99 	
;:~
gk 	
66 	

64



^^-»--^-"^*""-t
X
/
if
!
i
1
i
l
1
i
i


iii
6 6 7 6
    OF sazuaaiiaQ 6i.ora.izy
      633

-------
•6LUQJ2V  IAJI_feT

-------
   I2EACTIOAJ 1AJ  5CG.UBBEIZ 6UIZG.E TAAJtf.
U>
Ul

-------
                   6-TAC.k.  6.A/S
                   FfZOM I.Q.
              PAr2.TlCUL.ATE
U)
                                          HUMIDIFIED
                  PATZ-flCDLATE.
                                    To
                                    6JATET2
                                                                                  CLE-AAJEC?  GA-6  To
                                                                                  I2.EMEAT  AUD  5TAO!
                                                                           I
                                                                                           11
To

-------
-TO Mq CD
rze&Eju
PLAAJ-T

-------
                                1-760°
FEED
         V
   CPl
   w
   00
          l\\\\\\\\\\\
IZEACTOIZ
                          LtfcJE.0)
 FDEU
   Atrz
PrZEU&ATErz.
                                      1OQO
                           AlC

                           1
                                                                  •ro
                                                              rZEC.Ol/ETlE.0
                              &LOUJ
                                                                                       j,
                                                                     6J£T
                                        fZEGEUEfZATlQJSJ  PLAMT

-------
                                                                              r—
                                                                               — i 4
   ha?-:

.£
                                                                         ©
        i '
        I i
        ©
  JL.
fflol
s
/
1
V



r
i
i
i 	
*> V
"Jgd XCJ 1"»"l ^M^ XL
' I
..I/;© os . U
' N
i
i
^ri
f~
^ t^S '"K1^1 H-"^
— u
rw CUM rL
>
i
/
/
— =%*;

-"•""
                                                                ^ r'1
                                                                           t~-1
                                                          ^p
                                                         Efe
                                    34  11  ?*  « Jo  ii  i?  M  !4

-------
                    CAPITAL  COST
                            AJELJ 3OiUfcBlAJ£» UAJIT



 10 - $ 1 5/H.LJ     f2E6EAJEfZATlO/J U/UIT AlUP W5O  PLA/oT



FOfZ JZETfZOFIT - ADD A&OUT
                OPEflATlAJd   <105T
 Ol
 Jt
 o
   01/EI2ALL



   FACTO125  COAJ5lDEfZBD

         FUEL.                        ^O."70/MM

         POMEfZ                      -7 mil6/l^liJU


         FI^ED ^UAfZ.GE.5            15% af £apif
-------
      APPLICATION OF THE WELLMAN-LORD  SO2
RECOVERY PROCESS TO STACK GAS DESULFURIZATION
                        by

               Raymond T. Scheider
                Christopher B. Earl
                Davy Power gas,  Inc.
                 Lakeland, Florida
                         641

-------
                          ABSTRACT









     The Wellman-Lord S02 Recovery Process has been successfully




commercialized in the U.S.A. and Japan.  Six full-scale installa-




tions have now built up an aggregate operating experience of




eight years.  The fourteen emission sources controlled by the




process include sulfuric acid plants, claus plants and oil-fired




boilers.




     Since the start-up of the original plant at Paulsboro, New




Jersey for Olin,  the process has demonstrated long-term relia-




bility of operation and its ability to attain SO9 emission




levels as low or lower than required by the regulatory agencies.




     This paper describes the basic process and reviews the cur-




rent status of this technology.  The application to large fossil




fuel-fired steam generating plants is reviewed,and capital and




operating costs for a specific application are included.
                               642

-------
               THE WELLMAN-LORD SCU RECOVERY PROCESS
INTRODUCTION
Six commercial installations of the Wellman-Lord SO2 Recovery Process
are operating successfully in the U.S.A. and Japan.  These applications
include seven sulfuric acid plants, two oil-fired boilers and five claus
units  (see Table 1} .

The SO2 emission level first envisioned with the use of the process was
500 ppmv since this was a statutory requirement in 1969.  The operations
of the five plants have demonstrated the process's ability to maintain
emission levels at consistently less than 200 ppmv.

CHEMISTRY OF THE PROCESS

The process is based on a sodium sulfite/bisulf ite cycle.  The reactions
that take place in the process can be abbreviated for simplicity as
follows:

     ABSORPTION

               S02 + Na2S03 + H2O - »-  2NaHS03


     REGENERATION

               2N3HSO - »-  Na2SO  |  + S02 |  + H O 4


Apart  from the two major reactions above, sodium sulfate  (Na^SO,^) ,
which  is nonregenerable in the normal process, is  formed  in the
absorber as a result of solution contact with oxygen and  sulfur  trioxide
as follows:
                                     Na2S04
                            and
The sodium sulfate so formed is controlled at a  level  of approximately
five percent by weight  in  the  absorber feed  stream by  maintaining a
continuous purge  from the  system.   This purge  stream is discussed later.
    additional  source  of  sodium sulfate  and thiosulfate is the autooxidation
or  so-called disproportionation reaction which takes place in the regen-
eration  section:

                                  643

-------
                                                        TABLE I
                                               DAVY POWERGAS INC.
                                                                                                          Type Contract
                                                                                                          E - Engineer ing
                                                                                                          P - Procurement
                                                                                                          C - Contraction
                                                                                                          T - Turnkey
WELLMAN-LORD SOj RECOVERY PROCESS
Project
Number
3659


l<120
4126
412?
M37
4H
-------
               6NaHSO 	>- 2Na2S04 + Na S~0o + 2SO  + 3H-O
                     «3                 £ *• -^      ^     ^

Laboratory research and commercial experience have guided the selection
of operating conditions which minimize all these sources of sodium
sulfate formation.

A make-up of caustic is required to replace that lost in the purge
stream.  The caustic make-up solution reacts with the sodium bisulfite
in the absorber solution to form additional sodium sulfite:

               NaOH + NaHS03	»- Na2S03 + H2O

Soda Ash (Na2CO3) can also be used as the makeup source of sodium.

The simple regeneration scheme of the patented Wellman-Lord process
relies on the favorable solubilities of the sodium system.  The bisulfite
form has almost twice the solubility of sulfite at the temperatures con-
sidered for the process.  Because of this it is possible to feed the
absorber with a saturated sulfite solution, or even a slurry, without
any fear of additional crystallization or scale forming despite consid-
erable evaporation of water.  This is due to the fact that as S02
absorption proceeds, the composition of the solution is shifted in the
direction of increasing solubility.

There are several advantages to operating the absorber with highly con-
centrated solutions.  The solubility of oxygen decreases rapidly as salt
concentration increases and so the mass transfer of oxygen into the
solution is slowed appreciably.  This reduces oxidation of the sodium
salts to sodium sulfate.  The steam requirements of the process are
directly related to the quantity of water pumped around the system.
Operating at or near saturation with respect to sulfite thus reduces
the overall steam consumption.

The same solubility effect is taken advantage of in a revers fashion
in the regeneration section.  AS SO2 is stripped from the concentrated
solution,, the sulfite salt is formed and rapidly reaches its solubility
limit thus precipitating as crystals.  in effect the low vapor
pressure component  (sodium sulfite) in the solution being regenerated,
is removed from the equilibria balance.  Thus the  regeneration
operation can proceed with a constant high percentage of bisulfite
in solution permitting a considerable reduction in the stripping
steam requirements.

DESCRIPTION OF FLOW

The flow scheme for the basic process is shown in figure 1.  The SC>2
rich gas is contacted counter-currently in the absorber by the
sodium sulfite solution and passes out the top stripped of SO2-
The solution leaving the bottom of the tower, now rich in bisulfite,

                                 645

-------
                           FIG, I BASIC FLOW DIAGRAM
                                                 «<£•«
                                                   '%
                                                      ICONULNSER
                                                    77?k

                                                              PRODUCT
G
     ABSORBER
SOLUTION
 STORAGE
  VAPORATOR
CRYSTALLIZE.R
DISSOLVING
  TANK
               WELLMAN-LORD SO.,  RECOVERY PROCESS

-------
is discharged to a surge tank and then pumped to the regeneration
section.

Low pressure steam is used to heat the evaporator and drive off tJ02
and water vapor.  The sodium sulfite precipitates as it forms and
builds a dense slurry of crystals.

The overhead stream is subjected to a partial condensation to remove
the majority of the water vapor.  The condensate recycles to the
dissolving tank and the product SC>2 is discharged from the process.

A stream of slurry is withdrawn from the evaporator and is redissolved
by the overhead condensate.  The lean solution is pumped to a surge
tank prior to being fed back to the absorber.

    GAS PRETREATMENT

    Normally the flue-gas from a fossil fuel-fired boiler or a smelter
    will require cooling and particulate removal prior to absorption
    of the SC>2»

    In the case of a claus plant, the tail gas must first be
    incinerated under conditions which destroy the H2S, COS and
    CS2»

    This pretreatment step must be studied on a case-by-case basis
    to ensure the selection of the optimum design in relation to the
    overall facility.

    The type of pretreatment ultimately chosen will depend on the
    following parameters:

                           Gas Temperature
                           Particulate Content
                           Organic Sulfur Content
                           Sulfur Triodide Content
                           Acid Mist or Vapor Content
                           Humidity.

    SO? ABSORBER

    The absorber is a simple two or three stage contacting device.
    DPG commercial experience with both tray and packed towers
    permits selection of the design most suited to  each specific
    application.
                                 647

-------
For most gases treated the absorber will require recirculation
around each individual stage, since the quantity of feed solution
is insufficient to adequately wet the tray or packing.

During operations the recirculation rate can be throttled until
the SC>2 emission increases up to the design point.  This will
minimize oxidation since the absorption of oxygen is liquid-film
controlling and therefore proportional to liquid rate.  The
absorption of S02 is gas film controlled and therefore relatively
insensitive to liquid rate.

An inherent advantage of the Wellman-Lord Process is that the
absorption system can be physically separated by a large distance
from the regeneration system.  This also means that it is practical
to treat gases from more than one plant by installing separate
absorbers for each source of SO2» with all the absorbers being
supplied by a common regeneration system.

SOLUTION SURGE TANKS

The use of solution storage not only enables the process to
operate smoothly during frequent changes of gas flow or S02
concentration, but it also permits a complete shutdown of the
regeneration section to perform scheduled maintenance acitvities,
without any pause in S02 absorption.

This feature minimizes the amount of expensive spare equipment
required with no sacrifice in basic pollution control reliability.

CHEMICAL REGENERATION SYSTEM

The heart of the regeneration system is a conventional forced-
circulation evaporator/crystallizer.  The design parameters of
this unit have been developed such that long term continuous
operation is assured.  The evaporator can be designed to use very
low pressure exhaust steam which might otherwise be discharged
to the atmosphere.

In very large plants or in case of high cost steam it is
practical and economical to operate the regeneration system as
a double effect evaporator which will reduce steam consumption
by about 40-45%.

PRODUCT SULFUR DIOXIDE

The vapor leaving the evaporator is subjected to one or more
stages of partial condensation to remove water.  Existing
plants are operating with both air and water-cooled condensers

                            648

-------
     in this service.  The final product SO2 can be delivered at
     whatever quality is required for further processing.  It is
     suitable for conversion to high grade sulfuric acid, elemental
     sulfur or liquid SO2•

     Existing Wellman-Lord plant are recycling the gas back to
     sulfuric acid and claus plants or sending it to be processed in
     a satellite sulfuric acid plant.

     PURGE TREATMENT

     A small amount of the circulating solution is oxidized to a
     non-regenerable form and a purge stream is removed to control
     the build-up.  This stream can be dried for sale or disposal, or
     it can be treated to permit its discharge as an innucuous effluent,
     Other process steps are available which recover the sodium values,
     thus allowing the system to operate as a closed loop.

OPERATIONS

The first Wellman-Lord plants quickly proved the viability of the
process but were not without some start-up problems.  Needless to
say, the lessons learned in these plants were quickly incorporated in
subsequent designs.  An indication of this is that three of the first
five users already have second plants in operation or in the process of
being engineered.

     STANDARD OIL OF CALIFORNIA

     One of our recent projects for Standard oil of California started
     up in September, 1972 and from the first day has operated continu-
     ously in a most satisfying manner.

     This plant included large surge tanks which if necessary would
     permit a three day shut-down of the regeneration section, while
     the absorption section continues to run.  The reliability pro-
     vided by this feature was vital to protect the integrity of this
     large refinery located close to Los Angeles International Airport
     and surrounded by a residential area.  The regulations governing
     this situation require full air pollution control 24 hours a day,
     365 days per year.

     A side benefit of this relatively cheap means of providing
     reliability was that the plant could handle excursions of SO2
     concentration up to three times the design level while still
     attaining an SO2 emission less than design.  The unit mentioned
     aboye^has surpassed the performance guarantees, and has been
     officially accepted by standard oil of California.

                                  649

-------
     JAPAN  SYNTHETIC  RUBBER COMPANY AT CHIBA, JAPAN

     This installation  is  operating on the  flue gas  from  two  oil
     fired  boilers  each with  a capacity  of  286,000 Ibs/steam  per hour.
     The fuel  oil contains 4.0 - 4.2% sulfur and  produces  a flue gas
     varying from 1500  to  over 2000 ppm  SO2.  The plant has consis-
     tently reduced the SO2 emission to  below 200 ppm  since its start
     up in  June  of  1971.   The plant currently averages an  inlet SO2
     concentration  of 1500 ppm with an outlet emission 150 ppm.

     During the  period  June 1971 - March 1973, the boilers operated a
     total  of  12,972  hours.   During this same period of time  the
     Wellman-Lord process unit operated  12,533 hours for a 97% on-
     stream factor.   During the period May  9, 1972 - March 1, 1973,
     the boilers operated a total of 6,978 hours, and the  Wellman-
     Lord process unit was in operation  during all of these for a
     100% on-stream factor.

     In this plant  each boiler is connected to a  five meter square
     by 21  meter high sieve tray absorption tower, and the recovery
     section utilizes a double effect evaporator  for steam economy.
     The tail gas after the absorber is  reheated  to 130°C  for plume
     dispersal.

     The product SO2 gas is processed in a slightly modified,
     conventional contact sulfuric acid  plant, and produces high
     grade,  merchant quality,  98% sulfuric acid which is sold in
     normal markets.

APPLICATION TO POWER PLANTS

     THE NIPSCO DEMONSTRATION PLANT

     As a consequence of the successful operating units mentioned
     above,  Davy Powergas has  been awarded a full scale demonstration1
     plant  project  for Northern Indiana Public Service Company,
     funded jointly by the U.S.  Environmental Protection Agency and
     NIPSCO.  This facility will treat the total stack gas from a
     115 mw coal fire boiler at the Dean H.  Mitchell  station in Gary,
     Indiana.   The end product of this plant will be  elemental
     sulfur.  SO2 Removal from the flue gas  will be performed by the
     Wellman-Lord SO2  Recovery Process,  which will deliver the
     concentrated SO2  to a reduction facility that uses Allied chemical
     Corporation's technology  to produce high quality elemental  sulfur
     as the  final product.
                                 650

-------
 DEMONSTRATION  OBJECTIVES

 The broad  overall  objective  of  this  project  is  to demonstrate
 the applicability  of  the  above  two commercially proven  processes
 to large scale coal fired boiler  plants.   In order  to meet  the
 overall objectives of the project, specific  contractual guarantees
 are being  developed in accordance with  the following guidelines:

 1.  The process will  accomplish a minimum  of 90% removal  sulfur
    dioxide  from the  stack gas when  firing coal up  to 3.5%  sulfur
    content  producing an  emission of 200 ppm SC>2 or less.   3.5%
    sulfur content of the coal  is equivalent to approximately
    2300 ppm by volume sulfur dioxide in the stack  gas.

 2.  The aggregate  cost of steam,  electric  power, water, and
    natural  gas will  not  exceed a specified  amount.

 3.  The required quantity of make-up chemicals  will not exceed
    a specific amount.

 4.  No other water or air pollution  sources  will be created.

 5.  The sulfur produced will be of a quality suitable for the
    manufacture of sulfuric  acid  by  the contact process.

 6.  The plant will comply with all applicable federal, state and
    local  air and water quality regulations  in  effect at the time
    detailed engineering  is begun.

 PHASING AND SCHEDULING OF THE PROJECT

 Phase I of the project consists of preliminary  engineering and
 definitive cost estimates.  This  phase of  the project is
 essentially complete  at this writing.

 Phase II includes  the  completion  of  detailed engineering, pro-
 curement,  construction and start-up.   This phase of the work is
 scheduled to be completed during  the  latter  half of 1974.

 Phase ill,  following  acceptance tests at the conclusion of Phase
 II,  covers a one year period of operation by Allied Chemical and
will be funded entirely by NIPSCO.

 TRW/EPA TEST PROGRAM

Concurrent with Phase ill of the  project,  a comprehensive emission
 testing program will be conducted by EPA.   To implement this test

                             651

-------
and evaluation program the EPA has retained the services of TRW
Inc.  The primary objectives of the test and evaluation program
are:

1.  A validation of the emission reduction achieved by the
    WeiIman/Allied process.

2.  A validation of the estimated installation and operating
    costs of the demonstration unit.

3.  A determination of the applicability of the Wellman/Allied
    process to the general population of utility boilers.  The
    test program will include extensive measurements of those
    parameters that characterize the sulfur removal and economic
    performance of the Wellman and Allied process units.

    In addition to specific test work, TRW will conduct a survey
    for the purpose of identifying the potential users of the
    Wellman/ Allied process.  User boiler units having
    applicability potential are selectable based upon considerations
    such as expected useful life, capacity, average load factor,
    low sulfur fuel availability and the marketability of sulfur
    and sodium sulfate in the region where the boiler is located.

PURGE CRYSTALLIZATION SYSTEM

In order to meet the objectives stated above that no additional
water or air pollution sources would be created, a purge cry-
stallization section has been included in this project,  in this
section of the plant the purge stream will be subjected to
fractional crystallization in order to concentrate the sodium
sulfate and sodium thiosulfate components of the purge.  This
portion of the purge stream will be dried to produce a solid
crystalline byproduct of high sodium sulfate and sodium thiosulfate
content.  The resulting sulfite/bisulite stream from which most
of the sulfate and thiosulfate has been removed will be recycled
to the S02 Recover Process.

ALLIED CHEMICAL CORPORATION SO? REDUCTION SYSTEM

This project is unique in that it will produce elemental sulfur
as a by-product.   The Allied chemical Corp. Process is the
subject of the next paper and will not be covered here.

SITE CONSIDERATIONS AT THE D. H. MITCHELL STATION

The demonstration plant will be installed on unit 11 of NIPSCO's


                             652

-------
     Dean H. Mitchell Station.  This slide is a view of this station
     with unit 11 at the right.  Unit 11 is a Babcock and Wilcox
     pulverized coal feed boiler capable of producing 821,000 Ibs/hr
     of steam.  The superheater outlet pressure is 1870 Ibs/sq in.
     gauge at 1005°F.  It is equipped with an electrostatic precipitator
     installation rated at 98% collection efficiency.  The coal
     presently being fired in this unit is from Southern Illinois
     with an as received average analysis of 3.16% sulfur, 10.5% ash,
     and 11.8% moisture.

     NIPSCO clearly has reason to be interested in a method for removal
     of sulfur dioxide from the stack gas.  Such a method, if effective,
     reliable, and economically viable, permits NIPSCO to comply with
     emission limits while continuing to fire coal from assured
     relatively near by sources.
     Stack gas treatment processes can be diveded into two classes,
     those that produce a so-called throw-away product and those
     that yield a product that can in theory at least be marketed.
     The throw-away process was rejected by NIPSCO for ecological
     reasons.

     The Dean H. Mitchell Station is located on Lake Michigan in
     the Northwest corner of Gary, Indiana.  As this slide shows,
     its immediate environs are highly industrial, but farther out
     the plant community is residential and densely populated.
     Considerations of land and water use made a throw away process
     environmentally unacceptable at such a site.

     This slide shows an artists rendering of the S02 recovery and
     reduction installation that will be made at the Dean H. Mitchell
     Station superimposed on a photograph of that station.  The red
     structures are those of the Wellman-Lord process, and the
     Allied process is housed within the orange building.

CAPITAL AND OPERATING COSTS

Actual costs for power plant flue gas scrubbing and S02 recovery
will vary considerably depending on the specific application.  Local
site conditions and the, difficulty of retrofit can cause costs to vary
over a wide range.  It is very misleading to compare estimates that
are derived from different bases.

Capital and operating costs for applying the W-L process to two large
plants (500 and 1,000 MW) have been projected from the detailed
studies made of the 115 MW plant at the Dean H. Mitchell Station of

                                 653

-------
Northern Indiana Public Service Company.  The following table indicates
the basic parameters used in estimating these costs:

                                TABLE 2

                                   500 MW            1000 MW
     Absorber Design Flow
     SCFM                          3 @ 400,000 ea    6 d> 400>000 ea

     Regeneration Plant
     Lb/Hr SO2 Product             1 @ 18,000        2 @ 18,000 ea

     Sulfur Plant
     LTPD Sulfur Product           1 @ 92            1 @ 184

The studies of the 500 and 1,000 MW plants are based on 3.2% sulfur
coal and the flue gas flows are based on 20% excess air.

The regeneration plant and sulfur dioxide reduction plant sizing
correspond to an 80% load factor.  Absorber operation at peak capacity
would be handled by solution inventory in large surge tanks.

The annual utility costs presented in Table 3 are based on a continuous
80% load factor for 330 days per year which is equivalent to an over-
all load factor of 72%.
                                  654

-------
                 TABLE 3 - CAPITAL AND OPERATING COSTS
Capital Cost

Annual Costs

Chemicals
Natural Gas d>55C/MM BTU
Electricity @ 7 Mil/KWH
Steam © 50C/1000 Lb
Cooling Water @ 2C/1000 Gal

Labor © $5.80/hr
Supervision
Overhead
Maintenance © 4.5% of Capital
Laboratory
Miscellaneous Supplies
capital Charges © 14.5%
                       TOTAL
Sulfur Credit @ $18/LT
                    NET TOTAL

Clean-up Cost mils/KWH

C/MM BTU
  500 MW
$18,700,000

   $/Year

    956,000
    336,000
    499,000
    554,000
    111,000

    169,000
     47,000
    343,000
    842,000
     72,000
     45,000

  2,712,000

  6,686,000

    (546,000)

  6,140,000

    1.9

     20
   1000 MW
$33,200,000

   $/Year

  1,912,000
    672,000
    993,000
  1,108,000
    222,000

    229,000
     47,000
    506,000
  1,494,000
     72,000
     90,000

  4,814,000

 12,159,000

  (1,092,000)

 11,067,000

    1.7

     18
                                   655

-------
    APPLICATION OF 803 REDUCTION
IN STACK GAS DESULFURIZATION SYSTEMS
                   by

          William D. Hunter, Jr.
       Allied Chemical Corporation
       Industrial Chemicals Division
         Morristown, New Jersey
                   657

-------
     Allied Chemical's experience in the design and operation of
a large, single train plant for reduction of SO  to elemental
sulfur has made possible the application of new technology to
emission control in metallurgical operations and fossil fuel
combustion.

     This Allied Chemical technology has been proven in two
years' successful commercial operation where it functioned as
the emission control system at a Palconbridge Nickel Mines, Ltd.
facility near Sudbury, Ontario, Canada.  The quantity of sulfur
in the gas stream treated in the Allied unit at Falconbridge
(500 long tons per day) is equivalent to that contained in the
largest metallurgical sources, or in flue gas from 2000 megawatts
of electric generating capacity fired with 3% sulfur coal.  As
a consequence, future applications of Allied1s technology will
involve scale down of equipment size with full assurance of
optimum plant performance.

     Availability of a proven process capable of converting SO
to elemental sulfur offers freedom from the problems and cost 2
of disposal of the waste products of neutralization.  Likewise,
costly storage and transport of by-product sulfuric acid can be
avoided at locations remote from acid markets.

     The Allied Chemical S02 reduction process can be applied
directly to gas sources where the SO2 content is 4-5% or more
and where oxygen content is limited.  Various metallurgical
emissions can, therefore, be processed directly in the Allied
system.  Since oxygen reacts with the reducing agent in the
same manner as the SO2, it is desirable in optimizing the
system to achieve the lowest oxygen content possible in order
to minimize reducing agent consumption.  The Allied process
has the versatility of being adaptable to feed gases containing
any SO2 concentration up to 100% (dry basis).  Hence, in those
applications where the oxygen content is too high or SO2
concentration too low for direct application to the emission
source, the Allied SO- reduction process may be joined with one  of
several regenerable flue gas desulfurization process which
recovers the SO2 from the flue gas in a concentrated, low
oxygen form.
                              658

-------
     Adaptability of Allied's SC>2 reduction technology to a
feed gas containing 100% SO2 (dry basis)  will be demonstrated
at the D. H. Mitchell station of Northern Indiana Public Service
Co. at Gary, Indiana.  There, the process will be combined with
the Wellman-Lord SO  recovery process to provide a flue gas
desulfurization system for a 115 megawatt, coal-fired boiler
in a project jointly funded by NIPSCO and the Environmental
protection Agency.

     Engineering, procurement and construction of the entire
facility will be the responsibility af Davy Powergas Inc.
Allied is providing the S02 reduction process technology as
well as technical and start-up services under contract with
Davy Powergas.  Then, under a separate agreement with NIPSCO,
Allied will operate the entire flue gas desulfurization system
and will market salable by-products.

     The role that will be undertaken with NIPSCO is a familiar
one for the Allied organization which has for many years
provided services to oil refineries in the design, construction
and operation of satellite plants for the conversion of
hydrogen sulfide to elemental sulfur and the decomposition of
spent sulfuric acids,  in addition  Allied has performed services
to metallurgical industries in the building and operation of
facilities for the roasting of sulfide ores, and the production
and marketing of sulfuric acid.  The development of SO, reduc-
tion technology enabled Allied to extend the range of its services
directed to environmental improvement.

     The Falconbridge plant,  which involved the fluidized bed
roasting of pyrrhotite ore, could only have been built if
provisions were made to comply with Canadian regulations
requiring the removal of at least 90% of the sulfur values
contained in the roaster gases before their release to the
atmosphere.  Two alternatives were considered:  the manufacture
of sulfuric acid or reduction of the SO2 to sulfur.  There
is no substantial market for either in the immediate area, but
the recovery of elemental sulfur offered-the advantages of a
commercial product that is more easily stored than sulfuric
acid and has a much wider economic shipping range.
                              659

-------
     The  Falconbridge  installation demonstrated the required
 capability of  removing at  least  90% of the SO- from roaster
 gases   at a  rated  capacity of  500 long tons or sulfur per day
 in  the  incoming gases.  Reliability of performance was
 established  as well as flexibility in turndown to two-thirds
 and one-third  of design capacity with essentially constant
 operating efficiency at all rates.

     At Falconbridge,  Allied engineered and constructed the S02
 reduction plant.   The  facility was run by Allied supervisors
 utilizing an operating and maintenance staff assigned Allied
 by  Falconbridge Nickel Mines.

     Product sulfur was, for the most part, distributed to
 Allied's  sulfuric  acid manufacturing locations serving the
 U.S. merchant  acid market.  The  Falconbridge product was used
 interchangeably at these locations with sulfur from Frasch
 producers and other by-product sulfur sources.
COMMERCIAL PLANT DESCRIPTION

     A flow diagram of the Allied Chemical SO- reduction
process as it is applied to a sulfide ore roasting operation
like that at Falconbridge is shown in Figure I.  The hot
SO2 gas from the roasters is passed through hot gas heat
exchangers (1) and (2) where part of the heat content of
the gases is utilized to reheat other process gas streams.
These will be described in more detail later.  At this point
the roaster gas still contains fine dust particles as well as
gaseous contaminants which must be removed before the gas
reaches the reduction reactor.  This gas purification is
accomplished in a two-stage aqueous scrubbing system consisting
of a two-leg gas cooling tower (3) and a packed condensing
tower (4).  The bulk of the dust and other contaminants are
collected in the gas cooling tower while the gas is cooled and
saturated by contact with a recirculated weak sulfuric acid
solution.   The demister pad at the tower outlet is continuously
sprayed with weak acid from the condensing tower.  The under-
flow from the gas cooling tower is treated with lime to precipitate
dissolved metallic impurities removed from the gas, and
neutralize the acidity, before being delivered to a waste pond
where the solids are allowed to settle.
                             660

-------
     The process gas is further cooled in the condensing tower
(4) by circulating weak acid which is cooled externally in
impervious graphite heat exchangers (5) .   Entrained droplets
of acid mist are removed from the gas in electrostatic
precipitators  (6) .  Drips from the precipitators are returned
to the gas cooling tower.

     The temperature of the clean gas is then raised above the
dew point of sulfuric acid by admixing with a reheated stream of
the same gas in mist tower (7) .  This recycle gas stream is
heated by circulation through hot gas heat exchanger (2).  The
process gas is drawn through the wet purification system, then
forced by a centrifugal blower (8) through the balance of the
plant.  Natural gas, which serves as the reducing agent, is
introduced into the process gas stream at the blower discharge
and the mixture passed through hot gas heat exchanger (1) to
raise its temperature above the dew point of sulfur before
entering the reduction reactor system.

     The principal function of the catalytic reduction system is
to achieve maximum utilization of the reductant while producing
both sulfur and H2S, so the H^S/SO  ratio  in the gas stream
leaving the system is essentially that required for the subsequent
Glaus reaction.  Although the chemistry of the primary reaction
system is extremely complex and includes reactions involving
11 different elements and compounds, it may be summarized in
the following  equations:
     CH4 + 2 S02  - j.  C02 + 2 H20 + S2

     4 CH  + 6 SO2 - ^  4 CO  + 4 HO + 4 H2S + S2


     The preheated process and natural gas mixture enters the
catalytic reduction system through a four-way flow reversing
valve (9) and is  further preheated as it flows upward through a
packed-bed heat regenerator (10) before entering the catalytic
reactor  (11) .  The temperature of the gases entering the reactor
is held constant by continuously by-passing a varying quantity
of cold process gas around the upflow heat regenerator.  The
                              661

-------
heat that is generated  in reactor  (11) by the exothermic
reactions serves to sustain the overall heat in the system.  Upon
leaving the reactor, the main gas  flow passes downward through a
second heat regenerator  (12), giving up its heat to the packing
in that vessel before leaving the  catalytic reduction system
through flow reversing valve  (9).  A thermal balance is maintained
in the system by passing a minor flow of the hot gases from
reactor (11) around the downflow regenerator and the flow
reversing valve (9) and remixing with the main stream before
entering sulfur condenser (17)

     The primary function of the heat regenerators (10)  and (12) ,
then, is to remove heat from the gases leaving the catalytic
reactor (11),  and to utilize this heat to raise the temperature
of the incoming gases to the point where the S0«-natural gas
reaction will initiate.  The direction of flow through the two
regenerators is periodically reversed to interchange their functions
of heating and cooling the gases.  This is accomplished by the
flow reversing valve (9) and four water-cooled butterfly valves
(13), (14), (15)  and (16).  The valve arrangement shown in the
flow diagram Figure I is specially designed to maintain the gas
flow through the catalytic reactor (11) in one direction only.
All five valves are operated from  a central control system which
synchronizes their movement perfectly so that each flow
reversal is completed in less than one second.

     The elemental sulfur that is  formed in the primary reactor
system is condensed in a horizontal shell-and-tube steaming
condenser (17).  This represents over 40% of the total recovered
sulfur.   The process gas stream then enters the first stage
(18)  of a two-stage Glaus reactor  system where the following
exothermic reaction occurs:
     2 H2S + S02 	^   3/2 S2 + 2 H2O
After the first stage of Glaus conversion, the gas is cooled
and additional sulfur condensed by passage through a vertical
steaming condenser (19).  Further conversion of H2S and SO2 to
sulfur takes place in the second stage Glaus reactor (20).
This sulfur is condensed in a third steaming unit  (21).  A
coalescer (22),  containing a mesh pad, then removes entrained
liquid from the gas stream.  Molten sulfur from the three con-
densers and the coalescer is collected in a sulfur holding pit

                             662

-------
(23) from which it is pumped to storage.  Residual H_S in the
gas from the process is oxidized to SC<2 in the presence of
excess air in an incinerator (24) before being exhausted to the
atmosphere through a stack (25) .

APPLICATION IN COMBINED SYSTEMS

     The versatility of the Allied Chemical SCU reduction
technology can best be illustrated by considering its application
as a component part of several stack gas treatment systems for
electric utility generating stations.  As was mentioned earlier,
the Allied Chemical process can accommodate a wide range of SC>2
concentrations in the feed gas.  A process flow diagram of the
SOn reduction unit as it is applied to flue gas desulfurization
systems which produce strong SO  gas, say 40% or more (dry basis)
is shown in Figure II.  Since tne gases from S02 concentrating
systems are, generally cleaned by some type of wet scrubbing
operation before delivery to the SO- reduction unit, and are
saturated with water vapor at the scrubbing temperature, it is
necessary to raise the temperature of the gases sufficiently
above the dew point that no condensation of weak acid occurs
in the main blower.  After the preheated gases are compressed,
the reducing gas is added in the correct proportion and the gas
mixture then passed through a feed gas heater where its temperature
is raised above the dew point of the sulfur that is formed in the
primary reaction system.  The purpose of this is to preclude
sulfur condensation in the four-way reversing valve mentioned
in the earlier process description.  The source of this heat
will be discussed subsequently.

     The functions of the reduction reactor and the heat re-
generators are essentially the same when treating gases having
high SC-2 concentrations as when treating roaster gases.  The
exothermic reactions which occur in this primary reactor system
result in the gases leaving the system at a substantially higher
temperature.  A portion of this heat is used to heat the feed
gas in a shell and tube heat exchanger.  However, because of the
excess heat liberated in the reactor system, the exit gases are
too hot, and, as a result, are too corrosive to be passed directly
through this heat exchanger. Consequently, they are first tempered
by diverting a part through a gas cooler, and remixing this side
stream with the main stream before entry into the feed gas heater.
Some sulfur is condensed in the cooler and withdrawn to a sulfur
holding pit.  Low pressure steam is generated on the shell side
of the cooler.
                             663

-------
     The two-stage secondary reactor, or Glaus converter, system
 is  similar to that used  for roaster gases.  Residual H2S in the
 exit gas from the Glaus  system is oxidized to S02 in the presence
 of  free oxygen in an incinerator.  Instead of being released to
 the atmosphere, however, as in the roaster gas treating process,
 the incinerator tail gas is recycled to the flue gas desulfuriza-
 tion unit.  The S02 emission to the atmosphere is thus reduced
 to  the minimum level attainable.

     Still another adaptation of the Allied Chemical SC>2 reduction
 process is represented in the process flow diagram, Figure III.
 Here, the only significant difference is the quantity of  heat
 generated in the reactor-regenerator system,  with an 8% SO2,
 low oxygen gas, the temperature rise in the reduction reactor
 system is inadequate for the exit gas from the reduction reactor
 system to be used efficiently in the feed gas heater.  This makes
 it  necessary to use the  incinerator tail gas for this purpose,
 as  shown in the flow diagram.  The remainder of the process is
 the same as for the more concentrated SO2 gases.  The tail gas
 is  again returned to the front end S02 gathering system.

     For a given emission control problem, it will be apparent
 that the fixed amount of sulfur delivered into the Allied SO_
 reduction unit can be contained in greatly differing gas volumes
 depending upon the choice of flue gas treating system.  This
 is  illustrated in Table I.

     Note that the quantity of sulfur available for recovery
 in  the case of a 600 MW boiler fired with oil of 4% sulfur
 content is 130 long tons per day compared to the 500 long tons
 per day in the roaster gases at Falconbridge when operating at
 rated capacity.  It will also be seen that the volume of gas in
which that 130 long tons per day of sulfur is contained may
 range from as little as 2,500 SCFM to nearly 32,000 SCFM.  As
 an  added note of perspective, this latter value is substantially
 less than one-half of the roaster gas flow entering the S02
 reduction unit at Falconbridge.

 REDUCING AGENT REQUIREMENTS

     The reducing agent is the largest single element of operating
 cost associated with the SO., reduction process.  The consumption
varies directly with the quantity of SG>2 to be reacted, as is
 illustrated in Figures IV and V.

     The natural gas requirements for the reduction of SC<2
 recovered from coal burning electric utility stacks is shown in

                             664

-------
Figure IV.  The sulfur values on the curves are based on the use
of 10.6 tons of coal (containing 10% moisture)  per megawatt day.
Corresponding figures for electric power generation using 2%
and 4% sulfur content residual fuel oil are presented in Figure V.

     Natural gas was the logical reducing agent for the sulfur
recovery facility at Falconbridge because of the proximity of a
major trans-Canada pipeline.  One phase of Allied's continuing
effort in the development of advanced SO,, reduction technology
is focused on the substitution of other gaseous and liquid
hydrocarbons for the natural gas utilized in the present process.

CAPITAL COST COMPARISON

     The installed cost of an SO2 reduction plant is primarily
a function of the gas volume processed and secondarily the quantity
of sulfur contained in that process gas stream.  Figure VI shows
the comparative relationship of the estimated installed costs of
Allied Chemical SO2 plants applied to the three feed gas concen-
trations we have examined — 8%, 40%, and 100% S02* dry basis.

     The effect of S02 concentration on capital cost is readily
apparent, as is the fact that SO2 reduction is strongly favored
by increasing system size.

     Such a comparison does not, of course, reflect total flue
gas desulfurization system installed cost., In fact SO2 reduction
will be a minor fraction of the total installed cost in every case.

     Allied Chemical's studies to date have indicated that location
factors have a greater influence on total installed cost than on the
cost of the S02 reduction component alone.  As a consequence, it
does not automatically follow that the lowest capital cost S02
reduction unit will insure the lowest total system installed cost.

     Operating factors, particularly energy requirements and costs,
can vary widely among different combinations of  flue gas desulfuriza-
tions systems.  The optimization of operating requirements can only
be undertaken for the combined SO2 gathering and S09 reduction systems,

     Allied Chemical is prepared to work with organizations having
SO2 emission problems, making its commercially proven SO2 reduction
technology available for direct application or in combination
with SO2 concentration systems provided by others.

                             665

-------
ALLIED CHEMICAL S02 REDUCTION TECHNOLOGY TYPICAL ROASTER GAS APPLICATION
Fig. I
O>
REGEN
1
— JT
NATU
GA
MA
CO
rtEAT
ERAT(
S02 GAS
I HOT GAS HEAT , ,
1 EXCHANGERS 2
—3 L_ -JT- ' j L^
I'll
RAL J ^
5
EFFLUENT

t
i . . , „
--K ^1^ /^^ 6

3 ^ ^ x CONDENSING " frrf
^_____Z^:±:TI 4- TOWER
i — _, -^^-s- / \ luni-n .
^ 	 -^TJIC «. •
. 	 llflS •" , „- J Jl-H
in C0°L|NG -i ^^^
IU1 Tfiiurn .,/r i r"~7-N I
NEUTRALIZATION 1UiilJI ri^5^ ^| ~r

8 3 1^ f*~
N BLOWER «cmE ^P
__ J

LD GAS BY-PASS
r— M-— *-txi 	 ,
14 —^5 |
n REDUCTION
11 REACTOR
13 16
1 o
¥ C/*
j • 2j
17 Sr in HEAT
|R 1Z ^ '" REGENERATOR
_ i—
<=> i
ac f
A J
ix>
T9
— n ,, fltIO UDOUR !
DRIP ACID
1
MIST TOWER 1 "
_t \ TWn-STAGE A|n
CLAUS REACTOR ftm
^20\ NATUR
STEAM 	 *n CTC»U f7c
oltHm fc 	 STEAM UAb
/^^=J -^ STEAM — nr hir— ' ^^
i — Lt5=~t~77 m «, cumin ^21
i ^^^S' / CIHCIID IS 21 lULriiK 11*01*0
fiurae^i SULFUR ',?, r.1 rnNnrjjcrD **
SULFUR T rnunruccD 1' i iunutnatK 22
i/UNULNdLn — H — _
REHEAT GAS 1
1 SULFUR SULFUR \ SULFUR SULFUR
1 -
\ T 	 fc Tfl CTHDAPr
J f n •" olORAht
r — M
[ 23 1
SULFUR HOLDING PIT
x^
-^1
ELECTROSTATIC
.PRECIPITATORS
n
~^1
25
&_J»]
INCINER
-i ft CTi
P5
COALESCER


ATOF
CK
Allietl
rhemical

-------
ALLIED CHEMICAL SQ2  REDUCTION TECHNOLOGY  100% AND 40% SO?  FEED  (DRY BASIS)
                                                      Fig. II
SIEAM REDUCING
1 GAS ~] \ u=:=i i
tft nip 	 . 	 / \ • 	 3 	

olio bAo 	 •• — 1 	 	 f-c j 2 	 	 .
PRFHFATpT T Su FEED GAS
PRtHtATER f MAIN HEATER
BLOWER




CTF1U — -" — ~~ ~~~~~~~~-.
STEAM
i ^
, 	 T
^•fcj t^t>— ^ i
^-^K^^y J>7=CT
C|iir||R 1^7 ** iMii tTFAM *
oULrUK f J pnuutDTrnp T — *-oltABI
CONDENSER T CONVERTERS 8
cniniR ~^ cm FUR




10

outrun outrun
CONDENSER CONDENSER
°i SULFUR



. REDUCTION STEAM
REACTOR i
SYSTEM 	 1 	 Y^l^^^
REACTOR EXJT^J^]
GAS COOLER T |
1 1 1
(

STEAM i 	 1
***** FUEL *

n GAS 	 ^ 13 — '

1 , COALESCER L_ INCINERATOR
SULFUR
9 	 -Tn ^TnRARF & — A|R . —
                                                                                                  TAIL
                                                                                                  -GAS
                     SULFUR HOLDING
                          PIT       lc
12
            STEAM

-------
               ALLIED CHEMICAL SO, REDUCTION TECHNOLOGY 8% SO, FEED (DRY BASIS}
                           Fig.II!
                                                  TAIL GAS
SQgGAS
REDUCTION
 REACTOR
 SYSTEM
    4
                                                            SULFUR

                                                                TO  STORAGE
                                          SULFUR HOLDING  PIT
                         AUied

                          (jlemical

-------
_ 4
    3
GO
    1
    0
           NATURAL GAS REQUIRED FOR S02 REDUCTION
              AT COAL BURNING ELECTRIC UTILITIES
100       200      300      400       500
     ELECTRICAL POWER GENERATED-MEGAWATTS
                                               Fig. IV

-------
^  3.0

2  2.5
O
c
CO
-  2.0
1.5
00
oo
=  1.0
    0.5
     0
            NATURAL GAS REQUIRED FOR S02 REDUCTION
            AT ELECTRICAL UTILITIES BURNING FUEL OIL
               100       200       300       400       500
                    ELECTRICAL POWER GENERATED-MEGAWATTS
                                                          600

-------
$ (MILLIONS)
   10

   8
   6

   5

   4


5  3
                                                                            Fig.  VI
   2
   100
                                                      8% SO 2
                                                      40% SO2
                                                      100% S02
                                                             ALLIED CHEMICAL S02 REDUCTION
                                                                 TOTAL INSTALLED COST
                                                                  (Battery Limits Plant)
1 1 1 1 1 1
S02
1 1
COMPOSITIONS ARE DRY
1 1 1
BASIS

                      200        300    400
                     CAPACITY (MEGAWATTS)
600

-------
                            Table I
     FEED GAS TO ALLIED CHEMICAL S02 REDUCTION  UNIT
MEGAWATTS GENERATED
  100    300    600
EQUIVALENT SULFUR IN GAS (Long Tons/Day]
  21.7    65
130
% SO 2 IN GAS (Dry Basis)

          100%

           40%
VOLUME OF GAS (MSCFM)

  0.4    1.3    2.5
   1.0    3.1
6.3
                                             5.3   15.8   31.6
         Based on 4% sulfur in fuel oil assuming 90% recovery
         of SO 2 in the associated flue gas desulfurization step.

-------
 THE CAT-OX PROJECT AT ILLINOIS POWER
                    by

W. E. Miller,  Director, Environmental Affairs
           Illinois Power Company
              Decatur, Illinois
                     673

-------
                      THE CAT-OX PROJECT AT ILLINOIS POWER

ABSTRACT
The Catalytic Oxidation method developed by Monsanto Enviro-Chem Systems, Inc.
for removing sulfur dioxide from flue gas of fossil fuel generating stations has
passed the pilot plant stage.  The first commercially sized installation was
operated for a short period in September, 1972 on the 100 Mw Wood River #4 unit
of Illinois Power Company.  The project is being financed jointly by the Control
Systems Laboratory of the Office of Research & Monitoring of Federal EPA and by
Illinois Power Company.  This article describes why Illinois Power Company chose
Cat-Ox for a demonstration installation in an attempt to control sulfur oxide
emissions and also describes the operation of the Cat-Ox System.

INTRODUCTION
The removal of sulfur dioxide from the flue gas of fossil combustion units is a
major concern of public utilities since they rely heavily on fossil fuel for the
generation of electricity; therefore, much effort is being expended in close co-
operation with the chemical industry in the development of an acceptable method of
sulfur removal.  At the present time a proven method of removing sulfur dioxide
from flue gas is not available although many demonstration installations of the
scrubber type are now being installed by utilities.  The Cat-Ox System, the chemi-
cal conversion method described in this paper, has now been installed on a demon-
stration basis by Illinois Power Company.  A demonstration is considered to be a
successful installation whose scale and length of operation are great enough to
provide full determination of feasibility for commercial application.  It is hoped,
therefore, that the Cat-Ox System now installed on a demonstration basis will be
successful and can be offered for general use.
                                          674

-------
ILLINOIS POWER COMPANY




Illinois Power Company is a combination electric and gas investor owned utility




serving territories in central and southern Illinois.  To supply its electrical




needs, the five fossil-fueled generation stations shown in Figure 1 combine to




provide a capacity of 2200 Mw.









Baldwin #2, a 600 Mw coal-fired unit has been completed recently and Baldwin #3,




another 600 Mw coal unit, will be completed in 1975.  With the addition of Baldwin




#3, the company capacity will be 3,400 Mw in 1975.  Since coal is used for gener-




ating the major portion of this capacity, it is vitally important to Illinois




Power that reasonable methods of eliminating pollutants from coal either before




or after combustion be developed as rapidly as possible.









The Wood River Station of Illinois Power Company is located in Madison County




near East Alton, Illinois, which is in the St. Louis Major Metropolitan Area.




The Cat-Ox System is being installed, on Unit #4 at Wood River which has a nominal




rating of 100 Mw; the boiler is a Combustion Engineering reheat, natural circula-




tion, balanced draft, pulverized coal, tangentially fired, unit with steam condi-




tions of 1500 psig and 1005°F at the superheater outlet.  The Cat-Ox System is




designed for a flue gas flow rate of l,120,000#/hour which is equivalent to a




load of about 110 Mw.









The unit burns approximately 275,000 tons of coal per year with a sulfur content




of 3.1%.  Based on this, the Cat-Ox System theoretically should produce about




19,000 tons per year of equivalent 1007. sulfuric acid.






                                       675

-------
 CAPITAL AND OPERATING COSTS




 The capital costs involved in the design and construction of the Cat-Ox System




 were estimated at $7,300,000, including the costs of providing off-battery ser-




 vices.  Of this amount Illinois Power Company contracted to pay $3,800,000 and




 the Office of Research and Monitoring of Federal EPA contracted to pay $3,500,000.




 The battery limit services include supply facilities for natural gas and fuel oil




 for reheat burners, cooling water for acid coolers and miscellaneous uses, electric




 services for ID fans, electrostatic precipitators and pumps and facilities for ash




disposal.  Final costs now appear to be $8,150,000, giving us an installed cost of




 $74 per Kw based on the 110 megawatt gross capacity of the generating unit.








The Federal EPA will undertake an extensive test and evaluation program to fully




characterize the Cat-Ox System and to provide data on its emission reduction capa-




bilities as well as the operating costs for the system to form a basis for compari-




 son with other installations.  It is expected that the data will be made available




 to all interested parties and will be used by EPA, regulatory agencies, and members




of the utility industry.








For a period of five years, Enviro-Chem will be responsible for the disposal of




the sulfuric acid produced by the system with net revenues from acid sales credit-




ed 757o to Illinois Power  and 25% to Enviro-Chem.  Operating and maintenance costs




are expected to run $600,000 per year after the initial de-bugging period; assuming




a price of $8 per ton of  acid,  the net operating and maintenance costs should be




about $500,000 per year,  excluding fixed charges on capital costs.
                                          676

-------
THE CAT-OX PROCESS




The operation of the Cat-Ox System consists basically of the following  six




separate phases:




          1.  Fly Ash Collection




          2.  Conversion of S02 to 503




          3.  Heat recovery




          4,  Removal of sulfuric acid




          5.  Acid mist elimination




          6.  Acid storage and loading




These basic steps are shown diagrammatically in Figure 2 and are described below.




          1.  Fly Ash Collection




              The existing mechanical collector remains in service on Unit #4




to remove most of the fly ash from the flue gas.  A new Research-Cottrell electro-




static precipitator with a design efficiency of 99.6% has been installed to work




in series with the mechanical collector to remove essentially all the particulate



matter from the flue gas.  After leaving the electrostatic precipitator, the cleaned




flue gas is heated and passes into the converter of the Cat-Ox System or, during




start-up or unusual operation, can be by-passed directly to the stack.  The fly




ash collected by the precipitators is conveyed pneumatically to the existing ash




pit area.  The electrostatic precipitator installation was completed in February




1972 and has been operating with Unit #4 since that time.




          2.  Conversion of S02 to SO-j





              The temperature of the flue gas leaving the electrostatic precipi-



tator is 310°F and must be reheated to 850°F to allow a 90% conversion of S02 to




503.  It was proposed that this be done by two in-line reheat burners using




natural gas or No. 2 fuel oil and by recovery of sensible heat from the treated





                                          677

-------
 flue  gas.  The reheat burners were to be designed to maintain the 850°F conversion




 temperature regardless of boiler  load.  Modifications are now being made on this




 re-heat  system to permit operation totally on #2 oil.  Following reheat to con-




 version  temperature, the flue gas enters the converter where the Cat-Ox \^} catalyst




 (a vanadium pentoxide catalyst) reacts with the S02 to form 803.  The converter is




 designed so that the catalyst bed can be emptied onto a conveyor system for trans-




 port  to  a screening process after which the cleaned, catalyst is conveyed back to




 the converter.  About 2.57, of the catalyst mass is lost during each cleaning process




 which is anticipated to occur about four times per year.  About forty-eight hours




 is required for each catalyst cleaning.




          3.  Heat Recovery




              The treated flue gas, now containing 803, passes to a Ljungstrom-




 type heat exchanger where about 400°F sensible heat is recovered to heat the in-




 coming untreated flue gas.  As a result of heat recovery in this exchanger, the




 overall  need for fuel usage is to add 150°F of sensible heat in the Cat-Ox process.




 The temperature of the gas is maintained well above the dew point.  Normal flue gas




 leakage  in a regenerative heat exchanger of this type will allow about 5% of the




 flue gas to by-pass the unit thereby reducing the overall efficiency of S02 removal




 to approximately 85%.




          4.  Removal of Sulfuric Acid




              The flue gas is further cooled in a packed-bed absorbing tower




which operates in conjunction with an external shell and tube heat exchanger.




During cooling, the H20 and 863 combine to form sulfurlc acid which is subsequent-




 ly condensed.   The tower brings a cool stream of sulfuric acid into direct contact




with the rising hot flue gas.  Exit gas leaves the packed section at about 250°F




while hot acid is constantly being removed from the bottom of the tower and cooled
                                          678

-------
in the external heat exchanger and sent to storage.




          5.  Acid Mist Elimination




              Very fine mist particles of sulfuric acid are formed, in the gas




as it is cooled in the absorbing tower.  These mist particles in the flue gas




are removed along with some entrained droplets of circulating acid from the




tower by the Brink vS/ mist eliminator system.  The packed section of the absorb-




ing tower and. the mist eliminators are contained within one vessel.  The flue gas




leaving the mist eliminator to enter the exit stack contains less SOo than the




amount normally emitted from the combustion process.




          6.  Acid Storage and Loading




              The cooled acid amounting to 12 gallons per minute of 78% l^SO^ is




collected in two 400,000 gallon steel storage tanks.  An acid loading pump and




tank car loading facilities are provided adjacent to the storage tanks.  Tank




trucks may be loaded from this station if desired.






PROJECT TIMETABLE




The time schedule for the Cat-Ox System includes:




     Design and Capital Cost Estimates - Initiated June  1970




     Detailed Engineering & Equipment Procurement - Initiated November  1970




     On Site Construction - Initiated January 1971




     Electrostatic Precipitator - Placed in Operation February  1972




     Entire Cat-Ox unit scheduled for operation mid-1973




     Data Phase including one year operational testing by Federal  EPA  *




          estimated to be completed October 1974.






DELAY IN TIMETABLE




The Research-Cottrell precipitator was completed on schedule and was placed  in




operation in February, 1972.  The Cat-Ox System was scheduled to begin operation



                                         679

-------
 in June, 1972 but, due to construction delays,  initial start-up using gas as the




 in-line heater fuel did not occur until September 4,  1972.   Preliminary opera-




 tions in September and October indicate that the unit is meeting the performance




 criteria set for the system.   Conversion of S02 to 303 reached 93% which is  over




 the 90% guaranteed.  The Cat-Ox unit removed 89% of S02 admitted to it which is




 over the 85% guaranteed.  Acid of 78% concentration was produced when operating




 the absorbing tower at design temperature.   Acid mist exiting  the system was below




 0.3 mg/cu.  ft.  against a design criterion of 1.0 mg/cu. ft.  The system handled  the




 full 1,120,000  Ib/hr.  of flue gas when meeting  these  criteria.








 Over 1,400  tons of acid were  produced  during this period and two tank cars of the




 acid have been  sold to a fertilizer  manufacturer to assure its  acceptability.








 In  October  it became evident  that the  shortage  of natural gas  would  preclude its




 use as  a reheat fuel for the  Cat-Ox  unit.   The  existing burner  equipment in  the




 Cat-Ox  had  to therefore be modified  to  operate  totally on #2 fuel oil.   Design




 modification and equipment changes are  ia progress.   As soon as they are completed,




 the Cat-Ox  System  will  again  be brought on  stream.








 In  a  new generating unit, the necessity for  reheat  would  not exist.   The flue  gas




 would be taken  from the boiler at a  point near  the  economizer or super-heater  at




 a temperature of 850-900°F.  A hot precipitator  would then remove the particulate




matter  and  let  the  flue gas enter the converter  at  the  required 850°F.   For  this




 reason, based on the limited operation we have had  to date, we  believe the Cat-Ox




 System might be a feasible system for installation  in a new coal  fired station.




 Retrofit application will be proven or disproven by the Wood River #4 installation-




                                      680

-------
GUARANTEE AND OPERATING PROGRAM
One or two 24-hour tests by Enviro-Chem will be conducted to assure that the
Cat-Ox System meets the following guarantees:
     1.   The system is capable of operating with a gas flow of
          l,120,000#/hour entering the system at 310°F.
     2.   The system is capable of producing 60° Baume' (77.7% l^SO^)
          sulfuric acid.
     3.   The exit gas emitted to the stack does not contain, on the average,
          more than 1.0 milligrams of 100% sulfuric acid mist per actual cubic
          foot of gas when the system is operated at rated capacity.
     4.   The conversion of S02 to S03 of the gas entering the converter
          is at least 90% at rated capacity.
     5.   The system operates so that over 99% of the fly ash in the flue
          gas leaving the boiler is removed when operating at rated capacity.
     6,   The system removes 85% of the S02 in the flue gas entering the system.

Following the 24-hour test period, Illinois Power Company will operate the Cat-Ox
System for a minimum of one year as mentioned previously.  During this period, and
for a subsequent period of four years, if Illinois Power Company decides to continue
to operate the system, data will be obtained to evaluate the following items:
     1.   Operating characteristics and plant performance (relative to
          S02 and fly ash removals and to H2SO^ recovery).
     2.   Maintenance procedures, requirements, and costs,
     3.   Total process operating costs.
Figures 3 and 4 show the Cat-Ox installation as it nears completion.

SUMMARY
In an attempt to advance the frontiers of knowledge in the science of removing
sulfur dioxide from flue gas, Illinois Power, with Battelle Institute, conducted
                                       681

-------
 a comprehensive survey of possible sulfur removal systems.  The possibility of

 using low sulfur Western coal was considered but it was found to be less efficient

 than local coal, and precipitator efficiency declined with its use; in addition,  it

 cost three times as much.  As a result of these studies,  Illinois Power Company

 decided  that the Cat-Ox System, after both a pilot installation and prototype

 installation,  was most nearly ready for commercial demonstration.  In 1970  the

 capital  costs  being considered for Cat-Ox were  much higher than those proposed for

 other S02 control systems.   However,  as more actual experience has been gained

 since 1970,  it seems that the capital costs originally estimated by Enviro-Chem

 for  Cat-Ox were more realistic than those estimated for other systems,  primarily

 because  of the decade of  research and the advanced stage  of development of  the

 Cat-Ox System.


 The  Cat-Ox System is expected  to  remove 85% of  the sulfur dioxide from  the  flue

 gas  as 787o sulfuric  acid  and  to remove essentially all of the fly ash.   Sale of

 the  sulfuric acid  could offset  some of the  operating  costs.   Illinois Power Company,

 by removing  the pollutants  from the environment  and  conserving natural  resources

 by recovering  a product which  is  presently  being  thrown away,  is not  solving One

 pollution problem and  creating  another one.  While some start-up difficulties  have

 been  encountered  with  the Cat-Ox  System,  we  are  still  hopeful  it will prove to be

 a feasible method  and  that  it may  solve the  sulfur removal  problem in a  new coal

 fired generating unit  and. possibly even in  a retrofit  installation.


ACKNOWLEDGMENT

We hereby express grateful acknowledgment to the Control  Systems  Laboratory of

the Office of Research and Monitoring  of  the Federal Environmental  Protection

Agency for their financial assistance  and excellent cooperation  in  this major

project.
                                       682

-------
                              .'> .      qw«u_J:,	.L^;	r	„_!
FIGURE 1   TERRITORY SERVED  BY ILLINOIS POWER COMPANY
                           683

-------
THE REHEAT
   CAT-OX
   SYSTEM
    FLUE GAS
   FROM EXISTING
    ID FAN
       PRECIPITATOR
                              CONVERTER
REHEAT;*
BURNERn
                      -DAMPERS
t      I
                              GAS HEAT
                              EXCHANGER
                                SULFURIC
                                  ACID
                                STACK
 CAT-OX
  MIST
ELIMINATOR

ABSORBING
 TOWER
                     ACID
                    COOLER
                                RECYCLE
                                                       STORAGE
                            FIGURE 2

-------
                                              Monsanto
3
CO
tn
Monsanto
Envjrol
 IChem
   Systems Inc
                REHEAT CAT-OX SYSTEM

-------
      REHEAT CAT-OX SYSTEM
ILLINOIS POWER - WOOD RIVER UNIT NO. 4
                686

-------
      MITRE TEST SUPPORT
FOR THE GAT-OX DEMONSTRATION
            PROGRAM
                by

            G.  Erskine
           E. Jamgochian
       The Mitre Corporation
         McLean, Virginia
                  687

-------
                           ACKNOWLEDGMENTS






     The work described In this paper wae performed by the MITRE



Corporation under the sponsorship of Mr. G. S. Haselberger of the



Control Systems Laboratory of the Office of Research and Monitoring,



U. S. Environmental Protection Agency.  Members of the MITRE team



who participated in the program include:  Dr. J. S. Burton, Mr.  G. Erskine,



Mr. E. M. Jamgochian, Mr. N. T. Miller, Mr. J. P. Morris, Mr. R. Reale,



and Mr. W. L. Wheaton.  Contributions to the program were also made by



Mr. D. C. Simpkins of Consultants and Designers, Inc., and members of



the Midwest Research Institute under subcontract to the MITRE Corporation.
                                688

-------
                                ABSTRACT






     This paper describes MITRE test support for the Cat-Ox Demonstration




Program.  The overall scope of MITRE*s test program is outlined, and




specific task efforts accomplished to date are presented.  A detailed




description is provided on the individual tests of the Baseline Measure-




ment Program, and select test results are provided concerning:  grain




loading measurements, comparisons of continuous and manual measurements




of S02 with theoretical values, comparisons of continuous and manual




measurements of NO , sulfur balance measurements, and comparisons of
                  Jt



elemental concentrations of coal and ash throughout the steam generator




system.  The paper further discusses current activity in preparation




for the One-Year Demonstration Test Program.  In particular, the




measurement parameters and design of the continuous measurement instru-




mentation system for the evaluation of process performance are presented;




and the instrumentation and facilities for the measurement of gas con-




centrations, gas flow, and automatic recording of data are reviewed.
                              689

-------
                          MITRE TEST SUPPORT
                                FOR THE
                      CAT-OX DEMONSTRATION PROGRAM
Background

     MITRE's contract with the Environmental Protection Agency was

initiated in April of 1971.  Under this contract, MITRE is required to

provide technical support for the Cat-Ox Demonstration Program in terms

of:  1) program management assistance, 2) the development of test

plans, 3) the design, development, and installation of measurement

systems, 4) services related to the implementation of test programs,

and 5) evaluation of test results with application of the results to

the total industry.

     As a first step, a number of program plan alternatives were

developed and evaluated.  These plans, and the details of the overall

program plan which was selected, are described in our Management Plan

for Test Support for the Cat-Ox Demonstration Program (MITRE Technical

Report No. 6054 dated July 1971).  A summary of the MITRE task efforts

defined for the program is provided in Figure 1.


Major Task Areas

     The first of the task areas concerns the "Definition of Test

Requirements," in which MITRE examined the requirements of potential

users, reviewed the technical and economic capabilities of the Cat-Ox

process, and defined the "window" of test conditions at the steam

generator/Cat-Ox interface for the demonstration test (see Figure 2).

This task was completed in November of 1971 with the definition of

                               690

-------
         TASK 1
         TASK 2
         TASK 3
cr>
vo
         TASK 4
DEFINE TEST REQUIREMENTS





BASELINE MEASUREMENT PROGRAM





CAT-OX DEMONSTRATION



-  TEST PREPARATION



-  ONE-YEAR MEASUREMENT PROGRAM





EVALUATION OF RESULTS
(COMPLETED)





(COMPLETED)
TO BE COMPLETED AUGUST 1973



SEPTEMBER 1973 - SEPTEMBER 1974





SEPTEMBER 1974 - MARCH 1975
                                               FIGURE 1



                                           PROGRAM SUMMARY

-------
                                EXAMINE  REQUIREMENTS OF POTENTIAL USERS
                             •   SPECIFY  TECHNICAL &  ECONOMIC CAPABILITIES
                                OF CAT-OX PROCESS
o>                            •   DEFINE  "WINDOW" OF TEST  CONDITIONS AT
£                               STEAM GENERATOR/CAT-OX INTERFACE  FOR
                                DEMONSTRATION
                                              FIGURE 2

                               TASK 1  -  DEFINITION OF  TEST  REQUIREMENTS

-------
interface flue gas conditions (temperature, gas flow rate, SO. concen-



tration, and grain loading) for 89 existing steam generators which



were categorized as potential users of the reheat Cat-Ox system.



Interface flue gas conditions were also determined for 75 steam genera-



tors to be constructed by 1976 which were categorized as potential



users of the integral design Cat-Ox system.




     The second of our task areas concerns the "Baseline Measurement



Program."  The specific objectives of the Baseline Measurement Program




as shown in Figure 3 are to 1) characterize baseline performance in



terms of operability, reliability, and emission levels prior to



installation of the Cat-Ox process; 2) to determine the relationship



between control settings and operating conditions for the steam



generator, and flue gas properties at the Cat-Ox steam generator




interface; 3) to test and calibrate measurement procedures and hardware



to be used in the one year demonstration test; and 4) to obtain



quantitative data supporting the establishment of realistic performance



standards for pollutants other than SO- and particulates.




     The third task area concerns the performance of a "One-Year



Demonstration Test" wherein a measurement program is conducted which



will fully characterize Cat-Ox process emission control performance;



quantify the operating economics of the process; establish the




operability, reliability, and maintainability of the Cat-Ox process;



and determine the resulting effects on the steam generator with which



the process is integrated.
                                 693

-------
•  DETERMINE BASELINE EMISSION LEVELS, OPERATING EFFICIENCY, AND
   RELIABILITY OF STEAM GENERATOR IN ACCORDANCE WITH A.S.M.E.
   POWER TEST CODES
•  DEFINE STEAM GENERATOR CONTROL SETTINGS AND OPERATING CONDITIONS
   TO PRODUCE DESIRED FLUE GAS PROPERTIES AT THE CAT-OX INTERFACE
   TEST AND CALIBRATE MEASUREMENT PROCEDURES AND HARDWARE TO BE USED
   IN THE ONE-YEAR DEMONSTRATION
   PROVIDE EMISSION INFORMATION TO EPA TO SUPPORT THE ESTABLISHMENT
   OF REALISTIC EMISSIONS STANDARDS
                          FIGURE 3

                BASELINE MEASUREMENT OBJECTIVES

-------
     The fourth task area concerns the "Demonstration Evaluation" where



reduced test data are translated into quantified statements on the



technical and economic adequacy of the process.






Scope of Baseline Measurement Program



     To meet the objectives of the Baseline Measurement task (Task 2),



a test program was initiated with six preliminary tests wherein



necessary background information was obtained on isokinetic sampling



techniques, rates of particulate loading in sampling equipment, and




effects of power plant ambient conditions on the measurements.  This



information was then used to determine ranges of gas and particulate



concentrations to be encountered in the measurement program, to confirm



the sampling frequency and sampling positions utilized in the measure-



ment program, and to identify any changes in operating procedures or



modifications in test equipment necessary for the primary measurement



effort.  These preliminary tests were conducted over a two-week period



beginning September 28, 1971 (see Figure 4).




     The main measurement program was initiated on November 8, 1971




following the completion of the preliminary measurement effort and the



completion of all required facility modifications (ports, sheds, and



stack access platform),  Each of the twenty-one tests of the main



measurement program was of approximately 10 hours duration.  During



the 10 hour period of each test, the major steam generator parameters




(load factor, fuel type, soot blowing cycle, excess air, and burner



angle) were fixed at a predetermined operating level.  The combinations





                                 695

-------
en
(£>
                6 PRELIMINARY TESTS BASED ON:

                -  LOAD FACTOR:  35%, 100%
                -  EXCESS AIR:  MINIMUM, MAXIMUM
                -  SOOT BLOWERS:  NONE, RETRACTABLE AND WALL BLOWERS
                -  TEMPERATURE AND VELOCITY TRAVERSE AT ECONOMIZER
                -  TEMPERATURE, VELOCITY, AND GRAIN LOADING TRAVERSE AT AIR HEATER
             •   21  INDIVIDUAL TESTS BASED ON:
                 -   LOAD FACTOR:  35%, 50%, 75%, 100%
                 -   FUEL TYPE:  3 LBS S/M BTU, 1.5 LBS S/M BTU, 1.0 LBS S/M BTU
                 -   SOOT BLOWING:  NONE, RETRACTABLE BLOWER AND WALL BLOWERS
                 -   EXCESS AIR:  MINIMUM, NORMAL, MAXIMUM
                 -   BURNER ANGLE:  NORMAL
             •  2 TESTS BASED ON:
                -  TEMPERATURE, VELOCITY, AND GAS CONCENTRATION TRAVERSALS AT STACK AND AT
                    AIR HEATER
                -  LOAD LEVEL;  100%, 50%
                -  FUEL TYPE:  3 LBS S/M BTU, 1.5 LBS S/M BTU
                 1  TEST  BASED ON:

                 -  NO   GAS  CONCENTRATIONS AT STACK AND ECONOMIZER
                 -  BURNER ANGLE:  MINIMUM THROUGH MAXIMUM
                                               FIGURE 4

                                      BASELINE MEASUREMENT PROGRAM

-------
 of operating levels  were  selected so  as  to provide  the maximum of



 information in a minimum  number  of tests, varying the parameters on



 a "one-at-a-time" basis.



      Two  supplementary  gas  traversal  tests were also conducted to



 determine the pattern of  leakage at the  air heater  and the gas flow



 pattern midway in the stack.



      A supplementary test was also conducted in which all factors were



 held  constant except for  burner  angle, which was varied in steps from



 the minimum to maximum  position.





 Baseline  Measurement Parameters



      For  all  of the  tests, key steam  generator operating parameters



were  monitored, samples of coal  and ash were obtained at various points



 in the steam  generator, gas samples were manually obtained, particulate



 grain loadings were  determined by manual sampling, and temperatures,



pressures,  gas flows and  gas concentrations were monitored by a MITRE



designed  continuous measurement  system.



      These key parameters were measured as shown on Figure 5 at three



locations:  1) Location 1 - Prior to  the economizer, was selected to



provide data at the relatively high temperatures corresponding to the



inlet of an integral Cat-Ox system designed for installation in new



steam generators, 2) Location 2 - between the upper and lower sections



of the air preheater - selected as representative of a lover temperature



condition prior to any existing or planned flue gas treatment system,



and 3) Mid stack - approximately 135 feet above the foundation of the





                                697

-------
              LOCATION 1 PRIOR TO ECONOMIZER
              LOCATION 3 MIDWAY IN STACK
TEMPERATURE
PRESSURE
FLOW
GAS COMPOSITION (SO , CO, C02, HC,
   NO, N02, 02, H20^VAPOR)
01
V£>
00
              LOCATION 2 IN AIR PREHEATER
              LOCATION 3 MIDWAY IN STACK
TEMPERATURE
PRESSURE
FLOW
GRAIN LOADING
PARTICULATE COMPOSITION
PARTICLE SIZE DISTRIBUTION
ADSORBED MATERIALS
                                               FIGURE 5
                                      BASELINE MEASUREMENT PARAMETERS

-------
 250 foot stack,  selected to provide a condition representative of the




 flue gas emitted to  the atmosphere (and,  in this instance,  conditions




 which will be seen at the reheat Cat-Ox/steam generator interface).




 At  all  locations,  the ports and sampling  points were selected so as  to




 provide the proper number and distribution of measurement points in




 accordance with  ASTM Standard D 2928-71 ("Standard  Method for Sampling




 Stack for Particulate Matter").






 Results of Baseline  Measurement Program




      A  brief summary of the results of the Baseline Measurement Program




 is  provided in Figure 6.   This  paper summarizes a few of our  major




 findings,  whereas  the detailed  results of  our Baseline program are




 presented in a report to  be distributed by the Office of Research and




 Monitoring ("Baseline Measurement  Test Results for  the Cat-Ox Demonstra-




 tion Program," April 1973,  Report  No.  EPA-R2-73-189).




      Representative  grain loading  measurements are  shown in Figure 7.




 These measurements were taken at our position number  3,  midway in the




 stack.  All measurements  were taken using  the EPA particulate sampling




 train (with impingers)  as described  in the  Federal  Register,  Vol. 36,




No.  247, December  23,  1971.   The upper set  of data  shown in Figure 7




 summarizes  the results  from a series  of tests which were run  with




 coal from  the  Southwestern Mine of  the Peabody Coal Company.   Ranges




 are  shown  in the data set for the  sulfur content, ash  content,  and




 the measured grain loading, representing the  extremes  of  the  measure-




ment.   (As  for example, six tests were run at  the 100 mw power  level






                                  699

-------
                  COMPLETE CHARACTERIZATION OF POWER PLANT  IN TERMS OF:



                     •  TEMPERATURES,  PRESSURES,  AND GAS  VOLUME  FLOWS  THROUGHOUT  SYSTEM



                     •  GAS CONCENTRATIONS THROUGHOUT  SYSTEM AND EMITTED AT  STACK



                        -  S02>  02,  C02,  CO, HO,  NO, N02, H20 VAPOR



                     •  GRAIN LOADING  AT  AIR HEATER AND EMITTED  AT STACK



                     •  COAL FEED COMPOSITION: PROXIMATE, ULTIMATE, AND  ELEMENTAL ANALYSES



                     •  FLY ASH  PARTICLE  SIZE, COMPOSITION,  ADSORBED MATERIAL
-j
o

0                    •  BOTTOM ASH COMPOSITION
                        COMPLETE SULFUR BALANCE



                        "MAPPING" OF STRATIFIED GAS  FLOW AT AIR HEATER AND STACK



                        N0x EMISSIONS RELATED TO BURNER ANGLE







                                                FIGURE 6



                                        BASELINE MEASUREMENT RESULTS

-------
       ASH CONTENT
LOAD
FACTOR
100 MW
75 MW
50 MW
100 MW
75 MW
50 MW
FUEL
TYPE
3.24-3.40%S
2.87-3.37%S
2.74-3.47%S
1.65%S
1.83%S
1.68%S
OF COAL
(AS RECEIVED)
9.9-10.7%
9.8-10.4%
9.6-10.8%
16.7%
13.6%
16.0%
GRAIN LOADING
(GRAINS/ACF)
.6-. 95
.5-. 57
.6-. 75
.78-. 90
.59
.76
EMISSION RATE
(LBS/HR)
2,100-3,000
1,000-1,200
650-1,000
2,500
1,300
1,100
             FIGURE 7
BASELINE GRAIN LOADING MEASUREMENTS
        (MIDWAY IN STACK)

-------
for which the fuel for each individual test ranged from 3.24% sulfur




to 3.40%)  The lower set of data shown in Figure 7 summarizes the




findings from three tests each run at a single power level, using as




fuel a mixture of natural gas and coal from the Orient No. 3 mine in




Jefferson County, Illinois.  The coal and gas were fired in combination




so as to simulate fuel with a sulfur content of 1 Ibs Sulfur/10  BTU.




The values shown in Figure 7 do, however, represent the actual measured




percentage sulfur and ash of the coal constituent of the mixture.  The




data show that although the ash content of the coal in the lower set




of data was substantially higher than the upper set, grain loadings




and emission rates were comparable because of the diluting effect of




the natural gas fired with the coal.




     Data from the same tests are shown in Figure 8, compared with




respect to measured SO,, concentration.  Again, the lower data set




which shows sulfur content of the coal as 1.65%S, 1.83%S, and 1.68%S




is fuel equivalent to 1 Ibs S/10  BTU because of the blending with




sulfur-free natural gas.  The theoretical values shown for S0_ con-




centration (and the other measured values shown for SCL concentration)




are based upon the actual measured excess air, they are not corrected




to any standard excess air condition.  The manual SO,, values were




obtained using the procedures defined by EPA in the Federal Register




for determining sulfur dioxide emissions from stationary sources




(Federal Register, Vol. 36, No. 247 - December 23, 1971).  The




continuous S0? values were obtained using MITRE's continuous measure-




ment instrumentation system, which included a Dupont 460 UV analyzer





                                 702

-------
o
w
LOAD
FACTOR
100 MW
75 MW
50 MW
100 MW
75 MW
50 MW
FUEL
TYPE
3.24-3.40%S
2.87-3.37%S
2.74-3.47%S
1.65%S
1.83%S
1.68%S
MEASURED
EXCESS AIR
36-64%
27-52%
38-69%
41%
35%
34%
THEORETICAL
(PP&)
1760-1990
1780-2170
1760-2110
580
755
533
MANUAL
so,
(PPH)
1300-1850
1300-1430
360
425
445
385
CONTINUOUS
(PP&)
1755-2040
1740-2080
1780-1815
535
630
565
                                                 FIGURE  8

                                      COMPARISONS OF CONTINUOUS & MANUAL
                                MEASUREMENTS OF S02 WITH THEORETICAL VALUES
                                             (MIDWAY IN  STACK)

-------
for SO. measurement.  Both the upper and lower data sets shown in




Figure 8 are typical of the results of the total test effort where




good agreement was found between the continuous measurements and the




theoretical measurements, whereas the manual measurements were generally




lower than the theoretical values (averaged 76% of the theoretical




values).



     Figure 9 shows a comparison between continuous and manual measure-




ments of NO  for the same set of tests.  The manual measurements were
           j£


obtained using the standard EPA method for determination of nitrogen




oxide emissions from stationary sources as described in the previously




cited Federal Register.  The continuous measurement results were




obtained using a Dupont 461-C UV instrument which measured both NO and




NO..  The results shown in the continuous measurement column are




primarily NO.  No consistent patterns were found with respect to the



effect of test conditions on NO  concentrations, except for the tests




performed with gas and coal fuel firing which produced NO  levels
                                                         A



significantly lower than in the tests performed with the other fuel types.




     Representative sulfur balance calculations are shown in Figure 10.




For each of the tests, the average coal consumption rate was determined




utilizing coal scale readings.  The average sulfur content of the coal




(as determined by chemical analysis) was then used with the coal scale




readings to determine the rate of sulfur feed to the steam generator.




Average SO, mass flow readings from the continuous measurement system




were then used to determine the average sulfur flow from the stack.




The results are based upon measurement of SO,, concentrations in the






                                 704

-------
o
VI
LOAD
FACTOR
100 MW
75 MW
50 MW
100 MW
75 MW
50 MW
FUEL
TYPE
3.24-3.40*8
2.87-3.37ZS
2.74-3.47ZS
1.65ZS
1.83ZS
1.68ZS
MEASURED
EXCESS AIR
36-64%
27-52Z
38-69Z
41Z
35Z
34Z
MANUAL
NO
(Ppfi)
320-650
270-450
325-405
265-365
308
235-260
CONTINUOUS
NO
(PPfi)
285
330-345
345
105
240
150
                                                 FIGURE 9



                                COMPARISONS OF CONTINUOUS & MANUAL MEASUREMENTS

                                           OF NO  (MIDWAY IN STACK)

-------
                                             AVERAGE
                                             SULFUR FEED IN
AVERAGE
SULFUR FLOW FROM
VJ
O
LOAD
FACTOR
100 MW
75 MW
50 MW
100 MW
75 MW
50 MW
FUEL
TYPE
3.24-3.40%S
2.87-3.37%S
2.74-3.47%S
1.65%S
1.83%S
1.68%S
COAL
(LBS/MIN)
35.8-46.0
31.7-36.0
21.5-29.2
10.4
12,8
7.3
STACK AS SO.
(LBS/MIN)
31.8-45.5
26.5-35.1
19.8-22.9
10.0
11.3
6.3
                                               FIGURE 10

                                      SULFUR BALANCE MEASUREMENTS

-------
stack and do not include measurement of the sulfur exhausted from the



stack as SO,, or the sulfur adsorbed on the surface of the ash through-



out the system.  The results, however, show good agreement on the



sulfur balance leading to the conclusion that the total combined error




in SO2 and gas flow measurements was'low.  For all except two tests,



the sulfur feed rate exceeded the sulfur flow rate measured in the



stack indicating that there were, in fact, small unmeasured losses of




sulfur.



     Trace element concentrations were determined in four tests on the




coal feed, pulverizer rejects from the mills, bottom ash, and the fly



ash collected from the bottom of the air heater, the mechanical collector,



and in the duct at the air heater and at mid stack.  Trace element con-



centrations were also determined for samples of fly ash collected in




the duct at the air heater and at mid stack for four additional tests,



as well as trace elemental analysis of pulverized coal for six additional



tests.  In all cases, the analysis was performed for approximately 27




elements using atomic absorption.  A sample comparison from these



analyses is provided in Figure 11 for thirteen of the elements of



special interest aa potentially hazardous materials.  An exact material



balance of the elements cannot be performed because of the  fact that



certain material flows such as bottom ash and ash from the mechanical




separator were not measured and can only be estimated.  This  type of



balance is also precluded by the fact that many elemental concentrations



are expressed as "less than" values.  The results are, however, of




special value in estimating emission rates for a number of  elements





                                 707

-------
CO
ELEMENT
Ba
Be
Cd
Cr
Cu
Hg
Mn
Ni
Pb
Se
Sn
V
Zn
PULVERIZED
COAL
<.03
<.0002
.0006
.002
.002
<.0002
.009
.009
< 0.003
<.06
<.05
<.02
.11
PULVERIZER
REJECTS
<.03
<.0002
.0004
.002
.003
.00006
.007
.0009
< 0.003
<.06
<.05
<.02
.011
BOTTOM
ASH
	 _
.0004
<.005
.016
.007
.0003
.057
.013
.009
<.05
<.l
.04
.038
ASH FROM
MECHANICAL
SEPARATOR
.03
.0008
.002
.013
.007
.00004
.036
.040
< 0.003
<.07
<.05
<.02
,057
ASH FROM
AIR HEATER
DUCT
<.04
.001
	
.05
.010
<. 00002
.050
.050
0.020
<.05
<.05
.02
____
ASH FROM
MID STACI
<.04
.001
.002
	
.020
	
.080
	
0.020
<.05
<.05
.03
.090
                                               FIGURE 11

                               COMPARISONS OF ELEMENTAL CONCENTRATIONS IN COAL
                                          AND ASH THROUGHOUT SYSTEM
                                               (WEIGHT PERCENT)

-------
not usually examined In emission testing programs.  For this purpose,



the elemental concentrations mast be combined with the mass loading



measurements to provide the emissions data In useful form.   An estimate



of the overall accuracy of the measurements can be noted by comparing



the values given for the ash from mid stack and with ash from the air



heater duct.  The values from these two columns should be nearly



Identical, and as such represent essentially duplicate measurements.





Cat-Ox Demonstration Teat



     Task 3 of MITRE's support program concerns the Cat-Ox Demonstration



Test.  Preparation for the demonstration test Is currently underway.



This preparatory work Is scheduled to be completed In August of 1973,



after which MITRE will Initiate our one-year measurement program In



September of 1973.



     The major areas of concern which MITRE will Investigate In the



one-year test program are summarized In Figures 12. and 13.  A test



plan has been drafted which examines these areas In a systematic step-



by-step fashion.  The prime area for Investigation concerns the overall



operating characteristics and performance of the Cat-Ox system relative



to the removal of S02 and fly ash and the production of sulfuric acid.



Other areas, Include the effectiveness of the catalyst as a function



of time, the effectiveness of the mist eliminator and any special



requirements for maintaining its effectiveness, the degree of seal



leakage encountered in the Ljungstrom rotary heat exchanger,



corrosion rates of the plant equipment, and the response of the Cat-Ox
                                 709

-------
   OPERATING CHARACTERISTICS AND PLANT PERFORMANCE (RELATIVE
     TO S02 AND FLY ASH REMOVAL AND H2SO, RECOVERY)
•  LONGEVITY OF THE CATALYST


•  NECESSITY AND FREQUENCY OF MIST ELIMINATOR WASHING OPERATIONS


•  DEGREE OF REGENERATIVE HEAT EXCHANGER SEAL LEAKAGE WITH TIME



                         FIGURE 12

       CAT-OX DEMONSTRATION TEST - MAJOR AREAS OF CONCERN

-------
•  CORROSION RATES OF PLANT EQUIPMENT (WITH SPECIAL EMPHASIS ON THE
     ACID LOOP AND HEAT TRANSFER EQUIPMENT)
•  RESPONSE OF PROCESS TO FUELS OF VARYING SULFUR CONTENT


•  EFFECT OF CAT-OX SYSTEM FAILURE ON POWER PRODUCTION


•  COMPONENT PRESSURE DROPS (AS A FUNCTION OF TIME)


•  ABILITY OF POWER PLANT PERSONNEL TO OPERATE AND MAINTAIN SYSTEM
                        FIGURE 13

     CAT-OX DEMONSTRATION TEST - MAJOR AREAS OF CONCERN

-------
system  to  fuels of various sulfur content.  Also of interest in our




test program is the possible effect of Cat-Ox system failure on power



production, pressure drops across system components as a function of



time, and  the ability of power plant personnel to operate and maintain



the Cat-Ox System.






Manual  and Continuous Measurements



     Both  manual and continuous measurement instrumentation will be



used for evaluation of the Cat-Ox process performance.  Figure 14



states  the reasons for utilizing both methods.  With regard to the



measurement of particulates, sulfur trioxide and sulfuric acid mist,



there is presently not available completely automatic continuous



measurement instrumentation, and, therefore, manual techniques must




be employed.



     Continuous recording of data has several advantages.  One advantage



is the  ability to obtain immediate indication of results instead of



having  to  wait for several days until sample solutions are analyzed




in the  laboratory.  Also real-time availability of data provides an



early check on the performance of the measuring instrumentation so



that necessary repairs can be made quickly, thereby minimizing the



loss of data.   The installation of continuous measuring instrumentation



is cost effective provided the measurement program for which it will



be used is particularly long, as it will be for Cat-Ox (one year), so



that the initial cost of equipment purchase and installation is offset



by reduced manpower requirements during the test itself.  Furthermore,





                                 712

-------
MANUAL SAMPLING METHODS
-  PROVIDE BASIS FOR COMPARISON
-  SOLE METHOD FOR DETERMINING CERTAIN POLLUTANTS
-  METHOD FOR CALIBRATION
CONTINUOUS MONITORING SYSTEM
-  CONTINUOUS RECORD OF EMISSIONS
-  AUTOMATIC SAMPLING FROM TWELVE LOCATIONS IN PROCESS
-  AUTOMATIC CALIBRATION ON PERIODIC BASIS
-  IMMEDIATE DATA OUTPUT ON GAUGES, STRIP CHART, AND TELETYPE
-  DATA STORAGE ON MAGNETIC TAPE FOR COMPUTER PROCESSING
                        FIGURE 14

        CAT-OX DEMONSTRATION TEST - MEASUREMENT METHODS

-------
if one Is evaluating a complex process as is the case with regard to




Cat-Ox where simultaneous measurements are desirable from several




locations, continuous measurements provide a particular advantage




because of the manpower which would be required to man all the sites




if the data were to be obtained manually.  Continuous measurement




instrumentation also permits test conditions to be changed more quickly,




and permits observation of process performance during transitions from




one test condition to another.  Finally, once it has been decided to




utilize continuous recording instrumentation, there is a tremendous




advantage to storing the information on magnetic tape so that it can




be readily processed by computer.  The process of transferring data




from strip charts to punched cards is expensive in time and manpower




when large quantities of data are involved.






Cat-Ox Process and Measurement Points




     The continuous measurement instrumentation represents a major




part of the MITRE program so that the discussion which follows will




be devoted primarily to that area of the MITRE work.  The process




itself is discussed by W. Miller in a paper entitled "The Cat-Ox




Project at Illinois Power" included in these same proceedings and,




therefore, will be reviewed in this paper to the extent required to




explain the interfaces to MITRE Instrumentation.




     Figure 15 shows the flow diagram of the steam generator and




Cat-Ox process.  In this Figure, the dashed line separates the




steam generator from the process.  The Cat-Ox process, which is of






                                 714

-------
                                                                                                                                  •FLUE GAS
U1
ELECTROSTATIC
PRECIPITATOR
                                                  GAS HEAT
                                                  EXCHANGER






REHEAT
BURNER
SYSTEM "B"
(IN DUCT)
J






                                         BY-PASS TO STACK
                                                                               FIGURE 15
                                                                  STEAM GENERATOR & CAT-OX PROCESS

-------
 the reheat type,  has  been  installed between  the  existing I.D.  fans




 and stack.   The electrostatic precipitator is regarded as part of the




 process  because it was  installed with  the process, and is required by




 the process  to minimize contamination  of the catalyst in the converter.




      Flue  gas from the  boiler passes through the economizer and air




 heater,  and  then  enters the mechanical collectors where particulates




 are initially removed.  Particulates are further reduced to a  very low




 level by the electrostatic precipitator.  During the course of this




 flow, the gas temperature has dropped  approximately 400°F and  therefore




 must be  reheated  prior  to entering the converter.  The reheating is




 accomplished by burners "A" and "B" and the heat exchanger.  The SO.




 in  the flue gas is then converted to SO., and combined with water vapor




 in  the absorbing  tower  to formulate sulfuric acid.  The mist eliminate




 removes  sulfuric  acid mist which may escape from the absorbing tower




 with the flue gas.  Because of the pressure drop throughout the process,




 a second I.D. fan is required to'make up the losses and restore flow




 to  the stack.  Acid from the absorbing tower and mist eliminator are




 pumped,  cooled, and finally stored in a large tank from which  it can



 be  periodically removed.




     The numbers shown  in the flow diagram identify the locations at




which measurements will be made.  Points 1', 2f, and 14 are locations




where measurements were made during the baseline measurement program.




Measurements will be repeated at these three locations to permit
                                716

-------
correlation with the baseline results and to reevaluate improvements




which have been made in the operational efficiency of the mechanical




collector.  The other measurement points identified are pertinent to




the evaluation of the overall Cat-Ox process and major subsystems




of the process.






Purpose of Measurements




     Figure 16 shows the measurement points which have been identified,




the parameters which will be measured at each point, and the purposes




for making particular measurements.  The significant parameters which




will be measured by continuous recording instrumentation are gaseous




concentrations of SO-, HJ) vapor, 02, C02, NO , N02, and THC.  Also




differential pressure, static pressure, and temperature will be




recorded continuously to obtain gaseous volume flow, system pressure



drops, and gas temperatures.  In addition, manual measurements will




be made of gaseous S0~, H^SO, mist, and particulates.




     These measurements will be utilized to determine subsystem per-




formance of the power plant and Cat-Ox process as follows:




     •  Efficiency of the mechanical collector




     •  Efficiency of the electrostatic precipitator




     •  Pressure drops across process subsystems and overall system




     •  Efficiency of the heat exchanger




     •  Percent leakage of heat exchanger




     •  Efficiency of converter




     •  Gas distribution in converter






                                717

-------
        POINT          LOCATION

          1*       INPUT ECONOMIZER
       PARAMETERS
   , 0, C0, AP, P, T
       PURPOSE

CONFIRMATION BASELINE
TEST; NON-REHEAT PROCESS
CONDITIONS
                   AIR HEATER
MASS LOADING
EFF. MECHANICAL COLLECTOR
00
                   INPUT ESP
                   OUTPUT ESP
MASS LOADING ,AP, P, T
MASS LOADING ,AP, P, T
EFFICIENCY ESP; PRESSURE
DROP ESP/CAT-OX
EFFICIENCY ESP;PRESSURE
DROP ESP
         MANUAL MEASUREMENT
                                              FIGURE 16

                               MEASUREMENT POINTS ONE-YEAR DEMONSTRATION TEST

-------
         POINT

           4
    LOCATION
       PARAMETERS
                                                              PURPOSE
 INPUT HEAT  EXCHANGER     S02, H20, 0,,,  CO  ,AP, P,  T    EFFICIENCY/PERCENT LEAKAGE
                                                        HEAT EXCHANGER
                    OUTPUT GAS HEAT
                    EXCHANGER (INPUT
                    CONVERTER)
                          S02, S03  „ H20, 02, C02,AP,  EFFICIENCY/PERCENT LEAKAGE
                          P   T                          HEAT EXCHANGER; EFFICIENCY
                           '                            CONVERTER
to
          8
                    CONVERTER
OUTPUT CONVERTER
so3 , so2,
                                       , co
S02,
       3 , HO,

MASS LOADING , AP, P, T
                                                               , C02,
                                                                           GAS  DISTRIBUTION
                                                       EFFICIENCY CONVERTER VS.
                                                       FLOW RATE, T, AND TIME;
                                                       ASH ACCUMULATION; EFFICIENCY/
                                                       PERCENT LEAKAGE HEAT EXCHANGER
         MANUAL MEASUREMENT
                                           FIGURE 16 (CONTINUED)

                               MEASUREMENT POINTS ONE-YEAR DEMONSTRATION TEST

-------
        POINT
LOCATION
         10    INPUT ABSORBING TOWER
       PARAMETERS
                                           AP, P, T
       PURPOSE
                                                   EFFICIENCY/PERCENT LEAKAGE
                                                   HEAT EXCHANGER;  HjSO,  IN
                                                   ABSORBING TOWER
         11    OUTPUT MIST ELIMINATOR       H9S°4 MIST > Ap» p»  T
                                                   PERFORMANCE MIST ELIMINATOR
KJ
o
         13    INPUT STACK
         1A    MIDWAY IN STACK
                     P, T
S02,
                            S04 MIST , C02, 02,

                     N02, NO . THC,  H20, MASS
                     LOADING , AP, P,  T
                              PRESSURE DROP ACROSS CAT-OX
EMISSIONS
         MANUAL MEASUREMENT
                                           FIGURE  16 (CONCLUDED)

                              MEASUREMENT POINTS ONE-YEAR DEMONSTRATION TEST

-------
      •  Ash accumulation in converter




      •  Formation of H.SO, in absorbing tower




      •  Performance of mist eliminator




      •  Emissions from the stack




 The measurements performed at point 1' will be used to confirm the




 baseline test conditions and to obtain data at a location in the steam




 generator which is typical of that to  which an integral  type Cat-Ox



 process would be connected.






 Overall Continuous Measurement System




      Figure 17 shows  the overall continuous measurement  system.   The



 system consists of four  subsystems:  the  time-shared  gas measurement




 subsystem, the flow measurement  subsystem,  the continuous gas




 measurement subsystem, and the data  recording/control  subsystem.   The




 measurement points identified  previously  are shown  inputing  to each



 of  the  major subsystems.   There  are  additional locations  which have




 not been identified in this now diagram where temperature,  pressure,




 and humidity are measured  continuously as individual parameters.




 Some  of these  locations are discussed in subsequent text.




      Gas concentrations are measured on a time-shared basis  at seven




 different  locations by the  time-shared gas measurement subsystem.




 One of  these locations is the stack  (point 14) where a second gas




measurement  subsystem is dedicated to measuring gas concentrations




 at this  one point alone.  There is a common time slot when both



the time-shared subsystem and the dedicated subsystem both measure




gas concentrations in the stack simultaneously so that data from




both subsystems can be correlated.



                                721

-------
KJ
TO
TIME-SHARED GAS
MEASUREMENT
SUBSYSTEM

1




F
S


                                                  FLOW MEASUREMENT
                                                  SUBSYSTEM
                                                  DATA RECORDING &
                                                  CONTROL SUBSYSTEM
                                                                                                *
CONTINUOUS GAS
MEASUREMENT SUBSYSTEM
                                                     FIGURE 17
                                             OVERALL INSTRUMENTATION SYSTEM

-------
     With the exception of point 6 which permits access within the




converter, flow measurements are made at every point where gas con-




centration measurements are made so that mass flow rates can be computed.




In addition, there are several points where continuous flow measurements




are made without corresponding continuous gas measurements.  At these




points flow measurements are combined with manually measured gas con-




centrations or are used of themselves for subsystem evaluation.




     Data from the gas and flow measurement subsystems are automatically




recorded on strip charts, by printer, and on magnetic tape.  Strip




chart and printer data are utilized in real-time to assure that the




instrumentation is functioning properly.  Data recorded on magnetic




tape is used for subsequent computer processing eliminating the tedious




task of transferring data from strip charts to punched cards.  In




addition, the system is designed to perform certain automatic control




functions such as switching of the time-shared gas subsystem, gas




analyzer calibration, and blowback of sampling lines.






Time-Shared Gas Measurement Subsystem




     Figure 18 shows the flow diagram of the time-shared gas measurement




subsystem.  Flue gas is drawn into the analyzers through a filtered




probe which may be either an in-the-duct filter or an external heated




filter.  The gas from the probe is passed through a heated water trap




and then through a heated teflon gas line (Dekoron line).  The gas




lines are heated to prevent condensation of water vapor and hydrocarbons.
                                 723

-------
KJ
                          CALIBRATE £



HEATED
SAMPLE
HANDLING








I—fa?
: T W


. )
                                                               H20
                                                              VAPOR
                                                                        FIGURE 18
                                                            TIME-SHARED GAS MEASUREMENT SUBSYSTEM

-------
     Gas is sampled  from seven different  locations by  a multi-point



sequential sampler.  The gas is aspirated through the  seven  lines



 continuously  except  during blowback.   Each  line is  then  selected



sequentially by switching pneumatic valves drawing a fraction of the



gas into the analyzers.  The gas lines from the sequential sampler



to the combined S09-N0_/N0 analyzer and  the THC and water vapor analyzers
                  *•   £    X


are also heated for  the  same reasons given above.  The flue  gas to the



C02 and 02 analyzers is  first passed through a refrigerator-condenser



in order to remove the water vapor, thereby preventing water vapor



interference in the  CCL  analyzer and corrosion in the 02 analyzer.



     Operation of the time-shared gas measurement subsystem  is shown



in Figure 19.  The subsystem is sequenced automatically on a one hour



time base by the control subsystem.  Flue gas is drawn into  the ana-



lyzers from one particular  line for a period approximately 7 minutes,



and then that same line is  blown back by  high pressure air for approxi-



mately 1 minute to remove particulates from the ceramic filter.  Sub-



sequently, each of the remaining lines are sampled in succession in



the same manner until the sequence is completed.  Then all of the



analyzers are automatically zeroed against nitrogen (except  for the



SO--NO./NO  analyzer), and  then spanned against a calibration gas.
  w   fc   X


The SCL-N00/NO  analyzer provides the blowback air and is also designed
      £,   £»   3C


to zero on the blowback air which is passed through the sample cells



of the analyzer.
                                 725

-------
                            TIME PERIOD
POINT
1'

4

5

6

8

10

14

ANALYZERS

OPERATION
SAMPLE
BLOWBACK
SAMPLE
BLOWBACK
SAMPLE
BLOWBACK
SAMPLE
BLOWBACK
SAMPLE
BLOWBACK
SAMPLE
BLOWBACK
SAMPLE
BLOWBACK
ZERO
SPAN
(MINS . )
7
1
7
1
7
1
7
1
7
1
7
1
7
1
2
2
                               60
          FIGURE 19




OPERATION OF TIME-SHARED SUBSYSTEM
             726

-------
Continuous Gas Measurement Subsystem




     Figure 20 shows the flow diagram of the continuous gas measurement




subsystem which samples .gas continuously from the stack.  The principle



gas sampled is SO. so that a continuous record of SO. emission from




the stack is obtained.  For purposes of illustration, an external




heated filter is shown as opposed to an in-the-duct filter which was




shown in the flow diagram of the time-shared subsystem.  In addition,




a multi-point gas probe is shown which will be used at several of




sampling locations where gas concentration is expected to be non-uniform.




The advantage of the external filter is that the filter element can be




removed for cleaning without removal of the probe itself.  The continuous




gas measurement subsystem is synchronized with the time-shared gas




measurement subsystem so that both subsystems overlap in time while




sampling gas from the stack permitting correlation of data from both




systems.






Flow Measurement Subsystem




     The flow measurement subsystem is shown in Figure 21.  Gas flow




is determined by measuring differential pressure, static pressure, and




temperature which are combined in an analytical relationship to calcu-




late flow.  As the crossection of ducting is very large throughout the




steam generator and Cat-Ox process, it is necessary to measure these




parameters at a number of points within any particular duct in order




to obtain a representative measurement.  Therefore, the crossection of




the duct is divided into a number of sampling points based on the ASMA






                               727

-------
ro
CO
        MULTI-POINT GAS PROBE
              FLUE GAS
                                                                FIGURE 20

                                                      CONTINUOUS GAS MEASUREMENT  SUBSYSTEM

-------
tsj
IO
TEMP/PRES RAKE
IU iu iu iu
FLUE GAS

PRESSURE

L


PRESSURE
C= TRANSMITTER




^•^n
••—•to.
••*••

L^
— ^
— »•
— *.
»
»
— >•
>
— *•
— K
»
»
»
>
— *.
»
»
»
>
— »•






TEMPERATURE
TRANSMITTER


T
E
M
P
E
R
A
T
U
R
E


S
C
A
N
N
E
R

TEM
AMP







r


PERj
PEN
LIFI


, VOLUME
FLOW
CONVERTER


BAROMETRIC
PRESSURE
TRANSDUCER


ftTURE
ER


fc CHART
^ RECORDER
"|

. A-D CONVERTER




                                                        FIGURE  21
                                                 ROW MEASUREMENT  SUBSYSTEM

-------
power test codes, and an array of combined  temperature/pressure  rakes




are used to sense differential pressure, static pressure and  temperature




at the sampling points selected.




     The magnitudes of the differential pressure and static pressure




are measured by pressure transmitters and  are normally recorded




directly on strip charts and magnetic tape.  As temperature is measured




at many more locations in the system than  is required for flow measure-




ments, the temperature sensor, which is an iron-constantan thermocouple,




is inputed through a constant temperature  enclosure to a scanner.  The




scanner acts as a switch to connect thermocouples from various locations




to a temperature-compensated amplifier which amplifies, linearizes, and




temperature compensates the signal prior to entry into the A-D converter.




The constant temperature enclosure, which is not shown in the flow




diagram, maintains the connector junctions at constant temperature so




that emfs which are generated cancel out.




     In addition to recording the three parameters directly, a specialized




analog computer identified as the volume flow converter is used to calcu-




late the volume flow in real-time.   The volume flow converter can be




utilized at the output of any of the flow measurement instrumentation.



Gas volume flow is of particular interest because it is one of the




design parameters of the Cat-Ox process.




     There are nine flow measurement locations which are nearly identical.




However, at two of the locations, the economizer and stack, where dif-




ferential pressures are particularly low, it has been necessary to use




an electronic manometer in place of the differential pressure transmitter.





                                 730

-------
Data Recording and Control Subsystem




     The outputs of the analyzers, transmitters, and other sensors



are recorded on magnetic tape as shown in Figure 22.  The data



acquisition system proper has a capacity of fifty channels, which has



been expanded, as discussed previously, with an additional twenty



channels by means of a low noise temperature scanner.  Eight channels



are assigned to gas concentration measurements, 10 to static pressure,



9 to differential pressure, 1 to gas volume flow, 3 to channel identi-




fication and ambient measurements, and 14 to additional temperature



measurements.  The last 14 channels are tentatively assigned depending



on time available to integrate the necessary electronics.



     The scanner connects the analog signal from each channel in



sequence to the A-D converter which digitizes the analog signal.  The



data from the A-D converter is transmitted to the coupler which formats



it for recording on magnetic tape.  A printer is utilized for visual



recording of selected data.  In addition, a teletypewriter  (TTY) is



employed as an input/output device.  As an input device, the TTY




writes directly onto the magnetic tape via the keyboard.



     The scanner will normally be operated at 1 scan/minute but is



capable of being operated at better than 1 scan/5 seconds.  The



digital clock will generate time in days, hours, and minutes and will




provide reference signals to the function controller.  The function



controller will Initiate start—stop commands to perform remote control



functions such as sequencing of the time-shared gas measurement sub-



system, calibration of the analyzers, and blowback of the probes.






                                731

-------
                        GAS (8}
SJ
                               STATIC PRESSURE (10)
                                         DIFF. PRESSURE (9)
                                                  VOLUME FLOW
                                                  & SPARE <6)
                                                        I.D.&
                                                        AMBIENT (3)
                                                           TEMPERATURE (14)
                                            SCANNER
         INPUT FROM
         TEMPERATURE
         SCANNER
             CONTROL
             SIGNALS
                                                                           SYNC TO
                                                                           TEMPERATURE
                                                                           SCANNER
                                          FIGURE 22
                              DATA RECORDING AND CONTROL SUBSYSTEM

-------
Status of the Instrumentation



     Figures 23» 24, 25, and 26 list the equipment previously discussed



for each of the major subsystems and identify the manufacturer and



model number.  If the equipment is specialized to MITRE specifications,



model numbers have not been provided by the manufacturer.



     Installation of the continuous measurement instrumentation is



near completion and will be on line prior to. start of the one-year



demonstration test program.  Some of the hardware is shown in the




following photographs.  Figure 27 shows two types of gas probes.  The



probe on the left uses a heated water trap with an external heated



filter, whereas the probe on the right uses an in-the-duct filter



with an external heated water trap.  Figure 28 shows the interior



of the multi-point sequential gas sampler.  The gas from each of the



7 Dekoron lines is passed through teflon water traps shown in the



center of the photograph and then from the top of the traps through



pneumatic valves to the analyzer.  The center cabinet is heated to



prevent condensation of water vapor.  Figure 29 shows the interior




of the S92-H02/NOx analyzer.  The center cabinet is heated and




contains the N02/N0x gas cell.  The S02 cell is attached to the side



of the cabinet and is not in view.  Figure 30 shows the refrigerator-



condenser in the foreground with parallel gas pumps.  The rack in the



rear contains the 02 and C02 analyzers with sample handling and



flow balancing instrumentation.
                               733

-------
u>
                            FUNCTION
                    SEQUENTIAL SAMPLER
                    S02 - N02/N0x ANALYZER
                    THC ANALYZER
                       GENERATOR
                    H20 VAPOR ANALYZER
                    HEATED SAMPLE HANDLING
                    REFRIGERATOR-CONDENSER
                    SAMPLE HANDLING
                       ANALYZER
                    C02 ANALYZER
       EQUIPMENT




DUPONT





DUPONT 461C





BECKMAN 400





MILTON ROY ELHYGEN R 8320





MSA LIRA M202 (MODIFIED)





MSA





BENDIX





BENDIX





BECKMAN F-3





BENDIX UNOR-6
                                             FIGURE 23




                                  TIME-SHARED GAS MEASUREMENT SUBSYSTEM

-------
                             FUNCTION                             EQUIPMENT
                    S02 ANALYZER                                DUPONT  460






                    REFRIGERATOR-CONDENSER                      BENDIX


*j
LO


01                   SAMPLE HANDLING                             BENDIX
                       ANALYZER                                 BECKMAN F3







                                             FIGURE 24



                                  CONTINUOUS GAS MEASUREMENT SUBSYSTEM

-------
           FUNCTION




TEMP./PRES. RAKES




DIFFERENTIAL PRESSURE TRANSMITTERS




STATIC PRESSURE TRANSMITTERS




TEMPERATURE TRANSMITTER




VOLUME FLOW CONVERTER




TEMPERATURE SCANNER




TEMPERATURE COMPENSATED AMPLIFIER




BAROMETRIC PRESSURE TRANSDUCER




ELECTRONIC MANOMETER
         EQUIPMENT




UNITED SENSOR & CONTROL




LEEDS & NORTHRUP 1912




LEEDS & NORTHRUP 1912/1970




LEEDS & NORTHRUP 1992




LEEDS & NORTHRUP




MONITOR LABS 1100




IRCON DATA SYSTEMS 3J16F




ROSEMOUNT ENGINEERING 1331




CGS DATAMETRICS 1023
                                  FIGURE 25




                           FLOW MEASUREMENT  SUBSYSTEM

-------
-•J
CO
       FUNCTION


DATA ACQUISITION SYSTEM


FUNCTION CONTROLLER


STRIP CHART RECORDERS
      EQUIPMENT


DATA GRAPHICS  CAT-12


DATA GRAPHICS DGC-100


MFE M26 CAHA

L & N SPEEDOMAX M MARK II
                                           FIGURE 26
                              DATA RECORDING AND CONTROL SUBSYSTEM

-------
                 A
                   ..
                   I
                   I
-
   •  .
          FIGURE 27
GAS PROBES WITH EXTERNAL AND
    IN-THE-DUCT FILTERS

             738

-------
         F1:::
[-POINT  !         ..,        ' [P
           739

-------
       URE 29

SO -NO /NO  ANALYZI-K
  £-   *—    A
      7/10

-------

           FIGURE  30
TOR-CONDENSER,  O?  ANALYZER,  AND C02  ANALYZER
              741

-------
     Figure 31 shows a photograph of the temperature/pressure rake.




The support for the temperature/pressure sensors is aerodynamically




shaped to minimize turbulence effects at the pressure sensor.  This




particular rake shows two sets of sensors.  The longer of the two




sensors is the pressure sensor and is a conventional pitot-static




probe.  The temperature sensor is an iron-constantan thermocouple




encased in stainless steel protective tubing.  Figure 32 shows the rakes




mounted in six inch ports with manifolding to the pressure transmitters.




Figure 33 shows the differential pressure transmitter on the left and




the static pressure transmitter on the right with the zero/blocking




valve and drip pots below.




     Figure 34 shows the data acquisition subsystem and function




controller.  The TTY is on the extreme left.  The data acquisition




subsystem is in the rack next to the TTY.  The lowest panel of this




rack is the temperature compensated amplifier.  The next rack is the




function controller and the last rack shows strip chart recorders




and the volume flow converter mounted in the top panel.
                                 742

-------

       FIGURE 31




TEMPERATURE/PRESSURE RAKE

-------
-
                                                FIGURE 32




                                 6 INCH PORTS SHOWING MANIFOLDING OF RAKES

-------
                  FIGURE 33
') I PFF.RENTIAL AND STATIC PRESSURE TRANSMITTERS
                      745

-------
.
' -•-!$.-*
                                                                                 * m-**!*-**-^* -
                                                                                 iL  -

                                 DATA ACQUISITION  SI'BSY  -

-------
     DISPOSAL AND USE OF BYPRODUCTS
FROM FLUE GAS DESULFURIZATION PROCESSES
        INTRODUCTION AND OVERVIEW
                     by

         A.  V.  Slack and J. M. Potts
 Office  of Agricultural and  Chemical Development
          Tennessee Valley  Authority
            Muscle Shoals, Alabama
                       747

-------
  DISPOSAL AND USE OF BYPRODUCTS FROM FLUE  CAS  DESULFURIZATION  PROCESSES

                         INTRODUCTION AND OVERVIEW

                                    By

                       A.  V.  Slack  and  J.  M. Potts
              Office  of Agricultural and Chemical Development
                       Tennessee Valley Authority
                         Muscle Shoals, Alabama
           The removal  of  sulfur  oxides  from  flue gases poses several very
 difficult  problems  to  the owner  of  the  emitting plant--whether  the operation
 involved is  power production,  smelting,  sulfuric acid production or a Glaus
 plant.  The  overriding consideration  in many situations may be  the diffi-
 culty  in disposing  of  the sulfur once it has been gathered from the gas
 and  concentrated in some  solid or liquid material.  Whatever the product--
 a waste solid or a  salable material such as  sulfuric acid, elemental sulfur,
 or a fertilizer—there must be the assurance that it can be moved away
 from the plant as fast as it is  produced, except to the extent  that surge
 storage may  allow fluctuations.  Since  it is a byproduct rather than the
 primary one,  production cannot be planned in accordance with market demand.
 Instead, the byproduct flow must be disposed of in the best way possible,
 even though  it may  vary widely with production of the main product.

           For a new plant the problem is not as troublesome because byproduct
 disposal is  part of the overall  planning involved in determining project
 feasibility.   For an existing plant, however, finding a way to  dispose of
 a byproduct  that was never considered in the original project planning can
 be quite difficult  or  perhaps in some cases  impossible.

           One of the difficulties, both  from the country-wide viewpoint
 and  that of  the local  plant, is  the very large tonnage involved.  Various
 estimates  have been published on the amount  of sulfur oxide emission in
 the  United States;  a reasonable  order-of-magnitude figure for 1980 appears
 to be  50 million tons  of  sulfur  dioxide  from sources amenable to stack gas
 cleaning--and if it is  assumed that emission from the other sources will
 be reduced by fuel  cleaning, about 10 million tons would be added to the
 total  recoverable sulfur  dioxide.  Since complete removal is unlikely, a
 factor must  be applied; assuming 80% removal, the total is 48 million tons
 as S02--or 24  million  tons of sulfur.

           Obviously this  amount  of sulfur, or anything approaching it,
 would  produce  an almost overwhelming disposal problem no matter what the
 product.   If  converted  to  a waste solid  by CaO-CaC03 scrubbing  the resulting
 tonnage (wet basis) would be on  the order of JQQ million tons per year or,
more importantly,  the  sludge accumulation over the next 20 years would cover
 an area of almost 5000 square miles a foot deep.   Or if converted to salable
materials,  the amount of  sulfur  involved would be about 1.7 times the expected
 1980 consumption in  the United States.

                                    748

-------
          The local situation is even more difficult.   Assuming a modern
large plant such as TVA's Paradise station (2550 raw; about 4.2$ S in coal),
sludge production could be as high as 8500 tons per day and the disposal
volume about 1600 acre-feet per year (at 70$ capacity factor).   If the
sulfur were recovered in useful form the annual tonnage (as H2S04)  would
be on the order of 600,000 tons, which would generate such a sales problem
that for the particular location involved it might be necessary to ship
the acid so far that shipping cost would be higher than sales revenue.

          Thus the problem of byproduct disposal looms as a major obstacle
in the future of sulfur oxide emission control.'  In this symposium an effort
will be made to explore the problem from all angles and in depth.  One paper
will describe in detail the situation regarding disposal or use of the waste
solids from CaO-CaC03 scrubbing and others are concerned with the future
market for sulfur products, including new uses that are under development.
Finally,  one paper will describe the overall situation in Japan, where both
throwaway and recovery methods have been carried farther than perhaps any-
where else in the world.

          The present paper introduces the problem and presents an overview
of the current status of the technology, particularly from the viewpoint
of a utility such as TVA that is faced with the problem of process selection.
Most of the information presented has been abstracted from reports on research
work carried out by the TVA Office of Agricultural and Chemical Development
for EPA and/or the TVA Office of Power.  The work of other organizations is
also summarized to the extent that it has been made available to the public.

          Most of the producers who operate sulfur dioxide-emitting plants,
particularly the utilities, favor discarding a waste solid over recovering
a product for sale.  Ash disposal is an old technique to power producers
whereas sale of chemical products is not.  For this reason, the majority of
the full-scale projects now under way are of the throwaway type.  Full-scale
tests of recovery-type processes are also being carried out, however, mainly
in the United States and Japan.  It is not yet clear which type will be the
most economical.  The situation should be clarified somewhat in a major
process evaluation and cost study being funded by EPA; the study, to be
carried out by TVA, win compare those processes in the United States on
which full-scale design and operating cost is now available (or soon will
be).  Three recovery and two throwaway processes will be included.

          The ideal waste solid is one that is concentrated in sulfur content,
resistant to pile erosion, relatively dense (to reduce disposal volume), and
water insoluble.  Only three materials meet these qualifications well enough
to have received serious consideration—calcium sulfite, calcium sulfate, and
elemental sulfur.  Of these, sulfur has been considered only in a minor way
as a waste solid because it is expected that large quantities can be marketed.
There are areas, however, particularly in the Southwest, where the market
prospect is so bleak that the waste solid route is being considered.  Moreover,
if any major portion of the sulfur oxide emitted were recovered as a useful
product the resulting market depression could well push elemental sulfur  into
the waste solid category.
                                    749

-------
           In comparing  discard as a waste  solid with recovery as a useful
 product,  it  can be  said as a general observation  that recovery does not
 necessarily  avoid the solid waste problem  because a large part of the
 sulfur consumed ends up eventually as a waste, usually in the solid form.
 Over  half of the U.S. sulfur consumption is in the fertilizer industry,
 where the principal use is in making phosphoric acid.  In this process the
 sulfur is converted to  calcium sulfate, a  waste solid that is discarded
 in  settling  ponds.  Some sulfur, much less than in past years, is left in
 fertilizers  such as ordinary superphosphate and ammonium sulfate, and thus
 ends  up on the  farmer's fields as a waste  solid (since it is either already
 in  the form  of  calcium  sulfate or may become so in calcareous soils) except
 to  the extent that it supplies nutrient calcium and sulfur.

           Thus  recovering the sulfur in a  useful form does not eliminate
 the waste solid problem but merely diffuses it.  It is quite true, however,
 that  recovery reduces the overall quantity of waste solid since otherwise
 both  the  fertilizer industry and the power industry would produce a waste.
                Waste Sludge from Lime-Limestone Scrubbing


          Most of the sulfur dioxide removal systems operating or under
construction in the United States are based on CaO-CaC03 scrubbing with
disposal of the product sludge as a waste.  In planning these systems, dis-
posal of the sludge is a major consideration along with cost and reliability
of operation.  There are still several unsolved problems in these areas but
a major effort is under way aimed at solving them.  The utilities that have
pioneered in CaO-CaC03 scrubbing are quite active in this, as well as EPA
with its funding of research projects.  It is expected that many of the
needed answers will be forthcoming from the EPA-TVA-Bechtel test program at
TVA' s Shawnee station.
Sludge Tonnage

          The amount of sludge produced is governed mainly by the excess
lime or limestone used, the amount of fly ash collected from the gas in the
scrubber, and the degree of sludge dewatering.

          Since lime is more reactive than limestone, less excess is normally
needed to achieve the same degree of sulfur dioxide removal.  The actual
excess of each required for good operation, however, is not yet known with
any certainty.  In the Mitsui Aluminum CaO system in Japan (designed by Chemico),
the stoichiometry (mole ratio of CaO to S02 in entering streams) apparently
has ranged from 0.95 to 1-05 with 80-85$ removal.  In the TVA limestone slurry
pilot plant a stoichiometry of 1.5 has given about the same removal but there
is some indication (from the TVA pilot, the EPA pilot at Durham, and the

                                    750

-------
EPA-TVA operation at Shawnee) that the system might be operated at 1.2,  or
possibly lower, without much reduction in efficiency.   So much depends on
type of scrubber, gas velocity, liquor circulation rate,  and other factors
that no conclusions can be drawn at present regarding the minimum excess
of absorbent.

          The weights of dry solids obtained under various conditions are
shown in Figure 1.  The values given are approximate only since in actual
operation several factors can vary and change the tonnage to some degree.

          The increase in weight of solids resulting from fly ash inclusion
can vary over a wide range.  When large-scale application of CaO-CaC03
scrubbing first began in the United States a few years ago the emphasis
was on injecting limestone into the boiler, in which case all the fly ash
went with the lime.  This approach has now been well nigh abandoned, however,
in favor of lime or limestone introduction directly into the scrubber circuit.
For such a method of operation, the situation differs between existing plants
and those started after the EPA emission regulations came into effect.  Since
most existing plants have fairly efficient particulate removal systems already,
or will have before sulfur dioxide removal equipment is installed, the amount
of ash entering the scrubber would be relatively low.  (There are exceptions
to this, e.g., TVA's limestone scrubbing installation at Widows Creek, where
the existing precipitator removes only about 50$ of the ash.)  For new plants,
however, the regulations are such that in most cases scrubbers must be installed
initially.   In practically all cases so far, dust removal is being combined
with sulfur dioxide absorption in either one scrubber or two in series, on
the basis that the incremental cost for dust removal is lower than for installing
an electrostatic precipitator ahead of the scrubber.  There is some difference
of opinion as to whether or not this course actually is the most economical
but the consensus seems to favor it; much depends on the dust characteristics
and on the sulfur dioxide content of the inlet gas, each of which can favor
either, the precipitator or the scrubber.  Another consideration is that  some
utilities would like to bypass the scrubbers to maintain power production if
the scrubber system fails, and consider that this would be more acceptable
if the dust could be removed in an independent unit.  Some producers also
want to retain the option of marketing fly ash and therefore must provide
for dry collection.

          The type of boiler is also a factor, since the usual types—cyclone,
tangentially fired, and front-fired--introduce, in that order, increasing
proportions of the coal ash into the stack gas.

          Thus the tonnage of ash accompanying the calcium-based solids  can
vary all the way from an insignificant amount up to as much as 11 tons of
ash per ton of calcium solids (for coal containing 20$ ash and 0.6% s).

          The amount of water remaining in the sludge has an important effect
on total tonnage, and is significant mainly when the sludge is to be dewatered
and transported to the disposal area as a solid rather than as a slurry.  In
filtration tests at TVA, vacuum filtration gave a moisture content ranging
from about tfO to ^% for various pilot plant samples.  Centrifuging at
1000 x gravity gave about Zjjo.
                                   751

-------
  200
  150
o

_j
o
CO


a:
o

CD
100
 50
                               STOICHIOMETRY

                                    1.5

                                     4

                                    .3

                                    .2

                                    . I

                                    .0
1
1
1
2
1
3
1
4
1
5
                    % S IN COAL




                  FIGURE I

   EFFECT OF SULFUR  CONTENT OF  COAL ON

    AMOUNT OF WASTE  MATERIALS PRODUCED
                                           J
                     752

-------
          The degree of dewataring depends to a considerable extent on the
 ratio of calcium  aulfate  to calcium sulfite, since the normally larger and
more blocky nature of  the sulfate crystals makes them easier to dewater.
For example, in the Chiyoda process (water scrubbing to give sulfurous acid,
which is then oxidized to sulfuric acid and reacted with limestone to produce
calcium sulfate), centrifuging reduces the water content to 10-15$.

          Thua the wet tonnage of dewatered sludge is likely to be l.J to
 1.8 times the combined dry tonnage of fly ash and calcium solids.

          For the TVA Widows Creek installation.. (550 mw, 4-5$ S in coal),
which may or may not be typical, the dry tonnage of waste solids is expected
 to be about 2000  tons per day at full load and 500,000 tons per year at 70$
capacity factor.  Assuming 40$ moisture in filtered solids, this would amount
 to 5500 tons per day to transport away from the plant each day at full load
if a solids handling disposal system were used.  Added to this would be the
incoming limestone, about 1300 tons per day, or a total solids handling re-
quirement of 4600 tons per day for a 550-mw boiler.


Sludge Volume

          In waste solid  disposal systems of the type normally operated by
utilities and mining operations, the solids are slurried and sluiced to a
settling pond from which  the sluice water is either overflowed or recycled.
Since the pond must be abandoned when it becomes full of settled solids, a
key factor in planning is the ultimate pond volume required per ton of solids
 (dry basis).  This varies widely depending on the size, shape, density, and
gelling characteristics of the solid particles.

          The utility industry is fortunate in regard to fly ash disposal
because most ash types settle compactly, requiring only about 20 cubic feet
of pond volume per ton of ash (dry basis).  Probably one of the worst
situations is in the phosphate mining industry, where the clay-laden gangue
forms a gel structure in waste ponds and may occupy as much as 125 cubic
feet per ton.  The situation most analogous to lime-limestone scrubbing is
in the phosphoric acid industry, where byproduct calcium sulfate (gypsum)
is ponded in large quantities; the pond volume required is about 28 cubic
feet per ton.

          Unfortunately,  the pond volume requirement for CaO:CaC03 scrubber
sludge will likely be relatively high since the calcium sulfite tends to
crystallize in small, thin platelets that settle to a loose bulky structure
and occlude a relatively  large amount of water because of their gelling
tendency.   The crystal form is shown in Figure 2, which is from an electron
microscope study by McClellan and Mills (TVA).

          In other studies U), Davenport (TVA) noted three phases in the
settling of the slurry—(l) an induction period (slow settling while floes
were forming), (2) free settling, and (3) compression settling.  The average

                                    753

-------
Figure 2.  Single and stacked
calcium sulfite crystals shoving
the thickness of some of the
crystals.
Figure 5.  A rosette aggregate
of calcium sulfite crystals
formed by interpenetration
during crystal growth.
Figure 4.  Gypsum from Chiyoda
Chemical Engineering & Construction
Co., Ltd. (japan)

Figure 5.  A fly ash
particle with calcium sul-
fite on and next to it.

-------
settling rate over the first two phases was about 5 centimeters per hour.
Phase 3 was reached when the floes began to touch each other, at which
point a continuous gel was formed and settling proceeded at a much slower
and rapidly decreasing rate.  There was very little settling after about
48 hours, even over a period of several months.

          Measurements with a torsion wire gelometer showed that a rela-
tively strong gel was gradually formed.  A l6$> slurry had no gel strength
immediately after stirring but developed a strength of 6 g-cm after 50
minutes and >25 g-cm after 18 hours.

          Reported values for degree of compaction have varied over a fairly
wide range.   Many of the data on settling are reported on the basis of water
content in the settled sludge, whereas the value needed is weight of the
solid phase per unit volume of fully settled sludge; water content is a
factor in this but is not necessarily proportional.   From calculations and
estimates based on the various reports, the packing volume appears to range
between 45 and 75 cubic feet per ton of dry solids.   This covers a range of
roughly 30 to 60$ solids in the settled sludge.

          The factors affecting degree of compaction have not been identified
with any great degree of certainty and very little quantitative information
is available.   The following considerations are perhaps significant.

     1.   Hydraulic head.   The TVA work mentioned earlier indicated that
         increasing the height of the slurry column—from 13 centimeters
         to  100 centimeters--increased compaction by about 15$*  Thus a
         20-foot head;  say,  in an actual pond might be quite helpful.

     2.   Ash content.   Since the ash particles are relatively large in
         comparison with the sulfite,  sludges containing a large pro-
         portion of ash might settle more compactly.   Not much data
         appear to be available on the point,  however,  and the high gel
         strength would indicate that  a large proportion of ash would be
         needed for help  in compaction.

     3-   Degree of oxidation.   Since sulfate crystals are normally larger,
         levels of scrubber operating  parameters that increase oxidation
         might be advantageous.   However,  both EPA (2)  and TVA (%) data
         indicate that  a  high degree of oxidation is necessary if compaction
         is  to be increased very much.   Presumably the gel structure per-
         sists unless  the content  of sulfite crystals is reduced to a very
         low level,  and thus the situation is  analogous to the effect of
         fly ash content  as  described  in 2.

    4.   Agglomeration*   Agglomeration of sulfite crystals,  which presumably
         should help  compaction, has been noted in some tests; a typical
         agglomerate  is shown in Figure 3 /from the  work of McClellan and
         Mills (If)/.   However,  the scrubber conditions  that promote agglom-
         eration have not been identified.

                                    755

-------
      5.   Lime versus  limestone.   There has been  some  indication that
          lime scrubbing may produce  a more compact  settled sludge than
          does limestone scrubbing.   The Chemico-Mitsui Aluminum unit
          in Japan,  for example,  apparently produces a settled sludge
          containing 50 to 60$  solids whereas  the range for limestone
          scrubbing  has been more on  the order of 30 to 50$.  There are
          indications  that limestone  type and  grind may affect settling
          and filtration characteristics of slurry.

      6.   Double alkali operation.  In the "double alkali" type of
          operation—scrubbing with a clear water solution of an alkali
          followed by  reaction with lime or limestone  to precipitate
          sulfite and  sulfate—the calcium sulfite and sulfate are
          crystallized under somewhat different conditions as compared
          with slurry  scrubbing and thus might have different settling
          properties.  Exploratory studies indicate  that they may be
          somewhat improved.

          In the Chiyoda process  oxidation of  the sulfurous acid to sul-
          furic before reaction with limestone produces large crystals
          of gypsum  as the end product (Fig. 4)*  In Japan this is used
          in construction products but if the  process were used in the
          United States the  sulfate would probably be discarded, in which
          case the packing volume should be relatively low.

          From the  planning standpoint, therefore, a  relatively large pond
volume must be provided for CaO-CaC03 scrubber solids unless the situation
can be improved or  unless some process is used (Chiyoda, for example) that
gives low packing volume.    For Widows Creek,  the pond being provided contains
about 100 acres and it is expected that the ultimate depth of the stored
sludge.will be about  37 feet.  The initial storage capacity of the pond is
about 4.5 MM cubic  yards,  with the dikes designed so  that they can be increased
in height by 10 feet  to give an  increase in capacity  to 5.8 MM cubic yards.
The estimated total scrubber effluent ponding requirements over the remaining
life  of the plant (about 25 years) is 9-3 MM  cubic yards based on projected
load  factors  and current knowledge of settling characteristics.

          Pond volume requirement is an important factor in overall emission
control economics because of the high pond cost.   The cost per ton of solids
varies widely, of course,  with pond size,  topography,  depth of fill, pond
type, and many other  factors.  The Widows Creek pond,  which will accommodate
fly ash and sludge  in separate sections,  is expected to cost on the order of
$0.50-0.75  per cubic yard of capacity.


Methods for Increasing Compaction

          The discussion thus far has been concerned with the degree of com-
paction under normal  operation of the CaO-CaC03 scrubbing system.   There
are various  special measures that might be taken to get greater compaction.

                                    756

-------
           Slow Stirring;   Since  the  calcium sulfite crystals form a gel,
 breaking up the gel  occasionally by  slow  stirring allows further settling
 before the gel reforms.   In  TVA  tests  (l)  intermittent stirring for short
 periods was fairly effective.  A slurry that had settled to 3!$ solids was
 made to settle finally to 48$  solids by intermittent stirring (about 1
 minute stirring followed  by  8  hours  or so  of settling).  Continuous stirring
 was  not effective.

           These results indicate that  pond capacity might be increased by
 occasionally removing  the supernatant  layer completely and stirring the
 sludge by some method, perhaps by air  injection (which might help oxidize
 sulfite to sulfate)  or by towing across the pond a raft equipped with sus-
 pended stirring elements.  Thickening  from 50  to 50$ solids, for example,
 would increase storage capacity  by about
          Flocculation;  Addition of agents  to promote  flocculation should
 improve  settling rate but not necessarily  the degree of compaction; in
 earlier  TVA work on phosphate slimes (ore washing tailings) flocculation was
 not effective.  In tests of various flocculants added to limestone scrubber
 slurry (jj), settling rate was increased  to a major degree but final compaction
 was not  improved.  It should be noted that the normal,  unassisted settling
 rate  is  adequate for providing clear supernatant liquor (for recycle) in a
 large pond.

          Dewatering:  if the sludge is  dewatered by filtering or centri-
 fuging and transported to the disposal point as a solid, less disposal volume
 will  be  required than for settling in a  pond.  TVA filtration tests have
 given a  packing volume of 39 cubic feet  per  ton (59$ solids by filtration
 versus 38$ by settling).  EPA filtration tests on sludge made from another
 type  of  limestone gave a solids content  of about 70$.   Other exploratory
 tests (TVA) have shown that the "dry" solids do not expand to larger volume
 when  submerged in water or exposed to rainfall.

          Disposal as a solid rather than a  slurry will be discussed further
 in a  later section.

          Increase in Crystal Size;  It may be possible to increase the size
 of the sulfite crystals, or to promote agglomeration, by some technique
 applied  either in the scrubber or in a special vessel through which the
 slurry would flow on the way to the pond.  Apparently no work has been done
 in this  area.  It would be necessary, presumably, to increase the size to
 the extent that gel formation would not  occur.

          Oxidation;  If the sulfite could be oxidized  to sulfate at
 reasonable cost this might be the best answer to the problem.  Such oxi-
 dation is an established practice in Japan (see paper in this symposium
 by Jumpei Ando) where the objective is production of gypsum for sale.  The
method involves blowing air through the  slurry after it leaves the scrubber;
 the main problem is the slow rate of oxidation, which has dictated in many
 plants the use of special (and expensive) equipment to  increase the rate of
oxygen absorption.
                                    / J I

-------
          Extensive work has been done at TVA on oxidizing limestone
 scrubber slurry  (jj, 6).  The results can be summarized as follows:

     1.  Air introduction into the scrubber.  In one pilot plant
         test., air was drawn into the scrubber along with the stack
         gas.  Oxidation was increased to 90$ and both settling
         rate and degree of compaction were greatly improved.  Similar
         tests made since did not increase oxidation as much, and
         compaction was not improved.

     2.  Oxidation in separate unit.  A small grid-packed scrubber
         (6-inch diameter) in which air was blown through product
         slurry  (02:S03= mole ratio of 10) was built at the pilot
         plant.  Oxidation was quite slow and the tests were generally
         unsuccessful.

     3.  Spinning cup oxidizer.  A laboratory-scale "spinning cup"
         oxidizer such as used in some of the Japanese plants was con-
         structed and tested in the laboratory.  Approximately 80$> of
         the sulfite could be oxidized in about 1 hour if the pH were
         first reduced to about 5 by adding an acid.  (The reduced pH
         is apparently necessary in order to get an adequate supply
         of sulfite in solution.)  Nearly, complete oxidation was
         necessary to achieve any significant improvement in com-
         paction, presumably because this was required to avoid gel
         formation by residual sulfite.   The preferred acid was sul-
         furous (which perhaps could be obtained by passing part of
         the stack gas through the product slurry and then into the
         main scrubber).

     4.  Use of catalysts.   The oxidation could be speeded up, at a
         given pH,  by adding a small amount of catalyst to the slurry.
         Manganese and iron compounds such as the sulfate and carbonate
         were the most effective.  At a pH of ^.0 and with 0.1% of
         manganese oxide added, 100$ oxidation of a 1% calcium sulfite
         slurry was attained in 30 minutes.

          In general,  the work on oxidation to increase compaction has been
discouraging.  An analysis should be made to compare the cost of pond capacity
increase by oxidation as compared with building a larger pond or dewatering.
Pond Management

          The way a pond is managed over its lifetime has an important bearing
on costs and must be planned in advance.  Several questions should be con-
sidered.

      Will it be operated as a single unit or divided into sections?
     •
      Will the original depth  be  the  limit  or  can  the walls be built
      up by using  the  settled material?
                                    758

-------
       Shall  the  pond be partially  filled with water before operation
       begins?

       Can the  pond be  filled  to  the  top of the dike or must some
       freeboard  be allowed  for periods in which rainfall exceeds
       evaporation?

           The  usual method  of ash  pond management is to operate the pond
 (built by excavating and  throwing  up a low dike wall of earth) as a single
 unit with the  ash slurry  entering  at one end, flowing as a stream over
 settled ash  to a low point  at the  other end where a pool forms, and over-
 flowing supernatant water over a weir or through a standpipe.  This is a
 relatively simple method, feasible only because the ash settles rapidly
 and compactly.   Ash is seldom used to build up the pond wall because the
 spherical  form (Fig. 5) gives it a low angle of repose.

           In the phosphate  industry, waste gypsum is ponded somewhat
 differently.   The pond is often  divided into two or more sections by dikes
 and operated independently, one  filling while the other is drying and
 hardening.   Supernatant liquor is  drained from the one being filled by
 overflow into  a  standpipe that carries the liquor down through the settled
 solids and out to a collecting ditch extending around the pond (from which
 the liquor is  recycled to the plant).  When the active section is filled
 with solids  almost to  the top of the dike wall, the slurry flow is trans-
 ferred to  another section.  After  a  time the solids dry and harden to the
 extent that  excavating equipment can be used to build the wall up a few
 feet higher  with excavated  solids.   The cycle is then repeated, resulting
 in some  gypsum piles as much  as  100  feet high.

           Whether this system can  be used to reduce the acreage requirement
 for lime-limestone scrubber slurry ponds remains to be seen.  Small-scale
 tests  indicate that under some climatic conditions the sludge might harden
 adequately for such a  technique  and  that the angle of repose and stability
 of the wall  should be acceptable.

           The necessity for recycling may make it desirable to start with
 some water in  the pond to provide  recycle until the supernatant overflow
 can take over.   Otherwise the water  balance and pumping situation at the
 scrubber could be a problem.  This will affect the composition of any seepage,
 as discussed later.

           In most areas evaporation  and seepage from the pond should exceed
 the rainfall collected but  the local situation should be evaluated in planning.
 Seasonal variations  in the balance could make it necessary to keep some surge
 capacity for supernatant liquor  in the pond,  in which case the effective pond
 volume would be  reduced.

           In some situations  it may  be feasible to use a mined-out area as
 the pond,  as is  done in some phosphate plants in Florida.  The important
considerations are distance from the main plant and suitability of the area
for retaining  liquid.
                                     759

-------
          In other cases it may be possible to use existing waste ponds
 for  the calcium solids--for example, ash ponds in power plants, tailing
 ponds at smelters, and gypsum ponds at phosphoric acid plants.  The last
 of these would be especially appropriate because liquor-recycling facilities
 will already be available in most cases.  A complication, however, is that
 the calcium sulfite (produced by absorbing sulfur dioxide from the sulfuric
 acid plant tail gas) could decompose and give off sulfur dioxide if mixed
 with the acidic supernatant liquor in the phosphogypsum pond.
Landfill Disposal

          Since waste ponds are expensive and sometimes infeasible because
real estate is not available, the possibility of landfill disposal must be
considered.  The term "landfill" as used here includes both (l) dumping into
suitable excavated areas (mines, quarries) or natural depressions, and (2)
piling solids at some suitable point near the main plant.  All require de-
watering the sludge and getting the material in suitable physical form for
transport as a solid.  The discarded solids may or may not be covered with
earth later, depending on the situation.

          The attainment of suitable handling characteristics depends in the
first place on the degree of dewatering.  As noted earlier, about 25-40$
water is typical for filtration and centrifuging tests carried out so far.
With some sludges, there are indications that this degree of dewatering
would result in acceptable handling properties, but this no doubt depends
on factors such as proportion of fly ash, CaO versus CaC03, degree of oxi-
dation, crystal size, and type of limestone.  At any rate there is some
question at present whether simple dewatering will be adequate.  Some of the
utilities, Commonwealth Edison, for example \Jj], plan to add dry materials
to the sludge to improve its properties.  Dry fly ash is an obvious possi-
bility since it acts somewhat like a cement.

          The cost of dewatering, plus any treatment to improve handling
properties, is likely to be quite high.  Filtration rate has been fairly
good in TVA tests, typically 50-55 gal/hr/ft2.  Exploratory centrifuging
tests have indicated that lower moisture content can be obtained as com-
pared with filtering, and if satisfactory centrifuging techniques and rates
can be developed, this method of dewatering might be utilized if solids are
to be transported without reslurrying.  The optimum type of filter or centri-
fuge has not been established.

          For operation in which a relatively low solids content is carried
in the scrubber slurry, it may be economical to operate a thickener before
the final dewatering device.  Use of a thickener as the sole dewatering unit
would not seem indicated, however, since the relatively low degree of de-
watering would increase the amount of dry material necessary to give a
manageable solid.


                                    760

-------
          A further possibility is to "thicken" the water in the sludge.
Relatively small amounts of certain commercially available materials thicken
water to a solid or semisolid and thus could convert a soupy sludge to a
form more easily handled.

          Again, oxidation to sulfate should give a product with good
handling properties directly from the filter or centrifuge.  The Japanese
processes, both sulfurous acid oxidation (Chiyoda) and calcium sulfite
oxidation (Mitsubishi, Bahco, Ishikawajima-Harima Heavy Industries Company,
Ltd.) produce a relatively dry solid by centrifuging or filtering and very
likely could do so by thickener settling.

          Assuming production of a solid with good handling properties, the
next problem is planning the transport system--truck, rail, barge, belt,
overhead cableway, or other.  The cost of any of these systems will normally
be much higher than for sluicing to a waste pond, which is the reason that
phosphogypsum disposal seldom involves dry transport; in only two or three
of the world1s phosphoric acid plants, located in Europe, are the waste
solids transported in the solid state (by truck or conveyor belt).

          If the dewatered sludge can be piled satisfactorily, it may be
attractive to transport to a suitable area near the plant and there build
a mound something like the coal pile.  The saving in pond construction cost
and liquor recycle should be considerable.  For the relatively short distance
involved, it might be feasible to pump the dewatered sludge with a positive-
displacement pump.

          From the long-range standpoint, a good case probably can be made
for returning the sludge to the limestone quarry or the coal mine—thus
returning the nonusable constituents of the original raw material back to
the place from which they came.  There are obvious problems but the very
large tonnages may make such a course imperative.

          For both landfill and ponding, wind and water erosion of the
exposed sludge surfaces could become a problem.  If so, covering with earth
may be an acceptable solution but an expensive one.  Or established practices
for stabilizing the surfa * of waste piles may be applicable.

          One consideration ...• landfill is whether the dumped solids will
bear the weight of people or animals.  As noted earlier, ponded phospho-
gypsum will eventually bear the weight of even excavating equipment.  From
preliminary tests it appears that some CaO-CaC03 scrubbing solids may
eventually behave similarly.  It should be noted, however, that differing
process conditions produce waste solids of different characteristics.  Type
of limestone seems to have an effect, some producing waste solids much more
thixotropic than others.
                                    761

-------
Water Pollution

          Although calcium  sulfite and sulfate are relatively insoluble
compounds,  small amounts dissolve in the liquid phase of the scrubber
slurry; moreover, magnesium in  the limestone produces soluble salts and
soluble materials such as chlorides and nitrates are introduced with the
gas.   Also,  in  double alkali processes some of the alkali will usually be
left  in the sludge, dissolved in the liquid phase.

          The nature of the water pollution problem resulting from these
dissolved impurities depends mainly on whether the system is open or closed
loop  in regard  to water.  By closed loop is meant operation without any
purging or  "blowdown" of liquor other than that remaining in the discarded
solids.  In contrast, full  open loop is similar to current ash pond practice
in which all the sluice water is overflowed to a watercourse.

          In open-loop operation, fresh water must be added to the scrubber
loop  at a rate  sufficient to maintain the desired solids concentration in
the slurry  to the pond.  Since  this is usually 5-15$ (unless a thickener
is used), a large amount of water must be introduced--which, incidentally,
makes scrubber  operation much easier.  As a result, the chlorides and other
impurities  are  purged rapidly and the concentration of these constituents
remains at  a low level.  Sulfite and sulfate concentrations are not affected
as much because the solution is in contact with excess crystals all the time
and thus can become saturated.  The concentration is still below that for
closed-loop operation, however, because.eliminating the recycle from the
pond  prevents supersaturation from building up to the high levels encountered
in closed-loop  systems.

          Thus  open-loop operation produces a relatively low concentration
in the overflow, low enough to  meet many water pollution regulations.  A
typical liquid  phase analysis for open-loop operation, taken from runs in
the TVA pilot plant, is given in Table I.  For comparison, the overflowing
sluice water from a fly ash pond (TVA at Widows Creek) contains about
250 mg/1 of dissolved solids.

          The situation in  regard to water pollution regulations is quite
complex, making it quite difficult to predict the degree to which open-loop
operation can be tolerated.  In some cases the regulations are based on the
degree to which the impurity concentration in the receiving watercourse is
increased, which depends, of course, on the ratio of watercourse flow to
effluent flow,  the impurity concentration already in the watercourse, and
the amount  of pollutants introduced by the effluent.  Thus each situation
must  be evaluated separately in regard to open-loop acceptability.

          As an example of  the  complexity, the Illinois regulation applicable
to the Commonwealth Edison's limestone scrubbing system originally was such
as to  limit  the dissolved solids in the effluent to 750 ppm (7).  This was
changed recently to read that "total dissolved solids shall not be increased
more  than 750 mg/1 above background concentration unless caused by recycling
or other pollution abatement practices, and in no event shall exceed 3500 mg/1
at any time."   The latter is much more lenient and reportedly would allow
open-loop operation.

-------
                                 TABLE I

             Composition of Pond Liquor (Open-Loop Operation;


                  Constituent             Concentration, tng/1

             Calcium                             815
             Magnesium                           85
             Sulfate                             1450
             Sulfite                             70
             Chloride                            675
             Sodium plus potassium               59
             Iron                                0.17
             Barium                              0.15
             Cyanide                           
-------
situations and that closed-loop operation would unquestionably eliminate
any concern.  This is not necessarily true, however, because regulatory
attitudes differ so much that, although in some areas full pond outflow
is allowed, in others there is concern even about seepage from the pond.
The latter has become so much in question, in fact, that major research
programs are being mounted to determine the magnitude of the problem, if
any, and possible means for solving it.

          In closed-loop operation, all constituents build up in the liquid
phase until the purge afforded by the liquor remaining with the solids
(which must be replaced with fresh water) brings the concentration to a
steady-state level.  Thus the final liquor content of the fully settled
solids will determine the actual steady-state concentration, low residual
liquor in the solids giving high dissolved solids concentration in the
liquor.  If any water is placed in the pond before startup (because of the
pumping arrangement, as discussed earlier), this also will affect liquor
composition, making it less concentrated in the beginning of the pond
operation and more concentrated toward the end.

          Liquor composition will also vary widely with the composition of
both the limestone and the coal (or the oil, the ore, or other source of
sulfur dioxide).  The best situation is in scrubbing tail gas from a sul-
furic acid plant, for which the impurities come only from the limestone.
The coal and limestone compositions expected for TVA* s Widows Creek operation
are given in Table II.  Under closed-loop operation, it is expected that the
composition of the pond liquid phase will approximate that given in Table III.
Although the concentrations of some of the major constituents are higher than
for open-loop operation, the increase is not enough to cause a significant
effect on river water quality even if major seepage occurred.
                                TABLE II

               Expected Composition of Coal and Limestone

                    for TVA's Widows Creek Operation


                           Limestone, % by wt

                                                          Acid
       CaO   MgO  NaP0  KP0   CljA*  FegOa  _S _  COg   insoluble
       51.1  2.1  o.oi  0.17  0.66   o.4o   o.o4  4o.8    4.0


                            Coal, as fired

             _ _ % by wt
             Sulfur  Ash  Moisture    Heating valve, Btu/Ib

             4.5    25      5                10,000

                                    764

-------
                                TABLE III

            Composition of Pond Liquor (Closed-Loop Operation)


                 Constituent _         Concentre tion, mg/1

            Calcium                               830
            Magnesium                             230
            Sulfate                               11*00
            Sulfite                               1^5
            Chloride                              1200
            Sodium plus potassium                 50
            Iron                                  0.07
            Barium                                0.2
            Cyanide                             
-------
      5-   In some areas,  sandy soil conditions might make  an impervious
          pond difficult to construct.   (However,  a lining of clay or
          other impervious  material could  be used.)

           It should be noted that there are many  large waste ponds  over  the
 world that in some instances contain liquors much more impure than  that
 from lime-limestone scrubbing.   The liquor in phosphogypsum ponds,  for
 example,  has low pH and a  high concentration of fluoride.   Seepage  from
 such ponds does not appear to be a problem.

           For landfill the situation is somewhat  different  since the disposal
 area would not ordinarily  be prepared as  carefully as a pond.  A primary
 consideration would be protection against erosion of solids*   Dissolution
 in rain water,  however,  would not appear  to be  a  problem.   If it should
 prove to  be,  the methods described earlier for  improving  the handling proper-
 ties by a cementing or water thickening action  should also  reduce any leaching
 in landfill disposal.
             Waste Calcium  Sulfate  from Acid Neutralization


           There  are at  least  two  situations that might arise in which  sul-
 furic  acid made  in the  process of reducing sulfur dioxide emission could be
 justifiably neutralized with  limestone to make waste calcium sulfate.  One
 is  the situation in many  smelters,  where the stack gas is rich enough  in
 sulfur dioxide to make  acid production by the standard method the most
 appropriate way  for reducing  emission.  However, in some areas, particularly
 in  the western part of  the  United States, there is little or no market for
 acid beyond that already  being produced and therefore the acid would have
 to  be  neutralized.  The second is the situation in which a system has  been
 installed  for acid recovery and sale but the market fails or becomes inade-
 quate,  in  which  case the  acid must  either be stored or neutralized.  Since
 acid storage is  expensive (requiring 17.6 cubic feet per ton), neutralization
 would  appear to  be more economical.  Thus it may be desirable to install
 neutralization facilities even though sale of the acid is the principal method
 of  disposal.

           Work is under way on determining the best method of neutralization.
 The preferred method, if  feasible,  is to react strong acid (such as produced
 in  a standard plant) directly with  limestone to make a solid product,  thus
 avoiding the need for a solids separation step.  The problem is similar to
 that in making "ordinary  superphosphate" in the fertilizer industry, where
 phosphate  ore (mainly calcium phosphate) is reacted with strong sulfuric acid.
 The product slurry sets up  in a short time to a solid that can be handled
 and piled, but several  weeks  are  required before complete reaction of  the
 acid is attaiitpd.  In the fertilizer industry, the "curing" is done in a
 covered building.  For  the  sulfur dioxide removal situation, transfer  of the
 solid  to a landfill disposal  area could pose an "acid leach" problem until
reaction was completed.
                                     766

-------
           Another approach is  to  dilute  the acid, react with limestone,
 and separate  the  solid.   Complete reaction can be attained in the dilute
 system,  and the solids can be  washed, if desired, to remove occluded acid.
 Research is also  being carried out on this method.
                   Elemental  Sulfur as a Waste Material
          At  the moment  it seems unlikely that production of elemental sul-
 fur  as  a waste material  will become significant.  However, it is conceivable
 that, with  further  development, such a system would be more economical in
 some situations than production of calcium sulfite-sulfate or calcium sulfate
 as the  waste  product.

          The closest analogy would appear to be the current situation in
 western Canada, where large quantities of sulfur have been produced from
 sour natural  gas and stockpiled because of marketing problems.  The problems
 are  the cost  of reclaiming if the sulfur is ever marketed and also the
 emission of hydrogen sulfide from the pile.  Sulfur produced from sour natural
 gas  by  the  Glaus process may contain up to about 0.01% hydrogen sulfide which
 causes  an odor but  is not considered .very hazardous in outside storage;
 hydrogen sulfide evolution could present a problem in inside storage and the
 odor in outside storage  would be quite objectionable.

          There is  also  the possibility of wind and water erosion, plus
 oxidation that would produce acid constituents subject to leaching.  It
 appears that  these  problems have not been adequately evaluated for the
 situation in  which  sulfur is discarded permanently in the vicinity of plants
 in populated  areas.

          Assuming  that  these problems can be handled satisfactorily and
 that the economics  are acceptable, sulfur as a waste material has some
 attractive  features.  The volume requirement is only about 23 cubic feet
 per  ton, and  since  there are no diluting constituents the waste storage
 volume  required is  only  about 10% of that for lime-limestone scrubber sludge.
 The  potential (long tern) value of the material is also important; even if
 discarded,  it should remain available for use at relatively low reclaiming
 expense, if at some future time the demand were such as to make marketing
 attractive.
                   Use of Lime-Limestone Scrubber Sludge


          Although the sludge from lime-limestone scrubbing is a somewhat
unattractive material—wet, impure, and containing unreacted limestone--
there has been a considerable amount of research on finding a use for it.
Companies such as Combustion  Engineering,  G. and W.  H. Corson,  Dravo, and
others are involved in this, and EPA has funded work by West Virginia University
and by the Aerospace Corporation.
                                    767

-------
           The main approach has  been to  find bulk uses  for the material
 that require very little processing.  One  such use is as a soil amendment,
 analogous to use of waste gypsum from the  phosphate  Industry as "land
 plaster"  in the Far West.   The main value  for such use, however, is to
 improve the physical  properties  of  alkali  soils,  a problem that is of
 little significance in the areas where most of the sludge is likely to be
 produced.   Fanners would have little incentive to accept the material, and
 even if they did the  net result  would appear to be a relatively expensive
 type of landfill operation.  And the leaching problem would be magnified.

           Other efforts have been aimed  toward converting the sludge to a
 material  suitable for use as a highway base material.   It has been demon-
 strated that some quantity of sludge can be sold  for such use but it seems
 highly unlikely that  a market could be developed  for any significant pro-
 portion of the potential sludge  tonnage.   Special treatment would be
 necessary  since gypsum used as such has  been known to dissolve slowly and
 collapse under weight.

           Other uses,  such as for structural products, mineral wool pro-
 duction, and beneficiation to produce minerals, would appear to be almost
 hopeless.   The situation can be  compared with that of fly ash, which is an
 excellent  additive to  cement—used  in .large quantities  (10,000 tons in 19J2)
 by TVA in  power plant  foundations and in dam construction.  Notwithstanding
 an intense promotional effort, TVA  has been able  to move less than 1% of the
 annual fly ash production.   The  situation would thus appear quite difficult
 for moving scrubber sludge,  which is much  less useful and economical than
 fly  ash.
                       Gypsum for Construction Use


          The usefulness of lime-limestone scrubber product solids can be
improved if fly ash is kept out, unreacted calcium oxide or calcium carbonate
is eliminated, and calcium sulfite is oxidized to sulfate.  The result is
a fairly clean grade of gypsum, suitable for use in making wallboard or as
an additive to cement.

          The only use of this approach has been in Japan, where the general
use of oil as fuel eliminates the fly ash problem, acid is added to react
with any excess absorbent, and the sulfite is oxidized in special equipment
to sulfate.  One of the papers in this symposium describes this practice in
detail.

          Sulfite oxidation equipment is also being installed at the Mitsui
Aluminum plant in Japan, where coal is used as boiler fuel.  However, the
plant is equipped with very efficient electrostatic precipitators that keep
the ash content of the sludge at a minimum.

                                   768

-------
          It should be noted that even with oil firing there is some
residual soot that darkens the product gypsum.  Moreover, the cost of
oxidizing and of drying the product solid is not insignificant.  In the
United States, where there are large natural deposits of dry, relatively
pure gypsum, costs of oxidizing and drying scrubber sludge would appear
to be unacceptable.  Moreover, the dark color (in coal-burning plants,
from fly ash and other impurities)—as compared with the white color of
wallboard made of natural gypsum—could be a major obstacle to marketing.

          For the Chiyoda and Hitachi processes, the gypsum is made from
sulfuric acid and therefore the expensive oxidation step is not required.
This would improve the economics.
                              Sulfuric Acid
          Several of the proposed recovery methods give sulfuric acid as the
end product.  Catalytic oxidation (Monsanto) and some of the carbon processes
(Lurgi, Hitachi) are restricted to acid as the product, and the magnesia
methods (Chemico-Basic, Grillo, United Engineers), plus some of the carbon
processes (Reinluft, Bergbau-Forschung, Sumitomo), are better suited to acid
than to elemental sulfur.  Others, such as copper oxide (Shell, Esso-B & w),
sodium scrub (Wellman-Lord, lonics-S & w), and ammonia scrub (EPA-TVA),
usually give an essentially pure stream of sulfur dioxide that can be con-
verted either to acid or elemental sulfur as the situation dictates.

          Acid production methods can be said to be farther along than sulfur
processes, on the basis that acid production from sulfur dioxide is a well-
established technology whereas sulfur dioxide reduction to sulfur has been
operated commercially in only one plant.  The main problem with acid, other
than operability of the gathering processes for producing the sulfur dioxide
(or of the Monsanto and the carbon methods that produce acid in situ in the
main gas stream), is marketing of the product.

          One problem is that the acid from some processes (Monsanto, Hitachi,
Lurgi) is not commercial grade, either because of lower concentration than
normal or because of impurities.  There would be more difficulty in selling
such acid than for the standard 92-98$ water-white product.

          Even if the acid produced is high grade, however, the very large
tonnage, the high cost of storage and shipping, and the vagaries of the acid
market are major problems.

          This is particularly true in the Southwest and Mountain areas,
where the large sulfur dioxide emission from smelters and power plants, the
limited acid consumption in these areas, and the long distance to an adequate
market combine to make acid production for sale a very dubious proposition.

                                    769

-------
 In a study made by the Arthur  G. McKee Company  for EPA  (lo),  it was esti-
 mated that only about 60-65^ of the  sulfur  dioxide emitted  from western
 smelters  could be sold as  acid (5-0-5-5 million tons per year) and that
 this could be done only  if the acid  were priced at $4 per ton.  This leaves
 the remainder of the smelter acid  capacity  and  all of that  from the power
 plants  and other sources as  essentially unsalable.

           Probably the best  situation for acid  production is  that in which
 a  process  for using the acid is  operated contiguous to  the  power plant, thus
 avoiding  the  cost of shipping  and marketing the acid.  A phosphate fertilizer
 plant is  the  most likely prospect because most  of the sulfur  consumption is
 in the  fertilizer industry.  Phosphoric acid, triple superphosphate, and
 ammonium  phosphate are the logical end products.

           An  appropriate location for such  a joinder of processes is the
 upper Midwest,  where over  half of the phosphate fertilizer  produced in the
 United  States is  consumed  and  where many of the power plants  burning high-
 sulfur  coal are located.   It should  be noted, however,  that there are many
 drawbacks  to  such an Arrangement, the main  one  being that the sulfuric acid
 must be used  as it is produced (unless expensive surge  storage is installed)
 and thus  the  fertilizer facility would have to  be operated  even at times
 when otherwise  it would not  be economical to do  so

           The overall problem  of marketing  sulfuric acid is quite complicated,
 too  much  so for any full treatment here.   EPA is funding a market study at
 TVA  which  should  be useful in  evaluating the situation; the TVA power system
 will be used  as an example and an analysis made  of the potential quantity of
 acid, shipping  cost to various points, and  the  amount of acid that could be
 produced and marketed for various levels of netback (including zero and minus
 levels).   The  study is in  the  beginning stages.
                            Elemental Sulfur
          Sulfur has several advantages over acid as a product for marketing,
including lower cost of storage, higher concentration for shipping, better
marketing flexibility, and broader spectrum of use.  It seems likely, however,
that costs will be higher because of the need for a reducing agent.  Moreover,
the basic economics are questionable since the sulfur is in the oxidized form
in the stack gas and this is the form (as sulfuric acid) in which most of it
is used.   It does not seem reasonable to back and fill by oxidizing the sul-
fur compounds in the fuel during combustion, reducing the resulting sulfur
dioxide back to sulfur,  and then oxidizing it again to acid before use—unless,
of course, storages and shipping costs are overriding considerations.  It is
for these reasons that installing an acid-using facility at the power plant
seems economically desirable.

                                    770

-------
           The future of sulfur  versus  acid  as  a  recovery product is un-
 certain at this  time.   Much will  depend  on  the process economics as they
 finally develop--plus,  of course,  local  considerations that may favor one
 product over another.

           Some of the processes ^Westvaco (carbon), Institut Francaise du
 Petrole (ammonia scrub),  U.S. Bureau of  Mines  (sodium citrate scrubby
 produce sulfur only.   Others, as  noted above,  can convert  the gathered
 sulfur dioxide either to  sulfur or to  acid.
                          Fertilizer  Products


           Some  processes produce a  fertilizer material directly without
 sulfuric  acid or  sulfur being  involved as an intermediate.  An example is
 ammonium  sulfate, which is made in  Japan either by ammonia scrubbing  (with
 ammonium  sulfite  oxidation) or by the Mitsubishi manganese absorption process
 (reaction of manganese sulfate with ammonia to give ammonium sulfate).   Bufete
 Industrial, an  engineering and construction firm in Mexico, also has a process
 in which  ammonia  scrubbing is  followed by ammonium sulfite crystallization
 and oxidation of  the sulfite to sulfate in the dry state.

           There is some question as to the marketability of ammonium sulfate
 in large  quantities.  Even in  Japan, where ammonium sulfate has been a more
 important fertilizer than in the United States, the material is losing ground.
 The relatively  low nitrogen content, as compared with fertilizers such as
 ammonium  nitrate  and urea, is  a major handicap.  It is considered, however,
 that if ammonium  sulfate were made  in power plants located on the Mississippi
 River system, barge transportation  could be used effectively to reduce the
 adverse effect of low concentration on shipping cost.  By pricing the material
 below the present level, fairly large tonnages probably could be moved.

          One factor offsetting the low nitrogen content of ammonium sulfate
 is that sulfur is also a plant nutrient.  In the past the natural supply of
 sulfur compounds  in the soil, plus  that supplied incidentally in fertilizers,
made primary application unnecessary.   This situation is changing because of
 soil depletion and general use of fertilizers containing little sulfate.  Thus
 the sulfur in byproduct ammonium sulfate is likely to have increasing value
 as soils become further depleted.

          Other fertilizers,  including some nitrogen-phosphate combinations,
can be made in scrubbing processes  that do not involve intermediate production
and shipping of sulfuric acid.   These have been evaluated in a TVA study
carried out for EPA (ll).   It was concluded that a fair tonnage of such
products could be sold even in direct competition (same netback) with present
commercial fertilizers.


                                      771

-------
                    Future Market for Sulfur Products
          Whether sulfur products will be recovered and marketed in large
quantities depends on the cost of recovery processes versus the throwaway
type of operation (see earlier discussion of forthcoming EPA-TVA study).
This in turn depends, in part at least, on whether the recovered products
can be marketed and what netback they will provide.

          A major question in this is how well the demand for sulfur-based
products will keep up with the supply from byproduct sources.   This is a
complicated question, covered in a recent study for EPA by Esso Research
and Engineering and summarized by .a paper in this symposium.  Up until a
few years ago, the main production was from natural deposits and the supply
could therefore be adjusted more or less to comply with the demand.  With
the growing production from desulfurization of sour gas and now perhaps
from fuel and stack gas cleaning, the supply from byproduct sources is be-
ginning to outstrip the primary production.  As a result, the market situation
has become more complex, more difficult to analyze, and more discouraging to
those who need a reasonably stable market prospect if they are to invest
billions of dollars in stack gas cleaning installations.
                                    772

-------
                                References
 1.  Davenport, J. E.  Tennessee Valley Authority unpublished report.

 2.  Borgwardt, R. H.  Environmental Protection Agency unpublished report
     (November 1972).

 3.  Potts, J. M., Jordan, J. E., et al.  Tennessee Valley Authority
     unpublished reports.

 4.  McClellan, G. H., and Mills, M. E.  Journal of the Air Pollution Control
     Association 2£, No. 2, 122-2? (February 1975).

 5.  Nason, M. C.  Tennessee Valley Authority unpublished report.

 6.  Kelso, T. M. , Schultz, J. J. , et al.  Tennessee Valley Authority unpublished
     reports.

 7.  Gifford, D.  C.  Proceedings of Sulfur in Utility Fuels:  The Growing
     Pi lemna, a technical conference sponsored by Electrical World, pp. 2%5-87 (1972).

 8.  Tennessee Valley Authority.  Environmental Statement;  Experimental SOg
     Removal System and Waste Disposal Pond Widows Creek Steam Plant
     (January 15, 1973 J.    "

 9.  Goldschmidt, K.  Steinkohlen-Elektrizitat Aktiengesellschaf t (STEAG),
     private communication.

10.  Arthur G. McKee and Company.  Systems Study for Control of Emissions
     Primary Npnferrous Smelting Industry (June 1969).  Vol. I, Report No.
     PB lb^-88^;  Vol. II, Report No. PB 18^-885; Vol. Ill, Report No. PB 1&-886,
     Clearinghouse for Scientific and Technical Information, 5285 Port Royal
     Road, Springfield,  Virginia  22151.

11.  Tennessee Valley Authority.  Sulfur Oxide Removal from Power Plant Stack
     Gas:   Conceptual Design and Cost Study.   Ammonia Scrubbing Cl970jl
     Report No. PB 196 -Bo4, Clearinghouse for Scientific and Technical Information,
     5285  Port Royal Road, Springfield, Virginia  22151.
                                     773

-------
STUDY OF DISPOSAL AND UTILIZATION
 OF BY-PRODUCTS FROM THROW AW AY
    DESULFURIZATION PROCESSES
                by

             J. Rossoff
            R. C. Rossi
             J. Meltzer
     The Aerospace Corporation
       El Segundo,  California
                 775

-------
      STUDY OF DISPOSAL AND UTILIZATION OF BY-PRODUCTS
        FROM THROWAWAY DESULFURIZATION PROCESSES

                                 by

                             J. Ross off
                            R. C.  Rossi
                             J.  Meltzer
                     The Aerospace Corporation
                       El Segundo, California

                             ABSTRACT

       This paper summarizes an on-going Aerospace Corporation study
for the Environmental Protection  Agency concerning the ecologically sound
disposal of sludges produced by the  limestone scrubbing of coal-fired uility
boiler flue gases. Related utility disposal  problems concerning lime-coal
and limestone-oil sludges are also considered. In addition, the usage of
sludge in commercial products is briefly summarized.  A review of critical
factors which affect  disposal requirements  and techniques is given.   These
include items such as the chemical characteristics of process materials and
sludges,  sludge physical properties, water quality criteria, trace element
constituents, and liquor recirculation. Disposal methods described include
ponding and landfill-type operations.  The impact of input parameters on
disposal methods is  briefly discussed, as well as features of ponding and
landfill requirements.
                               776

-------
 1.      Introduction
        With the expanding use of coal by the electric power utilities, and
 the incorporation of lime or limestone scrubbers to desulfurize the flue
 gases,  the disposition of and resultant effects of the sludges produced by
 these processes has become a major concern.   The concern for the dis-
 position of the sludges is based on two principal factors:  (1) the sludges
 contain soluble salts and some compounds containing trace elements --
 all or part of which may leach or drain from ponds, landfills or other us-
 age areas to ground or surface waters and produce problems of water pol-
 lution,  and  (2) many lime/limestone/sulfur sludges are highly water reten-
 tive and are therefore not structurally acceptable in their natural form as
 landfill materials. At this time, little is known of the potential effects, if
 any, that may result from these sludges; however,  there has been consid-
 erable effort expended in the determination of technology to support the
 utilization of the sludges and to condition them for use as a landfill material.
 Since the sulfur scrubbing technology is just emerging from its pilot plant
 stage, representative power plant operational disposal data are not abundant.
 Additionally,  current water quality standards are not directly applicable in
 many cases, or are very difficult to apply to the leaching of sludges.  There-
 fore,  adequate data do not exist to  verify whether the sludges can or will be
 used or disposed of in any ecologically sound manner.
        The  EPA Control Systems Laboratory, Research Triangle Park,
 North Carolina, has designed a program to establish a valid understanding
 as to  whether the  sludges pose a water quality problem,  to assess solutions
 or approaches to solutions in the event the sludges  do pose a problem,  and
 to assess the  industrial  status of sludge usage and disposal in light of this
 analysis. As part of this program, a contract has been let to The Aerospace
 Corporation to perform  analyses and bench scale tests of lime stone-coal
 sludges to study the potential effects just mentioned,  and to assess the status
 of industrial usage and disposal technology.   Background for this  study is
fully described in an EPA position  paper {Ref. 1).
       The  Aerospace study is now in the early  stages of a program directed
toward an understanding of limestone  sludges from power plants burning
eastern coal and western coal.  That program* is the subject of this paper
 A potential expansion of the program to include additional samplings from
 other power  plants so that a broader data base can be obtained, and to in-
 clude a more detailed study of material conditioned for usage or disposal
 is being planned by the EPA (Ref.  1).
                                 777

-------
 wnich is a status report that essentially summarizes the program's objec-
 tives and progress to date.

 2.      Rationale
        Much has been written a'bout the projected abundance of sulfurous
 sludges and  the technical feasibility of their use as a constituent of commer-
 cial products or as a landfill material.  For example, in Reference 2,  Brackett
 presents data taken in a 1971 survey of the electric utilities  industry which
 estimates a  sludge plus bottom and fly ash production of 117 million tons in
 1976.  And,  in Reference 1, it is noted that a government interagency Sul-
 fur Oxide Control Technology  Panel has made a preliminary lime/lime stone
 sludge (50%  solids) production forecast of 86-112 million tons annually by
 1980.  That value is based on  75% of coal  burning power plants using a
 desulfurization  system, 75% of which will be lime or limestone  scrubbers
 (see Figure  1).
        Although these values are necessarily "soft," they highlight an
 acute by-product disposition problem.  It  is significant to note that the plants
 committed to lime or limestone scrubbers are  located in widely spaced
 areas  such as Pennsylvania and Nevada as well as in the North,  Mid-West and the
 South.  Thus it  is evident that  disposition  is not a matter of local concern, but
 definitely one of national interest.
        The concern for potential  toxicity problems  is emphasized by some trace
 element data presented by Aerospace in Reference 3.  These data, produced
 for the EPA  by the Shell Development Company and the Oak Ridge National Labor-
 atories, were taken from samples of ash modified by the dry injection of lime-
 stone into the boiler.  The elements of concern show little tendency to be con-
 centrated in  any size fractions, five of which ranged from an average diameter
 of 28. 7 //m to 1. 5 #m. In all  size fractions,  however,  heavy metals were re-
 ported at levels high enough to cause concern,  e.g.,  As: 50 to 200 ppm; Ba
 200 ppm; Pb: 200 to 500 ppm; V: 200 ppm.  If these elements are combined
 such that they are chemcially inert, there  would be little concern over their
 reaching ground or surface waters.  However,  application of wet lime/lime-
 stone scrubbing of flue gases from coal-fired boilers is just emerging and the
possibility of producing toxic water-soluble compounds has not been determined.
 The most often considered methods of sludge disposal are ponding ahd landfilling
                                778

-------
                          Figure 1
         COMPARISON OF ESTIMATED ELECTRIC
  UTILITY COAL GENERATION CAPACITY WITH CURRENT
AND FORECAST CONTROL BY LIME/LIMES TONE SCRUBBING
      CAPACITY, KNOWN CONTROL
            COMMITMENTS
      1972
1974
1975    1976
    YEAR
1977    1978    1979    1980
                 Courtesy of EPA Control
                  Systems Laboratory
                  (See  Reference  1)

                            779

-------
 and,  for commercial applications, as an additive in road base construction
 among other technically feasible products (Refs.  8 through 12).  Sludges
 used in these applications,  whether treated or not,  can possibly leach or
 drain toxic trace elements (as well as other contaminants) to ground or
 surface waters.  It is on this basis that a determination is considered nec-
 essary as to whether the sludges pose an ecological problem.

 3.     Aerospace Study Plan
       The plan on which the Aerospace study is  based was derived from
 information such as that previously mentioned (Ref. 1).  Supporting data re-
 garding the production of lime/limestone sludges are  given in an EPA paper
 "Control of Sulfur Oxide Pollution from Power Plants" (Ref.  4), an article,
 "Removing SO2 from Stack Gases, " (Ref.  5), and a previous  Aerospace
 study, "Technical and Economic Factors Associated with  Fly Ash  Utilization, "
 (Ref. 3).  From these and  other data, the EPA Control Systems Laboratory-
 defined a one-year program,  "Lime stone-Coal Sludge Characterization and
 Toxicity Study, " which The Aerospace Corporation is now conducting.  This
 study is currently limited to analyses of limestone-coal sludges from one
 plant burning eastern coal  and one  burning western coal; however,  surveys
 of disposal techniques and  utilization effects are not restricted to these two
 plants.  An EPA plan for the potential expansion of this effort to other plants
 including some which use lime as the absorbent material for eastern and
weatern coal-burning plants,  and limestone scrubbing at an oil-burning plant
 is given in Reference 1. This potential  expansion is designed to produce a
 reasonably broad data base such that adequate results can be obtained to apply
generally to lime/limestone scrubbers.
       Following is a listing of the program objectives of the current Aero-
space study and descriptions of each phase:
       a.     Chemical Characterization - chemical characterization  (includ-
             ing trace element identifications) and crystalline phase determina-
             tions are being made of sludge,  process  liquors, coal, limestone,
             make-up water,  and regular fly ash.  Special emphasis will be
             placed on the  chemical and physical  state of toxic  elements.
             These chemical analyses will determine  possible elemental
             losses that occur in the combustion or collection processes;
             they will  determine the possible toxic hazards associated with
             fresh materials that have been dewatered and with materials
             that have experienced subsequent treatment and they will identify
                                780

-------
      soluble components and sublimation components.   At this time,
      samples are being analyzed from the Prototype Turbulent Con-
      tact Absorber (TCA) scrubber at TVA Shavmee Unit  #10 in
      Paducah, Kentucky (Ref.  6) and the pilot TCA scrubber at the
      Mohave Station Unit #1 in Nevada (Ref. 7).
b.    Potential Toxicity Determination - A literature study is being
      conducted to determine any potentially toxic effects of the com-
      pounds identified in the chemical analyses.  This will  include a
      determination of the means of inhalation or ingestion in humans,
      and an assessment of the potential intake of toxic elements in
      plants and animals.
c.    Physical Properties - Physical property studies are being con-
      ducted on sludges that have experienced aging, i. e., drying,
      leaching,  exposure to  sunlight, to assess permeability,  com-
      pressive strength,  pozzolanic properties, and other factors
      which may have an effect on the use of the sludge as a landfill
      material.
d.    Detoxification - An assessment of the potential for detoxification
      of the sludge will be made.  This will include schemes such as
      oxidation, fixation, and heating.  Positive indications  will result
      in process definitions, estimated costs,  and the potential impact
      on handling  and disposal or usage.
e.    Disposal Methods - In light of any potentially toxic effects identi-
      fied, evaluations will be made as to the feasibility of various
      types of disposal such as ponding,  landfilling, and mine filling.
      Also, a limited bench scale program is being conducted tc define
      the properties that will affect the economics of transporting the
      sludge to disposal sites.   This includes corrosive properties,
      abrasion,  adhesion, thixotropic properties,  viscosities of sus-
      pension,  and bulk densities as a function of water content.
f.    Disposal Costs - An engineering study will be conducted to define
      the relative costs for the ecologically sound  disposal of the sludge
      as a function of variables such as transport type and distance,
      treatment processes,  receiving site preparation,  operations, and
      maintenance.
                         781

-------
        g.     Water Quality and Solid Waste Disposal Criteria Review - A
              review of federal and state solid waste disposal and water
              quality criteria, is being made.  The findings of this program
              will be incorporated and a reasonable interpretation will be
              made as to the impact of the disposal of sludges on the existing
              or proposed criteria.
        The potential EPA expansion of this program will include additional
 sampling sources representing a broader cross-section of disposal and
 treatment processes as well as other sorbent/fuel classes.  Particular em-
 phasis will be given to the assessment and evaluation of both on-going and
 developing treatment processes now being  employed by industry and govern-
 ment agencies.

 4.      Water Quality and Solid Waste Disposal Criteria
        In conforman.ee with the requirements of the Water Quality Act of 1965
 which amended the Federal Water  Pollution Control Act of 1965, all of the states.
 the District of Columbia, and the territories of Guam,  Puerto Rico and Virgin
 Islands established or are establishing water quality standards* for interstate
 (including coastal) waters.  In December 1970, the responsibility for adminis-
 tering the Water Quality Act of 1965 was transferred from the Secretary of the
 Interior to the Administrator of the EPA.   Most of the state  standards have
 now been written, and accepted by the EPA.  The state  standards are therefore
 the major sources of criteria by which the power plant  scrubber effluents  are
 to be judged at this time and they deal with the quality of the  receiving surface
 waters only.   Further, the state standards when citing  criteria  for domestic
 water supplies or for food processing,  generally repeat or refer to the Public
 Health Service Drinking Water Standards (Ref.  18) which apply to water dis-
tribution systems.
        Other legislation, e. g., the Federal Water Pollution Control  Act A-
 mendments of 1972 (PL92-500) which applies to both surface and ground waters,
 establishes a goal of zero  pollution discharge by 1985.   While calling for interim
 guidelines and standards to regulate pollution discharges,  it establishes the
 applicability of two definitions of particular interest to  future consideration  in
 sludge studies.   They are:
*
 Significant background information for the state standards can be found in
 "California Water Quality Criteria, " McKee and Wolfe (Ref.  16); "Federal
 Water Quality Criteria, " FWPCA, Dept.  of the Interior,  (Ref.  17); and
 "Public Health Drinking Water Standards," PHS  Publication 956 (Ref.  18).
                                 782

-------
       a.    "The term 'navigable waters' means the waters of the United
             States including the territorial seas. "
       b.    "It is the national policy that the discharge of toxic pollutants
             in toxic amounts be prohibited. "
       Summarizing  regulations,  we find the following:
       a.    Current standards are established by the states which specify
             water quality criteria for interstate (including coastal) waters.
       b.    Future  standards will regulate pollution discharges to all waters
             of the United States including ground waters.  And the discharge
             of toxic elements in non-toxic amounts may be allowed.
       The route by which potential pollutants can reach various water systems
is either by direct discharge or runoff,  or by leaching through the soil to
ground waters  or surface waters.  Because of the high concentration of dis-
solved solids (generally sulfates,  carbonates, sulfites and chlorides)regard-
less of other possible pollutants,  the process liquors generally cannot be
discharged without additional treatment and are therefore recycled.  Disposal
sites can be constructed to prevent drainage.  This leaves leaching as the only
other route (barring  accidents) to potential pollution.
       The effect of a leachate from a disposal pond or landfill on the quality
of a receiving body of water is generally a variable factor depending on local
weather, soil conditions, topography, the chemical and physical characteristics
of the leachate, and the flow and quality of the ground and surface receiving
waters.  Because this effect is so difficult to determine, regulatory  bodies
ordinarily  will prohibit the earth disposal of an untreated sludge if it contains
heavy metals considered toxic in concentrations exceeding the safe limits set
for drinking water.   This of course assumes that the toxic element(s) exists in
a. soluble compound within the sludge and that after leaching through the soil it
will appear in the receiving water in toxic  concentrations.  Such an approach
is undoubtedly  safe,  but the knowledge of when,  where and under  what condi-
tions these precautions are necessary is not known.
       It can be argued that the concern for toxicity in the  sludges may not be
critical since it will  be necessary to treat the material to cause a fixation
condition which will eliminate the formation of a bog.   This would permit the
landfill site to  be reclaimed for some useful purpose and it may effectively
encapsulate the undesirable constituents of the sludge.  This is no doubt true
in some  instances; however, a fixation  treatment isn't necessarily required
for all sludges and furthermore,  it is not known what the impact of the leaching
                                 783

-------
 of a conditioned sludge will be.  If it is considered harmful, then a problem is
 defined.  If it is not considered harmful,  then the conditions which are accept-
 able should be identified so that those who have both the problems of disposal
 and regulation can satisfy their requirements.
        Although the foregoing applies specifically to water quality, a few com-
 ments regarding solid waste disposal are in order.  The EPA Office of Solid
 Waste Management Programs is in the process of issuing guidelines for land
 disposal and thermal processing operations.  This is in response to  the Re-
 source Recovery Act of 1970 {PL91-513) which amended the Solid Waste Dis-
 posal Act of 1965 (PL89-272).  The presently proposed guidelines do not apply
 to hazardous  materials because of a lack  of sufficient information; however,
 hazardous wastes and sludges containing free moisture are considered special
 wastes which under certain circumstances may be accepted for disposal at a
 disposal  site but under the authority of the responsible agency.  Most state
 standards for solid waste disposal which may be applicable to sludges  are
 similar, i.  e., special permission is required of the responsible office, and
 finite criteria are not given.
        In light of the foregoing, the current Aerospace study effort has foc-
 used on determining the potential toxic hazard posed by the sludge as a con-
 sequence of the variations that arise in alternative disposal schemes and
 selected processing variables.  A potential toxic hazard from a pond or dis-
 posal site may exist in the  leachate, the run-off, a purge flow,  from vapors
 arising from a dry pond, or fine dust particles blown from a dry pond surface,
 or from anaeorobic attack from the bacteria in the containment soils.  Toxicity
 in any of these cases is defined by existing state and federal standards but is
 subject to the chemical considerations discussed in the following section.

 5,      Toxicity
        The determination of the toxicity of a sludge is dependent upon the
 factors relating to what kinds of elements it contains, how much are  present,
 and in what form they exist.  Many elements and substances that are acceptable
 or even essential to the body can be toxic  when consumed in large doses.  The
question of  what constitutes a large dose is often never well defined in physio-
logical terms but is instead Hie consequence of arbitarily defined factors.  In
a few cases* such as mercury,  arsenic, lead and cadmium,  where the metal
                             784

-------
serves no known biological function, a toxic dose and threshold has been
defined from physiological experience.  In most other cases,  the concentra-
tion threshold is less well definable and "safe" dosage levels tend to be  de-
fined by standards.  In this study an attempt will be made to correlate toxicity
with specific physiological effects when possible and also with existing state
or federal standards.
       An additional factor that determines the toxicity of a specific element
besides its concentration, is the chemical form in which it is fouund.  Even
the most toxic material can be non-toxic when  it is in a form that is unavail-
able to the body.  However, caution must be taken to ascertain that it does
not exist in a form that could be assimilated by plant or animal life and
thereby become available to humans.   Thus, a determination of a chemical
analysis, in itself, can not define toxicity but instead defines only concentra-^
tion.   In addition to chemical analyses, this study will determine the chemical
state of each potentially toxic element and its subsequent availability to humans.
       The study will start with a precise determination of the chemical
analysis of the sludge and its liquor,  and  the coal,  limestone and process
waters constituting the raw materials in the process.  A composite listing
of sampling points is given in Figure 2.  In Table I is an analysis of a com-
posite sample of western coal (Ref. 19) and an analysis of a specific sample
of an eastern coal from Aerospace  experimental data.  Only those elements
that could be potentially toxic were  selected.  Also in Table I is a  specific
analysis of fly ash and bottom ash from a western coal (Ref. 19).
             A comparison of the coal samples analyses reveals specific
differences.  Such differences can be found between coal samples within the
same region and do not necessarily indicate specific differences between
eastern and western coal.  In general, the western coals are  associated with
more basic minerals;  the eastern coals with more acidic minerals.  The
western coals tend to contain a broader spectrum of rare earth elements.
The trace elements are found usually as contaminants in minerals codeposited
with the  coal or in the organic matter from which the coals originated.  In
this regard, general statements referring to relative toxicity between the two
types is  not possible.
       The coal for which the fly ash and bottom ash were derived contained
approximately 10% ash.  As a consequence, the ash samples could be expected
                                 785

-------
                    Figure 2








 COMPOSITE SAMPLING OF TCA SCRUBBER SYSTEMS








1.   LIMESTONE






2.   COAL






3.   FLUE GAS PARTICULATES AT SCRUBBER INLET






4.   FLUE GAS PARTICULATES AT SCRUBBER OUTLET






5.   BOTTOM ASH






6.   SLUDGE AT SCRUBBER EXIT






7.   CLARIFIER UNDERFLOW






8.   RECYCLE LIQUOR






9.   CENTRIFUGE PRECIPITATE

-------
                         Table 1.   Selected Elements in Coals and Ash (ppm)
                        Eastern Coal
Western Coal
Composite
Element
Arsenic
Mercury
Antimony
Selenium
Cadmium
Zinc
Manganese
Boron
Barium
Beryllium
Nickel
Chromium
Lead
Vanadium
Aerospace
Data
N. D.
<0. 01
<0. 05
N.D.
N. D.
180
350
46
1800
<0. 01
N. D.
310
30
180
Sample
3
0. 05
0.17
1.6
<0. 5
0. 56
15
15
400
N. D.
25
5
4
9
Fly Ash
(Ref. 19)
15
0. 03
2.1
18
<0. 5
70
150
300
5000
3
70
150
30
150
Bottom Ash
fRef. 19)
3
<0. 01
0.26
1
<0. 5
25
150
70
1500
<2
15
70
20
70
N. D.  - Not detected

-------
 to have a concentration of a specific element 10 times that of the coal sample.
 (Since the western coal analysis was from a composite sample a direct com-
 parison between the coal data and ash data is not valid in this  case).  Compar-
 ing the fly ash and bottom ash reveals a. major difference  in heavy metal con-
 centrations.  Relative to the fly ash,  the bottom ash is low in  refractory metals
 and metals having high vapor pressures at combustion temperatures.  In con-
 trast, those base elements that easily slag,  are more predominant in the
 bottom ash.
        In Table 2 is an analysis of slew water that may be typically found in
 an ash  pond (Ref. 20) and an analysis of a scrubber liquor from Aerospace
 data.  The  analysis represents the solubility of some of the elements specifi-
 cally found in a boiler ash and a sludge.  Whereas only one of these elements
 (manganese) poses a health hazard at the given levels in those cases, they
 may not represent the actual case. In addition, factors existing in the  sludge
 processing technology may serve to concentrate other elements thereby posing
 additional hazards.
        As an example of the  concentration effect, the case of fixation technology
 will be  used. In general, this technology requires the  introduction of additives
 to the sludge which  react with the  dissolved sulfates to form a new  crystalline
 phase.  If toxic elements are accepted within this new phase, they become
 eliminated  from solution and no longer become available as a contaminant in
 leaching or run-off waters.   On the other hand,  if they are rejected by the
 new phase,  they could become concentrated within the remaining water
 and thereby pose a health hazard.   Details of the specific chemistry of each
 potentially toxic element have not  yet been determined  and the toxic hazard  of
 sludges in the various forms they  may be found has not yet been assessed.

 6.      Disposal
        Of the many million tons of sludge that will be produced each year,  the
 possibility of economic utilization  exists for only a small portion of the total.
 The major quantities of sludge produced will require disposal in an ecologically
 sound manner.  Presently,  the alternatives being considered are ponding and
landfilling; both disposal operations will require the establishment of procedures
that avoid hazards related to  health and safety or land use.
                                 788

-------
             Table 2.  Selected Elements in Solution {ppm)

Element
Lead
Antimony
Barium
Manganese
Mercury
Beryllium
Boron
Nickel
Cadmium
Selenium
Zinc
Arsenic

Ash Pond
(Ref. 20)
.01
.015
.07
.075
<.00l
.002
.5
.015
.01
.035
.03
.01

Scrubber Liquor
Aerospace Data
<. 01
N. D.
<. 05
1.6
N. D.
N. D.
11
.05
N. D.
N. D.
N. D.
N. D.
PHS
Drinking
Water
Standards
. 05
-
1.
.05
-
-
-
-
.01
.01
5.
.05
N. D.  - Not detected
                                  789

-------
        The requirements for disposal will be strongly dependent upon whether
 or not the sludges are found to be toxic.  If the  sludges are found not to be toxic,
 the problem of disposal is simplified,  but the problem of safety and land use
 must still be addressed.  An untreated sludge   will retain a considerable
 quantity of water.  Laboratory tests on a limestone sludge containing a rela-
 tively high sulfate content will  settle only to 45% solids content if under -
 drainage is not provided,  but if drainage to sub soil is allowed,  the  settled
 sludge will reach a content of 50% solids.  A maximum particle packing
 equivalent to a value of  about 70% solids content is reached upon air drying.
 When a dried sludge  is re-wetted, the sludge will bloat and decrease its bulk
 density  depending upon the quantity of water absorbed by the sludge.  When a
 sludge contains a high sulfite content,  the sludge will settle to a value of about
 35% solids whether underdrainage is provided or not.  A sulfite sludge does
 not dry as readily or settle as much during drying as a sulfate  sludge.
       A sulfate sludge  for which underdrainage is provided will produce
 savings  relative  to one with no such provision because the higher packing
 density allows for the disposal of about 20% more sludge in the same ponding
 volume.   However, if the  sub-soil is the only means of underdrainage, the
 possibility for soil plugging exists especially in the case of sulfate sludges
 containing high concentrations of dissolved solids.  Depending upon soil type,
 salts will precipitate within the soil, filling pores and preventing further
 water passage.   Thus, the advantage of underdrainage would be lost.  The
 more reliable system is one in which underdrainage is provided. Although
 in the case considered the sludge is assumed not to be toxic,  it is very likely
 that the leachate would not be of a quality to  be acceptable for discharge to a
 water course.
       Ponding  can be used not only for the final disposal of the sludge, but
 also as an interim measure to allow settling before removing the material to
 a landfill site.   The most fundamental  problem  in the ecologically sound dis-
 posal of  the sludge arises when the hazard exists for leaching toxic elements or
 soluble salts to ground or surface waters, or drainage to surface waters.
 The  immediate solution  to this  potential problem may require dewatering
and/or fixation of the sludge and  storing it in a clay lined pond  so that per-
meation  to the sub-soil is minimized.  Overdrainage and underdrainage
                                  790

-------
must return waters to the scrubber system thereby eliminating direct discharge
to surface waters.  Such a pond can serve as a final disposal site, but in most
cases because of the large volumes of sludge and water to be handled, the pond
can serve as primary settling basin from which the sludge can be periodically
removed and placed in a landfill.  A fixation treatment applied to the sludge
prior to placing it in the pond is an alternative technique  generally applicable
only if the sludge is to be removed to a landfill after curing.  This treatment
has a dual purpose, one, the removal of water and increasing compaction
qualities and, two, the minimization of leaching  by decreasing the perme-
ability of the material.
       If a sulfate sludge is found to be toxic,  it may be adequate to store it
untreated in a pond having an impervious base.   The use  of underdrainage and
overdrainage would be required to  prevent the formation of a bog and the
possibility of overflow, especially  in areas where  rainfall exceeds evaporation.
This, of course, requires the lining of a large area pond, monitoring, and
maintenance as  necessary.  A further problem would be  faced when a point  in
time is reached where the generating plant can no  longer accept all drainage
originating from rainfall from all its disposal ponds, and direct discharge to
streams would not be allowed,  or where the plant is to be abandoned.  Under
either  of those conditions, encapsulation by covering with an impervious
material may be the best solution.
       Numerous alternatives for  ponding with or  without pond lining or sludge
conditioning may be possible.  A cursory  survey of the industry has shown
that many of these alternatives are presently being considered and employed.
Alternatives,  such as treatment processes mentioned earlier,  produce a high
solids  content sludge material having acceptable structural qualities for a
landfill usage; a process for producing aggregate has been developed.  The
details of these  processes will be discussed by the various process develop-
ers. Future concern by Aerospace for processes  of this sort will be in the
assessment of alternative disposal systems regarding  potential ecological
problems.
       Some very rough approximations of costs for interim pond and fixation
disposal have been put at about $7  to $10 per ton of dry sludge, (Refs. 15, 21)
which includes labor,  trucking, additives,  supervision, and fixation process-
ing.  The cost of lining a pond varies  considerably.  For ponds in the 5 to 10
                                 791

-------
 acre size,  estimates have been put at $5, 000 to $20, 000 per acre for clay
 linings and stablized pozzolan base linings, respectively,  without piping -
 the clay lining will undoubtedly require maintenance to repair damage
 when the "dried" sludge is removed.  A soil covered plastic lining complete
 with all drainage can cost $25, 000 to $30, 000 per acre. At this stage of
 our study,  these values,  though only approximate, underline  the need for the
 determination of whether lined ponds with their attendant maintenance and
 monitoring (equipment) are necessary and where they should  be located.

 7.      Utilization
        The utilization  of sludges will be discussed by several speakers at
 this symposium,  therefore it is mentioned here only in terms of what the
 ecological implications may be and how that relates to the Aerospace study.
 Numerous programs directed toward sludge utilization have been conducted
 through government  sponsorship and by industry alone (see Table 3 ).   Some
 of the more significant of these are:  1) the successful development of the
 technology  for autoclaved calcium silicate products, mineral wool and soil
 amendments at the Coal Research Bureau,  West Virginia University under
 EPA contract (Refs.  10,  12); parking lot construction, Federal Highway
 Administration (Ref. 8) and I. U. C. S. (Ref. 11); landfill,  road bases,
 aggregate,  and structural shapes, I. U.  C. S. (Refs.  9,  11); characterization
 and multiple use investigations at Combustion Engineering; and landfill  and
 sludge properties by the Dravo Corporation (Ref.  13).
        A review of the many possible utilization schemes  indicates the  fol-
 lowing significant factors regarding their potential impact on ecologically
 sound disposal or utilization:
        1.    Applications to be considered  for potential pollution effects are:
             a.    Fixation treatments for landfill
             b.    Aggregate for either landfill  storage for future use, or
                  construction usage in  road building
             c.    Road base construction
             d.    Soil amendment
        The applications listed above are all subject to various degress  of
 leaching.  For item  l.a above, permeability values ranging from 2 x
  ^ C               *J
 10   cm/sec to  10~   cm/sec have been  reported.  In some industrial tests
 of the treatment of acid waste  sludges,  it has been claimed that the permeability
                       —5     —7
 has been educed from 10" to 10"  cm/sec  and the concentration of trace ele-
ments in the leachate was additionally reduced by about two orders of magnitude.
                                  792

-------
                                                     TABLE 3

                                      POTENTIAL UTILIZATION OF SLUDGE
     Technology

     Calcium-silicate Products; Mineral
     Wool; Soil Amendments
                                               Research Organization

                                               Coal Research Bureau,
                                               West Virginia University,
                                               Under EPA Contract
                                             Reference
                                             10, 12
     Parking Lot Construction
                                               Federal Highway Administration, Wash. D. C.   8
     Research Review
                                               Coal Research Bureau
                                               West Virginia University
                                             14
VD
Sanitary Structural Land Reclamation;
Compacted Road Base; Binder for
Stabilized Road Base; Aggregate;
Structural Shapes
I. U. Conversion Systems, Inc.
Plymouth Meeting,  Pa.
                                                                                                9,
     Utilization Survey
                                              The Aerospace Corporation
                                              El Segundo, California
    Sludge Properties
                                              Dravo Corp., Pittsubrgh,  Pa.
                                             13
    Characterization, Multiple Utilization
                                              Combustion Engineering, Windsor,  Conn.
    Lightweight Aggregate
                                              Michigan Institute of Technology

-------
Thus,  as a consequence of chemically combining trace elements into new
crystalline phases and the reduction in leaching ratio, soluble salts or toxic
contaminants that may be available to ground or surface waters are reduced
by about 10, 000 times over sludge that is not treated.
       2.    Aside from using the sludge as a landfill, the commercial appli-
cations listed in item  1 above constitute a potential outlet for the usage of vast
tonnages of power plant  sludges.  A survey of the potential fly ash market by
The Aerospace Corporation (Ref. 3} in 1970-71 indicated that the maximum
potential marketability of fly ash in the near term was approximately 25 per-
cent of the supply.  Major inhibitions identified were transportation economics
coupled with competition from other materials and the lack of control of fly
ash quality and supply.  A growing market for bottom ash in road construc-
tion brightens the total ash utilization picture somewhat (Ref. 2).  However,
with the flue gas scrubbing which increases the by-product tonnage by a factor
between two and three, a simple deduction identifies  "disposal" as the major
outlet for the sludge, particularly in the near term.
       A summary of potential utilization schemes for the sulfate sludges
(other  than sulfur or sulfur products), the  researchers,  and related liter-
ature is given in Table 3.  Sulfur and sulfur product recovery are considered
in other presentations at this  symposium.
                                 794

-------
8.     Summary
       The laboratory results presented herein represent only the initial
phase of experimentation.  Since these data were derived from the analyses
of samples taken from one plant on a given day, they should in no way be
considered necessarily representative or typical.  They do serve as a base
point, however, to which other experimental and analytical data will be
added so that a reasonable interpretation will be made of the potential im-
pact of the disposal of scrubber sludges on water quality and solid waste
disposal standards.
                                  795

-------
                         BIBLIOGRAPHY

 1.      "Waste Product from Throwaway Flue Gas Cleaning Processes -
        Ecologically Sound Treatment and Disposal, " J. W.  Jones,  R. D.
        Stern, Control Systems Laboratory, EPA, Research Triangle Park,
        North Carolina, for Presentation at Flue Gas Desulfurization Sym-
        posium,  New Orleans, La., May 14-17, 1973.

 2.      "Production and Utilization of Ash in The United States, " C.  E.
        Brackett, Southern Electric Generating Company,  Birmingham,
        Alabama, March 13-14,  1973.

 3.      "Final Report Technical and Economic Factors Associated with Fly
        Ash Utilization, " Aerospace Corporation Report No.  TOR-0059(6781)-1,
        El Segundo, Calif.,  July 26,  1971.

 4.      "Control of Sulfur Oxide Pollution from Power Plants, " Frank T.
        Princiotta, Environmental Protection Agency, March 14,  1972.

 5.      "Removing SO, from Stack Gases, " A. V.  Slack, Environmental
        Science & Technology, Vol. 7,  Number 2, February 1973.

 6.      "Test Program for the EPA Alkali Scrubbing Test Facility at the
       Shawnee  Power Plant, " M. Epstein, F. Princiotta, R. M.  Sherwin,
       L. Szeibert,  and I. A. Raben, Presented at the Second International
       Lime/Lime stone Wet Scrubbing Symposium,  New Orleans, La. ,
       November 8-12, 1971.

7.     "The  Mohave/Navajo Pilot Facility  for Sulfur Dioxide Removal, "
       J.  L.  Shapiro and W. L. Kuo, Presented at Second International
       Lime/Lime stone Wet Scrubbing Symposium,  New Orleans, La.,
       November 8-12, 1971.
                                796

-------
 8.      "Use of Waste Sulfate on Transpo '72 Parking Lot, " Russell H.
        Brink,  Federal Highway Administration,  Washington,  D. C., Third
        International Ash Utilization Svmoosium,  March 13-14, 1973,

 9.      "Structural Compositions Prepared from Inorganic Waste Products, "
                                                                 %
        L. John Minnick, G. & W.  H.  Corson, Inc.,  Plymouth Meeting, Pa.,
        Presented at the Annual Meeting of the American Association of
        State Highway Officials, Miami Beach, Fla., December 5-10, 1971.

10.      "Pilot Scale Up of Processes to Demonstrate Utilization of Pulverized
        Coal Fly Ash Modified by the Addition of Limestone-Dolomite Sulfur
        Dioxide Removal Additive, " Final Report Contract CPA 70-66,  Coal
        Research Bureau,  West Virginia University,  Morgantown, West
        Virginia, October 1971.

11.      "Multiple By-Product Utilization, " L.  John Minnick,  IU Conversion
        Systems, Inc., Plymouth Meeting, Pa.,  Presented at Third Interna-
        tional Ash Utilization Symposium, Pittsburgh, Pa., March 13-14, 1973.

12.      "Potential Utilization of Solid  Waste from Lime/Lime stone Wet-
        Scrubbing of Flue Gases, ".Linda Z. Condry, Richard B. Muter, and
        William F. Lawrence,  Prepared for presentation before the Second
        International  Lime/Limestone Wet-Scrubbing Symposiu,  New Orleans,
        La., November 8-12, 1971.

13.      "Properties of Power Plant Waste Sludges," Joseph G.  Selmeczi,
        and R. Gordon Knight,  Dravo  Corporation, Pittsburgh,  Pa. -
        March 13-14,  1973.

14.      "Review of Current Research  on Coal Ash in the United States, "
        John F. Slonaker,  and Joseph W.  Leonard, Coal Research Bureau,
        School of Mines, West Virginia  University, Morgantown,  West
        Virginia,  Presented at Third  International Ash Utilization Sym-
        posium (TIAUS), Pittsburgh,  Pa., March 13-14, 1973.
                                  797

-------
15.     "Will County Unit 1 Limestone Wet Scrubber, " D. C. Gifford,
        Commonwealth Edison Company, Chicago, Illinois,  Presented at
        Second International Lime/Limestone Wet Scrubbing Symposium,
        New Orleans, La.,  November 8-12, 1971.

16.     "Water Quality Criteria, " Second Edition, J. E.  McKee and H.  W.
        Wolf, The Resources Agency of California State Water Resources
        Control Board, 1963.

17.     "Report of the Committee on Water Quality Criteria, " Federal Water
        Pollution Control Administration,U. S.  Department of the Interior,
        April 1, 1968.

18.     "Public  Health Service Drinking Water Standards, " PHS Publication
        956, dated 1962.

19.     "Southwest Energy Study, " Report of the Coal Resources Work Group,
        February 1972.

20.     EPA Survey Data

21.     "Will County Unit 1  - Limestone Wet Scrubber, " D.  C.  Gifford,
        Presented at Electrical World Sulfur in Utility Fuels Conference,
        Chicago, Illinois, October 25-26,  1972.
                                 798

-------
     EXPERIENCE IN THE  DISPOSAL AND UTILIZATION
OF SLUDGE FROM LIME-LIMESTONE SCRUBBING PROCESSES
                           by

                W. C. Taylor, Supervisor
      Materials Management Research and Development
            Kreisinger Development Laboratory
                 C-E Combustion Division
              Combustion Engineering, Inc.
                    Windsor, Connecticut
                            799

-------
                         INTRODUCTION

      The  C-E waste  disposal program was  initiated  in  1969
with  the  primary objective of providing  ecologically  safe
methods for the disposal of solid and liquid waste materials
discharged from the C-E air pollution control system.  A
second objective was to investigate and  develop beneficiation
methods capable of producing useful products that would offset
the disposal cost.
      The  program was formulated after considering a number
of factors relating to the feasibility of beneficiation
processes and its effect on the smooth operation of the power
plant.  One of the major factors considered was the industry
experience in promoting the utilization  of fly ash.  Although
fuel  ash  was first used as an admixture  to concrete in 1936
with  favorable results, its utilization  in 1971 amounted to
only  about 201 of that produced.  This low utilization persists
despite the formation of a national organization with the
single objective of promoting the use of ash from utility
boilers.  This is in contrast to the relatively high utiliza-
tion  in other industrialized countries,  for example, Germany --
80%,  France -- 65%, United Kingdom -- 55%, etc.  Another
consideration was the desire of the utilities that any utiliza-
tion  or disposal scheme could not interfere with their primary
role  of producing power.  A third consideration was the require-
ment  that waste material from the APCS would have to be disposed
of within the ecological limits set by the Federal Environmental
Protection Agency as well as local control agencies.
                            800

-------
     Using the above factors as major criteria, a program
was formulated to develop ecologically safe and economically
sound disposal and/or utilization methods for APCS sludge.
The program was divided into five phases and are discussed
below.
              PHASE 1. THE ENVIRONMENTAL SURVEY

The Procedure
     The addition of a limestone wet scrubbing air pollution
control system to a coal-fired steam generating unit more
than doubles the amount of solids (modified ash) to be
disposed of by the utility.  In order to approximate this
effect on the ash handling and disposal capabilities presently
used and to gather other pertinent environmental data, a survey
of the present practices was made in September 1970.  The
survey also gathered available  data concerning pollution con-
trol limits of local regulatory agencies which were more
strict than Federal regulations.
     Forty-two utilities covering the continental U. S.
were reviewed and from these 22 were selected to be surveyed.
The companies were chosen on the basis of:
 a.  Geographic location
 b.  Population center
 c.  System capacity
 d.  Type of coal
     These companies account for about 301 of the electricity
generated by coal in continental U. S, A.  Questionnaires were
delivered to the utilities by C-E sales personnel at which time
                             801

-------
an appointment was made for future discussions between
C-E's engineers and the utility's environmental engineer.
During this latter  visit, the questionnaire was discussed
and filled out.
Data Analyses

General plant site data -- About 25% of the stations had
multiple means of fuel delivery.  The data showed that railroads
were used 67.51,,barge 39.6%, trucks 141, and conveyor 7% for
the transportation of coal.  Railroad delivery was not further
defined as to unit train or regular rail transportation.
     Furnace type data were reported on 215 units which represented
83.2% of the units surveyed.   Balanced draft units represented
71.7% and pressurized furnaces represented 10.7%.
     Firing method data were improperly stated on 29 of the
units.  The data indicated the following for the firing methods
reported:
     Tangential               40.5%
     Stoker                   15.8$
     Horizontal                8.41
     Vertical                  1.9*
     Cyclone                   4.71
     Improper response
        Pulverized coal       28.8%
Air data -- Air pollutant data (stack, wind, and emission
rates) were requested to permit determination of ground level
concentration.  However, insufficient emission data were
obtained to determine concentration by calculations.  It may
                             802

-------
have been more fruitful to request ground level concentrations



directly as well as emission rates. (One of the stations did



report ground level concentrations when emission rates were



requested.)  It is significant to note that only a very few



stations reported NO  emission rates.   This indicates that
                    J\


very few stations were determining NO  stack emission levels
                                     J\.


at the time of the survey.  (The regulations on NO  emissions
                                                  X


had not yet been promulgated and may explain this minimal



response.)



     The waste solids were transported from the dust collector



hopper to the disposal area by sluicing through pipes



by 74% or by trucking by 28%,  respectively, for the



stations reporting.  Approximately 24% of the stations were



planning future improvements in their equipment for control



of atmospheric emissions.



Land Waste Disposal Data    Plant waste solids disposal rates



varied from 80 to 440 tons per day.  The average solid waste



disposal rate of the 41 reporting plants was 800 tons per day



Fourteen plants reported sales of waste solids of from 1 to



741 of the ash produced.  Actual sales tonnage based on all



41 reporting plants was 1200 tons per day of 33,000 tons produced.



This represents the sale of less than 4% of the ash produced.



Maximum daily sales for any plant was 400 tons per day.



     Nearly 16% of the ash sales were made by stations using



plant acres for disposal.  Over 841 of the sales were made



by plants using off-site disposal areas.  Since plant site



solid waste disposal is used by 79% of the plants in the



survey it appeared that the economic pressure, cost of the  off



                             803

-------
 site disposal, may be a major factor in the need for developing
 sales outlets for solid wastes.
     Solid waste disposal costs were reported by 81% of the
 plants  in the survey.  The cost varied from $0.03 to $1.10
 per ton for sluicing compared to $0.11 to $1.33 per ton for
 trucking the solid waste to an off-site disposal area.  The
 average sluicing rate cost for 25 plants was $0.45 per ton
 while the average trucking rate for 10 plants was $0.51 per
 ton.
 Water Waste Data -- The boiler blowdown and demineralizer waste
 disposal data obtained were not amenable to specific statistical
 analysis.  This was also true for cooling tower and ash sluicing
 water data.
    Once-through condenser cooling water temperature rise data
 were obtained from 58% of the reporting power plants.  The
 average temperature rise was 16 F for cooling water flow rates
 varying from 0.5 to 7 million gallons per day.
     More than one-half of the utilities did not report the
 composition of the waste water streams from their stations
 This type of data had not been obtained, in most cases, at
 the time of the survey.
 Conclusions
     The electric power industry's response to the C-E environmental
 survey was very good with 86% of the canvassed companies com-
pleting the questionnaire.  The high rate of response is a
direct indication of utility interest and concern with environ-
mental problems.  As a result of the survey, a greater appreciation
for the complexity and diversity of problems associated with
                             804

-------
waste disposal was acquired by both C-E and the cooperating
utilities.

        PHASE 2, THE CHARACTERIZATION OF APCS SLUDGE

Methods of Study
     Since there was very little information available on
the physical, chemical, Land engineering properties of the
modified ash produced by the APCS, it was decided that this
should be one of the first areas to be investigated.  Hence,
the second phase of the program was designed to completely
characterize the APCS sludge.  The laboratory program initially
concentrated on the solids APCS discharge which was to be
followed by a study of the treatment methods for any waste
water stream associated with the system.
     Properties relating to the permeability, solubility,
compactibility, drying, etc. of sludge are important considerations
in the direct disposal of this material.  Thermal properties such
as   melting points, decomposition temperatures, sinterability,
etc. are important in many areas where utilization as building
products are considered.
     In order to supplement the classical analytical methods
and the X-ray fluorescence and diffraction spectrometers used
in the characterization of the sludge, a thermal analyzer was
obtained.   The equipment shown in Fig. 1 is capable of performing
simultaneously thermogravimetric analysis CTGD), differential
thermal analysis (DTA), and differential thermogravimetric
analysis (DTG).  Some of the engineering and rheological studies
                             805

-------
 that  require  specialized  equipment were  contracted  to  labora-
 tories  specializing  in  establishing  these properties.
 Analytical  Procedure
      Initially,  four sludges were selected  as  standards  for
 the characterization studies.   The  sludges were  selected  to
 represent the material  that would be produced  by  the different
 types of air pollution  control systems using the lime or  limestone
 scrubbing method.  This number was eventually  increased  to
 ten standard samples  when the variations in the scrubbing
 processes increased.  The chemical composition of these  ten
 standards,  identified in Table I, are listed in Table  II.
 X-ray analyses were  also performed on these ten standards  and
 the results are  listed  in Table  III.  Each of  the samples,
 which weighed about  50  Ibs., were dried to a constant weight
 at 100  F and stored  for future analysis.  Instructions were
 also given  to the contracting laboratories to  treat the  sludges
 that they received in the same manner so that  interlaboratory
 data pertaining  to a  particular  sludge could be compared.
 Table IV gives the range of the  specific gravity of the
 as they varied with  the water content.  A summary of the
 leaching studies using water at  two pH levels  is given in  Table V.
 Thermal Analysis
     In order to accurately characterize the changes that
 occur when  the sludges are heated to various temperatures,
 an extensive program  of thermal  analysis was set up.  The
 program started with  the determination of the  thermal proper-
 ties of some pure compounds found in the APCS  sludges.   This
was followed by a study of the effect of fly ash on these
                             806

-------
thermal properties.  The study was then extended to synthetic


mixtures of compounds formulated to simulate APCS sludge.


TGA curves   of the known materials were used to identify and


analyze these components in the actual sludges.  A summary


of the work on pure compounds and with fly ash added is


shown in Tables VI and VII and represented in Fig. 2, while


Table VIII compares the analytical results obtained with the


differential thermal analyzer to those found using wet chemistry


Figure 3 shows the rmogravime trie curve of some synthetic sludge



samples formulated in the lavatory, whiie Fig. 4 shows a differential


thermalanalysis curve of a representative sludge.





              PHASE 3, DIRECT DISPOSAL METHODS




     Based on the assumption that most utilities would


prefer the minimum modification of their present waste handling


practices, this program investigated methods of direct disposal


immediately after the characterization step had been completed.


Since direct disposal could take the form of either ponding


or recoverable land fill operations, C-E's study investigated


both of these.


Ponding Studies


     The ponding studies began by investigating the permeability


of settled and compacted sludge to determine its ability  to


act as its own sealant.  This was followed up by a  field  study


that is currently in progress.  Two  consultant hydrologists


were interviewed and one was selected to draw up a  test  program


to determine the effect of a sludge pond on the surrounding


ground water.                Oft_
                             807

-------
      After  preparation  of  the pond  site  and before the introduction
 of  the sludge,  a  series  of 14 wells were placed around the
 perimeter of  a  new APCS  installation.  Weekly samples were
 taken from  wells  for  two months prior to the introduction of
 sludge and  analyzed to  establish  the initial quality of the
 water.  Sampling  will continue for  at least one year and any
 change and  rate of change  in the  quality of the water will
 be  tabulated  and  analyzed.  A diagram of  the ponding site and
 wells is shown  in Fig.  5.   A description of the wells is con-
 tained in Table IX.
 Land  Fill Studies
      Laboratory studies have indicated that in order to
 produce a stable  land fill, an air pollution control system
 sludge would have to be dewatered to about 70% solids.  Studies
 have  also indicated that    conventional vacuum filters
 probably will not dewater  the sludge to that extent and other
 equipment or procedures will have to be used.  These other
 procedures may involve the mixing of the sludge with dry fly
 ash if enough of  it is available  to reduce the moisture con-
 tent  to the required level.  Other equipment under consideration
 includes a press  type filter, various drying equipment, and
mixing apparatus.   An extensive program involving laboratory
 characterization,  pilot plant testing, and full-scale field
 demonstration of processes  that would produce a solid suitable
 for land fill application  is planned.  Figure 6 shows one such
scheme.
                             808

-------
Road Construction
     In addition to the work with stable land fill material,
an extensive program dealing with the formulation and testing
of sulfate sludge for use in road construction is being carried
out.  During these studies, APCS sludges were dried and compacted
into test cylinders that have reached compressive strengths
of 5,000 psia after 28 days of curing.  Other sludges, however,
have exhibited compressive strengths of only 250 psi.
     Another phase of the study on the use of APCS sludge as
a highway construction material was C-E's participation in
a project at Dulles Airport in Washington D. C.  This project
is described in a later section.

               PHASE 4, BENEFICIATION METHODS
                   FOR SLUDGE UTILIZATION
Procedure
     Consideration was given to the possibility that in some
metropolitan areas a combination of high disposal costs and
a shortage of raw material for which APCS sludge could be sub-
stituted might lead to a situation where it would be economical
to develop some beneficiation process that would lead to the
development of products.  We believe that due to the volume
of the sludge involved, any by-product utilization schemes
would be similar to those with the normal fly ash.  Further,
it was believed advantageous to the disposal and/or utilization
of both the modified and unmodified ash if a systematic inves-
tigation was made of where the APCS sludge could be substituted
for fly ash and where the use of sludge would be detrimental
                              809

-------
 to  areas  where  fly  ash  is being  used.   In  this  study,  data
 collected by  other  organizations,  such  as  the Bureau of Mines
 and the West  Virginia Coal  Research  Bureau, were  considered.
 Table  X is  a  partial list of  the areas  where investigations
 in  varying  degrees  were made  for the utilization  of the APCS
 sludge.
 The  Most  Promising  Schemes
     A number of the more promising  utilization methods were
 investigated  either in our  Kreisinger Development Laboratory
 or  by  laboratories  already  working with fly ash in these areas.
 Figure 7  and  8  show some of  these processes.   A number of
 products  made from  the fly  ash involve  sintering.  Since
 laboratory  studies  have shown that with APCS sludge sintering
 is  accompanied  by the decomposition  of  the calcium sulfate
 and  subsequent  release of SC>2, a study  of  the economics of
 the  joint process involving the sintered product and sulfur
 recovery  was made by two sulfur-producing  companies.   One
 of  these  companies  developed processes  for the  production
 of  sulfur from  Gypsum.   Even while the  study was in progress,
 the  price of sulfur changed from about  $40 per  ton to  about
 $10  per ton making  any process depending on the sale of
 this material uneconomical  at this time.

              LARGE SCALE DEMONSTRATION STUDIES
     C-E participated in a project with the Research and
Development Division of the Federal Highway Department to study
the use of sludge as a highway construction material.  The
program was part of the U. S. Governments' International
Transportation Exposition (Transpo '72) that was held  from
                              810

-------
May 24 to June 4, 1972 at the Dulles Airport in Washington, D. C.
     In order to accommodate the 50,000 automobiles and 600
buses expected daily at the exposition, a 120-acre parking lot
was constructed.  The Research and Development Division of
the Federal Highway Administration used this opportunity
to demonstrate the recycling of waste products as highway
construction materials.  The major porti-pn of the parking
lot was paved with a mixture of fly ash and sulfate sludge
from an acid plant.  Other portions of the parking lot were
paved wtih waste products such as air pollution control system
sludge and acid mine drainage sludge.
     Combustion Engineering transported 75 tons of sludge from
Lawrence, Kansas to the Dulles site - a distance of about 1300
miles.  Since this was the first time that APCS sludge had
been transported by any vehicle, the project was designed to
obtain the maximum amount of information on possible materials
handling problems.  The sludge was dredged from a pond at
Kansas Power and Light where it had been stored for 6 months.
It was allowed to drain for 24 hours and loaded on flat and
round-bottom trucks for the non-stop trip.  Two C-E engineers
monitored the sludge during the entire trip taking measurements
and samples for laboratory analyses at various intervals.
     In general, there was no problem of sludge leaking from
the trucks while they were on the road.  Excess water did leak
from the track tailgate during the trip.  At Dulles, the
sludge was removed from the truck.
                             811

-------
      It was  considerably  easier  to unload  the  round bottom
 trailers  than  it was  those with  square beds.   All  of  the  sludge
 slid  out  readily from the former while the  latter  required
 manipulation with  the backhoe  to complete  removal  of  the  sludge.
      The  APCS  sludge  was  slurried with water,  mixed with  fly
 ash and lime,  and  placed  in  a  test section  of  the  parking lot.
 Evaluation of  the  suitability  of APCS sludge as a  highway
 construction and repair material is continuing.  In a recent
 report to the  third International ash association, the
 highway administration indicated that so far the performance
 of the material was satisfactory.  An analysis of  the densities
 of the sludges at  the bottom of  the truck  is given in Table XI.
 Figure 9 shows the loading, hauling, and unloading of the sludge
 material.
                       CURRENT STATUS

     At present, C-E  is applying its efforts to the area  of
 direct disposal.   Beneficiation methods and studies are limited
 to those that are necessary to dewater the slurry from 10%
 solids to the 70% for land fill operation.   To support this
 study, APCS sludge in 200-gallon lots has been shipped to
the laboratories of manufacturers of commercial drying equipment
for tests.  One spray drying test has proved to be very satis-
factory.
                               812

-------
     The more we get involved in the study of waste disposal
the more we learn the enormity of the problem and the effort
that must be applied before a solution is reached.  A portion
of the waste disposal dilemma is a result of our attempt to
supply the nation with energy without an adverse effect on our
environment.  C-E is not in the waste disposal business and
has nothing to gain from the sale of any products produced from
the sludge or from the various contracts for the sludge handling
that the utility may enter.  Utilities, suppliers, and the
federal government should share the expense of waste disposal
utilization studies.
                               813

-------
                            TABLE I


             IDENTIFICATION OF APCS SLUDGE STANDARDS


STD I    --  Flyash from Connecticut Light and Power Company's
             Devon Station

STD II   --  C-E sludge - CaC03, 1501 stoichiometry, 2000 ppm S02

STD III  --  Kansas Power and Light sludge

STD IV   --  C-E sludge - Ca(OH)2, 38% to 50% stoichiometry,
             50 to 60% SO, removal, slurry feed 220 gpm, recycle
             165 gpm with 55 gpm blowdown

STD V    --  Union Electric sludge

STD VI   --  C-E sludge - CaC03, 150% stoichiometry, 45 to 55%
             removal, no recycle
STD VIA  --  STD VI plus 50% STD I (flyash)

STD VII  --  C-E sludge - 300 to 325% stoichiometry, 64% S02
             removal, 300 Ib/hr flyash, 550 Ib/hr CaC03

STD VIII --  C-E sludge - 120 to 130% stoichiometry Ca(OH)2,
             90.8% removal, 120 gpm (Ca(OH)2 slurry underbed,
             inlet S02 860 to 840 ppm, outlet S02 80 ppm,
             145 Ib/hr Ca(OH)2, no flyash addition

STD IX   --  C-E sludge - 220 gpm H90 spray, 275 Ib/hr lime
                                   £*
             feed, 300 F reaction temperature
                                814

-------
                                      TABLE II






                      WET CHEMICAL ANALYSIS OF SLUDGE STANDARDS







STD I    STD II    STD III   STD IV   STD V   STD VI   STD VIA   STD VII    STD VIII    STD  IX
sio2
A12°3
Fe2°3
CaO
MgO
Na.O
K20
Ti°2
P2°5
C02
so2
so3
CaCO,
46.7
23.2
13.7
4.7
0.9
0.3
2.6
1.5
0.3
2.6

0.8
5.9
1.5
0.32
0.27
49.6
0.54
0.04
0.17
<0.02
O.OS
29.2
11.7
3.5
65.7
30.7
6.6
8.6
22.7
1.5
0.50
1.1
0.26
0.11
5.3
5.8
6.5
12.0
0.79
0.05
0.18
42.5
0.10
0.03
0.05
<0.02
0.06
3.7
38.8
3.3
8.4
19.4
6.8
5.4
27.6
3.2
0.08
0.24
0.32
0.08
7.2
2.2
12.3
16.3
1.1
0.01
0.09
52.5
0.52
0.02
0.14
<0.02
0.13
36.6
6.3
0.5
80.6
27.7
14.7
8.3
24.2
0.70
0.16
1.2
0.79
0.19
15.3
3.4
<0.1
34.7
4.6
2.3
1.6
40.1
0.20
0.05
0.29
0.11
0.08
13.6
S.4
24. P
30.9
1.2
0.48
0. 72
42.5
0.90
0.05
0.07
<0.02
0.06
11.5
24.1
8.4
26.1
2.0
0.45
0.72
46.2
0.40
0.04
0.21
<0.02
0.07
24.4
13.7
4.4
55.4

-------
                           TABLE III
                 X-RAY ANALYSES OF APCS SLUDGES
III
 IV
 IV
 VIA
VII
VIII
 IX
Major
sio2
3Al203-2Si02

sio2

2CaS03'H20
Si02
CaC03
CaC03
CaCOj
2CaS03'H20

2CaS03-H20
CaCOj
Minor Trace
Fe2°3 CaC03
Fe3°4
CaS04
Fe304 3A1203
CaC03 Ca(OH)
2CaS03'H20 CaS04
MgO
Si02
CaS04-2H20 Ca(OH)
Fe3°4
2CaS03 H20 CaS04
Si02
Si02 2CaS03
2A1203-2S102 CaS04
CaC03 2CaS03
Si02
CaC03 Ca(OH)
Si02




i
2

2
' 2Si(

'H2°
'H2°

2
2CaS03-H20
Si02
CaS04'2R20
Fe2°3
                                              Ca(OH)

-------
00
                                           TABLE IV



         THE VARIATION OF THE SPECIFIC GRAVITIES OF STANDARD SLUDGE WITH WATER CONTENT
STANDARD
H20
0
5
10
20
30
40
50
60
70
80
90
I
2.28
2.10
2.06
1.83
1.62
1.50
1.38
1.29
1.21
1.13
1.06
II
2.64
2.27
2.09
1.91
1.70
1.54
1.43
1.32
1.22
1.14
1.07
III
2.18
2.16
2.06
1.92
1.S8
1.47
1.40
1.28
1.21
1.12
1.06
IV
2.48
2.00
1.91
1.72
1.51
1.44
1.34
1.25
1.18
1.13
1.07
V
2.18
1.86
1.80
1.81
1.62
1.44
1.38
1.28
1.18
1.13
1.06
VI
2.62
2.41
2.27
2.00
1.78
1.57
1.43
1.32
1.22
1.14
1.06
VIA
2.33
2.21
2.09
1.92
1.66
1 54
1.43
1.30
1.21
1.12
1.07
VII
2.62
2.42
2.22
1.94
1.84
1.56
1.42
1.31
1.25
1.12
1.07
VIII
2.34
2.20
1.98
1.86
1.70
1.57
1.37
1.31
1.20
1.14
1.06

2
2
2
1
1
1
1
1
1
1
1
IX
.19
.19
.01
.84
.65
.48
.40
.30
.21
.13
.07

-------
                                                 TABLE V
                             * IEACHING PROPERTIES OF APCS SLUDGE STANDARDS
00
STD

I
II
III
IV
V
VI
VIA
VII
VIII
IX


Wt.
Loss,
g
0.8178
0.5515
3.2688
0.6599
4.6063
0.5258
0.5491
4.3096
1.9332
0.9464
*Air
pH
% Wt.
Loss
2.64
2.94
16.21
6.53
33.60
2.95
2.36
23. 77
11.91
6.55
5.90
Vol-
ume .
ml
6400
6250
5825
6810
8915
6870
6650
6890
8330
6655

Loss,
g/ml
0.127 x 10"3
0.088 x 10"3
0.561 x 10'3
0.097 x 10~3
0.516 x 10'3
0.077 x IO"3
0.083 x 10~3
0.625 x IO"3
0.232 x IO"3
0.142 x IO"3
dried uncompacted materials

Wt.
Loss,
g
0.2525
0.5821
1.8784
0.8218
3.1613
0.6936
0.3009
4.1662
1.3071
1.1391

PH
% Wt.
Loss
1.06
3.71
10.54
8.68
27.92
3.64
1.59
30.01
9.90
7.93

4.15
Volume ,
ml
6495
6750
6885
5175
6625
7295
5095
6315
7285
7155


Loss
g/ml
0.038 x
0.086 x
0.272 x
0.158 x
0.477 x
0.095 x
0.059 x
0.659 x
0.179 x
0.159 x


•
ID'3
ID'3
ID'3
ID'3
!0-3
io-3
io-3
io-3
io-3
io-3


-------
                                                 TABLE VI
                       THERMAL ANALYSIS OF SOME PURE COMOPUNDS FOUND IN APCS SLUDGE
CO
_4
10
Temperature and
 Events

Compounds
CaS04 •  2H20
CaS04 - 1/2H20
   CaSO,
                           0-500 C
                        Dehydrations,
                   Lit.

                   20.9

                    6.2
Exper.

18.7

 3.5
                           500-700 C
                      Oxidation of
                        ? Weight Gain
                                            Lit.
                                 Exper.
                                                                       700-900  C
                                                                  Decomposition of
                                                                  Limestone, % C0~
                                    1200-1400 C
                                 Decomposition of
                                 Calcium Sulfate,
                                     % SOT
                                                                   Lit.     Exper.    Lit.
                                                                                         46.5
                                                                                         55.2
                                                                   58.8
                                            Exper,
                                                                              46.7
                                                                              48.2
   CaS03 •  1/2H20      7.0
   CaCO,
            6.8
                                         13.3
5.0
                                                44.0
                                                                             43.5
                                                                                      62.0
                                                                                                 64.2
   Ca(OH)2
24.3
                              22.5

-------
                            TABLE VII
DECOMPOSITION TEMPERATURE OF COMPOUNDS WITH AND WITHOUT FLY ASH
COMPOUND
Without
Literature
CaS04 • 2H20 128
CaS04 • 1/2 H20 162
00
K>
o
CaS04 1200
CaSO, • 1/2 H,0 400
•j £
CaC03 825
Ca(OH)- 580
Decomposition Temperature, C
Flay Ash
Experimental
130
145



1400
400
730
460
With Fly
Experime
130
145



1200
400
730
460

-------
                                        TABLE VIII
             ANALYTICAL RESULTS OBTAINED BY WET  CHEMISTRY AND THERMAL ANALYSIS
              Total SO.
  Equivalent
Calcium Sulfate
CO,
Chemical
Standard I
III
CO
V
VI
VII
VIII
IX
0.
12.
13.
5.
32.
39.
24.
8
5
5
1
3
7
2
Thermal
1
14
14
6
34
29
22
.9
.5
.7
.5
.4
*
.3
.8
Chemical
1
21
22
8
55
65
41
.4
.1
.8
.6
.0
.7
.2
Thermal
2
27
27
10
58
*
49
38
.9
.0
.5
. 2
.4
.3
.7
Chemcial Thermal
2
4
4
34.6 34
13.5 11
8.8 6
23.8 21
.6
.1
.7
.9
.3
.2
.9
   Equivalent
Calcium Carbonate
Chemical  Thermal
            5.9

            9.1
                                                                        79.5
                                                                        30.8
                                                                        54.3
                                                  80.0

                                                  25.8

                                                  14.2

                                                  49.82
Did not reach maximum temperature during the thermal analysis

-------
                                 TABLE  IX






           SAMPLING WELLS AT THE FULL  SCALE  PONDING  STUDY  SITE
Well Number
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Depth*
(Feet)
28
32
29
34
23
29
27
31
27
30
28
32
28
32
Water Level*
(Feet)
23
23.
24.
24
20
20
23.
23
22
19.
23
23
25
24.

2
8



4


8



5
*  These values include 2 feet of pipe above the ground  level
                                   822

-------
                   TABLE X
UTILIZATION AREAS FOR FLY ASH AND MODIFIED ASH

   1.  Airport pavement mixture
   2.  Asphaltic filler and wear-course aggregate
   3.  Cement manufacture
   4.  Concrete admixture for construction
   5.  Grouting agent in wells
   6.  Filler for glass
   7.  Filler for fertilizer
   8.  Filler for paint
   9.  Filler for plastic
  10.  Filler for rubber
  11.  Fill material for land recovery
  12.  Fill material in abandoned mines for fire control
  13.  Fill material in abandoned mines for subsidance control
  14.  Neutralization of acid mine drainage
  15.  Manufacture of sinter bricks
  16.  Manufacture of autoclaved bricks
  17.  Manufacture of light weight aggregate
  18.  Manufacture of mineral aggregate
  19.  Production of cenosphere
  20.  Soil amendment
  21.  Soil stabilizer for road base, dam, embankments, etc.
  22.  Recovery of gypsum
  23.  Recovery of sulfur
  24.  Sealant in sanitary land fill operation
  25.  Sludge dewatering
  26.  Reclamation of polluted lakes
  27.  Sand blasting
  28.  Production of mineral wool
  29.  Recovery of magnesium oxide
  30.  Recovery of calcium oxide
                        823

-------
                                  TABLE XI
            DENSITY MEASUREMENT MADE ON THE SLUDGE IN TRANSPORT
       Miles
0
5-1/2 (Weigh Station)
11    (KPL-Weigh
       Station Round
       Trip)
      (Kansas Tpke
       Toll Station)
35

103

214
431
657
931
1375
      (Between K.C.
       § Mexico, Mo.)
      (Mexico, Mo.)
      (Effingham, 111.)
      (Richmond, Ind.)
      (Wheeling, W. Va.)
      (Dullas Airport)
                          Dike   Truck 1
                          51.5
                                  54.6
54.2
          I Solids [All readings for bottom
                    of bed except as noted]
          Truck 2        Truck 3
                                                           ,54.3  top
                                                           156.8  bottom
57.1
54.2
54.2
55.4
56.2
47.8
42.7
64.2
48.4 top
53.1 bottom
55.8
56.7
55.4


53.6
                         60.9
                                      824

-------
at
•
        ig- 1:   Simultaneous  analyses equipment

-------
00
to
150


120


110


100


 9(X


 8o


 70


 60


 50
                   Decomposition Of Calcium Sulfate In
                   Sludges Without Fly Ash

                   Decomposition Of CaS04 In Sludge With  Fly Ash
                   Phase Change In Calcium Sulfate
                                                        CaS04
                                                           CaC •*•  S02 +  1/2 02
  Decomposition Of Calcium Carbonate              CaCCs

  Phase Change In The Calcium Oxide (Endothermic)

\ Slight Wt. Gain Due To Oxidation Of CaSOa
                                                                            CaO 4
                J
            30C _
            20C .
            100
                    Phase Change In The Calcium Carbonate (Endothermic)
                   , ^composition Of Calcium Hydroxide              Ca(CH)s  —y CaC •+ H20
                    Decomposition Of Calcium Sulfite
                                                 CaS03 1/2H2C
                                                                                 1/2 H20
          Decomposition Of The Hemihydrate                CaS°4 1//2  Ha°
          Decomposition Of Gypsu-n                         CaSc4 2H2°—>
          Slight Wt.  Loss Miy Occur Due To Loss Of Surface Water
                                                                                            1/2 H20
                                                                                       1/2 H£0 + 1 1/2
                  j Slight Wt.  Gain May Occur  Due To Oxidation Of Some Components In The Sample


                                     THERMAL EVENTS TTJ A TYPICV. APCr i-MJDGE ?..Wi\£

      Fig.  2:   Summary of the work on  pure compounds with  fly ash  added

-------
oo
       10
    CO  20
    o
X
O  30
Ul
      40
      50
a
tr


	
>•
••••••••• »^


/b
\.
1




/c
..— .y
••••••^L«»t
i.
•
•
•


•••••••••«*




— — 33%CaC03l33%Ca(OH)2,33%CaSCV2-H2O
^— 50%FLYASH,IO%CaS042H20,30%CoS09,IO*CaCO,
••••••50%FLYASH,25%CaSO42H20,25%CaS{V2-H20
— 50%FLYASH, 30%CaS042H20,IO%CaS03,IO%CaC03











-.. ,d
— ,\
A
Vs
*
K
•
•
•
•
•





\-r
s^
\






•*-**>
X
^
	


                 200
                      400
600      800     1000
  TEMPERATURE-°C
1200
1400
160O
         Fig.  3:   Thermogravimetric curve of some synthetic sludge samples

-------
00
NJ
oo
               2OO
400
600     800     IOOO    I2OO
     TEMPERATURE-°C
1400
1600
1800
         Fig. 4:  Differential  th^rmalanalysis curve  of  representative sludge

-------
      EVEN NUMBERED WELLS ARE DEEP, ODD NUMBERED WELLS ARE SHALLOW
      DISTANCES ARE APPROXIMATE IN FEET
CO
KJ
                350'
                                      8 -*•
                                                   NORTH
                                   POWER PLANT
    Fig  5:  Ponding site and wells diagram

-------
                                          * /ROTARYX
  ARCS

SCRUBBER
                               ROTARY DRYER
          STORAGE

           HOPPER
                    TRANS PORT
                    SYSTEM
                                                             LAND FILL AREA
                                       ROAD BASE

  Fig. 6:  Scheme for producing suitable land fill

-------
CO
u>





WATER TO
* CLARIFIER™^
|
WET 1
m-£±-*\ VACUUM FILTER
SLUDGE ]















-»












•*

+



-»

^^
^^^






nci i CT17PD
rcLLcTlZcK

CEMENT K ILN



DRYING KILN

UIVCB
IWIAtfl







IMI^H

•»sc



—









_J
••w

>2^



•*






AIITOn AVF


SINTERING
MACHINE
+S02
S02
PROCESSING
FACILITY
SOzf
ELECTRIC
FURNACE

KILN










—



"""*

— *•




MINERAL
AGGREGATE

LIGHT
WEIGHT
AGGREGATE

SULFURIC ACID
OR SULFUR



MINERAL WOOL

ADDITIVE FOR
ASPHALT

1 AMD Pll L

             Fig.  7:  Sludge disposal schemes for a single scrubber system

-------
CD
U>
to
                   CALCIUM CARBONATE
             S02

           REMOVAL
^    CALCIUM
 W CALCIUM CA

 L     >•
 •^WATERHH
                          ICOO BACK TO SCRUBBER
^^•^
^^^^

1
1



•^

SINTERING
MACHINE

POZZOLAN
FOR
CONCRETE
t

A
C02
Kl
700 °C
1292 °F







S02
PROCESSING
FACILITY
s|2
.N
1200 °C
2192 °F
1
LIGHT
U/PI ftUT
wc.1 \yr\ \
AGGREGATE

MINE
LANDFILL DRAINAGE
NEUTRALIZATION
t t





           Fig. 8:    Sludge disposa^. schemes for a dual  scrubber system

-------
                         •
                   (a)
                      Cb)
                                                           (d)
                                                          Ce)
Fig. 9:  Sludge material handling
                                                   833


-------
 FIXATION AND DISPOSAL OF FLUE GAS
           WASTE PRODUCTS:
TECHNICAL AND ECONOMIC ASSESSMENT
                  by

L. John Minnick, Executive Vice President
      IU Conversion Systems, Inc.
           1500 Walnut Street
       Philadelphia, Pennsylvania
                  835

-------
                  FIXATION AND DISPOSAL OF FLUE GAS
                           WASTE PRODUCTS -
                  TECHNICAL AND ECONOMIC ASSESSMENT
                          L.  John Minnick-/
      The  three primary  by-products  produced  in the  burning  of
coal  in modern power  plants  consist of  fly ash,  bottom ash,
and sulfur  oxides.  Over  the years  considerable technology  has
emerged with  respect  to the  disposal and  utilization  of fly
ash and bottom ash, and a number  of commercially viable pro-
cesses are  available.   Unfortunately, however,  the  large
tonnages  require  that the major portion of these ash  constit-
uents must  be sent to dumps,  which  presents  a  disposal prob-
lem to the  power  industry.   As the  industry  complies  with the
regulations relating  to control of  sulfur oxide emissions,  a
similar situation will  exist if an  attempt is  made  to reclaim
large tonnages of sulfur.  Opportunities  for utilization will,
of course,  be available:   some sulfuric acid can be produced
and sold, and some elemental sulfur or  liquid  S02 may find  a
market, but the majority  of  the sulfur  will  unquestionably  be
faced with  a  lack of marketing opportunities and therefore
may have  to be discharged as waste  to disposal sites.
     The  lime and limestone  scrubbing processes which are
JL/ Executive Vice President,  IU Conversion Systems,  Inc.,
     1500 Walnut Street, Philadelphia, Pa.
                               836

-------
receiving considerable attention at "the present time result
in a by-product slurry or sludge and because of the lack of
any commercial market for this material, it is necessary to
transport this waste to disposal sites.  Therefore these
scrubber operations are currently referred to as "throwaway"
systems.
     The prime purpose of this paper is to outline the oppor-
tunities that are available to convert power plant waste mate-
rials through suitable processing steps to stabilized struc-
tural fill material or, if desired, to usable manufactured by-
products.  This concept converts the "throwaway" waste into
products that can be placed or used and recycled in an ecologi-
cally acceptable manner.
     The main concept of the system is based on modifications
of existing technology dealing with the stabilization or bene-
ficiation of fly ash.  To some degree  fly ash is being used in
the construction industry as an additive to Portland cement,
lime, and asphaltic compositions.  Millions of tons have also
been used throughout the world in the  construction of struc-
tural embankments,  reservoirs, dams, road base materials and
the like.  Large scale utilization programs have been carried
out by governmental agencies in England, France, and by a
number of organizations in the United  States; for example,
National Ash Association, University of West Virginia, and
                                837

-------
by private companies such as Chicago Fly Ash Company and



G. & W. H. Corson, Inc.



     Several years ago the Corson company discovered that the



addition of sulfur oxide salts beneficially improved the struc-



tural properties of stabilized fly ash.  While the original work



was done with waste sulfate sludges from the chemical and coal



industry  (from the neutralization of acid mine drainage), it was



also determined that the underflow sludges from lime and lime-



stone desulfurization scrubbers worked equally well.  The net



effect of this activity has been to provide the coal burning



plants with a practical and economical tail end conversion pro-



cess which includes the sulfur oxide sludges as an important



and integral part of stabilized structural materials.



               DESCRIPTION OF THE CONVERSION SYSTEM



     Basically the tail end process is a modification of



Poz-0-Pac®,  which has been used for many years in structural



applications such as road bases,  reservoirs,  dams, and embank-



ments.   The basic chemistry involved in Poz-O-Pac® is a com-



bination of (1) the pozzolanic reaction that occurs between



the silica content of the fly ash and hydrated lime, which is



added in small amounts (2 to 3% by weight),  to form cementi-



tious hydrated calcium silicate compounds;  (2) the more rapid



reactions that occur between the soluble salts present in fly



ash with lime and the alumina that is found in the fly ash
                               838

-------
glass; and  (3) the mechanical support derived from the addi-
tion of various aggregate materials which serve as bulk fillers
in the fly  ash-lime matrix.  Of considerable importance in the
design of Poz-O-Pac® is the water content of the mixture.  In
fact, whenever fly ash is used as a structural composition,
it is essential that optimum moisture contents be utilized and,
in addition, it is important to specify adequate compaction when
the material is placed.
     In contrast to Poz-O-Pac®, the stabilization of fly ash
using sulfur oxide sludge - commercially referred to as the
Poz-0-Tec* process - is based on additions of large quantities
of calcium and/or magnesium sulfite, calcium sulfate, and, in
certain instances, magnesium oxide which results when dolomitic
lime is used in the treatment of the sulfur oxide acids or
waste gases.  The presence of the sulfites and sulfates are
beneficial to the cementitious reactions and, in addition,
allow the reactions to proceed at a much more rapid rate.  The
engineering properties of the end products are improved to the
point where it is possible to place the stabilized fill compo-
sitions without the addition of aggregate.  This does not mean
that aggregate cannot be used; in fact, Poz-0-Tec* mixtures
have been produced with various types of aggregate including
bottom ash from coal burning power plants, by-product slags,

*A service mark owned by IU Conversion Systems,  Inc.
                              839

-------
culm, or mine  refuse materials,  as well as conventional con-
struction  aggregates.  Work has  also been carried out with
these mixtures to encapsulate potentially toxic materials such
as crushed storage battery cases.
     Much  of the chemistry involved in the Poz-0-Tec* process
can be compared with reactions which have been described many
times in the literature by investigators dealing with portland
cement.  Calcium sulfate reacts  preferentially with calcium
aluminates to  form hydrated calcium sulfoaluminates.  Hydrated
calcium sulfoferrites, analogues of the sulfoaluminates, per-
form equally well in the cementing process.  The Poz-0-Tec*
process involves a special and inventive application of these
established chemical concepts used by the portland cement
industry,  although the conditions under which these reactions
proceed and the ultimate performance criteria are different.
     Of considerable interest to the utilities is the effect
of sulfite ions in the mixtures  since calcium sulfite is pro-
duced in large quantities in the scrubber slurries.  The pres-
ence of sulfite provides several advantages in the conversion
process, an important example being the acceleration of the
cementitious reactions.  Figure  1 shows scanning electron
micrographs which illustrate visually the characteristic of
a fly ash sulfur oxide sludge mixture as freshly prepared and
a similar composition after processing and aging for a period
                             840

-------
of 6 months.  The original platelets that are made up primar-
ily of calcium sulfite (or sulfate) are substantially depleted
after curing.  The crystalline material which is formed assists
in the hardening or cementation process.  Figure 2 shows X-ray
diffraction profiles and demonstrates the rapid rate with which
available lime is used up in these compositions during the
aging process.
     Figure 3 shows a series of patterns developed using an
electron probe scan of a fly ash-calcium sulfite-lime mixture
which has been cured at room temperature for 10 days.  The
presence of aluminum in the scan indicates early formation of
calcium sulfoaluminate crystals which are developing within
the calcium sulfite platelets.
     Figure 4 is a transmission electron micrograph showing
the hydrated calcium silicate structures (tobermorite-like
material) that results in both Poz-O-Pac® and Poz-O-Tec*
compositions.  This cementitious product contributes sub-
stantially to the ultimate strength and durability of the
structure.
     Figure 5 is a flow diagram of the  tail end conversion
process.  In the processing of the fly  ash and sulfur oxide
sludge, it is usually necessary to modify the water content
of the discharge from the power plant which in some cases
can be a fairly major operation.  As indicated in the sche-

-------
matic drawing, the fly ash may be added into the primary de-
watering vessel as a wet slurry or, if the fly ash is dis-
charged from  an electrostatic precipitator, it can be added
in a dry or damp state, thereby assisting in the dewatering
process.  Depending on the type of scrubber that is used, and
the particular discharge system which is adopted, dewatering
may or may not require a secondary state (usually a vacuum
filtering step).  Dewatering devices of this type have been
used conventionally in many large scale applications related
to the coal and steel industries.
     The mixing and conditioning step is basically a modifi-
cation of a conventional Poz-O-Pac® plant.   Figure 6 shows one
of these plants of recent design.  This plant has a production
capacity of 600 tons per hour.   After applying minor changes
to a plant of this type,  the fly ash and sulfur oxide sludge
can be supplied to the mixer and adjusted in composition with
make-up additives which may include additional lime, limestone,
fly ash, bottom ash,  other sulfur oxide salts, and, if desired,
aggregate or other waste products.
     As indicated in the flow sheet, Sulf-O-Poz®, suitable for
placement as a stabilized fill material,  is discharged from the
mixing and conditioning device.
     Although Sulf-0-Poz® is primarily a disposal material, it
has certain minimal structural qualities which would allow its
                              842

-------
application in some cases as structural embankments,  land rec-



lamation projects, liners for reservoirs,  etc.



     As also shown in the flow sheet, Sulf-0-Poz® can be



further processed to produce other compositions such as syn-



thetic aggregate which are suitable for use in producing struc-



tural shapes,  Portland cement or asphaltic concrete,  or may be



supplied directly for use as a high strength monolithic road



base composition equivalent to stabilized road base required



in "primary" highways, airport runways, trucking terminals,



and the like.   This supplementary processing  is of interest



since it offers opportunity to convert the power plant waste



material from a "throwaway" to a "layaway" disposal system.



In this connection aggregate can be stockpiled and made avail-



able for commercial projects as they develop  in the future.



It has been well publicized that shortages of aggregates are



already of serious concern in many areas of the country.



Poz-0-Tec* aggregate  is therefore one  approach to solving



this critical problem.



               PROPERTIES OF THE CONVERSION PRODUCTS



     The two areas in establishing criteria for products pro-



duced in the conversion process consist of, first, the eco-



logical, and,  secondly, the engineering properties of the



materials.  Both of these properties relate to the mix design



and are involved with the ratios of  fly ash to sulfur oxide



sludge, concentrations of lime and water, and also to the
                              843

-------
basic chemical properties of the fly ash and lime.  For ex-
ample, fly ash made from lignite coal which contains high CaO
works quite differently than fly ash produced from bituminous
coal.  Underflows from limestone scrubbers present substan-
tially different design requirements than underflows from
hydrated lime systems.
Ecological Properties
     The evaluation of the various compositions as related to
effect on the environment is dependent primarily on measurements
of permeability and leachability, whether the compositions are
simply placed as a stabilized fill or used as a construction
material.  Where massive quantities of Sulf-0-Poz® are placed
as a stabilized fill,  land improvement, or road base construc-
tion project,  the specific ecological concern is with perme-
ability measurements and possibly with surface runoff.
Table 1 includes some results of permeability tests made on
compositions associated with the recent Transpo 72 project at
Dulles International Airport using standard falling head perme-
ability procedures.   Figure 7 is a graph that shows the reduc-
tion in permeability in sulfur oxide containing mixtures when
compared with those that have been previously made with stan-
dard Poz-O-Pac® compositions.  Permeability values of freshly
                                                    -5
prepared compositions usually run in the range of 10   cm/sec,
which quickly changes during the hardening process to values
                              844

-------
               •~       ~
ranging from 10   to 10   cm/sec.  Obviously a road base or


embankment with this low permeability provides excellent


protection against leaching of deleterious materials through


the mass.


     From the standpoint of making leaching measurements, while


no "standard" tests are available, it is recognized that all


materials can be subjected to special laboratory procedures


that will exhibit the presence of small quantities of soluble


salts.  Accelerated tests run on specimens weighing 500 grams


which were placed in plastic containers covered with two


liters of distilled water then shaken continuously for 48


hours are given in Table 2.  The table includes limits for


potable water as described in U. S. Health Service, HEW


Bulletin No. 956 on Drinking Water Standards.  Figure 8  is a


photograph of the equipment used in the shaking test.  For the


sake of comparison, materials other than stabilized fly  ash


compositions are included in this table.  It will be noted


that raw,  loose materials are subject to substantially higher


quantities of leachate in a test of this type.  This obser-


vation is a further indication of the beneficial effect  of


forming a monolythic structural Sulf-0-Poz® mass.


     The question of leaching of heavy metals from surface run-


off is one that has been studied extensively by the United States


Potters Association and the International Lead Zinc Research
                              845

-------
Organization,  Inc.   The object of these studies was to eval-
uate the effect of temperature,  time,  and pH on the amount of
heavy metal that could be leached from ceramic surfaces.
Recent studies at Rutgers University School of Ceramics using
a scanning electron probe has shown the pattern of lead con-
centration across the surface of the ceramic ware.  Figure 9
is an example of this type of investigation.  Careful exami-
nation of the lead profile shows that leaching has occurred
to a depth of approximately 1/2 micron.   Techniques of this
type may be of interest for stabilized fly ash materials since
this affords a good method of determining the total impact of
consolidated masses on the environment.   It is evident from
laboratory leaching tests of the type described that stabilized
structures can provide virtually non-significant quantities of
leachable ions to the environment.
     Runoff tests have also been carried out by preparing slabs
of the fly ash-sulfur oxide compositions.  Two liters of dis-
tilled water were poured over the surface of the specimens to
provide a runoff that could be tested by atomic absorption
techniques.  Table 3 shows some of the results of these tests.
Figure 10 illustrates the technique that has been used.
     It is beyond the scope of this paper to make detailed
recommendations as to the most appropriate method of test for
these stabilized compositions or to suggest absolute limits
                             846

-------
that might be acceptable from an environmental standpoint.
However, the values that have been obtained fall well within
limits that have been established by a number of governmental
organizations which have addressed themselves to this type of
situation.
     The Poz-0-Tec* compositions that have been studied not
only meet rigorous environmental requirements, but due to the
high pH of these compositions they, in fact, can serve to
"protect" the environment from attack.  It has been noted,
for example, that when spoil bank materials which contain
pyrites are encapsulated within Poz-0-Tec* mixtures that
leaching of sulfuric acid is avoided and the attack on vege-
tation in the spoil bank area is therefore eliminated.  This
type of chemical encapsulation is a phenomenon analogous to
studies that have been made in Portland cement compositions.
The technical literature provides numerous references which
describe the method by which soluble components migrate into
the lattice complexes of a cement structure becoming chemi-
cally bound, thus preventing the leaching of soluble ions.
A well known example of this is the lime-alkali-silica complex
resulting from the solid state diffusion of soluble NaOH or
KOH into the lime hydrous silica complex.  When in this form,
the Na+ and K+ ions are chemically entrapped thus preventing
potentially harmful leaching.
                              847

-------
Engineering Properties
     The most important engineering factor which must be eval-
uated is the strength of the reacted mixtures.  A correlation
exists between the formation of cementitious compounds of the
types described above and the development of strength and other
engineering properties in fly ash-sulfur oxide mixtures.
Figure 11 shows the effect of aging on the strength of a
Sulf-O-Poz® composition.  The method of measuring penetration
resistance is based on a modified vicat procedure, the pene-
tration needle having a cross section area of one-fortieth of
a square inch.  The measured penetration resistance represents
the pressure that must be applied in pounds per square inch to
cause a penetration of one inch.
     Figure 12 provides information on a similar composition
when cast into 2 x 2 x 2" cubes and tested for unconfined
compressive strength and Figure 13 presents California Bearing
Ratio values.   Based on results of this type, it is possible
to design compositions with strength properties commensurate
with intended applications.
     Table 1 presents compressive strength tests that were made
for design purposes at Transpo 72.  Table 4 shows results that
were obtained from a field project and is based on cores that
were removed after periods of 4 and 6 weeks.
     The major criteria which relate to the ultimate strength
                              848

-------
of these mixtures are composition, moisture content, and com-
pacted density.  Where a Sulf-O-Poz® fill or a land rehabil-
itation project is contemplated, strengths may not be required
to exceed those that are conventionally developed in embank-
ments, earthwork dams, and the like.  However, where the product
is to be used under heavy loads such as for foundations of
bridges or for road base, strengths must not only be higher but
should be required to reach specified values at earlier ages.
The moisture contents which can be used may range from 20% to
50%, depending on the end use which is contemplated and on
other factors.  In certain applications additives may be used
to accelerate the early strength.
     In addition to strength, a second factor that  is of con-
siderable importance from an engineering standpoint is struc-
tural integrity.  Procedures that have been used to evaluate
this characteristic are usually of two types; first, laboratory
measurements of dimensional stability, and, secondly, field
performance.  Data from a few of  the mixtures that  have been
evaluated in the laboratory are given in Figure 14.  These
measurements were made on moist cured bars  (1 x 1 x 10") cured
at 73°F.  In each case slight expansion is  developed at early
ages, although the degree of expansion reaches a constant
value after several weeks and does not change significantly
from that point on.  It is believed that this type of expansion
                              849

-------
is beneficial particularly where the material is used in mass



applications since it helps to control shrinkage cracks  (a



concept that is used when expansive portland cements are used



in mass structures).



     The structural integrity of the sulfur oxide-fly ash mix-



tures as observed in field applications to date has indicated



that these mixtures perform quite well.  By making comparisons



with early Poz-0-Pac® evaluation programs, it is indicated that



mass structures such as have been placed in reservoirs and dams



can perform for many years with little detectable deterioration.



Poz-0-Tec* mixtures which have been used for road base work have



proven to be superior to the Poz-0-Pac® compositions,  although



like Poz-O-Pac® they should be covered with an adequate wearing



course,  as the stabilized mixtures per se are not capable of



withstanding the action of excessive traffic.



                      ECONOMIC CONSIDERATIONS



     The costs of converting the by-product wastes of a power



plant depend on whether or not the end product is produced in



a marketable form.   The least conversion costs result when



Sulf-0-Poz® is produced since this product may simply be dis-



posed of as stabilized fill which is designed to meet full



ecological, but only minimum structural and engineering criteria.



In many cases,  depending on factors to be elaborated below,



Sulf-O-Poz® can be economically competitive with ponding and
                             850

-------
ecologically unsound disposal methods.
     In certain markets,  it may be economical to produce ana
sell or stockpile for future sale a usable end product such
as Poz-0-Tec* road base or aggregate.   Revenues from the sale
of these products might appropriately be credited to the
additional costs of converting the waste by-products to
stabilized structural materials.
     Since both the nonmarketable and the marketable approaches
result in an ecologically sound material, the decision as to
which approach to take is primarily an economic one.
     A second area of importance in developing cost estimates
is related to the nature of the installation, thus if the con-
version process is added on to an existing plant, the capital
costs and operating costs may be quite different than would be
realized in the case of a completely new installation.  In an
operating plant, fly ash may already be recovered in electro-
Static precipitators and this would obviously affect the capital
cost of the tail end system.
     The major factors which influence the cost of the conversion
may be summarized as follows:
     - Annual tonnages to be handled by the conversion
       process.
     - New boiler installation versus existing facilities.
     - The type of equipment selected for fly ash removal
       — for example, electrostatic precipitators versus
       wet scrubbers.
                             851

-------
     - The chemical analysis of coal — sulfur,  CaO,
       and ash contents.
     - Location of plant — on site versus off site.
     - Transportation costs — to and from conversion
       plant.
     - Redundancy factor — duplication of equipment
       versus emergency holding basins,  etc.
     - Type of scrubber — limestone versus lime.
     - Acquisition and cost of land.
     - Type of end product selected.
     In view of the fact that each of these items considered
alone or grouped together can cause major variations in both
the capital and operating costs,  it should be recognized that
specific estimates cannot be given unless each situation is
reviewed on an individual basis.   On the other hand, IU
Conversion Systems,  Inc. has had an opportunity to make numer-
ous cost evaluations based on the programs which are currently
underway,  and it is at least possible at this time to suggest
a few examples which may be helpful to the utilities in eval-
uating the conversion system approach in individual cases.
     In several installations in which lime scrubbers have
been specified for new plants in the general size range of
1000 to 2000 megawatts, the cost of conversion to Sulf-O-Poz®
fill material will range between about $1.50 per ton to
$2.50 per ton, the variation depending in large measure on
items such as listed previously.   Converting this to a basis
                              852

-------
of cost per million Btu for coal of approximately 3 to 4%
sulfur and 10 to 15% ash represents a range of about 2 to
3-1/3 cents per million Btu.  Adding the necessary equipment
and operating staff to convert from a simple Sulf-0-Poz®
stabilized fill process to one in which commercial products
are fabricated would increase these costs by approximately
20 to 50% before allowing for any offsetting credits from
the sale of commercial products.
     IU Conversion Systems, Inc. can provide the power plant
with the necessary hardware on a turnkey basis and will under-
take to operate the conversion plant on the basis of a long-
term contract.  In general, the power industry prefers to"
contract with an outside organization to take care of the
disposal of its waste materials since it is not normally
geared up to operate this type of plant.  What the industry
requires then is a reliable process which will give them the
assurances of low-cost disposal with adequate warranties and
guarantees.   It is also essential that the conversion system
shall not interfere with the operation of the power plant.
     At the presert time the cost estimates given above appear
to be quite attractive since they are, in fact, competitive
in many instances to what the power plant pays simply to hire
a trucking organization to  haul the waste material to a dump
or lagoon.  In view of the  concern of the environmentalists
                               853

-------
and the difficulty in obtaining permits for disposal of the



waste material, it is clearly indicated that a system which



converts the waste to ecologically safe compositions is to



the advantage of a power plant operation.



                              SUMMARY



     The extension of existing technology dealing with the



stabilization of fly ash into the area where waste sulfur



oxide sludges can be utilized provides a practical solution



to the question of fixation and disposal of flue gas waste



products.   The real effectiveness of the fixation process



depends on the concept of converting the waste materials into



compositions that comply with ecological and engineering



criteria that allow for the disposal or utilization of the



product in a manner which is beneficial or at least not de-



leterious to the environment.



     The performance of the stabilized mixtures as demonstrated



in laboratory tests and field evaluation programs has been



quite favorable.  The ratios of ash to sulfur oxide sludges



that are produced in the average coal-burning operation fall



within the ranges that can be utilized for the manufacture of



the various compositions.



     The conversion system is adaptable to power plants as a



retrofit application or to new installations.   Since the pro-



cess depends on the availability of fly ash,  it is not appli-
                             854

-------
cable to oil fired operations unless such operations are
carried out in the vicinity of stockpiled waste fly ash.
     Cost evaluations which have been made to date indicate
that for full sized plants the conversion system is usually
competitive with costs that would be incurred simply by
transporting the waste material to dumps or lined lagoons.
In specific cases commercial opportunities may be available
which would allow the utilization and sale of products
produced in the process.
     Based on the findings to date, it is indicated that a
realistic solution to the disposal of sulfur oxide sludges
from lime and limestone scrubbers can be attained.
                              855

-------
                                            TABLE 1

                  RESULTS OF TESTS OF SELECTED  STABILIZED ROAD BASE MIXTURES
                         PREPARED AT DULLES AIRPORT TRANSPO 72 PROJECT
CO
Moisture
Content
19.5
19.4
20.0
19.8
19.7
20.0
19.1
Dry
Density
(pcf)
98.8
98.1
98.3
98.2
100.6
98.8
100.4
Falling Head
Permeabilities
(cm/sec)
7 Days
2.4 x 10~6
N.D.*
2 . 9 x 10"6
6.5 x 10~6
5.7 x 10~6
1.0 x 10~6
N.D.
Compress ive Strength
at 100° F (psi)
2 Days
301
267
369
196
333
290
200
14 Days
732
586
630
458
772
761
868
28 Days
881
622
889
490
861
789
1091
                            *N. D.
Not Determined

-------
                                                                  TABLE 2


                                          ATOMIC ABSORPTION TESTS FOR LEACHABLE IONS ON SELECTED
                                                SPECIMENS SUBJECTED TO 48 HOUR SHAKING TEST
00
en
Total
Dissolved
pH* Solids
FEDERAL SPECIFICATIONS-MAX.
Individual Solid Specimen
Dulles Cylinder (13 Days)
Dulles Cylinder (22 Days)
Poz-O-Tec* Test Road Core
Poz-O-Tec* Test Road
Cylinder
Poz-O-Pac® Cylinder
Fly Ash Concrete
Cinder Block
Clay Brick
Asphalt Roofing Shingle
Aggregate
Argillite
Do lorn i tic Limestone
Calcitic Limestone
Steel Slag Aggregate
Pumice
Fly Ash-Sludge Aggregate
Cement Mortar Balls
Mine Tailings
Loose Powdered Materials
Fly Ash
Portland Cement
Water Samples
Tap Water
Snow Sample from Pittsburgh
Water Supply (Peggs Creek)
10.6

9.5
9.5
6.7

9.2
9.3
10.7
8.2
7.3
7.1

6.9
9.75
8.4
10.8
7.1
11.7
9.0
3.95

9.8
12.0

7.5
6.45
7.25
500

840
620
90

250
150
440
410
110
150

120
96
180
840
120
700
530
130

2900
3700

180
40
316
Sulfate Cl
250

100
120
16

136
44
170
60
28
46

28
8
8
16
< 1
< 1
27
6

1500
200

36
< 1
"
250

8
12
14

16
26
46
6
12
22

22
18
—
28
10
16
8
2

8
20

76
6
"
Al
None

.38
.37
.03

.05
.10
.22
.01
.03
.01

,07
.02
.02
.05
.06
.03
.04
.05

. 11
.05

.02
.06

Total
Iron
.3

.08
.08
.06

.10
.25
.01
.04
.10
. 12

,06
.36
1.8
. 15
2.2
.26
.17
. 15

.26
.44

< .01
.46
2.9
Mn
.05

. 18
. 16
<.05

.10
<.05
<.05
<.05
<.05
<.05

<.05
<.05
<.05
<.05
<.05
<.05
<.05
<.05

<.O5
<.05

<.05
<.05
<.05
Cu
1.0

.08
.08
< .01

.08
.08
.04
< .01
.01
< .01

.08
< .01
< .01
< .01
< .01
< .01
< .01
. 16

< .01
< .01

.08
< .01
.05
Zn Cd Cr+3
5.0 .01 .05

.02 <.01 .02
< .01 <.01 <.01
< .01 <.01 <.01

< .01 <.01 <.01
< .01 <.01 <.01
< .01 <.01 <.01
< .01 <.01 <.01
< .01 <.01 <.01
< .01 <.01 <.01

< .01 <.01 <,01
.02 <.01 <.01
.04 <.01 <.01
.02 .01 <.01
.03 <.01 <.01
.02 <.01 <.01
. 01 <. 01 <. 01
.02 <.01 <.01

. 01 <. 01 <. 01
< .01 <.01 <.01

.05 <.01 <.01
< .01 <.01 <.01
.02 <.01 <.01
As Hg
.01 .001

.02 <.001
.02 <.001
<.01 <.001

.01 <.001
.01 <.001
<.01 <.001
•c.Ol <.001
<.01 <.001
<.01 <.001

<.01 <.001
<.01 <.001
<.01 <.001
<.01 <.001
<.01 <.001
<.01 «.001
<.01 <.001
<.01 <.001

<.01 <.001
<.01 <.001

<.01 <.001
<.01 <.001
<.01 <.001
Pb
.05

.08
.09
<_ oi

.01
.02
<, 01
.02
<.01
<-01

<.oi
!o7
.03
.01
.06
.06
.03
.07

.06
.04

.04
.02
.02
Sn
None

. 10
. 10
< oi

<. 01
•c.Ol
<. 01
<. 01
<. 01
•S.01

<.01
<. 5
<. 5

-------
                                 TABLE 3

           ATOMIC ABSORPTION TESTS MADE ON SURFACE RUNOFF OF A
                     STABILIZED FLY ASH-SLUDGE MIXTURE



in
CO




pH*

Dulles Cylinder (13 Days) 7.0
Dulles Cylinder (23 Days) 6.9
Sulfite Beam 7.2
Total
Dissolved
Solids

100
96
85

Total
Sulfate Cl Al Iron Mn Cu Zn Cd Cr

26 12 .10 .22 <.05 <.01 <.01 <.01 <.01
32 18 .15 .06 <.05 <.01 <.01 <.01 <.01
8 18 .13 .06 .60 .12 <.01 <.01 <.01


As Hg Pb Sn

<.01 <.O01 <.01 <.01
<.01 <.001 c.Ol <.01
<.01 <.001 .03 <.01
+With the exception of pH, all values are reported in parts per million.

-------
                                           TABLE 4


                          RESULTS OF FIELD TESTS SHOWING  COMPARISON
                          OF POZ-O-PAC® AND POZ-O-TEC*  FORMULATIONS
           Description
                            Dry
                          Density
                            (pcf)
Compressive Strength
   at 100°F (psi)


   2 Days  7 Days
            Strength of Cores
            from Road1  (psi)


            4 Weeks  6 Weeks
CO
in
<£>
Standard Fly Ash Mix       121.2


Fly Ash-Sludge Blend A     121.4


Fly Ash-Sludge Blend B     120.8
     66


    348


    318
770


729


746
NCP2      NCP


NCP      1034


756      1089
                     Average temperature during curing period was 50°F.

                    2
                     NCP - No core possible due to insufficient strength.

-------
FIGURE 1 — Freshly prepared mixture of fly ash, sulfur oxide
             and sludge (Left View).

             Ettringite crystals formed after aging
             {Right View).
                            860

-------
                                     FIGURE 2 — Comparison of fly ash-calcium sutfite-Vime
                                               mixture showing depletion of free Ca(OH)2
                                               after aging. Upper profile — imaged.
                                               Lower profile — aged 2 weeks.
8
8
I
»

8
i
£
M
I
_c
'i
*-•
3
&
                                                  Ca(OH)2
                                  16
18
20
22
                                    24
                                                  Angle of Diffraction (26)

-------
                       2000X
FIGURE 3 — Electron probe patterns showing early reaction
            involving AI2O3 and calcium sulfite.
                                                               Al
                                                               Ca
                          862

-------

FIGURE 4 — Tobermorite crystals resulting from reaction
            between SiO,  in fly ash with lime.
                         863

-------
      I
      I
      i
      i.
00
  SULFUR
  OXIDES
UNDERFLOW
                                                                                           t	mill
                                                                                                        MAKE-UP PROCESS
                                                                                                            ADDITIVES
                                                                                           Alllllllttt
                                                                                            t       %

                                             PRIMARY DEWATERING                          *'*»***


                                               A                                             ^r
                                               • ••••ttlllllllllMllllllllllllllllllllllMIIIIIMIIIIIIlTlllllllllllllltiiiuu
                                              • ^                                                                  •
                                                          SECONDARY
                                                         DEWATERING
                                                               MIXING &
                                                            CONDITIONING
                                                                                                                                             Sulf-O-Poz®Products
                                                                                                                                                     Dams
                                                                                                                                                   Reservoir!
                                                                                                                                               Impermeable Liners
                                                                                                                                                   Road Base
                                                                                                                                                 Structural Fill
                                                                                                                                            •••••••••••••••i
                                                                                                                                             Structural Products
                                                                                                                                                   Aggregate
                                                                                                                                                   Concrete
                                                                                                                                                Structural Shapes
                                                          FIGURE 5 - Schematic diagram of Poz-O-Tec* process.

-------
FIGURE 6 - Fly Ash stabilization (Poz-O-Pac®) plant.
                      865

-------
  100
b
^
X
U
00
<
111

5
DC
LU
Q.
                                                Standard Fly Ash Mix
                                                Sludge - Fly Ash Mix
                               468

                              AGE-DAYS OF CURING AT 100°F
                     FIGURE 7 - Comparison of Poz-0-Pac® and Poz-O-Tec*

                               permeability values.


                                           866

-------

  ****-

                           id
FIGURE 8 - Shaking apparatus used for laboratory shaking test.
                        867

-------
FIGURE 9 — Electron probe scan showing lead profile in
            ceramic ware subjected to long time outside
            exposure.


-------
FIGURE 10 - Runoff test as conducted on sulfite
              stabilized fly ash specimens.
                       869

-------
      6000
      4000
-
3
8


I
I
2


c

I
         2000
                                                      Age (Weeks)


                                      FIGURE 11 - Penetration resistance for a typical fly

                                                  ash-calcium sulfate-lime mixture.

-------
 2000
f
I
w
5
's
I
o
u
3
  1000
        0
8
                                                    Age (Weeks)
                                  FIGURE 12 — Compressive strength fora composition similar
                                              to that given in Figure 11. Moisture
                                              content of composition is 35%.

-------
   200--
                 Sludge- Fly Ash Mix
CD
z
S   150
z
oc
O
                                 Standard Fly Ash Mix
    100-
                    1
                        AGE-DAYS OF CURING AT 70°F
              FIGURE 13 - California Bearing Ratio for Poz-0-Tec* road base.
                                    872

-------
          .012
   s.
00
^J
Ul
          .000
               0
8                 12
   Age. Weeks at 73° F
20
                                          FIGURE 14 — Early expansion of fly ash-sludge mixtures.

-------
   UTILIZING AND DISPOSING OF SULFUR PRODUCTS
FROM FLUE GAS DESULFURIZATION PROCESSES IN JAPAN
                         by

                    Jumpei Ando
          Faculty of Science and Engineering
                   Chuo University
          Kasuga, Bunkyo-ku, Tokyo, Japan
                         875

-------
           Utilizing and Disposing of Sulfur Products from

             Flue Gas Desulfurization Processes in Japan


                             Jumpei Ando


     Japan's energy supply has doubled in the past five years depending
mainly on growing amounts of Imported oil.  Even though nuclear energy
may eventually supplant petroleum, Imported oil and sulfur derived from
it are likely to keep increasing for some time to come.  Regulations on
SOn emissions are becoming more and more stringent, inducing greater
efforts for the desulfurization of fuel oil and waste gas.  While virtually
all of the by-products from the desulfurization—elemental sulfur, sodium
sulfite, sulfuric acid, gypsum, etc.—have been utilized so far, it will
not be long before supply of by-products runs far ahead of demand.  Gypsum
is generally regarded as the most rational by-product because of its
several uses including land filling and also because of its ease of aban-
donment.  Elemental sulfur is considered to be another logical by-product
because it may be stored for some time for future use.

           1.  Characteristics of desulfurization in Japan

     Of the total electric power (386 billion Kwhr) produced in Japan in
1971, about 62% depended upon oil, 12% upon coal, 22% upon hydraulic
power, and 2% upon nuclear energy.1   Heavy oil, a residue of atmospheric
distillation of crude oil, has been used as a major fuel in recent Japan.
Heavy oil consumption in 1971 reached 745 million barrels.  About 30% of
the heavy oil was burned in power plants and the rest in other industries
and buildings.1   Heavy oil from high-sulfur crude contains 4% sulfur.
In 1971 nearly one-fourth of the total heavy oil was subjected to hydro-
desulfurization giving 287,000 tons of elemental sulfur as by-product.
Still nearly 6 million tons of SO2 was emitted by burning the heavy oil,
constituting the chief source of SC^ emissions.

     At present in Japan, more than 100 commercial and prototype plants
for waste-gas desulfurization are on stream.  Most of them treat either
flue gas from oil-fired industrial boilers or waste gas from chemical
plants, smelters, etc.  Nine major power companies which produce more
than 70% of tbtal power in Japan have recently decided to build many
flue gas desulfurization plants.  The total capacity of the plants of the
power companies will increase from 357 Mw in 1972 to 2,700 Mw in 1974 and
to 4,800 Hw in 1976.

     A salient-feature of the desulfurization efforts in Japan is that
they are oriented toward processes that yield salable by-products such
as sodium sulfite for paper mills, sulfuric acid, gypsum for retarder
of cement' setting and for gypsum board, etc.  The feature is predicated
on the following .circumstances:  (1) Japan is subject to limitations In
domestic 'supply of sulfur and its compounds as well as in land space

                                 876

-------
 available for disposal of useless by-products.   (2) Most desulfurization
 plants treat flue gas from oil-fired boilers, waste gases from chemical
 plants, etc. which contain less  dust than does coal-fired flue gas and
 therefore are suited to the production of by-products of high purity.
 (3)  Most steam power plants are  located close to chemical plants; hence
 it is easy to utilize the desulfurization by-products in chemical plants.

      However,  since so many desulfurization plants are to be built, it
 is considered that in future the supply of the by-products will far
 exceed the demand,  necessitating the abandonment of a substantial portion
 of them.   Even now there are many small plants, (less than about 30,000
 scfm)  which produce waste solutions of sodium sulfate, sodium sulfite, and
 ammonium sulfate;  it is uneconomical to recover by-products at small
 plants.


                   2.   By-production of sodium salts

      Since 1966 more than 50 units have been built to recover S02 to
 by-produce sodium  sulfite for paper mills.  Some of the plants are listed
 in Table  1.   In most of them, S02 is absorbed with a sodium sulfite
 solution to form sodium bisulfite which is then treated with sodium
 hydroxide to produce sodium sulfite.  (By treating flue gas with a sodium
 hydroxide solution,  C02 is  also  absorbed.)  One-half of the sulfite solu-
 tion is recycled to  the absorber and the rest is concentrated to produce
 anhydrous sodium sulfite.   An example of the process is shown in Figure 1.
 Flue gas  is  first washed with water in a conventional scrubber for cool-
 ing  and dust removal;  60 to  80%  of dust which is contained in flue gas
 from oil-fired  boiler to an extent of about 0.1 grain/scf is removed.
 The  purity of the product ranges from 90 to 97%; small amounts of sodium
 sulfate,  dust,  etc.  are contained in the product.  Sodium sulfite solu-
 tion of about 20% concentration  is also produced in some plants and used
 for  paper mills.

     A new semi-wet process  for  sodium scrubbing has been developed
 recently  by  Hitachi,  Ltd.  (Figure 2).  Flue gas from oil-fired boiler
 at about  350°F  after  passing through a dust eliminator is introduced
 into a reactor  into which sodium hydroxide solution is fed.  By the
heat of the  gas, moisture is removed and powdery product consisting of
 sodium sulfite, sulfate  and  carbonate (for example, Na2SOo-60%,
Na2S04-20%,  and Na~C03-20%)  is formed which is caught by a dust separator.
The product  is usable for kraft pulp production.
                                 877

-------
CO
^•J
CO
                        Table 1.  Major Waste-Gas Desulfurization Plants that By-Produce Sodium Salts

                                                                                    (Charge: NaOH)
                                                                          Unit capacity
Process developer
Oji Paper
Oji Jinkoshi
Kureha Chemical
Kureha Chemical
Showa Denko
Showa Denko
Bah co -Ts uk i s h ima
Banco -Tsukishima
Hitachi, Ltd.
IHI-TCA
Mitsubishi (MKK)
Kurabo Ind.

Toyobo Co.
IHI-TCA

Product
Na2S03
Na2S03
Na2S03
Na2S03
Na2S03
Na2S03
Na2S03
Na2S03
Na2S03
Na2S04
Na2S04
Waste
Na2S04
Waste
NaoSO*
i •**
Waste
Na2S04
User
Oji Paper
Oji Paper
Kureha Chemical
Konan Utility
Aj inomoto
Asia Oil
Daishowa Paper
Daio Paper
Jujo Paper
Mitsui S. 0.
Asahi Glass
Bridgestone Tire

Toyobo Co.
Nissan Motor

Plant site
Kasugai
Tomakomai
Nishiki
Konan
Kawasaki
Yokohama
Yoshinaga
lyomishima
Miyako j ima
Sakai
Amagasaki
Tokyo

Shogawa
Oppama

(1,000 scfm)
805 (in 12 units)a'b
400 (in 4 units )a
176a, 176a
123a
159a
142a
129b, 65b, 26b
88a, 70a
57a
88a, 88a, 88a
130C
71a

24a
67a

Date of completion
1966 - 1972
1971 - 1972
1968
1972
1971
1972
1971
1972
1972
1973
1973
1972

1971
1972

   a: Oil-fired boiler
b:  Kraft recovery boiler
c:  Glass melting furnace

-------
00
^J
<£>
                      Water
                     (27Vhr)
           Flue  gas
(I?6f000scfm)
(S02 l.^OOppm)
                   To wastewater
                       treatment
                              W&ste gas 
-------
00
00
o
                t
Reactor
  Dust    \.

collector
                                                        To stack
                                   t
                                                                         NaOH
                                                                           i
                                                i^^.i

                                                I By-product
                     Figure 2    Hitachi semi-wet  sodium process

-------
     These sodium processes are simple and are operated with ease.
Investment cost is low.  Demand for sodium sulfite, however, is limited.
As shown in Table 2, production of the sulfite has increased rapidly
with the progress of desulfurization and has resulted in a considerable
price drop for the sulfite due to oversupply.

                                                               2
     Table 2.  Production and Price of Anhydrous Sodium Sulfite
                                   1967    1968    1969    1970    1971

     Production (1,000 t)           113     135     171     289     330

     Price ($/t)                     64      62      61      58      60

     Under such situation, the following ways of sodium scrubbing have
been developed recently:  (1) Several plants have been built recently
to by-produce salable solid sodium sulfate.  The sulfate is produced by
air oxidation of sodium sulfite solution.  Demand for the sulfate is
also limited.  (2) Many smaller plants have started to produce waste
solution of sodium sulfate or sulfite.  Some of these plants are listed
in Table 2.  (3)  Showa Denko as well as Kureha Chemical jointly with
Kawasaki Heavy Industries have developed sodium-limestone double alkali
processes which are described in the present author's separate paper
for the symposium entitled "Flue Gas- Desulfurization Technology in
Japan."
                                  881

-------
                          Table 3.  Major Waste-Gas Desulfurizatlon Plants that By-Produce Gypsum
CO
KJ
Process developer
Mitsubishi-JECCO
Mitsubishi-JECCO
Mitsubishi-JECCO
Mitsubishi-JECCO
Mitsubishi-JECCO
Mitsubishi-JECCO
Babcock -Hitachi
Chiyoda
Chiyoda
Chiyoda
Chiyoda
Chiyoda
Hitachi, Ltd.
Showa Denko
Kureha-Kawasaki
Kureha- Kawasaki
Nippon Kokan
Chemi co -Mitsui
Absorbent
Ca(OH)2
Ca(OH)2
Ca(OH}2
Ca(OH)2
Ca(OH)2
CaC03
CaC03
H2S04, CaC03
H2S04, CaC03
H,,SQ. , CaCOr
Z 4 ^
H_SO . . CaCO,;
2. 4 -S
H2S04, CaC03
Carbon , CaCO „
KaOH, CaC03
NaOH, CaCO 3
NaOH, CaC03
NH3,Ca(OH)2
Ca(OH)2
User
Nippon Kokan
Kansai Electric
Onahama Smelting
Kansai Electric
Tohoku Electric
Tokyo Electric
Chugoku Electric
Fuji Kosan
Mitsubishi Rayon
Daicel
Hokuriku Electric
Mitsubishi
Petrochem.
Tokyo Electric
Showa Denko
Shikoku Electric
Tohoku Electric
Nippon Kokan
Mitsui Aluminum
Plant site
Koyasu
Amagasaki
Onahama
Kainan
Hachinoe
Yokosuka
Mizushima
Kainan
Otake
Aboshi
Shinminato
Yokkaichi
Kashima
Chiba
Shintokushima
Sendai
Keihin
Omuta*
Capacity
(1,000 scfm)
37d
59a
54e
235a, 221a
224a
220a
170a
93a
S33
S9a
420a
413a
250a
340a
250a
220a
88b
226C
Date of completion
1964
1972
1972
1974
1974
1974
1974
1972
1973
1973
1974
1974
1972
1973
1974
1974
1972
1972
        *Producing waste calcium sulfite; gypsum will be produced in near future.
   a: Oil-fired boiler                    d: Sulfuric acid  plant
   b: Iron-ore sintering plant            e: Smelting furnace
   c: Coal-fired boiler

-------
                     3.  Gypsum and calcium sulfite
 3.1  Uses
     Gypsum is considered the ideal by-product for the time being.
Many processes have been developed in Japan recently that recover
salable gypsum with sufficiently good quality to make it available as
retarder of cement setting and for production of gypsum board (Table 3).
Six of the processes which seem to be of more interest for possible
application in the U. S. A. are described in the author's paper referred
to above.  Demand for and supply of gypsum in Japan are illustrated in
Figure 3.  Prices of by-produced gypsum (mainly phosphogypsum from
wet-process phosphoric acid production) are shown in Table 4.
                                                        2
            Table 4.  Prices of By-Produced Gypsum ($/t)
     All of the phosphogypsum and most of the other by-products have been
used for cement and board production.  As demand for gypsum has increased
substantially, supply is on the short side at present, resulting in some
price increase.

     In order to use by-product gypsum for wall board, proper crystal
size and shape and less impurity are required to ensure high strength.
Phosphogypsum from "the conventional dihydrate process usually has a
small crystal size (10-30 microns) with much impurity and gives less
strength.  Japanese phosphoric acid producers have developed hemihydrate-
dihydrate processes by which good crystals of gypsum (50-150 microns)
with less impurity are obtained.  Good phosphogypsum from these processes
has a strength equal to or better than natural gypsum.

     Such background has helped find outlets for by-product gypsum from
desulfurization.  For example, a wall board production plant of
Onahama-Yoshino Gypsum Co., Ltd., using gypsum (450 t/day) recovered
from the Mitsubishi-JECCO process is under construction at Onahama works,
Onahama Smelting and Refining Co.  Another example has been presented
by Mitsui Aluminum Co. which has produced waste calcium sulfite by
treating coal-fired flue gas (226,000 scfm) with the Chemico-Mltsui
process.  Mitsui Aluminum, jointly with Mitsui Toatsu Chemicals, Inc.,
has succeeded in tests to manufacture wall board from gypsum obtained
from the calcium sulfite.  The gypsum has a little less than 90% purity,
containing fly ash and other impurities.  Mitsui Aluminum plans to
install a prototype reactor within 1973 to convert about one-fourth of
its by-product calcium sulfite to gypsum.  It is likely that a full-
scale reactor will be installed in 1974.
                                 883

-------
CO
CO
        W

        O
        •P
O

CO

o

•H
rH

Q
           8 r
   6  -














—






—




•§•
B
Q



OU
B

C


rH
CL
Q.
W


OS
R


P


















OT1

B


C






OS
H



P













at
1
O

OTT






C


a.
to

OS

R



















OU

B




c



OS

H





p



•










           2  -
                   196?
                                          1973
            Demand

            Supply
1976
             1970


C: Cement   B: Board   OU: Other uses

Pi Phospho-gypaura    R: Recovered   OS: Other  sources
            Figure 3  Demand for and supplv of gypsum in Japan

-------
      By-product gypsum is available also as retarder of cement setting.
Gypsum  is added to Portland cement clinker to the extent of 3-4% at
grinding.  Nippon Kokan  (Mitsubishi-JECCO lime-gypsum process) has
sold  the by-product gypsum for cement since several years ago.  Gypsum
should  be charged continuously into the cement mill.  Powdery gypsum
with  more than 10% moisture tends to form a "bridge" in the hopper and
cannot  be charged smoothly.  By the Mitsubishi-JECCO process and the
Nippon  Kokan ammonia-lime process, the product gypsum from a centrifuge
has a low moisture content of about 10%.  Gypsum of about 90% purity
containing fly ash is useful, but alkali-rich gypsum by-produced from
the double alkali process would impair cement setting.  New uses of
gypsum  and calcium sulfite have been developed recently in Japan.
Mitsui  Toatsu Chemicals, Inc., jointly with Taisei Construction Co. has
started producing a new material "gypsum polymer composite" with gypsum
and methyl methacrylate.  Lion Fat and Oil Co. jointly with Idemitsu
Kosan has commenced production of synthetic paper from calcium sulfite
and polyethylene at a weight ratio of about 70:30.  By-produced gypsum
and calcium sulfite could be used if the dust in flue gas is removed
reasonably well so that the color is not too dark.

3.2   Discarding

      Gypsum is produced from calcium sulfite by oxidation.  Oxidizer is
virtually unnecessary in certain plants—such as Amagasaki plant,
Kansai  Electric Power (Mitsubishi-JECCO process) and Rumagaya plant,
Chichibu Cement (IHI-TCA process)--which treat flue gas with relatively
low S02 and high 02 concentrations; essentially all of calcium sulfite
is converted to gypsum in the scrubber.  But normally an oxidizer (for
oxidation with air) or a reactor (by the above mentioned Mitsui gypsum
process in which a catalyst is used) is required which adds some cost
to desulfurization.  Gypsum has advantages over calcium sulfite even
for land filling or discarding.  Gypsum can be grown into fairly large
crystals (50 to 300 microns); moisture content of the centrifuge dis-
charge  can be made as low as 10%.  On the other hand, the crystal of
calcium sulfite is normally very small (1-10 microns); centrifuge
discharge contains about 60% moisture and is like a paste.  The calcium
sulfite may not suit land filling because of the high moisture which is
not easily removed, while gypsum would be useful.  Moreover, calcium
sulfite has some danger of consuming oxygen in ambient water.  In
discharging in slurry form to a waste pond, gypsum precipitates much
more  easily in smaller volumes than does calcium sulfite, thus reducing
the required pond size.  In case of truck transportation, gypsum can be
handled with greater ease.  By the sodium-limestone processes developed
recently by Showa Denko and also by Kureha Chemical jointly with
Kawasaki Heavy Industries, calcium sulfite grows into much larger crystals
than usual so that the above mentioned problems might be solved.3
                                  885

-------
                        4.  Other by-products
 4.1   Sulfuric  acid and  sulfur

      There  are many processes that by-produce sulfuric acid and sulfur
 (Table  5).  By these processes SO^ in waste gases is first absorbed
 with  various absorbents and then recovered as SC>2 gas of 7 to 95%
 concentration  which is used for sulfuric acid or sulfur production.
 Among these processes,  the Wellman-Lord process is well known.  In
 Japan plants based on this process were constructed respectively by
 Mitsubishi  Chemical Machinery Mfg. (MKK) and Sumitomo Chemical Engi-
 neering Co. (SCEC) in 1972.  The former produces sulfuric acid and the
 latter  returns the recovered S02 gas to a Glaus furnace to produce
 elemental sulfur.  By magnesium scrubbing processes, magnesium sulfite
 formed  by the  reaction of magnesium hydroxide and S0_ is calcined to
 recover SCL for sulfuric acid production.  Some magnesium sulfate is
 also  formed which is not as easy as magnesium sulfite to be thermally
 decomposed.  The Mitsui Mining process features the by-production of
 some  solid magnesium sulfate for fertilizer and other uses.   Demand
 for magnesium  sulfate, however, is limited.  The Onahama Smelting
 process treats waste gas from a copper smelter containing about 20,000
 ppm S02 to produce 6,600 t/month sulfuric acid.  By the Chemico process,
 a new plant to treat 294,000 scfm waste gas from a Claus furnace is to
 be constructed by 1974.  By the Sumitomo Shipbuilding process, the SC^
 absorbed on activated carbon is expelled by heating it in a reducing
 gas,  to release SC^ gas of 10-20% concentration which is used for
 sulfuric acid  production.

      By the Shell process, SC>2 is absorbed with copper oxide to form
 copper  sulfate, which is then treated with reducing gas to expel SC^-^
 A commercial plant is scheduled to come on-stream in a few months at
 Yokkaichi.  The MHI-IFP process uses ammonia scrubbing with thermal
 decomposition  of ammonium sulfite and sulfate to regenerate S0«
 (Figure 4).6

     A  commercial plant based on the MHI-IFP process will come on-stream
 in 1974.  In these two processes, the recovered SO* is reacted with ELS
 to produce elemental sulfur.

      Supply of sulfur and sulfuric acid in Japan is shown in Table 6.
 Their prices are shown in Table 7.  Demand for sulfur and sulfuric acid
 is nearly equal to the supply.  There has been little import or export
 of sulfur and sulfuric acid.  Sulfur mines in Japan are small and
 sulfuric acid has been produced mainly from pyrite and smelter gas.
 Since elemental sulfur recovered from hydrodesulfurization of heavy oil
has increased resulting in the lowering of sulfur price, the use of
 sulfur, for sulfuric acid production has been started recently.

     An annual increase in the demand for sulfuric acid is estimated at
about 5%;  some more flue gas desulfurization plants by-producing sulfuric
acid will be built but not very many of them are expected.
                               886

-------
Table 5.  Major Waste Gas Desulfurization Plants  that Produce Other By-Products
Process developer
Wellman-MKK
Wellman-MKK
Wellman-MKK
Wellman-SCEC
Wellman-SCEC
Mitsui Mining
Onahama Smelting
Chemico-Mitsui
a, Sumitomo Shipbuilding
OS
•vj
Shell
MHI-IFP
Nippon Kokan
Mitsubishi (MHI)
Kurabo Ind.
Kurabo Ind.
Absorbent
NaOH
NaOH
NaOH
NaOH
NaOH
MgO
MgO
MgO
Carbon
CuO
NH3
NH3
MnO , NH..
A. J
NH3
NH3
Product
H2S04
H2S04
H2S°4
S
H2S04
H2S°4
H2S04
S
S
S

-------
   Voter

  Ammonia
Recovery
  Tower
 Absorber


     Cooler

       	1
    Waste
     $as	
           X
 3
X
TTTK
    Make-up
oo
OS
00
                    NH3
Evaj
ratoi
c
5O-
*
T

>
X
s*
J
t
S02,
/

           Dust
                        Reactor
                                                                      1

                                                                      I
                                                ~,I.F.P    I
                                                 t reactor  I
                                                    T
                                                    I
                                                    I
                                                    I
                                                    I
                                                    r
                                                                  Fuel
                                                               After-burner
                                                             Sulfur
Hydrogen  sulflde
                                                                      I
                                                                      L.
                                                                     Foul acid  gas
   Figure
                          Fl°w sheet of M.H.I.- I. P.P.  process

-------
            Table 6.  Supply of Sulfur and Its Compounds2

                                           (1,000 tons of  material)


                                  1968     1969      1970      1971     1972

     Sulfur
       Mined                       283      180       120        47       18

       Recovered                    97      179       297       414       526

       Total                       380      354       417       461       544

     Sulfuric acid
       from pyrite               4,576    4,524     4,303     3,348     2,747

       from smelter gas          2,719    2,972     3,242     3,770     4,343

       from sulfur                   0        0        17       210       295

       Total                     7,295    7,496     7,562     7,328     7,385

                                                             2
            Table 7.  Price of Sulfur and Its Compounds ($/t)

                                  1967-     1968

     Sulfur                         73

     Sulfuric acid                  26       25
     Elemental sulfur can be produced from recovered SO, easily at oil
refineries which have H2S.  For plants which have no H S, the production
may be costly because it is required to reduce two-thirds of the SO2 into
H2S; the desulfurization cost including production of elemental sulfur may
be close to that of hydrodesulfurization of heavy oil by the topped-crude
process which ranges from $1.1 to 1.5/bl oil to reduce sulfur from 4 to
1%.  When a large oversupply of sulfur and its compounds occurs due to
the development of desulfurization, the sulfur production process would
assume greater importance because elemental sulfur has the smallest
volume among sulfur and its compounds and requires the least transpor-
tation and storage space.

4.2  Ammonium sulfate

     There were several ammonia scrubbing plants a few years ago in
Japan treating tail gas from sulfuric acid plants to by-produce ammonium
sulfate.  All of the plants have been shut down because of the oversupply
of ammonium sulfate and also because double contact processes for
                                    889

-------
 production of  the  acid has been Introduced to reduce S02 emissions.  Now
 there  are  three ammonia scrubbing plants.  Two of them produce dilute
 ammonium sulfate solution to be discarded, and the last produces solid
 ammonium sulfate (Nippon Kokan process) although in this plant ammonium
 sulfite and sulfate are being converted to gypsum at present.  In
 addition,  ammonium sulfate has been produced at Yokkaichi Station,
 Chubu  Electric Power, using the Mitsubishi manganese process.  Due to
 the worldwide increase in the demand for nitrogen fertilizers, it is
 possible that ammonium sulfate will be produced in future from S0~ in
 flue gas at a considerable number of plants in several countries.  One
 of the problems in ammonia scrubbing has been plume formation, but the
 problem has been nearly solved in Japan.  For plants that produce waste
 solution,  ammonia  scrubbing.is less expensive than sodium scrubbing
because ammonia is much cheaper than sodium hydroxide.  It is likely,
however, that the  emission of ammonium sulfate solution will be restricted
because it could cause a eutrophication problem.
                               REFERENCES
 1.   Enerugi Tokei (Energy Statistics), Ministry of International Trade
     and Industry (MITI)  Japan,  1972.

 2.   Sekiyu to Sekiyu Kagaku (Petroleum and  Petrochemistry), Vol. 16,
     No. 8, 1972.

 3.   H.W.  Elder,  F.T. Princiotta,  G.A. Hollinden,  and  S.T. Gage,
     Sulfur Oxide Control Technology, Visits  in  Japan—August  1972,
     Interagency  Technical Committee U.S.A.,  Oct.  1972.

 4.   F.  Nishimi and  Y.  Ikeda,  24th Technical  Meeting,  the Sulfuric
     Acid  Association of  Japan,  Oct. 1972.

 5.   K.  Kishi and R.F.  Bauman, Kogai Boshlsangyo (Pollution Control
     Industry), Dec.  1972.

 6.   Haiendatsuryu no subete (All  about Waste-Gas  Desulfurization),
     Jukogyo Shimbunsha,  Nov.  1972.
                                 890

-------
 LONG RANGE MARKET PROJECTIONS FOR BY-PRODUCTS
OF REGENERABLE FLUE GAS DESULFURIZATION PROCESSES
                         by

                    M. H. Farmer
             Esso Research and Engineering
                 Linden, New Jersey
                         891

-------
                            ABSTRACT
            LONG RANGE MARKET PROJECTIONS FOR BY-PRODUCTS OF
             REGENERABLE FLUE GAS DESULFURIZATION PROCESSES
          If all sulfur emitted from utility stacks, smelters, etc.
were to be recovered in useful form, the quantity would be so large
as to have a major impact on sulfur's price and markets, with the
possibility of causing severe dislocations'.  This paper desribes a
quantitative study of this proposition, using a computer model, with
projections extending to the year 2020.  The purpose was to provide
the EPA with a planning tool for establishing the relative priorities
that should be given to development of technology for recovery of
'.'abatement sulfur" in (a) marketable and (b) non-marketable forms.

          The paper describes how various abatement schedules were
simulated, and the supply/demand/price implications of these schedules.
Because sulfur is an international commodity, the domestic abatement
schedules and their implications are placed in an international
context and related to world supply/demand/price relationships.  In
addition, the relative values of sulfur recovered in acid and elemental
form are discussed.
  Note:   The work on which  this  paper is  based  was  performed  pursuant  to
         Contract No. EHSD  71-13 with The Environmental  Protection Agency,
                                   892

-------
INTRODUCTION




          In its 1970 report on "Abatement of Sulfur Oxide Emissions  from




Stationary Combustion Sources," a joint panel of the National Academy of




Engineering/National Research Council remarked that it would be desirable




to conduct a study of the long range supply and demand situation with




regard to the several alternative (sulfur) by-products to aid in establish-




ing priorities for support of control and abatement technology.




          NAE/NRC's suggestion was implemented by the EPA through Contract




No. EHSD 71-13 with the Government Research Laboratory of Esso Research




and Engineering Company.




          The full report of this study has been issued as Document No.




PB 208993 by the National Technical Information Service, Springfield, Va.




22151.




PURPOSE OF STUDY




          To provide the EPA with a planning tool for establishing the




relative priorities that should be given to development of technology for




recovery of "abatement sulfur" in marketable and non-marketable forms.




TO BE DISCUSSED




     •  How various abatement schedules were simulated.




     •  The modeling of domestic and foreign sulfur supply and demand to




        permit projections to the year 2020.




     •  The effect of stockpiling elemental sulfur.




     •  The price implications of the supply and demand projections, as




        derived by linear programming (computer) calculations.




     •  The relative values of sulfur recovered in acid and elemental forms
                                 893

-------
 SIMULATION OF ABATEMENT SCHEDULES




           Abatement refers  to removal of sulfur  so  that  it  has  not  emitted




 to the atmosphere.   The degree of  abatement,  as  treated  in  this  study,




 refers to the percentage reduction from the  total quantity  of sulfur




 oxides emitted from utility stacks in 1970.   This was  the base  year for




 the study.   The degree  of abatement does not  include sulfur removed from




 fossil fuels,  e.g.,  by  desulfurization of petroleum, prior  to combustion.




 This  is because such sulfur was  treated separately  in  the supply/demand




 calculations,  so that the effects  of flue gas desulfurization could be




 isolated.




          The  term  "useful  form"   is used to  distinguish between sulfur




 removed in  elemental form or as acid from that recovered as  a "throwaway"




 by-product.  Theoretically,  a  given degree of abatement could be achieved




 with  no sulfur  in useful  form (if  only  throwaway by-products were produced),




 or  with 1007. recovery in  useful form,  or with an infinite number of inter-




 mediate cases.   However,  what  matters  in the  computer calculations  is




 the quantity of sulfur  in useful form,  rather than  the percentage that




 this  quantity  represents  of  the total  amount  abated or of the total abate-




 ment  potential.





          The total abatement  potential  in 1970 was the amount of sulfur




 actually emitted from utility  stacks in  that year.  For the other years for




which calculations were made  (1975,  1980, 1985, 1990, 2000,  2010, 2020),




 it was  necessary to project both the amounts and sulfur contents of the




 fossil  fuels that domestic electric  utilities would burn.  This led to the
                                894

-------
abatement potential without any stack-gas treatment.  From these




quantities, it was then possible to assume varying degrees of abatement,




e.g.., so that 1970emissions would not be exceeded! and other cases in




which various percentage improvements over the 1970 level would be achieved.




Having done this, further assumptions were made about how much of the




abated sulfur would be in useful form.  These are the quantities of useful




sulfur that were included in the supply/demand calculations.




          It is important to recognize that the study did not forecast that




any particular quantity of sulfur would be abated in any future year.




What" was done was to simulate a wide range of possibilities, and then to




assess the implications of these possibilities in terms of the calculated




value for recovered sulfur in different parts of the United States.  The




whole purpose was to provide the Environmental Protection Agency with a




quantitative tool for exploring the consequences of different policies and




technological approaches for the abatement of sulfur oxide emissions.




 MODELING OF DOMESTIC AND FOREIGN SUPPLY AND DEMAND




           For convenience,  the study has been referred to as a "long range




 sulfur supply and demand model."  In practice,  several models and sub-models




 were developed,  and one of these models was "computerized" for the purpose




 of calculating sulfur values  in different locations.   Two types of model




were used:




      -  a simulation or morphological model to represent the sulfur




         industry and the structural way in which abatement sulfur




         could impact on it.
                                   895

-------
      -  a  forecasting model to project the demand/supply/price framework




        of the industry at various times into the distant future.




           In preparing the simulation model, it was necessary to take into




account not only all of the major supply/demand regions under specific




study, but also the external sources of supply and demand that could inter-




act with the former.  Thus, the starting point was the "World model"




shown in Figure 1.  The next step was to construct the "North Americal model"




shown in Figure 2.  The numbered boxes in this figure are the U.S. regions




under specific study.  The cross-hatched boxes are the extra-regional




North American suppliers to the United States.  For example, Calgary




is the center for Canadian sulfur recovered from sour natural gas,




Coatzocoalcos represents Mexican Frasch production, while Aruba represents




production of sulfur from all Caribbean refineries.  The cross-hatched




boxes with the underscoring represent sources of sulfuric acid, rather than




elemental  sulfur, that is potentially available for shipment to other




regions.   Whether such shipment of acid will occur in practice involves




economic factors, particularly transportation costs, that will be discussed




later.




          Figure 3, an exploded map of the continental U.S., shows the numbered




regions of Figure 2 in their normal geographical perspective and also the




way in which the regions coincide approximately with the Federal Power




Commission's regionalization of electricity generating capacity.




          The elements of the forecasting model are listed in Table 1.




Unlike many long range projections that may be little more tihan 10-15 year




extrapolations,  the sulfur study is concerned with the next five decades.
                               896

-------
For this reason, it was considered necessary to start by making economic




projections, and then to use correlations to convert the levels of




industrial and agricultural activity into projections of sulfur supply



and demand.  The principal steps are indicated in Table 1.  It should be




noted that a key step was to translate the projections of economic activity




into energy demand.  This is necessary because it is from fossil fuels




that the greater part of the by-product sulfur ia coming, and will



continue to come, during the next several decades.  In addition, the demand




for sulfur, whether industrial or fertilizer demand, also correlates with




economic activity and economic growth.  Consequently, by basing the overall




projections on correlations with constant dollar G.N.F., there is some




self-compensation In the net statistics.  For example, if economic develop-




ment is overestimated this would have the effect both of overestimating




the potential for sulfur recovery and of demand for sulfur.




          Separate forecasts were developed for nuclear and hydro energy




that, when subtracted from projected total energy demand, gave the fossil




fuel energy requirements.  The latter were broken down into individual




projections for  natural gas, petroleum and coal.  These quantities,  in




conjunction with data concerning  the sulfur content of these  natural




resources in different parts of the world, led to the amounts  of sulfur



potentially recoverable from each.  Separate projections were  made of the




sulfur  recoverable from smelters, pyrites and miscellaneous sources.  And,




finally, estimates were made of the future availability of elemental




sulfur  from domestic and foreign  Frasch mines.
                                    897

-------
          On  the demand side, a set of GNP/Industrial sulfur correlations




was used to project industrial sulfur demand in different areas, while




separate correlations were used to project fertilizer sulfur demand.




Worldwide, the latter accounts for about half of the total demand for




sulfur, but,  because of differences in agricultural acreage in different




countries and other factors, does not necessarily move in step with



industrial demand for sulfur.




          The sum of industrial and fertilizer demand for a particular




region* is the region's'total demand.  The difference between this quantity




and the region's internal supplies of sulfur represents either a net demand




for sulfur or the capability of the region to supply sulfur to another



region.  For  the purpose  of the computer calculations,  a net foreign




demand was located in Northern Europe (symbolized by Rotterdam) while net




supply potentials were located in Canada, Mexico and the Caribbean.




These four foreign regions, together with the eleven regions of the





continental U.S., comprised  the "North American model" used for  the




computer calculations.




EFFECT OF STOCKPILING ELEMENTAL SULFUR



          By  1970, stockpiles of  elemental sulfur  in Western Canada  had




reached several million tons and  now approximate 8 million  long  tons.  This




sulfur is coming from sour natural gas,  and must be removed before  the gas




is delivered  to a commercial pipeline.   Thus, it is the demand for natural




gas, not the  demand for sulfur, that has been a major  factor in  the  rate
if  as defined in Figures  1 and  2.





                                 B98

-------
at which this by-product sulfur has been produced.  The reason for the




stockpiles is that production has greatly exceeded the demand of the




markets to which the Western Canadian sulfur can be delivered economically.




Recently, there have been additional restraints on movements of sulfur




through the port of Vancouver and on rail deliveries to the Midwest.




These factors have accentuated the stockpiling in Alberta, and have con-




tributed to a temporary tightness in worldwide sulfur supplies.  The word




temporary is used advisedly because the transportation bottlenecks can be




relieved and, once this happens, the Canadian stockpiles will again exert




a major influence on world sulfur supplies and prices.




          Transportation costs are a major factor in the sale of Western




Canadian sulfur.  Competition takes place at the point of delivery, e.g.,




Northern Europe.  The price obtainable here minus the transportation cost




gives the net-back to Western Canadian producer.  Clearly, it is not in




the interest of the producers to move sulfur into world markets that would




net little or nothing back to Alberta.  This is why Canadian sour gas sulfur




competes in some world markets but not in others .  In turn, this is why




the Canadian stockpiles of recovered sulfur are growing.  On the other




hand, the existence of the stockpiles means that this sulfur could enter




world markets on comparatively short notice provided that the price obtainable




for it was sufficient to cover transportation costs and give a tolerable net-




back to the producer.  This statement begs the question of what is meant




by "tolerable return."  An elliptical answer to this very difficult question




is that an incremental sale may cause the average price of all sales to be




depressed thereby lowering the overall net-back.  Thus, the optimum strategy
                                    899

-------
 from  the standpoint of a Western Canadian sulfur producer may require a




 combination of stockpiling and export sales.




          The Canadian stockpiles differ in function from those of the




 Frasch sulfur producers.  At present, domestic Frasch stocks are probably




 just  below the level of 4 million long tons and have been declining slowly




 as sulfur demand has grown and various production and transportation




 bottlenecks have occurred.  In principle, however, the Frasch stocks are




 merely a working inventory.  When this inventory rises, the Frasch




 producers will tend to cut production; when inventory falls, there will be




 attempts to expand production.  As discussed already, such flexibility




 of production is not enjoyed by producers of by-product sulfur.




          What has been happening in Canada illuminates what may occur in




 the U.S. as and when air pollution controls force the recovery of sulfur




 from utility stacks.  If a surplus of sulfur from this source is to be




 avoided, some combination of the following approaches may be necessary:




      (a)  recovery of large amounts of sulfur in the form of throw-away by-




          products .




      (b)  recovery and stockpiling of elemental sulfur.




Singly, or in combination, these approaches could be applied so as to limit




 the quantity of sulfur recovered in useful form to the amount that could




 be marketed without causing major dislocations to the industry.




          There is some evidence that the application of throw-away systems




may be limited either by logistics or by the possibility of water pollution




 from the discarded by-products.  This is the reason,  why the E.P.A. asked




 for the implications of national stockpiles of elemental sulfur to be
                                   900

-------
examined.  As treated in the computer calculations,  a national stockpile




of sulfur is a source of "negative supply" since it removes sulfur from




the market.  Additionally, a national stockpile would permit the re-entry




of sulfur in later years on the assumption that the general condition




of potential oversupply will eventually be corrected.  Conceptually, this




condition could result from a continuing increase in the demand for sulfur




and from an eventual decline in the use of fossil fuels for the generation




of electricity, general industrial purposes, etc.




LINEAR PROGRAM FOR CALCULATION OF SULFUR VALUES




          The computer calculations were made with IBM's linear program




"LP-MOSS."  This is a variant of the familiar transportation linear program.




In addition, a number of specially developed programs and sub-routines were




needed to get the information in and out of the computer.  A schematic




diagram of the entire procedure is given in Figures 4.




          The linear program calculates the way that each of the net




demand regions in Figure 2 can secure the necessary supplies at lowest




cost from other domestic and foreign regions that are in a net supply position.




          The net demand in a particular region is the region's gross demand




minus whatever supplies of sulfur are available from  within the region.




The base case of net demand excludes regional abatement sulfur from the




region's internal supplies.  Other cases then consider varying  quantities




of abatement sulfur in useful form.  Such sulfur is conceived to reduce the




net demand of the pertinent region before the possibility of shipment to




another region is considered (i.e., permitted by the computer program).




The LP program also requires the following input data;
                                     901

-------
      -  Quantity of supply available  from each  extra-regional  supplier.




      -  F.O.B.  price for each source  of  extra-regional supply.




      -  Transportation costs  for  each linkage between the net  demand




         regions  and the extra-regional supply points.




           In  addition,  certain special feastures were incorporated:




      -  Upper bounds  on the amount of acid  that may be shipped into a




         given region (in order to simulate  the capacity of the region




         to accept merchant acid).  The upper bounds were raised with




         time  in  order  to simulate an  increasing ability to accept sulfur




         in the form of  acid.




      -  An acid  equivalency credit, recognizing that the cost of manufactur-




         ing acid is  avoided by a customer who purchases acid rather than




        elemental sulfur.




        Lower bounds on certain supply likages in order to simulate the




        effects  of marketing strengths (e.g., captive terminals in the




        net demand  regions) and captive use of sulfur by the suppliers (e.g.,




        the manufacture  of P fertilizers by the producers of Frasch sulfur).




          The output data from the LP program include a matrix of extra-




regional suppliers and  net demand regions which shows how demand was




filled (i.e., who sold  to whom), and what the delivered cost was.  If no




sales were made by a particular supplier to a particular demand region,




the required reduction  in minimum delivered cost (and,  hence, in F.O.B.




value) for a sale to occur is printed in the matrix.  Other output programs




tabulate the sales by and netback to each supplier,  and the calculated




delivered value of S in each region.
                                  902

-------
          Each LF case calculates the value of an incremental unit of




supply rather than of supply increments of any size.  However, parametric



cases involving substantially different quantities of abatement S. supply




make it possible to obtain value estimates by interpolation.   The calculated




values are considered to be maxima because new suppliers would have to




shave their price in order to "buy into" an existing market.   These maxima




are referred to later as "maximum delivered values" (M.D.V.).  AS a first



approximation, it is postulated that a recoverer of abatement S in




elemental form might expect a plant netback of about $10/LT less than  the




M.D.V.  The $10/LT represents a composite of transportation and marketing




costs (including price-shaving to buy into the market).  Actual transporta-




tion and marketing costs are likely to vary greatly from one specific




situation to another.  Thus, the deduction of $10/LT from M.D.V. to obtain




the F.O.B. value (F.O.B.V.) of abatement S should be used with caution.




PRICE IMPLICATIONS OF SUPPLY/DEMAND PROJECTIONS




          The computer calculations permitted the construction of correla-




tion charts for the individual regions of the continental U.S.  Examples




are given in Figures 5 and 6.  Each chart indicates:




     (a)  The variation in net regional demand with increasing regional




          supply of abatement sulfur in useful form.



     (b)  The maximum delivered value  (M.D.V.) of abatement sulfur




          corresponding to a given level of supply in useful form.




In addition, a rough estimate of the F.O.B. value (F.O.B.V.) of abatement



S in useful form is shown on the charts.  In most cases, F.O.B.V. is




M.D.V.-10 ($/LT), i.e., it is merely the M.D.V. minus an arbitrary delivery
                                    903

-------
 cost of $10/LT.   In a few cases,  the F.O.B.V.  refers  to  net-back  esti-




 mated for abatement S shipped  to  another region.   There  are also  cases




 in which it was  not possible  to estimate either an M.D.V. or F.O.B.V.





          No charts were prepared for the year 2020 because the results




 obtained, while similar to those  for 2010, are judged to be too sensitive




 to  the assumptions made.  Each  figure is a composite of six charts




 covering  the years  1975 through 2010.  All of the charts have certain



 features  in common:




      •  Regional abatement 8 in useful form is plotted as an abscissa.




      •  Net regional demand is plotted as an ordinate, with the scale at



        the left.




      •  M.D.V. and F.O.B.V. are plotted as ordinates,  with a common scale




        at the right.




      •  The numbers 0, +5, +10 and +15 refer to parametric demand (as




        discussed in the complete report, but not here).




      •  All volumes are expressed in million LT of S equivalent.




      •  All values are expressed in $ per LT of S equivalent.




Chicago Region (Figure 5)




          This region has  a significant demand for sulfur but it has an even




greater potential for producing abatement S.  Markets  for all of this




potential cannot be conceived.  Hence, all of the M.D.V. and F.O.B.V.




curves have sharp downturns.  A calculated F.O.B.V. curve is shown for 1990,




based upon the assumption that Chicago region sulfur will be moving to




extra-regional markets by  then.  In retrospect, the contrast between the
                                 904

-------
1985 and 1990 charts suggests that a turn around may not come so quickly.

However, a national stockpile in the Chicago region could be a spur to

exports and could support the conditions projected in the 1990 chart.

Tampa Region (Figure 6)

           The  net demand in  the Tampa  region greatly exceeds  its abatement

potential.   But most of  the  demand  involves captive production  of acid

from purchased elemental S.  The  charts  in Figure  6 suggest  that reason-

able F.O.B.V.'s may be possible for local deliveries of  acid  in the

Tampa/Bartow area.*  The downtrend  in  M.D.V. for  the Tampa  region with

increasing supplies of abatement  S  in  useful form is due to  the general

level  of  recovery in the U.S. not to the impact of the Tampa region's  own

abatement  potential.

Implications

          The years 1980 and 2000 may be used to illustrate the results of

the computer calculations.  The choice permits significantly different

situations to be contrasted.  In  1980,  production of W. Canadian sulfur is

expected to be at peak levels and backs topped by huge stockpiles, world

markets will be under intense supply pressure,  and the fitting  (and

retrofitting) of abatement systems in the U.S. should be.beyond the

demonstration stage and into widespread use.  In contrast, two decades

later,  W. Canadian S is expected  to be a minor factor in world markets,

S recovered from petroleum refining will be of major importance while, in

the U.S., the policies applied to abatement S in prior years will have

determined the shape of the domestic sulfur and sulfuric acid industries

and their relationship to world markets.
*  Direct negotiations between potential producers of abatement acid acid
   users of fertilizer acid would seem to be required.

                                    905

-------
           Results that illustrate the conditions  projected  for  the years
 1980 and 2000 are given in Tables 2 and  3.   The overall  sulfur  demand
 of the continental U.S. is summarized in the top  segment of each  table.
 This is followed by estimates  of demand  in  two of the most  important
 regions of the Model,  Chicago  and Tampa.  In both cases,  sulfur supplies
 from within the region are deducted to give the net  regional demand.
 These sulfur supplies  exclude  abatement  S in useful  form.
           The third segment of each table shows cases that  illustrate
 the  effects  of different  levels  of abatement combined with  different
 percentage recoveries  in  useful  form and, hence,  different  quantities of
 abatement  S  in useful  form.
           The fourth segment of  each table  shows  the net  regional demand
 for  the Chicago and  Tampa regions,  after subtracting the pertinent quanti-
 ties  of abatement  S  in  useful  form from  the  base  case of net demand.  The
 reason that  the net  demand  in  cases  (B) and  (C) is the same  as  in case (A)
 is that the  two former  cases make  the  assumption  of no recovery of S in
 useful form.   Even though cases  (A),  (B) and  (C)  represent different levels
 of abatement,  they are  identical  in  terms of  the  assumption  that no useful
 sulfur is  recovered.  This implies different  levels of recovery in throw-
 away  form.  However, the  latter has  no direct effect on the value estimated
 for abatement  sulfur.
          The  fifth segment of each  table shows the estimated value of
 abatement S corresponding  to each quantity of abatement S in useful form.
Here,  it must  be pointed out that, for both years, cases (A), (B)  and (C)
 lie outside the area actually investigated by computer cases.  This is
because it is considered an unrealistic assumption that no abatement S
at all will be recovered in useful form.   With no  useful abatement S in
                                  906

-------
 the year 2000,  the M.D.V.'s  of  $31+ and  $30+ (obtained  by extrapolation  from
 calculated cases)  might be somewhat higher.   Additionally,  some  switching
 to processes  not requiring sulfur  would  be expected  since cases  (A),  (B)
 and (C)  correspond to a net  deficit in supplies  for  the continental U.S.
           It  may seem somewhat  of  a contradiction to have an M.D.V. for
 abatement S if  none is  assumed  to  be recovered in useful form, as  in  cases
 (A),  (B)  and  (C).   However,  the M.D.V. applies,  conceptually,  to the  first
 units  of abatement S that  would be recovered in  useful  form.
           For 1980,  case (D)  suggests  an M.D.V.  only slightly  less than  for
 cases  (A)  - (C)  in both regions.   However, a marked  difference should be
 noted  for  case  (E).   Here, a marginally  lower value  is  estimated for  the
 Tampa  region  while an  indeterminately low   value is estimated for the
 Chicago  region.  The explanation is  that abatement sulfur has hardly  changed
 the net demand of  the Tampa  region whereas,  in case  (E), it has  produced
 a  condition of gross  oversupply in the Chicago region,  namely a  net demand
 of minus 0.14 million LT.  Because of general conditions of oversupply
 corresponding to case 1980 E, no extra-regional  outlet  is envisaged for  the
 Chicago region's surplus of  0.14 million LT.
           The drop off  in estimated  M.D.V.'s  for the Tampa  region is  attri-
 butable not to the small quantities  of abatement S in useful form assumed
 to be  produced within the region but to  the  corresponding quantities  produced
 in other regions.
          For the year  2000,   it will be  seen  that  both  cases (D) and  (E)
represent a surplus of  supply over demand in  the Chicago region.  In spite
of this,  significant M.D.V.'s are estimated  for  both  cases.  The explanation
is that,  in contrast to the  1980 cases,  extra-regional  outlets for Chicago
                                  907

-------
 region sulfur are anticipated,  i.e.,  the Chicago region is  expected  to be
 an "extra-regional supplier" by the year 2000.   Nevertheless,  if the
 percentage of abatement S recovered in useful form were to  be  only slightly
 greater than in case (E)— e.g.,  60% instead  of  50% in useful  form — then
 the Chicago region would become grossly oversupplied  and  the estimated
 value of abatement S in useful  form would be indeterminately  low.
 RELATIVE VALUES OF SULFUR RECOVERED IN ACID  AND  ELEMENTAL FORMS
           U.S.  consumption of sulfur  is mostly  in the form  of  acid.   Thus,
 recovered sulfur has a  different  value depending on whether it is  recovered
 as  acid or as elemental sulfur.   Theoretically, acid has the higher value
 because it means that the customer  does not  have to bear  the cost  of manu-
 facturing it.   However,  the  cost  of transporting acid is  at least  three times
 agreater than for elemental  S per ton  of S-value.   In consequence, the
 greater value of recovering  acid  rather than elemental  sulfur  is critically
 dependent on assured, local  markets for all  of the  acid produced.  Unless
 these  conditions are fulfilled, the value of  acid may be  low and may  even
 be  negative.  Elemental  sulfur can  be  stockpiled; in  most cases at small
 cost.   Sulfuric  acid can not be stockpiled.
          At the end of  1965, just over  half  of U.S.  sulfuric  acid capacity
was in 6  states:  Florida, Texas, New Jersey, Illinois, California and
Louisiana.  Today, the industry is even more  concentrated on the Gulf
Coast  and  in Florida, with a relative  loss of capacity in the Midwest.  The
geographical trends  reflect  the importance of Florida's pebble phosphate
deposits and the development of chemical industry on the Gulf Coast.
          Most of the acid used to manufacture P fertilizers is produced
captively.  This is true of a significant percentage of industrial acid
                                  908

-------
as well.  Some of the acid manufacturers  also have  captive  production of




S values.  This applies not only to the special case of acid  recovered




by smelters but also to combinations of:




     •  Frasch S production and P fertilizer manufacture.




     •  S recovery from oil-and-gas operations and  F fertilizer




        manufacture.



     •  S recovery by chemical companies  and manufacture of industrial




        acid.




          The structures and geography of the elemental sulfur and acid




industries will make it difficult for abatement acid to enter the market.




The willingness of existing acid marketers and captive users to offtake




abatement acid is necessary if a significant outlet is to be developed.




The incentives for such offtake have not been clearly established.  Currently




the acid manufacturers, particularly those who merchant industrial acid,




stand to benefit if abatement S were to enter the market in elemental



form but to lose if entry were to be as acid.  On the other hand, a




significant amount of old acid plant capacity is being shut down and this




may provide opportunities for abatement acid  to enter  the market.




          The acid equivalency credit, discussed in  the computer program,




 is a device making  it possible to deal with  both acid and elemental sulfur




in the same calculation.  The credit was changed with  time to  reflect  the



difference between:




          (a)  having  to  shut down existing acid manufacturing capacity




               in order to be able  to purchase merchant acid




   and    (b)  purchasing acid incrementally  rather  than building new  acid




               manufacturing capacity.




                                   909

-------
          The structure of the U.S. sulfuric acid industry suggests that,




near term, it will be essential for many power stations to recover or




remove sulfur in forms other than acid.  The market for S02 is small,




and that for ammonium sulfate is declining.  Hence, apart from using low




sulfur fuels (if available), the only broadly applicable choices appear




to be elemental sulfur and waste gypsum.  This is not to say that acid




recovery systems will not be useful, but it does say that the larger




part of the current problem will require another solution.
                                   910

-------
                                           TABLE 1
                                       FORECAST BASIS
  •  Supply
  (1)  PROJECTIONS OF I        (2)  CONSTANT $ GNP
                               (3)  POPULATION
  (2)  CNP/ENERGY CORRELATION + (2) —*»(5) ENERGY FORECAST
  (6)  NUCLEAR AND HYDRO ENERGY FORECASTS
  (5)  - (6)  —» (7)  FOSSIL FUEL ENERGY FORECAST
  (8)  BREAKDOWN OF FOSSIL FUEL ENERGY BY SOURCE
  (7)  & (8)  —M 9)  NATURAL GAS FORECAST
             —V(10)  PETROLEUM FORECAST
             —+(11)  COAL FORECAST
 (12)  NATURAL GAS SULFUR CONTENT  & (9)—*(13)  NAT. GAS S RECOVERY POTENTIAL
 (14)  PETROLEUM SULFUR CONTENT  & (10) —»(15)  PETROLEUM S RECOVERY POTENTIAL
 (16)  COAL SULFUR CONTENT  & (11) —*(17)  COAL S RECOVERY POTENTIAL
 (18)  % RECOVERY IN USEFUL FORM FROM NATURAL GAS
 (19)  7. RECOVERY IN USEFUL FORM FROM PETROLEUM
 (20)  7. RECOVERY IN USEFUL FORM FROM COAL
 (13)  & (18) —+(21)  S RECOVERED FROM NATURAL GAS
 (15)  & (19) --*(22)  S RECOVERED FROM PETROLEUM
 (17)  & (20) —*(23)  S RECOVERED FROM COAL
 (24)  PROJECTION OF S RECOVERY FROM SMELTERS, PYRITE AND OTHER SOURCES
 (25)  PROJECTION OF FRASCH S AVAILABILITY
 (21)  & (22) & (24) & (25) —*(26)  TOTAL S SUPPLY (EXCLUDING STOCKPILES)

 •  Demand
 (27)  POPULATION  &  GNP PER  CAPITA/FERTILIZER S  DEMAND/RELATIONSHIPS
  (2)   & (3)  &  (27) —» (28)   FERTILIZER S DEMAND
 (29)  GNP/INDUSTRIAL S DEMAND CORRELATION
  (2)  &  (29) "* (30)  INDUSTRIAL  S DEMAND
 (28)  + (30) —*  (31)  TOTAL  S DEMAND
 •  Supply/Demand  Balance

(31)   - (21)  - (22) - (23) - (24)  —* (32)  NET FOREIGN DEMAND, INCLUDING ABATEMENT S BUT
                                           EXCLUDING FRASCH SUPPLY
(31)   -  (21)  - (22)        - (24)  ---%(33)  U.S. NET REGIONAL DEMAND, EXCLUDING FRASCH
                                           SUPPLY AND ABATEMENT S

        Note:  The symbol  &  signifies the joint consideration of one factor
               with  another, not  a simple arithmetic sum of the  factors.
                                         911

-------
                                 TABLE 2
              ILLUSTRATION OF U.S. SULFUR SITUATION IN 1980
•  Continental U.S.
          Fertilizer S Demand
          Industrial S Demand
               Total S Demand
Million LT

    7.1
    6.7
   13.8
•  Demand in Representative Regions

          Gross Demand
          Regional Supply (excl. Abate. S)
     (A)  Net Regional Demand
            Chicago

              1.66
              0.59
              1.07
TampA
 3.19
 0.10
 3.09
   Illustrative Cases of Abatement Supply
Total SOX
Emitted
(B) Same as 1970
(C) 40% Less Than 1970
(D) Same as 1970
(E) 40% Less Than 1970
• Net Regional Demand After
(B), (C)
(D)
(E)
% of Abate. S
In Useful Form
None
None
50
50
Inclusion of Abate.

• Estimated Maximum Delivered Value (M.D.V.) for
(A), (B), (C)
(D)
(E)
In de terrain
Quantity of Abate. S
In Useful Form (10<)
Nil
Nil
0.49
1.21
S in Useful Form
1.07
0.58
-0.14
Abate. S ($/LT)
26
25
ately Low
Nil
Nil
0.02
0.06
(106 LT)
3.09
3.07
3.03

23
22+
21
      Note:   Estimates of Value are in 1970 constant dollars.
                                   912

-------
                                 TABLE 3
              ILLUSTRATION OF U.S. SULFUR SITUATION IN 2000
•  Continental U.S.
          Fertilizer S Demand
          Industrial S Demand
               Total s Demand
                               Million LT

                                  11.0
                                  14.2
                                  25.2
•  Demand in Representative Regions (Million LT)
     (A)
Gross Demand
Regional Supply (excl. Abate. S)
Net Regional Demand
Chicago

  3.53
  2.92
  0.61
                                                                      Tampa
4.20
0.30
3.90
•  Illustrative Cases of Abatement Supply
         Total SO,
          Emitted
(B)  Same as 1970
(C)  40% Less than 1970
(D)  Same as 1970
(E)  40% Less than 1970
                      % of Abate. S
                     In Useful Form

                          None
                          None
                           50
                           50
 Quantity of Abate S
In Useful Form (106LT)
   Nil
   Nil
  2.35
  3.50
 Nil
 Nil
0.13
0.20
   Net Regional Demand After Inclusion of Abate. S in Useful Form  (106  LT)
(B), (C) Same as Base Case (A)
(D)
(E)
                                             0.61
                                             -1.74
                                             -2.89
                  3.90
                  3.77
                  3.60
   Estimated Maximum Delivered Value (M.D.V.) for Abate. S  ($/LT)
(A), (B), (C)
(D)
(E)
                                              31+
                                              26
                                              21
                   30+
                   29
                   28
       Note:  Estimates of Value are in 1970 constant dollars.
                                  913

-------
                                Figure 1
                            WORLD MODEL
y////////
 UAPAN/
tCOMMUNIST ASIAE
    FAR  EAST
 EXCLUDING JAPAN
                              CANADA:
^UNITED STATES
                                LATIN
                               AMERICA
EUROPEAN O.E.C. D/
        :u. s. s. R.:
                                                      EASTERN EUROPE^
      MIDDLE EAST
                                              AFRICA
 LEGEND
              ORGANIZATION FOR ECONOMIC CO-OPERATION AND DEVELOPMENT

              COMMUNIST COUNTRIES

              "DEVELOPING FREE WORLD"

-------
                                  Figure 2
VI
       (7)
    SEATTLE
         (8)
 SAN FRANCISCO
    LOS ANGELES
        (8)
                          NORTH  AMERICAN  MODEL
                           (fi]
                         OMAHA

                                                              (1)
                                                            BOSTON
                                      (4)
                                   CHICAGO   (DETROIT)
                                       (5)
MEMPHIS'
                             NEW ORLEANS/
                                                              (2)
                                                   (BUFFALO) NEWARK
                                                 TAMPA
                                                              (3)
                                                           NORFOLK
                                                          ROTTERDAM
                        /
                      COATZACOALCOS
                                       /
LEGEND
  (/)-(II) DESIGNATE U.S. REGIONS
  #     SOURCE OF "EXTRA-REGIONAL SUPPLY"
        IN LATER YEARS OF FORECAST
  (    ) SUB-REGIONS
                                                      EXTRA-REGIONAL
                                                      ACID SUPPLIER

-------
                                         Figure 3

                                      U.S. REGIONS
vo
                                 	)
                   \
   VTRO.

EAST NORTH

  CENTRAL OHI
WEST NORTH

 CENTRALS  __
~]uTAH"~f

I



/ MOUNTAIN
       i

-------
       Figure 4
COMPUTER CALCULATION
FT,
l_
^
> INPUT PROGRAMS
SUPPLY
FILE
1


PROGRAM: RRB
i
f SUPPLY
J CARDS
1
L.








16
— ii
•^
DEMAND
FILE




PROGRAM: RRB is
i
C DEMAND
]r CARDS







' CALCULATION PROGRAM
I
1—



^^

"III


JT

TRANSPORTATION
COST FILE






.



FOB
COST FILE

— h-

PROGRAM: RRB 15




r ACTIVITY
CARDS
1
JU
PROGRAM
L














f1



I^LP/CONTROL
| CARDS

1
III
:LP-MOSS

P SOLUTION
CARDS


LP OUTPUT OPERATIONAL
PROGRAMS
\


P


ROGRAM:RRBOT
i
LP OUTPUT
MATRIX






J











1
	 1
0
]




SALES
FILE
i











.








«
.

VALUE
FILE


1
ACID
EQUIVALENCY
VALUE FILE

1
("" COST
J CARDS
1
1





I

1
J-

1

1
1

NET- BACK
FILE
1
1


-------
•£>
 O
 UJ
 c;
 o
 o
 o
 u.
 o
 LJ
 (3
                              Figure 5

                           CHICAGO REGION
     1,4
     1.2
     1.0
     0.8
     0.6
     0.4
     0.2
     NIL
     -0.2
     -0.4
     -0.6
     -0.8
     -1.0
     -1.2
     -1.4
  2

  1

NIL

 -1

 -2

 -3

 -4
             (1975)
                            (1950)
   M.D.V.       _-_
• — — _        "" """"*•»»
      ^ "v      M.D.V.  N
 ^F.O.B.V.  \	^    ^
   "^"""^      F.O.B.V.N
             *
                    1     (I
               T    I
                   C1985)
                                 I    I
                            (1990)
                F.O.B.V.
                                                     25
                                                     20
                                                     15
                                                     10
                                                     5
                                                     0
           0   0.2  0.4  0.6  0.8    0.5  1.0  1.5  2-0  2.5
                                           I
                                        25
                                        20
                                        15
                                        10
                                        5
                                        0
                                                               _)
                                                               >
       2

     NIL

      -2

      -4

      -6

      -8

     -10
   +10
   +5J
   -0
    F.O.B.V.
  \     \

12000)   +15
        +10
          0
                       riii
                            (2010)
                                          30
                                          25
                                          20
                                          15
                                          10
                                          5
                                          0
                                                  8    10
               REGIONAL ABATEMENT SULFUR RECOVERED
                      IN USEFUL FORM (106 LT>
                                  918

-------
sO
 o
 o
 LU
 Q.
                           Figure 6

                        TAMPA REGION
5  o
        I    I


       (1975)
                                     1   I
                  (1980)
                     (1985)
                                           (1990)
      +5

F.O.B.V.
                       1     1    1     1   1
                                    1
                                                 1
        0   0.1   0    0.1   0   0.1  0.2
                                   0.1  0.2
 U_
 O
UJ
o


§  5
LL


S  4

f-
LU
2  3
    0
                   I     T
            (2000)
                             (2010)
      -  M.D.V.
      -  F.O.B.V.
                             M.D.V.
                             F.O.B.V.
                  1
                               1
            0.2  0.4
                                             1
                   0  0.2  0.4   0.6  0.8
           REGIONAL ABATEMENT SULFUR RECOVERED
                  IN USEFUL FORM (106 LT)
                                                       25

                                                       20

                                                       15

                                                       10

                                                       5
                                                       0   P
                                                  LJ
                                               35

                                               30

                                               25

                                               20

                                               15

                                               10

                                               5

                                               0
                             919

-------
         NEW USES FOR  SULFUR-
      THEIR STATUS AND PROSPECTS
                   by

H. L. Fike,  Director of Industrial Research
   J. S. Platou, Director of Information
          The Sulphur Institute
          1725  K Street, N. W.
           Washington, D. C.
                   921

-------
          Hew Uses for Sulfur - Their Status and Prospects

                     H. L. Fike and J^v S^ JPlatou2

 Introduction

 Prom statements and forecasts on the energy needs of our nation
 and the means available to fill these needs, it appears certain that
 high-sulfur coal will continue to pla^ an important part in the
 foreseeable future.  The Air Quality Criteria which are now being
 enacted and which are due to be fully implemented by 1977 make it
 equally certain that all of the sulfur content of this coal will riot
 be  allowed to escape into the atmosphere.  Thus, it appears highly-
 probable that large tonnages of sulfur will be recovered in one way
 or  another from high-sulfur coal.

 Two approaches are presently considered possible to prevent sulfur
 oxide emissions from coal:  stack gas scrubbing and coal gasification.

 The majority of the stack gas scrubber technologies near commerciali-
 zation recover the sulfur values as gypsum or sulfuric acid.  While
 these are acceptable intermediate term solutions, the problems
 associated with spent gypsum disposal and the general transportation
 and marketing problems with sulfuric acid make them unattractive and
 even unacceptable over the long run.  Recovery as elemental sulfur
 appears to be the ultimate solution, from the point of view of
 marketing, storage, and transport of the product.  Also, the processes
 under development for coal gasification generally recover sulfur in
 the elemental form.  The following discussion is therefore restricted
 to  the question of finding a market for the elemental sulfur recovered
 from coal as a result of air pollution abatement regulations.

 The Magnitude of the Problem

 Let us first put the recovered sulfur in perspective in relation to
 tonnage, time, and cost.  It is unlikely that large tonnages of sulfur
 from stack gas scrubbers or coal gasification will reach the market
 before the 1980's.  The first demonstration plants for recovery of
 elemental sulfur have recently been completed or are still under.
 construction.  It is likely that they will have to operate successfully
 for a year or two before the technology achieves general commercial
 acceptance.  This will then be followed by an estimated three-year
 period between placing the order and full operation.  However, by the
 mid-eighties it is quite conceivable that 500 million tens of high-
 sulfur coal per year will be consumed in the U.S. in power ntations
 and/or coal gasification plants equipped with scrubbers recovering
 sulfur.  Assuming an average sulfur content of 3$, about 15 mill?'-::..!
 tons sulfur per year could theoretically be recovered.

The disposal of this amount of sulfur becomes of considerable economic
 importance.  If it can be sold at a price of $25/tcri f .o.b. the
 Paper presented at the Flue Gas Desulfurization Symposium, New Orleans,
 Louisiana, May lk-17, 1973.
o
 Directors of Industrial Research and Information, respectively, She
 Sulphur Institute, 1J25 K Street, N.W., Washington, D.C.  2COOu.
                                  922

-------
 power plant*, the U.S. utilities industry could realize an annual
 income of about $375 million, to set off against the capital and
 operating costs of desulfurization equipment.  Conversely, if no market
 exists for this sulfur, the storage of large quantities in heavily
 populated areas may cause its own pollution problems.  Furthermore,
 the  costs associated with storage of sulfur could amount to $2-$5 par
 ton  per year—representing an additional cost to the industry.

 The  Sulfur Market

 What are the possibilities of marketing the large quantities of
 sulfur which will be recovered from stack gases or coal gasification?
 To answer this, let us look at the present and probable future situation.

 In 1972, Free World production of elemental sulfur reached 21 million
 tons.  This exceeded consumption by over 1 million tons.  In addition,
 nearly 2 million tons were imported from communist countries and,
 as a result, sulfur inventories increased by over 3 million tons.

 Over half of the sulfur presently produced is recovered from natural
 gas  or petroleum.  The demand for natural gas and low-sulfur petroleum
 products will accelerate the recovery of sulfur from these sources.

 The  traditional uses of sulfur, that is, in the fertilizer, chemical,
 and  other industries, are not expected to grow at an average of more
 than U-55W annually.  Most forecasts agree that production of recovered
 sulfur will grow at rates considerably in excess of this.  Thus, even
 without any sulfur from coal, the oversupply situation can be expected
 to continue in the 1980*3.  If recovery from stack gas or coal gasifi-
 cation reaches the estimated levels, sulfur production in North America
 in the mid-1980's would outstrip consumption by a substantial margin.

 Under these circumstances, the majority of utilities would be unable
 to sell substantial quantities of their recovered sulfur in the market.

 New  Uses

 There are new, potentially high-tonnage uses for sulfur which,
 if developed commercially, can absorb the sulfur that will have to be
 produced if the current air quality standards are to be met.  In the
 remainder of this paper, we will discuss with you the more promising
 of these new uses.

 A word here about the agricultural uses of sulfur.  Sulfur is essential
 in plant and animal nutrition, and is valuable as a soil amendment and
 pesticide.  The Sulphur Institute is working actively with agricultural
 scientists and the fertilizer industry to promote and develop the uses
 of sulfur in agriculture.  The potential market for these uses in the
U.S. is estimated at between 1.5 and 2 million tons per year above the
 present use, but as this is, strictly speaking, not a "new" use but an
 expansion of an existing one, we will not discuss it further in  this paper.
*This figure is given for illustrative purposes only and does not imply
 a recommendation or forecast of the sulfur selling price.
                                 923

-------
 To alleviate  the supply/demand imbalance and to "be commercially
 acceptable, a new use for sulfur should ideally satisfy the following
 criteria:
      1.  The  potential tonnage use, realistically assessed, must be
         large, at a minimum several hundred thousand tons per year.
      2.  The  time and cost needed to develop the use must be reasonable.
      3-  The  economics should be favorable, to attract industry
         interest and capital.
      U.  The  new use must be ecologically acceptable, i.e., it must
         not  create any pollution problems of its own.

 The Sulphur Institute has carefully considered a great number of possible
 new uses for  sulfur and its compounds, including  sulfuric acid and
 sulfur  dioxide.  Those uses which in one way or another utilize the
 mechanical properties of elemental  sulfur appear to have merits which
 match these criteria most closely.  Our discussion here will be limited
 to these uses.

 Pure  elemental sulfur is familiar to most people as a rather brittle,
 yellow solid.  However, it has a number of interesting mechanical
 properties, not all of which are, as yet, fully understood.  These
 properties are dependent upon the time-temperature history of the
 sulfur and can be greatly modified by additives.  Conversely, when
 sulfur is added to or incorporated into other structural materials, the
 mechanical properties of these materials are often improved.

 The following examples will demonstrate some of the commercially
 interesting utilizations of these properties.

 Sulfur-Asphalt Paving Materials

 Processes for improving the properties of asphalts by treatment with
 sulfur were first introduced over 100 years ago, but have generally
 not become commercially successful.  Most of this early work had as
 its objective to chemically react sulfur with asphalt.  The new
 technology described below relates to the physical addition of sulfur
 to  asphalt-aggregate mixtures.  Chemical reaction between the sulfur
 and asphalt is deliberately minimized to prevent the evolution of
    which generally accompanies the reaction of sulfur with hydrocarbons.
The composition of sulfur-asphalt paving materials varies considerably,
but a typical material consists of about 13$ sulfur, 6% asphalt, and
81$ sand, by weight.  This material is interesting for two reasons:
     1.  It has certain technical advantages over conventional paving
         materials.  In general, addition of sulfur to asphalt paving
         material can increase durability, strength, and resistance to
         water.  Also, there are indications that the material may be
         particularly suitable for direct application on weak subgrades,
         without prior excavation and backfilling.
     2.  In many areas of the country (Gulf Coast region, Pacific
         Northwest, Great Lakes region) good quality aggregate is
         becoming scarce and increasingly expensive.  Sulfur-asphalt
         permits the use of low-cost, widely available sand instead
         of aggregate.
                                  924

-------
 The composition and coat of sulfur-asphalt paving materials  compares
 with those of conventional asphalt paving materials approximately
 aa follows:

                                   Ibs/cu ft  % by wt  ^ "by vol  cost  (£/cu  ft
 Conventional asphalt material*
   Asphalt                              8         6        13       14
   Aggregate, crushed & screened      135        94        81       27
   Air voids                           —        --      	6       zz.

   Total                              ilj.3       100      100       kL

 Sulfur-asphalt material*
   Asphalt                              8         6        13       14
   Sulfur                              17        13        iU       21
   Sand                               105        8l        $5         2,5
   Air voids                           —        —      	8       —

                                      130       100      100       3T-5
 ^Assuming asphalt at $35/ton,  aggregate at $4/ton,  sand at $0.50/ton,
  and sulfur at $25/ton.

 There are indications that it  may be  feasible  to use  lower grade  asphalts
 in the sulfur-asphalt mix.  This  would further improve  the economics
 of this material.

 Approximately  22  million tons  asphalt are  used each year in the United
 States for road building.   For the non-communist world  as a whole, the
 corresponding  figure is  about  45  million tons.   Any sizable penetration
 of sulfur-asphalt into this market clearly could consume multi-million
 tonnages  of sulfur.

 Further work is needed in  several areas before  sulfur-asphalt paving
 materials  can  become a commercial reality.  Although HXS formation
 during manufacturing and transport does not appear  to be a problem, it
 must be studied further  before the technology  is released for general
 use.   Durability  comparisons of sulfur-asphalt  with conventional materials
 require considerable time,  and the introduction of  new paving materials
 into general use  is  often  a slow  process.

 The  most extensive work  on sulfur-asphalt  paving materials has been
 carried out  by Shell Canada, Ltd., in  British Columbia and Ontario.
 Trials  have  also  taken place in Prance,  Japan,  and  Scandinavia.  The
 Sulphur Institute has  recently concluded an agreement with Shell whereby
 the  development of sulfur-asphalt pavements in  the  U.S. will be carried
 out  under a  Joint government-private industry program.  Initial develop-
 ment vork is now  under way  at  the Texas Transportation Institute,
 Texas A & M University.  Following preliminary  familiarization vork
 this ; ear, road tests  are scheduled for 1974.

The sulfur-asphalt paving materials so far developed do not result in
 a decreased consumption of asphalt per unit of  road.  The existing
 and projected shortage of domestically produced petroleum may make it
necessary to limit the consumption of petroleum products, including
                                  925

-------
 asphalt,  where satisfactory  substitutes  exist  or can "be  developed.  As
 large quantities  of sulfur become  available  from increased use of
 doaestic  coal., a  partial  substitution  of sulfur  for asphalt would seem
 to make sense both from the  national economy and ecology points  of  view.

 There are indications that satisfactory  sulfur-asphalt materials containing
 less asphalt may  be possible.   If  the  asphalt  is eliminated completely,
 the result is a sulfur-sand  concrete,  and we will discuss  this next.

 Sulfur Concretes

 Mixes of  molten sulfur  with  sand and/or  aggregate set on. cooling to
 form concretes with properties  equal or  superior to those  of Portland
 cement-aggregate  concretes.  Sulfur concretes  set and reach full
 strength  in a matter of minutes, as opposed  to one or more days  for
 the conventional  materials.  There is  also the advantage that the large
 quantities of water required for conventional  concretes  are not  needed.

 Preparation of the mix  is simple:  the sand  or aggregate (70$ by weight)
 is pre-heated to  about  325°  P and  sulfur (30$) is added.   Mixing is
 continued until all the sulfur  is  melted and the mix is  then poured,
 tamped, and left  to harden.

 Using this technique, tiles, blocks, bricks, and other structural elements
 have been produced.  In 1972, part of  an experimental house was  built
 near Montreal by  McGill University staff, using  interlocking sulfur-
 concrete  bricks,  without  mortar or other joining material. Reports
 indicate  that the walls built using this technique are structurally
 sound and weatherproof.

 Sulfur concretes  may be particularly attractive  in countries with no
 indigenous supply of Portland cement,  but with a readily available  supply
 of sand and sulfur.  Northern Canada,  western  U»3., and  several Middle
 Eastern countries fall  into  this category.   Sulfur concretes may also
 have application  in remote areas where the cost  of casting and curing
 Portland  cement concretes is excessive.   Another potential use is as an
 emergency patching material  for highways.

 Sulfur concretes  are at an early stage of development.   Their eventual
 acceptance and the size of the  market  are matters for speculation.
 However,  the potential  market is very  large  if the technique is
 basically acceptable in housing and road construction.   For example,
 building  a 2-lane highway suitable for heavy traffic would require
 about 1,800 tons  sulfur per mile.

 Surface Bond Construction

 Low-cost  easily built structures -can be- erected l>y -aurface-bonding
 concrete  blocks with a  sulfur formulation.   This technique has
 interesting potential applications in  low-cost housing,  storage
 buildings,  silos,  etc.

The  sulfur-surface bond construction technique-ia simple both in
 concept and practice.   It is applicable  to structures utilizing
 lightweight concrete blocks- or  other- blocJt or -block-like materials.
                               976

-------
No mortar or Joining material is used between the blocks;  they are
simply stacked dry, one upon the other, until the desired wall con-
figuration is achieved.  A thin single coating of molten mixture,
consisting primarily of sulfur vith small percentages of glass fibers
and other additives, is applied over the outside and inside of the
wall.  Within seconds, the molten coating solidifies and forms a
hard, impervious surface.  If a co'lor other than yellow is desired,
pigments can be added to the formulation or the coating may be painted
over with conventional house paint.  Walls constructed using this
technique are much stronger than walls using conventional masonry
construction.  In particular, sulfur surface bond-constructed walls
have considerable strength in tension, whereas conventional masonry
walls primarily show strength in compression and very little in tension.

Using this technique, a building was constructed in 1963 at the South-
west Research Institute, San Antonio, Texas.  Unskilled labor with
no previous experience in applying hot coatings of this type carried
out the work, using regular paint brushes to apply the sulfur.

Concrete block walls are highly porous and must be properly sealed
after installation.  At the Southwest Research Institute, there are
conventional concrete block buildings, some of which are more than
15 years old, which have required sealing, caulking, and re-sealing
from time to time, and still leak when subjected to a hard driving
rain.  No leaks have ever occurred in the sulfur building, nor have
any cracks developed.  The building has not presented any problems and
as a consequence has not received any maintenance.  No cracking,
spelling, or other deterioration of the sulfur coating has occurred.

The performance of the sulfur building over a period of ten years
under the climatic conditions prevailing in San Antonio has convincingly
demonstrated the practicability of the technique.  Another sulfur-surface
bond building was recently completed at the U.S. Bureau of Mines
research facility at Boulder City, Nevada, using recently developed
spray equipment to apply the sulfur-fiber formulation.

The difficulty and cost involved in changing building codes and
practices probably preclude the near-term use of the technique for
human dwellings in the U.S. and other developed countries.  However,
it would appear to have considerable potential in the construction of
warehouses, utility buildings, silos, and similar structures.  A
proposal to evaluate the technique in Latin America and Africa has
been accepted in principle by U.S.A.I.D.  The economics of the technique
appear favorable, and the advantages of higher strength and improved
water resistance should make it attractive in many applications.

Sulfur as a Coating Material

Sulfur formulations similar to those used in surface bond construction
are being tested as low-cost coating and repair materials.  Sulfur
coatings can provide both chemical (corrosion) and mechanical (erosion)
protection.  For coatings of maximum mechanical strength, glass or
other fibers are added to the formulation, but in many cases this
can be dispensed with and coatings consisting essentially of sulfur
vith small amounts of organic modifiers can be used.
                                927

-------
 Sulfur coatings are being evaluated or  considered for a number of
 applications:   protection of  concrete structures in corrosive environ-
 ments, stabilization and erosion control  of  earthworks and storage
 ponds, and stabilization of mineral tailings piles.

 Sulfur is also being evaluated as a repair material for cracks
 in Portland cement and asphalt concretes, for example, in roads,
 airport runways,  and swimming pools.  Several federal agencies recognize
 that inexpensive  coatings and repair materials  can fill an important
 need, and are  investigating various applications.

 A different, but  potentially  very important  application of sulfur
 coating relates to its  use in fertilizer  technology.  In particular,
 sulfur-coated  urea (SOU)  shows signs of becoming commercially accepted
 for use on a number of  agricultural crops in several regions.  By
 coating urea with sulfur,  the dissolution rate  of this nitrogen fertilizer
 is considerably reduced.   This results  in more  efficient utilization
 of the nitrogen by plants and reduced loss of nitrogen to the environ-
 ment by leaching  and run  off.  Many other methods of providing
 slow-release nitrogen have been suggested, but  SOU appears to be the
 only one that  is  cheap  enough for general use.  The Tennessee Valley
 Authority has  been pioneering the development and testing of SCU,
 and is presently  considering  the construction of a full-scale plant.
 A typical SCU  product contains about 10-15$  sulfur.  With world production
 of urea for fertilizer  use at  about 15  million  tons annually, this
 clearly represents  a considerable outlet for sulfur.

 Sulfur Impregnation

 Ceramic and other porous  materials  can  be impregnated with sulfur,
 thereby improving the chemical resistance and the mechanical properties
 of the material.  Bonded  abrasive grinding wheels impregnated with
 sulfur have improved strength; the  sulfur also  acts as a lubricant and
 coolant during grinding.   Sulfur impregnation of ceramic tiles results
 in lower water adsorption, making the tiles  frost resistant for external
 applications.   Impact and compressive strengths are also improved.
 Corrosion of concrete sewer pipes  is a  serious problem in many localities.
 Impregnating the pipes with sulfur  has been  estimated to extend the
 life of the pipe by  a factor of  10, and to double the flexural strength.

 Impregnation of ceramic and concrete materials is generally carried
 out  by immersion in molten sulfur.  Use of vacuum to remove air from
 the  pores of the material prior  to  immersion in the sulfur speeds
 the  rate  of impregnation.

 Research  and development on sulfur  impregnation is in its infancy.
 Potential applications are numerous, particularly in building and construction.

Sulfur Foams

Using small amounts of additives, sulfur can be turned into a foam with
 interesting properties.  The foamed sulfur is lighter than water (typical
 densities are in the range 0.2-0.5); the compreasive strength is in
 the range 50 to several hundred psi, higher than typical organic
 polymer foams.   The foam shows excellent thermal insulation properties.
                                   928

-------
One promising application of foamed sulfur is as a sub-base for highways
and airport runways in cold climates.  Polystyrene foams are being used
experimentally to prevent freeze and thaw damage to highways and
runways, replacing costly and unreliable methods like deep sub-base
beds of stone or gravel.  Foamed sulfur is being evaluated as an
alternative to polystyrene foams in this application.  One inch of
polystyrene or sulfur foam is equivalent in insulating value to 24
inches of gravel.  Sulfur foams have higher compressive strength than*
polystyrene, they are less expensive and can be foamed in place.
Construction of a 1,000-mile, four-lane highway would require about
too,000 tons of sulfur.

Several other potential applications of sulfur foams are being investigated.

Conclusions

By the early 1980's, it is probable that large tonnages of sulfur will
become available from utilization of high-sulfur coal through stack
gas scrubbing, coal gasification, or both.  These tonnages, when added
to those resulting from other sources, will be considerably in excess
of sulfur demand.  Because the electric utilities will be producing
sulfur at a large number of locations often unfavorably situated for
the sulfur markets, disposing of this sulfur in an oversupplied market
can be expected to be initially difficult, and eventually Impossible.
Stockpiling the recovered sulfur would impose an added financial burden
on the industry.  However, if new uses were developed, this sulfur
could be marketed, resulting in added income for the industry,
partly offsetting the cost of desulfuriaation.

The potential new uses outlined above, could, if developed commercially,
absorb the foreseeable supply of sulfur.  They fulfill the criteria
stated earlier to the extent that they are large-volume uses and they
are ecologically acceptable.  They also appear to have favorable
economics, and the time and cost needed for commercial development do
not appear excessive.

It would be wishful thinking, however, to expect these new end-use
markets to materialize automatically, or that their development will be
eagerly undertaken by private industry.  In most cases, the potential
financial rewards will be insufficient to entice firms which do not
have an interest in sulfur or fuels to commit funds.  Up to now, the
burden of supporting the development of new uses for sulfur has been
carried by the main suppliers of sulfur, i.e., the Frasch producers and
the oil and gas companies.  As new sources of sulfur become increasingly
important, the originators of the new supply will have to bear e. portion
of the burden to develop the new uses needed to absorb this
increased production.  Several federal agencies, notably the U.S. Bureau
of Mines, have recognized, the implications of cutting down the
emission of sulfur oxides and are actively sponsoring or cooperating
in sulfur development programs.  The electric utilities industry,
potentially the largest sulfur supplier in the country, has a special
responsibility in this regard.  Although large-scale sulfur production
by the utilities may be several years in the future, now is the time
to research and develop the needed markets.  We invite the utilities
industry, in its own interest, to actively support sulfur development
work, either through the Edison Institute, through cooperative programs
vith Federal agencies, or through The Sulphur Institute.
                                  929

-------
            PANEL DISCUSSION:

    DISPOSAL AND USES OF BY-PRODUCTS
FROM FLUE  GAS DESULFURIZATION PROCESSES
                     931

-------
                             PANEL DISCUSSION

     A. V. Slack (TVA)— We have had an extended discussion of about all
the problems we can think of in. regard to disposing of the products  that
we will be making when we go extensively to recovery or to ridding stack
gases of the sulfur compounds.   We now have our panel  assembled,  including
Mr. Fike (K. L. Fike - The Sulphur Institute) co-author of the last  paper;
so we are ready for any questions that either the panel or the audience
may wish to pose.   I would like  to ask Dr. Minnick  (L. J. Minnick-IU
Conversion Systems) and Bill Taylor (W.  C.  Taylor - Combustion Engineering)
a question.  If we are able to convert the sludge to useful  construction
products (and it seems quite obvious that we can) what proportion of our
sludge  production do you think that we could ultimately  (say, within 10
years) convert to this useful purpose?

     L. J. Minnick— I do not think I can project the finished statistics.
I think that I  would like to comment to this extent.  The utilization of
the technology which was described and illustrated is something that is here.
It is our mission,  as  far as  IUCS is concerned, to take on projects  as they
develop and really turn out to be the garbage man, to take the problem
off the hands of the utility entirely.  That is our mission so that they
are not involved with the running of chemical plant or process.  Now the
thing that dictates whether or not 1t is a good idea to do 1t, of course,
ts the economics; and the answer to your question really 1s that each
situation has Its own logistics.  If you are going to consider stockpiling
aggregate and you have a place to stockpile it in an ecologically safe
way, fine.  If there is a market for aggregate, you may sell enough
                                   932

-------
of it to justify some costs,  If It.1s simply disposal  of a product that's
going to be put Into a landfill  or land improvement project,  what  land
improvement project or what kind of land is available?   If you  are near
a strip mine and you want to rehabilitate it, then it makes a lot  of
sense.  We have looked at a lot of projects across the  country.  Now
as I say, I can not give you a statistical number; but I can assure you
that many, many of the situations that are opening up in the southwest
and eastern United States do lend themselves to the conversion  system
approach.  The quantities of materials that are involved, the very fact
that there is lot of it, make it much more economical.   If someone says
that we have one million or two million tons of material to convert a year,
it makes a lot more sense than if you are just handling a small quantity.
So what the exact amount is, I can not answer that.
     A. V. Slack— All I was looking for was a guess.
     W. C. Taylor— Well, I agree with John (Minnick) in that the question
bears on many factors; but  I see no reason why you could not
use TOO percent of the sludge,  I said in my talk that  in Germany now
they use 80 percent of all  the flyash that they manufacture. About 80
percent of all of the construction of say less than 8-story buildings in
Europe now has a lot of the flyash in it; they are now  digging  up some of
the flyash that they buried to use in construction,  But there  are many
factors in this country that will determine whether or  not we use large
quantities of the stuff.  In my slide, I mentioned my one brief moment
of glory when we had thought that in the laboratory we  had proved the
technical feasibility of utilizing this material in a number of areas.
Several different useful products pertained to building and highway con-
                                    933

-------
 struction.   If you used it in sub-base for highways, you could use all
 the material.  If you look at the highway building program, you can see
 that we use  about 15 billion tons of stone in this country every year
 for construction,  But, like we said, the flyash brick that the University
 of West Virginia came up with was supposed to be economic and it was
 cheaper than the clay,  It was lighter and cheaper to ship.  Structurally,
 it was sounder.   Everything was better than the clay brick; but, you
 could not get anyone to buy it.  You could not get anyone to put up a plant
 to make it.  So this is the problem you have — the old inertia.  A guy
 has been getting his clay brick from a distributor for years.  He has estab-
 lished his rapport.  He is not going to change all of a sudden just because
 you  say,  "Look,  I've  got  some  bricks and  I can give you  2  cents off
 per brick."  I think it is a big marketing problem.  Frankly, we are now
 looking at some throwaway processes because we feel that, until  the marketing
 problem is solved, we will be throwing it away in the beginning.  But if
 someone has a lot of money and wants to make a lot more money, we have  a lot
 technically proven  processes  that  he can  use  and  he can  start buying some  sludge.
     A.  V.  Slack-" Are there any questions that the panel members now
would like to put to each other?
     J.  Ando(Chuo University)— May I ask a question of Mr. Minnick?
 I think the use of the sludge for road  construction is a very good thing.
 I must say I have English trouble and I am afraid I may not understand  you,
but would you say it Is mainly a matter of our economy? I think, it is
rather that if the material  Is very cheap and if its use 1s technically
feasible, it is economically feasible.
     1.  J.  Minnjck— Well, first of all, as we look at it, we do not attempt
to take  credit for the sale  of the product.   We feel  that the cost tif
                                  934

-------
conversion is a disposal cost.  Now one of the reasons that the aggregate
is useful is that It can be stockpiled.  It is important that the power
plant stay on stream; it must run; and every minute of the day the
sludge is coming out and you have got to do something about it.   Now if
you convert it to aggregate and stockpile it, in the construction
season the stockpiles are depleted; when the construction
season is oft, like in the wintertime or something of that sort,
you can still operate and just pile up the material.  But whether or no,
you sell It is a function of the market.  If you are in an area where
there is lots of cheap aggregate, you're not going to sell  it.  If you
are in an area where there is a depletion or shortage of aggregate, yes,
you can sell and sell it at a profitable price.  And again, it's the
logistics of it, the individual situation.  But in our approach to it,
we are looking at it as a disposal charge and then these other things
are credits that can come back and give you some economic trade-offs.
     A. V. Slack— I have a question here from Bill Richardson of Bechtel  who
is concernedi with about 6 mill ton tons of gypsum that we import each year
into this country.  So the  question  
-------
to get this 95 or >90 percent gypsum.  Here you can mine 99 percent pure
gypsum and you can float it down on a barge for practically nothing.  It
would cost, it seems to me, in this country about 5 times as much to get
90 percent gypsum from sludge as it would be to float 1t down at 99 percent.
And unless we are talking reverse economics, I do not think even (though
we are now looking at wall board) that so far it looks like 1t is much more
expensive to try to recover it than to just mine it.
     J.  Ando—Yes, I believe we agree that byproduct gypsum cannot com-
pete with your pure cheap natural gypsum, it is true; but I think that
in certain areas byproduct gypsum may be used because you are still  im-
porting some,  And in Japan, until about 10 years ago, we had plenty of
byproduct gypsum from wet process phosphoric acid production,  No one
used it and we were still importing because they had no experience using
ft; they thought the use of byproduct gypsum might give bad results  for
cement and plaster board,  I worked in the field and made some reliability
tests and showed It could be used, but no one believed 1t.  Now all  of
them are using it and now we have a shortage of gypsum,  Every bit of
gypsum is now used.   I think there is a similar situation in this country,
     A.  V. Slack—Of course, it does depend on what it costs to make
tfiis gypsum.   Since there are various processes in Japan for making  it,
I suppose we really need an economic comparison.  Now may we have questions
from the floor.
     John Salm.(Pioneer Service & Engineering)—On the matter of gypsum,
even the naturally occurring gypsum in the United States has various
degrees  of effectiveness when used as a moderator or regulator in the
setting of Portland cement.  I was wondering how much work has been done
                                  936

-------
 on  testing  the  gypsums that are formed in these various processes to see
 how reactive  they  are, how effective they are in actual use of Portland
 cement?   I  talked  with Dr. Ando in the intermission; he indicated that it
 has been  used successfully in Japan, but there have also been some
 problems  with the  use of the artificially produced gypsum.  I just wanted
 to  throw  that out  to find if there had been any actual research done in
 that regard over here?
     J. Ando—In Japan, phosphogypsum has been the problem if it
 can be used for cement or not.  Until about several years ago, cement
 companies did not  want to use it.  The main reason is  that some small
 amount of P205  (phosphoric acid) and also a fluorine hindered  the
 cement setting.  Since that time they have improved the process of the
 production  of gypsum so that now the P205 in the gypsum can be reduced
 to  below  12,1 percent and it can be used without any problem.  Also/many
 cement companies recently have tried to use byproduct  gypsum from coal-
 fired boilers produced by Mitsui in Omuta by the Chemico and other
 processes:  the Chiyoda process and the Mitsubishi-JECCO process.  All
 of  the byproduct gypsum can be used v/ithout any problem.  The only problem
 is,  if it contains too much moisture, it cannot be put into the cement
 mill; but if  the moisture is about 11-12 percent or less it can be used
without any problem.  In the Mitsui case,  if the flyash content in the
 byproduct gypsum is less than about 10 percent it can  be used in cement.
     Ken  O'Brien (R.  W.  Beck)—I have two questions.   The first one is
                   «
to Dr. Minnick.  I am wondering about the ratio of flyash to calcium sul-
                                   937

-------
fite or sulfate.  Mould you say that what comes out of an average coal-
fired power plant 1s compatible to make these building products  or 1s
there likely to be a large surplus or a lack of one or the other?  Jfy
other question Is to anyone who can answer It.  Assuming flyash  is one
of the products necessary to make a building product,  what do you do on
an oil-fired plant?  Involved in a particular project, we are oxidizing
to sulfate.  It probably wouldn't do the job'because the landfill would
be under water.
     L. J. Minnick^Both of those points were covered in the written
portion of my paper.  But very briefly, the answer is  "yes" to your first
question.  For most coals (those with 3, 4, even 5 percent sulfur),
assuming you have 8, 10, 12 percent ash, you have the  right proportion*
The only place where you get out of line is when you have a boiler that
produces a very low flyash; such as the centrifugal or,what do you call
it, cyclone boiler which produces most of the ash in the form of bottom
ash.  In this case, you are short of flyash.  In connection with oil,
the system that I presented depends on flyash; that's  the magic  ingredient.
And if you don't have flyash available somewhere, you  don't get  the
chemistry that was described.
     Max Schmidt (The university of Wurzburg, Germany)—I want to
ask Dr. Platou (0. S. Platou, The Sulphur Institute),  the last speaker.
Just one question.  You showed us very nice slides on  the use of sulfur
in the construction business.  You showed us a house which was built
in Te*as about 10 years ago and another one just under construction in
Montreal.  Now you are talking about sulfur, talking about millions of
tons of sulfur, but what do you mean by sulfur?  It 1s not pure  sulfur.
Can you tell us something about the additives, the cost of the additives
                                   938

-------
compared with the elemental sulfur itself?
     J. S. Platou^The sulfur mixtures, the sulfur formulations  that we
used in the surface bond construction used in constructing these  houses
are approximately 90 percent elemental sulfur,   The remaining 10  percent
i.s composed largely of fibers; glass fibers have been used mostly up
to now and a very small amount of organic additives in an order of a
few percent.  Does that answer your question?
     M. Schmidt —I did ask for the cost average of the organic additives
or thiocols.  What is the price of sulfur?  Not what is the price of the
additives.
     J. S. Platou—Harold (Fike), do you want to comment on that?  I  would think
that probably thiocols are not exactly what we would want to talk about now.
     H. L. Fike — Over the years, Prof. Schmidt, we have looked at a
good many additives.  As you recall, we did use many of the thiocols,
which cost roughly $1.00 per pound 8 or 10 years ago.  Now we have prf-
marily concentrated on some unsaturated hydrocarbons, particularly
the dipentene, the terpene-type products, the dicyclopentadienes.  We
have looked at various other crude aromatic unsaturated compounds which
come from refineries, some of which are available for 8-10 cents  per
pound,  we feel that if these additives are used in the amount of 3,  4,
or 5 percent they will still represent a rather large portion of  the
costs of the sulfur coating.  It will still be a coating which will be
perhaps slightly  less than most of the standard asphalt coatings  and much
less than the asphalt emulsion coatings.  In some of the cases where we
have used this, we have been more or less competing against the epoxy
materials and, of course, there it was only a fraction of the cost.
                                  939

-------
    A. V. Slack—Two of our panelists  may have  to  leave early and
I have a couple of questions I would like for them  to cover.  One to
Jerry Rossoff (Aerospace Corporation) and again  I am  concerned with
this dewatered sludge that we may produce and what  happens to it when it
rains?  Have you gotten into that problem much,  Jerry?
    J. Rossoff—Well, we have some  limited data and  we are working with
one sludge from the TVA Shawnee plant that shows it will absorb and re-
absorb water and bloat and hold it.   So as far as rain goes,  if
the sludge is not thick and you do not  have any  way to let it drain through
and carry the water off, it will  absorb the water again,
    J, W. Jones (EPA)  —I would like to make one comment on that.
That particular material was allowed to settle naturally and  then
air-dried; isn't that right, Jerry?
    J. Rossoff— Yes,
    J. W. Jones — There was no compaction or any treatment, any physical
treatment of it?
    J. Rossoff—Right, there was no treatment whatsoever.  It was allowed
to dry out.  And it will reach a saturation point and will hold so much
and that is it, but it will regain water.  I think  that  was  the question;
if it rains, it will pick it up and  bloat.
    A. V. Slack — And let me get one more question  in to Mr.  Farmer  (M. H.
Farmer, Esso Research) who may have  to  leave.  In your model, did you
include what we mentioned earlier, the  possibility  of building a ferti-
lizer plant at a power plant or smelter and thereby affecting the economics
in the market?
    M. H. Farmer— Well, in a way, yes.  Of course, there were many things
                                 940

-------
covered in the full report, a very heavy thing,  that  is  not  even  in
the handout material that most of you probably have,   The  Western smelters
were considered in some detail and some of them, of course,  particularly
those in the more northern copper mining states, are  tied  in with
fertilizer production.  But those in the southern part of  Arizona have
a particular problem.  But the study did not ge't into local  marketing
situations.  We wish that we had been able to do so,  but it  was outside
of the scope of the program.
    Fred Grozinger-(Potomac Electric Power Company}—I would like to ask
Mr, Farmer a question,  I am somewhat confused about the overall practi-
cality and objectives of the report that you did for EPA.   Could you
possibly explain if there's anything in this country in general?  Or
is it going to be  specifically a tool for EPA?
    M. H. Farmer — Well, I  hope that it will be of some benefit;to the
utility industry;  but of course the terms of work, the scope of work,
was decided by the EPA,  Working with Norman Plaks (EPA), we were able
to expand  the original  terms  somewhat  and I  think we  made it more useful
as the  project developed.    But  it had  a  specific  purpose to kelp them
 in establishing  their research priorities;  I  believe  that it has had
 some  impact on those decisions already.   But also  it  was  my hope at least
 that, by  placing the industry in  its world context,  it  would make it
easier  for a  utility perhaps  or  any other individual  to fit his  piece
of the  puzzle  into a broader context.   Now,  a  lot  of  times  a suggestion
 is made,  "Well,  surely such and  such  a thing could be done," and the
 suggestor does not realize what  the problem really is in  accomplishing
what  he wants  to do.  Just one quick example,   I read a suggestion
                                   941

-------
 some tfme ago that phosphogypsum should be used as a soil conditioner.
 Well, of course, most of the fertilizer gypsum is produced in Florida
 and most of the soil conditioning goes on in California,  So the
 transportation cost is a very, very important part in the utilization
 of any low unit value material,  I am sorry if I have not answered your
 question.
    F. Grozinger-~-That is fine.  I was just a little confused on the
 actual objective.
    J. $alm—-0n the subject of utilizing sludge as a construction material,
 what experience has there been with regard to the resistance of this
 material to freezing and thawing cycles?  Or to, say, gully washing
 type rains in the Southwest?  Have these materials actually been
 tested for all of these requirements?
    L. J. Hinnick"-Hen t yes.  As I mentioned, this is sort of an
 outgrowth that was done with Poz-o-Pac,  The test procedures have been
 established by Federal highway departments and ASTM and so on.  They
 are very well known, freezing and thawing being a case in point, and
 the answer is  "yes."  They have been fully evaluated from that standpoint.
    J. Salm — One other question, Mr. Minnick.  On the expansiveness
 of this material, is that a controlled expansion, like that of
 cements that can either be expanded or . , ?
    L. J. Minnick — Exactly.   That Is the way to look at it, as an expan-
 sive cement.   Ettringite formation in Portland cement can cause dele-
 terious reactions if the ettringlte forms after the cemfent sets.  In the
case that we have here, the ettringite Is the setting process and it
 forms during the hardening and therefore does not cause deleterious
expansions.   The expansions are beneficial, just like you have with an
                                   942

-------
  expansive  cement.  Same idea.
    J.  Salm—Could I ask one question of Mr. Platou, please?  On
the use of  sulfur  in the construction materials, you are talking about
yery  high percentages of sulfur.  What is the fire hazard?  Would
OSHA  agree  with it?
    J.  S. Platou — It is a good question, sir.  There are several aspects
of this, of course.  Before going any further, I would just say that
additives have been developed, they are even patented, that will make
sulfur  non-sustaining of combustion.  In other words, they comply with
one of  the  ASTM standards.  Self-extinguishing.  Another aspeet is that
when  you add a certain amount of aggregate or other non-combustible
material to the sulfur, this will of course also act as a heat sink
and again contribute to making the material self-extinguishing.  This
is a  problem that  we are aware of,  It is, of course, being evaluated
for these various  uses.  Do you want to add to that, Harold?
    H.  L. Fike — About the only thing I might add to this is the
thought that where many people think that they are going to be just
storing sulfur out in the country, it does pose a potential danger to
just  store  tremendous piles of it in the country.  It could catch fire
there.  And this is one of the reasons where I think the beneficial use
(for  example, in a sulfur-asphalt road or a sulfur foam as an underlay
for a pavement material) would not really create any problem as far as
being combustible.  But it is a problem which we would like to face up to
tnore  directly than we have.  In fact, we would like to come up with some-
thing which would  actually make the sulfur non-burning,  We have not done
this  as yet.  And  it is not really a very promising endeavor.
                                  943

-------
    Jacques Dulin (Industrial Resources)—We have heard today discussions
of taking care of the waste products from the calcium systems  (specifi-
cally, scrubbers) and also what to do with the sulfuric acid and the
sulfur.  Tomorrow we are going to hear something in connection with
sodium systems and ammonia systems,  I wonder if the panel  could jump
ahead just a little bit and perhaps give a few moments of discussion
to the state of the art in connection with the disposal of sodium sulfite
or sulfate from those types of systems,  There is one aspect that I would
like to mention for them to direct their attention to.  We have heard
here specifically of scrubbers and not of baghouses and the use of reactive
filter aids in connection with the baghouses to produce (rather to re-
act with) S02» and to take it out of the air along with the particulates.
I wonder if they could direct their attention to the disposal  of that
type of baghouse waste product material as well,
    A. V. Slack — Well the question of sodium sulfate as a throwaway
product I suppose is best discussed by Dr, Ando and he covered that to
some extent but , .  ,
    J, Ando—Again, there is a big difference in circumstances.
Japan is a small country and surrounded by sea,  We can just put it in
the sea.  We can deliver what is going directly to the sea, so sodium
sulfate is no problem,
    A. V. Slack — Would any member of the panel or any member  of the
audience like to discuss the throwing away of sodium sulfate in the
United States?
                                    944

-------
    W. C. Taylor-»»Hell. I would like to express Combustionvs  apologies
for not having considered that at all,  He seem to be pretty well  tied
up, looking at the calcium sulfate sludges.  While I am speaking,  I
jnfght clarify Combustion*s point of view on this in that we  at CE  are
really not In the waste disposal business.  We thought that  we had to
do this because at the ttme we began there was no one else worrying
about the disposal.  And there was a lot of talk going on that we  had
a potential problem, but no one seemed to be solving It.  We thought
we had to do something,  Once some of these other companies  seem to
be coming up with, processes for disposing of the sludge, we  would  be
yery glad to let them handle our problem.  We were only offering this
in-between service.
    ft. L. Fike-—Mav I make one brief comment on that as well?  The
Sulphur Institute is very interested in finding uses for sulfuric  add
and S02.  We just concentrated on the one effort today; but  I  think  I
would have to say that over the longer term we have not been able  to
see any future market for sodium sulfate in the United States*  As
you know multi-million ton quantities are used today by the  pulp and
paper industry.  From what I can see from the general direction that
this industry is going and from the people I talk with who are quite
knowledgeable in this, I certainly get the suggestion that they are
going to non-sulfur using pulp and paper processes.  Probably about  the
time the utility Industries might have this available, it will not be
used to any great extent in North America.
    Don 6ylfe-{ Atomics  International)—I'd like to answer part of  the
question of disposal of sodium sulfate in this country, a paatial
answer, anyway.  We have a system that uses sodium carbonate that
                                 945

-------
we are trying to promote.  It turns out that in the western states where
the sources of the trona are located ( in general  in such areas as dry
lakes in the California area), you can bring the spent product, the sodium
sulfate, back to such lakes,  It is being done  currently in Owens  Lake
in Owens Valley, in connection v/tth an operation there.    The  California
Water Resources Board has told us that a similar process could be  done
at Searles Lake.  That is not a general answer  on what you do with sodium
sulfate in all  locations, but at least in the western U.S.   It is one
possibility for an interim solution.
     G. A. Hollinden— I would like to ask Dr.  Ando one question.   When
the team was over there in Japan last August, we were told by MKK that
they were going to use sodium sulfate as a soap binder and they showed
us a product of that.  Do you know anything further on that, Dr. Ando?
    J. Ando — Yes, they may use it, but actually they are
wasting their sodium sulfate to the sea.  MKK is discarding the ....
The Wellman-Lord process waste water containing sodium sulfate is  still.  , .
    G. A. Hollinden —Are they discarding any other products over
there too, in the sea, Dr. Ando?
    J. Ando—Yes, MKK recently built another plant in which they
recover sodium sulfate.  It is sodium scrubbing of waste gas.
    H. L. Fike—• I believe there is a sizable quantity if you talk
in terms of millions of pounds of sodium sulfate added to detergent to
protect the aluminum in washing machines, but this would not utilize
the types of tonnages which the utility industry will produce.  But I
think most detergent formulations in the United States contain 3 or 4
percent sodium sulfate if I recall correctly.
                                946

-------
    J. Dulin — Seeing as to how we have heard a few plugs  from people,
our company is of course very deeply interested in  sodium  systems since
we have a couple hundred million tons of the stuff  to sell  out in Colorado,
specifically a sodium bicarbonate mineral  called nahcolite,   We have
addressed ourselves to that problem and have found  ways  in which to in-
solubilize the sodium sulfate as an easily dewaterable granular precipi-
tate, which is a double salt of sodium sulfate and  ferric  sulfate.  And
the material, you are not talking about 60 percent solids sludge,
you are talking about really a material that is 100 percent solids—the
water just drains fight out of it.  The water will  then  be suitable for
the very simplest type of landfill.  We can not say to you that it is
100 percent available today, because we are still doing  research on  it and
answering the hundreds of questions that people have in  connection with  it.
But it can be done.  We are also handling a couple of other methods as well.
I think this should alert the industry to the fact that  whereas there has
been a focus on calcium systems because of the end-product theoretical
insolubility, there has been an earlier recognition of the advantages of a
clearwater liquor type of scrubber or dry baghouses for use with sodium
systems.  The problem with the sodium systems has always been,  "What  do you
do with the sodium sulfate or sulfite that you  get at the back end?"  We
think we  have  an answer to that and  I would like to alert you  to the
fact  that sodium systems should be considered more carefully  than they
have  in the past.
     H. W. Elder (TVA, Muscle Shoals)— I have a question for  the Sulphur
Institute.  Have you considered how there can be equitable competition in
a market where there's an oversupply between a regulated and a  non-regu-
lated industry?
                                   947

-------
    J. S. Platou-—Since the Sulphur Institute has as its members both
 the Frasch industry and the industry which recovers sulfur, such as natural
 gas producers, I cannot comment on that question.  I am sorry.
     J. Ando— May I ask one question to Dr. Taylor?  You showed me that sludge
 is usable for the production of lightweight aggregate.  How could you make it?
    W. C. Taylor—- Yes, well, any product made from sulfate sludge that
 requires sintering would have to have a sulfur recovery process with it.
 We have considered that.  At the time we were looking at beneficiation
 processes, we looked at everything for which flyash had been used,
 Among these were making lightweight aggregate, making sintered brick,  or
 anything like that.  That is one of the reasons for our extensive thermal
 studies of the material.  And we found that you could not sinter the
 material without recovering the sulfur.  Again, at that time we were
 considering a sulfur recovery process with a number of thermal sintering
 processes for utilization of sludge,  Once you get off. the S02 you are
 right back to something like a lignite flyash.  And you can make the
 same type of sintered products that you could from the lignite ash,
    Russ Eberhart (John Hopkins Applied Physics Lab)—Question for
 the gentlemen from The Sulphur Institute.  What kind of landfill
 preparation, pitlining, or whatever would be necessary to stockpile
 sulfur?  That is, to protect (if any protection is necessary), ground-
water,  and so on.
    J. S. Platou^-Are you referring to possible oxidation of the elemental
 sulfur into sulfate and getting it into groundwater?
    R. Eberhart—Anything you can think of; that is, if a power company
 were required to stockpile this sulfur because the market were not
 available, what would they have to do in terms of preparing the ground
                                  948

-------
or  the  pit  in which the sulfur would be stored?
     J.  S. Platou— I  do not  know to what extent I am qualified to answer
that question as of now.  It would of course depend very greatly upon
the location of the sulfur stockpile, whether under average conditions
or  under conditions of high rainfall.  Probably the people who would have
most experience in  that field would be the Frasch industry who, of course,
maintained  large stockpiles of sulfur.  I do not really think I can answer
that question too well.  Harold, do you have any comment?
     H.  L. Fike —Probably all I can do is to add to the problem rather
than to solve it.  Sulfur is very susceptible to both chemical and bio-
logical oxidation, and as a result could cause a problem in certain areas
of  the  country.  I think that we should also recognize that practically
all  the sulfur sold in any large quantities in North America, particularly
the United  States, is now delivered in liquid form.  What they do now,
of  course,  is just put it in forms, vat it, and let it sit there.  If it
were going  to be sold, it would have to be remelted and put into tank
cars.   Now  to avoid that up in Canada, because they are no longer allowed
to  ship solid sulfur  through the port of Vancouver, they have gone into
a slating process.  In other words, as the liquid sulfur comes, they put
it  on a roll and actually make small slates out of it.   They can load
this  in the port of Vancouver and this minimizes the danger from dust
explosions which they had a real problem with in the past,  I only point
this out because I think your question, as one gets down to it, would
require a good many hours of thought and conjecture and certainly a great
deal more knowledge than I possess.
    A. V. Slack —WeTl. Harold, could you continue with that discussion
and cover the HgS emission problem?  I have heard about 1t, but I do not
                                   949

-------
know much about the details.  I have heard that on some of the, what
do you call it, blocks of sulfur in Canada, they are considering having
to put hoods over them to recover the H2S.  Is this true,  or what is
the extent of the problem?
    H. L. Fike—I do not know.   I am trying to check into that.
You, as with many of these problems, perhaps get a different answer from
different individuals.  However, I think it is generally accepted that
as it comes out of a Claus kiln, liquid sulfur will contain fairly
sizable amounts of hydrogen sulfide.  In the liquid sulfur, it is more
soluble hot that it is cold, so when it does  cool  down and crystallize
you now have a potential  problem of its emitting hydrogen  sulfide.  I
do not think this will be a problem in rural areas.  I do not think the
concentrations will perhaps be so large.  In metropolitan  areas I imagine
it could be a  problem, but  I could not quantify it closer than that, Archie.
    A. V. Slack —How does sludge smell?  I am not sure.
    John Raslle (Ebasco Services)  —We  have done some work on the disposal
of these waste sludges from a number of power plants.  We  have found that
there are in most cases,  in fact all  cases, concentrations of boron,
cadmium, nickel,  and silver which far exceed drinking water standards.
In all of effluents from  both wet particulate scrubbers and sulfur
dioxide absorbers, there  seems to be a  certain lack of experience in these
things.  I  just thought I might add that.   Somebody asked  a question
about lining.   In the few plants that have either proposed or put in
wet scrubbers  or sulfur dioxide absorbers, synthetic materials have been
required for the  lining.   Clay has been found to disintegrate under
the action  of  high calcium and magnesium salts contained in the scrubbing
liquor and it's not considered an acceptable lining material.  There-
fore, you normally have to go to something synthetic.  The Bureau of
                                  950

-------
Reclamation has actually designed and 1s requiring the Installation  of an
extensive monitoring system at the Mohave plant,  just to make  sure that
the synthetic lining does not fall or, 1f It does fall, to observe
when  it fails.  What I am trying to say, I guess, 1s that I do not think
there 1s going to be any acceptable disposal procedure which 1s not
going to be associated with lining and extensive monitoring.
    V, C. Taylor —I would like to make on comment on that,  I have
heard of this problem of potential hazards associated with sludge.   But
I would like to mention that you have been ponding flyash now for a number
of years and we are now talking about maybe 25-30 million tons a year
flyash in ponds.  Now the use of sulfur  from air  pollution control  sludge...
in other words, when you added the limestone or the calcium sulfate to it,
you have not increased these potential hazardous elements.  Any hazardous
elements that were there, were probably  associated with the coal and
it's  been with the coal long before you  started adding limestone to it.
So what?  You are not aggravating  any problem  here; you may be  accentuating
it or thinking about it.
    J. Rasile — He have not found  that to be the  case.  We have run
leachate tests on flyash that was  collected dry.  We have stirred vigorously
for   18  hours and determined  how much  leaching of the same trace
elements comes out and I think the highest we ever got on anything was
boron and  that was  1  part  per million,  on  trying to leach
ft out of dry flyash.  As a contrast, when you scrub  it out of  a wet
participate scrubber, when you are scrubbing the  particulates out, we
have  reached 60 parts per million  of boron with no trouble at all.
This  has been observed in about  three plants,  So there seems to be some
chemistry involved.   In other words,  it  is  not the  same procedure.
                                  951

-------
    W. C. Taylor— May I ask what are you adding to the flyash now?
 You are saying the scrubbing process itself makes the flyash mor6
 soluble?
    J. Rasile— No, I am saying, what I think is happening.  What perhaps
 should be determined is whether we are removing elements which used
 to go up the stack.
    W. C. Taylor— Well, it was a hazard in any case, wasn't it?
    J. Rasile — Well, except we are talking here about the disposal
 of waste products.  And it is too bad it was a hazard.  Yes, but right
 now somebody is going to be sitting there monitoring and if they did not
 monitor the stack well, somebody got away with a lot a while ago.   But
 it is not going to be the situation any longer.
    A, V. Slack — I might point out again, as I did earlier, that tests
 in Lunen, Germany, have not, I believe, established any leaching at all
 in test wells.   Tests from Ontario Hydro indicate,  I believe, that the
 liquid phase pretty well  stays occluded in the sludge and  doesn't get
 washed out.  These are just bits of data.
    J. Rossoff-—Well,  I  have my own observations on that.   I would  not
want to doubt your fears,  but I  have to agree with  Bill.   You know that
you are not adding any new elements.  They may be in a different form.
 I would just like to ask you whether you are basing your feelings on actual
 leaching test data to the  soil  or whether  you are basing it on concen-
trations within the solid  sludge itself.   Or are you talking about con-
centrations in  the liquid  form,  which are  now in solution?
                                 952

-------
     J. Rasile—The concentrations I  referred  to  are  those
which were in the supernatant scrubbing liquor,  the  clarified
scrubbing liquor.  The same trace elements exist,  for  example,
in flyash, in roughly the same proportions they did  in the original
coal.  They stay there.  I have no proof of this but my feeling at
this point in time is that we are scrubbing the gases  or vapors,
these volatilized materials, out of the gas. Me are not adding
anything.  But now that it is in the water (and it is in the waste
water) you are trying to discharge it at the rate  at which waste
water will leach.

     A.V. Slack —Well, of course.  May I ask,  though, that  if we
did agree that there is no leaching downward, are  you saying that
this waste water could leach sidewise because we planned to  recycle
it?
     J. Rasile—-By what definition is it not going to leach down-
ward?

     A.V. Slack — I said tests have indicated that.  Now, of course,
we need to corroborate those.  But if you assume this were true,
then would there be a problem?

     J. Rasile-"Well, about the tests, not knowing anything about the
tests, I would say you would have to know something about the soil
                                   953

-------
 characteristics  to determine whether it is going to leach downward or
 not.
    A.  V, Slack — Well, this liquor is staying in the solids.  It does not
 make  any difference about the soil  underneath.  Well, I think perhaps
 we  are  exhausting the subject.
    J.  Rossoff —-May I ask one more?  When Dick Stern (R. D. Stern, EPA)
 spoke here Monday, he made a plea for cooperation and help.  Since
 that  time I have had a lot of offers, gracious offers, to send me carloads
 and 55  gallon drums.  If you would  like to part   with that data   we
 would be happy to use it and compare with ours and include it.  I think
 it  would be very helpful,
    Jim Henderson(ASARCO)^  I don't quite understand  the opinion
 expressed that we are not adding anything, because we are certainly
 adding  lime rock.  And in looking at various lime rocks in the
 vicinities of our smelters, we see a wide variation in the trace
 elements present in the lime rock itself.   So I think we truly are
 adding  something, as compared to flyash.
    W.  C.  Taylor —Well, that is why I asked what were you adding? The
 limes that we have been adding and have been testing have not had any
 higher quantity of trace elements than we had in the ash.   So I imagine
you can get limestone in any purity that you want.   And maybe some of
 them are  contaminated.   But I  guess you would have to look at your
 limestone  then if you were concentrating  trace elements,
    J. Rasile — It is not the quantity of concentration I'm speaking
of.   The total  gross quantity of trace elements is  going to be increased
as compared to the flyash depending on what's in the limestone.
                                954

-------
    Howard Hesketh(Southern Illinois University)— It kind of surprises
me in one respect that we are talking about a problem here with not
only the sludge or the sulfur storage.  You realize that  here in  the
same discussion we have been talking about the fact that you can  use  the
sludge as a retainer for city water supplies, you can use the sulfur
as a waterproof barrier, and so on,  I really see that we have the
answer right here, should we want to use it,
    A. V. Slack — If we can get somebody to buy the stuff,  There is
a practical problem here.
    H. Hesketh—  I am talking about the power pi ants utilizing their
own material to solve their own problem.
    A. V. Slack— I think your material balance might be a little out
of perspective there.
    R. H. Borgwardt(EPA)"- I have a question for Bill Taylor or
anyone else who might know, concerning the magnesium that enters  the
scrubber with the limestone.  All limestones contain magnesium in some
degree or another.  In our pilot plant, the evidence seems to Indicate
that the magnesium leaves primarily in the solid (rather than in  the
liquid) purge.  We do not know; however, whether it is in the form of
magnesium carbonate (that is, the carbonate entering Just didn't dissolve
and leaves with the solid) or whether it dissolved and was precipitated
as possibly the magnesium sulfite and leaves as magnesium sulfite,
    VI. C. Taylor — Well, I'm sorry you asked me that.  But maybe  there 1s
someone here from Combustion.  Is there anyone here from Combustion that
could help me out on this?  I do know that the magnesium hydroxide 1s
fairly Insoluble and 1f you had a high enough pH, you would not have
considerable magnesium in your. .  . ,
                                  955

-------
    R. H. Borgwardt — I am talking about the limestone scrubbing system.
    Robert Van Ness (Louisville Gas & Electric)— One of the reasons
we are using carbide lime is because of this particular point.   Carbide
lime slurry contains probably the range we are using, probably from 0.04
to 0.07 percent MgO.  Now the lime that the carbide was originally  made
from was about 1 percent MgO and I believe the same thing would happen
in Japan because MgO is vaporized in the furnace, and it was already in
the atmosphere.  So you are not adding again.  If you are using limestone,
then you do have the problem.  But that is one of the reasons we chose
carbide lime.  Because it was cheaper also. But this is another point.
We would not have high MgO levels.
    Joe Se1meczi(Dravo Corporation)— As a "has-been" geologist I
would like to make a stipulation relative to the magnesium content.
It's one of the paradoxes of geology that calcium and magnesium carbonate
has a lower solubility than calcium carbonate or magnesium carbonate alone.
Yet, you cannot synthetically make dolomite.  Now when somebody is  trying
to use dolomitic limestone (where the combination is not a camouflaged
dispersion of the magnesium in a calcium structure, but a dolomite
structure) then the magnesium will go through undissolved because it
will be a lot less soluble.  In other words, equilibrium and kinetic-
wise, one way a field geologist determines the difference between calcium
carbonate and dolomite is to drip 4 percent hydrochloric add on a
stone.  Now if he observes a fast evolution of carbon dioxide, 1t 1s
                                     956

-------
limestone,  If the evolution 1s very slow,  1t 1s  dolomite.   I  do not
know how much this 1s worth but probably it has something to do  with
the way magnesium 1s leaving a system of limestone scrubbing.  Now
relative to the trace elements.  We have been working  with  western coals
(I am sorry, eastern coals) primarily at Dravo and we have  run several
Teachability tests on flyashes and the sludges generated primarily
by lime scrubbing.  Now we Pound that in all cases the trace elements
were lower than you would get from flyash leaching alone.  In one case
only the arsenic content was in excess of Federal regulations  for
drinking water.  And even in the case of arsenic, the flyash leachate
(the leachate from flyash) contained higher arsenic content than the
sludge.  One reason for that is perhaps that the sludge 1s  creating Its
own reducing condition and arsenic is not thermodynamically present as
arsenate in a reducing condition.  Now if you are slurrylng and a
dissolution experiment is done in a highly oxidizing condition, you may
dissolve more arsenic from your sludge.  Now western coals do contain
a great deal more arsenic than eastern coals.  As a matter of fact, mine
leaching  (mine discharge waters) in the West contain arsenic In excess
of what's permissible by drinking water standards.  Pedple we talked to
at EPA are not concerned about mine acid waters  in the East, as far as
arsenic content is concerned.
                                    957

-------
    Steve Smith (Koch Engineering Co)— I have a question for Bill
Taylor regarding the spray drying of  the sludges from the scrubber
with the waste heat of the boiler gas.  We think there may be some
trend for the unreacted limestone in the sludge to reactive SOp and
also it looks like there might possibly be some oxidation of the sulfite
to sulfate.  Has your work turned up anything on this?
    U. C. Taylor — We have gone to people like you and other spray dryers.
But address that to Rao again.  Rao, did you hear the question?
    M. R. Gogineni — I did not quite understand the question.  Could
you repeat the question?  Maybe I can help.
    S. Smith — Yes, in spray drying the sludge from a limestone
scrubber, have you found any reaction between the SCL and the limestone,
the unreacted limestone in the sludge?  And  have you found any oxidation
from the oxygen, the flue gas of the sulfite to sulfate?
    M. R. Gogineni ->-I think in saying we have done work, I think it is
not experimental work,  It is still in the paper evaluation and we
have done some pilot plant tests using the,  I think, air at the tempera-
ture of the flue gas.   But there was no S02  removed in the spray dryer as
a result of reacting with the unreacted calcium carbonate that was present
in the sludge.
    S. Smith-"Okay; I think, depending on the amount of limestone in
the sludge, of course, that this will vary.   I think there may be a little
                                    958

-------
bit  of  removal,
     M.R. Gogineni — Maybe,
     S.  Smith-Okay.
     M.R. Gogineni —Or maybe not, if it is solid.
     W.  V. Botts (Atomic International)— I might just comment on
spray drying, We have some data that suggest, even with the lime,
that you do get S02 removal.  We are selling systems of a spray
drying  nature using sodium carbonate.  A spray dryer is probably one of the
most efficient contactors that you can come up with for
contacting flue gas with a reactant, and we are using sodium carbonate.
In our  tests at the Mohave station, we ran some lime dilute solutions and
got  quite good S02 removal.  So if there is any lime in the sludge, you
do indeed get $02 removal,
     J.  Selmeczj — i do not think I have to introduce myself again,  but
I remember what I wanted to say,  As far as the heavy metals leaching
into the soil and the probes, valve points, or whatever devices you
try  to  determine, their transferral in the groundwater depends entirely
on the  soil conditions,  As geochemists know, humus is a very good  ion
exchanger which you probably find close to the surface.  As a matter
of fact, this is how we detect buried deposits of valuable minerals.
Clays also have fantastic capacities  for heavy metals in the cationic
form; not in the anionic form,  however,   Now, for example, with cesium
(which was  conducted under  the  supervision  or sponsorship of AEC),  a
                                    959

-------
 radioactive cesium was  found  to  be most  economically  removed  by  clay  or
 just natural  soil, mixing  radioactively  contaminated  water with
 natural  soil.   So  just  to  sound  a little bit of a hopeful  note,  heavy
 metals  if  they  do  leach into  the ground  and if the ground  is  not porous
 sandy soil,  but contains even a  small  amount of clays,  the clays will
 remove  the heavy metals.
     J. Ando—Let  me just  mention just one thing about  the  spray
 drying of  the sludge.   I don't think that desulfurization  takes  place
fry  the calcium  carbonate temperature which is not very  high,  But I am
 sure that  considerable  oxidation of calcium sulfite will occur in
 the  spray  drying.
     M.R. Gogineni—Regarding the spray  drying and S02  removal in the
 spray dryer.  I  think the  gentleman commented about the fact  that he was
 getting  some SO^ removal when he tried to dry sodium  carbonate.  Am I
 right?   It may  be  due to what's  in the solution rather  than what's in the solid
 form.
                                         960

-------
REMOVAL OF SULFUR DIOXIDE FROM STACK GASES
  BY SCRUBBING WITH AMMONIACAL SOLUTIONS:
          PILOT SCALE STUDIES  AT TVA
                       by

      Gerald A. Hollinden and Neal D. Moore
               Power Research Staff
            Tennessee Valley Authority
             Chattanooga,  Tennessee
                P.  C. Williamson
         Division of Chemical Development
            Tennessee Valley Authority
             Muscle Shoals, Alabama
                  D.  A. Denny
            Control Systems Laboratory
         Environmental Protection Agency
      Research Triangle Park, North Carolina
                         961

-------
           REMOVAL OF SULFUR DIOXIDE FROM STACK GASES

             BY SCRUBBING WITH AMMONIACAL SOLUTIONS:

                  PILOT SCALE STUDIES AT TVA
                          ABSTRACT
     The Tennessee Valley Authority and the Environmental Protection
Agency have pursued an ammonia scrubbing program at TVA's Colbert
pilot plant since 1969.  This effort began as a wholly funded EPA
program with the objective to fully characterize ammoniacal liquor
scrubbing of S02-laden flue gases.  Since 1971, the program has
become a jointly funded effort in which the operational phase of
ammonia scrubbing was coupled with a regeneration scheme for producing
a concentrated stream of sulfur dioxide.  The principal advantage in
the EPA-TVA process is economic regeneration by acidulation with
ammonium bisulfate produced from thermal decomposition of ammonium
sulfate.  This paper highlights the activities during the continuing
test program.

     In general, S02 recovery is excellent and NH3 losses are low
using ammoniacal scrubbing solutions.  However, the following problem
areas have been established:

           •Fume formation in the scrubber and after discharge
            from the stack

           •Fly ash separation

           •Ammonium sulfate separation

     Under proper operation of the system, it is possible to control
fume formation inside the scrubber while operating at relatively high
salt concentrations.  Avoiding fume formation on days of low tempera-
ture and high relative humidity may be impractical to achieve.   While fly
ash and ammonium sulfate separation have not been satisfactorily demon-
strated, several promising leads will be investigated in a modified
process sequence.

     The double alkali approach is another possible scheme to utilize
ammonia in recovering oxides of sulfur from stack gases.  This approach
also will be studied.


                             962

-------
Introduction

     The Tennessee Valley Authority and the Environmental Protection
Agency embarked on an ammonia pilot plant scrubbing project  4 years
ago.  The pilot plant treats 2600 standard ftVmin*slip stream from a
TVA coal-fired steam generating station located in northwestern Alabama.
The objectives of the study were to establish the ability of ammoniacal
solutions to efficiently remove dilute quantities of sulfur oxides from
flue gas with extremely low gaseous ammonia losses.  Trouble-free
operation of equipment during the first 1-1/2 years of testing
promises a highly reliable, low-maintenance scrubbing system.

     Because of the cost of ammonia and the extreme solubility of the
ammonium salts, the process is not amenable to throwaway operation.
In this light, a regeneration scheme ("bisulfate process") was adopted
that appears to have several advantages over other regenerative systems.

     This paper describes the pilot plant operation incorporating that
portion of the total system—absorption plus regeneration—that is
currently available from a 1-year period of testing.

Background

     Several regenerable or semi-regenerable ammonia scrubbing processes
have been developed and are, or have been, in full-scale operation.  The
earliest of these was pioneered by The Consolidated Mining and Smelting
Company (Cominco) in Trail, British Columbia.1  The process consists of
aqueous ammonia scrubbing followed by acidification of the sulfite
liquor with sulfuric acid to evolve S02 and produce ammonium sulfate.
The S02 is sent to a sulfuric acid plant and the ammonium sulfate is
further processed for sale as a fertilizer.  The process has been
operating continuously and reliably on smelter gas since the mid-1930's
and is still in operation.  Work on adapting the method to power plant
stack gas was carried out by TVA (pilot plant scale) in 1953-54.2  The
absence of a large market for ammonium sulfate, however, severely
limits the applicability of this otherwise excellent process in the
United States.

     Quite similar processes have been developed and are in full-scale
operation on sulfuric acid plant tail gases in Czechoslovakia  and
Romania.3*4  In these methods, the effluent scrubber liquor is acidu-
lated with nitric or phosphoric acid; and the process, therefore, must
be integrated with fertilizer plants to provide an outlet for  the
ammonium nitrate or ammonium phosphate formed.  Again, the limited
marketability of the fertilizer products and constraints as to the
location of the SC^-emitting plant near a fertilizer manufacturing
center limit their wide use in this country.
 *See English units and metric equivalents on page 996.


                                 963

-------
     A regeneration process producing SCL only was postulated and tested
 in  the 1930 's by H. F. Johnstone.   Steam stripping was employed to
 recover S02 and to regenerate the ammoniacal solution for reuse in the
 scrubber.  Since concentrated S02 is the major product, either sulfuric
 acid,  liquid S02, or elemental sulfur can be the final product, depend-
 ing on the need of the user.  Development of the Johnstone process has
 been vigorously pursued in the U.S.S.R.  A 140,000 standard ft3/min
 (60 Mw) unit was installed near Moscow in 1952 and operated
 continuously until 1967 at which time it was disassembled because the
 power  plant was converted from coal to natural gas.  It is reported
 that the cyclic ammonia process will be installed on a 200 Mw coal-
 fired  utility boiler in the U.S.S.R. in 1973.  Although the Johnstone
 process has been shown to be feasible and presupposes no link to a
 complex fertilizer plant, it possesses some undesirable characteristics.
 Energy requirements for the process are high, approximately 12 Ib steam
 per pound of S02.  Oxidation products may be difficult to purge from
 the system without loss of active species thereby resulting in a higher
 effective oxidation.  The occurrence of undesirable disproportionation
 reactions in the steam stripper further aggravates the oxidation problem.

 TVA-EPA Ammonium Bisulfate Process

     In the decomposition of sulfites with an acid, it may be possible
 to regenerate the acid, thereby avoiding the need for disposal of the
 acid salt.  Such an acid regeneration process*) has been known in the
 fertilizer industry since the 1920 's.  In this process, the ammonium
 sulfate is heated to drive off ammonia and produce acidic ammonium
 bisulfate, which is then used as an acidulant to release
     In more recent work, an engineering company has incorporated the
bisulfate technique in various fertilizer flowsheets? and has carried
out pilot plant work for converting ammonium sulfate to bisulfate.
The process involves direct heating of the ammonium sulfate with
combustion gas.  The decomposition reaction is highly endothermic.

     This work is being expanded currently to a large-scale test program,
on a scale equivalent to the size required for a 30 Mw power plant burn-
ing 3.5 percent sulfur coal.  Several other research organizations have
also worked on the conversion step but in a less extensive way.

     Application of the bisulfate technique to regeneration of ammonia
scrubber liquor in an S02 removal process was proposed by Hixson and
Miller in 1944. 8  Ammonium sulfite-bisulfite liquor is pumped from the
scrubber to an acidifier-stripper where the following reactions take
place:
           (NH4)2S03 + 2NH4HS04 -»• 2(NH4)2S(>4 + H2<> + SO  +      (1)

             NH4HS03 + NH4HS04  -»•  (NH4)2S04 + H20 + S02 t      (2)
                               964

-------
     Essentially all of the SO. is released from the concentrated
liquor in the acidifier-stripper.  If a condenser is employed on the
off gas, virtually a 100 percent stream of SC>2 can be obtained.  The
resulting stripped solution, containing mainly (NH^oSO^, is sent to
an evaporator-crystallizer where the water is removed.  Ammonium
sulfate crystals are then transported to the decomposer where the
(NH^oSO^ is thermally dissociated into NH.HSO, and gaseous NH3.  The
bisulfate is returned to the acidifier and the ammonia to the scrubber.

Pilot Plant Equipment and Operation

     The pilot plant is designed to treat 4000 actual ft3/min flue gas
at 300°F.  Stainless steels (304 and 316} and rubber are used in all
wetted sections for corrosion protection.  The gas ducts are all
mild steel.  For simplicity, the plant is divided into the absorption
and regeneration sections.

     Absorption:  The absorption section is shown in Figure 1.  The
scrubber is a three-stage, marble-bed unit which was available from
a previous ammonia-scrubbing study.  The three scrubber stages can
be operated to maximize the S02:NH3 ratio in the product liquor while
limiting the S02 and NH3 content of the scrubbed gases to 250 and
50 ppm, respectively.

     The source of the pilot plant flue gas is downstream of the
electrostatic precipitator on Unit No. 4 at TVA's Colbert Power Plant.
Most of the fly ash is removed in the precipitators.  However, the
remaining fly ash (up to 0.5 gr/ft3) enters the pilot plant scrubber
where a portion of the ash is removed and accumulates in the scrubber
solution at a rate of 1 to 10 pounds per hour, depending on the
efficiency of the precipitator

     Most of the fly ash settles out and accumulates in the scrubber
product storage tanks, F-5 and F-6.  (Each of the storage tanks has
a capacity for approximately 24 hours' absorber operation.)  Periodi-
cally, the settled material is purged from the system.  In later work,
the settled material will be sent to a tub filter and the cake washed
in a batch-wise operation.  The wash from the filter will enter the
absorber cycle as make-up water.

     An alternative method of fly ash removal is to scrub the flue
gas with water ahead of the absorption sections.  Wj.th this scheme,
a small amount of fly ash in the gas stream, as well as entrained mist,
will get through the fly ash scrubber even though a high efficiency
scrubber is used.  Thus, a portion of the carry-over fly ash will be
collected in the absorber, necessitating a purge cycle.  A portion of
the make-up water will be used to wash the fly ash cake  from  the purge
cycle.
                              965

-------
               WET ELECTROSTATIC
                 PRECIPITATOR
                           SCRUBBER
                                                                                 TO STACK
FLUE
GAS
.
	 V.
	 r
MAK
WAI
N 	
E-UP
PER
i
J





G-3
G-2
G-l
: 	 7
i i r i' v



VARIABLE SPEED BLOWER
J-23





f
•\ - /-
*v /•
    J-21
              J-22
AT/1MONIA FROM
REGENERATION
   SECTION
                                  VARIABLE SPEED PUMP          VARIABLE SPEED PUMP       VARIABLE SPEED PUMP
                                          J-l                           J-2                       J-3
                      REC1RCULATION TANK          RECIRCULATION TANK        RECIRCULATIONTANK
                                                                                PRODUCT LIQUOR TO
                                                                               REGENERATION SECTION
                             J-12
         Figure 1.  Absorption section ammonia scrubbing-ammonium bisulfate regeneration process.
                                                966

-------
     Regeneration:  The regeneration section shown in Figure 2 is
designed to process liquor produced in the absorber section.
Originally, the pilot plant regeneration process was designed to
include the following steps:

     1.  Acidulation of the absorber product with ammonium bisulfate
         melt to release S02 and form ammonium sulfate.

     2.  Stripping of the released S02 from the ammonium sulfate
         solution.

     3.  Crystallization of ammonium sulfate from the mother liquor.

     4.  Separation of ammonium sulfate from the mother liquor.

     5.  Generation of ammonium bisulfate for the acidulation step
         by heating a 1:1 mole ratio of H2S04 and ammonium sulfate.

     The acidulation is accomplished in a vessel made from a 6-foot
section of a 12-inch stainless steel pipe.  The vessel is coated
internally with DuPont's TFE Teflon for corrosion protection.  Mixing
of the incoming scrubber liquor and ammonium bisulfate is accomplished
in a cone mixer in the upper part of the acidulator.  Released S02 flows
from the acidulator to the stack aftei: joining the effluent from the
stripper vessel.  The acidulator is mounted so that the point of gravity
overflow from the acidulator to the stripper can be raised or lowered to
vary the retention time of material in the acidulator from zero to a
maximum of 4 minutes.

     The Teflon-lined stripping vessel is 1 foot in diameter by 6 feet
high and contains a 4-foot section of 1/2-inch ceramic Raschig rings.
Acidulated liquor enters the top of the vessel and flows countercurrent
to a stream of stripping gas entering the vessel near the bottom.  In
the pilot plant, provisions are made to use either steam, air or
scrubbed flue gas as the stripping gas.

     Ammonium sulfate solution from the stripping vessel flows by gravity
to crystallization tank F-7 where water is evaporated.  The concentrated
solution then flows to F-8 where cooling and crystallization take place.  The
saturated solution, containing crystals of ammonium sulfate, is centrifuged
to separate ammonium sulfate crystals from the mother liquor.  The filtrate
is returned to the absorber section.

     In a completely closed-loop ammonium bisulfate regeneration scheme,
all the ammonium sulfate produced in the acidulation step would be
thermally decomposed.  The resulting ammonium bisulfate is fed to the
acidulation step and the ammonia is returned to the absorption section
(to recirculation tank F-2).
                              967

-------
                                                                                 c
        SULFURIC ACID
ETANK
f~X
I
Jj
10 I)
II /
n"
U '



HOT
COMBUSTION
GASES I
T
^


/^



Ji
AMMON
AIR BISULF
HIK GENER/
^
                                                                           AMMONIUM SULFATE
                                                                               FEEDER
           J-ll
           AMMONIA STORAGE TANK
                                      PROPANE
                                        GAS
                                                                                         TO STACK
            (I
AMMONIA TO
ABSORPTION
SECTION
STORAGE AND SETTLING TANKS
                                                               ACIDULATCR
                      FLY ASH TO WASH AND DISCARD
                                                              STEAM
                                                                                      STRIPPING
                                                                                        GAS
                                                                      F-7
                                                                    AMMONIUM SULFATE    J*10
                                                                 CRYSTALLIZATION TANKS
                                                                      CENTRIFUGE
                                                                                       	|
   PRODUCT LIQUOR FROM
    ABSORPTION SECTION
                             J-8
                                                                        AMMONIUM SULFATE
    Figure 2. Regeneration section ammonia scrubbing-ammonium bisulfate regeneration process.

                                              968

-------
     In the present pilot plant study, the thermal decomposition stage
was omitted and the acid ion was furnished as sulfuric acid.   Ammonium
bisulfate will be the source of the acid ion later.  The required
ammonium bisulfate will be generated by heating a 1:1 molar mixture of
sulfuric acid and ammonium sulfate.  The sulfuric acid and the ammonium
sulfate (from the centrifuge) will be fed separately and continuously
to the ammonium bisulfate generator," D-l, a Teflon-lined stainless
steel vessel.  Here, sufficient heat will be added to the acid-ammonium
sulfate mixture to evaporate the moisture added with the acid and to
bring the mixture to about 350*F.  The ammonium bisulfate melt will then
be fed to the acidulator.

     Since ammonium sulfate is not decomposed in the present  study, no
ammonia is available from the regeneration section.  The makeup ammonia
is fed to the absorption section from the ammonia storage tank, F-9.
During the present pilot plant operation, the ammonia is fed  undiluted
directly to the points of use in the absorption section.  Provisions
will be made later to simulate ammonia recovery from a thermal decom-
position process.

     Vent System:  The scrubbed flue gas is vented to the atmosphere for
plume observation.  The recovered S02 from the acidulator and stripper
can be vented into the power plant stack.  The storage tanks  containing
ammoniated solutions can be vented to the absorber or to the  stack.

     Instrumentation:  The pilot plant is instrumented throughout so
that all pertinent liquid and gas flows are monitored and values are
recorded.  All signals are electrically transmitted from the  sensing
element to the recorder-controller.

     The gas flow through the scrubber system is monitored using a
flange orifice in the duct leaving the scrubber.  A differential
pressure cell senses the pressure differential across the
orifice and sends a signal to a recorder-controller on the
pilot plant instrumentation board.  Any deviation from the preset values
on the recorder-controller causes a signal to be sent to the variable-
speed drive mechanism on the induced draft blower (J-23) to correct the
deviation.  This arrangement assures that a constant gas flow through
the scrubber system is maintained even though the pressure drop across
the system may change.

     Sulfur dioxide levels in the gas to the scrubber and after each
scrubber stage is monitored using an ultra-violet analyzer^
The analyzer has three ranges of S02 values: zero-4000, zero-
1000, and zero-100 parts per million full-scale reading.   The
sample point is changed manually from station to station to avoid the
possibility of leaks from an automatic sample sequencing system.
Periodic checks by wet-chemical methods rarely differ from the analyzer
reading by more than 5 percent.
                                969

-------
      An  NOX  analyzer has been installed in the pilot plant but has not
 been  activated.   It is expected that this instrument will be available
 to monitor the inlet and outlet NO and NO- levels in future pilot plant
 runs.

      A smoke detector is used to monitor the intensity of the plume at
 the stack exit.  The instrument uses a light source and a photocell
 to measure the plume intensity.  The digital readout is in Ringelmann
 units.

      Gaseous ammonia is metered to the system as required by the use of
 a differential pressure cell coupled with a recorder-controller
 and a flow control valve.  Liquid flows are sensed by magnetic flow-
 meters which send electronic signals to recorder-controllers.  The
 required flows of recirculating liquor to the scrubber stages are
 controlled by variable-speed pumps (J-l, J-2, and J-3).   Variable-
 speed pumps are used for flow control instead of valves because fly ash
 removed in the scrubber could cause plugging and erosion of control
 valves.  Automatic flow control valves are used to control the flow of
 the remaining liquid streams.

      Temperatures throughout the system are sensed with thermocouples
 and are recorded on strip charts in the control room.

 Problem Areas

      Fume:  The most apparent problem surfacing during the past and the
 current TVA-EPA work with the ammonia-scrubbing process is that of fume
 formation.  Figure 3 shows a typical ammonia-based plume including steam
 emitted during routine pilot plant operations with no gas reheat (exit
 temperature is approximately 125°F).  The plume was formed during opera-
 tion using three scrubbing stages, no prior humidification, and while
 producing a liquor with a C of about 10 and an S/C of 0.8 (C = moles of
 ammonia as sulfite and bisulfite per 100 moles water*, S = moles of
 sulfite and bisulfite sulfur per 100 moles water).  Sulfur dioxide
 removal efficiency under these conditions is typically about 90 percent.

      Efforts to reduce or eliminate the plume through manipulation of
 the pH   of the scrubber liquor and the addition of a humidification
 step  ahead of the scrubber did not reduce the plume opacity signifi-
 cantly.    However, under these test conditions, the steam plume may have
masked any reduction in the ammonia-based plume.  Adding a wet electro-
 static precipitator after the scrubber failed to produce significant
 results at the normal gas flow rate of 2600 standard ft3/min. (The
precipitator was found to be 50 percent efficient in removing the
particulate mass at this flow rate--15 ft/s shell velocity through the
precipitator.)

*combined and uncombined
                                 970

-------





                                                                         ami
Figure 3.  Typical plume during routine operation; high scrubber liquor concentration (c = 10); no reheat.

-------
      Chemical  and petrographic  analyses of the plume collected in an
 impaction sampler  (Brinks mass  sampler] indicate that the major fraction
 of the ammonia-sulfur  salt  is ammonium sulfate.  It is probable that the
 particulate was  formed in the vapor phase as ammonium sulfite and then
 oxidized to the  sulfate form in the sampler.  A portion of the particulate
 has been analyzed as ammonium chloride.   (The coal used during these tests
 normally contained  0.1-0.2  percent chlorides.)  Tests showed that a water
 wash ahead  of  the scrubber  materially reduced the chloride content of the
 gas entering the scrubber.

      Personnel from the Southern Research Institute of Birmingham,  ~
 Alabama,  determined the fume particle size to be 0.25 micron with 10
 particles in the size  range of  0.005 to 0.5 micron per cubic centimeter.

      Reheating the  scrubbed gas  was effective in reducing plume opacity.
 These data also  show that plume opacity is greater at higher salt con-
 centration  (high C's)  than  at lower salt concentration (low C's).

      Solids Separation:  Other  problems identified in the pilot plant
 program  are not as  visible  as the plume problem but may be difficult to
 solve.   These  are the  problems  of fly ash separation and ammonium
 sulfate  separation.  As stated  earlier, the fly ash is expected to
 separate by gravity in the  absorber product storage and settling tanks.
 These tanks, which  have capacity for about 24 hours' retention time, are
 effective in removing  most  of the fly ash.  However, the small quantity
 in  the supernatant  liquor from  the tanks may be sufficient to adversely
 affect the  production  of crystalline ammonium sulfate further downstream
 in  the process.  Also  the fly ash which contains iron may catalyze the
 decomposition  of ammonia during  thermal decomposition of ammonium
 sulfate.  The  cracking of ammonium caused by iron and other contaminants
 in  the fly  ash during  the ammonium sulfate decomposition step will be
 examined by the Eugene Kuhlmann  Company in France.  The work will be
 done in  a pilot plant  sized to handle the production from a 30 Mw power
 plant.

      Attempts  to remove the product liquor from the settled fly ash by
 filtration  in  a tub filter failed because of blinding of the filter
media.   The fly ash was  settled  from scrubber product liquor with C's
 of  approximately 10 and containing from 10 to about 30 percent by weight
 ammonium  sulfate.  The blinding was caused by gelatinous, thixotropic
material  composed of finely divided fly ash and tiny needle-like crystals
 of  ferrous  ammonium sulfite which precipitated from the scrubber solution.
The  iron  in this compound is believed to come from fly ash.  The quantity
 of  iron dissolved in the solution is small—on the order of 0.01 gram
per  liter.  Use of  filter aids and precoats on various types of filter
media failed to prevent blinding.

     An equipment company specializing in solids separation made filtra-
tion  tests on  the scrubber product liquor containing fly ash and also
on the mud from the bottom of the settling tanks.  They recommended use
of drum filters and filter aids on both materials.

                                  972

-------
     Separation of ammonium sulfate crystals from the acidulated and
concentrated liquor in the regeneration loop, though difficult,  may
be easier than fly ash separation.  As stated earlier, the liquor
concentration and crystal growth take  place by evaporating water from
the solution in a tank at atmospheric pressure.  Crystal growth  cannot
be controlled in this equipment and attempts to separate the small
crystals in the pilot plant centrifuge and drum filter have failed.
Petrographic analysis indicated that ferrous ammonium sulfite was
present in the material blinding the separation equipment as was the
case in the fly ash separation tests.

     Bench-scale work using solutions from the pilot plant operation
produced adequate crystals for filtration.  Filtration rates of over
200 gallons per hour per square foot were obtained in these tests which
produced much larger crystals than those produced in the pilot plant.
The ferrous ammonium sulfite did not cement together the larger crystals
produced in the bench-scale work.  The use of an evaporator-crystallizer
is expected to produce the required crystals although the small amount of
fly ash in the solution may influence the design (and cost) of the unit.
The equipment manufacturers contacted concerning the sulfate crystal
growth and separation were of the opinion that the standard sulfate
crystallizer equipment would be sufficient.  However, all stated that
they could not guarantee this equipment, without making tests using the
actual solution containing fly ash.

     Once adequate crystals of proper size are produced, separation
would be accomplished by the standard centrifugation method used in the
ammonium sulfate industry.

Pilot Plant Test Results

     Routine Operation:  Aside from the aforementioned problems, per-
formance of the pilot plant in removing S02 from the flue gas and in
subsequently releasing the absorbed S02 has been good.  Table I shows
data from a typical pilot plant run when producing an absorber liquor
with a C of 10 and an S/C of 0.8.  The ammonia required to react with
the S02 was added to the second-stage absorber loop  (tank F-2).  Under
these conditions, the S02 removal was approximately 90 percent.  When
the S/C was raised to 0.85, the removal efficiency, as expected,
dropped (to 85 percent).  Approximately 13 percent of the absorbed S02
was oxidized to ammonium sulfate.  The Murphree tray efficiency for
the marble-bed scrubber was about 0.90.

     The regeneration loop was operated using sulfuric acid  (93 percent)
to acidify the liquor from the absorber section.  Typical data from
these tests are shown in Table II.  Sulfuric acid was used for acidu-
lation instead of ammonium bisulfate because no ammonium sulfate crystals
were available from the pilot plant for production of the bisulfate.
Clarified liquor from the absorber section was pumped to the acidulator at
approximately 0.4 gaVmin, the rate the liquor is produced in the absorber.

                                 973

-------
                                  TABLE I
                        Typical Absorber Loop Data
      Test Conditions

Gas to scrubber
  Flow rate, ft3/min at 32°F
  Temperature, °F
  S02, ppm
Gas leaving first stage
  Temperature, °F
  SO2> Ppm
Gas leaving second stage
  Temperature, °F
  SC»2, ppm
Gas leaving scrubber
  Temperature, °F
  SO2, ppm
862 removal, %
Stoichiometrya
Forward feed flow rate,*5
Liquor to top stage
  C
  S/C
  PH
  Specific gravity
Liquor to middle stage
  C
  S/C
  PH
  Specific gravity
Liquor to bottom stage
  C
  S/G
  pH
  Specific gravity
Product from scrubber
  C
  S/C
  PH
  Specific gravity
  Flow rate, gaL/min
 NH3  to  F-2  tank
(Normal  Operation!
        2650
         285
        2360

         127
        1440

         125
         360

         122
         280
          88
        1.45
         1.3

         1.4
        0.78
         6.1
       1.036

         5.1
        0.62
         6.8
       1,098

        10.8
        0.81
         5.7
       1.200

        10.4
        0.84
         5.7
       1.202
        0.41
a
 Stoichiometry is the ratio of moles of. gaseous ammonia added to the
 moles of S02 in the gas stream.
3The only forward feed added to the system was water.
                                   974

-------
                                TABLE II
                    Typical  Regeneration Loop  Data
     Test Conditions
Acidulation with
                                                                Only
Acidulator
  Liquor feed in
    C
    S/C
    pH
    Specific gravity
    Flow rate, gaL/min
  Sulfuric acid
    Flow rate, gaL/min
    Percent su If uric acid
  St oi chiome t ry a
  Liquor flow out
    C
    S/C
    PH
    Specific gravity
  Percent 862 release
Stripper
  Stripping gas
    Type gas
    Flow rate, ffVmin at 70°F
  Liquor flow out
    C
    S/C
    pH
    Specific gravity
  Percent S0  release
Overall
          S02 release
                                                      9.8
                                                     0.83
                                                      6.0
                                                    1.193
                                                     0.36

                                                     0.11
                                                       93
                                                      1.2

                                                      1.3
                                                      1.0
                                                      2.0
                                                    1.194
                                                       85
             Air
              12

             0.2
             1.0
             2.1
             .188
              83
              97
 Stoichioraetry is the ratio of moles of acid ion to moles NH3 as
 ammonium bisulfite and ammonium sulfite in the liquor to the
 acidulator.
                                 975

-------
 Sulfuric acid  was pumped  to the acidulator at a rate to give a ratio
 of 1.2 moles of  acid  (hydrogen) ion per mole of sulfite and bisulfite
 sulfur in the  feed  liquor to  the acidulator.  (The 1.2 ratio was used
 for convenience  in  these  preliminary tests; the theoretical require-\
 ment is  1.0.)  Approximately  85 percent of the absorbed sulfur as
 sulfite  and bisulfite  is  released from the acidulator.  The remaining
 15 percent of  the S02  flowed  to the stripper where air at 12 ft3/rain
 was added to remove the residual S02.  Approximately 83 percent of
 the residual S&2 to the stripper was removed by the stripping gas to
 give an  overall  removal efficiency of 97 percent in the combined
 acidulator and stripper.

      The acidulated and stripped liquor was concentrated to about 45
 percent  by weight as  ammonium sulfate and then cooled to near 100°F to
 precipitate crystalline ammonium sulfate.  However, as noted earlier,
 the ammonium sulfate  could not be separated because of blinding of the
 equipment by extremely fine crystals of ammonium sulfate and ferrous
 ammonium sulfite.
     Double Alkali Tests:  Production of liquors with low C's is
uneconomical for use in the ammonium bisulfate regeneration process
because of the energy requirements to concentrate the solutions in the
regeneration sections.  However, scrubbing with solutions of low C's
has possible application in the double alkali process being considered
as a backup system for TVA's Widow's Creek Power Plant located in
northeast Alabama.  In the double alkali process based on sodium as
the absorbent, the concentration of sodium in the absorber loop is
limited to about 0.17 mole per 100 moles of water because of solubility
limitations in the regeneration section.  Ammoniacal solutions, which
are not limited to low concentrations in the regeneration loop, would
offer economic advantages because the size of the regeneration equipment
could be reduced.  When using an ammoniacal solution with a C of 0.5,
the regeneration equipment size could be reduced to about one-third
that required when scrubbing with sodium solutions.

     A test series was made to determine whether a plume would be
present under scrubber conditions similar to the proposed Widow's Creek
operation.  The test series was made using the flow configuration shown
in Figure 4.  Water was used on the bottom stage for humidification and
fly ash removal.  Fresh makeup water was added to this loop at the rate
required to maintain the temperature of the gas leaving the bottom (G-l)
stage at 120°F.  Forward feed liquor (ammoniacal solution) was pumped
to the middle (G-2) stage at 30 gal/min on a once-through basis and
then drained to the product liquor hold tank.  In one test, two-stage
scrubbing was used to increase S02 removal.  The forward feed liquor
was pumped at a rate of 30 gal/min to each of the middle and top stages.
The gas leaving the scrubber was reheated to 150, 175, and 200°F during
most of the test series by direct heating with a propane torch.

                                     976

-------
 CHEVRON-TYPE
MIST ELIMINATOR
      G-3
  (MARBLE BED)
       G-2
  (MARBLE BED)
       G-l
  (MARBLE BED)
   INLET GAS
                                                  EXIT GAS TO ATMOSPHERE
         MAKE-UP
          WATER
                              REHEAT

                             «
ONCE-THROUGH
   LIQUOR
     TO
   STORAGE
       TO SEWER
Figure 4.  Flow configuration for once-through liquor on G-2 and water on G-1,
                              977

-------
     A steam plume, as well as a minor residual particulate plume, was
present when scrubbing flue gas with water only (Figure 5).   The
steam component of the plume was dissipated when the exit gas was
reheated to 175°F, but a light particulate plume remained (Figure 6).
An additional plume component was present when an ammoniacal solution
was used.  The plume produced when scrubbing with a solution having a
C of 0.5 or 1.0 was reduced to an opacity of 10 percent or less, below
the maximum specification of 20 percent, by reheating the stack gas to
175°F.  Figure 7 shows the severe plume while scrubbing with a solution
having a C of 1.0 and with no gas reheat.  Figure 8 shows the reduced
plume from this operation while reheating to 175°F.  The SC>2 removal
efficiencies varied from 75 to 85 percent during these tests.

     A second absorber stage was added in one test sequence to increase
S02 removal efficiency.  An average removal of 92 percent resulted when
operating with solutions having a C of 1.8 to both scrubbing stages.
Reheating to 200°F was required in this test sequence to lower the
plume opacity to 20 percent or less (Figure 9).

     From these tests it was concluded that the plume produced when
scrubbing with solution having low ammonia content is significantly
lower than the plume produced during routine scrubber operation.  Also,
the plume produced under these low salt conditions could be  made
acceptable by reheating the exit gas to temperatures necessary to
dissipate the steam component of the plume.  This temperature will
vary according to the ambient air conditions—temperature and relative
humidity.

     Modified Operation:  The absorption section of the ammonia scrubbing-
ammonium bisulfate regeneration process pilot plant was operated from
April 16 to May 4, 1973, to study additional methods of minimizing the
plume from the absorber.  Two basic flow configurations were used:  a
recirculation system using one, two, or three stages of absorption
(Figure 10); and a once-through system using one stage of absorption
(Figure 11).  This operation was carried out in cooperation  with Air
Products and Chemicals, Inc.,  Allentown, Pennsylvania.

     The tests were made to determine whether the plume leaving the
absorber could be reduced to an acceptable level by operating under the
following conditions:

          • Insulated absorber

          • A water wash ahead of the first absorber stage

          • A water wash after the absorber stage

          • Reheating the scrubbed flue gas

                                   978

-------
Figure 5.  Plume emitted while scrubbing with water only; no reheat.

-------
kfi
oo
o
                                  Figure 6. Particulate plume emitted while scrubbing with water only; 175° F reheat.

-------
to
00

                                    Figure 7.  Plume from operation with scrubber solution having C =1.0; no reheat.

-------
1C
DC
                                  Figure 8.  Plume from operation with scrubber solution having C= 1-0; 1750 F  reheat.

-------

Figure 9.  Plume  from operation with solution having C= 1.8; two scrubber stages; 200" F reheat.

-------
                                                  EX IT GAS TO ATMOSPHERE
 CHEVRON-TYPE
MIST ELIMINATOR
      G3
  (MARBLE BED)
      G-2
  (MARBLE BED)
       G-l
   (KOCH TRAY)
   INLET GAS
                              PRODUCT
TO SEWER
            Figure 10. Normal recirculating flow configuration.
                                      984

-------
                                                 EX IT GAS TO ATMOSPHERE
 CHEVRON-TYPE
MIST ELIMINATOR
       G-3
  (MARBLE BED)
       G2
  iMARBLE BED)
       G-l
   (KOCH TRAY)
   INLET GAS
                                   ONCE THROUGH
                                      LIQUOR
                                       TO
                                     STORAGE
         MAKE-UP
          WATER
       TO SEWER
TO SEWER
                Figure 11.  Once-through flow configuration.
                                    985

-------
     An  acceptable plume for the pilot plant was defined as one having
 a maximum of  5 percent opacity.  The opacity readings were made by
 qualified observers who successfully completed the EPA-prescribed
 visual emissions school.  The absorber and all downstream ductwork
 were insulated for reasons that Air Products and Chemicals, Inc.,
 consider proprietary.  The water wash ahead of the first absorption
 stage was used to humidify and cool the hot flue gas; to remove
 chlorides and, thereby, prevent formation of an ammonium chloride
 plume; and to remove fly ash.  It was expected that cooling the flue
 gas would reduce the ammonia-based plume containing sulfur.  The
 moving marble bed of the bottom stage (G-l) used in previous work was
 replaced with a multi-venturi FlexiTray manufactured by Koch Engineering
 Company, Inc.  No downtime was required when this type of tray was
 deactivated during absorber operation.  (The glass marbles had to be
 removed  from the marble bed scrubber when the unit was deactivated to
 prevent  damage from thermal shock should the scrubber liquor accidentally
 come in  contact with the hot bed.)

     The water wash after the absorption stage was to decrease the salt
 content  of the entrained mist leaving the absorber.  Reheating would
 have evaporated the mist and produced an increased concentration of
 ammonia  and sulfur dioxide in the exit flue gas which could have then
 reacted  to form a plume.  The top stage of the absorber (G-3) was used
 for this wash.  The gas leaving the'absorber was directly reheated with
 a propane torch to destroy the particulate matter present and eliminate
 a steam  plume.  The retention time of the heated flue gas in ductwork
 was 1.5  seconds.

     A summary of the test conditions and the opacity readings for each
 test are given in Table III.  These tests show that an acceptable plume
 was obtained during production of liquor with high C's when:

          • Water wash was used ahead of the first absorber stage.

          • Absorber and all ducts were insulated.

          • Reheat was applied as required to dissipate the steam
            plume.

     Comparison of Figures 12 and 13 shows the effect of the
water wash ahead of the first absorber stage.  In Figure 12, the water
wash was not employed and the opacity was 5 percent.  The water wash
was activated and the opacity dropped to 0 percent (Figure 13).

     The concentration of ammonia salts in the water wash (G-3) after
the absorption stage did not show any significant effect upon the plume
formed in the range studied (C = 0.1 to 3),


                                  986

-------



Summary
TABLE
III
of Operating Data for

the




Test Series, April - May 1973
Test No.
Date
Time
Liquor flow configuration
Flow
Liquor, gaL/m'.n
To G-l (bottom stage)
To G-2 (middle stage)
To G-3 (top stage)
Gas, ft3/miu
Liquor Ca
G-l-lb
G-l-OC
G-2-lb
G-2-OC
G-3-Ib
G-3-Oc
Liquor S/C^
G-l- fb
G-1-0C
G-2-fb
G-2-Oc
G-3-Ib
G-3-OC
Liquor pll
G-l-lb
G-1-0C
G-2-Ib
G-2-OC
G-3-fi
G-3-Oc
SO,, ppm
Entering G-l
Leaving G-l
Leaving G-2
Leaving G-3
Percent removal
KH3 leaving G-3, ppm
Temperatures, °F
Liquor
From G-,1
From G-2
From G-3
Gas
To G-l
From G-l
From G-2
From G-3
Exit gas
Ambient
Relative humidity, %
Predicted temperature at
which steam plume forms,
Reheat temperature, °F
Percent opacity
0
4/19
11:00 a.m.
•4— •


30
30
15
2700

28.1
-
9.52
9.48
1.35
1.88

0.76
-
0.58
0.59
0.62
O.dO

5.90
-
6 , 70
b,70
0.70
6.80

2600
1700
80
40
98.5
47


141
120
125

282
144
130
126
200
73
80

"Fe 152
200
25
1-1
4/20
1:00 p.m.

I-2A
4/20
1:45 p.m.
> Da*-* i T»r-i 1 1 n
I-2B
4/20
2:15 p.m.

I-2C
4/20
3:15 p.m.

-o

30
30
15
2700

13.21
13.09
4.90
4.73
-
0.10

0.82
0.83
0.55
0.57
-
0.86

5.75
5.75
7.25
6.95
7.20
7.10

2320
1860
400
100
95,6
47


131
127
126

290
132
127
125
150
84
49

120
150
65

33
30
15
2700

12.6
13.0
5.2
5.2
0.1
0.6

0.85
0.84
0.55
0.57
0.84
0.58

5.80
5.75
7.10
6.95
7.25
7.15

2440
2000
300
40
98.4
51


131
127
126

289
133
127
125
200
85
47

120
200
15

31
30
15
2700

14,0
13,8
4,9
5.0
0.1
0.4

0.81
0.82
0.56
0.57
0.85
0.61

5.80
5.75
7.10
6.95
7.25
7.15

2600
1880
16C
60
97.7
51


130
127
126

286
132
126
125
220
85
47

120
220
20

31
30
15
2700

.
_
_
.
_
.

_
.
.
.
.
.

5.80
5.90
6.80
6.70
6.50
6.50

2400
2040
160
60
97.5
51


131
126
125

290
133
127
125
225
82
42

124
22S
10
I-2D
4/20
3:40 p.m.
•^


30
30
15
2700

13.0
13.6
4.5
4.5
1.2
1.4

0.79
0.79
0.60
0.62
0.72
0.70

5.80
5.90
6.80
6.70
6. SO
6. SO

2400
1640
240
160
93.3
SO


131
128
125

290
133
127
126
200
84
42

120
200
15
 w>les of active ammonia as  ammonium sulfite-bisulfite per 100 moles of water.
blnlet sample.
C0utlet sample.
aMole ratio of S02:NH3.
Calculated temperature  at which steam plume  can  form regardless of dilution; does
 not include a solid particulate plume, that  is,  an ammonia salt plume.

                                           987

-------
                                          TABLE  in (continued)
Summary 'of
Operating
Test Series, April
Test No.
Date
Time 5
Liquor flow
1-3
4/20
:5S p.m.

1-4
4/20
6:05 p.m.

configuration 4 — Recirculating— *
Flow
Liquor, gaL/ndn
To G-l (bottom stage)
To G-2 (middle stage)
To G-3 (top stage)
Gas, ft3/nin
Liquor Ca
G-l-lb
G-l-OC
G-2-lb
G-2-Oc
G-3-Ib
G-3-Oc
Liquor S/Cd
G-l-lb
G-1-0C
G-2-lb
G-2-Oc
G-3-Ib
C-3-QC
Liquor pH
G-l-Ib
G-l-OC
G-2-Ib
G-2-Oc
G-3-Ib
G-3-Oc
S02, ppm
Entering G-l
Leaving G-l
Leaving G-2
Leaving G-3
Percent removal
NH3 leaving G-3, ppm
Temperatures, °F
Liquor
From G-l
From G-2
From G-3
Gas
To G-l
From G-l
From G-2
From G-3
Exit gas
Ambient
Relative humidity, %
Predicted temperature
at which steam plume
forms, °Fe
Reheat temperature, °F
Percent opacity


29
30
15
2700

0.85
1.50
10.6
10.4
0.14
0.31

0.95
0.93
0.78
0.79
0.99
0.95

5,10
5.70
5.90
5.90
5.70
4.90

2520
1760
920
920
63. S
87


12J
129
122

270
124
129
121
160
83
42


124
160
Nil


30
30
15
2700

0.85
1.50
10.6
10.4
0.14
0,31

0.95
0.93
0.78
0.79
0.99
0.95

5.10
S.70
5.90
5.90
5.70
4.90

2520
1760
920
920
63. S
-


121
129
122

270
124
129
121
-
83
42


124
None
5
1-5
4/23
2:15 p.m.





31
31
15
2700

0.30
0.50
10.1
10.1
0.21
0.38

0.93
0.88
0.77
0.78
0.72
0.81

4.70
5.90
6.00
6.00
4.70
5.00

2560
2600
880
680
73.4
47


125
130
130

290
127
130
125
ISO
72
84


159
150
20
Data for the
- May 1973
I-6A
4/23
1:45 p





32
32
15
2700

0.30
0.50
10.1
10.1
0.21
0.38

0.93
0.88
0.77
0.78
0.72
0,81

4.70
5.30
6.00
6.00
4.70
5,00

2480
2520
880
720
70.9
47


127
ISO
127

290
128
130
Z25
200
70
84


160
200
5
1-7
4/23
.m. 2:40 p.n.





0
30
IT
2700

.
_
9.1 '
8.9
' 0.30
0.41


_
0.77
0.79
0.93
0.90

_
_
5. SO
5.80
5.10
4.20

2480
2480
020
780
68.5
79


_
128
124

290
274
138
124
150
72
80


154
ISO
30
I-8A
4/23
3:05 p.m.





0
30
17
2700

_
_
9.1
8.9
0.30
0,41



0.77
0.79
0.93
0.90



5.80
5.80
5.10
4.20

2480
2480
920
780
68.5



_
128
124

288
285
128
124
200
72
SO


154
200
5
1-13
4/24
3:15 p.m.

fc



30
30
15
2700

0.12
0,11
1.9
1.9
0.33
0.29

0.50
0.70
0.76
0.81
0.78
0.91

2.40
2.30
6.00
5.80
3.10
3.10

2480
2520
480
560
77.4
37


125
126
125

288
125
126
125
150
65
93


181
150
90/15 at lip
aMoles of active ammonia as ammonium sulfite-bisulfite per  100 moles of water.
^Inlet sample.
C0utlet sample.
dMole ratio of S02:NH3.
Calculated temperature at which steam plume  can  form regardless of dilution; does
 not include a solid participate plume,  that  is,  an ammonia salt plume.
                                               988

-------
                                         TABLE III (continued)


Summary of
Operating Data for the
Test Series, April - May 1973
Test No.
Date
Time 3
Liquor flow
configuration 4
Flow
Liquor, ga.l/min
To G-l (bottom stage)
To G-2 (middle stage)
To G-3 (top stage)
Gas, ft3/min
Liquor Ca
G-1-IU
G-1-0C
G-2-Ib
G-2-Oc
G-3-Ib
G-3-Oc
Liquor S/Cd
G-l-Ib
G-1-0C
G-2-Ib
G-2-QC
G-3-Ib
G-3-QC
Liquor pH
G-l-Ib
G-l-0=
G-2-Ib
G-2-Oc
G-3-Ib
G-3-QC
SO,, ppm
Entering G-l
Leaving G-l
Leaving G-2
Leaving G-3
Percent removal
NH- leaving G-3, ppm
Temperatures, °F
Liquor
From G-l
From G-2
From G-3
Gas
To G-l
From G-l
From G-2
From G-3
txit gas
Ambient
Relative humidity, %
Predicted temperature
at which steam plume
forms, *F«
I-14A
4/24
:45 p.





30
30
IS
2700

0.12
0,11
1.9
1.9
0.33
0.29

0.50
0.70
0.76
0.81
0.78
0.91

2.40
2.30
6.00
5.80
3.10
3.10

2600
2600
440
560
78.5
39

125
126
125

286
126
127
126
200
66
93

181
200
Reheat temperature, °F 0
Percent opacity
aMoles of active ammonia as
Inlet sample.
C0utlet sample.
"We ratio of S02:NH.
era 1 MI 1 a t «d temnerat 111


,.
re &t •
1-15
4/24
m. 4:30 p.m. 4





0
30
15
2700

_
1.8
1.7
0.23
0.24

_

0.78
0.86
0.95
0.95

_
_
6.00
5.70
3.10
2.90

2600
2540
440
600
76.9
34

_
128
126

286
268
127
125
150
66
92

181
150
80/30
i8T. Up
I-16A
4/24
:10 p.m.

	 >•


0
30
15
2700

.
1.8
1.7
0.23
0.24

_

0.78
0.86
0.95
0.9S

_
^
6,00
5.70
3.10
2.90

2600
2540
580
64
75.4
34

.
126
125

284
255
127
124
200
66
92

181
20(1
5
1-17
4/26
3:05 p.m.





30
30
25
3000

0.03
0.03
11.6
11.8
0.29
0.38

_

0.74
0.75
0.93
0.93

2.30
2.20
6.00
5.90
5.30
5. .50

2620
2440
600
680
74.0
89

124
128
124

280
124
128
124
200
60
96

198
200
40/5
aTlio
ammonium sulfitc-bisulfite per



jhieh steam nil



LUBA can \



rrvrm raaar
1-17 1-18
4/27
1:45 p.m. 1


S/l
:00 p.m.

» m « f.
1-18
5/2
4:00 p.m.

^



31
30
26
3000

0.03
0.03
11.6
11.8
0.29
0.38

_

0.74
0.75
0.93
0.93

2.30
2.20
6.00
5.90
5.30
5.50

2840
2560
680
400
85.9
70

120
128
121

272
120
128
120
160
57
51

170
160
15/10
aTliD
100 moles of



dlnss of dill


30
30
25
2900

_
11.62
11.72
2.06
2.39

.

0.75
0.75
0.93
0.89

2.20
2.20
6.00
5.90
5.60
5.60

2640
2480
560
440
83.3
93

125
134
129

295
125
133
127
200
74
64

141
200
10/5
aT lip
water.



nt ion* dot


30
30
25
2700

0.04
0.04
7.61
8.39
2.94
3.21

_

0.82
0.79
0.95
0.92

2.40
2.30
5.80
5.70
S.40
5.40

2520
2400
560
800
68.3
85

110
125
122

294
115
124
121
160
80
56

131
160
5




is
not include a solid particulate plume, that is, an ammonia salt plume.
                                        989

-------



.
                                       Figure 12.  Operation with no water wash ahead of first absorber stage.

-------
c
-
                                        Figure 13.  Operation with water wash ahead of first absorber stage.

-------
     The  chloride concentration of the inlet flue gas was approximately
 30 ppm and  in the exit gas was about 3 ppm.  The effect of chloride
 removal on  plume reduction could not be separated from the effects of
 cooling and humidifying the gas ahead of the first absorber stage.

     Heat losses were minimized by insulating the absorber system.  The
 difference  between the liquor temperature and the outside absorber skin
 temperature averaged about 1 °F.  In operation without insulation,
 differences as great as 30 °F have been measured.  No definite conclu-
 sions were  drawn about how effective the insulation was in preventing
 plume formation.

     The reheat required to dissipate the steam plume is a function of
 the ambient temperature and relative humidity.  The reheat temperature
 required  to avoid formation of a steam plume can be predicted for a
 given gas composition and known ambient conditions.9  The predicted
 reheat temperature for each test is included in Table III.  Reheating
 the gas to  a temperature above the predicted reheat temperature resulted
 in an acceptable plume when a water wash was used ahead of the absorber.
 In one test (1-4), an acceptable plume was obtained without reheating
 (Figure 14).  In this case, the gas leaving the absorber had a higher
 temperature than the predicted reheat temperature.  In those tests
 without a water wash, reheating decreased the plume opacity and in two
 tests (1-8  and 1-16)  the decreased opacity reached an acceptable level.

     Tests  1-17 and 1-18 were extended runs designed to show the reheat
required as a function of the ambient conditions (temperature and
 relative humidity).  The plume opacity was maintained at a constant
 5 percent by adjusting the reheat temperature of the scrubbed flue gas
 (maximum temperature set at 200°F).   Data from these tests show that
 the reheat requirements for an acceptable plume increase with increased
relative humidity.  These data, as well as the predicted reheat tempera-
 ture necessary to avoid formation of a steam plume and the opacity
reading of the stack,are given below.
                      Predicted Reheat
                       Temp Required                 Opacity Reading, %
 Relative     Ambient   to Eliminate     Reheat        At        10 Feet
Humidity. %  Temp, °F  Steam Plume. °F   Temp,  °F  Stack Exit  Above"Stack

    94          62           189           125         50          60
                                       (No Reheat)
    94          62           189           200          5          30
    80          65           166           193          5          10
    69          63           162           196          5          10
    62          68           152           197          5           5
    53          57           170           180          5           5
    42          60           154           175          5           5
    32          82           124           158          5           5

                                  992

-------
\\       V    '
       r"    ,
                            Figure 14. Acceptable plume with no reheat.

-------
    Formation of a high-opacity plume several feet from the discharge
of the stack occurred on days when the relative humidity was high (see
above data).  Figure 15 shows a plume reforming downwind from a clear
stack.  Reheat temperatures required to avoid formation of a plume on
days of high relative humidity and low temperature ar.e impractical
to achieve.

Conclusions

    The ammonia bisulfate process is a promising candidate for second
generation SC^ removal systems.  The following conclusions are
reasonable for the bisulfate process:

         • Under proper operation of the scrubber, fume formation
           can be controlled in the scrubber while producing a
           liquor having a high salt concentration (C > 12).

         • For low salt concentrations, there is much greater
           flexibility in the manner of scrubber operation.
           However, such low concentration is prohibitively
           expensive in the bisulfate process, but very reasonable
           in the double alkali approach.

         • Avoiding steam plume formation outside the stack on days
           of high relative humidity and low temperature may be
           impractical to achieve.

         • Adequate separation of fly ash  and ammonium sulfate
           crystals has not been effective in the pilot plant
           equipment;  with proper equipment, adequate separation
           is expected.
                                 994

-------
-

                                                Figure 15.  Plume reforming downwind of stack.

-------
                            REFERENCES


    1.   Lepsoe, R. and  Kirkpatrick, W.S.  "SO- Recovery at Trail ,"
        Trans. Can.  Inst. Mining Met. XL, 399-404  (1937).

    2.   Hein,  L.B.,  Phillips, A.B., and Young, R.D.   "Recovery of
        SCU  from Coal Combustion Stack Gases."   In Problems and
        Control of Air  Pollution (Frederick S. Mallatte, ed) ,
        Reinhold, New York  (1955) pp. 155-69.

    3.   Sulfur  80 (1),  36-37  (Jan. -Feb.  1969).

    4.   Rumanian Ministry of Petroleum Industry  and Chemistry.
        "Ammonium Sul fate"  Brit. Pat. 1,097,257  (Jan. 3, 1968).

    5.   Johnstone, H.F.  "Recovery of S02 from Waste  Gases."
        Ind. Eng. Chem.  29  (12), 1396-98  (Dec. 1937).

    6.   Alabama Power Company, "New Process of Fertilizer Manufacture
        Announced."  Mfr. Rec. 92(26), 53 (Dec.  29, 1927).

    7.   Ruben, Allen G.  (Bohna Engineering $ Research Inc.).  Private
        Communication.

    8.   Hixson, A. W. and Miller, R.  "Recovery  of Acidic Gases."   U.  S.  Pat
        2,405,747 (Aug.  13, 1946).

    9.   EPA  study by L.  I.  Griffin, Jr.,  (to be  published).


           Conversion of English Units to Metric Equivalents


  Multiply                                                    To Obtain
English Unit                      B^                      Metric Equivalent
   ft /min                     0.0283                          m3/min

   lb                          0.454                           kg

   °F                    °C = |- (°F - 32)                      °C
        3                                                         3
   gr/ft                       2.288                           g/m

   in.                          2.54                            cm

   ft                          3.281                           m

   gal/rain                     3.785                           1/min

                                    996

-------
AN  EPA OVERVIEW OF SODIUM-BASED DOUBLE ALKALI PROCESSES
               PART I.  A VIEW OF THE PROCESS
         CHEMISTRY OF  IDENTIFIABLE AND ATTRACTIVE
                           SCHEMES
                              by

                         Dean Draemel
                        Research Branch
                   Control Systems Laboratory
             National Environmental Research  Center
                Office of Research and Monitoring
            U.  S. Environmental Protection Agency
            Research  Triangle Park, North Carolina
                               997

-------
                               INTRODUCTION

      "Double" or "dual" alkali scrubbing involves circulating a clear liquor
solution of a soluble alkali salt (Na, K or NH-), with scrubbing taking place
by absorption and reaction to form the bisulfite from the sulfite.  The spent
scrubbing liquor is treated with limestone and/or lime which precipitates the
absorbed sulfur as CaSO^ (and possibly some CaSOJ and regenerates the alkali
scrubbing solution.  Refinements may involve additional regeneration of the
sulfate which is formed by oxidation.
      Double alkali flue gas desulfurization processes have received increased
attention in recent years because of some potential advantages over competing
scrubber systems.
     The circulation of a clear liquor removes many errosion, corrosion and sol it
deposition problems.  The regeneration using limestone and/or lime is relatively
cheap and simple and the solid formed is acceptable as a throwaway product.  The
system appears to be versatile in terms of modes of operation and may even be
used  to produce a salable product.  Both installed and operating costs for
such  systems appear to be very competitive with other flue gas desulfurization
systems at comparable levels of development.  '
     A number of organizations, including the Environmental Protection Agency,
have become involved with major double alkali development efforts.  General
Motors has been heavily involved in the development and design of double alkali
                                             i?\
systems for use in its industrial facilities/ '  Other major double alkali
development efforts have been conducted by the Tennessee Valley Authority,
Arthur D. Little, Inc., FMC Corporation, Envirotech Corporation and others.  The
                                   998

-------
Environmental Protection Agency 1s funding a contract with Arthur D.  Little,  Inc.
to aid the development of this promising flue gas desulfuHzatlon technology.
Major aspects of the process chemistry and the attractive operating schemes
which are under development by various organizations will be discussed 1n this
paper.  The particular systems under development by the various organizations
are discussed 1n greater detail 1n Part II of this paper.
     The mention of company or product names 1s not to be considered as
endorsement or recommendation for use by the Environmental Protection Agency.
                                     999

-------
                            PROCESS DESCRIPTION
Absorption
     A simple flow scheme for a double alkali process is shown in Figure 1.
The clear alkali solution is circulated through a scrubber where SOp is absorbed
and some oxidation takes place.  The type of scrubber used depends somewhat on the
concentration of the alkali scrubbing solution, but the clear scrubbing liquor
circulated allows flexibility in scrubber selection.  Some of the effluent liquor
from the scrubber is recirculated and the remainder is sent to the regeneration
sys tern.

Regeneration
     The liquor from the scrubber is treated with either limestone or lime
(Ca(OH)2) to precipitate the absorbed sulfur as calcium solids.  After separating
the precipitated solids, the regenerated scrubbing liquor, which has a high
sulfite to bisulfite ratio, is sent back to the scrubber loop.

Oxidation and Sulfate Control
     Oxidation of absorbed sulfur may require additional regeneration of the
sulfate formed.  A separate stream from the primary bisulfite/sulfite regeneration
system may be treated specifically for sulfate regeneration to remove sulfate from
the system as an insoluble solid product.
                                        1000

-------
FLUE GAS OUT
1
            1
   FLUE
    GAS
     IN
                                                             OPTIONAL SOFTENING
                                        I	
        EFFLUENT
    RECIRCULATIONTANK
                                                                            »
                                                                          SOLIDS
                                                                                                          OPTIONAL SULFATE
                                                                                                            REGENERATION


                                                                                                                  SOLIDS
                                                      REGENERATION


                                Figure 1.  Double-alkali process flow scheme.

-------
                                 PROCESS  CHEMISTRY

      Generally,  sodium-based  double  alkali systems are  the only systems receiving
 major development  emphasis.   Since sodium-based systems are under major active
 development  in this  country and  appear to have promise with respect to flue gas
 desulfurization, this paper will deal only with sodium-based systems.

 Absorption
      The  scrubbing step  involves absorption of SCL in a scrubbing solution of
 sodium sulfite,  bisulfite and sulfate.   The absorbed S02 reacts with the sulfite
 in  solution  to form  bisulfite (reaction  1).  The scrubbing solution is actually
 a buffer  system  of sulfurous  acid (reaction 2).  The pH range over which the
 scrubber  systems may operate  is  from * 6 to 8.5.  Figure 2 shows the major species
 in  solution  as a function of  pH.  Solutions of higher pH, with sulfite as the
 major species, are circulated into the scrubber while the lower pH (higher
 bisulfite concentration) spent solution  from the effluent redrculatlon tank Is
 sent  to the  regeneration system.
       Na^SO.,  +  SO,  +  H,0  	*   2 NaHSO,                              (1)
        t  J       c.      £.   \    *        j
 S02(aq) + H?0 -—* H2$03  —^   H+  +  HSOg —^   2H+  +  SO*           (2)

Regeneration
     Spent scrubbing solution, with a low sulfite to bisulfite ratio, which is
sent to the regeneration system is treated with limestone or lime to precipitate
the absorbed S02 as calcium solids by reactions 3 and 4.  The bisulfite is
neutralized and sulfite is  regenerated as the  "active" scrubbing agent.
                                       1002

-------
o
o
                                         2-   Distribution of aqueous sulfite species as a function of pH.

-------
    CaC03  +  2 NaHSOg  -—*  CaS03  +  Na2$03  +  H20  +  C02              (3)
    Ca(OH)2  +  2NaHS03 -—*•  CaS03  +  Na2S03  +  2 HgO                    (4)
      The EPA has  performed  batch  laboratory experiments to characterize reactions
 3  and 4 in simulated scrubber effluent solutions.  '  The results of these
 experiments are consistent  with laboratory and pilot plant results of other
 organizations involved in double  alkali development programs.  Hold times and
 utilizations using  limestone and  lime are mainly dependent on solution concentra-
 tions, temperature, agitation level and reactant stoichiometry.  Generally, using
 limestone, hold times on the order of an hour or more are needed and utilizations
 on  the order of 75-85% may  be realized.  Generally, using lime, hold times on
 the order of around 10 minutes may be used and utilizations of 90% or more may
 be  realized.  The benefits  of using lime for the bisulfite/sulfite regeneration
 may be offset by  higher chemical  costs, slaking requirements, scaling potential
 and possible pH control problems.
      In addition  to the bisulfite/sulfite regeneration, lime may be used to
 react further with the sulfite and sulfate present in solution by reactions
 5 and 6.  The reaction between lime -and sulfate is limited by the equilibrium
 hydroxide ion concentration of approximately 0.15M.* '
      Ca(OH)?  +  Na?S07  	*  CaSO,  +  2NaOH                             (5)
           £       £  0  ^~'"' • •      j
      Ca(OH)2  +  Na2S04  -—^  CaS04  +  2NaOH                             (6)
 Sulfite is the more active  species and if more hydroxide is formed by reaction
with sulfite (reaction 5) than the approximately 0.15M equilibrium value for
 reaction with sulfate (reaction 6), the reaction between lime and sulfate will
not occur.  The effective regeneration of sulfate using lime thus depends on the
                                     1004

-------
sulfite ion concentration present and the hydroxide ion concentration  formed
from reaction 5.  Batch scale laboratory experiments conducted by the
Environmental Protection Agency have indicated that increased sodium sulfate
concentrations tend to favor reaction with sulfate over sulfite when solutions
are treated with an equilibrium amount of lime.  With 1.78M (^ 20 wt?) sodium
sulfate, roughly 50% of the final (3-hour batch reaction) hydroxide is from
reaction with sulfate at 0.066M initial sulfite.  With 0.67M HO wt£) sodium
sulfate, roughly 50% of the final (3-hour batch reaction) hydroxide is from
reaction with sulfate at less than .055M initial sulfite.  Figure 3 shows the
relative amount of reaction between lime and sulfite and sulfate in solution
as a function of initial sulfite concentrations.  Batch reactions were conducted
at 52°C (125°F) with 1.78M (•>, 20 wU) sodium sulfate solutions and roughly
equilibrium  (0.078M/1) lime addition.  This indicates that lime can be used to
regenerate sulfate in the presence of dilute sulfite concentrations.
Oxidation and Sulfate Control
     Oxidation of bisulfite and sulfite  in the scrubber  liquors occurs by
reactions 7  and 8.  The sulfate formed by oxidation must be removed from the
system either as a solid product by- regeneration or as soluble salts  by a purge.

        2 NaHS03  +  1/2 02 -—* Na2S04 +  S02  +  HgO                  (7)
        Na,SO,  +  1/2 0,  _—* Na,S04                                   (8)
          L.  3          <_  V~~      L.  *r
     The effective removal of  sulfate as a solid product is highly  desirable
from the standpoint of environmental acceptability.  Completely  closed loop
operation is an ideal situation, while the opposite extreme would  Involve a
balance between sulfate  formed by oxidation and sulfate  lost  as  soluble salts
                                      1005

-------
o
o
en
U.1D

0.10
.", moles/1 itei
° 0.05

0
0
I

•• •*
i
j
j ;
- 7
//
«
.02 0.04
1
TOTAL OH'
	 . 	 — —
	 • 	
OH'FROMSOj2' ,
/OH"FROMS042' J
/ ii
i «
|l
li
0.06 0.
1
— 	 — 	 '
1
1
1
1
1
i !
i ii
08 0.10 .0.1
                                                    S032' CHARGED, moles/i900 cc


                            Figure 3.   Batch reactions-Ca(OH)2/Na2S03, Na2S04[equilibrium OH".

                            1.60 moles/900 cc Na2S04 @ 52° C (125° F)T,

-------
leaving the system with the solids or possibly as a simple purge.   Considering
the volume of sulfur removed and the fractional oxidation levels  expected,  the
loss of all oxidized sulfur from the system as soluble salts represents  a
serious water pollution potential.  Attempts must be made to regenerate  the
sulfate formed and remove it as an insoluble solid product.
     Oxidation by reactions 7 and 8 is a function of solution concentration,
oxygen mass transfer to the scrubbing solutions and probably a number, of other
factors such as traces of catalysts or inhibitors that may be present from sources
such as fly ash.  At least one oxidation inhibitor for this type of process is
                    (4\
presently available/ ;  The oxidation reactions are not well defined or well
understood.  So far, most efforts have attempted to cope with oxidation  problems
without developing a complete understanding of the factors involved.  The EPA
recently funded a grant (No. 800303) with the University of Illinois to  determine
the mechanisms and kinetics of oxidation in soluble alkali scrubbing systems.
     Sulfate regeneration with lime is possible in the presence of low sulfite
ion concentrations (<0.08M) subject to the limitation of the equilibrium hydroxide
ion concentration mentioned previously.  The low sulfite ion concentrations
necessary for effective sulfate regeneration using lime are a disadvantage because
of the large volume of the dilute scrubbing liquor necessary for effective S02
removal.  As an alternative to sulfate regeneration using lime, other sulfate
removal techniques are under development which may be used with concentrated
scrubbing liquors (sulfite concentrations of ^ 0.5M).
     Sulfate removal would generally be carried out by treating a side stream
from the scrubber liquor loop.  An example of this approach is shown by the
                    fa\
following reactions.v/
                                       1007

-------
    H2S04  +  CaS03 * 1/2 H20 + 1/2 HgO •* S02 + CaS04 '  2 HgO            (9)
S02 + CaS03 '  1/2 H20 + Na2S04 + 5/2 H20 + CaS04 '  2H20  + 2 NaHS03       (10)
     A side stream from the scrubber discharge is mixed  with calcium sulfite
solids and sulfuric acid is added.  The net effect  is that sodium sulfate
in the scrubbing liquor is converted to solid calcium sulfate and sodium
bisulfite is generated for recycle to the main regeneration system.   Additional
limestone or lime is required to remove the sulfur  added as sulfuric acid, i.e.
the bisulfite  generated by reaction 10.  Roughly 125-150% of the theoretical
sulfuric acid  is required, based on the amount of sodium sulfate reacted.  It
is estimated that with 7% oxidation, roughly 9% of  the solids produced would be
derived from the sulfuric acid.  This would add to  the operating cost of a
throwaway system.
     Another  alternative for sulfate removal being considered is the selective
crystallization of sodium sulfate decahydrate by cooling of scrubbing liquors,
in which the decahydrate will crystallize out at about 32°C (90°F).   The crystals
could then be  separated and somehow treated to regenerate active sodium and
precipitate a  relatively insoluble solid product.  Alternative methods of sulfate
control are being considered.

Scaling
     Scaling may occur if conditions develop in the scrubber which cause super-
saturation of calcium sulfate.  The solubility product,  K  , for calcium sulfite
is roughly 2 orders of magnitude less than the K   value for calcium sulfate
although steady state sulfite and sulfate concentrations may develop for which
either solid may precipitate.  Supersaturation and  solids precipitation is desired
                                    1008

-------
in the regeneration section but calcium 1on concentrations must be controlled
in the liquor returning to the scrubber to guard against supersaturation and
precipitation in the scrubber.  The relative solubilities of the various
calcium compounds are shown in Figure 4.
     The presence of solid calcium hydroxide and/or solid calcium sulfate at
equilibrium with the liquid in the regeneration system produces high calcium
ion concentrations on the order of 300-400 ppm.  These high calcium ion con-
centrations must be reduced considerably to guard against calcium sulfate
supersaturation and scaling in the scrubber.
     Calcium ion concentrationsmay be controlled by "softening" reactions
11 and 12.
           Na2C03  +  Ca(OH)2  -—*  2 NaOH  +  CaCOg                    (11)
           C02  +  H20  +  Ca"1"1'  -—*  CaC03  +  2H*                     (12)
     Treatment of scrubbing liquor with sodium carbonate (reaction 11) or
carbon dioxide (reaction 12) before being sent to the scrubber leads to precipita-
tion of calcium carbonate and thus removes calcium ions from solution.  The
reduction of calcium Ion concentrations in this manner ensures against scaling
in the scrubber.  Addition of sodium carbonate has the advantage of both
softening and replacing sodium losses from the system.  Addition of Na2C03 has
the disadvantage of requiring a possible sodium purge if the amount required
for softening exceeds the amount required to replace sodium losses.
     Addition of carbon dioxide, which is acidic, also requires Increased use
of regeneration chemicals for neutralization.  The net effect of adding either
sodium carbonate or carbon dioxide 1s to reduce calcium ion concentrations

                                      1009

-------
o
o
                          0.28
                          0.24
                          0.20
                       o
                       s
                          0.16
ro
>^"
                       5  0-12
                          0.08
                          0.04
                                                                                    i        i—r
             	Ca(OH)2
             — -CaS03»2H20
             -^— CaCCU
                                      I
                                                              I
                                     10
                      20
30
                       40       SO       60      70      80
                        TEMPERATURE,0 C
Figure 4.   Relative solubilities of calcium compounds.
90     100

-------
entering the scrubber to a level  below the saturation value for calcium
sulfate.
Process Chemistry—Summary
     Before going on to discuss possible modes of operation for sodium-based
double alkali systems, a brief summary of the process chemistry discussion
will be made.  Sulfur oxides are absorbed into a sulfite/bisulfite buffer solution
This absorption shifts the pH down and increases the bisulfite concentration.
The liquor from the scrubber may then be treated with limestone to precipitate
calcium sulfite and neutralize the bisulfite.  The liquor from the scrubber
or the limestone reaction vessel may be treated with lime to precipitate calcium
sulfite and possibly calcium sulfate.  Oxidation of absorbed sulfur requires
the regeneration of sulfate; a few possible methods have been discussed.  Calcium
ion concentrations in the scrubber can be controlled by softening steps in the
liquor loop.
                                      1011

-------
                       DOUBLE ALKALI OPERATING SCHEMES

     Five relatively distinct modes of operation have been identified for
sodium/calcium double alkali scrubbing systems.  These modes will be discussed
very briefly with some advantages and disadvantages noted for each.  More
detailed discussion of these modes, with emphasis on commercial developments,
is presented in Part II of this paper.  The five operating schemes to be
discussed are:
     1.  Limestone and lime regeneration, dilute active alkali, with sulfite
         softening.
     2.  Lime regeneration, dilute active alkali, with sulfite softening.
     3.  Lime regeneration, concentrated active alkali with side stream
         sulfate treatment (removal).
     4.  Limestone regeneration, concentrated active alkali with side
         stream sulfate treatment (removal).
     5.  Lime regeneration, dilute active alkali, with carbonate softening.

Limestone and Lime Regeneration, Dilute Active Alkali,with Sulfite Softening
     The first mode identified is a double-loop (limestone/lime), dilute alkali
system.  Dilute alkali indicates that the sulfite or alkali concentration entering
the scrubber will be less than approximately 0.08M.  Dilute alkali scrubbing
solutions allow sulfate regeneration with lime in this mode of operation.
"Double loop" refers to the regeneration method.  Spent scrubbing liquor is
treated with limestone to neutralize the bisulfite and precipitate calcium
sulfite.  A stream from this limestone reaction vessel is then treated with  lime
to precipitate calcium sulfite and sulfate and generate sodium hydroxide.
                                        1012

-------
The relative amount of reaction between lime and the sulfite and sulfate
is determined mainly by sulfite and sulfate concentrations and lime stoichiometry.
The liquor from the lime reaction vessel is returned to the limestone reaction
vessel to ensure unsaturated calcium ion concentrations going to the scrubber.
Desaturation is accomplished because of the presence of sulfite ion in the
limestone reactor.  The reduction of calcium ion concentrations by calcium
sulfite precipitation is referred to as sulfite softening.  Sodium carbonate
may be added to provide additional softening and make up any sodium losses.
     This scheme has the disadvantage of requiring relatively long hold times
for the reaction with limestone, requiring two major reactors, and having to
circulate large volumes of relatively dilute scrubbing liquor.  The advantages
appear to be the use of relatively inexpensive limestone and lime for regeneration
of both bisulfite and sulfate, and relatively simple equipment requirements.

Lime Regeneration. Dilute Active Alkali, with Sulfite Softening
     The second mode of operation is a double-loop (lime only), dilute alkali
system.  This system arrangement is very similar to the first system discussed.
Spent scrubbing liquor is treated with lime to neutralize the bisulfite only.
A stream from the "lime/bisulfite reaction vessel is then treated with additional
lime in a separate reaction vessel to react with sulfite and sulfate, forming
calcium solids and sodium hydroxide.  The liquor from this second reaction vessel is
returned to the lime/bisulfite reaction vessel to ensure unsaturated calcium ion
concentrations with respect to calcium sulfate by "sulfite"softening of the
liquors before entering the scrubber.
     This flow scheme has the disadvantage of difficult pH and calcium ion control
in the regeneration loop because of the addition of a relatively concentrated

                                      1013

-------
strong base (lime) to a dilute solution of a weak acid.  The 'regeneration
loop requires two reactors, and large volumes of relatively dilute scrubbing
liquor must be circulated to the scrubber.  The advantages of this system
appear to be the relatively short hold times necessary for regeneration using
lime and the relatively simple equipment requirements.

Lime Regeneration, Concentrated Active Alkali, with Side Stream Sulfate Treatment
     The third mode of operation involves lime regeneration, concentrated alkali
solutions and side stream treatment for sulfate removal and control.   Lime is
used for the bisulfite neutralization of the concentrated (^ 0.5M) alkali
scrubbing solutions.   A side stream is taken from the liquor loop and treated
specifically for sulfate removal.   The major difference between dilute (<0.08M)
and concentrated (% 0.5M) alkali systems is that lime cannot be used to regenerate
sulfate in concentrated (greater than ^O.OSMJ sulfite solutions because of the
equilibrium hydroxide ion concentration mentioned previously.  Side stream treatment
                                                                 (3)
has been accomplished with proven  technology (reactions 9 and 10)v '  but sulfate
removal and control is the subject of intense development efforts since proven
technology is limited in this critical area.
     This third flow scheme has the disadvantage of possible sulfate regeneration
complications.   The advantages of this flow scheme are:  the relatively short
hold times necessary for the bisulfite/sulfite regeneration With lime, high
reactant utilizations, an advanced state of development and high S02 removal
efficiencies with low flows of the concentrated alkali scrubbing liquors.  This
third mode of operation appears to be very promising but universal large-scale
successful application depends on  effective treatment for sulfate removal and
control.
                                     1014

-------
Limestone Regeneration, Concentrated Active Alkali,  with  Side  Stream  Sulfate
Treatment
     The fourth mode of operation involves limestone regeneration,  concentrated
alkali solutions and a side stream  treatment for sulfate removal  and control.
Limestone is used for the bisulfite neutralization of the concentrated (^0.5M)
alkali scrubbing solutions.  A side stream is taken  from  the liquor loop and
treated specifically for sulfate removal.  This scheme,  like the previous one,
has the possible disadvantage of sulfate regeneration complications.   This scheme
also has the disadvantage of requiring long hold times for the bisulfite/sulfite
regeneration using limestone.  Advantages of this flow scheme appear to be high
S0« removal efficiencies with low liquor flow rates  and high reactant utilizations,
Successful large-scale application of this flow scheme also depends on effective
treatment for sulfate removal and control.

Lime Regeneration, Dilute Active Alkali, with Carbonate Softening
     The fifth mode of operation is lime regeneration, dilute alkali  with
carbonate softening for calcium ion (scaling) control.  The spent scrubbing
liquor is treated with lime to neutralize bisulfite and react with sulfite and
sulfate.  The dilute active alkali allows simultaneous reaction with .both the
sulfite and sulfate by reactions 6 and 7 even though the equilibrium hydroxide
ion concentration for sulfate regeneration (reaction 7) limits the extent of the
reactions.  The calcium ion concentration in the liquor from the lime treatment
tank would be high and represent serious scaling potential unless it is reduced
considerably below saturation values for calcium sulfate and hydroxide.  Carbon
dioxide may be added to the system to precipitate calcium carbonate  (reaction 12)
and significantly reduce calcium 1on concentrations in the liquor before entering
the scrubber.  Sodium carbonate may be used to soften (reaction 11) and make
                                      1015

-------
up sodium losses.  This scheme has the disadvantages of possible Inadequate
sulfate regeneration, possible scaling potential and having to circulate large
volumes of relatively dilute scrubbing liquors.  The advantages appear to be
the ability to deal with high oxidation levels, relatively simple equipment
requirements and an advanced state of development.
     Modes 2 and 5 are very similar except the double-loop regeneration method
of mode 2 would probably provide both sulfate regeneration and calcium ion
control without a major softening step involving the addition of chemicals
such as sodium carbonate or carbon dioxide.  All advantages and disadvantages
discussed serve mainly to compare these five modes relative to each other and
not to other scrubbing systems.  The advantages and disadvantages of these double
alkali systems relative to other scrubber systems have been discussed in the
introduction.
                                      1016

-------
                                 SUMMARY
     The major process chemistry for sodium/calcium double alkali  scrubbing
of SO  1s well defined and reasonably well understood.  A number of relatively
     A
distinct modes of operation have been Identified and are at various stages
of development.  Development efforts are presently being directed at effective
sulfate regeneration In concentrated sulflte solutions, control  of scaling, and
control and understanding of the oxidation reactions.  SolIds characteristics
such as settleabllity, cake moisture and cake "fixing" are also being studied.
Many minor variations are possible in the flow scheme configurations discussed;
however, the description of these schemes was meant mainly to show the versatility
of double alkali systems and the range of development efforts.
                                     1017

-------
                                REFERENCES

1.  Rochelle, G., Economics of Flue Gas Desulfurization, paper,  Flue Gas
    Desulfurization Symposium, May 14-17, 1973, New Orleans, La.

2.  Phillips, R. J., Sulfur Dioxide Emission Control  for Industrial
    Power Plants, paper, Second International Lime/Limestone Wet Scrubbing
    Symposium, Nov. 8-12, 1972, New Orleans, La.

3.  Draemel, D.  C., Regeneration Chemistry of Sodium-Based Double-Alkali
    Scrubbing Process, Environmental Protection Technology Series,
    EPA-R2-73-186, March 1973.

4.  Elder, H. W., Princiotta, F. T., Hollinden, D.  G.  A. and Gage,  Dr.  S. J.,
    Sulfur Oxide Control Technology Visits in Japan -  August 1973,  U. S.
    Government Interagency Report, Muscle Shoals,  Ala.,  Oct. 30, 1972.
                                  1018

-------
   AN EPA OVERVIEW OF SODIUM-BASED
       DOUBLE ALKALI PROCESSES

                 PART II

STATUS OF TECHNOLOGY AND DESCRIPTION
        OF ATTRACTIVE SCHEMES
                   by

              Norman Kaplan
        Control Systems Laboratory
     Office of  Research and Monitoring
      Environmental Protection Agency
  Research Triangle Park, North Carolina
                   1019

-------
                              ABSTRACT
                   AN EPA OVERVIEW OF SODIUM-BASED
                       DOUBLE ALKALI PROCESSES

                 PART II.  STATUS OF TECHNOLOGY AND
                  DESCRIPTION OF ATTRACTIVE SCHEMES
     Important criteria for evaluation of double alkali schemes are
given.

     Flow sheets for potentially attractive schemes of operation
are presented and described with emphasis on evaluation of each with
respect to the criteria given.

     A brief description of selected pilot plant and prototype
experience and future plans for further development of double alkali
systems is given.  In addition, an effort is made to categorize each
of the known commercially available double alkali systems presently
being marketed, in accordance with the identified schemes.

     The EPA/A. D. Little double alkali development program plan and
general philosophy are described.

     Double alkali technology, although less advanced, is presented
as being potentially more reliable and less costly than lime/limestone
technology.
                                 1020

-------
                          ACKNOWLEDGEMENTS
     The author wishes to express appreciation for assistance in
preparation of this paper to Frank T. Princiotta, Chief, Engineering
Test Section, for assistance with technical presentation, to
Charlotte Bercegeay for typing and to Beverly Tilton for assistance
in preparation of diagrammatic material; all of these personnel are
assigned to Environmental Protection Agency components in the
Research Triangle Park area of North Carolina.

     Appreciation is also expressed toward the many personnel repre-
senting Arthur D. Little, Inc., Chemical Construction Corporation
(Chemico), Combustion Equipment Associates, Envirotech Corporation,
FMC Corporation, General Motors Corporation, The Southern Company,
Utah Power S Light, and Zurn Industries  (Zum Air Systems) for their
informative input and cooperation, without which a presentation of
this type would not be possible, and to  Dr. Ando for his summary of
some of the technology developed by Japanese companies.
                                  1021

-------
                                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, NERC/RTP is providing con-
    version factors for the particular non-metric units used in the
    document.  Generally,  this paper uses British units of measure.


    For conversion to the  Metric system, use the following equivalents:
         British

         5/9 (°F-32)

         1 ft
         1 ft3

         1 grain

         1 in.

         1 in. 2

         1 in.3

         1 Ib (avoir.)

         1 ton (long)

         1 ton (short)

         1 gal
Metric

°C

0.3048 meter

0.0929 meters2

0.0283 meters3

0.0648 gram

2.54 centimeters

6.452 centimeters2

16.39 centimeters3

0.4536 kilogram

1.0160 metric tons

0.9072 metric tons

3.7853 liters
                                 1022

-------
                     INTRODUCTION AND BACKGROUND
     Based on presently known firm commitments of utilities,  the prevalent
method of stack gas cleaning for reduction of sulfur oxide emissions  fr
fossil fueled power plants is the wet lime/limestone scrubbing process.
Unfortunately, this method of control is not the perfect solution to  the
problem.  Problems have been identified with wet lime/limestone scrubbing
of flue gas which are mainly due to the fact that the absorbent and the
products of the absorption reaction are present in the scrubbing equipment
in the form of a slurry which has a tendency to cause solids  build-up,
scaling and erosion in the equipment; these can reduce the overall relia-
bility of the system.

     In order to circumvent the problems inherent with the use of an
absorbent slurry, scrubbing with a soluble alkali is a logical first
consideration; however, by itself, this method is not satisfactory since
it would necessitate the disposal of large quantities of liquid containing
dissolved sulfur compounds, thus creating a water pollution problem.   The
"Double Alkali" process, on the other hand, combines the advantages of
wet lime/limestone scrubbing with those of soluble alkali scrubbing by
scrubbing with soluble alkali and following with regeneration of the
soluble alkali and precipitation of the absorbed sulfur oxides as insoluble
calcium salts.  The regeneration reaction involving the use of calcium
compounds, which are considered to be potentially scale producing, is caused
to occur outside the scrubber system where scaling tendencies can be  more
adequately controlled.  In addition--whereas with lime/limestone systems,
S02 removal is limited by solids dissolution rate--no such limitation is
inherent with soluble alkali scrubbing systems.  The net effect of the use
of a double alkali process is the same as that of a wet lime/limestone
process in that limestone and/or lime is consumed and calcium sulfite and
sulfate are generated as a disposable waste product.  An important excep-
tion to the previous statement is the fact that the disposable waste may
be moderately contaminated with soluble salts which can potentially cause
a water pollution problem.

     The term "double alkali" is used generally to describe systems which
employ a soluble aqueous alkali (e.g., Na+, K+, NH4+ based) to scrub
acidic sulfur oxides from flue gas and then produce an insoluble throwaway
product by reacting the scrubber effluent with limestone and/or lime.
This paper will address only sodium-calcium systems as these are the
prevalent schemes under study and commercially available in the United
States.  An overview of these systems is presented with only some reference
to the chemical complexities associated with them.  The chemistry of these
systems has been discussed in detail in part I of this paper; in summary,
however, the principal scrubbing reaction in the many variations of the
double alkali process discussed involves the absorption of sulfur dioxide
by an aqueous solution of sodium sulfite, converting the sulfite to
bisulfite.

                                1023

-------
                      IMPORTANT CONSIDERATIONS
     In the discussion of various double alkali schemes, it is considered
helpful to compare these in relation to lime/limestone scrubbing systems.
The double alkali process has been considered a second generation version
of the wet lime/limestone process and therefore comparison is considered
reasonable.

     Sulfate Removal
     As in the operation of a lime/limestone system, some of the sulfur
absorbed in double alkali processes appears in the system in the oxidized
form as sulfate.  In order to operate a double alkali system in a steady
state continuous manner, provisions must be made to remove sulfate from
the system at the rate at which it is formed in or absorbed into the
system.  This is not the case with lime/limestone systems since the
oxidation product is insoluble gypsum.  Failure to allow for sulfate
removal in double alkali systems will ultimately result in precipitation
of sodium sulfate somewhere in the system:  quite possibly in the scrubber
circuit.  The known possible methods of removing sulfate from the system
are:

          1.  purging a waste liquid stream containing dissolved
              sodium sulfate;

          2.  removing wet calcium sulfite/sulfate/flyash sludge
              containing dissolved sodium sulfate in the occluded
              water;

          3.  removing solid sodium sulfate from the system for
              sale or other disposal by selective crystallization
              or operation of another concentration process; and

          4.  precipitating the sulfate as insoluble calcium sulfate
              to be disposed of with the waste product sludge.

     The first method is ecologically undesirable due to its contributing
to water pollution by emission of dissolved sodium salts.  In addition,
depending on the concentration of active alkali (sulfite/bisulfite), the
liquid discharge may have a high COD value.

     The second method disposes of the sulfate in a more subtle manner
than the first.  Here a liquid stream is actually purged; however, it
is much less obvious since the liquid is contained in the solid waste
product, which appears to be relatively dry even when containing up to
40 percent moisture.  The potential contribution to water pollution in
this case is due to the leachability of the waste product sludge and
water run-off from the sludge disposal area.  A highly leachable sludge
will result in slow contamination of the ground water table in the
disposal area while water run-off would lead to contamination of surface
water by soluble salts.  An acceptable alternative to direct disposal of

                                1024

-------
wet solids as outlined above, would be disposal of a treated sludge
which contains the required amount of soluble sulfates but in a "fixed"
solid, unleachable, innocuous form.  Although at present there is no
demonstrated commercially available system to accomplish this, inclusion
of double alkali sludge as part of a current program concerned with
ecological treatment and disposal of lime/limestone wet scrubbing sludge
is under consideration by EPA.(2)

     The third method cited for sulfate removal involves the separation
of sodium sulfate crystal from the scrubber liquor by a heating-cooling
cycle or possibly by operation of a vapor compression cycle such as the
system being marketed by Resources Conservation Corporation as a brine
concentrator.  To date, these methods have not been applied to double
alkali systems and therefore more evaluation is necessary to determine
their applicability.  One possible problem created in removal of
sodium sulfate crystal is the ultimate disposal of this soluble salt.
Since the market for large quantities of sodium sulfate is sparse at
best, dissolution and treatment with lime to produce insoluble calcium
sulfate (gypsum) may be the acceptable short-term solution to the
problem.

     The fourth method, precipitating calcium sulfate directly from
the scrubber liquor, can only be accomplished under carefully controlled
conditions.  Basically, calcium sulfate and sulfite will not be pre-
cipitated in appreciable quantities, simultaneously, in neutral or
basic solution with lime treatment, unless the [S04=]/[S03=] concentration
ratio in solution is relatively high.  A simple explanation for this is
that CaSOj is much less soluble than CaS04 and thus is preferentially
precipitated from solution.  Since the concentration of sulfate is
limited by the solubility of sodium sulfate (approximately 3 molar) the
active alkali, sulfite, must be relatively dilute in order to maintain
a high [S04=]/[S03=] ratio and thus precipitate calcium sulfate upon
treatment with lime.  (See part I of this paper.)

     A variation of the fourth cited method, precipitating calcium
sulfate from the scrubber liquor, was developed by the Japanese.  In
this variation, a small portion of the scrubber liquor (a "slip stream")
is treated with sulfuric acid and calcium sulfite to precipitate calcium
sulfate.  Adjustment of the pH with sulfuric acid allows the calcium
sulfite to go into solution as soluble calcium bisulfate, thus increasing
the calcium ion concentration in solution until the solubility of calcium
sulfate is exceeded and calcium sulfate is precipitated from solution.
The major advantage of this variation is that concentrated active alkali
can be used in the scrubber loop.  The obvious disadvantage is that
sulfur is added to the system as sulfuric acid, procured from another
source,  and must be removed along with the sulfur in the flue gas.
                               1025

-------
     Scale Prevention

     One of the primary reasons, and probably the most important, for
development of double alkali processes was to circumvent the scaling
problems associated with lime/limestone wet scrubbing systems.
Therefore, an obvious consideration in any double alkali system is
the ability of the system to operate in a non-scaling manner.

     In order to eliminate scaling tendencies, the calcium concentration
in the scrubber liquor must be held to a minimum.  Recirculation to the
scrubber loop of supernatant liquor from a lime reaction vessel which
is controlled to precipitate sulfite and sulfate is not acceptable.  The
calcium concentration of this liquor would be high enough to cause
scaling problems since it is saturated with respect to calcium sulfate
and calcium hydroxide.  Figure 1 shows (with straight line interpolation
between two data points) relative solubilities in water of the various
calcium compounds present in a typical double alkali system.  The
figure shows that CaSO, and Ca(OH)2, although only slightly soluble,
are about two orders of magnitude more soluble than CaCOj and CaSC^.
In order to minimize the calcium concentration in the scrubber loop, the
liquor returned to the scrubber from the regeneration process should be
saturated with respect to CaSO- or CaCQ$ but not with respect to CaSO^
or Ca(OH)2.

     It should be noted that Figure 1 is not meant to specify the exact
concentrations to be expected in double alkali systems but only to
point out the direction to go in order to minimize scaling tendencies.
The actual solubilities of these calcium species in real systems will
be greatly affected by such important factors as pH, ionic strength
and the presence of other chemical species.

     In many cases scaling problems are not only due to high soluble
calcium values related to saturated solutions of CaS04 and Ca(OH)2 but
are also possibly due to supersaturation with respect to any of the
calcium compounds.  In order to reduce scaling due to supersaturation,
provisions to maintain high concentrations of suspended solids in the
reaction zones of the regeneration equipment are necessary.  This can
be accomplished by recirculating precipitated solids to the reaction
zones.   This technique is stressed in the Envirotech system which is
described under "Status of Technology" later in this paper.  It has
also been referred to as a crystal seeding.

     In general, efforts to minimize scaling rely upon:

          1.  carbonate softening (i.e.,  reduction of soluble calcium),

          2.  sulfite softening, and

          3.  crystal seeding techniques.

                               1026

-------
0.30
  0  '	
                              40           60
                              TEMPERATURE, °C
                 Figure 1. Solubility of calcium compounds.
                                1027

-------
     Water Balance

     In order to operate a closed system to avoid potential water
pollution problems, system water balance is a primary concern.  Water
cannot be added to the system at a rate greater than the normal water
losses from the system.

     Generally there is a tendency to add fresh water to a scrubbing
system to serve many purposes.  These include:

          1.  saturation of flue gas,

          2.  pump seal needs,

          3.  demister washing needs,

          4.  slurry make-up needs, and

          5.  waste product washing.

     On the other hand water should only leave the system in the following
ways:

          1.  evaporation by the hot flue gas,

          2.  water occluded with solid waste product, and

          3.  water of crystallization in solid waste product.

     Careful water management, part of which is the use of recycled
rather than fresh water wherever possible, is necessary in order to
operate a closed system.

     Waste Product Washing

     As previously indicated, disposal of wet solid waste containing
soluble salts is ecologically undesirable.  In addition, allowing
active alkali or sodium salts to escape from the system is an important
operating cost factor.  Sodium make-up to double alkali systems is
usually accomplished by adding soda ash (recently quoted at $42 per ton)
at some point in the system.  Thus, both ecological and economic con-
siderations dictate that waste product washing is desirable.  With lime/
limestone systems, there is no need to wash the solid waste product.

     S0? Removal

     Generally,  the concentrated active alkali systems have a greater
capability for S02 removal than the dilute systems.  Based on pilot
testing of concentrated systems in Japan, S02 removals of up to 98 percent


                                   1028

-------
have been reported. J  With the dilute systems the range of S02 removal
to be expected is probably in the 80-90 percent range.  In general,  St^
removal with double alkali systems is more efficient than with lime/
limestone systems, since the active alkali is soluble.

     Costs

     Based on cursory design and cost analysis, with the assumption
that a single scrubber device (e.g., a single stage venturi) can be
used to remove particulates and 862 to the extent required to meet new
source performance standards, the capital and operating costs of double
alkali systems appear to be significantly less than those for a lime
or limestone system designed for the same requirements.^ '

     Generally, the dilute active alkali systems tend to be higher in
capital costs than the concentrated systems since both equipment size
and process flows are required to be larger to accomplish the same
degree of desulfurization.  Methods used for sulfate removal and for
reduction of scaling tendencies will affect both capital and operating
costs.
                                 1029

-------
                        SCHEMES OF OPERATION


     Five basic schemes of operation considered to be potentially
attractive, based on design considerations discussed previously, are
described below and illustrated with schematic diagrams.  It should
be noted that many variations on these five basic schemes can be produced
by varying the many parameters associated with the systems such as:

          1.  concentration of active alkali;

          2.  water make-up addition points;

          3.  regeneration with lime, limestone or lime plus limestone;

          4.  sulfate removal method; and

          5.  method employed to reduce scaling tendency.

In addition, although the schematics shown are specific with respect
to the type of equipment employed, this is done only for the sake of
establishing the equipment function at a glance.  Thus, the systems
presented are not limited to the specific equipment shown, namely:

          1.  scrubber type,

          2.  solids separation equipment,

          3.  settling and thickening equipment, and

          4.  reaction vessel type.

     Limestone and Lime Regeneration, Dilute Active
     Alkali with Sulfite Softening (Scheme 1)

     This scheme of operation is typified by Figure 2.  In this system
sulfur dioxide is absorbed by the active alkali, aqueous sodium sulfite,
in the scrubber to form sodium bisulfite.  Limestone is used to regen-
erate sodium sulfite, by reacting with bisulfite and precipitating
calcium sulfite in the limestone reaction vessel.  Lime is used to
precipitate sulfate as gypsum with regeneration of active hydroxide ion
in the lime reaction vessel.  The hydroxide is recirculated to the
limestone reaction tank where it reacts with bisulfite to regenerate
sulfite as does the limestone.

     The product from the limestone reaction vessel is split into a
clarified liquor stream for recirculation to the scrubber loop and a
slurry stream to feed the lime reaction vessel for sulfate removal.
                                 1030

-------
CLEANED
  FLUE
  GAS
   EFFLUENT
   HOLD TANK
 WASH
WATER     ROTARY
          VACUUM
          FILTER
                                            REGENERATED RECYCLE
                                                  LIMESTONE
                                                    SLURRY
                                                   MAKE-UP
                                                    TANK
                                            LIMESTONE
                                            REACTION
                                             TANK
                          Figure 2. Limestone and lime regeneration, dilute, sulfite softening (Scheme 1).

-------
 Supernatant  liquor  (containing hydroxide  ion) from the lime reaction
 vessel  is  recirculated  to  the limestone reaction vessel while the
 remaining  slurry  is processed to remove waste product solids from the
 system.

      Sulfate is removed from the system as gypsum, by precipitation
 with  lime; but in order to accomplish this, the system must be operated
 with  dilute  active alkali  in order to maintain the needed high
 [S04=]/[SOj=] ratio.  Dilute active alkali is somewhat of a disadvantage
 with  double  alkali systems in that higher liquor flow rates are needed
 and scrubbing efficiency is generally lower than with concentrated
 active  alkali systems.

      The scaling  problem is minimized through the use of a sulfite
 softening  technique.  This is accomplished by recirculating (to the
 scrubber loop) only supernatant liquor which is unsaturated with respect
 to CaS04 and CafOH)? from  the limestone reaction system.  Sulfite
 regenerated  in the limestone reaction system tends to limit the calcium
 ion concentration to the low value established by the CaS03 solubility
 product.   Sulfite softening is also advantageous when compared to car-
 bonate  softening  in that a carbonate supplying compound (e.g., C02 or
 Na2CO_) need not  be added  to the system at additional cost (except as
 needed  to  make up for sodium losses).

      Fresh water  addition  to the system is indicated only for:

           1.  lime slurry  make-up, and

           2.  waste product washing.

 Note  that  the limestone slurry is prepared with recycled water, and the
 gas-is saturated  in the venturi with recycled water.  It has been
 reported(5)  that  lime cannot be slaked successfully with recycled water
 containing high concentrations of sulfate ion.

      Operating cost for this system may be lower than for some of the
 others to be discussed  in  that limestone, a very inexpensive calcium
 compound, is used as the major source of calcium.  On the other hand,
 capital costs for this  type of system may be higher than for some of
 the others since  separate  systems must be installed for handling lime
 and limestone.  Limestone handling generally includes crushing and
 grinding equipment in addition.   Also, regeneration with limestone, as
 opposed to that with lime, requires a much greater residence time to
 attain good utilization, usually on the order of 1-2 hours.(3)

      It is anticipated that most of the water addition to this system
will be done via a waste product wash step, thus reducing solubles in
 the solid waste product and conserving expensive sodium in the system.

                                  1032

-------
     It is interesting to note that the use of both lime and limestone
for regeneration in this scheme is limited by the rate of sulfate
formation (sometimes referred to as oxidation rate) in the system.   At
a 50 percent oxidation rate, no limestone would be consumed, since  the
hydroxide produced by the sulfate removal step will consume all of  the
bisulfite in the limestone reaction system.  As a corollary, at oxida-
tion rates greater than 50 percent, the system could be prone to scaling
unless other provisions are made to reduce scaling tendencies.  Obviously,
if the expected oxidation rate is 40 percent or more, scaling could be
a major problem and the advantages of using both limestone and lime are
greatly diminished.

     Lime Regeneration, Dilute Active Alkali
     with Sulfite Softening (Scheme 2)

     This scheme of operation is illustrated in Figure 3.  Basically,
this scheme is the same as scheme 1 except that only lime is used for
both regeneration of sulfite from bisulfite and for removing sulfate.

     The sulfite softening is shown to take place in a reactor clarifier,
a piece of equipment designed to combine the functions of a reaction
vessel with those of a clarifier.

     Comparison of this scheme with the previous one shows:

          1.  greater use of fresh water tor lime slaking; and

          2.  an overall simpler system, probably leading to
              lower capital cost and possibly higher operating
              costs due to the use of lime only as opposed to
              lime and limestone.

     Lime (Scheme 5) or Limestone  (Scheme 4) Regeneration,
     Concentrated Active Alkali with Sulfuric Acid Sulfate Removal

     These schemes of operation are illustrated in Figure 4.  The schematic
is taken from the flow sheets developed and pilot tested by two
Japanese companies, Showa Denko K. K. and Kureha Chemical Industry
Company.

     Since there appears to be a good market for gypsum in Japan, the
processes developed by the two Japanese companies include an oxidation
step to covert all of the calcium sulfite to gypsum for sale.  Limestone
has been selected as the source of calcium by both Japanese companies,

     In the United States, the market for gypsum is relatively smaller
than in Japan.  In addition, there is a tendency to avoid production
of a salable sulfur waste product due to the inherent problems associated

                                 1033

-------
          FLUE
   VENTURI
  SCRUBBER

CLEANED
  FLUE
  GAS
t-
         DEM1STER
o
o
LU
0=
K
Ul
to
CO

K
o
                  EFFLUENT
                  HOLD TANK
                                REGENERATED RECYCLE
                                                                 LIME
  LIME
SLAKING
  TANK
                WATER
         LIME
       REACTION
         TANK
                                   WASH
                                  WATER
                             ROTARY
                             VACUUM
                             FILTER
 WASTE
PRODUCT
                             Figure 3.  Lime regeneration, dilute, sulfite softening (Scheme 2).

-------
CLEANED
  FLUE
  GAS
                                                           LIME
                                                            OR
                                                         LIMESTONE
     VENTURI
    SCRUBBER
                             SLURRY
                              TANK
CENTRIFUGE
 OR FILTER
                                                                                                          WASTE PRODUCT
                                                                                                              (CaS03)
                                                                                                       A  SULFURIC
                                                                                              CONVERSION
                                                                                                 TANK
                              REACTION
                                TANK
                                                                                                              WASTE
                                                                                                             PRODUCT
                                                                                                             (GYPSUM)
                                                          CENTRIFUGE
                                                           OR FILTER
  EFFLUENT
  HOLD TANK
Figure 4.  Lime (Scheme 3) or limestone (Scheme 4) regeneration, concentrated,
                                                                                           sulfate removal.

-------
with marketing a by-product.  Due to these considerations, Figure 4
does not show the oxidation step designed for gypsum production.

     In these schemes sulfate is removed by precipitation as gypsum
from an acidic solution in which CaSO, is soluble.  The overall sulfate
removal reaction is represented as follows:

          Na2S04 + 2 CaSOj + H2S04 + 2 CaS04  4- + 2 NaHSOj

                   (water of hydration not shown)

     The sulfuric acid used to acidify the reaction mixture must be
procured from an outside source and added to the system.  Scaling
tendency is circumvented again in these schemes by sulfite softening.
(There is excess sulfite present in solution in the liquor recycled
to the scrubber circuit, thus maintaining a low calcium concentration.)

     The fact that these systems can be operated with concentrated
active alkali tends to reduce the capital costs.  Liquor flows can be
greatly reduced and most of the equipment can generally be smaller in
size than that required by the dilute active alkali systems.  On the
other hand, anticipated operating costs for these systems would be
higher than for dilute systems due to the cost of sulfuric acid and
the additional cost of calcium required to remove the sulfur added by
the sulfuric acid.  In the United States, where there does not appear
to be a large market for gypsum, these processes become prohibitively
costly if the oxidation rate is high.  Based on data from the Kureha
pilot plant, however, only about 9 percent of the sulfur in the waste
product is due to sulfuric acid added to the system.  This figure is
primarily affected by rate of oxidation in the system.  It should be
noted that the factors affecting oxidation rate in any double alkali
system are not clearly defined and various systems have reported
oxidation rates varying over a large range.

     Lime Regeneration, Dilute Active Alkali with Carbonate
     Softening (Scheme 5)

     This scheme is illustrated in Figure 5.  Although this scheme is
similar to schemes 1 and 2, there is a subtle difference.  Here clarified
liquor from the sulfate precipitation reaction is fed back to the scrubber
loop, but only after its calcium content is significantly reduced by
reaction with carbonate and after another clarification step.  The
carbonate is generally supplied to the system in the form of soda ash
(Na2C03) or carbon dioxide.

     The advantage of using CO? is that no excess sodium is added to the
system.  A major disadvantage is the need to provide a complex system
for adding the C02> thus increasing capital cost and possibly leading to
complications in the operation of the system (chance of CaCOj scaling).
                                 1036

-------
 VENTURI
SCRUBBER
               EFFLUENT
               HOLD TANK
                           Figure 5.  Lime regeneration, dilute, carbonate softening (Scheme 5).

-------
     Using soda ash, on the other hand, requires much simpler equipment
and, therefore, would probably lead to lower capital cost.   The potential
problem, however, is that the amount of sodium carbonate added for
softening purposes is in excess of that required as make-up for sodium
losses.  As a result, this excess sodium would have to be purged from
the system, as sodium sulfate in order to establish steady  state concen-
tration.  Obviously, a high purge rate is unacceptable from the stand-
point of potential for water pollution by these emissions.   A good
compromise may be to use both C02 and soda ash, thus achieving the
advantages of each.

     The operating cost for this scheme would probably be slightly
higher than for schemes 1 and 2 due to the additional cost  of the
carbonate-supplying reagent.

     An advantage of this system over the other dilute systems discussed
is that the active alkali recycled to the scrubber loop would be
hydroxide rather than sulfite as is necessary in the systems employing
sulfite softening.  Feeding hydroxide rather than sulfite to the scrubber
loop results in a higher scrubber liquor pH, and thus more  efficient
scrubbing.

     Comparison of Schemes

     Table 1 presents a qualitative comparison as a summary of the
various double alkali schemes described.  The ratings used  are very
subjective rather than absolute since some of the systems described
are hypothetical.  A comparison of actual systems would be  greatly
dependent on the specific equipment used.

     Capital costs are shown to be higher for dilute systems only
because it is assumed that the dilute systems will generally require
larger size equipment and greater flow rates.  Operating costs are
shown to be higher or lower, based only on the need to supply additional
calcium or other reagents exclusive of normal sodium make-up requirements.

     Oxidation limitation of the systems, although tabulated as a
specific oxidation rate, should only be taken to represent  an attempt
at quantification of this factor based on the chemistry of  the system;
this is commonly known as a "Guestimate."

     Sulfur dioxide removal capability is relatively higher with the
concentrated active alkali systems since more SO  can be removed per
given volume of absorbent liquor than with dilute active alkali systems.
In theory, however, both concentrated and dilute active alkali systems
should be able to attain about the same S02 removal given the right
conditions.  In practice, cost becomes the controlling factor.

     Additional reagents required is meant to include reagents in
addition to those required to make up for minimal unavoidable sodium
losses.
                                1038

-------
Table 1.   COMPARISON OF ATTRACTIVE DOUBLE ALKALI SCHEMES
SCHEME
1

2
3
4
5


Ca"
CONTROL
(SOFTENING)
SULFITE

SULFITE
SULFITE
SULFITE
CARBONATE


ACTIVE
ALKAU
DILUTE

DILUTE
CONC.
CONC.
DILUTE


S04= REMOVAL
AGENT
UME

UME
•W
H.SO,
UME


CALCIUM
SOURCE
UME AND
UMESTONE
UME
UME
UMESTONE
UME


ADDITIONAL
REAGENTS
REQUIRED
NONE

NONE
H2S04
H2S°i,
Na2C03
AND/OR
CO,
ESTIMATED
OXIDATION
UMITATION
40%

50%
30%
30%
NONE


RELATIVE COST
CAPITAL
HIGHER

HIGHER
LOWER
LOWER
HIGHER


OPERATING
LOWER

LOWER
HIGHER
HIGHER
HIGHER


RELATIVE
SO2 REMOVAL
CAPABIUTY
LOWER

LOWER
HIGHER
HIGHER
MEDIUM



-------
                        STATUS OF TECHNOLOGY
     A number of organizations in the United States and abroad have
studied and tested various double alkali schemes on systems ranging
in size from the laboratory bench scale to 30,000 CFM.  A summary of
the status of development of these various systems will be presented
in order to familiarize potential users of these systems with what is
commercially available and what is planned for larger scale installation.

     Every effort has been made to present a status report that is
accurate.  Information presented in an overview report of this type is
necessarily "second hand," and obtained from a multiplicity of sources
including process and equipment vendors, engineering companies and
potential system users, all of whom did not always yield consistent
and/or specific information due in part to the complexities and changing
conditions associated with development programs.

     FMC Corporation(6)

     The requirement to control SO- emissions from large reduction kilns
operated by FMC in Modesto, California, led to the development of a
workable double alkali system.  This full scale system has been in
operation since December 1971.   The double alkali system processes a
combined 30,000 ACFM off-gas stream containing 5000-8000 ppm S02 and
about 5-8 percent oxygen from two kilns.  The scrubber at the Modesto
plant consists of a vertical column packed with 9 feet of Intalox
saddles.  A wire mesh entrainment separator is used in series with the
packed column.  The absorbent liquor contains a high concentration of
active alkali (Na2S03/NaHS03) and sodium sulfate.  Sulfate formed in
the system as a result of oxidation is purged from the system, dissolved
in the liquor adhering to the CaS03 waste product.  A rotary vacuum
filter effects the final solids separation and produces a waste product
containing about 50 percent moisture, CaS03, dissolved Na2S04/Na2S03 and
kiln ash which is disposed of in an on-site pond.

     The Modesto system is reported to operate continuously without
scaling problems.  The system is capable of attaining S02 removal in
excess of 95 percent with simultaneous removal of kiln ash.

     Figure 6 is a schematic representation of the FMC system.  Comparing
the FMC scheme to the five schemes previously presented, it is most
closely related to scheme 3.  The major difference is that the FMC scheme
makes no provision for removal of sulfate as a solid innocuous waste
product.  Sulfate is removed from the system as a dissolved solid in the
liquor entrained with the CaS03 waste product.

     Although the FMC scheme makes no provisions for removal of sulfate
as an insoluble solid, FMC reportsC6) that tests indicate a reduction in

                                 1040

-------
o
                                                                                                  FILTER   PLANT
                                                                                                 VACUUM   IATER
                                                                                                  PUMP
                                                    LIME
                                                   STORAGE
                                                    BIN
                  REHEATER

                     CONDENSATE
                                                                                   FILTRATE
                                                                                   RECEIVER
                                   SODA ASH
                                   STORAGE
                                     BIN
                                                                                                  ROTARY FILTER
                                                   PLANT
                                                   WATER
        ENTRAPMENT
         SEPARATOR
                                                                     LIME
                                                                   REACTOR
                                                                                                           SOLIDS TO
                                                                                                           LANDFILL
                                                                                    THICKENER
                                                                                    UNDERFLOW
                                   RECIRCULATION
                                       TANK
                                                                     OVERFLOW
                                                                       PUMP
RECIRCULATION
    PUMP
                                              REGENERATION
                                              CIRCUIT PUMP
                                                                                           _ -FILTRATE
                                                                                          0 1 RETURN
                                                                                                PUMP
                                                 Figure 6. FMC schematic.

-------
sulfate formation when the system is operated with high ionic strength
absorbent liquor.  As a result, provisions have been made for washing
the solid waste product to the extent consistent with maintaining a
steady state sulfate level.

     In an effort to develop their system for application to industrial
and utility boilers, FMC has continued testing their system at the
pilot plant level.

     A 2,000 CFM pilot plant was operated at an FMC chemical plant in
South Charleston, W. Va., for a period in excess of 6 months starting
in June 1972.   The flue gas supply was taken from an underfed stoker
type steam boiler rated at 80,000 Ib of steam per hour and burning
3.5 percent sulfur coal.  The oxygen content of this flue gas was varied
up to 13 percent in order to obtain conditions conducive to oxidation
and to observe effects on the system.

     FMC has also recently operated a 3,500 CFM pilot plant housed in a
40-foot trailer at the Mossville engine plant of the Caterpillar Tractor
Company.  The flue gas supply for this testing was taken from a steam
boiler at the plant which burns typical high sulfur Illinois coal.  This
plant was equipped for filter cake washing to reduce sodium losses from
the system.  The waste product from this system contained about 40 percent
moisture under optimum conditions and appeared to be relatively dry and
easy to handle.

     After initial testing in March 1972,  in a 5,000 CFM unit at the
Modesto Plant, all of the FMC pilot plant programs have employed the
"FMC-Link Belt Dual Throat Variable Scrubber."  This is a single stage
venturi type scrubber coupled with a cyclonic type separator.

     Typical operating parameters for the FMC system are:

          S02 removal efficiency                           70-95%

          Approximately alkali composition                 15-20% Na2S04
                                                           5-7% Na2S03/
                                                                NaHS03

          Ca:S stoichiometry                               1.0

          Scrubber pH range                                6-7

          Na.CO^ requirements                              8 moles Na2COj
                                                             moles Ca(OH)

          Ca++ in scrubber liquor                          < 5 ppm

                                  1042

-------
     Capital and operating costs for the FMC system are  estimated by
FMC to be:

                                           Generating Capacity, MW

                                             8      20       200


          Capital costs, $/KW              184     100        30

          Operating costs, $/ton of coal    11       8         5
     FMC is presently negotiating with a large industrial manufacturer
in Illinois for a contract to install the FMC process to control  SOX
and particulates from a system of steam boilers approximately equivalent
in steam generating capacity to a 40 MW power plant.  It is presently
anticipated that this system will be in operation by June 1974.

     Envirotech Corporation(5)

     Envirotech has operated a 3,000 CFM double alkali pilot plant using
various equipment arrangements for approximately a year, since early
1972.  The pilot plant is located at the .Gadsby station of Utah Power
and Light.

     After testing various equipment arrangements, and fighting the
scaling problem with some, Envirotech developed the scheme shown in
Figure 7.  This scheme might be the typical scheme which would be
proposed by Envirotech for control of SOX and particulates from a
utility or industrial boiler.

     Examination of this flow sheet indicates that it is related to
scheme 3 or 5, depending on the concentration of active alkali in the
liquor.  The concentration range of sulfite, bisulfite and sulfate in
the liquor is considered by Envirotech to be proprietary information
and, therefore, has not been disclosed.  Operated with concentrated
active alkali, this system is similar to scheme 3 without provisions
for removal of sulfate (with sulfuric acid) as an insoluble solid.
Operated with dilute active alkali, the system is similar to scheme 5,
using soda ash for carbonate softening.  In this case there is the
necessity to purge from the system a certain amount of sodium dissolved
in the liquor adhering to the waste product filter cake.

     In the pilot plant, a perforated plate and an expanded metal tray
were tested in the tray scrubber.  The preferred arrangement was a
2-stage expanded metal tray scrubber.  The pilot plant is reported to
have operated in a scale-free manner for a period of 6 months with

                                 1043

-------
                     RAN H20
i CLEAN GAS
         *      rrsf
         KENER         }  k
                           CLARIFIER  LIQUOR
                                   STORAGE
              COOLING
              TONER
              BLON-
              DONN
                                              PROCESS
                                              NEEDS
                      NASTE
Figure 7.  Envirotech schematic.

-------
provisions for soda ash softening, inlet gas quenching with fresh water,
and recycle_of precipitated solids to the reaction tanks and reactor
clarifiers."  Quenching with fresh water, as pointed out previously,
could lead to water balance problems and thus require purge.  Use of
cooling tower blowdown and sending demister wash water to plant use
are examples of good water conservation attempts.

     Envirotech lists the problems they consider significant in their
operation of the system as follows (not necessarily in order):

          1.  scaling;

          2.  gas distribution at low pressure drop;

          3.  over-liming (relating to pH control);

          4.  mechanical problems - valves, pumps, piping;

          S.  flow regulation; and

          6.  inlet wet/dry interface accumulation.

     A list of important process variables and their typical values from
various pilot tests is given below:

          S02 in inlet flue gas                         350-400 ppm

          Scrubber pressure drop                        4-9  in. H20

          Calcium content of scrubber liquor            400  ppm

          pH of scrubber liquor at exit                 7

          Scrubber L/G ratio                            12  to  36  gal/
                                                          1,000 ft3

          S0^=/S03= concentration in scrubber liquor    not  disclosed

          S02 removal efficiency                        90%

     Additional pilot plant testing, 24 hours/day,  5 days/week,  for
2-3 months will be geared to investigate:

          1.  simultaneous removal of SCL and particulates;  and

          2.  system operation at inlet S02 levels  up  to  1500 ppm.

     Envirotech will attempt to demonstrate their system next at the
100-200 MW level.

                                  1045

-------
     General MotorsC7»8)

     The double alkali system developed by GM is»schematically illus-
trated in Figure 8.  After considering other possible alternatives, GM
decided that a double alkali system was most practical for their S02
control requirements in that it fit the criteria of being relatively
economical and uncomplex.

     Pilot plant testing was conducted at the Cleveland Chevrolet plant
using a 2,800 CFM cross-flow packed scrubber supplied and operated by
Ceilcote for the joint development effort.  Flue gas for testing was
an isokinetic sample representing 10 percent of the total flue gas flow
from a boiler having a steaming capacity of 80,000 Ib/hour and burning
2 percent sulfur coal with 100 percent excess air.

     In the pilot plant, it was found that a 1 molar sodium solution
would give reasonably good SO? absorption, while also regenerating
caustic and precipitating sulfate as gypsum.  It was found that a maximum
concentration of 0.1 M hydroxide could be regenerated in a 1 M sodium
solution.  Increasing the sodium ion concentration above 1 M did not
give appreciable increase in regenerated hydroxide concentration.  Lowering
the sodium ion concentration, however, gave a decrease in regenerated
hydroxide concentration.

     Optimum lime utilization was found to occur with high speed mixing,
using near stoichiometric quantities of lime.  Eighty percent conversion
of lime was attained in 5 minutes.

     Early in the pilot plant program, GM ran into the calcium plugging
problem; but later, using soda ash softening, they were able to alleviate
it.  Reportedly, using soda ash for sodium make-up, they were able to
reduce the calcium content in the scrubber loop from 400 to 250 ppm.

     A high degree of confidence in the double alkali system has prompted
GM to construct: a full-scale system at the Cleveland Chevrolet plant.
The plant presently has four stoker-fired boilers, equivalent in steaming
capacity to a 32 MW electric generating plant.  The double alkali system,
however, is designed to handle the flue gas equivalent of a 40 MW plant
in order to accommodate possible future expansion.  The system is being
built at an approximate cost of $3,000,000 and is scheduled to start up
in December 1973.

     The system in the full scale plant will consist of four parallel
Koch tray (valve tray)  scrubbers.  Provisions for softening with soda
ash and/or carbon dioxide will be included.  Sodium make-up will be as
sodium hydroxide and/or soda ash.  The active alkali will be dilute to
ensure removal of sulfate as insoluble gypsum by lime treatment.  The
anticipated sodium ion concentration will be in the range of 1-2 molar.

                                  10/16

-------
                        SULFUR-FREE
                          FLUE GAS
o
4?
-j
                (flETGAS
               ))SCRUBBER
                      Na2S03 SOLUTION
                                                        REACTOR
                                                       CLARIFIERS
   HIGH-SULFUR
     BOILER
     FLUE GAS
FILTER |
                     FILTER CAKE
                    (TO LANDFILL)
                        SLURRY
   CaCOs FOR      _
NEUTRALIZATION-*-
                                        Figure 8.  General Motors schematic.

-------
 Depending upon  control of  the operation, this system should be capable
 of producing  an environmentally acceptable solid waste product, at least
 with respect  to contamination by dissolved sodium salts.

      This design is comparable to scheme 5, previously discussed.


      Zurn Air SystemsC9»10)

      Zurn contracted with  Southern Research Institute to conduct a
 laboratory study of the  regeneration process for a lime/sodium, dilute
 double  alkali system.  The study, completed in 1972, basically confirmed
 the information reported by General Motors.  This study was conducted in
 order to  obtain a firmer basis for the design of a marketable double
 alkali  system.   Figure 9 is a schematic representation of the type system
 Zurn is prepared to market.  This flow diagram is comparable to scheme 5,
 previously discussed, with soda ash softening.

      The  Zurn system would employ a "Dustraxtor" scrubber, briefly
 described as  a  multitube entrainment contactor capable of achieving high
 internal  liquid/gas ratios.

      It should  be pointed  out that the system would be capable of
 removing  sulfate as insoluble gypsum by precipitation with lime; however,
 due to  the lack of provisions for softening with carbon dioxide, a
 certain amount  of sodium and sulfate would necessarily have to be purged
 from the  system in the moisture associated with the waste product filter
 cake.   Provisions for filter cake washing are shown.  The system is
 anticipated to  operate at  a sodium concentration Of 1-2 molar.

      A  Zurn operating cost estimate, exclusive of amortization of equip-
 ment, was given as $4.86 per ton of coal burned.  Zurn is presently
 negotiating with a large industrial manufacturer for a contract to
 install the equivalent of  a 20 MW system on a coal-fired industrial
 boiler  system.

      ChemicoC11)

      Chemico tested a double alkali system in a 1,500 CFM pilot plant
 located at the Mitchell  Power Station of Allegheny Power Service Company
 Testing lasted  approximately 6 weeks, and Chemico was able to operate
 the  pilot unit  in a scale-free manner, using either lime or limestone
 as  the  calcium  source.   Figure 10 is a schematic representation of that
 pilot plant, but not necessarily of the system which might be marketed
 by Chemico.

      In the pilot plant,   there were no waste product cake washing
provisions, so dissolved solids were purged from the system in the
moisture associated with the cake.  In addition, since make-up sodium
was added to the system as a mixture of sodium sulfite and bisulfite
                                 1048

-------
o
vo
                                                                                        CLEAN GAS
                                                                                         AND H20
                                                           S02, S03, AND
                                                          PARTICULATES
                                                            I
        MECHANICAL
                                               DUSTRAXTOR
                                                SCRUBBER
   DUST
COLLECTO
                                      DILUTE
                                       NaOH
                                        NaHS03,
                                        Na2S03,
                                      Na2S04, AND
                                     PARTICULATES
           WASTE
          PRODUCT
                   VACUUM
                    FILTER
                                                                             LIME
                                                                             MIX
                                                                            TANK I
                                                                            Na2C03
                                                                           MIX TANK
        SLUDGE
       STORAG
               ^VACUUM
     TO \/ # PUMP
  LANDFILLT  SEAL H20
     
-------
                                                                               Na SALTS
VENTURI
                                                                                        H20
                                                                                    PRODUCT
                                                                                     LIQUOR
                                                                                     TANK
                                                                               FUE.   •SE
                          Figure 10. CHEMICO Pilot Rant schematic.

-------
on a regular schedule, it necessarily follows that there must have been
a significant purge from the system.  Testing in this fashion was
attributed by Chemico to be due to the fact that testing at such a
small scale with separate pieces of equipment which were not sized to be
consistent with each other tended to necessitate a somewhat open system.
Typically, in small unit testing, leakage from the system is generally
enough to simulate open system operation.

     Based on testing in the pilot plant, however, Chemico developed
other flow sheets for application to utility and other boiler systems
which are apparently similar to schemes 1 and 2, previously discussed,
provided that the systems are intended for operation with dilute active
alkali.  If these systems are operated with concentrated alkali, a purge
of sulfate from the system in some form would be necessary.


     Arthur D. Little/Combustion Equipment Associates^12-*

     ADL and CEA are involved, per a joint venture agreement, in developing
and marketing double alkali systems.  Generally speaking, CEA is an equip-
ment and systems designer and manufacturer, and ADL participates in a
developmental and engineering capacity, in the joint venture agreement.

     A 2,000 CFM pilot plant has been in operation since January 1973,
testing various double alkali configurations at the ADL facility in
Cambridge, Massachusetts.  The pilot plant unit was built by CEA for
testing in connection with any of the ADL/CEA joint ventures.  This pilot
unit is supplied with flue gas generated by combustion of natural gas.
There are provisions for addition of S02, flyash and excess air.  The
scrubber system consists of a'variable throat venturi, tray separator
and radial vane demister.  The solids handling facilities include a
circular clarifier or settling tank with revolving bottom rake arm
mechanism and a rotary vacuum filter with provisions for filter cake
washing and drying.  The plant is equipped with seven miscellaneous
tanks, four of which are provided with mechanical agitation.  These tanks
serve as scrubber recycle tanks and reaction vessels as necessary.

     ADL/CEA have contracted with The Southern Company (a utility combine
producing electricity in the southeastern states, consisting of Alabama
Power Co., Georgia Power Co., Gulf Power Co., Mississippi Power Co.,
Southern Electric Generating Co. and Southern Services, Inc.) for
installation of a 20 MW S02 and particulate scrubber prototype system.C13)
Flue gas supply for the prototype will be half the flue gas produced from
the Unit No. 1 boiler of the Scholz plant operated by Gulf Power Company.
The scrubber systems to be tested will be various combinations of a
venturi, spray tower, tray tower and packed tower.  The system will be
designed to test double alkali operation with lime only or lime and
limestone.(14)  in addition, the system is designed to test lime and
limestone slurry scrubbing.  The exact plans for operation of this system

                                  1051

-------
with respect to alkali concentration are not known; however, in order
to produce an environmentally acceptable waste product, the system
would probably have to operate, as previously described, by schemes 1,
2, or 5, since no provisions for sulfuric acid addition were revealed,
as would be required for operation as depicted by schemes 3 and 4.

     It is anticipated that construction of this prototype facility
will be initiated in late 1973, and that the system will be in operation
in June 1974, or earlier.

     Kureha Chemical Industry Co. (Japan)

     Kureha has recently developed a sodium/calcium double alkali system
which operates in a manner similar to that described by scheme 4,
previously discussed.  After testing in a small pilot plant, a larger
pilot plant (3,000 SCFM) was built in a joint development effort with
Kawasaki Heavy Industries and has been in operation since July 1972.
Two commercial plants using this scheme of operation will be completed
in 1974 to treat flue gas from oil-fired boilers at two different
electric power plants each having a 150 MW generating capacity.

     The proposed Kureha system consists of a venturi for particulate
removal, followed by a grid-packed scrubber for S02 removal.  The SC>2
absorber operates in the pH range 6.0 - 6.5.  With an inlet concentration
of 1,500 ppm S02, 98 percent removal is achieved.  The scrubber liquid/
gas ratio is 7 gal/1,000 ft3 of gas.

     The limestone reactor operates at a temperature somewhat higher than
the scrubber temperature and is designed to provide a 2-hour residence
time.  Centrifuges are used for the solids separation steps.

     Composition of the absorber feed and discharge liquor is reported
below:

          Absorber feed -

               Sodium sulfite                          20-25%

               Calcium                                 30 ppm

               pH                                      7-8

          Absorber discharge -
                  4

               Sodium sulfite                          10%

               Sodium bisulfite                        10%

               Sodium sulfate                          2-5%

                                  1052

-------
     Since there is a good market for gypsum in Japan,  an oxidation
step is added to the process described previously by scheme 4.    Calcium
sulfite is reacted with air at atmospheric pressure in  an oxidizer
developed by Kureha.  The product gypsum,  removed by centrifuge from
the oxidizer liquor, is suitable for use in wallboard and cement.

     Operating costs associated with this  process include additional
costs for sulfuric acid and the incremental cost for additional
limestone required for precipitation of the sulfur added in the form
of sulfuric acid.  The cost penalty is aggravated if gypsum cannot  be
sold.  As pointed out previously, the sulfuric acid requirement for
decomposition of the sulfate is 125 percent of the theoretical  amount;
thus, about 9 percent of the product gypsum is derived  from sulfuric
acid.

     Capital costs for this process can be split as follows:

          Absorption system                            30%

          Limestone reaction system                    30%

          Sulfate removal                              10%

          Oxidation                                    30%
                                                      100%

The system is reported to operate reliably with no scaling problem,
while removing sulfate in an environmentally acceptable manner.

     Showa Denko K. K. CJapan)

     The Showa Denko process is very similar to the process described
for Kureha and again comparable to scheme 4 with the additional provision
for oxidation of all the CaS03 to gypsum  (for sale) in a separate oxida-
tion step.  This company uses a vertical cone type absorber which
operates with liquid/gas ratios ranging from 7-14 gal/1,000 ft3  and with
a pressure drop range from 8-15 inches of water.

     Showa Denko has operated a 5,900  SCFM pilot plant using this process
since 1971 to desulfurize flue gas from an oil-fired boiler.  A  commercial
plant is currently under construction  and expected to be in operation in
June  1973.   This plant  (Chiba plant of Showa Denko) will  treat  flue gas
equivalent to that  emitted by a 34 MW  electric generating  plant.

     Summary of Commercial Systems

     A comparison of the double alkali systems developed by the various
organizations discussed in this paper  is presented in Table 2.   The
systems described by the table generally represent the systems developed,

                                  1053

-------
                               Table 2.   COMPARISON OF COMMERCIAL DOUBLE ALKALI SYSTEMS
COMMERCIAL
SYSTEM
FMC
ENVIROTECH
GENERAL
MOTORS
ZURN AIR
SYSTEMS

CHEMICO
A.D. LITTLE/
COMBUSTION
EQUIPMENT
ASSOCIATES
SHOWA DENKO
K.K.
KUREHA
ACTIVE
ALKALI
CONC.
UNKNOWN
DILUTE
DILUTE

UNKNOWN
DILUTE

CONC.
CONC.
SULFATE
REMOVAL
PURGE
WITH WASTE
PURGE
WITH WASTE
UME
UME
TREATMENT.
SOME PURGE
LIME OR
PURGE
UME OR
OTHER

„,*>,
H2SO,
CALCIUM
CONTROL
*>*'
<»3=
co3=
cof

S03=
S03= OR
co3=

—
»,'
CALCIUM
SOURCE
L
L
L
L

L AND/OR
LS
L AND/OR
LS

LS
LS
ADDITIONAL
REAGENTS
H.,00,
Na2C03
COj AND/OR
*.,«,,

DEPENDENT ON
OPERATION
DEPENDENT ON
OPERATION

H2S04
H2S04
LARGEST UNIT
TESTED, cfm
30,000
3,000
2,800
LAB SCALE
.
1,500
2,000

3,000
5,900
PLANNED
INSTALLATIONS
SIZE
40 MW
-
40 MW
20 MW

-
20 MW

34 MW
150 MW
150 MW
DATE
7/74
-
12/73
'

-
6/74

6/73
1974
1974
o
en

-------
tested and/or available for marketing by each of the listed companies.
In some cases several systems are implied.   The reason for this is that
most of the companies have tested several double alkali schemes during
their development efforts.  In some cases,  the schemes tested are not
necessarily the ones which would be marketed by the company.  Information
is presented in the table to allow the reader to identify each of the
commercially developed systems with the attractive schemes of operation
previously discussed in detail.
                                  1055

-------
              THE EPA DOUBLE ALKALI DEVELOPMENT PROGRAM
     In an effort to more fully test and characterize double alkali
systems, EPA has contracted with Arthur D. Little, Inc., (contract
value approximately $500,000) to conduct a two-task study.  The
proposed study is entitled, "A Laboratory and Pilot Plant Study of
the Dual Alkali Process for S02 Control."  Technical administration
for this study will be split between the Research Laboratory and the
Development Engineering Branches of the Control Systems Laboratory -for
administration, respectively, of the laboratory and pilot plant tasks.
The test program and study are anticipated to last for 15 months.  One
of the main objectives of the program will be to develop several methods
of producing environmentally acceptable waste while operating in a
reliable trouble-free manner.

     An outline of the program by task is given below:

     Task 1 - Laboratory Study

     1.   A comprehensive literature survey will be performed stressing
         kinetics and thermodynamics of regeneration chemistry, S02
         absorption, and solids separation operations related to
         calcium/sulfur compounds.

     2.   A laboratory study of the  regeneration chemistry will be
         conducted, using lime and/or limestone.

     3.   A complete laboratory bench scale simulation of the absorption
         and regeneration systems will be constructed and tested in
         various attractive configurations to investigate S02 removal,
         scaling problems, residence times,  optimum solution concentra-
         tions  and solids settling  and filtration rates.

     Task 2 - Pilot Plant Study

     1.   Attractive schemes of operation will be tested at the
         ADL/CEA 2,000 CFM pilot plant in Cambridge.

     2.   A 9,000 CFM pilot plant at the Reid-Gardner station of Nevada
         Power  Company will be modified to accommodate testing of
         attractive schemes of operation.   (This pilot plant was
         formerly used to test once-through  sodium carbonate scrubbing
         to firm up the design of a 250 MW desulfurization system now
         under  construction at the  Reid-Gardner plant by CEA and
         anticipated to start up in mid-1973.)   The scrubber system
         in this pilot plant is the same as  in the 2,000 CFM system
         located at the ADL facility.


                                 1056

-------
     3.  Attractive schemes will  be tested  at  the  9,000 CFM pilot
         level.

     4.  Data from the pilot plant operations  will be.assembled
         to provide input for an  economic study to estimate capital
         and operating costs for  various  full  scale  systems.

     5.  Particular emphasis will be given  to  the  important considera-
         tions previously discussed:

         •  Sulfate removal methods

         •  Scale prevention

         •  Water balance

         •  Waste product washing

         •  SC>2 removal

         •  Costs

     Figure 11 is a simplified schematic of the "Double Alkali" process
which is used by ADL to illustrate the system in a very general way.
It should be noted that sulfate treatment is shown in a general manner,
illustrating the possible application'of any of a number of techniques
to be tested to cope with the problem of removing sulfate in  an
environmentally acceptable way.  There is additional significance to
the general philosophy of treating for sulfate removal  in a slip stream
off the main liquor regeneration  loop:  methods developed to  handle
this problem would be applicable  to other double alkali systems.

     Conduct of the Test Program  and Objectives

     Testing will be conducted at three successive levels:

         1.  laboratory bench scale,

         2.  2,000 CFM pilot plant, and

         3.  9,000 CFM pilot plant.

     The successive levels of testing will be used to screen and select
the most attractive schemes from  the many initially showing potential.
Testing in this manner should allow testing of many schemes at the bench
scale level, fewer at the 2,000 CFM pilot level and still fewer at the
9,000 CFM level where long-term,  reliable,  closed-loop operation will be
stressed.  Equipment at each level of testing will be designed with
maximum flexibility consistent with budgetary constraints of the program.

                                 1057

-------
   I
  FLUE
  GAS
BY-PASS
-  FLUE
S  GAS '
cS  FEED
     (*OHk
               SCRUBBED
                  GAS
               SCRUBBER
         JL
   n
           MIXING
            TANK
                                                                      SCRUBBER
                                                                       FEED
                                                                                                               MAKE-UP
SCRUBBER
EFFLUENT
                               U
                            CAUSTIC1ZER
                                                  H
                                             H
                                                                            THICKENER
                                                                                        NASH
                                                                                        HATER
                                                  SULFATE
                                                 TREATMENT
                                                 (GENERAL)
                                                                WASTE
                                                               CALCIUM
                                                                SALTS
                                                                                                             HOLDING
                                                                                                               TANK
                                           Figure 11.  Arthur D. Little general schematic.

-------
Pilot plant capabilities will  include provisions  for:

    1.  solid waste product washing,

    2.  control of SC>2 content of the inlet flue  gas,

    3,  flyash addition,

    4.  sodium make-up as sodium sulfate,

    5.  recirculation of precipitated solids,

    6.  flue gas quenching, and

    7.  testing of at least five attractive schemes.

Following is a list of important program objectives:

    1.  demonstrate reliable system operation,

    2.  demonstrate high S02 removal, 95 percent  desirable,

    3.  demonstrate high particulate removal,

    4.  demonstrate environmentally acceptable  sulfate
        removal schemes,

    5.  minimize disposal of soluble material,

    6.  demonstrate closed-loop operation (or close approach
        to this ideal), and

    7.  obtain complete material balances.
                            1059

-------
                             BIBLIOGRAPHY
 1.  F. T. Princiotta and N.  Kaplan, "Control of Sulfur Oxide Pollution
     from Power Plants."  Presented at LASCON '72, Washington,  D. C.,
     October 18, 1972.

 2.  J. W. Jones,  R. D. Stern,  F. T. Princiotta, "Waste Products from
     Throwaway Flue Gas Cleaning Processes - Ideologically Sound
     Treatment and Disposal."  Presented at F.PA Flue Gas Desulfuri zation
     Symposium, New Orleans,  La., May 14-17, 1973.

 3.  J, Ando,  "Recent Developments in Desulfurization of Fuel Oil and
     Waste Gas in  Japan (1973)."  !:PA report EPA-R2-73-229,  May 1973.

 4.  G. T. Rochelle, "Economics of Flue Has Desulfurization."  Presented
     at EPA Flue Gas Desulfurization Symposium,  New Orleans,  La.,
     May 14-17, 1973.

 5.  D. Dahlstrom  and others, L;imco Division, Fmvirotech Corp.,
     personal  communication,  1973.

 6.  J. Brady  and  others,  FMC Corp., personal communication,
     November  1972-May 1973.

 7.  R. J. Phillips, General  Motors, personal communication,  1973.

 8.  R. J. Phillips, "Sulfur  Dioxide Emission Control for Industrial
     Power Plants."  Presented  at Second International Lime/Limestone
     Wet Scrubbing Symposium, New Orleans, La.,  November 8-12,  1971.

 9.  G. G. Langlcy, Zurn Industries, personal communication,  1972-1973.

10.  "Sulfur Oxide and Particulate Control System."  Zurn Industries,
     July 1972.

11.  I. S. Shah and S. Sawyer,  Chemico, personal communication, 1972.

12.  C. R. LaMantia, ADL,  personal communication, 1972-1973.

13.  J. M. Craig,  The Southern  Company, personal communication, 1973.

14.  I. S. Shah, CEA, personal  communication, 1973.
                                  1060

-------
      STONE & WEBSTER/IONICS
SO2 REMOVAL AND RECOVERY PROCESS
                  by

   N. L. Foskett and E. G. Lowrance
Stone & Webster Engineering Corporation
         Boston, Massachusetts
           Wayne  A. McRae
          Ionics,  Incorporated
       Watertown , Massachusetts
                  1061

-------
 1.   IMTRODUCTION

 The Stone & Webster /Ionics  S02 Removal and Recovery Process  is based on the
 absorption of sulfur dioxide in a caustic  solution  which then is regenerated
 using Ionics' electrolytic  cells (electrolyzers), the key component of the
 process.   This patented process -is applicable  to gaseous effluents from
 stationary power plants burning fuels  containing sulfur and  to tail gases
 from sulfur recovery plants, smelters  and  sulfuric  acid plants.

 The process consits of three essential steps arranged in a closed loop:

      1.    Sulfur dioxide (SOg)  is absorbed from flue gas in  a solution
            of caustic soda  (NaOH)  in water to  produce a solution of
            sodium bisulfite (NaHS03) containing some sodium  sulfite
      2.   The  sulfite-bisulfite  solution is mixed with sulfuric acid
            (H2SOl|),  causing release of 802 ft*011 the solution.  The SOg
            is  recovered for sale as such, or converted to sulfur or
            sulfuric  acid.  The solution is converted to sodium sulfate
            (Na2SOif ) .

      3.   The  sulfate solution is converted by an electrolyzer into two
            solutions, one containing caustic, and the other sulfuric
            acid.  The caustic solution is recycled for use in step 1;
            the sulfuric acid solution is recycled for use in step 2.

Under a joint  program co-sponsored by the Environmental Protection Agency
and Wisconsin  Electric Power Company, the process will be tested at
Wisconsin Electric 's Valley Station in Milwaukee.  Phase I of the potential
three-phase demonstration program is currently underway and consists of
design, installation and operation of an integrated pilot plant, development
of prototype scale electrolyzer  system, preliminary design of a prototype
system and  development of test programs and operating schedules.  Phases II
and -III of  the program would involve the design, procurement and installation
of a  prototype facility treating the flue gas from one of the four 75 MW
coal  fired  boilers at the station, and its start-up and long-term operation,
respectively.

2.  HISTORY OF THE PROCESS

Several years  ago, Stone & Webster, because of its involvement in the design
and construction of power plants, undertook to examine the various sugges-
tions which had been advanced for 802 removal from flue gas and came to the
early conclusion that scrubbing with an alkali solution was a promising and
probably enconomic approach, provided that appropriate methods could be
developed for  regenerating and reusing the solution without excessive cost
and without the production of additional pollutants.  The advantages of
electrolytic regeneration as proposed by Ionics, Incorporated of Watertown,
Massachusetts, appeared particularly interesting.
                                  1062

-------
Joint activity by Stone & Webster and Ionics commenced in 1966, and has
been directed toward the engineering development of an economic system for
an electric generating plant to remove S02 from flue gas by absorption
in a caustic solution, the caustic being electrically regenerated and
reused in a closed circuit.  A major commercial installation of an
electrolyzer system employing technology similar to that practiced in
the Stone & Webster/Ionics S02 Recovery System was installed in a plant
of Monsanto Textile Corporation in Decatur, Alabama, in 1967, for the
reduction of acrylonitrile to adiponitrile, an intermediate in nylon
manufacture.

An SOg removal pilot plant was installed at the Gannon Station of Tampa
Electric Company in Florida in 1967 and was operated for approximately
6 months.  It treated approximately 200 cfm of flue gas from a boiler
operating on pulverized coal.  The flue gas contained about 0.25 vol
percent sulfur dioxide.  The pilot equipment included a flue gas quench
which cooled the gas to about 120° F and removed most of the fly ash.
This was followed by an absorber for removal of the sulfur dioxide, and
an electrolyzer system for regeneration of the caustic.  The pilot program
confirmed the soundness and operability of the process, and pointed the
way toward further development of both the electrolyzer and the flue gas
absorber.

As result of the pilot work, electrolyzer materials of construction were
revised by Ionics to improve reliability and to reduce costs.  In addition
some design modifications were introduced to reduce electrical energy
requirements.  Subsequently, the electrolyzer was scaled to its full
commercial height, and a six month test period of operation was carried
out in a second electrolyte pilot plant at Ionics' Watertown Laboratory
using this "tall cell model."  A third pilot plant has been built and
successfully operated in Japan.  Its purpose is to demonstrate the
regeneration of caustic for removal of S02 from the flue gas of oil fired
boiler under Japanese conditions.

Following the Tampa pilot work and recognizing that considerable effort
had to be expended on the design of absorbers for large steam plant
installations, the National Air Pollution Control Administration (NAPCA)
engaged Stone & Webster to make a study of existing absorber concepts.
The study was based on removing sulfur dioxide from flue gas produced  in
a coal-fired power station rated at 1,200 MW and consisting of four boilers,
with special emphasis on minimizing the oxidation of sulfites to sulfate in
the absorption system.  It was concluded that a packed tower was best  suited
to the Stone & Webster/Ionics Process.

Having carried through the laboratory experiments at Ionics, scale-up
engineering and seni-commercial studies for the various system components
within the process, the next logical step would be to undertake a demon-
stration program of the entire process using each significant component in
                                   1063

-------
 its presently developed state.   However,  considering the small size
 (200 CFM)  of the Tampa Pilot  Plant  and the large thruput of even a modest
 sized prototype plant, say 75 MW, which would have stack gas volume of
 about 225,000 CFM,  it was  decided to  install an intermediate sized pilot
 plant of 2,000 or 3,000 ACFM.  This should provide data less subject to
 gross error than a  scale up from 200  to 225,000 CFM.

 Such a program is now under way  at  Wisconsin Electric's Valley Station in
 Milwaukee.

 3.   PROCESS DESCRIPTION^1*

     3.1 General

 The process involves  three  basic steps.  These are:

      1.  Absorption

      2.  Sulfur Dioxide  Recovery

      3.  Absorbent  Regeneration and Oxidation Product Rejection

 The flow scheme is  shown on Pigs. 1-1 and 1-2.

         3.1.1  Absorption

 The flue gas  is first  cooled by direct water quench in the bottom of the
 absorber tower, after  which it is contacted with a dilute (&%) caustic
 solution containing sodium  sulfate.  At the top of the tower, the flue
 gas,  with most  of the  S02 removed,  is reheated if necessary and returned
 to  the stack.   The  absorbing solution  is converted to a sodiun bisulfite-
 sodium sulfite-sodium  sulfate mixture which contains the 802 and 803
 removed  from  the  flue  gas.

         3.1.2  Sulfur Dioxide Recovery

The bisulfite-sulfite mixture is reacted with a dilute mixture of sulfuric
acid  and sodium sulfate recycled front the electrolytic cells (electrolyzers).
This  reaction forms sodium  sulfate, S02 and water, and is carried out prior
to  entering a stripping tower.  In the latter, SOg is removed as the over-
head, leaving as tower bottoms a solution of sodium sulfate.

         3.1.3  Absorbent Regeneration and Oxidation Product Rejection

The sodiun sulfate  solution is sent to two types of electrolyzers which are
described in detail later in this section.  In both electrolyzers, sodium
hydroxide (caustic)  containing sodium sulfate is generated.  This caustic
is recycled to  the absorption tower for reuse as the absorbing medium.
Also  in both electrolyzers a mixture of sodium sulfate and sulfuric acid is
generated and recycled as the dilute acidic solution required in the Sulfur
Dioxide Recovery step mentioned above.

      (1) Description is general but relates more specifically to
          prototype plant rather than pilot plant.

                                 1064

-------
     STACK
                                                                  STRIPPER
                                                                                STRIPPER
                                                                                 REFLUX
                                                                               CONDENSER
FLY ASM SLOWDOWN TO ASH TRANSFER SYSTEM
                                QUENCH
                               CIRCULATION
                                 PUMP
ABSORBER CIRCULATION
     PUMPS
 STRIPPER
REFLUX PUMP
                                                   FIGURE 1-1

                                          S02  REMOVAL  SECTION
                            STONES WEBSTER/IONICS S02 REMOVAL G RECOVERY PROCESS

-------
                               ELECTROLYTIC   CELL  SYSTEM

                          "A" BANK
CELL FEED TANK
CCLL FCCO
•»
5
r»
P>
•CCTCLE
RECYCLE
p
J
ACID 1ULFATE
CAUSTIC
r


~?—
k



f
ANODE CATI

	 1 1
D M
• i • 1 1


V B
ANOI
OR


f
ANK
.YTE
UM
,
«••__
I-'



f
J
WOE CATI
-
ANODE CATHODE
0 M
I 1
J
—
ANODE CATHODE
	 1 1
D M
— 11
.1

"A"/'B"BANK
CATHOLYTE
DRUM
V BANK
ANODE
COOLER





(



c
I



T-'




f

CATf


CATt
•
«H>C | UK
1 1 1
It D M
1 1 1
J L-

K>DE I AN(
1 1 1
M D M
1 1 1
J L_

IOOC f AW
1 f 1
MOM
1 1 1
^1 1^
IDE


>OE


»DE
—
V/'B'BANK "B'BANK
CATHODE ANODE
COOLER COOLER






;







f


VB
ANOL
OR
}





' — r
ANK
YTE
JU
)
—

ACID SULFATE
RECYCLE f-
DRUM '
( )





^



W




^
t                                                                                                     RECYCLE
                                                                                                     DRUM
                                                                                                         ••Ten
                                                                                                      10%
  CELL FEED
    PUMP
             CIRCULATION
               PUMP
"AW CELL CATHODE
   CIRCULATION
      PUMP
B' CELL ANODE  ACID SULFATE
 CIRCULATION    RECYCLE
   PUMP       PUMP
CAUSTIC
RECYCLE
 PUMP
                                               FIGURE  1-2
                                     REGENERATION   SECTION
                       STONE & WEBSTER /IONICS SO, REMOVAL & RECOVERY PROCESS

-------
Oxygen is generated at the electrolyzer anodes as a pure gas.  Hydrogen
is generated at the cathodes as a pure gas*  If not recovered as
"byproducts, the hydrogen is burned and the oxygen is vented to the
atmosphere.
In the four compartment electrolyzers , and anode product is dilute
equivalent to about 6 to 8 percent of the SO^ absorbed.  Being commercially
pure, it is vithdrawn from the system for marketing, or other disposal.
The sulfate ion so removed as dilute acid represents the method by which
S03 contained in the flue gas from the boiler plus the S02 oxidized in the
absorber to 803 are removed from the system.  The total amount of SO^
determines the required number of four compartment electrolyzers and the
amount of dilute acid produced.

    3.2  Technical Considerations

         3.2.1  Absorption

In the absorption step, capital investment and operating cost are influenced
by causxic utilization and by sulfite oxidation.  The following sections
explain and describe the influence of each.

                A.  Caustic Utilization

The first important process variable is the caustic utilization in the
absorption system, that is, the mols of SOa reacted per mol of caustic
(NaOH).  The 8 percent caustic at the top of the absorber will first react
with the C02 in the flue gas to form sodium carbonate.  The latter will
then react with the S02 in the flue gas to form sodium sulfite (NajjSO^).
The partial pressure of S02 above a solution containing sodium sulfite is
essentially zero, so it is possible to absorb virtually all of the SOg from
the flue gas.

In the lower part of the absorber, the entering flue gas reacts with the
sodium sulfite as follows:

                Na2S03 + S02 + H20 - >• 2NaHS03 (sodium bisulfite)

As the conversion of sodium sulfite to sodium bisulfite occurs, the
absorption work accomplished by each mol of caustic fed to the absorber
increases.  For example, if the fluid removed from the tower is a  solution
of sodium sulfite in water (neglecting the sodium sulfate recirculated and
any formed in the tower), one mol of S02 can be removed from the flue gas
for each two raols of caustic entering the tower.  If all of the sodium
sulfite is converted to sodium bisulfite, one mol of S02 can be removed from
the flue gas for each mol of caustic entering the tower.

There is, however, an inherent constraint that limits the extent to which
sodium sulfite may be converted to sodiura bisulfite in a practical situation;
it is that the partial pressure of S02 in the flue gas entering the absorber
must exceed the equilibrium partial pressure of S02 exerted by the sodiursi

                                   1067

-------
 bisulfite solution.   The partial pressure of 802  in the  flue gas entering
 the absorber depends on the sulfur content of the boiler fuel and the
 amount of excess air fired.  The equilibrium partial pressure of S02
 exerted by the sodium bisulfite solution depends  on the  concentration of
 bisulfite in the solution and the absorption temperature, both of which
 variables can be influenced by the engineer designing the system.  Addi-
 tionally, the design of the absorber tower, which the engineer controls,
 influences the rate  of transfer of S02  from the flue gas to the solution.

 Accordingly, the absorption system can  be designed  to optimize the production
 of sodium bisulfite,  taking into consideration absorber  capital investment,
 solution recirculation expense,  and electrolyzer  system  investment and opera-
 ting cost.

                 B.   Oxidation

 The second major consideration in the absorber design is the extent of
 oxidation of sulfite  to sulfate  by oxygen absorbed  from  the flue gas in
 the absorber.   502 in the sulfate form  cannot  be  released by neutraliza-
 tion with recycled dilute acid solution.   Sulfate generated by oxidation
 in the  absorber poses no  technical problem but imposes some economic
 penalty,  reflected in both the investment and  operating  costs of the
 entire  process.

 Operating costs  are increased because more caustic  is required to remove
 S02 as  sulfate  than as bisulfite.   Capital investment also is increased by
 the need  to  include in the process a four compartment electrolyzer for
 rejecting the excess  sulfate  as  dilute  pure I^SOi^.  This four compartment
 electrolyzer, described in section 3.2.3,  achieves  rejection of sulfate
 without loss by  the overall system of valuable sodium ion but with some
 additional sodium ion recirculation.

 Since absorbed oxygen reacts  slowly to  form sulfate, the engineer's objec-
 tive  is to design an  absorber  that  will not absorb much  oxygen, but will
 rapidly absorb a substantial portion of the S02 as bisulfite.  Based on
 performance data from packed tower  in S02  removal service, a system
 removing  S02 to the 150 to  300 ppm  range  should experience an oxidation to
 sulfate of approximately 6  to  8 percent of the S02  transferred fron the
 flue gas  to the liquid.

          3.2.2  Sulfur Dioxide Recovery

The liquid effluent from the absorber is mixed with a recycled dilute acid
 solution before entering the stripper.   The following reactions take place:
             2NaIIS03
                                  1068

-------
The S02 is taken overhead from the stripper, sufficient trays being
provided to insure that 99.5 percent of the S(>2 generated in the above
reactions is removed from the liquid.

The stripped sodium sulfate solution is cooled, adjusted in pH, filtered,
and fed to the electrolyzers.  It is desirable to maintain the feed slightly
alkaline in order to keep the cation membrane resistance at a minimum.
Therefore, the pH of the stripped sodium sulfate solution is adjusted vith
caustic.  Addition of the caustic causes heavy metal ions (leached out of
fly ash in the absorber) to precipitate from solution as hydroxides, which
are removed by filtration and ion exchange.

Absorber output varies with plant load, and average electrolyzer operation
is a function of average caustic production.  In a plant with a large peak
demand, it is possible to shut off cell power during the peak and run at a
slightly higher rate when off peak.  This can be done by providing storage
for sodium sulfate solution, caustic and acid, thus allowing absorber and
stripper operation to follow power plant loading, and the electrolyzers to
operate on the spinning reserve power available from the generators during
off-peak periods.

         3.2.3  Absorbent Regeneration and Oxidation Product Rejection

The unique feature of the Stone & Webster/Ionics Process is the electrolyzer
system in which the caustic is continuously regenerated.  Reference has been
made above to the use of two different designs, a three compartment elec-
trolyzer and a four compartment electrolyzer.  The former is the basic
design which converts sodium sulfate into two streams, sulfuric acid and
caustic soda.  The four compartment design represents a further refinement
and is the means by which excess sulfate is removed from the recirculating
liquid system as dilute, pure sulfuric acid.

A schematic diagram of the three compartment (Type A) electrolyzer is
presented on Figure 2—1.  The main components are an anode, a microporous
diaphragm, a cation-transfer membrane, ar.4 a cathode.  These components are
separated from each other by flow directing spacers which also provide the
required gasket ing.  The sodium sulfate is fed to the central compartment
at a concentration of about 20 wt percent.  Sodium ions migrate through the
cation-transfer membrane toward the cathode under the influence of a direct
current voltage impressed across the electrolyzer.  At the cathode, water
Is electrolyzed to hydrogen gas and hydroxide ion as follows:

             HO + e~ - » ?#  («) * OH~ (cathode)
The hydroxide ions are electrically balanced by sodium ions entering the
cathode compartment through the cation-transfer membrane.  Thus, the catho-
lyte effluent consists of NaOU plus I^SOfc which, after disengaging hydrogen,
is sent to the caustic storage tank and then to the absorption tower.
                                   1069

-------
 The function of the cation-transfer membrane  is  to prevent physical mixing
 of the catholyte and center compartment  feed  streams.  Essentially, only
 sodium ions from the center compartment  pass  through the membrane to combine
 with hydroxide ions produced at  the cathode.   In industrial applications
 other than S02 removal  by the Stone & Webster/ Ionics Process water may be fed
 to the catholyte compartment in  sufficient  quantity to produce  6 to 10 per-
 cent caustic.   However, the Stone  & Webster /Ionics Process is an essentially
 closed system.  Therefore,  to maintain water  balance in the system it will
 generally be necessary  to feed recycled  sodium solution to the  catholyte
 compartment instead.  This  procedure does not introduce any problems.

 The center compartment  feed stream passes through the microporous diaphragm.,
 into the  anode compartment.   At  the anode,  water is electrolyzed to hydrogen
 ions and  oxygen gas:

                 H0 - *  %Q (g)  + 2H+ +  2e~ (anode)
The hydrogen  ions combine with sulfate ions to form sulfuric acid.  The
diaphragm flow is designed to prevent hydrogen ions from migrating across
the cat ion- transfer membrane.  Such flow must give enough linear velocity
through the diaphragm to sweep hydrogen ions back into the anolyte compart-
ment.  The diaphragm should have an hydraulic resistance adequate to insure
that such flow is substantially uniform over the entire surface.  A flow
which is sufficient to sweep back most of the hydrogen ions carries with it
about half of the sodium in the center compartment.  Thus at the anode, only
part of the sodium sulfate in the feed stream is electrolyzed to sulfuric
acid.  The anode product is therefore a mixed solution containing both
sulfuric acid and sodium sulfate.  This product is the dilute acid solution
that is recycled to release S02 in the stripper.

In the four compartment (Type B) electrolyzer, a schematic diagram of whicn
is presented  on Figure 2-2, the reactions and operation of the catholyte
compartment are exactly those described for the three compartment electro-
lyzer.  The catholyte effluents from both electrolyzer types are combined
before being  pumped to the absorber.  The center compartment feed flows
through the porous diaphragm, into the "mid-anolyte compartment," from
whence most is discharged from the electrolyzer.  This stream contains some
sulfuric acid due to the inability of the membrane to exclude completely the
hydrogen ion.  It is combined with the anolyte stream from the three compart-
ment electrolyzers .

The important reactions of the four compartment electrolyzer occur at the
anode, where  water is electrolyzed to hydrogen ions and gaseous oxygen.
These hydrogen ions combine with sulfate ions entering the anode compartment
to form a sulfuric acid anolyte.  The function of the anion-transfer membrane
is to exclude sodium from the anode compartment while allowing sulfate to
enter.  It also, to a large extent, prevents hydrogen ions from leaving the
anode compartment.  Thus, the anode reactions permit removing sulfate ions
without losing sodium.
                                      1070

-------
                     1/2 02
  H2S04
                          D      CM
                          * Na2S04
   D-POROUS DIAPHRAGM
  CM* CATION SELECTIVE MEMBRANE
                              M
                               Na2S04
NaOH
Na2S04
                         FIGURE 2-1
     SCHEMATIC DIAGRAM OF "A" ELECTROLYTIC CELL
                      H2S04
                1/2 02  Na2S04
H2S04
* J >
NaOH




1
© AM D CM 0




1/2 02
j fc
f —
4. "*"
-»*2H
t
+ Na2S04
-S04« •*— S04« -

1
>
	
20H-*-
It



1
Na2S04



t
t
POROUS DIAPHRAGM Na2S04
 H20
 CM • CATION SELECTIVE MEMBRANE
 AM • ANION SELECTIVE MEMBRANE

                         FIGURE 2-2
     SCHEMATIC DIAGRAM  OF "B" ELECTROLYTIC CELL
                           1071

-------
 A large number of materials  of construction have been tested in order to
 optimize the life and performance of the electrolyzer components.  Anodes
 are a lead alloy; cathodes are steel.  The  microporoua diaphragms and ion
 transfer membranes are polymeric organic materials.

 Each electrolyzer consists of  an anode, a cathode, at least one membrane,
 and one diaphragm vith appropriate  separators and internal flov distributors.
 A second membrane is  placed  between the diaphragm and the anode in the Type B
 electrolyzers.  Fluid is  internally manifolded and is in parallel.  The
 electrolyzers are arranged into a number of modules within a single struc-
 tural frame.   The electrolyzers within a module are in parallel electrically
 and the modules are in series.  In  order to conserve space and minimize field
 piping,  three frames  are  arranged in a vertical array to form a bank.  The
 internal fluid manifolding is  brought to headers in the frames.  The headers
 are provided  with auick disconnect  fittings to the electrolyzer fluid distri-
 bution and collection mains.   Typical arrangements of modules, frames and
 banks are shown schematically  in Figures 3-1 and 3-2.

          3.2.U  Electrolyzer System Auxiliaries

 Hydrogen produced by  the  cathodes is gathered in standpipes and passed to
 gas  collection mains.   Caustic is collected from the bases of the stand-
 pipes and pumped  to the caustic surge tank.  Oxygen is similarly collected
 from anode  fluids.

 Since Ohmic losses must be dissipated from  the electrolyzers, heat exchangers
 are  provided  for the  recirculating  anolyte  and catholyte streams.  The anolyte
 coolers  have Karbate  tubes and the  catholyte coolers carbon steel tubes.
 The pure, dilute acid, produced in the Type B electrolyzers, will be approxi-
 mately 9 percent acid.

      3.2.5      Potential Problems and Process Limitations

A number  of problem areas have been identified and potential solutions
postulated.  These include:

     1.  Oxidation of S02 to 803}  amount and affect on economics.

     2.  Effect of oxidation inhibitors if used, on system and cells.

     3.  Effect of NOX on system and cells.

     4.  Effect of fly ash on system;  which minerals, if any, affect
          oxidation, absorption, filtration, cell performance and
          cell life.

     5.  Effectiveness of quench section and absorption section in
          removing fly ash.
                                  1072

-------
     6.  Effectiveness of ion exchange in removing heavy metals from
          solution.

     7.  Deleterious impurities in recovered S02; nature, amount,
          and effect.

     8.  Efficiency of absorption, stripping, and electrolysis.

     9.  Materials of construction.

    10.  Disposal of weak acid from "B" cells.

None of the above problem areas appear to be without a reasonable solution.

The potential process limitations mainly lie in the area of efficiency in
absorption, stripping, mimization of cell power, and the ability of the
system to produce a high bisulfite to sulfite ratio.

Pilot plant work in Tampa and at Watertown has given indications of what
may be expected but long run testing on commercial sized equipment will be
required to give reliable answers to many of these areas.  Corrosion test
coupons that are being installed in the WEPCO pilot plant will aid in
selection of materials for the prototype plant.

U.  PILOT PLANT

The pilot plant at Wisconsin Electric- is being installed primarily to
verify operability of the Stone & Webster/Ionics process in an integrated
operation under Wisconsin Electric conditions, and to confirm the process
design of the 75 MW prototype plant.  The pilot plant will process a slip
stream of about 2,900 ACFM of flue gas at 300 P containing about 1,000 ppm
of sulfur dioxide from a coal fired boiler.  It is designed to remove a
maximum 95 percent of the S02*

As in any coal fired boiler, it is to be expected that the sulfur content
in the flue gas will vary with the quality of the coal.  At Wisconsin, the
amount of sulfur is known to vary from as low as 1/2 percent to as high as
5 percent.  Since it is desirable to operate the pilot plant at reasonably
uniform conditions during any one test, a by-pass from the discharge of
the absorber to the suction of the forced draft fan has been installed to
permit adjusting the S02 content of the flue gas to a fixed value for each
test.

The absorber contains three stages of contact with provision to operate on
two only if desired.
                                   1073

-------
           H2S04



>
I"'
T;

k JV *|
ft

JJ
'T
i
i
i
i
n!

J
y
fl

i
LI
IT,

r
11
ft

|
Lj
i
i
i
i
i

u
r
i
It

u
u
ft

J(
•I
1




                                                   NoOH
 M'MEMBRANE
 D-DIAPHRAGM
FIGURE 3-1
MODULE
                                                   ,so.
II20W4 "•— * 	
^ 	 1± t
NoOH
TOP
FRAME
MIDDLE
FRAME
i
__c
__/
BOTTOM
FRAME
No!


^
4
1
i



I
i
M
^
i
M
ft
^^



^
L
M
t t



> ^ ^
••

^

t
M
ii

\ \

>
t
M


I

V
j
^
1
n*T
M
f

>
k
4
M


J

>i>
M
1*

4 * 4t A t

\
k


M

^ h
M
>lf-

— «

T t r

' k , _ _^ ^ ^
>t >
.
^ k
M
'A ^
M
h
J
	 J-*-H2S04
' ? **'
S
M
1

^
<
<

~^~s
	 3
c 	 s
M
T T T

r

IY
«
T^^^"T^^.

t


]
k ^
-n*
M
^
t ^
-n*
M

t



J
k
M




M

T '
J1+ l-n*

M
k ' M >
M
M
j
l
M
T * * *
FIGURE 3-2
/
t
^ ,
t


No2SC
\ —
4
\
i_-_
U
V
»
t
M*MODULE
  BANK
                             1074

-------
The pilot plant contains all the elements that present knowledge indicates
will be required in the 75 MW prototype plant.  The equipment, in so far
as possible, is of a type that can "be obtained f.or the larger plant.
Materials of construction have been selected based on previous experience
to be suitable for commercial operation.  In addition, test coupons of
various materials are installed at strategic locations so as to provide
information on alternate materials of construction.

The testing program is expected to begin in May 1973 and extend for about
six months.  Data will be obtained on:

     a.  Mass transfer coefficients in the absorber and stripper;

     b.  Extent of S02 oxidation to 803;

     c.  Effect of operating variables on S02 removal and on
          sulfite-bisulfite ratio in the rich liquor;

     d.  Performance data on electrolyzers, including power
          consumption for both types of electrolyzers under
          field conditions and over a range of operating
          variables.

Concurrent with the pilot plant testing, prototype electrolyzer modules
similar in design to that intended for installation in commercial plants,
will be operated at Ionics1 tfatertown facility to obtain data for the
design of the full-sized electrolytic "cell system.

5.  75 MW DEMONSTRATION (PROTOTYPE) FACILITY

Following successful demonstration of the pilot plant, Stone & Webster will
proceed with design and installation of the 75 MW Demonstration Facility,
assuming that the estimated costs of the system are acceptable.

    5.1  Basis of Design

The basis of design is:

     1.  Process all the flue gas from one 75 MW boiler at 100#
          load factor; approximately 225,000 ACFM at 300° F

     2.  Average sulfur content at 2,5 wt % in the coal.

     3.  Flue gas to be desulfurized to 200 ppm or lower.

     k.  Oxidation of S02 to be a maximum of 10 percent.

     5.  Minimum ratio of mols SO2 absorbed per atom of recirculated
          sodium equal to 0.87.
                                   1075

-------
 Figs. 1-1 and 1-2, referred to earlier, show the major equipment and flow
 schemes for the 75 Mtf unit.

 The Facility will be designed to remove a minimum of 90 percent of the S02
 from the stack gas.  A minimum of 90 percent of the absorbed SOg will\be
 recovered as S02 and the balance removed as dilute silfuric acid.

 Anticipated pressure drop across the absorber,  including duct losses, is
 eight inches of water.

 The recovered S02 vill be dried, compressed, and liquefied for sale.   It is
 believed that during the test program an outlet can be found for the  9 wt %
 sulfuric acid purge stream, amounting to about  five tons per day as lOOjJ
 H2SOJ|.

     5.2  Operating Program

 The primary purpose of the prototype plant is to provide information  on the
 operating and maintenance characteristics of the process, and to develop
 conclusions with respect to its  technical and economic feasibility.  The
 program includes:

      1.   Systematically varying  the  primary process variables over
           the practical operating ranges,  and collecting the data
           required to characterize the process;

      2.   Developing correlations to  assist with  scale-up of the
           process  equipment and  estimating performance under a
           variety  of operating conditions;

      3.   Determining the electric power consumption of the
          electrolyzers;

      kt  Evaluation of  the  operator  work load and maintenance in
           the  electrolyzer  and non-electrolyzer  sections.

      5.  Evaluating  the  effects  of corrosion, to serve as  a basis
           for  selecting  materials  of construction for  future
           installations.

6.  PROCESS ECONOMICS

Since many large boilers  supplying steam for  power  generation are in  the
UOO to 500 MW range, we  have developed  process economics  for this paper
for a 1*00 MW unit instead of the  smaller 75 MW unit  planned for  Wisconsin.
We believe this will be more representative of the  average  power plant
installation.  Cost of utilities  do  not  necessarily  reflect  WEPCO costs
but are those generally used in assessing  costs  of  competitive processes.
                                  1076

-------
An operating load factor of 80 percent for the UoO MW unit has been used
for the utility and a k percent sulfur coal has been assumed.  Credit
taken for the byproduct sulfuric acid is deemed to be reasonable for the
sale of the acid by the utility.  Fuel value only is assumed for the
hydrogen even though there may be particular locations at vhich hydrogen
could be sold at a much higher value.  No credit is taken for the byproduct
oxygen.

Approximate installed costs and operating costs for Stone & Webster/Ionics
S02 removal unit for a UOO Mw coal fired steam generating unit are given
in Table 1.  Included is the cost of facilities to convert the S02 to
commercial grade sulfuric acid.  Alternatively, the S02 could be dried and
sold as liquid product or could be converted to sulfur.

The fixed charge rate assumes 100 percent financing by industrial revenue
bonds at 6 percent interest, straight line pay backs of principal,
accelerated depreciation under IRS Section l69» and a 7 percent discount
rate for all cash flows.  Such industrial revenue bonds and accelerated
depreciation have become very popular for financing pollution control
plants.  It is assumed that these provisions will remain in effect.  They
permit a fixed charge rate of 7.1 percent per annum after taxes on the
20 year life assumed for this project.

The indicated annual cost of about $2 million for S02 removal is influenced
by the rate the station charges for power used in the electrolyzers and by
the credit assigned to the by-product sulfuric acid.  As shown in the foot-.
note of Table 1, the net annual cost of the plant may be defrayed by (a) an
average increase of 0.73 mills in the price of sold power; (b) an increase
of $15.00 per ton in the price of sold sulfuric acid; (c) a decrease of
9 mills in the station service electric rate; (d) a decrease of $1.83 in
the price of purchased coal; or (e) any combination of these.  Each of
these alternatives includes a provision for income taxes of 50 percent.
As indicated by the footnote, the trade-off point between S02 removal and
the purchase of low sulfur coal is a premium of $1.86 per ton for the latter,

The energy consumption of the electrolyzers is a unique type of station
service account.  The absorption section must process flue gases as
generated, but the sulfur dioxide recovery and the absorbent regeneration
and oxidation product rejection may be decoupled from a real-time, opera-
tional standpoint.  (They may be geographically removed as well, connected
with the absorption section by relatively small, liquid-handling pipelines.)
Electrolyzers, when on line, can be instantly interrupted without damage,
either in blocks or all at once at the sole control of the utility dispatch
center.  Power pool agreements as well as plant operating principles provide
for spinning reserve capacity which can be made available from generators
Vhich are spinning and generating power at a rate less than nameplate
Ratings.  Such reserve is usually sufficient to cover a predetermined per-
                                   1077

-------
centage of total system load or a percentage of the load on the largest
generator in the system.  If the generator capacity represented by this
reserve is under the sole control of the utility dispatch center, there
should be no bar to using this reserve for performance of useful work.

Such an interruptible station service account is instantly available in
comparison vith the usual "interruptible" customer and should be regarded
as loaded spinning reserve.  It therefore seems reasonable to charge the
removal system only for the marginal cost of fuel, labor, and maintenance
materials vithout capital charges.  If this approach is followed and the
electrolyzers are operated on spinning reserve power, the S02 removal cost
might be reduced by about 0.3 mills per kwhr.

The regeneration subsystem is amenable to evolutionary development.
Significant reductions in energy and maintenance costs of the electrolyzers
may be confidently predicted, based on pilot plant results to date.  Such
improvements should result in the electrolyzer characteristics shown in
Table 2.  The 1975 components should result in 15 to 20 percent saving in
electrical energy consumption.
                                  1078

-------
                                             APPROXIMATE ANNUAL OIERATING COST
                                         STONE & WEBSTER/IONICS 302 REMOVAL PROCESS
                                       Basis
o
^j
IO
  Installed Cost

    302 Removal Unit
    Sulfuric Acid Unit
Annual Cost

  Fixed Charges after Taxes

  Operating Cost - S0% Removal Unit
    Steam
    Electricity (cells, pumps, blowers & lighting)
    Cooling Water
    Process Water
    Deionized Water
    Operating Labor
    Maintenance
    Subtotal

Operating Cost - H2S04


Credits
  Sulfuric Acid
  Hydrogen

    Subtotal
Total Annual Cost
                                           400 Mw Station Burning 4.0 Wt
                                                  Operating Load Factor
 $14,000,000
   2,500,000
 $16,500,000

 Unit  Cost

& per  annum
                                                  $0.60 per 1,000 Ib
                                                  $0.007 per kwhr
                                                  $0.02 per 1,000 gal
                                                  $0.05 per 1,000 gal
                                                  $1.00 per 1,000 gal
                                                  $10.00 per hrU)
                                                  4^> per annum
                                                      $2.00 per ton
                                                      $15.00 per ton
                                                      $0.60 per MM Btu (LHV)
    Notes
      Equivalent Cost per Sold Kwhr
      Equivalent Increase in Sold Sulfuric Acid Price
      Equivalent Decrease in Station Service Rate
      Equivalent Decrease in Cost of Purchased Coal (l,083,OOO ton)

      (l)  Includes  overhead, and supervision
      (2)  All  costs including maintenance are represented.
                                                                             S Coal
                                                                                    Quantity
                       70,000 Ib per hr
                       31,900 kw
                       8,500 gpm
                       650 gpm
                       40 gpm
                       2 oper/shift
                                                                              465 tons per day
                                                                              465 tons per day
                                                                              24.7 MM Btu per hr
                                                                            0.73 mills or
                                                                            $15.00/ton or
                                                                            iiO.OOQ/Kwhr or
                                                                            $1.88/ton
                                                                                                     Annual Cost

                                                                                                     $1,174,000
   313,000
 1,570,000
    76,000
    14,000
    18,000
   175,000
   560.000

$2,726,000

   271.000
 4,171,000


(2,035,000)
  (io4tooo)

(2,139,000)

$2,032,000
                                                                                                                   I
                                                                                                                   sr

-------
                       ELECTRQLYZER CHARACTERISTICS

               STOME & WEBSTER/IONICS S02 REMOVAL PROCESS

                        Basis:  Plant of Table I
                                                                     Table 2
Elect rolyzer
    Type

  1973 "A"

  1973 "Bn
1975 "A1

1975 "Bf
Voltage

  4.6
                          3.8

                          4.2
   Current
Efficiency

    95
                       95
                                                                Kilowatts AC

                                                                     24,400
               Total 1973   2


                            21,300

                             2.600

               Total 1975   23^900
(l) First number is overall current efficiency.
    Second number is current efficiency for the pure

(2) Cell requirement only.   Added 3,000 few in Table 1 for
    pumps, blowers, etc.
                                  1080

-------
FW-BF DRY ABSORPTION SYSTEM
             FOR
     FLUE GAS CLEAN UP
              by

         W. F. Bischoff
   Foster Wheeler Corporation
     110 South Orange Avenue
     Livingston,  New Jersey
              1081

-------
ABSTRACT
               The information contained in this paper is directed
to those individuals in private industry and governmental agencies
concerned with the reduction of S02 in flue gas emissions.  It is
intended to serve as a means of presenting information on the history
of the dry adsorption process developed by Bergbau-Forschung, GmbH,
process description, technical details, pilot plant experience,
demonstration plant design and economic factors.
                              1082

-------
                      FW-BF Dry Adsorption System
                                 for
                          Flue Gas Clean up
                                 by
                           W. F. Bisehoff
                     Foster Wheeler Corporation
I.  Introduction

           Foster Wheeler is a major manufacturer of steam generation

equipment and also a major process plants contractor.  In both of these

roles we have been actively concerned with pollution problems and their

proper solutions.  To this end the Corporation has participated in consid-

erable research and development work in order to solve the pollution pro-

blems of our customers.

           Our research program has included) in part, the extent*j..~

technical and economic evaluation of various pollution control systems.

One of the systems on which we have expended considerable time is the

subject of this paper.

II.  Background of Bergbau-Forschung* GmbH

           Bergbau-Forschung, GmbH, (BF) is the central research institute

for the German coal mining industry.  They are industry supported and there-

fore analagous to our own Bituminous Coal Research Institute..

           In 1963 Bergbau became interested in the  application of chars

produced from coal to remove sulfur dioxide from stack gases.  The chemical

reactions involved for adsorption were well known but, the means for doing

it successfully on a continuous basis were not.  Bergbau  therefore determined

to produce a char from non-coking coals which would  meet  the  system require-

ments they had conceived.  The char had to have the  following characteristicst

           1.  High mechanical strength

           2.  High S02  adsorption capability

           3.  High ignition point

           li.  Low pressure drop when put in a  travelling bed reactor

                                  1083

-------
                   •L'iK^fSy^f^f:'.- -;•-.* ;-*K,
                   vH,:*.•;>•?>?'>: *>j<* -^ > ».' - ••-• -.'t*' ;-~- X -Jinf* .-
                   ...•v."o -'i- » '>-M> - -,^-<---. :,•- ->.;*--V^-:-..
••v::.;^: x-^w^-^-':
.•-X^MV-; ^'^-S- •:•'-.-:
 -f " • ' ^: -i \ •''--,- ^-' •'• ^' ' •

                                              A6 R AM^::a«^SSS:^:"
                                              ** "  -'. ^ -"' • *•?-, *^*V-^' 'j--** •  " '•' --r '• ' '• J '*- '- fc-'i1'' -'if^^r^ \ '.'-' •'•  -r' -  * -
                                              J" -..•.."••••• ' " "*-%'-i:^l»" .j7 - vv * •", " Tj.rj,* •  •", " • '"-'"'V Ji ^'ii *'"<"" ;i-r:.'3 " . ^" •••'.-.-." ' •':".	
o
00
                   BOILER
                                         CHAR FROM
                                       REGENERATOR
              DUST

            COLLECTOR
                              w
ADSORBER
                             I
                                         CHAR  TO
                                       REGENERATOR
                                                                /STACK
                                                                         FW-I3I-004



                                                                             Fig. 1

-------
Bergbau was successful in this endeavor and then built a laboratory scale
pilot unit to test their system.  After the successful operation of the
laboratory unit Bergbau built a pilot plant of about 1.$ MW equivalent which
was operated in a slip stream from a coal fired utility boiler.   The gas flow
through the pilot plant was 3000 m^/hr. (10J>,000 SCFH).  This pilot unit was
run continuously for two years during which time all of the correlations
developed during the laboratory work were confirmed.  Although the system  was
basically developed for  removing S02, it was discovered in the pilot plant
work that the system also has capability to significantly reduce NO^ and
particulars in the flue gas stream.
           As a result of the work done by Bergbau as well as our own develop-
ment efforts on the system Foster Wheeler is convinced that the Bergbau system
is the most feasible for application to a wide variety of boiler flue gas pollu-
tion problems.  Because of this conviction Foster Wheeler has become fully
licensed under the Bergbau technology and patents.
           During the period that Foster Wheeler has been licensed for the
process, development work has continued both in Germany and in the United
States.  The development work has resulted in what is known as the Ftf-BF Dry
Adsorption System.
III.  Process Description
           Figure 1 shows how the FW-BF System fits into a typical utility
boiler system.  The adsorber is the key to this system.  It consists of
parallel louvre beds which support and contain the char during the adsorption
phase.  The char moves slowly downward in plug flow while the pollutant laden
gases pass through the adsorber bed in a cross flow.  The fan indicated at the
gas discharge of the adsorber is utilized to make up the pressure losses which
occur during the passage of the gas through the adsorber bed.  Clean gases
                                  1085

-------
o
oo
                FLUE GAS (250-300 °F)
S0
                            I
                             CHAR PELLET

                             3/8" DIA. X3/4"LG.
                                                 FW-131-005



                                                     Fig. 2

-------
leaving the fan are directed to the stack for disposal.  Once the char has



progressed through the adsorber it is taken off and sent to the regeneration



section.  Once regenerated the char is recycled to the top of the adsorber.



This process scheme allows for the continuous adsorption of pollutants and



the continuous on site regeneration of the adsorbent media.



            The reactions that are taking place in the adsorber are shown



in part in Figure 2.  The rectangular block represents a single char pellet



which is approximately 3/8" in diameter x 3/8" - 5/8" long.  Boiler flue



gases containing S02, NOX, water vapor and particulate matter come into contact



with the char pellet.  The S02, oxygen and water vapor are adsorbed into the



char pellet.  Once in the char pellet the S02 is oxidized to SO^ and subsequently



catalytically converted to sulfuric acid which is firmly held in the interior



pore system of each char pellet.



          Nitrous oxides are likewise adsorbed in the char pellet and held



until regeneration occurs.  Particulate matter is collected on the surface of



the char pellets.



          The reactions in the adsorber are designed to take place through a



range of temperatures above 2£0*F.  The lower limit of 2$0°F is specified



because of dew point considerations.  The relationships between adsorber size,



adsorption efficiency and operating temperatures indicate that optimum



efficiency is achieved at temperatures around 27£° to 300°F range.



          As a result of the adsorption phenomena and the dust collection



capabilities of the system, it is possible to reduce pollution emission levels



as indicated below:
                                   1087

-------
                 Pollutants            Removal Efficiency
                 S02                         80-9#
                 NOx                         UO-60J6
                 Particulate matter          9
           The S02 removal efficiency is listed as a range to indieate\
flexibility of design.  In some cases, dependent on the sulfur in the fuel,
it may not be required to achieve a 95% removal efficiency to be in accordance
with the required codes.  Therefore economies in the design can be realized
by virtue of the lower removal efficiency requirement.  NO^ efficiencies are
indicated as a range for the same reasons.  The particulate removal efficiency
shown above is meant to indicate that the system is capable of removing from
90 to 95/5 of the particulate matter received in the inlet of the adsorber
after prior treatment with high efficiency electrostatic precipitators.
           The regeneration of the saturated char can be accomplished by wash-
ing (wet regeneration) or by contacting it with hot sand (thermal regeneration),
Wet regeneration produces dilute sulfuric acid (lS>% by weight) as a by-product
material.  Because of the limited use-for this material thermal regeneneration
is the more practical method of regeneration.
           In thermal regeneration an enclosed refractory lined vessel is
utilized to contact the saturated char pellets with hot sand which has been
heated to approximately 15>00°F.  Regeneration is a function of vessel size,
retention time, sand to char ratio, and sand temperature.  As the char pellets
become heated the reactions that took place in the adsorber are reversed pro-
ducing a concentrated stream of S02, 1^0, C02, and N2-  In other words the
I^SO^ produced in adsorption is converted to H^O + SO?.  The SO^ is reduced
to SC>2 in the presence of the hot char and produces C02 by the combination of
the liberated oxygen with the carbon in the char.  In a like manner the oxides
                                   1088

-------
             :.-^>JV^
             l$&&.-&'
             "i^*^^
              i$t's|
:^--^H'-':^i;^;?wr\*^.:-'i?^ "'*-£•- ':'• "-.••:'' ' . *:"-r.-^J^-< ^Af^K'^' If-'- ' < .- •'•• '-'^1'^VA^

                           isaiPSiii^
o
00
vo
        FROM
        DUST
        COLLECTOR


        CHAR FROM
      REGENERATION
                                       N.
                             HAR TO
                         REGENERATION

                          FLUE GAS
                         " TO STACK
                                                   FW-I3I-008


                                                     Fig. 3

-------
 of nitrogen are reduced to N2 with the resultant production of additional
 003.
          The concentrated stream of off-gas is then directed to a proprie-
 tary Foster Wheeler OGT system for conversion of the S02 to elemental sulfur.
 IV.  Technical Details
          The FW-BF System consists of three basic sections:
          1.  Adsorption
          2.  Regeneration
          3.  Off Gas Treatment
          The adsorption and regeneration sections of the system are those
which were developed and designed originally by Bergbau.  The Off Gas Treat-
ment section is a proprietary FW design for which we have made patent
application.
A.  Adsorption
          The adsorption section, as stated previously, is the heart of the
system.  It is here that the reduction of pollutants take place.  As can be
seen in Figure 3 the adsorber is a vertical bed in which the char moves slowly
downward at a rate of from 0.5 to 1.0 ft .per hour.  This slow but variable
movement allows for operational flexibility as well as guaranteeing low abra-
sion rates in the adsorber.  The louvres are designed to not only support and
contain the moving char bed but also to give direction to the incoming and
outgoing flue gases.  Flow distribution of the gases across the bed is main-
tained by means of adequate pressure drop.  The pressure drops that have been
measured for these types of adsorber beds vary from k" K2Q to li;" H20 depend-
ing upon the adsorber configuration.
          The adsorber has been designed based on the information generated
during pilot plant operation.  During this pilot plant operation several diff-
erent bed cross sectional areas were utilized varying from 0.£ meter x O.J> meter
                                   1090

-------
to 1 meter x 1 meter.  Commercially sized beds vary from U1 x li1 to 6' x 6*
thus keeping the scale-up factor for adsorption beds within reasonable lijnits.
The height of adsorber beds is varied to achieve the necessary frontal bed
surface area to bed volume ratio.  For large gas flows a series of beds are
operated in parallel so as to achieve the desired removal efficiency,  ty
designing large adsorbers as a series of modules it allows for maximization
of shop fabrication and minimizes field erection.
           The flow of char through the adsorber beds is controlled at the
bottom of each bed by a vibratory feeder which is controlled as a function of
the amount of S(>2 entering the adsorber.  This mode of operation gives flexi-
bility to the system not only from the standpoint of variations in load but
also variations in the percent of sulfur in the fuel.  As load decreases the
amount of S(>2 entering the adsorber also decreases and therefore it is possible
to reduce the flow rate of char so as to maintain a consistent removal effi-
ciency.  The reduction of char flow then results in a lowering of operating
costs for the unit.  For a given boiler load with an increasing percent sulfur
in the fuel the char flow rate can be increased so as to achieve a reduction
efficiency compatable with the fuel being fired.
           Upon leaving the feeder at the bottom of each adsorber bed section
the char is discharged onto a natural frequency or oscillating type conveying
system.  The end of this conveyor is fitted with a screening section  which
screens off the majority of the fly ash collected in the adsorber.  The fly ash
removed is in a dry form.  In order to eliminate fly ash losses from  the system
the entire system is totally enclosed.  Once the char and fly ash have been
separated the char is elevated by means of an enclosed bucket elevator and sent
to the regeneration section.  Upon completion of regeneration the char is
returned to the adsorber section, elevated in an enclosed bucket elevator  and
distributed evenly across the top of the adsorber beds.
           The entire adsorber and its associated distribution system is
                                   1091

-------
o
10
N»
       SATURATED
       CHAR FROM
       ADSORBER
                REGENERATOR
   REGENERATED
   CHAR  TO
   ADSORBER ""
 SAND
HEATER
                                                     !

                                                 C02  I

                                                 H20  f

                                                 N2 J
                  TO OFF-GAS
                  TREATMENT
                                                                FW-I3I-009


                                                                  Fig. U

-------
enclosed so as to provide a sealed system and eliminate any possible second-
ary pollution sources.
B.  Regeneration
           In Figure k the saturated char from the adsorber enters the
regenerator vessel where it is mixed with hot sand.  The char temperature is
raised as a result of this contact to a level where the reactions that have
occurred in the adsorber are reversed as previously discusseo.
           Sand is utilized as an inert heat transfer media and as such does
not take part in the reactions occurring within the regenerator.  Its sole
function is to supply heat so that the reactions may take place.
           The mixture of hot sand and hot char pellets flow slowly downward
through the regenerator and their flow is controlled at the discharge of the
regenerator by a char-sand feeder and separator.  This feeder-separator not
only controls the flow rate of the materials through the regenerator but also
acts to separate these two materials after regeneration has taken place.  The
char pellets, being considerably larger than the sand, flow overhead  on the
char-sand separator and are subsequently cooled and are then returned to the
adsorber for reuse.  The sand passes through the screen section and is returned
via an enclosed hot bucket elevator to a fluidized bed sand heater.   The
fluidized bed sand heater restores the heat lost to the char and recycles  the
reheated sand to the top of the regenerator.
           The production of C02 in the regenerator results in a chemical
consumption of the char pellets which accounts for approximately 90%  of  the
char makeup requirement.  The remaining 10$  char makeup is as  a result of
mechanical or abrasion losses.  As the char  is consumed chemically or reduced
in size mechanically it eventually becomes too small to pass  over  the top  of
the screening section in the char-sand separator.  At  this point the  small
quantity of small char particles enter the hot sand side  of the cycle and
                                  1093

-------
make  their way to the fluidized bed sand heater.  There, in the presence
of excess oxygen, they are consumed as a fuel.  However, the quantity of these
particles represent a negligible savings to the fuel requirement for the sand
heater.
           The hot flue gases eminating from the top of the sand heater
contain S02 as a result of fuel combustion and also contain a source of
recoverable heat.  These hot gases are piped back to the inlet of the boiler
air preheater where they are mixed with the boiler flue gases.  The heat in
this  stream is utilized in the air preheater to preheat incoming boiler air.
The additional S02 is now a part of the incoming flue gases to the adsorber
and is adsorbed in the same manner as the S02 from the boiler flue gases.
           That portion of the fly ash which was not removed by screening at
the discharge of the adsorber is removed from the char in the regenerator by
the gentle interaction of the sand with the char.  This material passes through
the char sand separator and enters the hot sand loop where it too acts as a
heat  transfer media.  The quantities of fly ash removed in this manner are
extremely small but eventually constitute a buildup in the inventory of material
in the sand loop.  It is therefore necessary to periodically remove a portion
of the sand fly ash mixture and replace it with a smaller quantity of new sand.
           The regenerator vessel utilized in this section is identical to
a vessel utilized in the production of the adsorbent char.  This regenerator
has been in operation for five years in a commercially sized char production
facility in Germany.  Therefore the problems associated with its design and
operation have already been solved.
           The fluidized sand bed heater utilized in the regeneration section,
while unfamiliar to many in the utility field, has been used for many years
by the process industries in applications identical to those contemplated here.
C.  Off-Qas Treatment Section
           Figure 5> schematically illustrates Foster Wheeler's proprietary
                                  1094

-------
             OFF-GAS TREATMENT
              CRUSHED COAL
       OFF-GAS
       FROM
       REGENERATOR
to
in
          REACTOR
                                          TO
                                          ADSORBER
                                            FW-131-010


                                              Fig. 5

-------


                                               •'-.-•-• x •••• *"'•'; '-'-' •. i^^i-i-^T, ••% ^*'--V." ' '"
                                               •'.'V^iCv-'^1 ;4\ "••:"*''''•*?•''J. :'*'•'' "'

                                               ^^^^•^
o
o
o»
                         »yii.vj^ .•*
          DUST COLLECTOR
          •••••••••••••••••••

              ^:^r?^^j;-;^^.,.::'l f/^'JC  "




          .'-•• •. &. v-:--jj••* '••''^•Oi^a*'%iiii
                                                                                 ^K'Svi- .Vr\^0|5^^p^t-^


                                                                                                c^ii^>Si
                                                                                                    .K. ^; .JI4...TT
                                                                                        ••• .•' •  'j- ^ '--' ,^-**S>SL.".3*
ASH
                                                                                        -- . : >;• j-^1""•?''?.*'    •-"??-^ •-- 'V^-'- |"-t-«*-:
                                          • -    •jK »;./.
                                                                                                         •
                                                 REGENERATION
  OFF  GAS TREATMENT
    - ^ ,-  •
                                                                                                                     iMi.
                                                                                                            Fig. 6

-------
design for the off-gas treatment section.  In this section concentrated
off-gas, rich in SC^, is received from the regenerator and converted to
elemental sulfur.  The reaction takes place in a reactor vessel which is
filled with crushed coal.  The SC^ stream is reduced to gaseous elemental
sulfur and the liberated oxygen combines with a portion of the coal tc forn
carbon dioxide.  The gases leaving the reactor enter a sulfur condenser where
the sulfur is condensed and carried off as molten elemental sulfur.  The
sulfur is collected and stored in an insulated tank which is equipped with
steam heating to make up for heat losses through the insulation system and
thereby maintaining the sulfur in a molten form pending shipment via heated
tank car or heated truck.  These two methods of sulfur transfer are common in
the sulfur industry and have been utilized for many years.  As an alternative
the sulfur could be pumped to cooling pits and allowed to harden for shipment
as a dry solid.
           The gases leaving the condenser consist of carbon dioxide, water
vapor, elemental nitrogen and those remaining sulfur values not converted to
elemental sulfur in the Off-Gas Treatment section.  These gases are recycled
to the front of the adsorber where the sulfur values are adsorbed on the char.
The C02, H20 and N2 are non-reactive materials at 33>0°F and therefore pass
through the adsorber bed and enter the atmosphere as non-polluting gases.
D.  System Recap
           The three sections of the FW-BF Dry Adsorption System are shown
together in Figure 6.  As can be seen the system is composed of a series of
closed loops which allows only cleaned gases or by-product materials to leave
the system.
           The requirements of the system are relatively simple.  Cooling
water is required to cool the char and also condense the sulfur.  The major
portion of this water is recoverable and recycleable.  In addition tc cooling
water the system requires char, dand, clashed jp.ant ocal ar.o aurf lifcj-y DTWS -.
                                   1097

-------
 V.  Maf1or Areas  of  Application
            The  FW-BF  system is applicable to all fossil fuel fired utility
 applications.  Primary application would be jji the area of large coal fired
 control  station  utility boilers because- of the system's inherent ability to
 handle all three of the pollutants associated with coal fired boilers.  Size is
 another  factor.  It is felt at least for the present that applications above
 150 MW size range will show superior economic advantages.  Of course it is possible
 to apply the system to smaller units which could be connected in parallel to a
 single FW-BF unit thereby achieving the economics of larger sized units.
            Similar units in refinery or chemical process plants could also
 be handled provided that similar size considerations are taken into account.
            Work is presently underway to determine application to metals
 smelting operation.  Initial studies, particularly those involving wet regenera-
 tion, indicate good potential application in this area.
 VI.  Demonstration  Plant Design
            At the  present time two demonstration plants for the system are in
 the design stages.
            One  contract is being executed in Germany which has been financed in
 part by  the West German government.  This contract involves the installation of
 a BF unit as a 35 MW equivalent slipstream off of a large central station boiler.
 The system will  use the thermal regeneration scheme and will be ready for
 operation in fourth quarter of 1973.
            At the  same time Foster Wheeler is designing an FW-BF unit to be
 installed at a major southeastern utility.  This unit is designed to meet State
 and EPA  code for S02 removal when burning 3% S coal.  The adsorber will be
capable of handling a 50/6 slipstream from a nominal UO MW coal fired boiler.
The thermal regeneration and off-gas treatment sections are being sized for
the full 1*0 MW flow.  The combination of half sized front end and full sized
                                   1098

-------
back-end will allow a comprehensive  testing program utilizing fuels varying

in ash and sulfur contents while minimizing capital expenditures during the

demonstration program.  At the completion of testing it is anticipated that a

second 20 MW adsorber will be installed thereby converting the unit to a full

sized lj.0 MW unit.

            High efficiency electrostatic precipitators capable of meeting

state codes for particulate emissions will be installed upstream of the FW-BF

system.  During the testing program sections of these precipitators will be

selectively de-energized thereby imposing higher dust loadings on the FW-BF

unit.  Measurement of dust loadings will be taken at the inlet and outlet on

the FW-BF unit to determine its actual efficiency as a particulate removal

system.  It is anticipated that the test work previously conducted by Bergbau-

Forschung will be re-confirmed on this large scale test and that the particulate

removal efficiency of the FW-BF unit will be firmly established.

VII  Capital and Operating Costs

            The capital and operating costs of the system vary as a function

of the unit size (expressed as the MW rating of the boiler) and the sulfur con-

tent in the fuel.  Capital costs represented as dollars per kilowatt can be

plotted against the megawatt rating of the boiler and the percent sulfur in the

fuel.  This plotting will result in a family of curves which  stabilize and become

relatively level after a boiler size of 150 to 200 megawatts  is reached.  Beyond

the 200 megawatt level relatively little decrease in the curves is witnessed.

This indicates that the most economic application of the system is in a range of

boiler sizes above l£o to 200 megawatts.  Below this level the fixed engineering

costs cause the cuves to rise more sharply so that the cost for units below

50 megawatts are uneconomic.  This uneconomic situation can be overcome, however,

by the connection of several smaller boilers to a single treatment system which

would then have a rating equivalent to the sum of the individual boilers connected

to it.
                                     1099

-------
             The capital costs in the range above 200 megawatts vary from $20 to
 $UO depending on the percent sulfur in the fuel.  $20 per kLllowat represents a
 system utilizing ,9% sulfur fuel and $UO per killowat represents a system utilizing
 U.3JC sulfur fuel.
             Operating costs vary from 80# per ton of fuel fired to $2.60 per
 ton as a function of the sulfur content in the fuel and within the limits of
 those sulfur contents described above.
 VIII.  Conclusions
             In conclusion we can say that the FW-BF Dry Adsorption System for
 Flue Gas Clean-up will:
             1.  Meet or exceed EPA and State codes for S02 removal
             2.  Significantly reduce NOX emissions
             3.  Significantly reduce particulate emissions
             Further we can summarize the advantages of the FW-BF Dry  Adsorption
 System as  follows:
             1.  No stack plume
             2.  No reheating of stack gases
             3.  Minimum water consumption
             k»  Low auxiliary power
             5>.  Minimum waste disposal
             6.  No handling of wet material
             7.  Handles coals of variable ash and sulfur contents
             8.  High S02 removal efficiency and the ability  to deal with S02, NOX
                 and particulates in one system
             9.  Low operating costs
            10.  Low capital costs
We therefore say that the FV-BF Dry Adsorption System is a second generation
gas cleanup system of superior capability.
                                        1100

-------
     THE ATOMICS INTERNATIONAL MOLTEN CARBONATE
      PROCESS FOR SO 2 REMOVAL FROM STACK GASES
                            by

W. V. Botts,  Program Manager, Environmental & Utility Systems
      R. D. Oldenkamp, Manager, Environmental Systems
                Atomics International Division
              Rockwell International Corporation
                   Canoga Park, California
                             1101

-------
         THE ATOMICS INTERNATIONAL MOLTEN CARBONATE
           PROCESS FOR SO^ REMOVAL FROM STACK GASES
                           I.   INTRODUCTION

The molten carbonate process (MCP) was invented at Atomics International
in 1965,  and has been under continuous development there since then.   In
the first phase of the process development program,  emphasis was placed
on process chemistry and materials corrosion studies and on bench scale
engineering tests.   This phase was completed in 1971; the results were
encouraging,  and the process was developed to the point where a pilot plant
test under  actual power plant operating conditions was both warranted and
required for further progress.   The initial process development work was
funded by Rockwell International Corporation; subsequent first-phase work
was done under contract to the U. S.  Environmental Protection Agency.
Con Edison has undertaken to help finance the pilot plant program, in  con-
junction with Northeast Utility Services Company (Hartford, Connecticut)
and Rockwell International Corporation,  United Engineers and Constructors
(Philadelphia,  Pennsylvania) are construction managers for the pilot plant.
The pilot plant phase of the process  development program is now underway.
A pilot plant has been designed and is now being built at the Consolidated
Edison Company's Arthur Kill Power Generating Station.  This pilot plant
will treat a side stream of 20, 000 scfm of flue gas from an oil-fired steam
generator (equivalent to about 10 mw of generating capacity).  This paper
describes the molten carbonate process in general and the pilot plant design
in particular,  and gives a status report on the pilot plant program.
                                  1102

-------
                       II.  PROCESS DESCRIPTION

A.    BASIC PROCESS FEATURES
In the molten carbonate process, a molten eutectic mixture of lithium,  sodrum,
and potassium carbonates is used to scrub the power plant gas  stream.   The
sulfur oxides in the gas stream react  with the molten carbonates to form
sulfites and sulfates, which remain dissolved in an unreacted excess of
carbonate melt.  The molten carbonate (sulfite-sulfate) mixture is then regenerated
chemically, converting the sulfite and sulfate back to carbonate and recovering
the sulfur as hydrogen  sulfide.  The regenerated carbonate is recirculated to
the scrubber to repeat  the process cycle,  and the hydrogen sulfide is converted
to elemental sulfur in a Claus Plant.
The regeneration of the carbonate is done in two  steps:  first the sulfite and
sulfate are reduced to sulfide, and then the sulfide is converted to carbonate
plus hydrogen sulfide.   The reduction is accomplished by reaction with a form
of carbon,  such as petroleum coke.  The conversion of the sulfide to carbonate
is accomplished by reacting the melt with steam  and carbon dioxide.
The basic process fluid is the molten- eutectic mixture of 32 wt.  % Li£
33 wt.  % Na2CO3, and  35 wt. % K2 CO3.  This mixture melts at 747 °F
(397 °C) for form a clear, mobile, non- volatile liquid.  At 800 CF, the melt
has the physical properties listed in Table I.

                                 TABLE I
               PHYSICAL PROPERTIES OF MELT AT 800°F
                     Viscosity             12 cp
                     Specific Gravity       2. 0
                     Specific Heat          0. 40
                     Thermal Conductivity ~ 0. 3 Btu'/hr. , ft. ,  °F (estimated)
The melt can be pumped and sprayed just like any other liquid.
                                     1103

-------
 In the process, the melt composition (and freezing point) changes as  sulfur
 compounds are formed and reacted.   The process will be controlled so that
 melts with freezing points above 850°F are avoided.  This limitation corres-
 ponds to melts containing about 30 wt. % sulfur compounds.
 The freezing point limitation makes  it necessary to keep the temperature
 above 850°F in all process equipment in which melt is handled,  including the
 scrubber.   The flue gas being treated must also be above 850*F when it enters
 the scrubber.  The ramifications of  this requirement are discussed further in
 the following sections of the paper.

 B.   PROCESS FLOW DIAGRAM
 The basic  process flow diagram is shown in Figure 1.   Each step in the flow
 diagram is numbered, and discussed in the corresponding section below.
 1.    Gas  Preparation
      The  gas to be treated is removed from the boiler at a temperature above
 850°F, or else it is reheated to above 850T.   In a new plant designed for this
 process and in some retrofit situations, the gas will be removed from the
 boiler at the superheater or reheater outlet or economizer inlet at a nominal
 temperature of 850°F.  If the boiler  is burning coal, the flue gas will be passed
 through an 850°F,  high-efficiency electrostatic precipitator where nearly all
 of the fly ash will be  removed.  Such electrostatic precipitators with
 efficiencies of greater than 99% are presently available.  If hot gas cannot be
 removed from the boiler, or if the heat needed to reheat the flue gas  can be
 used or recovered, the flue gas can be removed from the power plant air
 heater (or  low temperature precipitator if coal is burned) and reheated to
 a nominal  850°F.   The reheat can be done by direct firing with gas or oil,
 or by indirect heat exchange.  In any case,  the gas preparation step provides
 the flue gas stream at 850°F, with as much fly ash removed as possible.
 2.    Scrubbing
 The flue gas  stream then enters the scrubber, where it is contacted intimately
with the molten salt.  The sulfur oxides react chemically with the molten
                                  1104

-------
GAS TREATMENT-*-1—MELT PROCESSING AND SULFUR RECOVERY


 FLUE GAS       1     CARBONATE
 RETURN        I     MAKEUP
 TO BOILER 850oF |
                                              S  CLAUS
                                                  PLANT
                                       SULFUR
  SCRUBBER
850°F
PRECIPITATOR
 FLUE GAS
 FROM BOILER
                             REGENERATOR
                                                       CARBONATE
                                                       ANDSULFIDE
                                            QUENCH TANK
     EXCHANGER
                              COKE FILTER
                           ASH FILTER
                                       CARBONATE,
                                       SULFITE
                                       AND SULFATE
CAKE TO
REPROCESSING
    ^^
    10
                                                      AIR
                                                           PETROLEUM
                                                           COKE
              MELT FLOW
              GASES
              SOLIDS
                       Fig. 1.  Process Flow Diagram

-------
 carbonate to produce sulfur compounds and evolve carbon dioxide:
       M2C03 
-------
burning coal,  the ash is removed by filtering the side stream of melt being
processed.  If the power plant is burning oil, the particulate matter  in the
flue gas is mainly carbonaceous.  In this case, it should be possible to
eliminate the  ash filter, since the carbon will be consumed in the reducer
and the oil ash can be removed in the coke filter  (see Figure 1).  The need
to filter oil particulate s out of the melt upstream of the reducer will be
determined in the pilot plant.
The ash filtration step  is carried out in centrifugal basket-type filters,  which
produce a continuous "dry" cake discharge.  A filter aid,  such as petroleum
coke particles, may be used to enhance the filtration.  The filter cake is
collected and  treated to recover its lithium content, using a recovery process
described below (Item 10).
4.    Reduction
The melt next enters the reducer, where the sulfur compounds are reduced to
sulfide with carbon.  The sulfite disproportionates to sulfate and  sulfide as
the melt enters the reducer  (or even before) according to the reaction:
                                                                        (3)
The reduction reaction is thus actually that for sulfate only:

      M2S04(I)  +  2C (s) - —M2S(£) +  2C02(g)                        (4)

The melt temperature is raised from 850 °F to about 1600'F to enhance the
rate of reaction 4.   The heat required for this, plus the endothermic heat of
reaction, is obtained by re-oxidizing some  of the product sulfite to sulfate
with air:

      M2S(£) +  202(g) - »-M2S04                                      (5)

The net result of the heat generation reaction is  the sum of reactions 4 and 5:

      2C(s) +  202(g) - *-2C02(g)                                    (6)

The heat is thus actually generated  by the indirect combustion of excess carbon.
                                    1107

-------
In actual practice the reduction step will be carried out in a two-zone vessel,
using petroleum coke as the reducing agent.  The coke will be consumed in
the first (reduction) zone,  and air will be injected into the second (oxidation)
zone to generate heat.  Internal circulation will carry heat and reactants be-
tween the two zones.  This is shown schematically  in Figure 2.   The vessel
will be lined with  corrosion resistant high-density alumina refractory,  since
the melt is very corrosive at these elevated temperatures.
The reducer produces a molten mixture of  carbonate  and sulfide,  plus carbon
dioxide.   The carbon dioxide is sent to the  regenerator (Step 8) where it is
used as a reactant.
5.    Quenching
The reducer produces a molten mixture of  carbonate  and sulfide,  at about
1600°F.   This mixture  must be cooled back to about 900°F before it under-
goes further reaction.  The major part of this cooling is done by mixing the
hot melt stream with a  pool of cooler melt  (about 900°F) in the quench tank.
This procedure makes  it possible to cool the melt to below 1000°F without
using a direct-contact heat exchanger.
The quench tank is a large tank containing a melt pump; it is also used as  a
drain tank for melt storage during temporary shut-downs.
6.    Coke Filtration
Cooled melt from the quench tank is filtered to remove coke ash and coke
particles  carried over from the reducer.  Equipment similar to that described
for ash filtration (Item 3) is used.
The filter cake from this step is combined with that from the ash filter (if one
is used) and treated to recover its lithium content.
7.    Cooling
The filtered melt is then cooled further,  to  about 850° F,  in a heat exchanger.
The cooled melt  discharge  stream from the  heat exchanger is divided; about
90% is recycled to the quench tank to cool melt from the reducer, and the
remainder goes to the regenerator.
                                  lino

-------
FEED MELT
FROM SCRUBBER
      850°F
C02 TO
REGENERATOR
WASTE GAS
TO STACK
REDUCTION
              M2SO4 + 2C-+M2S + 2CO2
                              1400°F,
                      t
                                        OXIDATION
                       2O2-»M2SO4
                          600°F
                                        AIR AND
                                        COKE
                           REDUCED MELT TO
                           QUENCH TANK
                           AND FILTER
               Fig. 2.  Reducer Schematic
                         1109

-------
 The heat exchanger is a shell and tube design, with melt on the tube side and
 air or water on the shell side.  Provision must be made to prevent the melt
 from freezing,  even under upset  conditions.
 8.    Regeneration
 In the regenerator, the sulfide-containing melt is reacted with a gaseous
 mixture of carbon  dioxide (from the reducer) and steam.   The reaction is
      M2S(I)  +  C02(g)  +  H20(g)^^lM2C03(i)  +  H2S(g)              (7)

 The reaction is exothermic,  and  reaches equilibrium rapidly, with the
 formation of hydrogen sulfide being favored by lower temperatures.  To
 attain a high degree of regeneration,  the reaction is  carried out in a multi-
 stage countercurrent gas-liquid reactor, such as a tray column.
 9.    Sulfur Recovery
 The hydrogen sulfide from the regenerator is sent to a Glaus unit, where it
 is converted to  elemental  sulfur.  A conventional Glaus unit is suitable;
 however, the  feed  gas stream has a high water content, so it is sent to a
 condenser-cooler before undergoing the  Glaus reactions.  The tail gas
 from the Glaus plant is incinerated  and ducted to the  scrubber inlet and
 scrubbed along with the flue gas,  so there is no Glaus plant air pollution.
 1 0.   Lithium Recovery
 The melt used in the process is relatively inexpensive  except for lithium
 carbonate, which it is desirable to recover from the  process filter cake.
 An aqueous process has been developed for this purpose.  The filter cake
 is slurried with water and filtered,  to extract the very soluble sodium and
potassium carbonates; lithium carbonate remains with  the ash since it is
 relative insoluble (1 wt. %) under  these conditions.  The ash-lithium carbonate
 cake is then re-slurried in water and the lithium is solubilized by treating  it
with carbon dioxide to convert it to  the bicarbonate.  The ash is removed by
filtration and the soluble lithium bicarbonate in the filtrate is precipitated
as the insoluble carbonate.  The lithium carbonate is separated by filtration
and returned to the process stream; the saturated lithium carbonate filtrate
 is recycled to conserve lithium.  Laboratory tests have demonstrated that
over 90% of the lithium can be  recovered from fly ash filter cake by this technique.
                                   1110

-------
 C.   PROCESS MATERIALS AND EQUIPMENT

 1.    Materials of Construction
 Extensive testing has been done to determine the rate of corrosion of steels
 and other materials in the different process melts.  The results of many tests,
 including one-year dynamic tests in rotating capsules under various  conditions,
 have shown that the 300-series austenitic stainless -steels are suitable up to
 temperatures of about lOOOT.   Type 347 stainless steel is most corrosion-
 resistant, followed in order by 321, 304, and 316.  Type 304 was chosen for
 the pilot plant,  because of its availability and its well-known welding and
 fabricating technology.
 The corrosion resistance of the 300-series stainless steels in the carbonate
 melts is due to a protective LiCrO_ film which forms a compact, tenacious,
 and self-healing layer.  This film forms in about 500 hours, and decreases
 the corrosion rate to a few mils per year below 1000'F.  However,  at
 temperatures of 1500 to 1600°F (corresponding to the reducer internals) the
 protective layer breaks down and steels are rapidly and severely attacked.
 At these temperatures, the only corrosion-resistant material available is
 dense,  high-purity alumina.
 Corrosion testing is continuing in a forced-circulation  test loop which is in
 operation  at Atomics International.
 2.    Pumps^
 Three pumps are required to circulate the process melt in the  flow  scheme
 of Figure  1.  These pumps are vertical cantilever-shaft centrifugal  pumps,
 designed to operate with no seals or bearings in  contact with the melt.  The
 design features of this type of pump are shown in Figure 3.  One of these
 pumps has been operating satisfactorily for over 3200 hours in  the test loop
 mentioned above.
 3.    Valves
 Conventional stainless steel valves can be used,  employing a corrosion-
 resistant stem packing material such as Union Carbide's "Grafoil".   Both
hand and pneumatically actuated valves have been tested successfully in the
test loop.
                                   1111

-------
SUCTION
              o
              o
                                JISCHARGE
       Fig. 3.   Molten Salt Pump
                  1112

-------
4.     Instrumentation
Process instrumentation is needed for temperature,  pressure, flow and
liquid level measurement and control.  Temperature measurement is done
with stainless steel or inconel-sheathed thermocouples, protected with
alumina thermowells in the reducer.  Pressure and flow measurement
is  done with strain gage or variable reluctance transformer sensors,
designed for use at high temperatures.   Liquid level measurement is done
with conventional float displacement instruments designed for high temperature
operation.  These components are currently being tested in the test loop.
5.     Trace Heating
The process equipment and piping must be heated up to about 800°F for
startup, and then kept at this temperature during operation.  This heating
is  done electrically, with resistance heaters fastened to the vessel walls and
pipe walls, under  the thermal insulation.  SCR-type  heater controllers are
used to maintain the proper temperatures.
                                   1113

-------
                      III.  PILOT PLANT PROGRAM

The process described above will be tested extensively in a pilot plant being
built at the Arthur Kill Power Station of the Consolidated Edison Company,
on Staten Island, New York.  The pilot plant program has the following goals:
      1)  Demonstrate the integrated behavior of the process in
          a real environment;
      2)  Provide a test facility to check out and modify equipment
          or internals of equipment in order to evaluate design options
          or improve operation;
      3)  Obtain equipment design criteria, performance, and reaction
          yield data for scale-up;
      4)  Establish maintenance,  reliability,  and availability data; and
      5)  Provide data for an economic assessment of the process.
The pilot plant will process a side stream of stack gases from a 335 Mw
boiler at the Arthur Kill Station.  This station will be burning 0. 3% S oil
during all of the test program.  It  is planned to vary the SO2 concentration
at the inlet to the  scrubber in order to map process performance over  a
wide range of conditions.

A.    PILOT PLANT DESIGN
The overall size of the pilot plant was fixed as small as possible, com-
mensurate with the primary consideration that each component be large
enough to yield meaningful data for scale-up to full-scale systems.   The
limiting component is the regenerator column,  where an 18-in.  ID tower
was determined to be the minimum size.  To operate a column of this
diameter  at its optimum capacity will require a throughput of 100 Ib/hr of sulfur

-------
as sulfide in the melt.  This requirement then set the throughput for the reducer,
and its size.  The sulfur throughput required for  optimum operation will be
obtained from the flue gas and partially (up to 85%) from recycling sulfur
dioxide from the  Glaus plant incinerator.
1.    Flow Diagram
The flow diagram for the pilot plant is shown as Figure 4.  This flow diagram
differs from that of Figure 1 only in the following features;
      1)   The flue gas temperature is raised to 850°F by reheat;
      2)   Sulfur  dioxide will be intentionally recycled to the  scrubber
          from the Claus plant; and
      3)   No provision is made for on-site lithium recovery from the
          filter cake.
2.    Flue Gas Supply
The pilot plant treats  16, 700 scfm of flue gas from the oil-fired boiler,
after  the gases have left the boiler and passed through the electrostatic
precipitator.  The gases are nominally at 250°F; they are reheated to
about 850°F with an inline burner firing the same No.  6 fuel oil as the
boiler.   The total quantity of gas from the boiler is increased by 3400  scfm
from the>burner, to a total flow of 20, 100 scfm (equivalent to about 10 Mw).
A blower upstream of the burner provides the pressure head to force the
gas through the ducts, burner, and scrubber.  A damper is used to adjust
the flow rate to the test requirements.
The flue gases produced by the boiler will contain '—200 ppm sulfur oxides.
This concentration is very low compared to the 2000 ppm typical of gases
produced by boilers burning fuel containing  3% sulfur.  In order to provide
the flexibility needed to operate over  a range of sulfur oxide concentrations,
the pilot plant is designed to recycle sulfur  dioxide from its  Claus plant
incinerator.

-------
             CLEAN
             GAS TO
             ATM.
               t
                           SULFUR
FLUE
GAS
    BLOWER
                 SCRUBBER
                                      N.
NO. 6OIL-*
           cO
          M2CO3 + M2SO3
          A
       BURNER
M2SO3
                    PUMP
FILTER
                             REDUCER

                             M2CO3
                                              M2S
                   PUMP
                   TANK
                               CAKE
                  COKE
                  HOPPER
                                 QUENCH
                                 TANK
                                             PUMP
                                             TANK
                           AIR '•  ' '   CHEATER

                          COMPRESSOR  *™ NO> 6 OIL
                        Fig. 4.  Pilot Plant Flow Diagram

-------
The flue gases produced by burning the fuel oil contain about 0. 05 grains/scf
of particulates after passing  through the electrostatic precipitator.   In high
temperature electrostatic precipitators, such as would be used in a full-
scale molten carbonate process installation, the particulate removal
efficiency will be much higher.  The quantity of ash to be handled in the
pilot plant,  therefore, is conservatively high.  It is possible that the
particulates can be burned in the reducer, thereby allowing future simplification
of the process by elimination of one filtration step.
3.    Component Design
a.    Gas Preparation System
The side stream of flue gas is  removed from the power plant ducting through
a 4-ft duct, and passed through a blower fan and direct-fired reheater.   A
recycle line also brings SC^-rich gas from the Glaus plant to the blower inlet.
The blower fan is of conventional design, with a 200 HP, 1800 RPM motor.
The reheater  is a refractory-lined unit which burns the same No.  6 fuel oil
as the power plant.  The burner discharges its 850°F gas stream directly into
the scrubber.
b.    Scrubber System
The pilot plant scrubber is shown in Figure 5.  It is a simple  spray chamber,
10 ft. in diameter and 25 ft.  tall.  The gas enters tangentially near the
bottom,  and is discharged from the top after passing through a 1 ft. thick
wire mesh mist eliminator.  The gas velocity is nominally 11 ft. /sec.
The melt is sprayed in through 3 tiers  of spray nozzles, and drains out
through the bottom to the scrubber pump tank.  The scrubber pump has a
capacity of 33 gpm, allowing a recycle of up to 31 gpm of melt to the scrubber
inlet.  Flow control is provided by pump drive speed control.
The scrubber design is flexible, allowing variation in the melt and gas flow
rates, the  number, type and configuration of the spray nozzles, and the
type and thickness  of the mist  eliminator.
                                   1117

-------
                        GAS
                        OUTLET
SPRAY NOZZLE
MANIFOLDS
INLETS
                                                 MIST
                                                 ELIMINATOR
                                                 VIEWPORT
                                                    GAS
                                                    INLET
                                    MELT
                                    OUTLET

                Fig. 5.  Pilot Plant Scrubber
                            1118

-------
c.    Ash Filters
Continuous filters with dry cake discharge are desired for the process; how-
ever, development work will be required to make them available.  For the
pilot plant, batch-type cartridge filters using wire-wound filter elements were
selected. Two filters are provided; one is in operation while the second is
being cleaned.  A filter is shown in  Figure 6.  Each filter has an active area
of 160 sq. ft. ,  provided by 25,  2-3/4 in. diameter by  9-ft. long  filter tubes
in each 2 ft.  diameter shell.
d.    Reduction Systems
The pilot plant reducer is shown in Figure 7.  It is a steel vessel 9 ft.  in
diameter and 16 ft. tall,  lined with blocks  of high density alumina.  An inner
concentric wall of alumina blocks divides the vessel into an inner cylindrical
oxidation region and an outer annular reduction region.  Holes in the inner
wall allow melt circulation between the two regions.  The melt is introduced
into the reduction region  from the  top, and leaves the  same region from the
bottom through an underflow-overflow weir.  Air and petroleum  coke are
introduced into the bottom of the oxidation  region, and the spent  gas from this
region is collected at the top and carried back to the scrubber inlet, where
it is scrubbed before entering the atmosphere. The carbon-dioxide-rich gas from
the reduction region is collected at the top in a separate manifold, cooled to
850°F by liquid water injection, and then ducted to the regenerator.
An air compressor provides the reducer air supply.  The air is  preheated
to 600°F in a gas-fired indirect heater, and then picks up coke particles
introduced through a star feeder.
e.    Quench Tank
The pilot plant quench tank is a horizontal  cylindrical  tank, 6 ft.  6 in. ID
by 15 ft. long.  It is of conventional stainless steel construction, and has
sufficient excess capacity to serve as a reducer drain tank during shutdown.
It is mounted in a pit directly below the reducer.
S.     Coke Filters
The coke filters are identical in design and operation  to the ash  filters.
                                    1119

-------
                                                      2 in. OUTLET
THERMOCOUPLE
PENETRATION
                                                             1/2 in. LOWER VENT
                      Fig.  6.  Pilot Plant Filter
                                    1120

-------
PREHEA
INLET
OVERFLOVy
WIER
 STEEL
 SHELL
                   SPENT GAS
                   OUTLET
                                         PREHEAT
                                         EXHAUST
                                         INNER
                                         WALL
 REFRACTORY
 LINING
                                         REDUCTION
                                         REGION
MELT
CIRCULATION
PORTS
                         AIR AND COKE
                         INLET
             MELT
             DISCHARGE
              Fig. 7. Pilot Plant Reducer
                         1121

-------
     g.  Melt Cooler,  The melt cooler is shown in Fig. 8.  It is a 750,000
 Btu/hr, air-blast type heat exchanger, with recirculation of the air stream
 to provide temperature control.

    h.  Re g e ne rat or.  The regenerator column is  shown in Fig. 9.   It is
 a sieve-tray column,  18 in. in diameter by 36 ft. tall.   It  contains 15 trays
 spaced 2 ft.  apart.  Provision has been made to vary the number and
 spacing of the trays during the test program.

    i.  Glaus Plant.  Sulfur recovery from sour gases  by  the Glaus process
 represents commercial technology and does not require testing or demon-
 stration in the pilot plant. The Glaus unit, however, had  to be specially
 engineered as its capacity of 100 Ib.  of sulfur per hour is  smaller than any
 built  commercially.  It provides a convenient way of recovering the sulfur
 removed by the pilot plant rather than burning the H S  in the regenerator
 off-gas to SO and  returning it to the power plant stack.  The Glaus plant
 is shown in Fig. 10.

 4.  Component Arrangement

 The pilot plant component arrangement plan is  shown in Fig. 11.  The pilot
 plant  is built on a  68-ft.  by 100-ft. concrete slab,  and has a high-bay filter
 room and reducer enclosure and a control room and laboratory.  The quench
 tank and make-up tank are mounted in a pit, and serve as drain tanks
 during shut-down.  There is also a large, 40-ft. tall coke hopper to hold
 the reducer coke supply.

 B.  PILOT PLANT STATUS REPORT

As  of November 1,  1972, the pilot plant program was well underway.  The
 design and engineering work was completed, and all major components had
been fabricated, delivered, and set in place at the site.  The steelwork

-------
Fig, 8.  Melt Cooler
        1123

-------
— ^
MELT INLET j[
THERMOCOUPLE NOZZLE "^
4
GAS IN LET
NOZZLE _-f
rr^'
—t
(
L
y.1


^j
V
f
to-
p1
I
O

® © © @ @ 0 7^ ^ I J e Vj> S i- w— » *» UUII.CI 1 1t THERMAL "~ INSULATION 6 in. THICK 10 \ 1( ft ft 28ft {141 )ft MELT OUTLET 3ft _ 1- FRAYS) "\^ -*- 1 ft 6 in. Fig. 9. Regenerator 1124


-------
NJ
in
        SULFUR
        CONDENSERS
                                                  WATER
                                                  CONDENSER
                                                      HEAT EXCHANGER
    SNUFFING
    STEAM
SULFUR
TANK

                                                           ACID
                                                           GAS
                                                           BLOWER
                                                   AIR BLOWER
   C
<£%^
                                              SALT BATH
                                              HEATER
                                         REACTION
                                         FURNACE
                               Fig. 10.  Glaus Plant

-------
       EXISTING
 ELECTRICAL SUBSTATION
                  V
       o
   n
   8ft-
NEW 4 ft x 4 ft DUCT-

     	60ft-
 J
TRUCK DOOR
   )
  GRATING

I
\
v^
t—fc
c



LT^
NK



^
\s
urriuc
AREA
LAB
/
HIGH
BAY
r
1
^L



L_U
v ;
PUMP
10^
o£
ol:
o"-
u
u
^HMH^^W

F
\ f

©
SCRUBBEI
	 ~*T
tEHEATER
• j
-, U
REGENE
BLOWER
1
i

CLAUS
PLANT
ORATOR
QUENCH TANK
PIT /
~» D(

_ /~*
Df
. m

Vp- REDUCER
1001 ^"" 1
\
\
COKE
HOPPER
TRUCK /^v
DOOR f J





68ft



0

O BLOWI
D



1


=R



S>- CONCRETE PAD 20 ft
 EXISTING
FLUE DUCT
."NEW SUPPLY
 DUCT TIE-IN

      20ft
                                         ASPHALT
                                     100ft
                                     CONCRETE PAD
                            SITE FENCE -
                                                     25ft
                                                     ASPHALT
              Fig. 11.   Pilot Plant Equipment Arrangement Plan
                                     1126

-------
for the platforms, supports, and building was being installed, and the refractory
lining of the reducer vessel was about to start.  Electrical heaters were also
being mounted on tanks and vessels.  All of the major subcontracts had been let,
and construction was scheduled to be completed by February 1,  1972.
Photographs showing the construction work in progress are presented in
Figures  12 to 15 inclusive.
C.    TEST PLAN
The test plan covers  a one-year test program.  The schedule for this program
is shown in Figure 16. The test period is divided up into four phases:  1) Start-
up  and Stabilization,   2) Design Verification,  3) System and Component Parametric
Studies,  and  4) Reliability and Operability Studies.
The startup and stabilization phase will require about 4. 5 months  to complete
and will consist of a thermal test,  a hydraulic simulation, fill and circulation
of the molten carbonate, and activation and stabilization of the process
chemical reactions.
The second phase of the test program will be the design verification, which
will consist of recycling SO£ from the Glaus plant to the scrubber and operating
at the design  sulfur throughput of the pilot plant.  About two months is  scheduled
for this  phase, the results  of which will be used to  verify the design point of
the pilot plant and establish system and component  performance characteristics
and operating cost data at the  design conditions.
The third phase will involve system and component parametric  studies.  It is
scheduled for 3. 5 months.   This will consist of general parameter studies of
the performance of the system as a whole and of each major component.  These
studies will establish performance characteristics  over a range of operating
conditions, maximize and establish  capabilities of  the system and individual
components,  and establish design and operating criteria for a large demonstration
plant.
The fourth and last phase of the currently planned test program will be reliability
and operability studies.  Reliability and operability of the pilot  plant will, of
course,  be under study throughout the one-year test program.  This phase will
be  a special two-month test at the end of the year to determine minimum
operator supervision and maintenance requirements.
                                    1127

-------

Fig. 12.   Pilot Plant Foundation Under Construction

-------
(SJ
10
                                              Fig.  13.   Glaus Plant

-------
co
o
                               Fig. 14.  Pilot Plant Scrubber Being Installed

-------
Fig. 15.   Pilot Plant Equipment Being Installed

-------
CO
NJ
      I   1   I    2    I   3    I   4   1
    Months


6   j    7   i  8     [    9
10
11
12

STARTUP & STABILIZATION
DESIGN
VERIFICATION
, SYSTEM & COMPONENT
PARAMETER STUDIES

RELIABILITY/
OPERABILITY
STUDIES
THERMAL
TEST
3 WEEKS
HYDRAULIC
SIMULATION
4 WEEKS
C03 FILL fc
CIRCULATION
4 WEEKS
PROCESS ACTIVATION
AND STABILIZATION
8 WEEKS
                              Fig. 16.  MCP Pilot Plant Test Plan and Schedule

-------
   SUMMARY OF FLUE GAS
DESULFURIZATION SYMPOSIUM
           1133

-------
                         SYMPOSIUM SUMMARY

     E. L. Plyer  (EPA)—.The last part of our symposium is the
summary.  As you can imagine this has to be a very awesome and formidable
job and I think we have a very capable person in Frank Princiotta who
has agreed to do this.  I would like to make a few remarks concerning
this.  Frank, as you might realize, has had to put this-together sort of
on the run here.  So even in summarizing something like a symposium of
this type, there's bound to be some subjectivity that comes in.  And
I'm sure that there will be a little bit of it here.  You might not
agree with everything Frank is going to  say  here. And what Frank has to
say I don't think you can take as being EPA policy.  But we do think
that it would be quite helpful to you to get a summary of the
highlights of the symposium.  Necessarily, he will not be able to
cover, even briefly, everything that has been given here.  And he will have
to concentrate his remarks, I think, on the large-scale demonstrations and
some of the commercial units which we have tried to concentrate on in
our symposium.  Since Frank has already been introduced as the session
chairman for the lime/limestone wet scrubbing processes, I don't think
I need to repeat any biographical information on him.  I will just
mention that he is the Chief of the Engineering Test Section of the
Control Systems Laboratory.  He's had about 2 years experience there
and due to his enthusiasm and drive he has become, I think, one of
our real experts in flue gas cleaning.  I think he will do a good
job in summarizing our symposium for us.
                              1134

-------
     Frank Princiotta (EPA)—Thanks,  Bill,  for the introduction
and disclaimer.  I can't get in trouble now.   As Bill  indicated,
keeping in mind the difficulty of summarizing a symposium of the
comprehensiveness that took 4 days in  half an hour. I'm bound to
miss some important highlights, so please bear with me.  First of
all, we've heard that sulfur oxides are among the most dangerous
air pollutants to human health.  Dr. Newill  described  how morbidity
and mortality can be correlated to ambient SO^ levels.  He
also pointed out some relatively new information that  can have
important repercussions.  Namely, that particulate sulfates appear
to be even more dangerous to human health than S02 alone.  He stated
that in cities where the primary ambient levels of S02 are met, high
daily sulfate levels have been associated with aggravation of symptoms
of heart and lung disease patients.  Although further  data are necessary,
the sulfate particulate problem could lead to more stringent ambient
SOX levels and therefore SOX emission control.  Time will tell.
     I think Dick Harrington (EPA) had some relevant
ooints about the importance of controlling industrial   boilers.
He indicated that despite the fact that area sources and industrial
boilers represent a relatively small percentage of total SOY emitted,
                                                           /\
they contribute due to meteorological  factors.  He stressed the
importance of trying to utilize stack gas technology developed for
power plants on these smaller sources.  It certainly seems like a
                              1135

-------
 reasonable suggestion and we're hopeful that systems like double
 alkali, lime scrubbing, and others could fulfill this role.
     Steve Gage  (Council on Environmental Quality) presented
 a very interesting paper on the alternatives to flue gas
 cleaning—the technological alternatives.  And really, in summary,
 there are none in the short term.  He reviewed coal liquefaction,
 gasification, and advanced combustion processes.  He indicated
 that these are in an embryonic stage with years of development
 necessary.  Probably they will make an impact somewhere between
 the 1980 and 1983 time period.  All of these processes, though,
are exoensive and, as most of you know, as you approach true
hardware, they tend to get more expensive.
     Therefore, flue gas desulfurization is the only alternative to
 flue switching now and will remain competitive for many years in the
future, particularly on retrofit applications and in the important
 industrial boiler size applications.
     We then heard from Dr. Ando (J. Ando, Chuo University) who
presented the Japanese situation and I think it's of great relevance
to the U.S.  situation.  They have a severe air pollution problem
aggravated by a high population density and concentrated industrialization,
The great majority of their power plants are oil-fired and, therefore,
there are differences that have to be remembered.  In the past, S0y
                                                                  A
regulations  in Japan were met primarily by hydrodesulfurization of

                                1136

-------
fuel oil.  They import high sulfur oil  and  then  desulfurize  it.
However, the recent tighter regulations,  equivalent  to  less  than
1 percent sulfur flue gas desulfurization,  have  led  to  plans for
at least 100 air pollution control units  on power plants, Glaus
plants, sulfuric acid gas treatment,  industrial  boilers, etc.
Most of these units produce sulfur products and  are  either  presently
installed or will be installed in the near  future.   Sulfur  products
produced by these systems in Japan include  sodium sulfite,  sulfuric
acid, sulfur, and (most importantly in recent months)  gypsum.   However,
Dr. Ando points out that by 1976 he expects that significantly more
gypsum will be produced than can be used; so they, too, may start worrying
about throwaway processes.  The most important Japanese systems appear
to be:  the Mitsui lime scrubbing unit - I  will  mention it  briefly
later; the Mitsubishi Heavy Industry-JECCO lime  scrubbing  process, which
this unit has operated with high reliability for 9 months  on a 35-Mw, oil-
fired, closed-loop unit; and the Mitsubishi Chemical Machinery - Wellman
Lord process, which has operated reliably for 1.5 years on a 75-Mw,
oil-fired power plant.  Also, there are three double-alkali systems
under active development in Japan:  the Kureha/Kawasaki and Showa Denko
sodium double-alkali systems, and the Nippon Kokan ammonium double-
alkali system.  These are considered to be very important and should
be followed closely.
                                1137

-------
    Moving on to costs, I think Gary Rochelle (EPA) presented a very
useful scheme for the rough estimation of capital and operating costs
for flue gas desulfurization systems.  Briefly, the following are
some of the operating costs he came up with.  (I will just talk about
total annualized operating costs that are perhaps more important be-
cause they include capital charge considerations.)  For lime/limestone
and magnesium oxide processes, he found costs for a new 500-Mw, 3.5-
percent sulfur coal unit to be very close for both processes at about
2.5 mills/kw-hr.  Costs for the Wellman Lord system and the Stone and
Webster/Ionics System under the same groundrules were about 2.7 mills/
kw-hr.  Operating cost for the Cat-Ox system was about 2.75 mills/kw-
hr.  He indicated that about 75 percent of existing plants can be retro-
fitted within 3 mills/kw-hr.  Comparisons of this estimating technique
with actual costs for six full scale installations, most of them first
of a kind, indicate that he's about 9-21  percent low in his capital cost
estimates,  ffy own feeling is that this is due primarily to the fact
they are first of a kind and that these estimates will come even more in
line as there's more widespread application of these systems.  However,
there were discrepancies,   You heard them and I heard them.  The Will
County unit and the Widow's Creek unit in particular have had recent
cost escalations which make Gary's estimates look even a little bit
farther off.
                                  1138

-------
    Most of these systems,  reqenerable and throwaway,  look  pretty
similar to Gary as far as operating costs are concerned  and he  feels
the key parameter which can turn the tide is sludge disposal  costs.
And he feels these operating costs are equal at $3 or  $4 per ton of
wet sludge.  Additionally,  interestingly, he also indicates what's
obvious I guess when one thinks about it: the low load factor systems
strongly favor the throwaway processes economically.  And this  has
obvious implications for industrial boilers which ordinarily operate
at low load factors.
     Getting into lime/limestone scrubbing systems, I  think Irv Raben
(Bechtel Corporation) presented an excellent paper and I recommend that
everybody review it carefully.  He pointed out that 21 full—scale units,
now comprising 9600 Mw of capability, will be or have been installed  in
this country.  In particular the Ohio Edison Bruce Mansfield, the Mohave,
Navajo, and Northern States units are the real biggies that contribute
most to that 9600-Mw figure.  His costs were in line with Gary's costs,
and he estimates for a hew 500-Mw unit on a similar basis for Gary,
about 2.3 to 2.5 mills/kw-hr.   Irv feels relatively confident, hopeful
I guess 1s the word, that reliability will  be demonstrated soon in the
United States based on all the  units that are going to come on to line
relatively soon.
                                 1139

-------
      Talking about some limestone systems in particular, the Chemico
 lime-Mitsui aluminum plant has to be mentioned.  Of coarse this
 paper was presented by Mr. Sakanishi (Mitsui Aluminum Co.).  Bob Quig
 (Chemico) made perhaps a controversial introductory statement.
 when  he  expressed his belief that the operation to date has indicated
 this  technology is now demonstrated by anyone's definition, including
 the National Academy of Sciences.  As you know this system has removed
 90 percent of the S02 on a coal-fired plant for 14 months with reliable
 operation; no problems of any significance were reported.  Briefly
mentioning, though, there were people on our panel, assembled to discuss
 the relevance of this system to the U.S., who felt that they have to be
 careful  in trying to extrapolate this experience to the U.S., primarily
 because  the system apparently operated open-loop for at least part of
 its time in Japan.  It's a base-loaded unit with few excursions, and it's also
 a 2 percent sulfur coal  system, which is somewhat lower than many eastern
 U.S. applications.  Other panel members,  however,  felt that it is a very
 relevant system and has many characteristics in common with many potential
eastern U. S.  coal applications.
     Moving on to Dr. Weir's (A. Weir,  Jr,, Southern California Edison)
paper on the Mohave power plant results.   This ties in to full-scale units.
These results, like those at Shawnee and other places, indicate that the
lower the S02  inlet in the flue gas, the easier things are as far as
                                  1140

-------
removal is concerned.  He was able to get down to 20 ppm  and  less  in
outlet for many of the eight scrubbers he looked  at under many
conditions.  He was also able to get 0.001  to 0.002 gr/scf  outlet
for several of his scrubbers (keep in mind that this is downstream
from a relatively efficient precipitator).  As you know this information
(and perhaps data from two big modules which will be built
at Mohave) will influence the selection of the very large Mohave and
Navajo plants.  They add up, by the way, to over 3000 Mw of capability.
So that's a real slug of capability.
     Talking a little about the B&W systems, the Will County unit  was
described to us.  It initially was a 175-Mw unit designed with a limestone
tail end system for 80 percent removal.  It started up about a year ago and
has had many operating problems, primarily minor problems with apparently
no inherent system-type problems.  Perhaps the most serious of the
problems has been demister pluggage.  However, there
appears to be relatively high probability in the near term  that thts
system will be made to operate reliably.,  The next B&W unit we heard
about was the Kansas City P&L La Cygne unit, which is kind of a version of
the Commonwealth unit multiplied by about 7.  There are seven
parallel circuits in this 840-Mw system, which just recently started up.
I think the boilers are now at about half load and I'm personally going
to keep an eye out for that system.  I suspect the unit could be
very reliable since  I think many of  the problems  they had  at Will
County have been corrected on that particular unit.
                                1141

-------
     I think we should mention some of the Combustion Engineering
units as well.  But I think we should mention the Kansas City Power &
Light units briefly.  They have been changed to a tail end
configuration; one of the two Hawthorne units (Unit No. 3) has been
changed from boiler injection to a tail end limestone system.  I think
Irv Raben indicated that he felt this was a trend in the industry; I
think this change is indicative of this trend.  There has not been
sufficient operating time on this tail end limestone Hawthorne unit
to get much information yet.  Our understanding is that they've been
burning some low-sulfur coal lately.  The Louisville Gas and Electric
Paddy's Run Station, a 70-Mw carbide sludge tail end system, though,
has started up.  It has about 1 month's operation now.  And, as you
heard, things look pretty good.  Reliability so far looks promising.
502 removal nas been reported as very high, although I haven't heard
a number yet.
     Let's move on briefly now to the EPA Shawnee lime/limestone program.
We heard from Dr. Epstein (M. Epstein, Bechtel Corporation) about all
the SOg data that's been generated as a function of system parameters;
this should certainly help design these systems in the future.  There's
information also on particulate removal; its range is from about 0.01 to
0.03 gr/scf for all three scrubbers.  You might remember these are three
10-Mw scrubbers in parallel  to test three different types of scrubbers.

-------
I'll summarize reliability information and perhaps  even add  a
bit to what Bill Elder (TVA, Muscle Shoals) said.   This
is hot off the press.  In general, reliability tests  started about
2 months ago.  And we are very pleased so far with  the reliability
seen to date.  We have had little evidence of scaling or demister
problems.  We feel the trick is low pH operation,  high L/G's,  and
backing off somewhat on gas velocity.  The venturi  after-spray scrubber
has operated for 1 month at 75 percent removal without any
problems of any type.  The TCA has also operated for  about a month with
from 85 to 90 percent removal; no demister or scaling problems occurred,
although we have seen some erosion of balls and grids and some
solids buildup in the inlet duct.  We feel confident  these problems
will be resolved shortly.  The hydrofilter had very few problems  over
about a month's tests until lately when the nozzles apparently have
eroded and led to some bed and demister problems.
     Moving on now to the first of the regenerate systems or  saleable
product systems.  We heard quite a bit about the magnesia oxide process.
Gerry McGlamery (TVA) described the system in detail.  He
compared the costs to those of wet limestone systems  and found them relatively
close although the magnesia oxide process was perhaps somewhat more
expensive; again, sludge costs were very important for any cost
comparison.  He also indicated that magnesium oxide systems  mak-
ing sulfuric acid makes most sense in metropolitan areas where
sludge is difficult to handle and sulfuric acid can be marketed.

-------
     Of course the most important Mag-Ox system is the Boston
Edison system.  This unit started up in April 1972 at Boston
Edison's Mystic Station, a 150-Mw facility producing 50 tons per day of
sulfuric acid, which is processed at a separate location in Rumford,
Rhode Island.  Operations to date have demonstrated each process step.
Greater than 90 percent S0« removal has been achieved.  But, as you
heard, there have been several "nitty gritty" mechanical problems
which have prohibited a long-term test which is really needed to check
out the system.  Such problems have included dryer problems, calciner
seal leakage, calcined material loss of reactivity, solids handling
problems, and so forth.  And Chemico has indicated that top priority
is going to be given to performance of some long-term testing relatively soon
to really home in on the reliability of this important system.
     Other Mag-Ox systems were mentioned.  The Philadelphia Electric
Eddystone system designed by United Engineers is a 120-Mw system that
will start up in the fall of 1973.  And the PEPCO Dickerson No. 3
coal-fired station, a 100-Mw station, will  start up in several months.
     Perhaps the most reliable of the regenerable systems,
in terms of operation to date, is the Wellman Lord process of Davy Powergas.
This process, at least based on what we heard today, has demonstrated the
greatest reliability, I think it's safe to say, of any flue gas desulfuri-
zation system to date.  There are five units presently operating reliably;

-------
eight are under construction on power plants,  sulfuric acid treatment
systems, and Claus treatment systems.  Perhaps the most significant
is Japan Synthetic Rubber's Chiba plant, the 75-Mw, oil-fired system
that has operated for over 1.5 years at a reported availability of 95  percent.
Sulfur dioxide removal has been at about 90 percent.   EPA is quite
interested in the process and has co-funded with Northern Indiana
Public Service Co. for installation of the Wellman unit at the NIPSCO
station, a 115-Mw system.  It's a coal-fired boiler that will produce
sulfur directly using the Allied SOg reduction process which uses  nat-
ural gas as a reductant.  This sulfur-production technology was success-
fully demonstrated at the Falconbridge Nickel  Plant near Sudbury,  Canada.
It should be noted, though, that for about a 1000-Mw system you need
about 20 Mw worth of natural gas.  Of course this is a problem in  many
locations; however, Allied indicated they are working very intensely for
substitutes for natural gas as the reductant.   Another problem attendant
to that system is the requirement for a sulfate purge based on a 4,5,
or 6 percent oxidation of sulfute/bisulfate to sulfate.
     The Monsanto Cat-Ox process should be mentioned.  This is another
regenerate process somewhat unique in that it's based on the well-known
contact sulfuric acid process.  Flue gas is heated to 800°F and oxidized
in the catalyst bed to 850°F where the sulfur dioxide is oxidized to sul-
fur trioxide to produce about 78 percent sulfuric acid.  You heard the
story about the Wood River station of Illinois Power.  This  is where
the system has been installed; the system started up back in September
of 1972.  I should add that  a 99.6 percent efficient ESP, which is
                                   1145

-------
necessary upstream of the system, was installed in February
of 1972 and has been operating well since then.  Initially the system
worked well when it was using natural gas as the reheating combustion
medium; however, after about 17 days of successful operation, reheater
problems with oil associated with soot buildup (with potential bed
pluggage problems) shut down the system.  As you heard, there will be
modifications and it's expected the unit will be back on line in late
summer of this year.  One of the problems with the system is the sale
of dilute sulfuric acid.
     Other processes that appear to be very well  along in their develop-
ment and very important are the double alkali processes. I was pleased to
hear the summary by Norm Kaplan and Dean Draemel  (EPA) of all the varia-
tions of this process, and there are quite a few of them.  And I think
it's wort,h everyone's while, who's interested in these systems, to care-
fully look at a summary of these systems and see just what these differences
are — dilute versus concentrated, various ways of handling the sodium sul-
fate, etc.   But they all seemed to have the potential for low capital and
operating costs and for high reliability.  Gary Rochelle indicated that
the costs were approximately 15-20 percent lower than an equivalent lime/
limestone system.   This was based on an assumption which will be validated.
                                  1146

-------
     Norm Kaplan mentioned some of the corporations involved ~ FMC, Envirotec,
General Motors, A. D. Little, Showa Denko, and Kureha.   Let me just
mention some of the more significant full-scale units that are planned.
The 20-Mw unit at Southern Services  Scholz station of Gulf
Power is of real significance; this will be an A. D. Little/Combustion
Equipment Associates unit.  General Motors' 40-Mw double alkali system
for four stoker boilers at the Cleveland Chevrolet plant will
start up late this year; this will be the first of the U.S. full-scale
installations.  In addition, there are full-scale units planned in
Japan, one of which will start up in June,  the Showa Denko 110-Mw,
system on an  oil-fired  boiler.  Two  additional  150-Mw  units will  begin opera-
tion in 1974.  There is quite a bit  of activity  in  the double alkali
area and it looks like  it will be a  real viable  alternative to lime/
limestone scrubbing as  far as the throwaway processes  are concerned.
     Let me just briefly mention other important processes
which were mentioned today.  I'm sure they are probably pretty fresh  in
your mind.  The Foster Wheeler-Bergbau Forschung process, TVA-EPA
ammonium bisulfate, and Stone & Webster/Ionics molten  carbonate processes
were described and could play an important role  several years
from now when these systems are able to be applied  to  full-scale commercial
units.
     Let me briefly summarize some of the things we heard about sulfur
product problems.  One  can easily write a  book  on  this, so  I'll try to
be brief.  The  large quantities of sulfur
                                1147

-------
inherent in flue gas must end up as a low-value sulfur product.
Generally this is either a throwaway sludge, which you can call  a
zero value, or negative value material as opposed to a low-value
sulfur or sulfuric acid type of material.  Quantities are somewhat
staggering.  Dick Stern and Julian Jones (EPA) indicated that a
100,000-Mw capacity for 20 years of sludge storage to
a depth of 10 feet would require 150 square miles of area for settled
sludge.  They also made some very provocative comparisons between the
affected areas associated with sludge based on predictions of lime/
limestone,utilization and areas associated with coal strip mining.
For sulfuric acid production, the same 100,000-Mw capacity will
produce acid at a rate equivalent to the present total U.S. production,
about 28 million tons per year.  Sulfur, as far as volume and quantities
are concerned, is certainly the most desirable end product.
     The EPA/Aerospace program will hopefully present a complete, com-
prehensive evaluation of the toxicity and water pollution problems
associated with throwaway sludges, and will also evaluate some sludge
treatment processes.  We'll look at lime/limestone and double alkali
sludges, and Dick Stern has asked all relevent parties for information.
     Also to be evaluated are the sludge fixation (treatment) processes.
We've heard quite a few of them mentioned; IUCS and Dravo perhaps being
the most important processes.  These fixation processes basically
                                 1148

-------
involve a recipe requiring the inclusion  of flyash.  The recipe
calls for flyash, calcium sulfite and sulfate sludge, additional
lime-based material, and perhaps an accelerator to make a cementitious
type material which might have attractive landfill application.  And
perhaps, at least in lUCS's opinion, a potential for sale as  aggregate
and cementitious materials.  Preliminarily, data presented  by IUCS  indicate
that these processes lead to decreased Teachability  and permeability  and
therefore reduce water pollution potential.  We hope to validate
such claims during our Aerospace program.
     Let me mention quickly a little bit about sulfuric acid.  This has
been the product of choice for the regenerable systems  so far in this
country.  But it appears that only selective applications will be possible
due to difficulty in selling large quantities of sulfuric acid.  Each
1000 Mw produces approximately 1 percent of the total  U.S.  production.
However, there seem to be obvious urban applications where such systems
make quite a bit of sense.
     Let me briefly discuss sulfur as the end product.   As
I mentioned, the storage volumes are certainly  attractive; there's no
question about that*  Sulfur  can also be  stored for eventual  use as opposed to
sludges, which must just  lie  there  with  no potential use except as
landfill.  However, some problems associated with the storage of sulfur
were mentioned in this symposium which were new at least to me.
                                1149

-------
They include potential flammabi.lity problems, potential
HgS evolution, and perhaps wind and water erosion and other related
problems.  Hopefully this will be evaluated because sulfur certainly
seems (at least preliminarily) to be a very important end product.
Also we heard some potential uses of sulfur.  They include sulfur in
asphalt materials, sulfur concretes, etc., although it should be pointed
out there's quite a bit of technical and marketing work ahead before
these use any great quantities of sulfur material.
     As far as the symposium conclusion is concerned, I would like to
be  a little bit subjective now.  I have tried to be objective during
this presentation although somebody may disagree with that.  Let me now
quote from the SOCTAP report.  You might be familiar with this.  This
is the Sulfur Oxide Control Technology Assessment Panel, a
governmental interagency group which reported to Mr. Ruckelshaus
(W.D. Ruckelshaus, former EPA Administrator).  Its charter was basically
to look into flue gas desulfurization and come up with a reasonable, objective
assessment.   This report is available, I might add from the Air Pollu-
tion Technical  Information Center, Research Triangle Park, N.C.  Let
me just read briefly the most relevant conclusions.  Now keep in mind,
these are the conclusions of the report that I feel accuarately reflect
the sum total  of what we have heard at this symposium:
     "We (the SOCTAP group) have examined the status of stack gas cleaning
     technology in the United States and Japan and have concluded that
     sulfur dioxide removal from stack gases is technologically feasible
     in commercial-sized installations.  We have concluded that technologi-
     cal  feasibility should not now be considered a decisive element in
     utilization of these systems and that a large fraction of the nation's
     coal-fired steam electric plants can ultimately be fitted with com-
     mercially available stack gas cleaning systems...  The reliability
     of currently available systems has been the subject of some question.
                                   1150

-------
     We concur that SOX control  systems must exhibit the high degree of
     reliability required by the utility industry.   We believe that the
     required reliability will  be achieved with early resolution of a
     number of applications engineering problems related to specific hard-
     ware components and system design parameters....   In view of the
     fact that a number of large scale plants scheduled for operation in
     the U.S. in the near future will provide additional 18 months opera-
     ting experience (or by 1974) should effectively remove engineering
     barriers to the application of stack gas cleaning to many facilities1.'

     The report emphasized the need for additional  R & D primarily in ad-

vanced processes to cut back costs and help produce less noxious solid waste

problems and for a solid waste sludge disposal evaluation program.  Thank you,
                                      1151

-------
                                TECHNICAL REPORT DATA
                         (Please read Instructions on the reverse before completing}
  REPORT NO.
EPA-650/2-73-038
                           2.
                                                      3. RECIPIENT'S ACCESS!ON>NO.
4. TITLE AND SUBTITLE
Proceedings, Flue Gas Desulfurization Symposium-
  1973
                                                     5. REPORT DATE
                                                      December 1973
                                                     6. PERFORMING ORGANIZATION CODE
 . AUTHOR(S)

Miscellaneous
                                                      8. PERFORMING ORGANIZATION REPORT NO.
                                                      10. PROGRAM ELEMENT NO.
                                                        1AB013
. PERFORMING ORGANIZATION NAME AND ADDRESS
Miscellaneous
                                                      11. CONTRACT/GRANT NO.
                                                       ROAP 21ACY-30
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
NERC-RTP, Control Systems Laboratory
Research Triangle Park, N. C.  27711
                                                     13. TYPE OF REPORT AND PERIOD COVERED
                                                       Proceedings      •	
                                                     14. SPONSORING AGENCY CODE
16. SUPPLEMENTARY NOTES
  . ABSTRACT
         Tne proceedings document the 30 presentations made during the symposium,
attended by 430 representatives of electric utilities , pollution control system supp-
liers , engineering and construction firms , state and local pollution control agencies ,
and EPA Headquarters, Regional Offices,  and NERCs. The objective was to present
the current status of flue gas desulfurization technology for full-scale power plants ,
the only near-term alternative to the use of low-sulfur fuels in meeting air quality
standards.  It emphasized lime/limestone scrubbing, magnesia scrubbing, sodium-
based scrubbing with thermal regeneration, and catalytic oxidation.  Disposal and
uses  of SOx control process by-products and the second-generation or advanced
SOx control processes were also discussed.  The symposium filled the need for up-
to-date information in support of federal, state, and local air pollution control
activities.
17.
                             KEY WORDS AND DOCUMENT ANALYSIS
                 DESCRIPTORS
                                          b.IDENTIFIERS/OPEN ENDED TERMS
                         e. COSATI Field/Group
Air Pollution
Flue Gases
Desulfurization
Scrubbers
Sludge Disposal
Economic Analysis
                    Marketing
                    Sulfur
                    Sulfur Oxides
                    Electric Power PlantsjSodium
                    Calcium Oxides
                    Limestone
18. DISTRIBUTION STATEMENT
 Unlimited
Air Pollution Control
Stationary Sources
Utility Boilers
       -Based Scrubbing
Catalytic Oxidation
Molten Carbonate
	orption
19. SECURITY CLASS (ThisReport}
    Unclassified
13B, 14A, 2IB
                                                                   21. NO. OF PAGES
                                                                        1161
                                          20. SECURITV CLASS (Thispage)
                                             Unclassified
                         22. PRICE
EPA Form 2220-1 (»-73)
                                        1153

-------
                                                        INSTRUCTIONS

     1.   REPORT NUMBER
         Insert the EPA report number as it appears on the cover of the publication.

     2.   LEAVE BLANK

     3.   RECIPIENTS ACCESSION NUMBER
         Reserved for use by each report recipient.

     4.   TITLE AND SUBTITLE
         Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, if used, in smauer
         type or otherwise subordinate it to main title, when a report is prepared in more than one volume, repeat the primary title, add volume
         number and include subtitle for the specific title.

     5.   REPORT DATE
         Each report shall carry a date indicating at least month and year. Indicate the basis on which it was selected (e.g., date of issue,  date of
         approvcl, date of preparation, etc.).

     6.   PERFORMING ORGANIZATION CODE
         Leave blank.

     7.   AUTHOR(S)
         Give name(s) in conventional  order (John R. Doe, J. Robert Doe, etc.).  List author's affiliation if it differs from the performing organi-
         zation.

     8.   PERFORMING ORGANIZATION REPORT NUMBER
         Insert if performing organization wishes to assign this number.

    9.   PERFORMING ORGANIZATION NAME AND ADDRESS
         Give name, street, city, state, and ZIP code.  List no more than two levels of an organizational hirearchy.

     10.  PROGRAM ELEMENT NUMBER
         Use the program element number under which the report was prepared. Subordinate numbers may be included in parentheses.

     11.  CONTRACT/GRANT NUMBER
         Insert contract or grant  number under which report was prepared.

     12.  SPONSORING AGENCY NAME AND ADDRESS
         Include ZIP code.

     13.  TYPE OF REPORT AND PERIOD COVERED
         Indicate interim final, etc., and if applicable, dates covered.

     14.  SPONSORING AGENCY CODE
         Leave blank.

     15.  SUPPLEMENTARY NOTES
         Enter information not included elsewhere but useful, such as: Prepared in cooperation with, Translation of, Presented at conference of,
         To be published in. Supersedes, Supplements, etc.

    16.  ABSTRACT
         Include a brief (200 words or leaf factual summary of the most significant information contained in the report. If the report contains a
         significant bibliography  or literature survey, mention it here.

    17.  KEY WORDS AND DOCUMENT ANALYSIS
         (a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
         concept of the research and are sufficiently specific and precise to be used as index entries for cataloging.

        (b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc.  Use open-
         ended terms written in descriptor form for those subjects for which no descriptor exists.

         (c) COSATI FIELD GROUP - Field and group assignments are to be taken from the 1965 COSATI Subject Category List.  Since the ma-
        jority of documents are  multidisciplinary in nature, the Primary Field/Group assignment(s) will be specific discipline, area of human
        endeavor, or type of physical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follow
         the primary posting(s).

    18.  DISTRIBUTION STATEMENT
         Denote reusability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
         the public, with address and price.

    19.8.20. SECURITY CLASSIFICATION
         DO NOT submit classified reports to the National Technical Information service.

    21.  NUMBER OF PAGES
        Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any.

    22. PRICE
        Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Form 2220-1 (9.73) (RBVim)                               115U       *U.S. G.P.O. ;  1974—747-791/340,  Region No. 4

-------