::!S::i:::ij:^
Si:!:-::::::::::::::::^
       U.S.     E  ^ -—^                                        N     AGENCY

-------

-------
              PROCEEDINGS


         SECOND INTERNATIONAL


            LIME/LIMESTONE


        WET-SCRUBBING SYMPOSIUM



               VOLUME II
          November 8-12, 1971
        Sheraton-Charles Hotel
        New Orleans, Louisiana
   ENVIRONMENTAL PROTECTION AGENCY
      Office of Administration
Research Triangle Park, North Carolina
              June 1972

-------
The APTD (Air Pollution Technical Data) series of reports is
issued by the Environmental Protection Agency to report tech-
nical data of interest to a limited number of readers.  Copies
of APTD reports are available free of charge to Federal employ-
ees, current contractors and grantees, and nonprofit organiza-
tions - as supplies permit - from the Air Pollution Technical
Information Center, Environmental Protection Agency, Research
Triangle Park, North Carolina 27711 or from the National Tech-
nical Information Service, 5285 Port Royal Road, Springfield,
Virginia 22151.
                    EPA REVIEW NOTICE


These proceedings have been reviewed by the Environmental
Protection Agency and approved for publication.  The contents
of this report are reproduced herein as received from the
authors.  Approval does not signify that the contents neces-
sarily reflect the views and policies of the Environmental
Protection Agency, nor does mention of trade names or commer-
cial products constitute endorsement or recommendation for
use.
                    Publication Number APTD-1161
                                  11

-------
                       PREFACE
     The Second International Lime/Limestone Wet Scrubbing
Symposium was held November 8-12, 1971, in the Claiborne
Room of the Sheraton-Charles Hotel, New Orleans, Louisiana,
sponsored by the Environmental Protection Agency, Office
of Air Programs, Control Systems Division.

     The Symposium, under the chairmanship and vice-chair-
manship of Messrs. E.L. Plyler and D.R. Mayfield, began
Monday morning with the official welcome and opening remarks
by OAP's Sheldon Meyers, followed by an introduction by
Mr. Plyler.

     The Symposium consisted of nine sessions, divided into
five different areas: fundamental research, pilot scale
research and development, prototype and full scale tests,
panel discussion on scaling, sampling and analytical methods.

   Sessions 1 and 2, chaired by Philip S. Lowell, were
concerned with fundamental research.  Frank T. Princiotta
of OAP was the chairman for Sessions 3 and 4 and the first
three presentations of Session 5, dealing with pilot scale
research and development.

     The last two presentations of Session 5 and Sessions 6
and 7 were chaired by H.W. Elder.  The 17 papers presented in
these sessions were concerned with prototype and full scale tests

     The panel discussion on problems related to scaling in
lime/limestone wet scrubbing,  chaired by A.V. Slack, consisted
of the five papers of Session 8.  OAP's J.A. Dorsey chaired
Session 9,  sampling and analytical methods, which concluded the
symposium early Friday afternoon.

     AH papers presented during the symposium are included
in these proceedings except those which were given by notes
and for which there exists no written text.
                            111

-------

-------
                              CONTENTS

Title
                           VOLUME I
     Preface
     Philip S. Lowell
     SUMMARY: FUNDAMENTAL RESEARCH - PARTS I AND II ...........    1

                  FUNDAMENTAL RESEARCH - PART I

     C.Y. Wen and S. Uchida
     Simulation of SC>2 Absorption in a Venturi Scrubber
        by Alkaline Solutions .................................    3

     M. Epstein, C.C. Leivo, and C.H. Rowland
     Mathematical Models for Pressure Drop, Particulate
        Removal and SC>2 Removal in Venturi, TCA, and
        Hydro-filter Scrubbers ................................   45

     Delbert M. Ottmers, Jr.
     A Model for the Limestone Injection-Wet Scrubbing
        Process for Sulfur Dioxide Removal from Power Plant
        Flue Gas ..............................................  115

     James L. Phillips
     Precipitation Kinetics of CaSC>4. 2H20 .....................  1 51

     D.C. Drehmel
     Limestone Types for Flue Gas Scrubbing ...................  1 67
                  FUNDAMENTAL RESEARCH - PART II

     J.M. Potts, A.V. Slack, and J.D. Hatfield
     Removal of Sulfur Dioxide from Stack Gases by
        Scrubbing with Limestone Slurry: Small-Scale
        Studies at TVA 	  195

     L.H. Garcia
     Absorption Studies of Equimolar Concentrations of NO
        and N02 in Alkaline Solutions 	  233

     J.D. Hatfield and J.M. Potts
     Removal of Sulfur Dioxide from Stack Gases by Scrubbing
        with Limestone Slurry: Use of Organic Acids 	  263

-------
                     CONTENTS (continued)

JL                                                            Page

  J.S.  Morris
  Potential  Water Quality Problems Associated with
     Limestone Wet Scrubbing for S02 Removal from
     Stack Gases 	    285

  Linda Z. Condry, Richard B. Muter and William F. Lawrence
  Potential  Utilization of Solid Waste from Lime/Limestone
     Wet Scrubbing of Flue Gases 	    301
 Frank T.  Princiotta
 SUMMARY:  PILOT SCALE RESEARCH AND DEVELOPMENT - PARTS I,
     II, AND III 	    315

          PILOT SCALE RESEARCH AND DEVELOPMENT - PART I

 B.N.  Murthy,  D.B.  Harris and J.L. Phillips
 Sulfur Dioxide Absorption Studies with EPA In-House
     Pilot Scale Venturi Scrubber 	    325

 I.S.  Shah
 SO2 Removal Using  Calcium Based Alkalies Pilot Plant
     Experience 	    345

 R.A.  Person,  C.R.  Allenbach, I.S. Shah, and S.J. Sawyer
 A Pilot Plant Test Program for Sulfur Dioxide Removal
     from Boiler Flue Gases Using Limestone and Hydrated
     Lime	    373

 R.J.  Gleason
 Limestone Scrubbing Efficiency of Sulfur Dioxide in
     a Wetted Film Packed Tower in Series with a Venturi
     Scrubber 	    391

 T.M.  Kelso, P.C. Williamson, and J.J. Schultz
 Removal of Sulfur  Dioxide from Stack Gases by Scrubbing
     with Limestone  Slurry: TVA Pilot Plant Tests.
     Part I - Scrubber-Type Comparison	    437

 N.D. Moore
     Part II — Experimental Design and Data Analysis for
     Spray and Mobile-Bed Scrubbers 	    462
                                 VI

-------
                         CONTENTS  (continued)

Title                                                            Pag

            PILOT SCALE RESEARCH AND DEVELOPMENT - PART  II

     A. Saleem, D. Harrison, and N. Sekhar
     Sulfur Dioxide Removal by Limestone Slurry in a Spray
        Tower 	    48

     J.L. Shapiro and W.L. Kuo
     The Mohave/Navajo Pilot Facility for Sulfur Dioxide
        Removal 	    50]

     J.H. McCarthy and J.J. Roosen
     Detroit Edison Pilot Plant and Full-Scale Development
        Program for Alkali Scrubbing Systems—A Progress
        Report 	    52!

     A.L. Plumley and M.R. Gogineni
     Research and Development in Wet Scrubber Systems  	    54'

     D.E. Reedy
     Lime Scrubbing of Simulated Roaster Off-Gas 	    56'

                          VOLUME  II
           PILOT SCALE. RESEARCH AND DEVELOPMENT - PART III

     John M. Craig, Burke Bell, and J.M. Payadh
     Mobile Pilot Plant Study of the Wet Limestone Process
        for SO2 Control 	    57!

     Robert J. Phillips
     Sulfur Dioxide Emission Control for Industrial Power
        Plants 	    60:

     Ivor E. Campbell and James E. Foard
     Sulfur Oxide Control at the Copper Smelter 	    63?
     H.W. Elder
     SUMMARY: PROTOTYPE AND FULL SCALE TESTS - PARTS I, II,
        AND III 	   65 •

               PROTOTYPE AND FULL SCALE TESTS - PART I

     James Jonakin and James Martin
     Applications of the C-E Air Pollution Control System	   657
                                 vii

-------
                       CONTENTS  (continued)

'itle                                                            Page

    E.G. McKinney and A.F.  Little
    Removal of  Sulfur Dioxide  from Stack Gases by
        Scrubbing with Limestone  Slurry: Design Considerations
        for Demonstration Full-Scale System at TVA	    673


              PROTOTYPE  AND FULL SCALE TESTS - PART  II

    M.  Epstein, F.  Princiotta, R.M. Sherwin, L.  Szeibert,
    and I.A. Raben
    Test Program for the EPA Alkali Scrubbing Test Facility
        at the Shawnee Power Plant  	    693

    R.M. Sherwin, I .A. Raben,  and  P.P. Anas
    Economics of Limestone  Wet Scrubbing Systems 	    745

    J.D. McKenna and R.S. Atkins
    The RC/Bahco System  for Removal of Sulfur Oxides and
        Fly Ash  from Flue Gases 	    765

    Gerhard Hausberg
    The Bischoff-Process—Initial  Results  from a Full-Size
        Experimental Plant 	    785

    J.J. O'Donnell  and A.G. Sliger
    Availability of Limestones and Dolomites  	    799


              PROTOTYPE  AND FULL SCALE TESTS - PART  III

    Tsukumo Uno, Masumi  Atsukawa,  and Kenzo Muramatsu
    The Pilot Scale R&D  and Prototype Plant of MHI Lime-
        Gypsum Process	    833

    Lyman K. Mundth
    Wet Scrubber Installations at  Arizona  Public Service
        Company  Power Plants 	    851

    J.H. McCarthy and J.J.  Roosen
    Detroit Edison  Full-Scale  Development  Program  for
        Alkali Scrubbing  Systems   (The material in  this
        paper was included in Paper No.  4c.)  	    863
                                 Vlll

-------
                          CONTENTS (continued)

Title

     Robert R. Padron and Kenneth C.  O'Brien
     A Full-Scale  Limestone  Wet  Scrubbing  System for the
        Utility Board of  the City of  Key West,  Florida  	    865

     J.A. Noer and A.E. Swanson
     Air Pollution Control at the Northern States Power
        Company Sherburne County Generating Plant 	    877

     J.W. James
     Ontario Hydro's Prototype Limestone Scrubber for SO2
        Removal from Clean Flue  Gas  	    899

     D.T. McPhee
     La Cygne Station Air Quality System  	    907

     J.F. McLaughlin, Jr.
     Sulfur Dioxide Scrubber Service  Record Union Electric
        Company—Meramec  Unit 2	    915

     B.C. Gifford
     Will County Unit 1 Limestone Wet Scrubber  	    917

     H.P. Willett  and I.S. Shah
     A Summary Report—Chemico's Commercial Systems
        Installations at  Electric Power Generating Stations  ...    931
     A.V.  Slack
     SUMMARY:  PROBLEMS  RELATED TO SCALING  IN LIME/LIMESTONE  WET
        SCRUBBING  	   943

        PROBLEMS RELATED TO  SCALING  IN  LIME/LIMESTONE WET SCRUBBING

     A.V.  Slack and J.D.  Hatfield
     Removal of Sulfur  Dioxide from  Stack  Gases by Scrubbing
        with Limestone  Slurry: Operational Aspects of the
        Scaling Problem 	   947

     Bela  M. Fabuss
     Calcium Sulfate  Scaling 	   965

     Joan  B. Berkowitz
     Review of Scaling  Problems in Limestone Based Wet
        Scrubbing  Processes  	   975
                                   IX

-------
                         CONTENTS  (continued)

Title                                                             Page

     J.R. Martin
     Deposition Problems and Solutions  in the  Combustion
        Engineering Lime/Limestone Wet  Scrubbing  Systems  	    983

     Philip S. Lowell
     Use of Chemical Analysis and  Solution Equilibria  in
        Predicting Calcium Sulfate/Sulfite Scaling  Potential  ..   1001
     J.A. Dorsey
     SUMMARY: SAMPLING AND ANALYTICAL METHODS  	   1013

                  SAMPLING AND ANALYTICAL METHODS

     Klaus Schwitzgebel
     Development and Field Verification of Sampling and
        Analytical Methods for Shawnee 	   1017

     E.A. Burns and A. Grunt
     On-Stream Characterization of the Limestone/Dolomite
        Wet Scrubber Process  	   1053

     Terry Smith and Ronald Draftz
     Particulate Emissions from Two Limestone Wet Scrubbers  ...   1073

     Terry Smith and Hsing-Chi Chang
     Design Criteria for a Size-Selective Sampler for Lime/
        Limestone Wet Scrubbers 	   1083

     R.M. Statnick and J.A. Dorsey
     Instrumental Methods for Flue Gas Analysis  	   1097

     Gene W. Smith
     EPA Recommended Source Test Methods for New Source
        Performance Standards Testing  	   1109

-------
                  MOBILE PILOT PLANT STUDY
                           OF THE
                   WET LIMESTONE PROCESS
                            FOR
                        S02 CONTROL
                             BY

                   JOHN M. CRAIG, Ph. D.1
                         BURKE BELL^    ?
                    J. M. FAYADH, Ph. D.
              Prepared for Presentation at the
            SECOND INTERNATIONAL LIME/LIMESTONE
                  WET SCRUBBING SYMPOSIUM

                   NEW ORLEANS, LOUISIANA
                    NOVEMBER 8-12, 1971
1  Present Affiliation - Southern Services Inc., Birmingham, Ala.
2  Zurn Environmental Engineers, Washington, D. C.

-------
                        INTRODUCTION

     One of the earliest processes utilized for the removal of
sulfur dioxide from combustion erases involved scrubbing with a
slurry of lime or limestone in water.  This process was inves-
tigated in England in the 1930's and with some significant
variation has been studied since then by interested parties
throughout the world.  The federal government^ utilities, and
hardware manufacturers have recognized the potential of lime-
stone scrubbing and are actively engaged in research and develop-
ment programs to develop the process and demonstrate its
commercial application.
     A critical step in the commercial utilization of this
process is, of course, experimentation at the pilot plant scale
to develop firm data on design,  operation, and economics and
to re-examine laboratory data in a more realistic environment.
The pilot plant stage allows a more critical review of factors
such as process chemistry, mass transfer mechanisms, scale
formation and control, effect of limestone characteristics,
optimum scrubber design, and identification of water pollution
problems.
     Zurn Industries, Incorporated along with the Office of
Air Programs, Environmental Protection Agency conducted a joint
research project which involved a pilot plant study of wet
limestone scrubbing: utilizing a Zurn Air Systems Division
                             576

-------
 "Dustraxtor" scrubber.  The pilot plant was mobile and was
 installed at two electric generating stations, an oil-fired
 station  in Key West, Florida and a coal-fired station in Paducah,
 Kentucky.  The study was applied in nature and investigated
 variables important to the success of the limestone scrubbing
 process.  Data were developed which will ultimately be utilized
 in a demonstration project to be discussed later during this
 symposium.  Since Zurn Industries has an interest in the
 commercial development of this type of system, we contributed
 significantly to the cost of the project through the design
 and construction of the pilot plant.

            CHEMICAL CONTACTING IN THE DUSTRAXTOR

     The objective of flue gas scrubbing for mass transfer is
 to remove as much material as economically possible out of the
 gas and  into the scrubbing liquid.  There are physical, chemical
 equilibrium, and rate relationships that limit the solubility
 of material in liquid that control the mass transfer relations.
 However, the amount of mass transfer not only depends upon the
 equilibrium relationships, but also on the contacting scheme.
 The "Dustraxtor" employs many contacting schemes in its operation.
     The "Dustraxtor" scrubbing unit employed in the pilot
 plant study is a type of turbulent contact scrubber and is
 shown in Figures 1 and 2.  The unit consists of a flooded,
 collecting tube through which the flue gas must pass.  The
 collecting tube is installed vertically in the inlet plenum
 chamber directly above the recycle hopper so that the bottom
 of the tube is a short distance above the liquid level in the
hopper.
                              577

-------
     Flue pases enter the inlet plenum chamber where they sweep
over the surface of the scrubbing; liquid and are directed up
the collecting tube.  The velocity of the gas passing beneath
the collecting tube and above the collecting bonnet draws the
scrubbing slurry from the bonnet surface and upward into the
collecting tube.  The shearing action of the gas atomizes the
scrubbing slurry into a dense spray as the gas-slurry mixture
continues up through the collecting tube.  The result of this
action is a highly turbulent mixing zone which provides the
intimate contacting and time necessary for the chemical reaction
to occur.  As the gases are discharged from the collecting tube,
they are directed into a curved liquid deflector which acts as
a separator by forcing the slurry downward onto the tube sheet.
This shower effect provides an additional mixing zone for
absorption with chemical reaction to occur.  The cleaned gas is
discharged through the exhaust stack and the scrubbing solution
is returned by gravity to the recycle hopper.  When employed
for limestone scrubbing, this built-in recycle of the scrubbing
slurry provides sufficient time for the limestone to go into
solution and react with the sulfur oxides.
     The stages of contact in this type of unit are:
     1.  The initial shearing action as the flue gas passes
         through the slot between the collecting bonnet and tube,
     2.  The highly turbulent mixing zone within the collecting
         tube where the gas is in intimate contact with liquid
         droplets.
     3.  Impingement of the gases and liquid upon tne surface
         of the deflector.

                             578

-------
     J4-.  The passage of the gas through a highly turbulent curtain
         of liquid being discharged from the liquid deflector.
     The contacting: schemes found in this type of unit include
counter-current,co-current, and cross-current.  Since there are
many contacting schemes taking place at the same tinie in this
unit, a theoretical analysis of the total mass transfer mechanism
would be very complicated to perform and may prove to be only of
academic interest.
     The unit normally operates at a pressure drop of from 6
to 8 inches WG.  This pressure drop is regulated by either
raisinpr or lowering the adjustable level control weir.  One
advantage of this type of  unit in limestone scrubbing applications
is that there are no moving parts, spray nozzles, or pumps
necessary for operation.   By variation in weir height, bonnet
spacing or tube selection, a wide range of operating conditions
may be met at maximum efficiency.
            DESCRIPTION OF THE MOBILE PILOT PLANT

     Zurn Industries cooperated with the Office of Air Programs,
Environmental Protection Agency on a project in which a mobile
limestone scrubbing pilot plant was designed, manufactured, and
operated.  The major objectives of the pilot plant study were
to evaluate the potential of this type of turbulent contact
scrubber in the wet limestone process, investigate process
chemistry and kinetics, and generate design data for scale-up
purposes.  The first phase of the pilot plant study was conducted
in Key West, Florida, at the municipal power plant which burned
approximately 1-2 percent  sulfur content fuel oil.  After this
                                 579

-------
phase of the project was completed, the mobile pilot plant was
moved to the TVA Shawnee coal-fired power plant for the second
phase of the study.
     Figure 3 is a schematic flow diagram of the pilot plant.
The scrubber had a nominal capacity of 1500 cfm.  The dry
collector permitted operation with and without fly ash when
used during the second phase of the study.  Other salient
features of the pilot plant were capabilities to handle a wide
range of gas and liquid flow rates, construction materials to
withstand a wide range of pH (pH 3 to 9), and provisions for
ready access to the scrubber to check on scale buildup and for
scale removal, if necessary.
     Six types of limestone reactants were evaluated (coral marl,
aragonite, Predonia Valley limestone, dolomite, precipitated
calcium carbonate, and lime).  A background series of tests were
conducted with salt water.  The stoichiometry of the reactant
and the effect of slurry concentration were studied for selected
limestones.  Stoichiometric requirements were based upon the
molar requirements for the reactant to form CaSO~.  For selected
reactants, the influence of particle size on sulfur dioxide
removal efficiency was studied.  The effect of an iron catalyst
on sulfur dioxide removal efficiency and on the composition of
the spent liquid slurry was also evaluated.
     The scrubber variables evaluated with respect to the
efficiency of sulfur dioxide removal for the most effective
reactant combinations include gas flow rate, system pressure
drop, stoichiometry, particle size, and slurry concentration.
Other factors analyzed during the test program included but

                              580

-------
were not limited to:
     Corrosion of the various metals utilized throughout
       the system.
     Scale formation.
     Operating; and maintenance problems associated with
       operation of the system.
     In order to evaluate these variables, the following data
were collected during each experimental run:
     Boiler operating conditions.
     Inlet and outlet gas samples including sulfur oxides,
       nitrogen oxides, particulates, flow rate, temperatures
       (wet bulb and dry bulb).
     Compositions of the slurry feed to scrubber and scrubber
       slurry discharge including solids concentration, calcium
       and magnesium, sulfate and sulfite, nitrogen compounds,
       chloride, pH. temperature, and flow rate.
     The equipment utilized for this study included a modified
Dustraxtor containing one tube.  The scrubber was modified by
the addition of an external weir to enable measurement of the
quantity of liquid drawn up the tube.  The scrubber was designed
such that various tube diameters could be tested in the system.
(12 inch I.D. and 8 inch I.D. tubes were utilized during this
study).  A continuous record of the inlet and outlet sulfur
dioxide concentration was taken during each experimental run
utilizing a Dynasciences Model SS-130 analyzer connected to a
strip chart recorder.  The analyzer was calibrated twice daily
utilizing a known standard of certified calibration p;as.
Intermittent nitrogen oxide and particulate samples were taken
                                 581

-------
utilizing standard stack sampling techniques and analyzed wiun
standard procedures.  Gas flow rate was determined with a
calibrated "Annubar" flow meter.
     The mobile pilot plant was installed on the Number 3 unit
of the Key West Municipal Power Plant during the first phase
of this study.  This unit has a generating capacity of 20 MW
and was burning 1 to 2 percent sulfur content fuel oil during
the experimental period.  During the second phase of this study,
the mobile pilot plant was installed on the Number 9 and 10
units of TVA's Shawnee generating station.  Each unit has a
generating capacity of 150 MW and was burning 2| to ^ percent
sulfur content coal during the experimental period.  The Number
10 unit at the Shawnee plant was also in the dry limestone
injection configuration and this gas was scrubbed in the mobile
pilot plant.
                 RESULTS OP THE TEST PROGRAM

     This section of the report will discuss the results from
the three major phases of the test program as well as operating
and maintenance problems associated with the pilot plant.
Generally speaking, the pilot plant performed satisfactorily
with only the normal problems that might be expected in this
type of operation.
     A detailed final report on this project is currently being
prepared and will be available soon.  This presentation will
only summarize some of the tests conducted on an oil-fired boiler,
a coal-fired boiler, and a limestone-injection boiler.  The
results were favorable and indicate that this type of scrubber
                              582

-------
should be considered for use in the wet limestone S02 control
process.
Key West Tests
     Initial tests were conducted at the Key West site to
determine the quantity of liquid that was drawn up into the
tube.  Figure ^ summarizes the data accumulated during this
phase of the test program.  As can be seen from Figure ^, the
quantity of liquid drawn up into the tube is a function of the
gas flow rate and pressure drop across the scrubber.  Liquid
to gas ratios of 100 to greater than 500 gallons per MCFH were
attainable with this scrubber configuration.
     Figure 3 illustrates the scrubber configuration and sample
locations utilized during the limestone scrubbing phase of the
Key West test program.  A fractional factorial design was
employed to determine the major parameters which effect SOp
removal in this type of scrubber.  Tables 1 and 2 summarize
the field test data associated with the experimental design for
each primary reactant, coral and Fredonia Valley limestone.
Based upon the experimental design, stoichiometry, gas flow
rate and pressure drop are highly significant factors in the
design of a S02 -limestone scrubbing system using this type
of scrubber.  Figure 5 summarizes the results obtained with the
two primary reactants studied.   Generally, at a stoichiometric
ratio of 1:1, a pressure drop of 6.5 inches H20, and 2000 ACFM
gas flow rate, efficiences of 50-75 percent were attained
depending: upon the firrind of the reactant studied.  An increase
in scrubber pressure drop to 12 inches H20 and a decrease in
    flow rate to 1000 ACFM improved removal efficiency to a
                                  583

-------
level of 75 to 90 percent.  However, it should be pointed out
that the scrubber was in a once-through configuration and was
only operated for a period of 3 to ^ hours per test.
     Figure 6 summarizes the effect of pH upon SOp removal
efficiency during this phase of the test program.  Due to the
location of the fresh slurry feed line, in relation to the
spent slurry discharge, improper mixing occurred in the scrubber,
and the discharge pH was not as reliable an indicator of this
effect as the weir pH.
     Another test series was conducted to determine the effect
of gas flow rate upon S02 removal efficiency.  The test conditions
were fixed, and only the gas flow rate through the scrubber
was varied.  Figure 7 summarizes the results obtained during
this series of tests.  The results were as expected, when the
L/G ratio for a given tube diameter increased, the S02 removal
efficiency increased.  The liquid flow rate utilized in the
L/G calculation is actually the liquid drawn up into the tube
based upon Figure k and not the actual liquid flow to the
scrubber.
     Tests were conducted using other reactar.ts, dolomite, lime,
and precipitated calcium carbonate.  The results of this phase
of the test program are summarized in Table 3.  All reactants
ran satisfactorily except for precipitated calcium carbonate
which caused heavy foaming at the weir and inside the scrubber.
Shawnee Unit No. 9 Tests
     Table ^ summarizes the Fredonia Valley limestone tests
conducted while the mobile pilot plant was connected to the
No. 9 unit at the Shawnee station.  Due to pressure drop
                            584

-------
consideration across the entire system, it was necessary to



conduct all tests with an 8 inch diameter tube.  A fractional



factorial design similar to that used during the first phase



of the test program was utilized for the primary reactant.



Commercially available grinds were used to determine the effect



of particle size on SG>2 removal efficiency.  Each experiment was



conducted with the primary dust collector in operation as



depicted in Figure 3.  The experimental technique was identical




to that utilized in Key West, wherein the liquor fed to the scrubber




was based upon the test stoichiometry, gas flow rate and SOp




concentration in the entering gas.  Figure 8 summarizes the




results from this test series.  The highly significant parameters




for this system were found to be pressure drop, gas flow rate,




and stoichiometry.  Particle size was not a significant factor




due to the large amount of 100 mesh material passing the 325




mesh screen.  This situation occurred in previous studies and




made it very difficult to determine the effect of a coarser




grind upon S02 removal efficiency.




     During testing on No. 9 unit, little or no mechanical or




operating difficulties were encountered.  Scale formation was not




a problem due to the installation of a device to continually




wash down the tube.  Based upon the limited number of tests




and operating hours on a coal-fired unit, removal efficiencies




of 35 to 80 percent may be expected in a once-through





                                  585

-------
configuration depending upon gas flow rate, stoichiometry,


particle size and pressure drop.  However, additional testing


would be required prior to any large scale-up of this scrubber


for control of S02 from a coal-fired station.


     An abbreviated test series was conducted inorder to


evaluate the effect of ionic strength upon S02 removal.  A


increased ionic strength slurry solution was tested at a


similar pressure drop, gas flow rate and stoichiometry as was


used during the first phase of this study.  The results are


shown below:


        Location                       SOg Removal Efficiency %


     No. 9 Unit                                  69.6


     No. 9 Unit, Simulated Ionic Strength        75.5
                 (12,000mg/l Cl~)


     Key West (24,000 mg/1 Cl~)                  84.2


Shawnee Unit No. 10 Tests


     Figure 9 illustrates the configuration of the mobile


pilot plant when installed on the No. 10 unit at Shawnee.  All


gas to the scrubber, laden with calcined limestone from the


limestone injected boiler, by-passed the primary mechanical


collector and entered the scrubber.  Table 5 summarizes the


results from this test series.  As can be seen, S02 removal
                              586

-------
efficiencies of 60 to 9^ percent were attainable depending upon
stoichiometry, gas flow rate, and pressure drop.  However, due
to the design of the scrubber, the entering gas and limestone
impinged upon the tube and built-up until the cake fell off due
to its weight.  A re-design of the entrance to the scrubber would
be necessary before this unit could be applied on a limestone
in.lection boiler.  The low inlet S02 concentrations found
during this test program can be attributed to leakage of
ambient air into the system.
Operating Problems
     In addition to the operating problems discussed previously,
two other problems were associated with the mobile pilot plant:
corrosion and scale build-up on the blower.  Since the scrubber
was initially tested utilizing salt water, severe corrosion
problems occurred in the scrubber and discharge piping.  However,
once the salt water tests were concluded and fresh water was
utilized in the system, no additional failures due to corrosion
were encountered.
     Scale build-up on the blower blades caused serious operating
problems more than once during the life of this project.  The
scale build-up would occur on the blower blades, causing an
imbalance in the system, and in some cases, ultimate failure
of the blower.  This problem could have been avoided by using
a "hot-side" blower instead of locating one down-stream of the
scrubber system.
Scale Formation in the Scrubber
     Scale formation occurred at two points in the scrubber
system.  Initially, scale was discovered at the tip of the tube

                                 587

-------
in an area of high turbulence and then at a later date, some
slight crystal formation was noticed on the scrubber housing
at the water line.  Heavy scale formation on the tube lip
could develop into an operating problem if not corrected, since
the crystal growth occurred directly in the gas path; therefore,
precautions were taken to eliminate scale formation at this
location.  Table 6 indicates the major compounds found in a
sample of the scale formed on the tube.

                           TABLE 6
                   ANALYSIS OF TUBE SCALE
          Compound                        Quantity (wt %)
        CaSO^ •  2H20                          30.5
        CaCCU                                 68.0
        MgC03                                  1.5

Rate of scale formation appeared to be a direct function of
the slurry concentration; however, no effort was made to
quantify this conclusion.  An appreciable difference was not
noticed in the rate of scale formation caused by the primary
reactants tested, coral or Predonia Valley limestone.
                              588

-------
   BALANCE VENT LINE



          DEFLECTOR





    COLLECTING TUBE




  COLLECTING BONNET





          WIER




    WATER IN




LEVEL CONTROL
      DUST AND WATER OUT
        FIGURE 1.  TYPICAL CROSS SECTION OF DUSTRAXTOR
                        589

-------
            CLEAN AIR OUT
                                                  OUTLET to
                                                      ROUND BOOT CAS UK
                                                           WATH DDLBCKM
                                                      DIRTT GAS IM MOD WATH
                                                       HOPPER CLEANED MANUALLY OB

                                                       BY ADDITION Of MECHANICAL
mm®
      COLLECTED MATERIAL

      TO BOTTOM OF HOPPER
               FIGURE 2.   DIAGRAMMATIC VIEW OF SCRUBBING
                         ACTION WITHIN DHSTRAXTOR

                                       590

-------

-------
                          FIGURE
 f   *
 I   5
^ i
-------
8°--
60--
So--
                                                              £> LlHSSTofJ*
                                                              v  Line
                                                              B $AiT l^fTf^.
                                                              A  /xoi^vr^
4-0 • •
Jo--
 /o --
                                   FIGURE  5
                                                593
                                                                7

-------
 1
I
 •a:
 «
                         o
                        >"
                        -oil
§   v,
<   s
                                                              C
                                                           1  s?
Hj     K
*   § 5
vi   8 >
                                  «S
                                                                         f N
                                                                                    W
                                                                                    M
                        ^
                        >N
                        '«!
                        in
                                                                         I N
                                                                               I
                                                                        4- tv.
                                       fi
                                   594

-------
fe
 1
 I
^3
             -3
                                    
                                                                                                   O
                                                                                                   M
                                                                                                   PH
                                                                                        I
                                                                                        A
                                               595

-------
   /oo 4-
   fc
i
I
   *  \-
                     = 6.C
                          4

                                                             -6 ^
                             FIGURE 8

                                  596

-------
                                                   O\

                                                   w
                                                   K
                                                   £>
                                                   c5
                                                   M
                                                   fr.
597

-------
I
<
S3
o
M
EH
O
<
     W
     CO
                                           598

-------
TABLE 2
SUMMARY OF KEY WEST FRACTIONAL FACTORIAL DATA
FenoM WLity L'^e-sro^f- FIELD TEST DATA ^/?"^ -^-j^-)
,cp
7
IL,
1
U-.'
\ — '! —
L '
f Jl'f
Vl
H
_j
• j
_i...
if
Tji
. o!
\ i;
i o
tl ',
', r~
•i'
-9
, 1 (
: ^
'V
:
00 in -
vA • ••
+' o' «o,
; *j £ ? _
! *^ O. (Tl
r S?I - w
i
or' -i .« ; ui
•^•j -S . —
- 	 I
1 t- 1 -
°i °. -
CT rO «o' \n
r ~"l ~'
o
J in frt
A^9. -1 *
or ifl KV
i
J) 0
c~ — i l!i ~
, • i
t w <":
r sil i! *
1 ;
to o' ~
^ ~i •• in
± -, ~
Vo o —
£"••• —
 fQ
£— -*9 *"~
, i
°° ^. £ _
i i •
P •'•'* —
' ol rO
! rO ' *! VO - .
1 I ^ vi — *"
i j
' U.
! i
1
1 !
•1 ••
l -
i
1 to
r!
o
•H
f« jj
! 0) -M
' B ; o
i .P j->
w w
Ig-tl
& 0! rf, ,
J' ci •• *°
| ! '
"o', —
CM
x o ' r.
•^ .-HO
H • -P -H
«,H n rt ' 4^
o -H K ' c)
CO *— i,
•*— ' O 4-5
a -H c:
0) O (.. 01
4J ^. 4J Cl
-
Iti t/l o t(
to -H f,
 01 o TJ
« »-, 4J rH
O Pi C/l 10
1 -. o S ° Ul
o ,9-2i ^Sy^r ° cot/1 oo^o
- w - w ' --<,| ^ * fi; , ^ 1/5 In
1 !
o; *0 tv-
°; tC —
cxi ' \n f |O
•tO , ' N, -
, -: i ly
I v/i ; o ! o^
r^ i — c**
10. . , CM -
Oj i <0 K>
Q ' ' Vl |O
— j I 1 M ~
1
, ift| J' u>
01 o W
Oj . o fJ
i ; ^
1 :
m ' • i o" o
N : • — to
"O . CM —
! •
0 j j 0 , C-1
• i
or ! o:vp
o i : &• cy
,
0^ i , ° *
"i ! .cv'
| i ;
! ' '"^
OS ' cP! tJ
°i ' fO, |rt
"i ! «° ~
101! ! rj' f"
M.1 ' - 01
-u) O 4J E-i E-i
^: O £
o (=-.
Ci, fl
- c
0 0
O o
Vn l/i
— . i " J I 1 y 0
•Til loS^^'i ivj|a3-; r^iiriviln^
i • < i : l
1 i ! i
VP *si 9
IT) ' ^' '•^
^L...^:.1^":..
1  ^
, v. — — —
1 L 31
• - "I — , ~r '
0 J «3 0
!£ I i~ ^.
j 	 J 	 . ,_
•*' l 00 f>
"* : o r); i f~ t-j o-
O i co — • ' c~ in ^*
i » ' i
r . j i M 0) .
"to i " 0!~«:~o"
! ift o £ Ji ^, «:
.-. T -u°°1 .7i_ ^i i W
' ! ! ' -J
j 
£ SI -: -,
in ' ' in
Ul ' Qi Q t/1
1 "~^ «^\' 1
C-j | -; -(
~ ' J 00 jj
y> ' j] «-; 1^
£ l i «.S N l
°0 — , cN|
: i !
•^i t-J oo. to i
^j- ' i 10 • «j' in
5 -_ 00 o 0
•* CP s
6 °° • — c^ ^ «f
. : , !
- : J f* j - ."
r>i r-: Sr ^ in ^
r -s ° J *t *
ui *"! — ' r1 vr> IP
i :
. . . ^ .... .
^ .. w' ET .5
r; ; C~ ; 0 *
—I : — «* ' (r f tr
o< ! . ,-! o. >«:
I ^- '  C E •—
r r, t-i i, fj r; o P ' ^ ; i
0)01 ^^Clfll-HtlOlf^-W 4J
OO 4->4->004->— f^Ofl 'C
dC cd o c c -H ^s^^oi UP.
OO ^^.OO'dOlJJ 3 pi 01
oohrcJa>oo^4Jrt'di-t T)I-IP-
C o, a, o ci f, 01 tn «i <,< P.
c« « -H r K CM x' o (> o w ^ ti •. o f-
01 t, rH Cl K
^ f< fr. EH P.
3
r-l
:io
o'
1 • •*.!
•*5 o
CO 
1 -
«~J lo :
In lo i1
' i.
01
01
Cl f~
3 <0
599

-------
                           TABLE 3

           RESULTS OF SECONDARY REACTANT TEST PROGRAM
Reactant   Gas Flow Rate Stoichiometric  Pressure     S02 Removal
               (SCFM)	Ratio    Drop (in H?0) Efficiency (%)
                                                        86.4

                                                        94.9

                                                        93.4

                                                        97.6



                                                        32.6

                                                        51.2

                                                        51.7

                                                        51.8



                                                        67.4

                                                        85.5

                                                        81.9

                                                        a
Lime
1445
790
782
1455
Dolomite
1430
1430
772
772
Precipitated Calcium
1431
778
773
1435

1:1
1:1
1:3
1:3

1:1
1:3
1:3
1:1
Carbonate
1:1
1:1
1:3
1:3

6.5
12.0
6.5
12.0

6.5
12.0
6.5
12.0

6.5
12.0
6.5
12.0
a  Sample contaminated due to heavy foaming and carry-over.

NOTE:
1.  Slurry concentration - I"o by weight
2.  Particle size - no control due to use of commercially
    available materials.
                             600

-------
                          °.  r>
                                            cr -
                                                                              03
                                                                                  —    
                                      ,J  0  0
                                      in  -  
oO j
oo i;
                                                                                   5
                                                                                  if
                                                                                           Ir
                                                                                               ti
                                                                                                (5
                                                                                                     .
                                                                                                     In
0, 0
a o i
— r—
1,
1 o. , o. 0 ^ ~ | |
.1 , {Q 	 C>- V>
                                                                                                0
                                                                                                •*
, CO
K
O
1 T4
t, -p

CO CO
, 0) tt)
1 E-< EH
O
CM
X
> 0
t, --I
I-. 4J
3 t,
r-l CD
CO P^


CO
o
*H
4^>
•H
•C)
1 O
' O

CO
„ (IS
'' C

hf
R
«H
tn
 4^
(tf CS
t, ^
(1) 
'*- V,
O1
J



p
r-d

(3
t
ID
P
[3
tt)
f-i





ature (Twb)
Mcantratlon (ppm)
t. O
0) LI
p,
1= CM
a) o
EH CO





ncentration (ppm) ,
O i
o
X1
§1
r
f
^
i

tn •—
§ a
,H hr
-p —
^H
•d a)
C -P
O 0)
0 (-
>! O
t. rH
r< It,
3
<— 1
IO
r

a
Q)
EH



• ft?
0
? s
>— fl)
3
Tj H
0) VH
CU  0)
fc W
0)
P It)
p. P
0 3
3J H





                                                        601

-------
m
2
      (K
     W
     EH
     O
     H
     EH
     O
     W
     •-3
W

o
EH
     W
     W
     fc
     O
IS
a

w
                o
               e
               u
              £
                i
               o
               a.
               o



I



1
U-
I
i—
i
j;

§
• i
h
J
Ql
H
r.!.
,
1 *"
ift
i ct-
I1"



U
iS
fr

h

,T
H
i ^

I




• '



K

4)
u
0)
H
"
j
1
i
1
1


—
•«»•
t
f\


cr
0

5
^
.£•
tsi
-
•
"-"
o
6
0
I'l
~~
VI

• 0





o
^4
a
W

chlometrlc
^4
o
4J
l/J










	
1

1
1

1
~1
1

,
1
I



1

|


I

M.
J
jc
NuO

C
0
tH
01
t.
ry Concent
i.
3
i-l
t/}


i#-i .
i
1






-
O

0
o
0
o
o

*^'


_3
—





*— s
£l
t*
*+-*

empersture
H





»>-^i»

r~



* •


M

o
SM
o
•a-
C\l
»o
N

N
M
	 .
C
O
O

o
(N
*
t>
M

(XT
O
M
c-

i
r

0
^

e
P,

N^

C
0
^H
tl
(,
02 Concent
0)





r








i


,

'
—
i
i

•
i
•

i

^

I


)


R
P.


^
fi
+j
01
^
Ox Concent
^





•-

i
1











—






















•A


U
d

^



--








f

r
O
£
~
0
—
— '
0
/^
-
vo
(0

s
?
K)
£1

0
"11
E

r?
•

Jo
—





«•»«.
/^
^?
..^

enperature
H














r-

r-
r

f
—
—

«
Q


JS.J
-
le

"r^


o
—





^•n
^Q
F?


P,
a
01
H





• -





~


Vi
rt^

•O
0

0
—
o

o
in
ui


"VJ
«
s
Vi

^

^>
CO

R

^^

f.

•H
n

Oo Concent
Vi





-•








1

i
i

i

i
\

i
-
i


i
i

(


\


a
P,
P,


c
o
.p
01

Or Concent

U-

^.'
ii

^:











Hopper






—






1

I

I
^

1

I

t
\
t


I
I

.


\











4>
(1
O
«
W
V
3
H






—
;




k
I
..
\
i
i
Tj

'
— ;
!
1'
\

1

','


1 ;
) j
j
* i
!l

\t



;,
||
[J
j>
L

D
I]
J
t J
'i
i|




                                                 602

-------
SULFUR DIOXIDE EMISSION CONTROL

            FOR
  INDUSTRIAL POWER PLANTS
      Robert J.  Phillips
       Project Engineer
  GM Manufacturing Development
      GM Technical Center
       Warren,  Michigan
        NOVEMBER, 1971
              603

-------
   SULFUR  DIOXIDE EMISSION  CONTROL
                       BY
               WET  SCRUBBING
Recent limitations on the emission of sulfur dioxide
have forced  industrial users  of coal to convert  to
natural gas  in order to meet  the new codes.   This has
been necessary because economic means of controlling
the emission of sulfur dioxide has not been  forthcoming.
Many approaches have been offered for the control of S0»
from large utility boilers, but these approaches are,
in almost all cases, based on the recovery of sulfur in
some salable form.  The small industrial coal user cannot
usually justify the expense of such recovery equipment for
the small quantity of sulfur  involved.
                        604

-------
General Motors is a heavy user of coal for industrial
powerhouses consuming approximately 2.3 million tons/yr.
Because of the long term availability of coal and with
the higher cost and deteriorating supply of natural gas,
General Motors has made it a policy to continue to use
coal wherever economically possible and do it within
existing pollution control codes.
                        605

-------
Most of the boilers in the corporation range from
steaming capacities of 50,000 to 150,000 Ib/hr.
with an average plant steaming capacity of 250,000
Ib/hr.

Any SO- control process that might be selected had
to first meet two corporate criteria.
First - The process must provide an economic
        incentive over conversion to an alternate
        sulfur-free fuel.
Second- The prime function of the powerhouse is to
        make steam.  The SO2 control process cannot
        be so complex as to compete with the steam
        generating function of the plant.  In other
        words , the S09 control process must be
                     £*
        nearly control and maintenance free.
                         606

-------
We made an evaluation of possible approaches to the
control of S0~.  Based on some promising results from
tests conducted by the coal industry and subsequent
tests run by TVA, it was first decided to try a dry
additive injection system to reduce SO0 emission from
                                      ^*
our boilers.  The dry additive injection system utilizes
limestone, dolomite or red mud, either fed in with the
coal or injected directly into the boiler.  The pulverized
stone is calcined in the boiler and reacts with the S02
to form a solid calcium sulfate.  This sulfate is then
collected by the dust collectors.
                            607

-------
    40
E  30
h—
o
    20
 CSI
O
CO
o
C£
UJ
Q_
    10
     0
                   50
100
150
200
25C
                        PERCENT  STOICHIOMETRY
          This system presented a very economical approach.
          Tests were conducted at the Chevrolet-St. Louis
          powerhouse.  These tests showed only  a 30% reduction
          in S09 but, at  the same time, particulate emissions
          were increased  beyond code limits.  Because of the low
          efficiencies and  increased dust loading, it was
          concluded that  the dry additive injection system is
          inadequate.
                                 608

-------
 Flue
 Gas -
 Feed

 H20 -
Co(OH)2
 or
CaCO,
         Scrubbed
         Gas
         li
       Scrubber Feed
t


	

ber




r

i .
i
i
i
Scrubber 1




i






i

1
Thickener






*j















ill
Slur
Tank


1
I

"2°
'H
y

J






>.
                Effluent
i
1-

Mixing
Tank


	 »•




1
H
T

Holding
                  LIME OR LIMESTONE SLURRY SCRUBBING
  At  the same time as our  studies were taking place,
  other  studies involving  wet scrubbing of boiler
  flue gases with lime were  also being carried out.
  The obvious advantage of scrubbing with lime is  that
  CaSOo  is  an ideal form for the disposal of the sulfur
  since  it  forms an inert  solid suitable for land  fill.
  By  far the overwhelming  problem encountered with the
  lime scrubber is excessive plugging of the system with
  insoluble CaSCU.  For this reason, we felt that  because
  of  the high maintenance  required, lime scubbing  is  not
  the answer to SO^ control  for industrial size boilers.
                            609

-------
           SODIUM  CARBONATE  SCRUBBING
            i
     SCRUBBER
                                             v   v
                                             MIXING
                                               TANK
A more practical  means of controlling SO- can be
utilized by substituting soda ash for lime in the
scrubber.   The  obvious advantage of this system is
the formation of  a soluble product thus eliminating
plugging problems.   Of course, the formation of a
soluble product is an unsuitable form for the sulfur
because a water pollution problem has now been created.
                       610

-------
 Flue
 Gas ~
 FeeJ

 H20 -

Ca(OH)2
        Scrubbed
        Gas
                       Scrubber Feed
t


ber
_-


Soda Ash
Make-up








i


Thickener

r



\ „

-v


411
Ca
Prec
Tank
2°


	 .W.

"] •
r
Mixing
Tank
i

	 »•



Cau


Wash
Water
           CaCOo
                   SODIUM HYDROXIDE SCRUBBING WITH LIME REGENERATION
  We  still  felt that wet scrubbing  would offer an economic
  means  of  controlling SO,, emissions  in industrial boilers..
  The  optimum scrubber system  should  involve the best parts
  of  both  the lime and soda ash  system.  For this reason,
  we  selected a system using caustic  soda for scrubbing with
  regeneration using lime.  In this system a caustic solution
  is  sent  through the scrubber.  The  sodium sulfite formed
  is  then  sent to a regeneration tank where lime is added.
  The  sulfur  is precipitated from the system in the form of
  calcium  sulfite and sulfate.   In  the process of precipitating
  the  sulfur, the lime also serves  to regenerate the sodium
  scrubbing solution.  In this lime regeneration system, the
  only product leaving the system is  an inert cake.
                             611

-------
After some favorable laboratory scale tests, it was
decided to run a pilot plant of the system on an
actual operating boiler.  A critical piece of
equipment in the system is the scrubber itself.  All
of our boilers are equipped with dust collecting
equipment so that the scrubber would not have to act
as a primary dust collector.  Therefore, it was decided
that a low energy scrubber would be preferred for the
system.  Nevertheless, the scrubber would still be
exposed to an environment of fly ash and also it must
stand up to possible boiler upsets and soot blowing which
would increase the fly ash loading a considerable amount
for short periods of time.  We finally decided that two
pilot systems should be run, one utilizing a scrubber that
could be classified as a gas absorber for maximum S0~
removal.  The second scrubber was selected for its ability
to handle particulates and was considered to be marginal
in absorption characteristics.
                         612

-------
The first pilot system was installed on a powerhouse
boiler at Chevrolet-Cleveland.  Arrangements were
made to use a 2800 cfm packed bed cross flow scrubber
for the tests.  This scrubber utilizes a plastic
packing noted for its good absorption properties.  The
scrubber contains a packing depth of four feet.  2800
cfm of flue gas (approximately 10% of the total flow)
is drawn out of the exhaust stack into a transition piece.
                         613

-------
The transition piece was positioned so that it points
directly into the gas stream and its diameter was
selected so that the gas velocity into the transition
duct approximates the stack velocity.  This insures
that the scrubber is seeing a fairly representative
dust loading.  The flue gas first enters a quench
chamber where the 580° F flue gas is cooled to 130° F
by spraying city water thru a series of baffle plates.
                         614

-------
The saturated gas then passes horizontally through
the four feet of packing.  Scrubbing solution is
simultaneously being sprayed vertically through the packing.
The scrubber is plastic and controls have been installed to
insure that the scrubber never sees a high temperature.  The
caustic scrubbing liquor is sprayed into a series of low
pressure full cone nozzles at a rate from 25 to 45 gpm.
                          615

-------
The spent liquor from the scrubber is sent to the
regeneration tank which contains both a mixing and
a settling section.  Lime is added to the mixing
section using a lime slurry feeder.  The sulfite
precipitate is filtered and dried using a vacuum
filter.  The regenerated liquor is sent back to the
scrubber.
                       616

-------
The boiler being used for the study is a 80,000 Ib/hr.
spreader stoker equipped with a multiclone dust
collector.  A coal with an approximate sulfur content
of 2% is used at the plant.  The boiler operates with
very high excess air (over 100%) resulting in S02
scrubber inlet concentrations of about 1000 ppm.  The
boiler is not equipped with an  economizer or air preheater
resulting in a flue gas temperature of 580° F.
                         617

-------
S00 concentrations were monitored using an infra-red
  £
analyzer.  Test ports were located at both the inlet
and outlet of the scrubber.  This monitoring capability
gave us the chance of chap'j, /,g variables and measuring
their effect immediately, thus greatly reduced the time
necessary to obtain the needed scrubber design data.
The first tests on the scrvbber were run to determine
scrubber performance under different operating conditions.
The variables with the most affect on scrubber performance
were the liquid and gas rates to the scrubber.
                         618

-------
               REMOVAL OF S02 VS.  GAS RATE
o
   40
 •  39
   38
o
   37
ct:
 CM
O
oo
   36
                 L- 2000
   1400
1600
1800
2000
2200
              SCRUBBER GAS RATE (LB./HR.FT )
In general, the  scrubber performance  increases as the
gas rate is reduced  in  a scrubber.  Our tests have
shown that at very high gas  rates  2150  Ib/hr. ft.2,
this characteristic  no  longer holds.  It was  found that
the scrubber efficiency actually increased from under
37% per foot of  packing to 38.5% as the gas rate was
increased from 1880  Ib/hr. ft.2 to 2150 Ib/hr. ft.2.
It is theorized  that at a relatively  high gas flow, the
increased turbulence that results more  than compensates
for the reduced  contact time in the scrubber.

The significance of  this result is that high  turndown
rates on the gas side can be accomplished without loss
in scrubber efficiency.  This is of considerable
advantage for fluctuating boiler loads.
                         619

-------
                       PERCENT S02 REMOVAL
                       PER  FOOT OF PACKING
LENGTH  (FT. )012345678910

S02 CONC.    1000  600   360   216   130   78    47    28    16    10    6

% REMOVAL    0   40   64   78   87   92    95    97    98    99  99.4
SINCE THE  FORSEEABLE REGULATIONS WILL REQUIRE LESS THAN  90% REMOVAL
THE PRELIMINARY ESTIMATE OF LENGTH IS FIVE FEET.
       Our conclusions  are that the packing efficiency will
       be approximately 40% / ft. of packing.  In  the light
       of forseeable  emission regulations, a 90% S02 reduction
       efficiency will  be adequate.  This will require that a
       5 ft. bed of packing be used.  It is interesting  to
       observe that if  99% efficiency is required,  the length
       of the scrubber  packing must be doubled to  obtain this
       9% increase.
                               620

-------
            EFFECT OF CAUSTIC  CONCENTRATION
              ON S02 REMOVAL EFFICIENCY
o
SULFITE-BISULFITE REGION
   40 h
<
>
o
 CSJ
O
   20
   10,
                _L
                _L
                        _L
          4567
             Ph OF SCRUBBER OUTLET
                                               SULF1TE REGION
                                                10
    Another factor influencing the scrubber  operation is
    control of caustic in the system.  Chemical  conversion
    of the caustic in the scrubber takes place in  the
    following manner:
    1st;    NaOH + SO,
      	             4.
           pH   11.5
                    Na2S03 + H20
                    pH  9.0
2nd:
              SC>
     pH   9.0
                                   NaHSC>3
                                   pH  4.0
   The reaction  can  be  controlled at any point merely by
   supplying caustic to maintain a selected scrubber water
   outlet pH.
                           621

-------
As can be seen, there is no loss of scrubbing efficiency
until the scrubber outlet pH falls below pH = 6.0,
which is greater than 50% into the bisulfite region.

There is considerable advantage in scrubbing into the
bisulfite region because of the limitations in producing
a high caustic concentration in the regeneration system.
If the system operates 50% into the bisulfite region,
then the caustic requirements become 1.5 moles NaOH/mole S
instead of 2.0 moles NaOH/mole S on a sodium sulfite cycle.
This reduction in caustic requirement to the scrubber can
be accomplished without loss in scrubber efficiency.  The
advantage of such operation will become more apparent
after a discussion of regeneration chemistry.
                        622

-------
Regeneration

Complete regeneration of the spent scrubbing liquor
requires that the captured sulfur be precipitated
from solution as calcium sulfite and calcium sulfate,
These reactions are shown in the following equations:
                     Ca(OH)2  -»•  2NaOH  +  CaSC>3  ,
     2)  Na2 S04  +  Ca(OH)2  -»•  2NaOH  +  CaS04  ,
It has been found that regeneration of sulfites
-(equation 1) presents no serious problems.  Sulfites
are formed by the absorption of S0~.  The relative ease
of sulfite regeneration results from the fact that the
solubility product  (Ksp) of calcium sulfite is very low
(6.25 x I0~8(gm moles) 2  /£2  @ 1150 F)t  when the
product of the dissolved calcium and sulfite ion concentration
exceeds the solubility product, calcium sulfite is
precipitated.

     3)  (ca ++)  x   (S03 =)  =  6.25 x 1CT8

If only sulfites needed to be removed from solution,
regeneration would be relatively simple.  A portion of
the sulfur absorbed in the scrubber appears as sodium
sulfate.   The percentage of sulfate formation may vary
from as low as 10% of the sulfur absorbed to as much as
30% depending on the amount of oxidation that takes place.
                        623

-------
Sulfates do not regenerate as readily as sulfites.
The reason for this difficulty is that the
solubility product of calcium sulfate is 3.7 x 10
times as great as calcium sulfite.  This high calcium
sulfate solubility shifts the reaction in favor of
sulfate over caustic formation.  In order to establish
an equilibrium favoring caustic formation, an excess
of sulfate ion must be present in solution.  This can
be done by keeping an excess of sodium sulfate in
solution.  This excess Na2SC>4 in solution is necessary
for the precipitation of CaSC>4 and formation of a
dilute caustic solution.
                           624

-------
Design of Regeneration System

To effectively operate the regeneration system, there
is a minimum sodium requirement needed to precipitate
sulfates from the system and form a suitable concentration
of caustic.  The minimum sodium required to remove a
gm mole of sulfur from the system can be summarized as
follows:

         1)  Free sodium (NaOH, Na2 003) required to
             react with absorbed sulfur (SCU / SOo)
             in scrubber.  This includes 2.0 moles
             Na/mole S03 plus,

             a)   2.0 moles Na/mole SO*, (sulfite cycle).
             b)   1.0 moles Na/mole SO- (bisulfite cycle).

         2)  Sodium sulfate required to reach a favorable
             solubility product of CaSO* in solution.
                          625

-------
                EQUILIBRIUM CAUSTIC FORMATION
               IN Ca(OH>2 - Na2$04 SOLUTIONS @ 120°F
o
t—
<
I—
z
LU
O
z
o
   .2
   .15
   .1
^£ .05
*—
to
3
<
    0
               I
                    I
                         I
                              I
                                    I
                                         1
               .4    .6    .8    1.0    1.2   1.4
                SODIUM CONCENTRATION (gm. moles/liter)
                                             1.6
1.8
Tests have been  conducted to determine  the  equilibrium
caustic formation in sodium sulfate - calcium  hydroxide
systems.  The  figure above shows the maximum amount of
caustic that can be formed at 120° F in regeneration
solutions as a function of sodium concentration.   As can
be seen, increased Na+ (indicating increased Na~S04)
concentrations decreases the Ca++ concentration  allowing
more OH~ to exist at equilibrium.  Beyond 1.0  molar
sodium the additional amount of OH~ that can be  formed
does not increase appreciably.

This equilibrium data shows that no more than  a  0.14 molar
caustic solution can be produced in 1.0 molar  sodium
solutions.

The figure does  show the limits of caustic  that  can be
formed in the  regeneration cycle.  It does  not indicate
the rate of caustic formation which is needed  for  practical
design of the  regeneration system.
                           626

-------
           EFFECT OF LIME AVAILABILITY ON THE RATE OF CAUSTIC FORMATION
                 IN 0.5 MOLAR Na2$04 SOLUTIONS @ 120°F
  .13
s
= .11
«
o
t .09
  ce
  I
  .07
  u
  2 .05
  t/>
  ID
  <
  O
                                        0.25
                                            Hydroxide
                                            Available in
                                            Lime gm.moles
                                            OH'/liter
                                        0.125
                                        0.07
                      6     8     10    12
                      MIXING TIME (MINUTES)
                                        14
16
An additional series of  tests were run  to  determine the
rate of  caustic formation in 0.5 molar  Na^SO,  (1.0 M Na+ )
solutions.
These  tests were run by adding a given  amount of
calcium hydroxode to 0.5 M Na-SO, solution  (a) 120° F
and drawing samples at  2 , 5 and 10 minute mixing intervals,
The samples were immediately filtered and titrated with
0.08 N HpSO..   In all,  three different  initial lime
concentrations were evaluated.
The figure above shows  that the rate of  lime conversion
can be maximized if lime  is fed at a concentration near  the
equilibrium hydroxide concentration  (0.14  M).   In the
0.125 M OH~ lime feed,  80% conversion of the  lime was
obtained  in 5.0 minutes and 0.1 M OH- was  achieved.  When
0.25 M OH - lime was added, 44% lime utilization was achieved
in 5.0 minutes and only 0.11 M OH - was  formed.
                         627

-------
These tests concluded that:
         1)  Ca(OH)2 should be fed at or near the
             equilibrium caustic concentration of
             0.14 gnu moles OH~/liter.

         2)  Lime mixing efficiency can be 80% in
             5.0 minutes.  Recycling of the sludge
             will have to be evaluated to determine
             if increased utilization can be obtained.

         3)  With 5.0 minute mixing, it is felt that
             0.1 molar NaOH can be produced for use
             in scrubbing.
                           628

-------
A main advantage of this process is that S02 can be
removed in the scrubber as a soluble salt.  This allows
us to use a low energy scrubber for the process.  Any
calcium carry over from the regeneration cycle will
result in deposition in the scrubber.  This carry over
must be avoided if the process is to retain its
advantages over lime slurry scrubbing.

Calcium plugging can occur from two sources -

1st - Carry over of suspended solids.

2nd - Precipitation of dissolved calcium.

Suspended solids can be handled by a gravimetric
settling system.  A more serious kind of plugging can
result from carry over of dissolved calcium into the
scrubber.  The dissolved calcium is at its solubility
product with respect to sulfite and sulfate.  Exposure
                        629

-------
to S02 an^ SO- in the scrubber will cause a small
amount of precipitation of CaSO_ and CaSO, in the
scrubber if free hydroxide is allowed to enter the
scrubber.

We have evaluated one means of eliminating
precipitation in the scrubber by using Na2COo make up
as a water softener prior to entry into the scrubber.
We were successful in reducing the calcium concentration
from 400 ppm to 250 ppm.  This, however, will not
completely avoid precipitation of sulfites in the scrubber.

Another means of eliminating precipitation in the scrubber
would be to operate the scrubber on a low pH loop.  The
scrubber recycle tank could then be used to precipitate
dissolved calcium leaving the recausticizing section.  To
operate under these conditions, a scrubber with excellent
gas absorption characteristics and high liquid to gas flow
rates would have to be used.  We are now evaluating the
possibility of using such a mode of operation.
Pilot Study of Regeneration System

It was found in the pilot plant that production of
0.1 molar NaOH is easily obtainable in 1.0 molar
sodium scrubbing solutions at 120° F.  During 130 hours
of pilot operation no difficulty was encountered in
recausticizing the spent scrubber solution.  It did
become apparent that three variables had a significant
influence on the concentration of caustic produced.

          1)  Mixing time
          2)  Degree of agitation
          3)  Rate of lime addition
                            630

-------
 The mixing section  of  the  solids contact reactor
 utilized  a slow 16  rpm paddle  for mixing.   It
 was found that  30 minutes  of mixing was required
 to obtain the desired  caustic  concentration.  It
 was also  found  that lime usage was only 70%.

 Simulated regeneration  tests using near stoichiometric
 lime  addition and high speed flash mixing showed that
 the lime  usage  was  increased beyond 80%.  Also, the mixing
 time  required to produce the same quality caustic was
 reduced from 30 minutes to 5 minutes.

 The amount  of lime  present in  the reaction  zone also was
 shown to  have an effect on the concentration of caustic
 produced  as was demonstrated in the laboratory tests.
 Although  higher amounts of lime in the reaction zone
 produced  higher caustic concentration, lower lime
 utilization was realized.

 To further discourage  the use of high lime  concentrations
 in the reaction zone,  it was found that the spent sludge
 was chemically  inactive even though it contained as much
 as 30% unreacted lime.  Therefore, high lime concentration
 in the reaction zone would only reduce lime usage because
 sludge recycle does not appear to improve lime usage.

 The final test conclusion was that near stoichiometric
 lime addition, together with high speed chemical mixing,
must be used to optimize lime usage.
                          631

-------
                    Cake Analysis
                  Na2S04         2.6 %
                  Na2CO3         1.13%
                  NaHCO3         0.73%
                  A12(S04)3      1.84%
                  Ca(OH)2        0.16%
                  CaS04.2H20    39.30%
                  CaS03.2H20     2.93%
                  CaC03          4.85%
                  H0           46.5 %
Sludge Settling & Filtering

The CaS03 - CaS04 precipitate settled very rapidly.
At a clarifier rise rate of 1 gpm/ft.  no turbidity
was noticed in the clarified liquid.  The thickened
sludge contained 20% solids which was taken to 50%
solids in the rotary drum vacuum filter.

An analysis of the filter cake was made.  This analysis
gives only an indication of the content of the cake and
cannot be considered representative.

A considerable period of time elapsed before the cake
was analyzed.  The degree of sulfite oxidation and
carbonation cannot be determined.  The high amount of
gypsum does indicate that high oxidation takes place in
the scrubber.  There is no explanation for the high
Al?(S04)3 content in the filter cake.  Silicon was
                           632

-------
also identified in significant quantities in the
sample indicating that aluminum may be present in the
form of aluminum silicate rather than sulfate.

This cake analysis also indicates better than 80%
lime usage where system material balances indicated
only 60 to 70% usage.

An evaluation of the settling of the precipitated
CaCCs in the softener was not made in the pilot plant.
Instead, an in-line filter was used to remove the
precipitated CaC03.
Livonia Scrubber

Concurrent with these tests being conducted, another
scrubber is being tested at the Chevrolet-Livonia
powerhouse.  This unit is a wet cyclone scrubber.  It
is primarily known for its ability to handle particulate
matter.  The unit has the same capacity as the packed
scrubber unit and operates with a multiple stage impingement
wash of the flue gas.  The pilot scrubber we are using
contains four stages with a capacity for two additional
stages.
                          633

-------
The scrubber is located on the exhaust of a 80,000 Ib/hr.
spreader stoker equipped with a multiclone collector.  The
boiler has an economizer so the scrubber inlet temperature
is about 350° F.

The gas is cooled and saturated in the first scrubbing
stage rather than using a pre-guench as in the packed
scrubber.  There is also no regeneration cycle in this
system since it was decided that all necessary information
for regeneration could be gained with the packed scrubber.
We mainly are interested in seeing whether this scrubber,
with marginal gas absorption characteristics, is capable of
removing sufficient amounts of SO-.

Sulfur dioxide removal tests were run on the 4 stage
scrubber.  Results showed an 87% SO- removal efficiency.
This efficiency was higher than expected.  The scrubber
pressure drop was 5.0" H-0 during the tests.  An additional
two scrubbing stages is now being installed and the scrubber
will be re-tested.  It is hoped that the additional two
stages will bring the overall SCU removal efficiency above 90%.
                          634

-------
                     BENEFITS
         1.  Greater than 90% sulfur dioixde
             removal efficiency.

         2.  Low energy scrubber (3" - 7" H^

         3.  Soluble scrubber (no plugging)

         4.  Inert by-product

         5.  Economically feasible
It is our conclusion that the caustic-lime regeneration
system presents a practical means for control of SO2-
Efficiencies in excess of 90% are easily obtained with
standard low energy scrubbing equipment.  The scrubber
is not in danger of plugging as in a lime system because
the lime is restricted to the regeneration tank.  The
sulfite cake formed can be readily disposed as sanitary
land fill.
                          635

-------
                 PROCESS ECONOMICS
    Installed Capital
          Costs                 $ 3.20/lb. steam/hr.
    Operating Costs             $ 3-4/ton of coal
An economic evaluation of SO- control methods was
                            t£t
made.  The results indicate that the caustic-lime

regeneration shows favorably over other possible

alternatives.  Fixed capital costs for the system
are estimated at about $3.20/lb. steam.  Operating
costs will be approximately $3.00 - $4.00/ton of coal,
                          636

-------
Another benefit of  the  system  is  that particulate
emissions now below code  in  the 0.2 - 0.3 grains/scf
range will be reduced by  70% +.

In our tests we have found no  appreciable formation of
nitrite or nitrate  ion  in water analysis.  We have
concluded that oxides of nitrogen are not being absorbed
in the system.

To sum up — the conclusions from our work are that we do
have a workable and  economic process for the control of
sulfur dioxide emissions from coal burned in industrial
sized power plants.  This should be of considerable benefit
to many industries  in the very near future, and it certainly
will be of great benefit to our environment.
                          637

-------


o
CO

E

11


O

E

+

-------
         SULFUR OXIDE CONTROL AT THE COPPER SMELTER
                              by

                      Ivor E. Campbell
                     Technical Director
            Smelter Control Research Association

                             and

                       James E. Foard
                   Senior Smelter Engineer
                    Metal Mining Divisions
SECOND INTERNATIONAL LIME/LIMESTONE WET SCRUBBING  SYMPOSIUM
                  New Orleans, November 8-12, 1971
                            639

-------
                 SULFUR OXIDE CONTROL AT THE COPPER SMELTER










     Although the basic chemistry of t-he principal reactions in various



sulfur dioxide control processes is the same regardless of the source



of sulfur dioxide, conditions prevailing at the copper smelter relevant



to sulfur dioxide removal processes differ in several significant respect



from those prevailing at other industrial sources of sulfur dioxide.



     This paper reviews the state of the art of sulfur dioxide control



as applicable to the copper smelter and the Smelter Control Research



Association1s-pilot-plant program.  Some of you may not be familiar



with the Smelter Control Research Association.  The Association was



founded and incorporated as a nonprofit association by the eight primary



copper producers of this country for the specific purpose of developing



processes for removal of sulfur oxides and particulates from smelter



stack gases and more especially for conducting pilot-scale investigation



of the removal of sulfur dioxides from reverberatory furnace off-gases.



The reasons for emphasizing the reverberatory gases will be discussed



later.



     The eight founding companies of the Smelter Control Research



Association are: American Smelting and Refining Company, The Anaconda



Company, Cities Service Company, Copper Range Company, Inspiration



Consolidated Copper Company, Kennecott Copper Corporation, Newmont



Mining Corporation, and Phelps Dodge Corporation.  These companies



account for all of the primary copper smelter production in the United



States.



     Results of the Association's experimental programs will be published



and will be available to the public.  It should be emphasized that the
                                   640

-------
programs of the Association supplement extensive in-house development



work being conducted by the individual copper companies.






     As background for our discussion of sulfur dioxide control as



applied to copper smelters, it will be helpful to first describe, in



general terms, a typical copper smelter operation and to note the opera-



tions in which sulfur dioxide is generated and the conditions under which




it is formed.





     To simplify the description reference will be made principally to



the flow of the copper-bearing materials and of the sulfur.






     Most copper ores today are much too low in copper content to be



processed directly.  Therefore, it is necessary to separate the copper-



bearing minerals from other materials in the ore in which they occur.



After crushing and grinding of the ore, the copper-bearing minerals are



concentrated by a flotation process.  The composition of the concentrate



varies from smelter to smelter, but the copper content is usually in the



range of 10 to 30 % and the concentrate contains, on an average, a pound



of sulfur or more for each pound of copper since it is produced from



sulfide ores.  The concentrate also contains a number of trace consti-



tuents, some of which are known to affect certain of the sulfur dioxide



control processes, particularly those involving the use of catalysts.






     Smelters throughout the world rely on oxidation processes to remove



the sulfur and this oxidation generates sulfur oxides.  A typical smelter



produces approximately two pounds of sulfur dioxide for each pound of



copper, since a pound of sulfur forms two pounds of sulfur dioxide.
                                 641

-------
     Foreign smelters employ flash, electric, and blast furnaces, as



well as reverberatory furnaces; but our domestic practice, which has



been keyed to the available raw materials, is based on the use of



reverberatories.  At some smelters, the concentrate goes through a



roasting operation before being fed to the reverberatory furnace.



The roasters, typically fluid bed reactors, generate a relatively



uniform, concentrated stream of sulfur dioxide, i.e., 10-15 % sulfur



dioxide.  In most cases, the concentrate is fed directly to the



reverberatory furnace where it is fused in an oxidizing atmosphere,



forming a slag and a molten mixture of copper and iron sulfides known



as matte.  Approximately one-third of the sulfur in the concentrate



is eliminated at the reverberatory furnace as gaseous sulfur dioxide



which is discharged to the stacks.  The matte is discharged from the



reverberatory furnace into ladles in which it is transported to con-



verters o  In the converters, the matte undergoes further oxidation,



additional slag is termed, and the balance of sulfur is discharged



as sulfur dioxide.  The converter product is approximately 97 to 98 %



pure, and is known as blister copper.  The blister copper is further



refined to produce the high purity product required for industrial



applications.



     The reverberatory furnaces are large, rectangular, refractory



chambers, usually in excess of 100 feet long and 30 feet wide.  They



are fired with gas, oil or coal.  The reverberatory furnace off-gases



contain, on  an average, around 1 % sulfur dioxide, along with a variety



of trace constituents as noted earlier.   The gases, on leaving the
                                   642

-------
furnace, are usually in excess of 2000°F, but a feed stream from a



typical fluesystem would probably be on the order of 400-500°F.



     The converters are cylindrical, refractory-lined vessels, usually



30 feet long and 13 feet in diameter, although a few larger units are



now in use.  The operation of the converter is cyclic, and the con-



centration of the sulfur dioxide in the exhaust gases varies widely



during the operation.  The average sulfur dioxide content of the ex-



haust gases during the blowing operation is usually in the range of



4 to 6 %.  This variability in sulfur dioxide concentration is a



definite handicap in the sulfur dioxide control process.  Smelters



operating sulfuric acid plants try to program their converter opera-



tion, coordinating the cycles of converters in order to provide a more



uniform flow of sulfur dioxide to the acid plant.  This objective is



difficult to attain and is a significant constraint on the smelter



operations and does not, in any event, produce a highly uniform flow



to the acid plant.



     It will be seen, then, that two distinct levels of sulfur dioxide



concentration occur at the copper smelter.  The converters and roasters



produce off-gases relatively concentrated in sulfur dioxide,  averaging,



in the case of the converters,  around 4 - 6 % sulfur dioxide.   The



reverberatory furnace produces a much more dilute gas, ranging from a



fraction of 1 % to somewhat in excess of 1 % sulfur dioxide.






     Technology is commercially available for converting the nulfur



dioxide in the more concentrated streams, i.e.,  from the converters



and roasters,  to sulfuric acid.  Seven acid plants are in operation



in the copper industry today,  and additional plants are either under
                                  643

-------
construction or in the planning stage.  There is, however, no commer-
cially proved technology available  for the removal of sulfur dioxide
from the more dilute gases produced in the reverberatory furnace.

     The Smelter Control Research Association program is, accordingly,
directed as noted earlier, specifically to the development of processes
for removal of sulfur dioxide from reverberatory furnace off-gases.

     It should be noted that although reverberatory gases are more
dilute than converter gases, they are on an average at least 3 to 5
times as concentrated in sulfur dioxide as typical utility stack gases.
In the wet-lime and wet-limestone scrubbing processes, this higher
concentration represents a significant difference, particularly with
respect to control of scale formation in the scrubbing system.

     It is fortunate that there are strong similarities in the problems
involved in removal of sulfur dioxide at the three concentration levels
we have referred to, i.e., the 0.1 to 0.3 % and lower level prevailing
generally at utility and industrial boilers,  the 1 % level at the
reverberatory -furnaces, and the 4 to 6 % and higher levels found at
the converter and the roaster.  However, sulfur dioxide removal systems
for each application present distinct and separate problems.  Verifica-
tion of any process under consideration must, therefore, be carried out
under conditions valid for the proposed application.  Therefore, pilot-
scale tests on actual smelter gas are considered essential before the
very substantial commitment that would inevitably be involved can be
made to a given control process.  For this reason the Association's
pilot-plant tests, as we noted earlier, are to be conducted on actual
smelter gas.
                                  644

-------
     In establishing priority to be assigned to investigation of various



possible alternatives for removal of sulfur dioxide from low-strength



gases  at the smelter, the stringent time schedule being imposed on the



industry for sulfur oxide control dictated that priority be assigned to



alternatives deemed closest to commercial availability, even though



other alternatives might provide a more satisfactory long-range solution.






     The time schedule and the economic facts of life dictated that the



control system be compatible with existing facilities, i.e., that the



system be what is commonly referred to as an "add on" system.  After



reviewing more than 100 processes or variations of processes, the Smelter



Control Research Association selected wet-limestone scrubbing for its



initial pilot-plant study on the basis that this process, which has



been investigated extensively by the Environmental Protection Agency



and by the utility industry, was closest to commercial availability.






     A pilot plant has been constructed and is in operation at the McGill,



Nevada smelter of the Kennecott Copper Corporation.  The pilot plant



is rated at 4000 SCFM and is of sufficient size to permit extrapola-



tion to full commercial scale.  (The extrapolation would be approximately



20 to 1.)



     The pilot plant has been designed with a view to optimum versatility



and will be used in testing wet-lime and sodium sulfite scrubbing as



well as limestone scrubbing.  In addition, lime regeneration of the



scrubbing lig_uid in the sodium sulfite system is to be evaluated.  The



tests on the sodium sulfite system will, hopefully, with information



available from other operations on sulfur dioxide desorption permit



evaluation of the sodium sulfite absorption-desorption process vis-a-vis




                                   645

-------
the throw-away process.






     The plant employs a Turbulent Contact Absorber in series with a



venturi.  Limestone-water slurry flows countercurrent  to the gas



stream from the Turbulent-Contact Absorber to the venturi.  A bleed



stream from the venturi is sent to a centrifuge where the reaction



products are separated and discharged, the liquid being returned to



make-up tanks for recycle.  Operation with lime will be essentially



the same as with limestone.






      When sodium sulfite is used as the absorption medium, the bleed



stream from the venturi will be sent to causticizing tanks where



calcium sulfite will be precipitated on addition of lime.  The product



will then be sent to the centrifuge for separation of the calcium



sulfite which will be discharged as a slurry, as in the limestone



scrubbing system.  The liquid from the centrifuge will be returned to a



holding tank where make-up caustic will be added to provide a fluid



suitable for recycling.





      The pilot plant has been in operation with limestone approximately



two months, on a 24-hour per day, 5-day per week basis.  In initial



operation, the scrubbing liquid was not recycled.  Once-through



operation was employed to minimize plugging of equipment with reaction



products.   This was done to facilitate operations while the operators



were familiarizing themselves with the system.  The plant is now a



closed-loop operation.
                                    646

-------
     It is too early to present any final conclusions  with respect to



;he effectiveness of the system in smelter gas  control.   Results to



3ate may be summarized as follows:






      The average sulfur dioxide recoveries have  been  approximately



'0  per cent from 0.7 per cent sulfur dioxide streams and approximately



>5  per cent from 1.0 per cent streams.






      A high calcium limestone has been  used in all tests.   Bench-scale



:ests of a number of high calcium limestones indicated that the  lime-



 tones tested were equally reactive.   Bench-scale tests  of additional



.imestones are planned.






      Limestone usage has been varied from 1.2  times stoichiometric



.o  2.0 times stoichiometric.






      Both 90% minus 200 mesh stone and  95% minus 325  mesh stone have been



 sed in the pilot plant.   Any difference in the reaction rates with the



 wo grinds is apparently masked by other process  variables.   Although



 ne might anticipate increased reactivity with  the finer grind,  no



 ignificant improvement in sulfur dioxide recoveries has been noted to



 ate with the use of the finer grind.






      The sulfur dioxide recoveries do appear to  decrease as the



 oncentration of the sulfur dioxide as the feed stream  increases.






      In this regard,  it should be noted that the concentrations of



 ilfur dioxide in the feed stream varies and extended  runs  are required



 i  order to obtain meaningful data.  Over shorter periods variations in



 he gas concentration may mask the effect of other variables.
                                     647

-------
      Our current efforts are directed to establishing the factors



controlling sulfur dioxide removal efficiencies in order to optimize



sulfur dioxide recovery.  Development of techniques for effective



control of scale formation is also a major objective.






      We would like to emphasize again that there are significant



differences in the problems involved in sulfur dioxide control at the



smelters and at other industrial sources of sulfur dioxide.






      The gas flows are generally lower than at the utilities.  A



typical commercial unit on a reverberatory furnace would handle



approximately 80,000 SCFM.






      The sulfur dioxide concentration of reverberatory gas is much



higher, averaging around 10,000 ppm, i.e., 1%, and is much less uniform



than that of industrial boilers.






      The total dust loading at the smelter is lower than that at many



utility boilers because of the lower gas volumes, but the concentration



is somewhat higher than at typical coal-fired boilers.  Trace constituents



of the smelter gas stream will not only be different than at the



utilities but will vary from smelter to smelter.  These constituents



may have little or no effect on some control processes but are known to



affect others, and their possible effect must be determined in actual



pilot-plant tests.






      The percentage of o::ygen in reverberatory gas streams is highly



variable, and is normally higher than that in exhaust gases from



utility boilers.
                                     648

-------
      The gas flow from the converter is cyclic.  VThile the gas flow

from the reverberatory is much more uniform, it, too, is subject to

variation.

     From the standpoint of disposal of waste products, the smelters

have, in most cases, for the moment at least, available sites to establish

sludge ponds, and this is not as serious a problem as it would be in

more heavily populated areas.  However, the problems involved in handling

the very substantial quantities of raw materials consumed in the control

process operation and the tonnage of waste products that must be

disposed of do present a substantial materials handling problem.

      YJhile the situations at the smelter, with respect to disposal of

the throw-away product, may be more favorable than that at the utilities,

the situation with respect to disposal of potentially marketable

products such as sulfuric acid and sulfur is much the reverse, since

the smelters, in many cases, are remote from potential markets.

     As noted earlier, the Smelter Control Research Association

supplements extensive in-house developments by the individual Member

Companies.  A few of these may be briefly mentioned:

      1.  A 1,000 SCFM ammonia scrubbing plant was operated for about

          eight months at a copper smelter to determine whether the

          process, which had been in successful operation for a number

          of years at a lead-zinc smelter was, in fact, adaptable to

          copper reverberatory gas.

          The sulfur dioxide recovery achieved in the pilot plant was

          in accord with recoveries from available theoretical data as

          well as from prior practice.  The problems encountered were:

          (1) aggravated corrosion; (2) high sensitivity to irregular

          draft conditions;  (3) a stack discharge of highly visible

          particulate matter which required additional collection
                                    649

-------
    devices;  and (4)  requirements for high volumes of low



    temperature water,  an extremely important consideration in



    many smelter locations.   Perhaps the most compelling factor,



    however,  is that the operation of a full-scale plant would



    produce large quantities of soluble ammonium compounds which



    would create disposal and pollution problems of their own



    unless they had utility as usable or salable products.



2.  A 400 SCFM pilot plant utilizing sodium citrate as the



    absorbent for sulfur dioxide was operated for about six



    months at another copper smelter.  The process calls for the



    sulfur dioxide loaded absorbent solution to be treated with



    hydrogen sulfide to form elemental sulfur, part of which



    would be reacted with methane to make up the hydrogen sulfide



    requirement.  The pilot operation, which included only the



    absorption and regeneration steps, was beset with mechanical



    problems, and solution losses were incurred which prevented



    reliable evaluation of citrate consumption, an obviously



    important parameter.  Recovery of sulfur dioxide from the



    reverberatory gas stream ranged between 90 and 99 per cent.



    The .hydrogen sulfide generation step in the process was not



    investigated.



3.  Another copper smelter is currently installing an industrially



    sized operating module to explore the commercial feasibility



    of absorbing the sulfur dioxide in reverberatory gas with



    dimethylanaline  (DMA), and springing the sulfur dioxide in



    highly concentrated form to feed an acid plant, or perhaps a



    sulfur plant.  DMA absorption has been successfully demonstrated
                             650

-------
          in small lead smelter applications treating gases above



          3.5 per cent sulfur dioxide.  The unit under construction



          will be treating a much leaner gas and is expected to be



          on-line by mid-1972.



      4.  A $1.7 million sulfur reduction pilot plant designed to



          produce 8 to 20 tons of elemental sulfur per day has been



          recently placed in operation as a joint venture of two copper



          smelting companies.  Its main objective will be to perfect a



          suitable process for reacting sulfur dioxide with reformed



          natural gas to form elemental sulfur.  The pilot plant will



          operate on gases ranging from 12 to 100 per cent sulfur



          dioxide.



      5.  A number of smelters have conducted bench-scale tests or



          even operated small pilot plants to e;:plore the potential



          of absorbing sulfur dioxide in naturally occurring or



          industrially produced materials readily available to them.



          Such waste products as fly  ash concentrate tailing pulps,



          furnace slags, and others have been tried.  In general, the



          sulfur dioxide loading capacity of these sorbents is not



          high', and their usage would require prohibitatively large



          tonnages unless they were effectively regenerated, and this



          step has not been successfully demonstrated.






      In addition to their work in the five areas mentioned, individual



smelters have been exploring, in pilot plants already completed or



being planned, changes in the smelting process itself by which pollution



abatement may be furthered.  This test work relates particularly to the



development of methods of continuous smelting, which would greatly



.enhance the efficiency of appurtenant sulfur recovery processes.





                                   651

-------
      In addition, the potential of electric smelting is being  explored



by several companies and a 7500 KVA electric furnace is presently being



installed at a large chemical plant with a small copper smelting annex.






      It is evident that the copper smelting companies are on the move



in their efforts to control sulfur o::ide eraissiona.
                                   652

-------
        PROTOTYPE AND FULL SCALE TESTS


            H.W. Elder, Chairman


               Participants:
James Jonakin and James Martin
E.G. McKinney and A.F. Little
M. Epstein, F. Princiotta, R.M. Sherwin, L. Szeibert,  I.A. Rabei
R.M. Sherwin, I.A. Raben, and P.P. Anas
J.D. McKenna and R.S. Atkins
Gerhard Hausberg
J.J. O'Donnell and A.G. Sliger
Tsukumo Uno, Masumi Atsukawa, and Kenzo Muramatsu
Lyman K. Mundth
J.H. McCarthy and J.J. Roosen
Robert R. Padron and Kenneth C. O'Brien
J.A. Noer and A.E. Swanson
J.W. James
D.T. McPhee
J.F. McLaughlin, Jr.
D.C. Gifford
H.P. Willett and I.S. Shah
                        653

-------
                     PROTOTYPE AND  PULL  SCALED-TESTS
                         (PARTS I, II, AND III)

      Second International Lime/Limestone Wet Scrubbing Symposium
                         New Orleans,  Louisiana
                           November 8-12, 1971

                         H.W.  Elder,  Chairman

                                SUMMARY
          The increasing interest in preserving a quality environment
in this country and abroad has caused an accelerating movement toward
enactment of legislation to limit discharge of undesirable materials to
the atmosphere.  The existing and pending laws have  started a somewhat
frantic search for methods to control emission of materials classified
as pollutants.  The participants and attendees at the Second International
Lime/Limestone Wet Scrubbing Symposium assembled for a critical  analysis
of one of the more promising methods for removal of  S02 from stack  gases.

          The session of the symposium on prototype  and full-scale  appli-
cation of the process covered topics from plans for  full-scale installa-
tions to availability of raw materials.   The main theme that evolved as
the session progressed was the concern over reliability of the scrubber
system as compared with the power plant.  The problem of waste disposal
was a popular subject and, as might be expected, economics was a con-
troversial issue.

          Participants in the prototype and  full-scale  session  are
listed below.

        	Name	         	Company	

        Masumi Atsukawa          Mitsubishi Heavy Industries, Ltd.
        A. 0. Blatter            Union Electric Company
        D. C. Gifford            Commonwealth Edison Company
        J. W. James              Ontario Hydro
        Jim Jonakin              Combustion Engineering, Inc.
        Ulrich Kleeberg          Gottfried Bischoff  KG
        L. K. Mundth             Arizona Public Service
        J. H. McCarthy           The Detroit Edison  Company
        J. D. McKenna            Cottrell Environmental Systems
        B. G. McKinney           Tennessee Valley Authority
        D. T. McPhee             Kansas City Power and Light
        J. A. Noer               Northern States Power Company
        J. J. O'Donnell          M. W. Kellogg Company
        R. R. Padron             Key West Utility
        F. T. Princiotta         Environmental Protection Agency
        R. M. Sherwin            Bechtel Corporation
        H. P. Willett            Chemical Construction Corporation
                                 654

-------
          Use of lime or limestone scrubbing for control of S02 emission
in stack gas is of interest because it is relatively simple compared with
processes that produce salable products, both fly ash and S02 may be
removed, limestone is available in most parts of the country, earlier
work has shown the process is feasible, and the complexities of marketing
sulfur products are avoided.  A better understanding is needed of factors
that influence chemical scaling, solids deposits, slurry settling rates
or mechanical dewatering, erosion, and mist carryover.

          Limestone can be added to the system in two ways—either addition
in the boiler where the carbonate is decomposed or addition directly into
the scrubber circuit.  The calcined material is more reactive and should
result in a lower raw material requirement.  However, there is evidence
that scaling is more likely to occur during closed-loop operation with lime
than with limestone.  Also, occurrence of calcium-containing deposits in the
convection pass was reported for operation during limestone injection.  It
is likely that relatively wide tube spacing would be required to prevent
plugging.  The risk of tube fouling could be avoided by addition of cal-
cined lime directly in the scrubber, but the cost of lime is high and the
scaling tendencies are a deterrent.  However, the process apparently has
been developed successfully in Japan and a 20-mw prototype unit is being
installed on an oil-fired boiler.

          The major interest appears to be in use of limestone in the
scrubber.  Plans for equipping over 3000 mw with tail end processes were
discussed.  The plants range in size from 37-580 mw and scheduled startups
vary from January 1972 to May 1977.  It was interesting to note that all
stressed the importance of reliability and many have included scrubber
bypass systems, redundant installed equipment, and modular construction.
Methods of solids disposal varied from use of only settling ponds to barge
transport of slurry to a remote location followed by mechanical dewatering
and return of the clarified liquor.  The concensus was that waste disposal
is likely to be a major cost item.  Stabilization of sludge by the addi-
tion of fly ash was reported.  Materials of construction varied but in all
cases the scrubbers are lined with polyester glass or rubber.

          One presentation dealt mainly with a fairly detailed cost
analysis for a tail end system.  The capital requirement for a l80-mw
retrofit unit amounted to $^9/kw and the operating cost was estimated at
about $^/ton of coal.  Others reported estimated capital costs ranging
from about $20-$6o/kw.

          In spite of the increasing commitment to full-scale application
of the lime/limestone scrubbing processes many questions remain unanswered.
The EPA-funded, prototype-scale demonstration of the process at TVA's
Shawnee Steam Plant is designed to fully characterize the process with
limestone added in the scrubber circuit, calcined lime added in the scrub-
ber, and limestone injected into the boiler followed by wet scrubbing.
                                  655

-------
The system was designed by Bechtel Corporation.   Three parallel scrubbers
each designed to handle 30,000 acfm of gas will be operated simultaneously
with closed-loop operation.  Each train is equipped with a thickener and
the underflow from a single unit can be further dewatered in a centrifuge
or filter.  A separate settling pond is provided for the test facility.

          The test program will begin with a break-in phase to establish
system integrity followed by statistically designed screening experiments
to identify the effects of variables and will be concluded by long-term
operation at optimum conditions to verify reliability.  The system is
extremely flexible and well instrumented to obtain maximum information.

          It was reported that limestone is available in most parts of
the country in quantities that are sufficient for the installed generating
capacity.  Moreover, because of the abundant reserves, the price is likely
to be relatively stable for the foreseeable future.
                                 656

-------
                 APPLICATIONS OF THE C-E

              AIR POLLUTION CONTROL SYSTEMS
                      JAMES JONAKIN

                       JAMES MARTIN

                  C-E Combustion Division
                   Windsor, Connecticut
                       Presented at

SECOND INTERNATIONAL LIME/LIMESTONE WET SCRUBBING SYMPOSIUM


                   November  8-12, 1971

     Sheraton-Charles Hotel, New Orleans, Louisiana
                            657

-------
                    APPLICATIONS OF THE C-E

                 AIR POLLUTION CONTROL SYSTEM

                         James Jonakin

                          James Martin
Much has been said in the technical press on the subject of air
pollution from power plant boilers and the C-E systems for meeting
this problem.  A brief review will serve to introduce the full-
scale commercial activities C-E has been engaged in.

Requirements for electric power in the U. S. will increase from
1-5 trillion kwh in 1970 to over 3 trillion kwh by 1980.  To meet
this demand for power, generation capacity must increase from the
present 3^0 million kilowatts to about 710 million kilowatts and
all available fuel sources - nuclear, coal, oil, and others -
must be used.  By 1980, nuclear energy is expected to supply about
21$ of the total installed capacity while a substantial portion of
the remaining capacity will be supplied by coal.  In the face of
this continuing demand for electric energy from fossil fuels, utili-
ties and equipment manufacturers must work toward decreasing pollution
while at the same time keeping the cost of the end product, electricity,
at a reasonable level.

POLLUTION EMISSIONS

As coal is burned in the furnace of a steam generating unit, a number
of combustion products are formed, of which 98.7$ are non-pollutants.
The primary concern of the power industry is particulate matter, sulfur
oxides, and nitrogen oxides.  Particulate matter accounts for 80$ of
the polluting effluent, while sulfur oxides and nitrogen oxides make
up the remaining 20$.  These three pollutants represent only 1.3$ of
the total stack effluent, a rather small portion but a very signifi-
cant quantity.

When all combustion processes in the U. S. are taken into account
and the figures are "based on yearly emissions of pollutants into
the atmosphere, then the total is of concern, as shown in table 1.
These quantities will increase in proportion to the amounts of
fossil fuel consumed unless additional control methods are developed
and applied.  Pollution from carbon monoxide and hydrocarbons are
almost two-thirds the total; however, their source is primarily from
automobiles, trucks, airplanes, etc. and will not be discussed in
this presentation.  Note that the tons of particulate matter emitted
is the smallest of the totals listed.  Although a "heavy" potential
polluter accounting for 80$ of the polluting effluent from a steam
generator, present technology is generally adequate to reduce
emission of particulate matter to the level shown.  Electrostatic
precipitators are available with a removal efficiency of better than
95$-
                                    658

-------
CONTROL OF SULFUR OXIDES

The control of sulfur oxides emissions, however, presents some
difficult problems.  Trace quantities of sulfur oxides have to
be removed from millions of cubic feet of gas on a continuing
basis.  As an example, over 2 million ft  of gas is discharged
from the stack of a 700,000 kw steam generating unit each minute.
The gas resulting from the combustion of a coal containing 3%
sulfur contains only 0.2/J sulfur oxide.

Sulfur oxides can be controlled by two general methods:

1.  by burning low sulfur fuels which occur naturally, or
    are produced by removing the sulfur from the fuels
    before combustion, or

2.  by removing sulfur oxides from flue gases after com-
    bustion.

Natural gas contains no sulfur, but the supply of this fuel in the
U. S. is limited and the proven resources are shrinking rapidly.
The use of this premium fuel by utilities is decreasing and cannot
be considered a general solution to the pollution problem.

Fuel oil produced from crude oils originating in North Africa and
the Far East contains less than 1% sulfur and many utilities on our
east coast are now burning this fuel.  However, drilling and tanker
capacities are somewhat limited and its cost has increased as the
demand has increased.  In addition, the rapidly changing political
situation in these countries jeopardizes this source of oil.  The
high bid fur Central America and Middle East oils are more economical
and the supply is more stable.

The coal used by electric utilities in the USA contains an average
of 2.2% sulfur but ranges from <~L% to >5$-  In the eastern and mid-
western states where about 80% of the total power is produced, most
coals contain >3% S-  There are large supplies of low sulfur coal
in the Western U. S., but for the most part, it ic located far
from major population centers or market centers, therefore, it has
not been mined commercially on a very large scale.

STACK GAS CLEANUP

Of the two methods of controlling SOx emissions, the present emphasis
of industry is on stack gas cleanup rather than removing sulfur from
the fuel before conbustion.  There are numerous schemes for removal
of SO  and particulate matter from stack gases.  They have been des-
cribed in some detail in various publications.

Wet scrubbing has received the most attention because of its low
cost and simplicity.  Combustion Engineering is offering two wet
bcrubbing systems:  "furnace injection" and "tail-end".  These
systems can be termed "throw away" systems since the process is not
dependent on the recovery of a salable chemical by-product.
                              659

-------
SULFUR RECOVERY AND WASTE DISPOSAL

The descriptive term "throw away" implies indiscriminate rejection
of by-products, and carries the stigma of conversion of one form
of pollution to another.  C-E is not blind to this situation and
continues to investigate all aspects of the problem.

A cost comparison of the C-E wet scrubbing process vs sulfur
recovery processes is shown in table 2,     In the U. S., there-
fore, the present systems for stack gas cleanup must be used at
least as an interim measure.  At the same time C-E has established
research programs for the investigation of by-product utilization
and waste disposal.  This subject is covered in another C-E paper
at the conference, entitled, "Research and Development in Wet
Scrubber Systems".

FURNACE INJECTION SYSTEM

The furnace injection process involves injecting an additive which
contains a large percentage of calcium or magnesium, such as lime-
stone or dolomite, into the furnace of a steam generating unit.
The schematic arrangement has been shown in many papers.  Here,
the additive is calcined, producing a more reactive compound
which then chemically reacts in the dry state with some of the
SO  and SO  in the combustion gases to form compounds of calcium
ana magnesium.  About 20 to 30 percent of the sulfur oxides,
including all of the SO  , is removed in the dry state.

Next, the flue gas containing unreacted SO  and calcined additive
flows to the wet scrubber, a large tank-like structure containing
water sprays and a bed containing glass spheres.  In the scrubber,
the remaining calcined additive chemically reacts with the water
and the remaining SO .  At the same time, the fly ash is scrubbed
from the gas.  The solution containing the coirroounds formed by
these reactions and fly ash drains out the bottom of the scrubber
to a clarifier or pond where the solids settle.  Clarified water
is then available for recirculation.  The cleansed flue gas
passes through a mist eliminator for removal of moisture and
then is reheated for fan protection and reduction of stack plume.

The development of C-E's air pollution control system started in
196^ with the construction of a small pilot facility in our labora-
tories.  This was followed by a second pilot application on a unit
at the Detroit Edison Company in 1966 and 196? •

APCS CONTRACTS - FURNACE INJECTION SYSTEM

1.  Union Electric - Meramec No. 2 - lUO MW

2.  Kansas Power and Light - Lawrence No. h - 125 MW
                               660

-------
    Installation of the APCS systems for these tvo existing
    units was completed in late 1968.  Operation of these
    demonstration units has revealed problems , and modifica-
    tions have been made to the scrubbers, water control
    piping, and the additive injection procedure.  Our
    Research and Product Development Department has been
    actively involved in the solutions to these problems,
    which are described in detail in their paper on wet
    scrubber systems.  Data obtained to date show that
    SO  emission when burning coal with 3-5% sulfur has
    been reduced to the equivalent of burning a 1% sulfur
    coal, and that at least 99% of the particulate matter
    is removed.

    These units are pictured in Figures  1  and  2  res-
    pectively.

3.  Kansas Power and Light - -Lawrence No. 5 - ^20 I'M

    The contract for this unit, obtained in 1968, is for a
    new C-E boiler.  This unit is just starting up and has
    fired natural gas during startup.  Coal will be fired
    beginning about mid-November 1971 and the APCS will be
    tested.

    This unit is pictured in Figure  3.

k.  Kansas City Power and Light - Hawthorne No. 3 - 100 MW

5.  Kansas City Power and Light - Hawthorne No. h - 100 MW

    The contracts for these two units were obtained in 1970.
    They are presently being installed and are scheduled for
    operation next year.

TAIL-END SYSTEM

The other C-E air pollution control system, designated "tail-end", is
similar to the furnace injection method except a slurry of pulverized
limestone or slaked lime is used to scrub the flue gases.  No furnace
injection of additive is required.  Results obtained in prototype
units show that oO% or more of the SO  can be removed with a particulate
matter removal efficiency of 99% or better.

The schematic arrangement for the contracts with this system is
approximately similar to that for the KDL prototype unit, shown
in the paper on wet scrubber systems.  The system is designed for
a variety of situations in which furnace injection can be eliminated.
Some of these situations  are:

a.   Providing  flexibility in the  selection of the most
     suitable additives.
                                661

-------
b.  Supplementing existing electrostatic precipitators  on
    older units.

c.  Replacing precipitators for the removal of particulates
    as well as SO , as in the case of one contract described
    below.

APCS CONTRACTS - TAIL END SYSTEM

6.  Louisville Gas and Electric - Paddy's Run No.  6 - 70 MW

    The contract for this unit, an existing boiler built by  a
    competitor, was obtained in 1970.  Of interest is the  fact
    that this contract will use carbide sludge, a waste material,
    as the additive for reducing SO  emissions.  The utility
    will obtain the carbide sludge from a local industrial firm
    which has large quantities of this waste product produced
    in the manufacture of acetylene.  In other words, a "throw
    away" product from another industry is recycled to improve
    the environment.

    To ensure that the carbide sludge would be an adequate
    additive, tests were conducted in our new laboratory
    prototype unit.  This 12,500 cfm unit, the largest
    laboratory test unit of its kind, has been running
    almost constantly since its startup in early 1970.   A
    tremendous amount of research data has been collected
    which has helped us in adapting systems to the specific
    requirements of customers.  Most important, however, it
    has helped us to accelerate progress in the development
    of a commercially operable APCS.

7-  Union Electric - Meramec No. 1 - 125 MW

    This contract was obtained in 1971-  The boiler is a
    duplicate of Meramec No. 2, for which a furnace injec-
    tion system was supplied, in item 1 above.

8.  Northern States Power - Sherburne No. 1 - 700 MW

    This contract, also obtained in 1971> is significant
    for two reasons:

    a.  It is the largest contract so far, for either
        of the two C-E systems.  The boiler is a 1970
        contract, a C-E controlled circulation unit
        burning a sub-bituminous coal.

    b.  The coal has a sulfur content of less than 1%
        and experience has shown that this fact is detri-
        mental to the efficiency of electrostatic preci-
        pitators .  It is for this reason that the wet
        scrubber system was selected and the primary
        object is particulate removal, although about
        half the sulfur in the fuel will be removed by
        the scrubbers.  This system and the reason for
        its selection will be described in a paper to be
        given by Mr. Noer.

                                 662

-------
APCS CONTRACTS - FIELD EXPERIENCE

The initial operation of the air pollution control system has
previously "been discussed.  Subsequent operation has revealed
additional problems and also has allowed C-E to further refine
the modifications made to the original APCS design.  The fol-
lowing is a summary of the modifications made to system com-
ponents .

Sootblowers

The installation of half-track sootblowers on the scrubber inlet in
1969 has eliminated the problem of massive inlet deposits.  These
sootblowers have been in service at Union Electric Meramec #2 and
Kansas Power and Light Company Lawrence ffh for over 7000 hours of
coal operation.  Only one serious buildup of deposits has occurred
during that time.  The oscillating motor on one of the inlet blowers
failed and was not discovered immediately.  In approximately 2U
hours a large deposit developed, which had to be removed manually.

The sootblowers have been operating at normal blowing pressures and
have not required any abnormal maintenance.   All future C-E - APCS
are being supplied with inlet sootblowers similar to those on the
existing unity,

Ladder Vanes

The installation of ladder vanes on the units at KP&L Co. , Lawrence #U
and U.E. Co., Meramec #2 helped solve the problem of gas distribution
in the scrubber but in the final analysis could not successfully be
kept clean.  The ladder vanes were removed in 1970.

Further work has been done at our Kreisinger Development Laboratory
to determine if a less complicated system of ladder vanes can be
used to improve the scrubber's gas distribution but not provide such
a large surface for deposition.  In addition, more sophisticated
wash systems are being designed.

Spray Piping

Deposition on the underbed spray piping has  been reduced dramatically
by the use of strategically placed wash nozzles and the installation
of some quenching nozzles at the inlets of the scrubbers.

The use of some synthetic materials to prevent corrosion has led C-E
to the conclusion that the material best suited to the environment
of the scrubber is fiberglass.  Piping, either made from fiberglass
or coated with fiberglass hasbeen installed in the existing field
units.  In addition, the use of fiberglass coatings on other scrubber
surfaces has been tried successfully.
                              663

-------
Nozzles

Underbed spray nozzle pluggage is another problem we have  encountered.
Deposit formation in the bed area is directly related to proper
operation of spray nozzles and distribution of spray water.   Nozzle
plugging with bits of scale and miscellaneous debris which happen
to enter the spray water system as well as a need for proper mixing
of spray water    and recycle slurry water led to the development
of a special nonclogging nozzle.  There have been no serious plug-
gage problems in the field units since installation of the nozzle.
Both overbed and underbed recycle of slurries have been accomplished
with these nozzles.

Bed Plugging

Bed plugging of two types has been experienced at both existing
demonstration units.  They are defined as follows:

1.  Deposits formed by mechanical means.  Deposits formed  because
    of low gas velocities in sections of the scrubber bed  fall in
    this category.  These deposits may be cementitious in  nature
    because of the mixture of solids found in the scrubber,  but
    usually can be washed away with a fire hose.

2.  Deposits formed as a result of precipitation of a slightly
    soluble calcium salt such as calcium sulfate, calcium  sulfite,  or
    calcium carbonate.  The scales formed by these deposits  are
    always crystalline in nature and are usually very difficult
    to remove.

The control of these two types of bed pluggage has been accomplished
by the use of higher liquid to gas ratios, to prevent mechanical
deposits, the addition of extra make-up water to the system to dilute
the concentrations of the slightly soluble calcium sulfate,  and  the
implementation of closer controls on the system's chemistry to prevent
calcium sulfite and carbonate scale.

Work in our research laboratory is just being completed which is
expected to eliminate the use of the extra make-up water for cal-
cium sulfate scale control.

Mist Eliminators

The mist eliminators above the marble bed became coated with deposits
during early runs.   The  throughput gas velocity was raised, and
consequently,  the  dewatering capability was greatly reduced.  The
demisters were cleaned, and a wash system was installed above the
demister section to be operated off-line, as required.  Deposition
on the demister vanes has been reduced  during operation under steady
conditions at moderate loads.

Laboratory studies were undertaken in 1971 to develop an improved mist
eliminator which could sustain higher gas velocities than present
designs and also lend itself to washing better than the existing demister.
                                664

-------
Water table studies suggested that a two stage L-shaped demister
would "be more satisfactory than the present Z-shaped demister.
Air-water models were constructed in our laboratory at Windsor,
Connecticut.  The evaluation showed that the double L-shaped
demister would have considerably more throughput capacity than
the existing design and that it could be washed much more easily.
This two stage L-shaped demister is presently being installed at
Kansas Power and Light Company on both units and has been incor-
porated on all future APCS.

Reheaters

The reheaters have been "plastered" with deposits a number of times
during previous operation.  The improvement in scrubber bed gas  dis-
tribution and the installation of half-track sootblowers have made
it possible to maintain the gas side pressure drop across the reheater.
Operating experience has'led to the conclusion that, in the damp atmo-
sphere above the demisters, a steel-on-steel construction was necessary
to provide fin strength to withstand blowing pressures required  for
Clean operation.  The initial reheaters were replaced with an arrange-
ment of the heavy duty finned tubes to allow for efficient cleaning by
scheduled sootblower operation.

Our experience during the last twelve months since the redesigned
reheaters were installed has revealed that the present reheater
design coupled with half-track sootblowers are adequate if the
demister is performing as designed.

On two or three occasions during the last year, improper demister
operation has caused excessive reheater deposition at both Kansas
Power and Light Company, Lawrence #U and Union Electric Company,
Meramec #2.  The new demister design previously described should
eliminate this problem.

Drain Line Scale

During June, 1969, operation of the APCS at Union Electric, a calcium
sulfite scale was found in scrubber drain line to the clarifier.  It
was theorized that some of the calcium oxide entering the scrubber was
falling out in the scrubber bottom.  This CaO would then begin  to hydrate
as it was conveyed down the scrubber drain line to the clarifier.  This
hydration would cause a rise in the pH of the scrubber effluent.  Since
the scrubber effluent contains a high concentration of calcium  bisulfite,
an upward shift in pH would cause a conversion of calcium bisulfite to
sulfite and a shift in solubility and probably produce the observed cal-
cium sulfite scale.

To eliminate this problem, the bed overflow pots were tied off and
piped to the clarifier separately thereby insuring no shift in  solu-
bility.  The calcium oxide that fell out in the scrubber bottom was
conveyed to a well-mixed recycle tank and then pumped up above  the
scrubber bed in order to utilize this additive.

This system has undergone extensive testing at both Union Electric
and Kansas Power and Light Company.  It has been determined that
                                   665

-------
the calcium oxide cannot be injected in the "bed without calcium
sulfite scale occurring.  A modification to the recycle system vas
made at KP&L Company in the spring of 1971.  A portion of the pot
effluent slurry was pumped to the recycle tank and thoroughly mixed
with the calcium hydroxide slurry.  It was determined that if the
proper amount of pot effluent slurry was used (i.e., utilized to control
the recycle tank pH essentially) no scale was produced when the
material was recycled and yet significant improvement in sulfur
dioxide absorption could "be achieved.  This improvement ranged
from 200-^00 ppm additional sulfur dioxide removal depending on the
stoichiometric amount of limestone being fed.

Furnace Operation

Kansas Power and. Light Company has been able to operate the APCS
for periods of four weeks or longer at a time without serious
plugging developing in the back-pass of the boiler.  Additional
sootblowers have been ordered but have not been installed to
date.

Union Electric Company Meramec #2 has experienced back-pass plugging
which  has  required high pressure cleaning.  The economizer of this
unit is extremely close spaced and is normally difficult to clean.
In addition, the tubular air heater does not lend itself to the
furnace injection APCS as well as the regenerative air heater.  The
application of the furnace injection system to existing units requires
a thorough evaluation of the back-pass of the boiler to prevent prob-
lems similar to those experienced at Meramec #2.

Other Field Work

During 1970 the application of wet-scrubbing for particulate removal
from boilers fired with low sulfur and low ash coal was evaluated at
the Naughton Station of Utah Power and Light.  A scrubber having one
marble bed and two marble beds in series, and a rod type scrubber
were tested.  Results indicate a particulate reduction to 0.02-0.035
grains per Standard Cubic Foot (SCF) of dry gas for the marble bed and
rod type scrubbers operated at a 6-10 inch pressure drop.  Increasing
the pressure drop across the rod type scrubber resulted in a reduction
of the particulate matter below 0.01 grains per SCF of dry gas.

Field testing was continued at Meramec Station #2 of Union Electric
Company  and Lawrence Station ffh of Kansas Power and Light Company
to evaluate a new nozzle design, variations in recycle methods and
limestone addition on the removal of SO  and deposit formation on
the scrubber components.  S0? removal up to 70 percent was achieved.
Interrupted operation due to change to gas firing to meet load demands
did not permit completion of scheduled testing program at Lawrence.
(This unit is capable of operation at higher loads on gas than coal.)

Following equipment revisions to incorporate new knowledge and experience
test programs were undertaken at both field units in 1971  in the follow-
ing general categories:

1.  Operation without recycle


                                 666

-------
2.  Operation with overbed recycle
3.  Operation with underbed recycle

k.  Simultaneous operation of both underbed and overbed recycle

Results were:

Kansas Power and Light Lawrence #U

The test program on recycle optimization was completed.   Tests were run
with above and below bed recycle, with and without dilution (bed over-
flow water directed to the recycle tank) at QO MW.

In general,results show:

1.  The APCS can be run with above  bed diluted recycle without scaling
    at or below stoichiometric additive feed rates.

2.  Modifications are required (additional punrp capacity) to fully test
    underbed recycle.

3.  Dust efficiency is between 98.5 and 99-2 percent.

k.  Overall SO  removal efficiency ranged from 50 to 75 percent depending
    on mode of operation.

Approximately 1500 hours of operating time was accumulated in 1971 up
to mid-June 1971 despite a three week annual outage.

The Environmental Protection Agency ran a weeklong series of tests at
our installation in Kansas in order to set national standards in terms
of particulate and gaseous pollutant emissions.

Kansas Power and Light Lawrence #5

The unit has been in service since March, 1971, on gas operation.
The APCS under went preliminary check-out in September, 1971-  The
unit was operated on a partial coal firing basis for five days prior
to a scheduled one-month unit outage.  It is presently planned to
return the unit to service the week of November 15, 1971, at which
time full coal operation will commence.

During the preliminary operation of the APCS at the end of September,
two problems were observed.

Initial operation revealed poor gas distribution in the scrubbers.
High velocities were found in the rear of the beds and conversely
the front of the beds (side above the inlet) were not active.  Tests
on gas flow models at our research lab indicate that the installation
of three ladder vanes will give the necessary improvement in gas dis-
tribution.  These systems are currently being installed with a wash
system to minimize deposition.  This ladder vane assembly is a sim-
plified version of those previously tried  at  Union Electric and KP&L
Lawrence #k.
                                  667

-------
A second problem was encountered during this brief period of operation
in September.  The distribution of furnace injected limestone to the
six scrubbers was found to be very poor.  A modification to the lime-
stone injection system is currently being made to improve the distri-
bution of the additive in the furnace.

The correction of the imbalance in limestone addition at the point of
intoduction should allow us to cope with small variations, the scrubbers
by employing the recycle concept previously mentioned.

CONCLUSIONS

In summary, C-E is one of the pioneers in the development of methods
to minimize emissions from stacks of power plants.  For example, initial
work on the C-E pollution control system for controlling SO  and dust
emissions began some seven years ago  at our Kreisinger Development
Laboratory to meet the needs of our electric utility customers.  The
project was undertaken without government funding.  Today, after the
sale of a number of systems, our program is still going strong.  Our
efforts have been expanded and new areas are now being explored.
                                  668

-------
                           TABLE 1

                    AIR POLLUTION MISSIONS

                     IN THE U.S. FOR 1969

                                        MILLIONS
POLLUTANT                               TON/YEAR      %_

Sulfur Oxides                              26       18.3
Nitrogen Oxides                            13        9-2
Particulate Matter                         12        8.U
Carbon Monoxide                            72       50.7
Hydrocarbons                               19       13.k
                           TABLE   2

              COST COMPARISON OF CE WET SCRUBBING

             PROCESS VS SULFUR RECOVERY PROCESSES

                         CAPITAL COST             OPERATING COST
    PROCESS                  $/KW                    MILLS/KWH

CE                         13 to 25                        0.2

Sulfur Recovery            30 to 60                 0.7 to 1.5
                               669

-------
Fig.
-1
   670
                                             CO
                                                o
                                                o
                                             O
                                               o

-------

                                   ^!

         CE-APCS
KANSAS POWER & LIGHT CO.
     LAWRENCE, KANSAS
       LAWRENCE it
        Fig.   -2
           671

-------
                CE-APCS
KANSAS POWER & LIGHT CO.  - LAWRENCE 5
           LAWRENCE, KANSAS
               Fig.   -3
                   672

-------
          REMOVAL OF SULFUR DIOXIDE FROM STACK GASES

             BY SCRUBBING WITH LIMESTONE SLURRY;

DESIGN CONSIDERATIONS FOR DEMONSTRATION FULL-SCALE SYSTEM AT TVA
                              By
                       B.  G.  McKinney
              Division of Power Resource Planning
                  Tennessee Valley Authority
                    Chattanooga, Tennessee

                        A.  F. Little
               Division of Chemical Development
                  Tennessee Valley Authority
                    Muscle Shoals, Alabama
                 Prepared for Presentation at
  Second International Lime/Limestone Wet Scrubbing Symposium
       Sponsored by the Environmental Protection Agency
                    New Orleans, Louisiana
                      November 8-12,  1971
                              673

-------
               REMOVAL OF SULFUR DIOXIDE FROM STACK GASES

                  BY SCRUBBING WITH LIMESTONE SLURRY;

      DESIGN CONSIDERATIONS FOR DEMONSTRATION FULL-SCALE SYSTEM AT TVA

                                   By

                            B.  G.  McKinney
                   Division of Power Resource Planning
                       Tennessee Valley Authority
                         Chattanooga,  Tennessee

                             A. F.  Little
                    Division of Chemical Development
                       Tennessee Valley Authority
                         Muscle Shoals, Alabama
                                ABSTRACT


          Design considerations are presented for the limestone slurry
scrubbing system planned for TVA1s Widows Creek unit 8 (550-mw; plant
located in Northeast Alabama).   The scrubber type will be selected on
the basis of results from pilot plant tests.   Basic design premises for
the overall installation are given and the following major design factors
are discussed.

    'Limestone handling and grinding.

    'Scrubber design (e.g., gas velocity, entrainment separation,
     turndown, liquor flow rate, limestone amount and particle size,
     solids content of slurry).

    *Particulate removal.

    'Equipment arrangement.

    *Flue gas handling and conditioning.

    ^Instrumentation and control.

    *Materials of construction (for resistance to both corrosion and
     erosion).

    *Method of solids disposal.

-------
               REMOVAL OF SULFUR DIOXIDE FROM STACK GASES

                  BY SCRUBBING WITH LIMESTONE SLURRY;

    DESIGN CONSIDERATIONS FOR DEMONSTRATION FULL-SCALE SYSTEM AT TVA

                                   By

                             B. G. McKinney
                   Division of Power Resource Planning
                       Tennessee Valley Authority
                         Chattanooga, Tennessee

                              A. F. Little
                    Division of Chemical Development
                       Tennessee Valley Authority
                         Muscle Shoals, Alabama
          The Tennessee Valley Authority (TVA) has a major research and
development program under way, mainly in cooperation with EPA, on methods
for removing S02 from power plant stack gas.  Small-scale and pilot plant
work, plus the EPA-TVA large-scale test program at the TVA Shawnee Steam
Plant, have been described in other papers in this symposium.  In addition,
TVA decided in late 1970 to install a full-scale S02 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.  Hopefully, a removal system can
be designed and demonstrated that will serve as a model for any future
installations required.

          Among the reasons for selecting Widows Creek No. 8 as the demon-
stration unit was the fact that the electrostatic precipitators there are
quite inefficient.  Since additional dust removal capacity was needed anyway,
it was decided to install a wet scrubbing process and thus remove both the
residual dust and the S02.  Thus the dust problem was a major factor in the
decision to remove S02, which otherwise might have been delayed until design
data were available from the EPA-TVA Shawnee project.

          In selecting a process,  the first choice to be made was between
recovery and throwaway operation.   Since no proven recovery methods were
available, and marketing problems were considered quite difficult, throwaway
was selected even though it involves a solid waste disposal problem.

          The next choice was between lime and limestone as the absorbent.
Lime is more active but can be produced economically only by injection of
limestone into the boiler and catching the resulting lime in the scrubbers.
Several major problems made this course unattractive:
                                    675

-------
     1.  Experience by TVA and others has shown that injection of
         limestone into the boiler can cause fouling and plugging
         of boiler surfaces and tube assemblies.

     2.  The use of lime, particularly when introduced with the gas,
         aggravates the scaling problem.  Scaling can be reduced or
         eliminated by operating the scrubber at low pH or by high
         blowdown and replacement with fresh water.   Both are unde-
         sirable, the latter particularly  because it entails outflow
         of scrubber liquor to watercourses.  It was decided that the
         TVA system must operate as a "closed loop," that is, with
         return of all liquor to the scrubber.

     J.  Since it was desired to retain the precipitators, the presence
         of lime in the gas from the boiler would have complicated
         operation.

     4.  Dry grinding of limestone creates a dust problem.

     5.  If lime is generated in the boiler, there is the problem of
         getting the same CaO:S02 ratio to all of the scrubbers serving
         the boiler.

          It was decided, therefore, to introduce the limestone, after wet
grinding, directly into the scrubber circuit.  Experience by others has
indicated that this reduces the scaling problem by a major degree.  It also,
of course, avoids the other problems listed for boiler injection.

          An attempt is made in this paper to define and evaluate the design
considerations associated with limestone scrubbing.   Design areas considered
are limestone handling and grinding, scrubber design, particulate removal,
equipment arrangement, stack gas handling and conditioning, instrumentation
and control, materials of construction, and method of solid disposal.

          Figure 1 shows the general area at the Widows Creek Steam Plant,
with the expected location of equipment.  The new waste solids pond is also
shown.
Basic Design Premises

          In preliminary planning for the scrubbing facility, it was first
necessary to establish major design premises.  The more important of these
are as follows:
                                676

-------
\
                                                                       w
                           677

-------
     1.  Coal analysis (as fired basis)

         a.  Ash content, 25$

         b.  Sulfur content, 4.3$

         c.  Moisture, 5-0$

         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 1,600,000 acfm at 280°F
             (5,325,000 Ib/hr)

     3.  Sulfur dioxide removal

         a.  Percent removal, 80

         b.  Inlet concentration, 3^-0 PPm (wet basis); 37^-0 PPm (dry basis)

         c.  Outlet concentration, 650 ppra (wet basis); 750 ppm (dry basis)

     k.  Particulate removal

         a.  Inlet particulate loading,2 5.6 gr/scf (dry); 5.! gr/scf (wet);
             3.6 gr/acf (280°F)

         b.  Particulate level at scrubber exit,2 0.020 gr/acf (l25°F saturated)
             0.022 gr/scf (wet); 0.026 gr/scf (dry)

     5.  Stack gas reheat temperature, 175°F (50°F rise)

These premises have been used for preliminary engineering purposes but will
be reviewed and updated as required prior to detailed engineering.
1 Based on a total of 33$ excess air including air heater leakage.
2 Based on the conventional ASME sampling technique.
                                   678

-------
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.  Capacity of the latter is
sufficient for 3 days'  operation.

          Limestone is conveyed from the live storage silo to a wet ball
mill where it is ground from the purchased size (probably 1-1/2 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 re-
cycled to the ball mill.  The product slurry (50-60$ solids) from the
classifier goes to a hold tank from which it is pumped to the scrubbing
system.

          Final selection of the particle size has not been made, but it
will not likely be coarser than 70$ minus 200 mesh or finer than 90-95$
minus 325 mesh.   Selection of the size will be made after completion of
additional pilot plant tests.  Hopefully, the additional tests will produce
sufficient data to permit determining whether the advantages of a finer
grind justify the increased grinding cost.  The potential advantages are
primarily (l) less erosion of piping and equipment and (2) less limestone
requirement for a given S02 removal rate, other conditions being equal.

          Final selection of limestone feed stoichiometry has also not yet
been made but probably will be between 1.2 and 1.5.  The final selection
will depend upon results of additional pilot testing and selection of a
particle size.

          The limestone source probably will be deposits in the vicinity of
the Widows Creek plant.   Estimates have indicated that the cost of shipping
limestone from greater distances would not be economical unless some major
advantage could be demonstrated,  which has not yet been done.  The local
deposits are fortunately quite low in magnesium,  which could cause a ground
water pollution problem if present in large amounts.   The composition of the
limestone is given in another paper in this symposium ("Removal of Sulfur
Dioxide from Stack Gases by Scrubbing with Limestone Slurry:  TVA Pilot Plant
Tests—Part I, Scrubber-Type Comparison" by T.  M.  Kelso,  P.  C.  Williamson,
and J.  J.  Schultz).
                                  679

-------
o
                                                                                                                         CVJ

                                                                                                                         3
                                                                                                                                en


                                                                                                                                ff
                                                                                                                                •r-l
                                                                                                                                T3
                                                                                                                                C
                                                                                                                                •rl
                                                                                                                                M
                                                                                                                                O
                                                                                                                                a)

                                                                                                                                t>0

                                                                                                                                •1-1
                                                   680

-------
          The flowsheet shows recycled pond water as the liquid portion
of the mill feed.  Use of recycled water at this point reduces the amount
available for washing the mist eliminators in the scrubber, for which a
substantial flow of clear liquid is needed.  Further attention will be
given to the problem of portioning out the available clear liquor for the
various needs.
Scrubber Design

          Several scrubber designs and configurations for limestone
scrubbing have been considered in studies conducted at TVA.  Three types
remain under consideration for the Widows Creek facility:  (l) a venturi -
spray tower system, (2) a venturi - mobile-bed combination, and (3) the
mobile-bed type alone.  Flow diagrams for the three are given in Figures 3>
ij-,- and 5-  In each of the three schemes, four scrubber trains will be used--
one for each of the four stack gas ducts from the boiler.

          In the venturi - spray tower system (Fig 3)» tne Sas exiting
the electrostatic precipitator is cooled and presaturated prior to entering
the absorption tower.   The latter is an open spray tower equipped with a
venturi in the bottom of the tower.  The venturi, through which a portion
of the scrubber circulating slurry is injected, serves to remove fly ash
and to atomize the scrubbing liquor for S02 absorption.  The remaining
circulating slurry is injected into the top portion of the spray tower
through spray headers equipped with nozzles.

          Under the current plan, limestone slurry from the mill, recycled
pond water, and makeup river water are introduced into the absorber circu-
lation tank (however,  as mentioned earlier, it may be necessary to use all
available water and clear liquor to wash the mist eliminators).   The resulting
slurry is circulated from the tank through the presaturator,  venturi, and the
two spray tower headers, with liquid to gas ratios (L/G) of about J, 20, 20,
and 20 gal/Mcf, respectively, to the four points.

          The pressure drop across the venturi is 10-15 in. H20.  Recent
pilot tests indicate that a lower pressure drop than this is adequate for
dust removal.   However, the tests also indicate that the higher pressure
drop is necessary for adequate S02 removal in the venturi, presumably because
the pressure drop gives a more effective spray pattern in the lower part of
the scrubber.

          The superficial design velocity for the spray tower is 6-8 ft/sec,
and the solids content of the circulating slurry is 10-15$ by weight.
                                  681

-------
                 QQQ
oo oo
OO
ru
                                  OOO  OOO
             ru
    Dri
                                             s
                                                 •H

                                                 O
                                                 (1)

                                                 4-1

                                                 CO
                                   I


                                   to
                                   tJ
                                   a,
                                                 4-1

                                                 c
  O   O
  O   O
                682

-------
OO 00

OO
                                           OOO  OOO
                                                              g
                                                              M
                                                              toO
                                                              W
                                                              •H
                                                              Q


                                                              s
                                                              o
                                                              CO

                                                              >>
                                                              CO
                                                              T)

                                                              (1)
                                                              A
                                                              5
                                                              c
                   o   o
                        683

-------
oo oo
oo
          5
                     ooo ooo
                               «J
                               •i-l
                               P
                               §
                               •d
                               
-------
          In  the venturi - mobile-bed system (Fig k), stack gas from the
electrostatic precipitators enters the venturi where it is cooled,
saturated, and the fly ash removed.  The cooled gas then flows to the
mobile-bed scrubber where the S02 is absorbed.   Limestone slurry and
recycled pond water (see earlier discussion) are added to the absorber
circulation tank.  Slurry is circulated from this tank to the upper of
the two beds  through a spray nozzle header (low pressure) and flows down
through the scrubber.  Slurry overflows from the absorber circulation
tank to the venturi circulation tank from which it is circulated through
the venturi.

          The absorber is operated with a liquid:gas ratio of about
kO gal/Mcf and a pressure drop of about 6-8 in. H20.  Corresponding values
for the venturi are 20 L/G and if.-5 in. H20.  The superficial design
velocity in the mobile-bed scrubber is 12-13 ft/sec.  The solids content
in the circulating slurries has been set tentatively at 5 and 8$, re-
spectively, for the mobile bed and the venturi.  The 5$ was selected because
general practice is to avoid higher contents in order to limit wear of the
plastic balls.  However, the pilot plant has been operated at 10-15$ solids
to reduce scaling, improve S02 removal, and reduce pumping load for trans-
porting the waste solids to the pond and the separated liquor back again.
Finer limestone particle size may allow use of the higher solids content
without excessive ball wear; this is being tested in the pilot plant.

          In  the scheme involving use of mobile beds alone (Fig 5); tne
flue gas is cooled and saturated in a spray chamber presaturator before
entering the mobile-bed scrubber, which removes both fly ash and S02.
Limestone slurry and recycle pond water are added to the absorber circu-
lation tank from which the slurry is circulated to the upper of the three
beds through a spray header and flows down through the other beds.

          The absorber is operated at an L/G of if-O-^O gal/Mcf and 9-12 in.
H20 pressure drop.  An L/G of about 5 is used for the presaturator.  The
absorber superficial gas design velocity is 12-lJ ft/sec.

          The three beds are specified because this was the arrangement
tested in the pilot plant.   There is some evidence from the pilot plant work,
however,  that the upper bed contributes very little to S02 removal.  Further
tests will be made in an effort to find the best combination.

          In each of the three scrubber arrangements, the absorber is equipped
with an entrainment separator.   Provisions will be made for continuous or
intermittent flushing with recycled pond water.  The selection of the entrain-
ment separator is further discussed later under the problem of particulate
removal.
                                  685

-------
          Some of the advantages and disadvantages of each of the three
schemes are as follows:

     1.  The venturi - spray tower has a lower superficial gas
         velocity, which results in a larger scrubber cross-
         sectional area than the other two schemes.

     2.  Erosion of the plastic spheres in the two schemes using
         mobile beds is a major potential problem.

     3.  The turndown capacity of the mobile-bed scrubber is limited,
         which may lead to pluggage and other problems at reduced
         loads.  Also, fly ash removal may be a problem at reduced
         load in use of the mobile-bed scheme in which the scrubber
         must remove both fly ash and S02.

     4.  For the large cross-sectional areas involved, there may be
         the problem of poor sphere distribution in use of the mobile
         beds.  It may be possible to keep the spheres in place by use
         of vertical partition screens but this has not been proven
         in practice.

     5.  The higher gas velocity in the mobile bed results in a higher
         loading on the entrainment separator.

     6.  The mobile-bed scrubber is a better contacting device than
         the spray tower, which gives better S02 removal efficiency
         or possibly a lower limestone consumption rate.

     7-  Spray nozzle erosion is more severe in the venturi- spray
         tower combination because spray nozzles with higher pressure
         drop are required.

     8.  A higher pressure head on the pumps is required for the venturi -
         spray tower system.

     9.  The venturi - mobile-bed system is the most flexible of the
         three because dust and S02 removal units are separated.

          Alterations to the configurations of the above schemes are under
consideration that may have potential advantages.

     1.  Packed section in spray scrubber.  A short, open packed section
         in the upper part of the spray tower should increase S02 re-
         moval and perhaps be free of scaling if open enough.
                                 686

-------
     2.  The  three-bed TCA scrubber performed fairly well in the
         pilot plant with the balls removed.  Thus a scrubber with
         several horizontal, open, thin grids (similar to the re-
         taining grids in the TCA scrubber) spaced from the scrubber
         top  to bottom, might give adequate S02 removal without the
         problems associated either with the usual type of fixed
         packing or with the mobile-bed type of scrubber.

          The choice between the various scrubber schemes will be made
after completion of additional pilot plant tests.   The investment and
operating costs of the three will be compared before a decision is made;
preliminary estimates, however, indicate little difference among them.
It is likely  that the primary factor influencing the final selection will
be the experience gained regarding operating reliability.
Particulate Removal

          Pilot tests conducted at Widows Creek on unit 8 stack gas indi-
cate that the fly ash can be removed satisfactorily with a pressure drop
of ij--5 in. H20 and relatively low liquid:gas ratios, about 10-15 gal/Mcf.
The main problem, as discussed earlier, is entrainment of slurry solids
into the gas stream.  Thus development of efficient mist elimination
equipment is essential for the Widows Creek project.

          In most of the pilot plant tests to date, a vane-type entrain-
ment separator has been used—selected because of low pressure drop and
large open area which minimizes pluggage and is easy to flush.  Plugging
and poor efficiency have been major problems.  Further studies will be
conducted to determine the best type of entrainment separator for the
scrubbers at Widows Creek.  One course being explored is use of two mist
eliminators in series, the first irrigated with clear liquor to replace
slurry mist with clear liquor mist and the second to remove the clear liquor
mist.  Use of the upper TCA bed, which apparently does not contribute much
to S02 removal, as a mist eliminator may also be helpful since the movement
of the balls should prevent plugging.  Also, if a fixed-bed section in the
top of the scrubber should prove feasible, mist carryover probably would be
reduced.
Equipment Arrangement

          A typical equipment arrangement study drawing is shown in Figure 6.
The general arrangement will be similar for each of the three scrubber
schemes under consideration.  Detailed equipment arrangement drawings will
be made after a specific scrubber scheme is chosen.
                                687

-------
f
                                  utvimoit
                                                                               SOUTH UtVMION
                                            FIGURE 6




                                    General Arrangement Study
                                                 688

-------
          The general plot plan shown in Figure 1 shows the relative
location of the three major areas (scrubbing system, limestone handling-
grinding, and solids disposal) with respect to each other and the
unit 8 powerhouse.
Stack Gas Handling and Conditioning

          As noted earlier, the gas will be humidified and cooled prior
to entering the S02 absorber.  Cooling the gas prior to the absorber is
desirable primarily for two reasons:  (l) better S02 absorption and (2)
elimination of temperature problems in materials selection for the absorber.
The gas entering the absorber will be cooled to a temperature of 150°F or
less.

          The gas leaving the scrubber will be reheated to desaturate and
to provide buoyancy.  The reheater will be designed to provide up to 50°F
temperature rise (125-175°^)-  Both direct reheating by burning oil and
indirect heating with low pressure steam are being considered.  Primary
consideration is being given to indirect steam but no decision can be made
until the economics are further evaluated.

          Soot blowers or liquor sprays will be installed where there is
a likelihood of deposits forming, particularly in the presaturator and
the reheater wet-dry interface areas.

          The gas will be moved through the system by induced-draft fans
located downsteam of the reheaters to permit operation of the fans with
a "clean" unsaturated gas.  Consideration is being given to designing the
fans with sufficient static pressure to permit converting the boilers from
the present pressure type to balanced draft.  A plenum will be provided
connecting the four ducts to the scrubbers to permit operation of the boiler
up to 75$> load with one scrubbing train out of service.   Also, bypasses
will be provided around the scrubbers to the stack to prevent an undue
amount of boiler downtime because of scrubber malfunction, particularly
during the initial shakedown operation of the system.

          Provision for scrubber turndown is a potential problem.   If the
individual scrubbers cannot be turned down without operating difficulty,
then provision must be made to cut one or more scrubbers out of the system
at low boiler load so as to maintain adequate gas velocity through the
active scrubbers.   This could require frequent operator attention to cut
scrubbers in and out of the system and would put emphasis on finding
dampers that open and close easily,  are resistant to sticking, and close
tightly.
                                  689

-------
Instrumentation and Control

          In general, the scrubbing system will be well instrumented to
permit flexibility in control and monitoring of the system since it will
be a demonstration unit.  Also, adequate controls are required to mini-
mize operating personnel requirements.

          Since there are four similar trains,  one of them will have
more instrumentation than the others to permit flexibility in determining
optimum operating conditions.  In addition to being useful in optimizing
the Widows Creek facility, this should give data useful in design and
operation of any subsequent installations.

          Operating variables to be controlled include limestone feed
rate to the scrubber and solids concentration in the circulating scrubber
slurry.  The limestone rate will be regulated to maintain the desired
stoichiometry, either directly or indirectly.  The solids content of the
slurry will be maintained at the desired level by regulating the slurry
purge rate to the pond.   Additional considerations on instrumentation
and control of the system are as follows:

     1.  The induced-draft fans will be instrumented and controlled
         similarly to induced-draft fans for balanced draft furnaces.
     2.
Analyzers will be provided for monitoring the scrubbing
system inlet and exit S02 concentrations.

One of the four trains will be equipped for regulating the
slurry circulation rate to the scrubber.  Amperage recorders
will be provided for the scrubber circulation pumps on all
four trains for rough monitoring of liquid circulation rates.

Sampling points, temperature recorders, pressure and differ-
ential pressure recorders, etc., will be provided as deemed
necessary for monitoring and controlling the system.
Materials of Construction

          Selection of the most economical materials for adequate resistance
to corrosion-erosion is a major problem.  Attempts have been made in the
pilot plant to evaluate various materials of construction including metals,
coatings, and linings.  However, it is very difficult, to obtain conclusive
corrosion-erosion data in the pilot plant because of the short running time
of the tests.  Evaluation of coatings and linings is particularly difficult.
                                  690

-------
          The equipment items most susceptible to erosion-corrosion are
pumps, piping, nozzles, and areas where there is slurry impingement against
surfaces.  Corrosion is generally not thought to be very severe in the
system except for its enhancement by erosion.  The material that appears
to be most promising for the high erosive areas is carbon steel lined with
a soft elastomer such as neoprene with a 55-60 Shore A hardness.  However,
in addition  to being expensive, neoprene-lined carbon steel has some distinct
disadvantages; field modifications and repairs are difficult, expensive,
and time consuming.  Such linings also generally have a maximum allowable
continuous operating temperature of only about 150°F.

          For tanks where erosion is not as severe, carbon steel lined or
coated with  a less expensive material than neoprene should be satisfactory.
It may also  be feasible to use unlined steel, as was done in the early ICT
work in England--particularly where the pH is above 6.0-


Solids Disposal and Water Recycle

          The solids purge stream will be pumped to the pond without any
concentration of the solids.  The supernatant clear liquor will be recycled
from the pond to the scrubbing system for reuse.  River water makeup will be
used as necessary to maintain the water balance in the system.   Water is lost
from the system by evaporation in the scrubber, hydration of reaction products,
pond water evaporation, pond seepage, and entrainment with settled solids as
interstitial water.  Water enters the system as rainfall into the pond and as
river water makeup to the scrubber.   An overall water balance is given in
another paper in this symposium ("Removal of Sulfur Dioxide from Stack Gases
by Scrubbing with Limestone Slurry:  Operational Aspects of the Scaling Problem"
by A.  V.  Slack and J. D. Hatfield).

          Thickeners could be used to concentrate the purge slurry and reduce
the pumping rates to and from the pond.   However, the savings in pumping costs
do not justify the additional capital investment required for thickeners.
Also there is some doubt as to how effective thickeners would be in concen-
trating the purge slurry because of the very poor settling characteristics of
the product solids.

          One of the undesirable aspects of the solids disposal system is the
possible low bulk density of the settled solids in the pond.   Data from the
TVA pilot plant indicate that the settled solids contain about $0-60% by
weight water.  The corresponding bulk density is 89-82 lb/ft3 of slurry, which
results in a large pond volume requirement for storing the accumulated solids.
Further work is planned aimed at reducing the settled solids volume.
                                     691

-------

-------
              TEST PROGRAM
                 FOR THE
  EPA ALKALI SCRUBBING TEST FACILITY
                  AT THE
          SHAWNEE POWER PLANT
                    By
M. Epstein
F. Princiotta
R. M. Sherwin
L. Szeibert
I. A. Raben
Bechtel Corporation
Environmental Protection Agency
Bechtel Corporation
Bechtel Corporation
Bechtel Corporation
                 Presented at the
     Second International Lime/Limestone
          Wet Scrubbing Symposium
            New Orleans, Louisiana
            November 8-1 2, 1971
                     693

-------
                         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 limestone  scrubbing for removal of sulfur  dioxide and particu-
lates  from boiler flue gas.  The Office of Air Programs (OAP) of the
Environmental Protection Agency (EPA) is  sponsoring this program,
with Bechtel Corporation of San Francisco as the major contractor.

Phase I of the test  program, which has been completed, consisted of
preliminary engineering, equipment evaluation and site  selection.
Phase II, which is  close to completion, involves the detailed design
and construction of the facility and the development of the test-plan
and mathematical models for predicting system performance.  Phase
III -will comprise, primarily,  the testing portion of the program.

The test facility consists of three parallel scrubber systems, each
capable of treating approximately 30, 000 acfm of flue gas, which are
integrated into the  flue gas ductwork of an existing coal-fired boiler
at the  Tennessee Valley Authority (TVA)  Shawnee Power Station,
Paducah, Kentucky.  The facility was designed for maximum flexibility
and has a high degree of instrumentation for control and recording of
data over a wide range of operating conditions.   Construction is approxi-
mately 80% complete, with Phase III start-up presently  scheduled for
March 1972.
                                 694

-------
EPA has utilized the capabilities of several organizations to maxi-



mize  the efficacy of the total program.  For example, Bechtel



Corporation,  as the major contractor,  has prepared the detailed



design of the test facility, is developing the test program and mathe-



matical models and will direct the test efforts.  TVA is constructing



the facility and will operate the unit during the test program.  Other



contributors include, Radian Corporation, who has provided impor-



tant support in the  analytical determination areas, and McCrone



Associates, who participated in the development of the particulate



mass loading  and size distribution measurement procedures.
                                 695

-------
                 TEST PROGRAM OBJECTIVES


The overall objective of this program is to evaluate the feasibility and

economics of closed-loop limestone wet-scrubbing processes.  The

following are the major goals of the program:
        Investigate and solve operating and design problems,
        such as scaling,  plugging,  corrosion and erosion.
        Generate test data to characterize scrubber and system
        performance as a function of the important process
        variables.

        Study various solid disposal methods.
        Develop mathematical models to allow economic scale-
        up of attractive operating configurations to full-size
        scrubber facilities.   A parallel economic study will be
        performed to enable estimation of capital and operating
        costs for the scaled-up system designs.

        Determine optimum operating conditions for maximum
        SO  and particulate removal,  consistent with operating
        cost considerations.
        Perform long-term reliability testing.
                                 696

-------
                        TEST FACILITY


The following discussion provides a brief description of the test

facility.  Reference 1 provides a recent,  more complete description.


The test facility consists  of three parallel scrubber systems,  each

•with a different scrubber  design and its own slurry handling system.

Scrubbers will  be of prototype  size,  each capable of treating approxi-

mately 30, 000 acfm of flue gas  from  the  TVA  Shawnee coal boiler #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 units.  The 30, 000 acfm scrubber train was

judged to  meet these requirements.


SCRUBBER SELECTION

The major criterion for scrubber selection was the  capability of remov-

ing both sulfur dioxide and particulates with high efficiency (sulfur
dioxide removal greater than 85% and particulate removal greater
than 99%).  Other factors  considered in the selection of the scrubbers

were:
       Ability to handle slurries without plugging  or excessive
       s caling.
                                 697

-------
     •    Cost and maintenance requirements.
     •    Ease of control.
     •    Pressure drop.


Based on the limited information available in the literature,  the follow-
ing scrubbers were selected:


     (1)   Venturi followed by an after-scrubbing absorption section
         (spray  or pall-ring packed-bed).
     (2)  Turbulent contact  absorber (TCA).

     (3)  Marble-bed absorber  (Hydro-Filter)


The venturi (manufactured  by  CHEMICO) contains an adjustable plug,
which permits control of pressure drop  under a wide range  of flow

conditions.  The TCA (manufactured by  UOP and described  in Ref. 2)
utilizes a packing of low density plastic  spheres  which are free to move
between retaining grids.  The Hydro-Filter  (manufactured by National
Dust Collector Corporation and described in Ref. 3) utilizes a packing
of glass spheres (marbles)  which are normally in slight vibratory
motion.  A "turbulent layer" of liquid and gas above the glass spheres

increases mass transfer and particulate removal.


Models describing pressure drop, particulate and sulfur dioxide removal
within the three scrubbers  have been presented in Ref. 4.


SYSTEM  FLEXIBILITY

The test facility has been designed to achieve great flexibility:
                                   698

-------
         The three scrubbers can be tested in parallel.

         Scrubber internals and piping configurations can be changed.
         For example, the TCA can be operated as a one,  two or  three
         stage unit, with a variety of liquor flow piping arrangements.

         The facility can be operated under various alkali addition
         modes.  These include: limestone in the scrubber circuit,
         hydrate in the scrubber circuit and limestone (or dolomite)
         injection in  the boiler.  For the scrubber circuit addition
         modes, alkali can be added at two locations, the Scrubber
         Effluent Hold Tank or the Process Water Hold Tank.

         Various solid disposal configurations can be evaluated.   They
         include: Clarifier/Pond
                  Clarifier / Centrifuge /Pond
                  Clarifier / Filter /Pond

         Each  scrubber has been furnished with a quench spray to
         permit humidification and cooling of the  flue gas prior to
         contact with the scrubbing slurry.

         A heat exchanger has  been provided  to permit cooling of  inlet
         scrubber liquor (only  one scrubber at a time).

         Each  scrubber has its own  oil-fired  reheater (and stack)  to
         increase the temperature of the exit gas over wet-bulb values.

         Each  scrubber will accommodate either  a chevron or centrifu-
         gal mist  eliminator (demister).
SYSTEM RELIABILITY


Difficulties in achieving reliable  operation of wet limestone scrubbing
systems attributable to scaling, plugging,  erosion and corrosion have

been reported by many investigators.


The following are among the design and test  features included to maxi-

mize system reliability:
                                    699

-------
    •    The scrubbers are either of stainless steel or Neoprene-lined
         carbon steel  construction.

    •    All major piping,  pumps and tanks are lined with rubber or
         fiberglass reinforced polyester.

    •    Variable speed pumps have been selected, where practicable,
         for flow control to avoid the solids build-up which can be
         produced by throttling flow controllers.

    •    Sootblowers have been installed at each scrubber inlet to mini-
         mize solids deposition.

    •    The flue gas has been heated downstream of the induced draft
         fan to reduce solids deposition within the fan.

    •    The ability to operate over a wide range of independent
         variables has  been designed into the facility,  to increase the
         likelihood of achieving operating conditions which are free of
         scaling and other operational problems.

    •    Scrubbers have been selected which are relatively immune to
         plugging and are designed  to minimize solids build-up at
         liquid contact zones.
SYSTEM DESCRIPTION

Typical schematic flow diagrams for the Venturi, TCA and Hydro-Filter

systems are shown in Figures 1,  2 and  3, respectively.   Not shown are

flue gas  saturation sprays, heat exchanges for cooling of liquor,  flow

controls, and process details.  These figures are intended to illustrate

the range of operating conditions  possible at the test facility.
In Figure 1, a venturi system is shown with a packed-tower (Pall-ring)

after-scr ubber, with hydrate (Ca(OH)  ) addition to the Scrubber Effluen
                                     L,
Hold Tank and a Clarifier/Pond combination for waste disposal.
                                  700

-------
In Figure 2, a TCA system is  shown with the TCA in a three-stage



configuration,  with limestone addition to the Scrubber Effluent Hold



Tank and a Clarifier/Filter/Pond combination for waste disposal.  In



Figure 3,  a Hydro-Filter system is shown with limestone addition to



the boiler  and  a Clarifier/Centrifuge/Pond combination for waste dis-



posal.  Any of the scrubber systems can be operated with any of the



solids disposal systems  and with  several piping  flow configurations.








For all configurations,  gas will be withdrawn from the boiler  ahead of



the power  plant particulate  removal equipment so that the entrained



dust, including lime during injection-scrubbing tests, can be introduced



into the scrubber.  Gas flow rate to each scrubber will be measured



by venturi flow tubes  and controlled by dampers on the induced-draft



fans.  Concentration of sulfur dioxide in the inlet and outlet gas will



be determined continuously. The efficiency of the process for removal



of nitrogen oxides also will be  determined by periodic checks  of inlet



and outlet  concentration.








Control of the  scrubbing systems  will be carried out from a central



graphic panelboard.  An electronic data acquisition  system will be



utilized to record the operating data.  The system is hard wired for



data output in  engineering  units directly on magnetic tape.  Onsite



display of  selected information will be available.  Also, important



process control variables will be  continuously recorded and trend



recorders  will be provided  for  periodic  monitoring  of selected data



sources.
                                    701

-------
.2>  s  £
                                                      II



r>
°£i_
LU t
°ogq*
§3oi
QliZlxS
"uj
CO
                           -®
                                                         "
                                                      gll
               A
                                      A
16
u



lAl
                                                      O O- 01



                                                      00®
                              702

-------
ro  S  j«

«  2  £
^  M.  C»


.S°  «  ^
;H  u  |_
   a.
                                                          •  —
                                                          (/I O

                                                          si
                                                          O _i


                                                           I

                                                           I
                                                           II
                                                            •*
                                                             fe
                                                          oo®
                             703

-------
   o

   5
CO
   1  =
i,
E-UP WATER^>— ^



(/l
E£g*
£«°i
%* = *-
(•)
j
/*
^
bo »J
frt  o
^  gf
      I
      o
           A
           A
                                                         ra 4-*
                                                         £ ^
                                                         II
                                                          1
                                                          081
                                                          II
                                                          '
                                                         |i|
                                                         O OL (/l



                                                         O®@
                               704

-------
Batch samples of coal,  limestone,  slurry and gas will be taken periodi-
cally during each test run for chemical analysis, particulate size
sampling and limestone reactivity tests.   The locations of these sample
points are indicated shown on Figures 1,  2 and 3.

The Shawnee pilot facility contains five major areas: (1)  the  scrubber
area (including tanks and pumps), (2) the operations building (including
laboratory area,  electrical gear,  centrifuge and filter),  (3)  the
thickener area (including pumps and tanks), (4)  the utility area
(including air compressors,  air dryer, limestone storage silos, mix
tanks, gravimetric-feeder, and pumps),  and (5) the pond area.  Figure
4 is a perspective sketch of the pilot facility.  The pond  area is shown
at the extreme right; the three Clarifiers are near the center, adjacent
to the operations building;  and the scrubber area is in the foreground
with the  boiler facility to the left.
                                  705

-------
/

-------
                         TEST PROGRAM
DESIGN CONCEPTS
Data from the test program (Phase III) will be utilized to:
         Develop mathematical models for process and  equipment
         scale-up to commercial (multi-hundred megawatt plant)
         scrubber systems.
         Evaluate the feasibility and long-term reliability of closed-
         loop limestone wet-scrubbing systems.
The following test periods have been defined:


    (1)  Air-Water Tests.

    (2)  Sodium Carbonate Scrubbing of Air-SO? Gas Mixtures.

    (3)  Break-In Testing for Wet-Scrubbing of Boiler Flue  Gas
        with Limestone and Hydrate in the Scrubber Circuit and
        with Limestone and Dolomite in the Boiler.

    (4)  Screening Testing for Wet-Scrubbing of Boiler Flue Gas
        with Limestone and Hydrate in the Scrubber Circuit and
        with Limestone and Dolomite in the Boiler.

    (5)  Primary Testing for Wet-Scrubbing of Boiler Flue Gas
        with Limestone and Hydrate in the Scrubber Circuit and
        with Limestone and Dolomite in the Boiler.
                                  707

-------
Air-Water Tests.  The experiments with air and water are designed

to:
    (D   Determine pressure drop model coefficients for all three
         scrubber systems without solids.

    (2)   Observe the fluid hydrodynamics within the scrubbers in
         a clean system, e. g. attempt to determine the droplet
         sizes in the  venturi throat region.

    (3)   Determine the quantity of entrained liquid in gas leaving
         scrubber demister.
Sodium Carbonate Tests.  In these tests,  water solutions of sodium

carbonate will be used to scrub SO^ from mixtures of air and SC^.

The purpose of these tests is to determine uncertain model coeffi-
cients  for gas-side controlling mass transfer for all three scrubber

systems.  These data may also be used to evcJuate and design other

systems which rely on sodium carbonate aqueous scrubbing.


Break-In Tests.  The break-in test periods will be used to identify
and resolve possible  operating problems prior to initiation of screen-

ing testing.   The primary objectives of the break-in tests are to:
    (1)   Determine  scaling and plugging tendencies and evaluate
         various methods for scale removal.  The limits of the
         levels of the most important scale-related variables which
         must be set during subsequent testing in order to prevent
         scaling will be identified, e. g. suspended solids concen-
         tration in scrubber inlet slurry,  effluent hold tank resi-
         dence time. An attempt will be made to develop reliable
         methods  for detecting potentially deleterious scaling
         before complete pluggage or major scaling  has  occurred,
                                  708

-------
         e. g. pressure drop information,  liquor conductivity
         measurements, visual observation.  Mechanical methods
         (chiseling, scraping) as well as chemical methods (muri-
         atic acid, oxalic acid, soda ash) will be tested as means
         for removing scale.

     (2)  Determine solids separating and handling sapabilities.
         Various solids separating configurations will be  evaluated
         for their effect upon settling characteristics, scaling ten-
         dencies, dewatering capability, reliability and the ability
         to approach steady state within the time constraints
         imposed during screening testing.

     (3)  Establish criteria for achievement of steady state condi-
         tions and ascertain the time required to achieve  steady
         state for various operating changes.

     (4)  Determine the limitations of the  system with respect to
         control of the independent variables  over their required
         levels.  Also,  determine the replicability of dependent
         variable  response to selected settings of independent
         variables.

     (5)  Evaluate the  efficacy of the  systems-data measurement
         capability.
     (6)  Supply information regarding the  selection of levels of
         some  of the less important independent variables which
         are to be held fixed during the screening testing.  Also,
         evaluate selected combinations of scrubber/liquor piping
         flow configurations.
     (7)  Determine system reliability for subsequent screening
         and primary  test requirements.
Screening Tests.  The screening tests are designed to:
    [1)   Characterize, as completely as practicable,  the effect of
         important independent variables on particulate removal
         and SC>2 removal for the three  scrubber systems.
                                   709

-------
    (2)  Compare model predictions for pressure drop,  particu-
         late removal, and SO,,  removal with the  data and deter-
         mine the best-fit values for uncertain constants and
         coefficients within the models.
The screening tests are divided into two general categories:
    (1)   Factorial tests where the most important independent
         variables and selected "secondary" variables are treated
         in a fractional factorial test matrix.
    (2)  Sensitivity tests where less important variables  are
         tested in an abbreviated test design.
A certain number of undefined tests will be allotted for the screening

tests.  This is to allow flexibility by permitting selection of further

test runs influenced by results  of prior factorial and sensitivity test-
ing. Also,  some of these runs can be used to replicate a previously

tested set of conditions for a relatively long term test (1-2 weeks).
This would  allow preliminary verification on a longer term basis of
potentially attractive operating modes, prior to the primary testing
effort which is performed as the  last portion of the test program.


Primary Tests.  The  objectives of the primary tests are  to:


     (1)   Perform selected short-term testing to ascertain close-
         to-optimum operating  conditions for each scrubber type
         and alkali injection mode.

     (2)  Perform long-term testing (two to four months) on the
         most attractive  operating modes for each scrubber
         system to determine reliability of operation and  to
         develop data  for process economics and for scale-up
         to  larger systems.
                                   710

-------
TEST SCHEDULE


The Break-In and screening testing will be divided into three blocks

which are to be tested sequentially.  They are:


     (1)   Block #1 -  Limestone in Scrubber Circuit

     (2)  Block #2 -  Hydrate in Scrubber Circuit

     (3)  Block #3 -  Limestone in Boiler


The preliminary schedule for the test effort is presented in Figure 5.

The estimated date for  initiation   of System Check-Out is March 1,

1972.  The presented schedule should be  considered approximate only,

since it may not be feasible to adhere to  a rigid schedule in a research

and development program of this  type.


Table 1 presents a description of  the reports which are presently

scheduled for general distribution.


TEST PROGRAM VARIABLES AND LEVELS


The levels of the independent variables for the air-water, sodium
carbonate and screening experiments for all three scrubber  systems

are given in Tables  2 through 12' .  The following conventions have

been adopted in these tables (see Figures 1,  2,  and 3):
     These represent the levels of the statistically designed runs.
     A small number of runs in the test sequence have not been
     statistically designed.
                                 711

-------
          a
          UJ
          I

          O
O   <

iZ   X
     O_  111
                     8

                    £?
                     s


                    I
                     m

                    !


                    I


                    1
                     a


                     s
o



£>•


CO


IX



•O


IO
ROGRAM FUNCTION
                                                                                  i
                                      i
                                                                 i
                                                                           i
                                                                                  i
I
                                             I
                                                                                                 ae     m
                                                                                                 M     2
                                                                                                 «     w
                                               at:
                                                               =
                                                               ot=
                                                                             o
                                                      at
                                              si
                                             si
                                                                                                               fc;
                                                                                                               o
                                                                                                               o

                                                                                                               z
                                                                                                               o
                                                                                                                             ..    6
                                                                                                                       G
                                                                                                                        fe   ^

888
Si   5*   3J
                                                                712

-------
 0)
I—*

.a

 rt

H
      2
      H
      u
      w
      w
      Q
      O
      (X
      w
      Q
4)
4->
nl
P
a
•o .2
4> •*->
-u rt)
nl U
a 3
*•§
m /if
W &
i-H
at
rH
4)
G
4)
O
fH
4)
^
nt
co
rC
^J

O
a
o
r>
£
H

*
to
C
O
•rH
4J
3
i-H
O
CO
/*%

CJ
O
• |H
-l_>
Informa
TJ
0
nt
CO
C
C
4)
i— i
rQ
0
rH
ft
1— 1
n)
C
o
of operati
lx
rH
h
^
3
to
F—1
=fc
M
CJ
0
r-l
«
m
0
d
o
• rH
4-»
4)
r-l
n.
If
0
O
fH
4>
-M
?
fH
•iH
«4
CO
«rH
O
CO
H->
r-l
f*
S3
CO
4)
fH

*
CO
c«
bo
•H
CO
CO
Tl
H->
CO
4)
-M
1— 1
nl
3
H->
CJ
n)
T)
C
nl
T3
0)
a
C
trl
(U
rH
ft
CO
H-*
to
4)
H

•rH
1
M

fH
«


r-H
4)
T3
O
g
H
fH
o
£
*rt
(V
4->
nt
TJ
««
O
C
o
ilizati
4->
3









i
fH
4>
H->
• H
rC
H^
• H
fe
r
bo
•rt
H->
CO
4)
H->
C
•H
ri
nt
4)
rH
rO
M-l
O
CO
be -^
•S 3
-M 4)
: *
-^ .
rT«
O 4)
u a
j^k
rt o
£
T3
C
nt
r-l
4)

4)
T)

























i
rt
T3
<+H
O
C
O
•rH
-U
n)
4>
fH
ft
                                                o
                                                o
                                             o  a
                                             4)  4)  bO

                                             4) 7;  C

                                             fH  ft "H


                                            £  C  0)

                                                O  nj
                                             s«w
                                             £ 4! -H
                                            H  n! *
                       bO

                       CJ
                      • rH






                    S  "  CO

                    O  n  co



                   1  «  g
                                             w
                                                w
                                                   ft
                                    i

                                    a
                                    o
                                   *H <«  bo

                                   §  °.s
                                   g  c  a
                                   C  o  4)

                                   o 33  £

                                   i^  «  n
                                   r   <-*  °
                                   EH  ft t/)
                 fi  r*

                 0  nt

                •H  4)


                 3  ^
                r5  rO


                 S  «H

                 m  O
                 4)

                 fH  CO
                                                                      nt
                                                                     H->

                                                                      nt
                                                                      ^4

                                                                      S
                                                                      °
                                                                  rt -
                                                                      nt
                                                                            a
                                                                            O

                                                                            0
                                 5  °.S
                                 c  c  d
                                 C  O  4>

                                 o 33  £

                                 ^  «  M
                                r  •—'  °
                                H  ft OT
fH
T3
C
nt
CO
a
4)
r-l
rO
O
fH
ft
bo
C
• rH
H->
rt
fH
4)
ft
O
WH
O
>^
fH
p
a
P
w
^
CO
4)
fH
•s
CO
CJ
be
•H
CO
4)
T)
4->
CO
4)
-u
i— i
nt
3
-u
O
ni
TJ
C
nt
TJ
4)
CJ
CJ
nj
r— 1
ft
*M
O
CO
B
nl
H->
CO
•k
nt
-u
nl
T3
«*H
O
s
o
interpretati
CO
4->
CO
4)
-4-1


H->
1 developmen
4>
T3
O
a
TJ
C
n)
CO
a
4)
i-H
rCJ
O
fH
ft
r- 1
nt
a
o
• rH
H->
re)
rH
4)
ft
O
MH
O
rT
nt
H
a
$
3
CO
4>
rH
CO"
c;
bo
•H
CO
4>
T)
HJ
CO
4>
4->
ed and actual
JH
nt
t— i
ft
bo
• rH
4->
CO
4)
H->
screening
<+H
O
CO
4->
r-H
3
CO
0)
fH
*
bo
CJ
•|H
•*J
CO
4)
•»->
H-«
4)
fH
ft
rH
4)
-»->
CJ
• rH
•elopment,
r>
4)
Tl
I-H
4)
T)
0
a
CO
m
4)
U
0
fH
ft


above
ro
=tfc
4->
fH
O
ft
4>
fH
CO
nt
4)
a
$
                                                                o


                                                                ^§
                                                                4)  ™
                                                               HJ  fH
                                                               *+*  bo
                                                                n!  o
S  o

a«
C  O
                   4)

                   i-H
                   -I-)
                   • rH


                   H
-t->  r;
r-l  ™
3  °


1  "


                                               c!

                                               O
                   CO
Te


- L:
                            o c
       H->
       • H
       3
       u

 £-  -
•S *  u

 s -s  ^

 2l|


»?1
«:  o  «
                •H (M

                C! =tfc

                41  .>
                                                            2
                                      .

                                      o
                                                               CO

                                                               4)
bO
fl
•s
4>
4)
fH
0
CO
-o
CJ
m
CJ
Break-i
-*'

CO
=fe
ty
rX
O
O
rH

r4
«2
CO
H->
1— 1
3
CO
4)
PH
4-J
CO
4)
H


•
fH
4)
t— i
• rH
O
W
CJ
•iH
Limestone
HH





>-J
fH
O
ft
4)
PH
i-H
n)
C
• H
fc
*
in
                                                       713

-------
•s
    I
    h
    CU
    a
    x
    W
    0)
    •l>

    cd
    >

    ^
     8.
     ID
    •a




s
•4->
CO
CJ
4-»
•— 1
[r!
O

TJ
$










g
tn
^,
w
^
u
H










§
»
03
3

C
9)











Level!







01
1
|H
£





1









CU
1
h
cd
>



CO
V
J









01
3
2
M
rt



1
0
O O O
O O 0
o o o
o o o
^H CM CO

^H CM CO





tl
4-1
a)
?
O
S
*"


60
Clt
en
^H CO
—1 CM






CO
 >

~t CM



c
o
15

a

c
o
u
Q)
^2
|
U


1
o —• N ^

^-i ro m TJ<


i
1>
"Ji *
M 2
o a
(4-1 .

|?
J
1
O
o o o
o o o
o o o
o" o" o
-H CM CO
*-< CM CO








tt)
1
Ci
h
^


1
U
o o o
o o o
o o o
o o o
^ CM CO

- CM CO








CU
K
O

.r^


S§§
O I-H (M

- CM -0


,

4-*
cd

O >•
^ 2

o •
2-3
J


M O O O
O O fM

^ Mm ^






11
*{
CO
0 %
•-H CQ
^ 1-1
O m
If
^



1
60 o o o
O i— i ro ^O

~H ro ro -


rt
t-i
O
Q
2*



c
00
in ^H

-H CM
0)
**
g

TJ
CU
(U
a
en

en
CO
ct)
O


S ^
rt "
^ CM






CU
4J
at
T3
»
O
a
A



a.
o o

^ CM CO





1
« t.
*J "

^ !o
"^-t CO
n ^t
3 
-------
       c
       D

      I
       |H

       8.

      w
       0
      XI
       ai
      U
id
H
       01
      —4
      XI
       nj



6
en
£
0
i
a










s
0»
m
l/>
U
H





















i
4J
01
ji
c
0)









en
a;




0)
3
2
^
CD




CO
r-4
01
01
J







in
XI
nl

u
cd









to
r-4
0)

J










0)

XI
a
rd
>



**-<
U
0 0 O
o o o
o o o
00*0
.-« OJ PO

- CM m

0>
nl
o
E

^,
^

V
5

.-I (M
-1 CM




CO
(U

(D
CO

"3
h
XI
s

§,
V
XI
1)
0> fc
C U n
_o to 3
CD 00 &
•- .3 £
Is 2

4) fll pl
t> ft to

— t OJ CO

(3
0
'•s
a

a
o
(J

0>
3

w
s
a
00
o o o
o un o
-4 (NJ Tf

- N fO

U
"rt m
r> fn
1 s-
<*H C
0 °
3 Ji
a- o
J
1
U
o o o
O O 0
o o o
o in o
r-t — 1 OJ
~* OJ CO






0)
4-1
rt

g
o
<«


^





s
"-H
o o o
0 O O
o o o

o un o
--( I-H CM

<-* OJ CO






"rt
f-l
o

.tt
^

1
O —< OJ

^H OJ CO

rt
o ^*
JH ft
o »

^ 0
J

a
o o o
o o o
oJ m oo
i-t OJ CO




CO
4->
2

o ^
^ t-t
o «*
SH O«

"~







1
00
o o o
o to o
-* rj ^

-I (M fl




£
Id
s
o
o

s*
J

00
in ^H

-. N
0)
4J
01
0)
0)
xi
a
CO
CO
fl!
a


fi ^
"*
""
-H (M




01

0)

•0
00
1
U

p,










o o o
Tf v£) 00

^H CM CO






c
0
•y
to
o
a.
00
3
ft

c
CM fo in

-H CM fO
00
c
3
U
n)
a
o
S
00

S


c
'"* O ^f
^ rf -H
-H CM c<1


00
C
U
a
MH
o

no
0)
u

rt
w

c
O CO sO

i-H OJ CO
4J
bp
QJ
XJ
•- .s
£ a

s *^
o














































                                                                                                                           B1
                                                                                                                           0)
                                                                                                                           S
                                                                                                                           P.
                                                                                                                           P.
                                                                                                                           nt

                                                                                                                          T3
                                                                                                                          XI

                                                                                                                           O
                                                                                                                           C
                                                                                                                           01
                                                                                                                           U
                                                                                                                           C
                                                                                                                           0
                                                                                                                           U

                                                                                                                            (M
                                                                                                                          O
                                                                                                                          in
                                                      715

-------
w>
-S3

n
c
4)
g


t


'?>
.£
n
C
4)
W
1
n
V
V
J
n
C
0)
£
"C
I
s

• H

t)
n
4>
4)
J
g
£
0
O
^
r
f-







D
^ 00
0 ^C
t. JH

I-M
=Ji ^
Venturi 1
Venturi 1



i-i CM
00
c
*?
(0
OH
Venturi &



t<
v
^



o ^
-?•£
h t-
OJ V
'(H h
JS jS
u u



•— CM





TD
c
0
p.

Clarifier







fl
o
a
4J
n)
5







T)
C
ft
Clarifier/


C
o
rt
h
p
S1
Solids handling con












 D
£ £
O O
O O

OC 00
2 3
0 0
f-i JH
0 O



^ N
m
0)
£
O
cO
00
P
0
0
a-





4)
£
O
O

00
p
o
£





V
N
'm
Limestone particle











h
W













h
W









H
Rl

^H fS]








H
ffi
w



*^
d

'o
ft
C
Limestone injectio












l-H CO



- CM









CO









CO










CO







Excess air












< ffl



~H rsj



















<


















«J
O
U

sajq-ei-iBA. A-r-epuoDag












0













6









6










o







1 Saturation sprays
|











o













o









o










o






t-,
o
p
cr
"c
Vn
O
bo
fl
O
0
U











o













0









o










o







I Coagulant
P3XT^t
PT3H
                                              716

-------
If
h U
en
C
V
£
s
Q.
X
W

>*•

_^
n
C
OJ
CO
n
!
•iments 1
a

W

5

0
u
o
a>
a>
J
H
a
0
U

G
,
nl
U
£
*-i
u>
D
flj
^
a
g
0
c
7
m

3
O



en
D

O
o
on
"S>
?0% throu





Limestone particle size
H
a
w








H
X S
w ft


-1 ISi







H
a
w








H
H X
a ^'
w cu




#
Limestone injection point













Is
— fN]







CO










sS
00








CO





Excess air













< £0
-M


















<














V
a
Ir
•a
0
U













00
..







O










c
O








c
Q





M
ft
h
a
CO
c
0
rt
1













a> o
— N







o
Z










0
Z








o





0
1
u
'o
a
"o
0
U
™—









c
0
CL
flj
Clarifi.




T)
C
0
O-
Clanfiery






T3
C
O
D.
Ll
Clarifie






C
-I1
QJ
rt



§

n)
3
00
g
u
W)
1
^
n
T2
o













o
2








o










0
Z








o





Coagulant












c









c










c
rj








G





Packing diameter













c
0








c
o










G
O








c
0





M
C
,5
u
S.
op
cu
u
S
w
paxij;
                                    717

-------
I  s?
HII
   M
   •D C
   TT 1-*
m
C
&)
E
4)
fi1
w

•*j
'>
.*H
(0
a
4)
4)
,4
m
a
1
S
s*
M
"3
'E
0
-4J
U
rt
Ji
CO
4)
i— I
S1
o
O
~"~"
c
i— i
nj
4)
^
CQ
FH
CQ
4)
5
_
A.
0
O

(5
""I*
^
a
^
ffl
4>
4)
|4



[0

O
•4-1
U
Controlla











o
in










m m
co r--

-~







o
in













in o m o
«M m r- o


I-H rj co •*






.2
"rt
Stoichiometric










o










00 r-<

-M







£2


" N










•^ CO 1-4 «-H


i— i f\] CO Tj<



•o
0
(1)
u
'o
4)
Percent solids









g
u
o
o
o
o*
PO








p
£
u
o o
o o
o o
in o

-co





g
u
o
o
o
o"
rO









g
is
u
o
o
o
in










4)
m
o
C
to
o










1
ac
0
o










B
p.
bo
o o
0 0

•-H CJ






PH
o
o











E
a
be
o o
o o
CO Tf


•-H f\J

CO
rt
PH
I/I

g
0
"o
4)
rt
O
S
er










c
1
O










c
E

^H ro






-S
1
o












c
E
U") O


•-j (M






4}
O
c
4)
O
(H
H
W










C










c
Tf CO

-< ro







^













c



c
u
p-
D
>
0
rt
be
Overflow weir '.











tsj
CO










ro co
CO rO

-ca







S













ro co
rO co


1-1 ro






*
C
O
Flow configura



4)
bO
^
m *^
C C
0 4)
P. O
Clarifier,
Clarifier,


r-n ro




c
0
'-^
Clarifiei


4)
CJ
IM
TJ -M S
O 4) ^
a u c!
0) 4) 0)
rt rt «j
u G G


«-• ro ro





o
a

OJ
^H
rt
5




*
C
c
'•£
CO
(H
ao
c
u
Solids handling













- ro

















<













<










Limestone type


^
CQ
0)
E
O
O
CO
A
M
0
1
o




tn to
E E
O O
O O
oo rj
& -C
2 2
0 0
O O

i-- (M
to
4)

O
o
CO
_
be
0
o



-C
4>
g
O
0
CO
_
be

o
tn

0









to
4)
U
Limestone part










•-H CO


^ ^







CO









CO ^D


-H CM










CO










*
Excess air










< CQ


- ro

















<













<










0>
P-
C
U










<§§


I-H ro







o









o













0









w
Satura tion spra












j



o
Z










o









0













in
4) O
^ Z


I-H ro






o
cr
C
C
be
C
o
0
U


c










0









QJ O
>• z


" N










c










*
Coagulant










U
Z










o.
00
o o
0 0
rH CNJ

^H fvj







O
z











g
a
00
o
o






^
m

a
P.
0
IH
0
C
3
cr
3
— —














H
K
W










H
EC
W








H
EC
W












f-H
EC
W







_!_,
C
'3
a
c
0
u
Limestone inje





CO










c
CO








c
CO
in












j
CO
IT









lameter
ID
01
OJ
p.
w
y>
a
0








c
cO










c*
CO









C
CO













C
CO









C
U
ra
P-
bC
0)
K
PI^H
                                          718

-------
     «  6
     c £
a
H
     •
     rt  i
     h U
     T3  I.
     C  u
     0) W5
     a
     D  B
     "O '
     C  tl
     01  >.
riments
(X
X
U
• Factoria
s,
u>
0>
OJ
J




,3
i
rt
ffl
o
[0
QJ
J



M
o
4->
O
(U
Controllab!


tower
£
D M
1 '
Venturi a
Venturi £




— < ro
0)
o
4-1
nt
t-.

CO
Venturi &








o
"rt
t-t
Scrubber configu




O CO
o *-*




^H ro





ON ^ CO i«


- M CO ^






O
fl)
Stoichiometric r




^ (NJ
CO ^




-H N





6^ {NJ ^


-H N m *


"U
•£>


£
O
0)
Percent solids r



1
u
o o
0 0
0 0
iD O
^H CO




-H (N]




a
u
0 0
o o
o o


-H rvj








0)
rt
o
CO
nJ
O




tx
GC
O o
0 o




^ N





a
O
o









4-i
G
>
1)
"rt
fn
O
'^-^
O




a
GO
0 0
O O
CO O




- N





ex
GO
o o
o o


- 
-------
    W u
    K H
0)

H

CO
C
g
fn
OJ
W

15
'£
O
u
o
to
0)
(U




5
^
rt
0)
M
CQ
*n
5
to
!



Controllable Factors










CO
00
rt
to








to
00
rt
Cfl






Number of stages











o m
O — i










. . . .

^H (M r<1 -^



Stoichiometric ratio











^ tM
00 -H



— « (M






^ (M xO

•— « CM cO ^
TJ
1)
"rt
Percent solids recircul
..„,„„








g
u
o o
o o
0 0
m o







1
u
0 0
0 0
0 0
ITi O

~N



+H
0
00
c
'o
o
u











CO










CO





Excess air
PT
sajqT



























u
O.
-H
rt
o
U
3H










o
Z










o





Coagulant
L










C
CM









C
CM
~*





Packing diameter

                                                                                                        on
                                                                                                        c
                                                                                                       X

                                                                                                       c
                                                                                                       H
                                                    720

-------
SI
B «
s|
£ 2
° n





B
"








Height of packing

PT^H
                                                                      an
                                                                      B
                                                                     .c
                                                                     B
                                                                     H
                                                                     *
                                     721

-------
is
C -M
^
CO
cu
§

cu
a
W
£
r Sensitivi
0
en
4)
cu
J
n
•M
c
cu
£
i-t
rx
w
• Factor la
0
CA
O

TJ

jj
1>

0
O
f~l
bC
0
O





in tn
£ £
O 0
O O
OO fM
QD W)
0 0
!H J-(
0 O



-i 00
VI
cu
g
o
o
X
00
o
(-1
4-1
o









cu
CO
CU
o
rd
a
CU
TJ







CO O
—i m



(-1 fM






CO










CO









Excess air







< CQ



-i fM



























g.
4J
•3
O
U







O 0



-> rg






O










c
o









Saturation sprays









0
Z










o
Z










o
z









0
u-
cu
"fj
o
oc
G
"o
0
u







o










o










o









-t-1
c
oJ
o
u

^,
£
0
a)
a
cn
Venturi &





a
o
.
nj
(-1
(X
Venturi &




V
o
>>
(Tj
tx
M
"G









Scrubber configuration
paxij
PT3H
                                            722

-------
en
C
g
JH
CO
^"
W
>x
• Sensitivil
o
{0
CO
CO

CO
"c
CO
SH
(X
Factorial E
IH
O
CO
CO
4)







r
re
£
0
OJ
V


Factors
to
_Q
Controlla









07
00









0)
(U
00
re
(M rO













w
re
-u
en
(M




CO
CO
re

Number of :








o
m









un m
r\] r-





-H rq








m o m o
rj m r- o

^ ^ ^ ^


o
re
u
'£
Stoichiomet








o









00 "-<














^f CO -H -H

-H (NJ fO Tf
TJ
"rt
3
u
o
tu

T3
Precent sol







1
u
o
o
o
o









u
0 O
o o
0 O
un O





-H t\]






g
u
0 O
0 0
0 O
LO O

i— i C.J




0)
re
o
n)
O








a
00
0
o









a
0 O
O IN]
0 -H





— i (M







£
a
0 O
O r\]

""
(T)

a.
a.
0
"re
^
0
o
cr








d
B
0









g
in o
— < sO





-H CNJ








C
g
m o
~-, -.£,

-.


OJ
£
OJ
u
CU
"H
H
W


















< CQ














<




cu
rt
"a
0
-u
0)
O
sajq-ei IBA A.-i'eujT j^








rj
CO









N r^
tN] CN]














CN oo

— ' CM


*
O
"S
^
GC
c
o
o





TJ «
O "^
a >*-
Clarifier ;
Clarifier;




^H (N]



•a
c
0
a
tn
 (M



















„.







OJ
a
U
sa^q-eij-eyv Aj-epuooag








O









c
O














0





rt
a.
Saturation s








o
Z









o
Z














o




S-.
o
3
CT
CU
*^i
Cooling of i








o
2









o
Z














0
Z







Coagulant








C
rj









C
(M













C
fN]





-------
     M h


     S 2
     •*H M






-    1' h
     M a>
w    rf -^
3    > o




     •O •"

     g S

     £"!"




     "3 ^3
CO
V
1

V
OH
)4
w
>.
|S
•"*



-« (M










- CM m Tf



'O
0)

^5
Percent solids recircu








1
u
o
0
0
o*







1
u
o o
o o
o o
m o
-* CO



f-4 f\]





g

O
o o
o o
O 0
in* o"


- (M







0}
O
<*-<
CO
O

83







g
U)
o
o








a
two
O O
o o
(M •*











1
DO
O O
O O


*" N
m

ro
ft
CO

£
0
41
ro
H
O
g-
j

iq«!"








c
1
O
sO








C
e
If) O



-H (M







q
e
in o


- CM






ii)
E.H.T. residence tirrn

A11W








C
sO








.3











C



DO
u
rt
ft
cu

O

Overflow weir height a

ud








<








< n






















Boiler injection plane










tM
CO








(M m



-i CM







CM m


— i CM







*
Flow configuration










0
Z








i
HO
§§
^ CM



r-l CM







1
O
O






2


CL
in
a
(U
re
In
&
O
tr








4)
§£
Limesi
Dolomi

^ CM





4)
Limes ton










ri
^imestoi










Additive type



CO
a>
£
O
o
CO
00
3
0
O


10 CO
S S
o o
o o
<*"> fM
bO bC
2 2
A j3
^ ^
o o



<-> CM
CO
V
S
o
rO
^
M)
y
0
JH
O










a>
N
CO
U
nJ
a
(U
.s

8TqB"T








-H CO

1-1 CM






00











CO










Excess air

A








< ffl

^ rg





























0)
a
o
U

"""








0 0

-H rxi






O











fi
o










Saturation sprays










in
0) 0








o
Z











0










Cooling of inlet liquor







C
0
ft
^
Clarifi*






•8
O
ft
Clarifier/







13
a
ft

>larifier ,
w




c
O


(_,
Solids handling configu










o








0











o
Z










Coagulant

F
sa^








C
CO








g
CO
in











C
00










Glass sphere diameter
pax^
PH
qi2T it








C








S











.c










00
c
u
rt
a,
Q
tUD
OJ
[
-
                                                                                    724

-------
     (1)  Limestone type A and B refers to limestone from two
         different quarry sites.
     (2)  Coal type A and B  refers to coal of two different  sulfur
         concentrations, i.e. high and low sulfur coal.
     [3)  Boiler injection plane A and B refers to two different
         temperature regimes,  for additive injection, within the
         boiler.
     [41  E. H. T. and P.  W. H. T. refer to the Effluent Hold Tank
         and the Process Water Hold Tank, respectively.
     5)  N. C.  refers to  a variable which is "not controlled. "
The flow configuration numbers  refer to specific flow configurations

for each scrubber system.  As an example, the TCA flow configura-
tion numbers 22 and 23 are shown in Figure 6 and 7, respectively.


Selection of Independent Variables


Generally, the independent variables chosen for a specific set of runs
were those variables considered to have a major effect on the cri-
terion (dependent) variables in question.  These judgments were based
on the limited existing pilot plant data and the results from mathema-
tical models which predict criterion variables as a function of the
independent variables (see Ref.  4).  Also,  the number of variables
vhich can be tested within a set of runs is limited by the  schedule time
limitations and the desired statistical "resolutions" (see  next section).
For the break-in  experiments, the dependent variables are SO?
removal, particulate removal, plugging and scaling of critical pieces
of equipment, solids separating capabilities (settling, etc. 1, length
of time to reach  "steady state" and system reliability.
                                  725

-------
©r
                                    J
                                     O
                                    Hz
                                    «i O

                                    «s
                                    w 2
                                    B S
                                    ^i2
                                    H In
                                    Z Z
                                    W O
                                    W U
                                    U
                                    S w


                                    si


                                    is
     x;

     <
     h


    is

    Is
    Z B!
    S w
    a H
    « <
    9*
    J UJ
    o y5
    w w



    ^
    ww
     a
       j
       <

       Q
       W
     OS

     O


     g

     3

    W H 
-------
             7;
           H
           a D
                II
                       J
                       2
           2 Z
           w o
           w U
           w
           BJ W
           a
           Q
    z OS    w
  s  3s  *  5
  2  ||  g  a

  5  §»  2  5
  S  08  J  „
    m u ., (., * a
      fiW'<;>i(Ui
    BS X S3 ^ U H <
    p g=> JH  .5
    " o, g a w 3 w
                              -s
           53
           -1 < H
           Q  OS O 5
           32<«<
           O g p W M
           OT H W ^ 'j
 KM W
 b< OS
W 2 0.
SOU
H U OS
    o c M ti <
    p p°.3^(
    h ^ Z ffl w ,

    ££££3
    2- Z ^ U S (
    wo?wg
                      PS H
                      o z
                      u. w
                    ° i s
                    ° s
             £ §
                 « 3 S
                             S -
             = -
             S 3
                            S 2 5
                  I I
                      -  i  1
                      S  •"  c
                 K g S 5  g
                          >

                          f
727

-------
For break-/in experiments with limestone in the  scrubber circuit
(Tables 4, 5 and 6), the designations of Group I  and Group II refer
to two different experimental sequences (designs^, based upon dif-
ferent groupings of the variables to be  investigated,  e. g. in the ven-
turi system, the limestone injection point is studied (varied) in Group
II and not in Group I.  All of the probable variables which may affect
the criterion variables have been included in these designs (Group I
and II).  For example:
    •    The E. H. T.  residence time (studied in Group I for TCA
         and Hydro-Filter) may affect the degree of supersaturation
         in the slurry system and,  hence, influence scaling.

    •    Addition of coagulant (see Group II, Hydro-Filter) can
         affect clarifier settling characteristics and solids build-
         up within the system.

    •    Solids handling configurations (see Group II, Hydro-
         Filter) can affect solids separating (dewatering) and
         handling characteristics.  Also, different  configura-
         tions can affect reliability and the duration of the  approach
         to  steady-state after system upsets.

    •    Stoichiometric ratio and the percent solids recirculated
         can directly affect  system plugging tendencies,  clari-
         fier settling characteristics,  SO^ removal and system
         overall chemistry.

    •    Excess  air (studied in Group  II, Hydro-Filter  can affect
         the degree of oxidation of sulfite to sulfate and, hence,
         the system settling characteristics,  scaling tendencies
         and SO2  removal.

    •    Plug position (Group I, venturi) can affect SG^ removal,
         particulate removal and scaling within the venturi.
                                    728

-------
For break-in tests with hydrate in the scrubber circuit and lime-
stone/dolomite in the boiler, (Tables 7 through 12) the variables to be
studied are fewer, since it is assumed that many of the operating
problems will have been solved during testing with limestone in the
scrubber circuit.


For the screening experiments, the  criterion variables are SO~
removal and particulate removal.  The variables considered to have
significant effects upon these dependent variables  are shown in Tables
4 through 12.  There are two types of variables considered in the
screening experiments:   (1) primary and (2) secondary variables.
Primary variables are those considered to have a major  effect upon
the criterion variables while secondary variables  are considered to
have lesser, though significant effects.  The screening experiments,
as mentioned previously, are of two types:
     (1)   Factorial tests where the primary variables and "selec-
         ted"  secondary variables are treated in a fractional
         factorial test matrix.

     (2)   Sensitivity tests where the remainder of the secondary
         variables are tested in an abbreviated test design.
Selection of Levels for Independent Variables


The levels of the independent variables have  been chosen using engi-
neering judgment, pilot plant test results,  the restraints of the  system
and results from mathematical models which relate the criterion and
independent variables.
                                   729

-------
In many of the experimental sequences it was preferable to vary cer-



tain controllable factors over their maximum design ranges.  For gas



flow rate,  the stable operating  range is from about 10, 000 to 30, 000



cfm in each scrubber system.  For liquor flow rates to the venturi



and after-scrubbers,  the controllable operating ranges  are from



about 100 to 600 gpm.  For the  Hydro-Filter bottom and top sprays,



the allowable limits of liquor flow rates  are from 100 to 600 gpm and



100 to 400 gpm, respectively.   For liquor flow rates to  the TCA, the



controllable  operating range is from about 200 to 1200 gpm.








As mentioned previously,  two of the stated objectives for the break-



in periods  are: (1) to provide information regarding  the limitations



of the system with respect to control of the independent variables



over their  "desired" levels and (2) to define the limits of the levels



of the important scale (or operating-problem)  related variables.



Therefore,  severe (or unfavorable) operating ranges have been  selec-



ted for  a number of the independent variables, e. g. stoichiometric



ratio and percent solids recirculated, during the break-in testing.








For the  screening experiments, the  variations in stoichiometric ratio



from 1. 25 to 1. 75 and from 0. 9 to  1. 3 for limestone addition and



hydrate  addition,  respectively, were based upon limited pilot plant



data, which indicated reasonable SO,, removals at these levels.   The



selected variation of percent solids recirculated from 8 to  12% was



also based on limited data, which  indicated that severe  scaling or



plugging would likely occur outside of these  limits.  Maximum ranges



of gas  rate variation (10, 000 to 30, 000 cfm)  and liquid-to-gas ratio
                                 730

-------
variation (about 10 to 60) were chosen in order to determine scale-up

effects  as accurately as possible.   Obviously,  the choices of the  values

of many of the variables  listed in Table 4 through 12 may have to be

modified once break-in testing has  been completed.


STATISTICAL DESIGNS


The test sequences (presented in Ref. 5) are, generally, partial  fac-

torial designs based upon the chosen  variables and levels and the

restraints of time as outlined in Figure 5.   For the screening (fac-

torial) experiments,  the  statistical designs are partially replicated

fractional-factorial sequences of Resolution IV   which have been selec-

ted with the understanding that an additional few experiments will be
                                                  ?'; ;'=:
run, if necessary, in order to achieve Resolution V   designs for the

significant variables.   It is assumed,  therefore, that the number of

"significant" variables will actually be less than the number of pri-

mary variables listed in  Tables 2 through 12.


TEST SCHEDULE


The total number  of scheduled test  runs  are given in Table 13 and the

estimated duration (line-out and "steady-state") of the runs in Table

14.   The complete test  schedule is presented in Ref.  6.
    A Resolution IV design is a plan in which no main effect is con-
    founded with any other main effect or two-factor interaction, but
    where two-factor interactions are confounded with one another.
    A Resolution V design is a plan in which no main effect or two-
    factor interaction is confounded with any  other main effect or
    two-factor interaction but two factor interactions  are confounded
    with three factor interactions.
                                   731

-------
             Table  13




NUMBER OF SCHEDULED RUNS IN



       EPA TEST PROGRAM

Test Program Function
Venturi
Air-Water 20
Sodium Carbonate 24
Limestone in Scrubber (Block #1)
Break-In 32
Factorial 40
Sensitivity 8
Undefined 8
Hydrate in Scrubber (Block #2)
Break-In 16
Factorial 32
Undefined 8
Limestone in Boiler (Block #3)
Break-In 19
Factorial 40
Sensitivity 10
Undefined 8
Number of Runs
TCA Hydro -Filter
20 20
20 20

30 32
32 32
10 10
8 8
16 16
32 32
8 8

19 19
40 40
10 10
8 8
Primary Testing To be defined
                      732

-------
P

W
J
D
O
u
                       o
                        C  t3
                        S  <«

                       H  5
                      oiracooitncnwtntntntnoiw
                      >^>,>,>,>x>N>,>,!>,>,>,;>,>,
                           —i
oj  O
O  O
O  H
m  m
H  y
<  H
                       a;  w
                       w 3
*  a,

 tn  tn

H
                     O   -H   —<
Q
W
H
en
W
                       0
                       s
                                                                                M   _    _    t!

T

e
fin
                                                                                733

-------
In Table 15, examples of TEST SCHEDULE SHEETS for TCA system



factorial experiments with limestone in the scrubber circuit (Block



#1) are presented.  The second column refers to the "statistical



design" table  and run numbers,  which have not been presented.



Generally, the test design for all three scrubber systems must be



"meshed, " so that the scrubber systems can run simultaneously.



Variables to be meshed (i. e. , variables which are common to all



scrubber systems) include limestone size,  limestone type,  and coal



type. In addition, stoichiometric  ratio must be meshed for runs



where limestone is injected into the boiler (Block #3).








The  s-ame conventions used in Tables 2 through 12 have been used in



Table 15.








ANALYTICAL SCHEDULE








Batch samples of coal, limestone  slurry and gas •will be taken peri-



odically  during each  test run for chemical analyses, particulate size



sampling and  limestone reactivity tests.  Locations  of sample points



have been shown on Figures 1,  2,  and 3.  In addition, clarifier settling



tests, filter leaf tests, and filter and centrifuge "operational tests"



will  also be conducted periodically,  during the break-in periods.  A



typical analytical schedule for break-in experiments with  limestone



in the scrubber circuit is shown in Table 16.  Normally, there will be



a separate analytical schedule sheet for  each  group or  sub-group of



tests.
                                    734

-------
  W

  O ~
  H 2
  w W
  W £~*
  B! <
W o u
§r
 i " ."
 IHH

g«8
OT § «
H OJ u
HX w
  WS
  H
  U
             6*

            3 IS
              .3
             
              U
            jfi
            £  P-
               ,
             § a
             cr rt
            "1 Pi
            •H  qj
                  HH
                                                 HH
                                                             H  H
                            H  PH
                                     K  ffi  ffi  ffi

                                     W  W  W  W
                                                       W  W

                                                       co  oo  ro  ro
                  no  ro  CM

                  ~-<  ro  to
                                       ro  ro  r\) ro
                                        i   i   i   i
                                       co  a*  o ~H
                                                    ro  ro  ro  ro  ro
                                                                                   .2 °

                                                                                   % ^

                                                                                   u 2 ;

                                                                                a
                                                                                >
                                                                               £
                                                                                a
                                                                               >
                                                                               U

                                                                               •D
                                                                               >
                                                                                  U J IX tn
                                                         735

-------
p ft
  c
  o

^1
0  .

•O -H

« s
ll
nj rt

O rf
o «
                                    HHHH
ffi  K


W  U
            £  S     ^  K  K  K  ffi  ^

            p,'  pi  CU  A  W  W  H  M  pi
                                             &  X  X

                                             A  W  W
                                                                         •0 a
                                                                         •3 o

                                                                          a X
                                                                         tn W
                                                                              .2 ™'
                                                                              3 g

                                                                              DO .5
                                                                              EH
                                                                                ro

                                                                              a ^ .c


                                                                              W J. CO
                                                                          ^_ oj aj ^ *

                                                                          3 .§ I S «
                                                                          U J J (X m
                                          736

-------
 c s as
 S B XS
M
• ii
!M
ill
  5
§88
Ssl
IS§6.
 a, > O Sj <
 ss 5 5 J
JZ25* I
   H
- 5

i 5
 w
2 S
Is
        o

        ti •

        £-2


        'It
= E "8 3
2 2 P

!l<
u i. X *J
8.8.^-fi
 C

_«>  .

II
 o n
^ "

Jj

*° p^
_£*•-!(
£ "8-

JP

jll
                                                                   «£ 2
                                                                   .  IU UJ
                                                                   3  2 s
                                                                   U  DC D
|o§
C   x

I   I










1 <^
i.'
I 	 "31SM
if
COAL
LIMESTONE
HVDRATE
PARTICIPATES aCUEB t.^E

s

™

POND RETURN ' IQUtC

NOWWOD
5*1
n
! PARTICULATES.FLJE GAS FEE
^1 i irt ^
PA«T,CULAI£S,SCRJB«ED GA
PARTICULATES,SCRU»MD GA
i
S
1^
o
z

s
^
iTii^
Q
?
i


SSiS,
3 > >•
"" at1 or
PROCESS WATER HOLD TANK
AFm-SCRUMER OUKET iLU
AFTER-iCRueMR OUTLET SLU
CLARIFIER UNDERFLOW"







' a 1^1
CIARIFIER OVERFLOW
FILTRATE RETURN LIQiX>R
CENTRATE RETUBN LIQJJOR
FILTER CAKE
1

>HniN3A
H
o
S
PAHTICULATES.SCBUMED GA


•5
ix-
PARTICULATES,SCIUMED GA
RECIRCULATING TANK INLET
, 3

!
> 3
ri
>- i-
RECIRCUIATING TANK OUTL
5
o
1
1
jjj
u



^
W
j
8
<
J
i




TCAQUILtl SilJRftY 14}
TCA OUTLET SIU«RV (5)




CLARIFIER UNO£RKtj»'
CLARIFttR OVERFLOW




FILTRATE RETURN LiCuOR




CENTRATE RETURN LIGUOP
FILTER CAkE




<
J


1
n

m
CN Csi

9


1
O
•?












PARTtCULATES.FLUE GAS FEE




PARTICIPATES, SCRUBBED GA

CO



9



01 '
fA«T(CULATES,SC«U«SfO GA
OVERFLOW WEIR OUTLET SLL*

s

1 11,
i- u O ' ^
SLAKING TANK O'HlET SLUR
PROCHSWATER HOtDTANK
PROCESS WATER HOLoTANK
a 311




HYOHO-FUJER OUTLfT StUflR
CLARIFIER UNDERFLOW
"CIARIFIER OVERFlOW
FILTRATE RETURf- .lQUG«
CINTRATI RfTJ*' -I1,|0«
FIL'TEB CAf-E
CENT»lfUG£ rAM
-O«OAH
CO

•*

r


                                       737

-------
In order to meet the formidable slurry,  coal, and alkali analytical


requirements of the facility at reasonable costs, equipment has been


selected that minimizes manpower.  For example, an x-ray fluores-


cence unit has been purchased, which  should allow for  comprehensive


slurry analyses with reasonable manpower  requirements.  A summary


of batch analytical methods for determining important species in


slurry, coal and alkali is presented in Table 17.





In the particulate sampling area, equipment will be available to mea-


sure mass loading in the scrubber inlets and outlets.  If the results


of an EPA sponsored program with McCrone Associates are success-


ful, this data will be  supplemented with particulate size distribution


data obtained with a multi-cyclone device.
Equipment has been carefully selected for gas analyses.  For


analysis, six "Du Pont" photometric analyzers will be utilized for con


tinuous operation at all three scrubber inlets  and outlets.   For HO,
                                                               LJ
    >  N   and Q£ determinations,  two "Process Analyzer" gas chro-


matographs have been purchased for intermittent operation at each


scrubber inlet and outlet as well as two "Environmetrics" NO


analyzers.  pH will  be monitored on a continuous basis using fifteen


"Universal Interlox" pH analyzers.  Three "Universal Interlox"


electrolytic analyzers will be used to monitor electrical conductivity.
                                    738

-------
                           Table 17

    FIELD METHODS FOR BATCH CHEMICAL ANALYSIS
         OF SLURRY,  COAL AND ALKALI SAMPLES
Species Desired

Sodium

Potassium
Calcium
Magnesium
Choride
Total Sulfur
Field Method
Atomic Absorption
  X-Ray Fluorescence
Total Sulfite and Bi-Sulfite
Dead Stop lodometric
Total Carbonate and Bi-
 Carbonate

Nitrite
Nitrate
Infrared Analyser
Ultra-Violet Technique
                               739

-------
DATA ACQUISITION

Over 150 pieces of data (flow rates,  temperatures, pH's etc. ) are to
be recorded automatically and continuously onto magnetic tape at the
pilot plant site.  At Bechtel Corporation, in San Francisco, all field
process data (including results of analytical analyses and manually
recorded  information) will  be placed into a "random access" computer
file.  The data will then be printed and plotted in suitable formats for
distribution and assessment by Bechtel,  EPA, and TVA personnel. A
computer program will average data during the "Steady-state" run
sequences,  identify possible  erroneous data points, etc.  Another
program will  "adjust" data which are functions of temperature and/or
pressure  and  convert certain data to more useful units.

DATA ANALYSIS

Statistical computer  programs  will analyse the data to  determine, for
the major dependent  variable (SO? removal),  the magnitude of the main
effects and the two-factor interactions for all the major independent
variables, as  well as the magnitude  of time trends, systematic
upsets, and random error.

Model predictions for the dependent  variables  of pressure drop, par-
ticulate removal, and SO_ removal (see  Ref.  4) will be compared
•with the data and the "best-fit" values for uncertain constants and
coefficients within the models calculated.  For pressure drop and
particulate removal, the form of the proposed models can also be
checked.  All models will be modified, if necessary, for better fit
to the  data.
                                    740

-------
                            SUMMARY

The Office of Air  Programs (OAP) of the Environmental Protection
Agency (EPA) has sponsored a program to fully characterize wet
limestone scrubbing for removal of sulfur  dioxide and particulates
from boiler flue gas.  Thes test facility consists of three parallel
scrubber systems, each capable of treating approximately 30,000
acfm of flue  gas,  which are integrated into the flue gas ductwork  of
an existing coal-fired boiler at the TVA Shawnee Power Station,
Paducah, Kentucky.  Construction at the TVA power  station is approxi-
mately 80 percent complete, with start-up presently  scheduled for
March, 1972.

Bechtel Corporation, as the major contractor, has prepared the
detailed design of the test facility, is  developing the test program and
will direct the test efforts.  TVA is constructing the facility and will
operate the unit during the test program, which is scheduled to last
30 months.

Three basic test periods have been defined:
    (1)   Break-In Tests
    (2)   Screening Tests
    (3)   Primary  Tests
                               741

-------
"Break-in" testing will ascertain operational problems and variable
control limitations and establish some attractive operating configura-
tions.

"Screening" testing will define the influence of process independent
variables on the dependent (performance) variables.  Results from
these tests  will be used to perfect the mathematical models, which
will then be valid  for optimization and commercial scale-up.

Primary testing will be used to ascertain conditions for maximum
performance of the scrubber systems and to demonstrate long-term
process economics and reliability.
                               742

-------
                      REFERENCES
Bechtel Corporation, Alkali Scrubbing Test Facility - Operating
Manual for the EPA Alkali Scrubbing  Test  Facility at the TVA
Shawnee Power Plant, October 1971

Universal Oil Products, Air Correction Division, Bulletin No.
608, "UOP" Wet Scrubbers",  1967

National Dust Collector  Corporation,  General Catalog,
December 23,  1968
M.  Epstein,  et al,  Bechtel Corporation,"Mathematical Models
for  Pressure Drop, Particulate Removal and Sulfur Dioxide
Removal in Venturi,  TCA,  and Hydro-Filter Scrubbers, pre-
sented at Lime /Limestone Symposium, New Orleans,  Nov-
ember 8-12,  1971

Bechtel Corporation, Alkali Scrubbing Test Facility - Statistical
Design of Experiments for Venturi, TCA, and Hydro-Filter
Scrubbers, to be published.
Bechtel Corporation, Alkali Scrubbing Test Facility-Test Pro-
gram Manual for the EPA Alkali Scrubbing Test Facility  at the
TVA Shawnee Power Plant, October 1971
                            743

-------

-------
ECONOMICS OF LIMESTONE WET SCRUBBING SYSTEM
              R.M. Sherwin
              I.A. Raben
              P.P. Anas

           Bechtel Corporation
             Prepared for
 Second International Lime/Limestone
      Wet Scrubbing Symposium
       New Orleans, Louisiana
        November 8-12, 1971
                 745

-------
 ECONOMICS OF LIMESTONE WET SCRUBBING SYSTEM
                      R. M.  Sherwin
                      I. A. Raben
                      P. P. Anas
                  of Bechtel Corporation
ABSTRACT

This paper describes factors influencing the costs of full-scale
limestone scrubbing installations.  Data on capital costs are
presented, together with process and mechanical design criteria
which most greatly affect these costs.  Among the latter are
corrosivity,  erosion,  scaling, local site factors, environmental
impact, weatherizing, front and back end logistics,  retrofit
limitations,  operating control philosophy.  A case history is
cited for an eastern utility planning retrofit scrubbers on a
multiple-boiler installation.  Operating costs are broken down
into their components,  and a range for each of these is given.
The importance of these cost considerations in planning programs
for compliance is examined.
    Prepared for presentation at the International
    Symposium for Wet Limestone Scrubbing,
    Nov.  8-12,  1971,  New Orleans,  Louisiana
                           746

-------
 INTRODUCTION AND BACKGROUND

 The Environmental Protection Agency, acting in accordance with the
 mandates of the 1970 Clean Air  Act,  has  set in motion the regulatory
 mechanism to achieve primary ambient standards  of air quality within 3
 years  following June,1972.  The states are busy preparing their  programs
 for implementing this time table,  and these are to be submitted to  EPA
 by February,  1972.  Many of the major urban areas have already enacted
 portions of these plans into local regulations which must be complied
 with by June,  1975.

 Major operators of stationary combustion equipment are concerned with
 the effects of this activity on their overall operations.   Not only  are major
 capital outlays indicated,  over and above those traditionally associated
 with fossil fuel equipment, but extra  operating costs as well.   Those of us
 who are following the progress of stack gas cleanup the closest would be
 less than frank if we did not recognize  that these costs are exceeding
 estimates made during the early,  conceptual stages of this program.
 Nevertheless, we should keep one thing in mind: the same factors  contri-
 buting to spiraling costs here are to some degree at work in the  other
 avenues available for emission control.  While of scant comfort to  a major
 utility, this should be somewhat reassuring to the  engineer charged with
 finding the lowest-cost means of compliance.  The  recognition of what
 these cost components are,  furthermore,  provides a basis for challenge,
 comparison,  and evaluation of progress being  made elsewhere in the whole
 spectrum of pollution-control technology.

 Early this year a major eastern utility retained Bechtel Corporation to
 prepare a study and estimate  of  fuel-burning alternates for one of its  major
 existing facilities.   Involved in the picture were four identical 180 Mw
 pulverized coal boilers burning  3-1/2% sulfur  coal.  Because of  scheduled
 availability only one unit could be converted at a time,  and it was necessary
 to make the conversion over a three-year period,  in time for EPA  deadlines.

 The  company was confronted with a number of possibilities,  including
 (a) high-sulfur coal with both  particulate and SO scrubbing, as well as,
 (b) low-sulfur coal with particulate scrubbing only.  The existing rquipment
was  laid out in a typically close-coupled configuration, raising many questions
 of ducting, fan placement, and tandem vs. booster  staging of fans.   The
capital  costs  for these two alternates are shown in Exhibits I and II  (slide).
It will  be noted that a system  for particulate removal only  involved about
 60% of the  outlay of a system  for particulate and SO together, using rear-
end limestone addition.
                                  747

-------
The following factors helped contribute in a major way to the capital cost
of this facility,  and these  should be particularly noted by anyone who leans
on generalized cost concepts,  such as the convenient (but undependable)
"0. 6 factor" approach for SCX removal in power plants.

Raw Material Handling.    Limestone was delivered by barge 9 months
out of the year.   This required a totally new conveying system, dedication
of area for stockpiling,  apart from the coal pile,  and displacement of
existing secondary service facilities.  Supply logistics dictated a relatively
high investment of this type to provide maximum long-term limestone
economy and reliability of supply.   Conveyors were sized for high-speed
discharge  rates to minimize conflicts  in scheduling with coal barge
deliveries. High reclaiming  rates with large 24-hour silos were indicated
to minimize the use of non-day shift
Under conditions of delivery by truck, this portion of the investment could
have been lower by $2/kw.

Limestone Milling.    A complete, integrated milling operation was planned.
The delivery of pulverized limestone by pipeline or truck from a central
milling point would have lowered costs.  However, the feasibility of such
an operation was felt to be conjectural without a more extensive study.

Should future study demonstrate the  practicality of off- site milling, the
investment in both milling and raw stone handling could be transferred to
a second party.

Scrubbing System.   The scrubbing  system is the heart of the operation and
incurs the major cost burden, being 60-65% of the total investment.  Some
of the design-cost factors which have confronted system designers are
listed below:

   (1)    Should a  booster fan be added in series with existing fan,
         or  should a completely new fan be specified?

   (2)    Should "pusher" fans be used, at the risk of eroding the fan
         blades, or should induced-draft fans be employed, with the
         possibility of high dewpoint corrosion and imbalancing?

   (3)    What corrosion mechanism is most  likely, and what material
         is most suitable for  fan blades following the scrubber?
         Stainless steel is costly and may be susceptible to stress
         corrosion.   Peripheral speed considerations limit the use of
                                 748

-------
      rubber-lined rotors.   Certain alloys of interest,  such as
      Corten, cannot be welded with safety.  Deposits of slurry
      may build up excessive unbalanced forces on the fan blades,
      requiring periodic washing.  Can washing be done on the
      line?

(4)    Are  reheat temperatures dictated by plume-drop considerations,
      (minimal reheat) or by dew-point corrosion factors (maximum
      reheat)?  Is the existing stack properly lined with impervious
      insulating material so as to accept a flue gas with dewpoint
      characteristics?

(5)    What is the most economical and lowest maintenance method
      of reheat —direct injection of combustion gas,  or steam
      coils ?

(6)    What is the most suitable corrosion-resistant medium for
      field fabrication and assembly?   For example,  a shop-applied
      soft  rubber lining appears to be the ideal protection for bi-
      sulfite  slurry environments.  (Field-applied rubber linings
      are generally more costly and less reliable than those applied
      under shop conditions. )  For large vessels, however, there
      is no choice but field  fabrication (of the steel),  which means
      field application of the coating.  Under such circumstances
      the cost of rubber linings is hard to justify.

(7)    How  far can one go in achieving economy of scale?  For
      present scrubbers, the size limit for a single train is thought
      to be about 400, 000 cfm.  There are a number of reasons for
      this,  including non-uniform gas distribution problems (e. g. ,
      if scrubbers are bigger, inlet ducts must be bigger)  and the
      limited availability of matching  fan capacity.

      The size of I. D. fans  will be physically limited by the maximum
      flow  and pressure drop of the largest boiler units now being
      designed.

      For ball mills,  there  is almost  no limit,  although at through-
      puts above 100 tons per hour (enough for 1000 Mw), there is
      greater concern for reliability.   With skilled maintenance,
      bearing and liner repairs can normally be taken care of within
      48 hours.
                                749

-------
   (8)   What should the philosophy be in terms of providing standby
         pumps, mills,  and scrubbers?  For example, one major
         installation includes seven large circulating pumps, although
         the theoretical number needed is only four.  Proposed EPA
         standards for new sources do not permit loss of control (i. e. ,
         as stated in the allowable Z-hour limit) more than once a year.
         This is a serious and costly constraint in terms of capital
         investment.

   (9)   Does damper design need to be positive shut-off, and if so,
         is double-dampering a reliable answer?  Double dampers with
         intermediate air purging are preferred by some, whereas
         others insist upon a heavy crane-supported gate.

  (10)   To what extent  is winterizing necessary?  Here  is an installation
         without winterizing (slide) and here is the same  installation
         with enclosure  (slide).  Closing in  of the equipment reduces
         exposure to freezing in small lines during down  periods  and
         also provides greater comfort during unscheduled maintenance.
         It may also encourage closer attention to the equipment, a
         matter of paramount interest during the early stages of equip-
         ment start-up.

  (11)   What type of instrumentation is most likely to satisfy conditions
         of power plant  reliability?  What are the response characteristics
         of fan-damper  combinations under  conditions requiring emergency
         by-pass?   During  such a period the operator must isolate 30 inches
         pressure drop  out of a system total of 40 inches  or more without
         pressurizing the boiler.  This problem relates directly to the
         aforementioned question of maximum sizing for  fans.  If you are
         using a booster fan in series with an existing I. D. fan on a
         retrofit job, the problem may be compounded.

  (12)   What method of solids thickening is to be used?  Does it stop at
         a pumpable consistency, or must filtration be employed? Can
         hydroclones or centrifuges be used without fear  of slimes build-
         up?  Are clarifiers the only safe answer?  So far we have en-
         countered no proposals which stop  short of complete ponding or
         clarification.

All the foregoing considerations were realistically evaluated and taken into
account in the aforementioned  study.  The results were neither excessively
conservative nor overly optimistic.  For instance, a successful working design
throughout  was  anticipated as a matter of fact,  despite the total lack of
operational experience.   On the other hand,  reasonably generous  sparing
was provided.
                                  750

-------
 Solids Handling.   One of the more difficult aspects of limestone scrubbing is
 how to handle the waste solids.  As typical wet filter cake,  these account
 for approximately 40% of the coal tonnages.  The rheological properties  of
 these materials have proven to be somewhat unpredictable.   In a scrubbing
 system where the alkali  efficiency is low, virgin limestone  comes  through
 unaltered,  a relatively high degree of dewatering will occur, the density
 and solids content will be high, but the cake may still slump.  At higher
 efficiencies dewatering is incomplete,  the density and solids content drop;
 on the other hand, the cake may retain its stiffness.  How to stockpile,  reload,
 and transport this material is a major challenge.  Liquidity and plasticity
 of the filter cake are important criteria for design  of such systems, and are
 variant.

 For the plant in question, ponding was conceived as the best method for
 disposal —at a point four miles distant.  (An alternate consideration involved
 total dewatering, folio-wed by day-time reloading, and truck disposal.) Rubber-
 lined pipe  delivered underflow from a battery of liquid cyclone thickeners out
 to the pond, and carbon steel pipe delivered the pond effluent back  to the
 process.

 One of the major cost elements here -was a decision to provide all of the
 diking required for the life of the pond in advance, rather than in small incre-
 ments involving a greater earthmoving total over the long haul.  The argument
 here is manifested in a discount cash flow (DCF) trade-off of operating costs
 vs. capital costs.  The capital costs shown in Exhibit I could be substantially
 reduced at the expense of increased operating costs in subsequent years.
 Diking,  incidentally, accounted for $3.40 per kilowatt, or about 35% of the
 total solids disposal costs.

 In summary, we have so  far  described a retrofit conversion job -which would
 cost $43 per kw if done all at once. Since that is impossible,  it is  necessary
 to add escalation factors.   These bring the cost to approximately $50/kw.

A  large-scale modern installation would hope to achieve economies of layout,
 scale,  and logistical planning such as to reduce all elements appreciably.
                                   751

-------
OPERATING COSTS

It is important to note the overwhelming effect which capital investment
has upon operating costs.  Exhibit III shows what may be expected for
operating costs of limestone scrubbing across a wide range of power plant
sizes (slide).   (This data is typical only,  and does not represent information
from operating records or projections of Bechtel clients. )  These are shown
in terms of cost per ton of coal, as these are  the units in which operations
people think.  Costs are divided into two categories  —those which are a
constant expense burden,  regardless of load factor,  and those which vary
with the amount of fuel burned.

Capital charges account for about 95% of the non-varying costs. As percent
of investment they may range from 12 to 15%.   We have assumed a figure of
14%.  Financial and accounting  people might object to calling this a non-
varying cost.  It is really an average cost, made non-variant by a process
of amortization.  A typical exercise  showing how such a figure is developed
has been presented by H. W.  Elder in a previous paper (1).

Prediction of maintenance costs is always a tricky business, since these are
traditionally developed from operating experience.   It must be  remembered
that  such experience cannot yet be brought to bear on this type  of installation.
 Chemical plants typically allow for a percentage of investment ranging from
less than 2% to over 4%.  A more rational approach  is the  "investment-year"
basis (2), a typical example of which is shown in Exhibit V (slide).   If the
experience of SO_ scrubbers proves equal to coke plants,  for example, one
might approach 2% maintenance after the fifth year.  For the purposes shown
we have been cautiously optimistic in assuming a figure of this magnitude.
This  compares with a  rate of 1. 4 to 1. 7% for many power plants today.

The  number of operators required, -while low, depends on  the size and number
of units and upon considerations of retrofit vs. advance-planning.

We have not attempted to assess non-varying costs on a "per ton of coal" basis,
as these change too much with size and capital investment.  However, for
each individual case one can approximate a unit cost, which will be somewhere
in the range shown.

Among the variable costs, the largest unit item, next to limestone,  may be
waste disposal if trucking is considered.  This is not included in Exhibits III
and IV because the station has access to ponding.  As previously noted, the
cost of amortized pond preparation ends up instead in the capital charge
account.
                                  752

-------
Power is a significant item for high-energy scrubbers  —approximately 4%
of boiler capability.  The cost of reheat may require 2-1/2% of steam
capability, unless the plant resorts to direct oil-or-gas fired injection.  In
this  case station efficiency is not affected, but fuels more costly than coal
are required.  There are other potential maintenance and  operating advantages
to this type of reheat —for example,  freedom from tube scaling.   The
ultimate trade-off still remains to be seen.  We have yet to  encounter a
regenerative reheat system, although the economics  are claimed to be
favorable.

We have seen that the capital costs of scrubbing for particulate removal only
run about 40% less than scrubbing for both SO  _and particulate removal.
Exhibit IV (slide) shows what the  combined effect of altered  technology and
reduced investment does to operating costs.  This demonstrates that the
reduction in operating costs for particulate removal only is  likewise about
40%.
                                 753

-------
COMBINATION OF PRECIPITATORS FOR PARTICULATE REMOVAL WITH
SCRUBBERS FOR SC>  REMOVAL
A frequently-asked question is:  "Are there cost advantages to an electrostat'
using a precipitator for particulate removal,  followed by a wet scrubber for
SO  ? "  Some of the disadvantages are:
   L*

   o    an excessive  space requirement, particularly in retrofit
        installations;

   o    the need for two separate waste handling systems.

On  the other hand, there may be some advantages, such as:

   o    a  reduction in capital cost (compared with 2-stage,  all-
        wet  scrubbing);

   o    an improvement in handling characteristics of waste solids,
        obtainable by blending of the lime solids filter cake with the
        precipitator fly ash.

From what we have just seen, if we were to leave off the second stage, i. e. ,
omit SO_  scrubbing,  we would achieve about 40% reduction in costs.  Looking
at it in  reverse,  omission of the first stage and retention of the second might
realize a  cost  reduction of $10-20 per kw which could be applied toward an
investment in precipitators.  This does not automatically establish a pre-
ference for 1st-stage electrostatic precipitators.  On the other hand, the
deliberate collection of dry fly ash for blending with filter cake may offer
important savings in material handling costs.  Exhibit  VI (slide) shows the
approximate relationship between limestone efficiency and the degree
of mechanical  moisture typically encountered in a filter cake.   The greater
the limestone efficiency, the higher the moisture content to be  trucked  away.
As previously  noted,  this material may be much more manageable in terms
of stiffness,  low plasticity,  etc. ,  than one of lower water (but higher
limestone) content.   The handling problem can be eased somewhat by blending
dry solids with wet filter cake,  but its feasibility depends entirely on local
circumstances.
                                  754

-------
OPPORTUNITIES FOR COST REDUCTION

There is great interest in knowing how truly representative these costs will
be for other  situations, particularly for new,  large-scale fossil plants at
the 500 Mw level and higher.  We are making no claims on this score, as
we believe each situation must be considered unto itself.  It is, however,
only being realistic to limit the installation of first-generation scrubbers
to retrofit situations.  These are at a distinct disadvantage in terms of:
economics of scale, cramped hind quarters, and protracted construction
schedules.

The following criteria deserve continuing investigation,  as they would appear
to offer the biggest opportunities for cost reductions.  Detailed study is
already underway in many quarters on these possibilities:

   o   Limestone mills - With good logistics and economics of
       scale, large modern power plants should not require more
       than one, or at most,  two mills for their entire supply.

   o   Use of lime -  If scaling problems can be solved, the potentially
       greater efficiencies associated with lime will reduce the
       tonnages of raw material and waste solids.  This could open
       the door to over-the-fence supply economics and do away
       with milling requirements altogether.

   o   Elimination of clarifier - Some day, when process constraints
       have been more fully appraised,  it may be found possible to
       by-pass the clarifier.  This was proposed many years ago
       by ICI-Howden,  but the conditions for its success have not
       been fully demonstrated.

   o   Plant layout - Progress in reduction of cost by appropriate
       planning offers hope for appreciable savings.  Structural
       work  and duct fabrication are major cost components so
       affected.

   o    Pond  management - Much remains to be done  on the flow
       classification  properties of the various process slurries
       obtainable under different conditions before ponding practices
       are optimized.  These findings will help dictate future operating
       practice.  There is precedent in the ore beneficiation field,
       where the management of pond tailings has been developed into
       a fine art.
                                   755

-------
                        REFERENCES
(1)     "Economics of Limestone - Wet Scrubbing"; Elder, H. W. ,
       International Symposium on Lime/Lime stone Scrubbing
       for SO  Control, NAPCA, Pensacola,  Florida, March
       1970.
(2)     "Predicting Maintenance Costs"; Chem.  Eng. ,  July 13, 1959.
                               756

-------

ffi Q CO
H H Z
£ H 0
^2£
H W n
CO O Q
O pj Q
U On <


O O O O
O O -
•*


                      H
                                    H
                                    CO
                                    o
                                    U
vD
        o
        -*
                ro
                in
                        in
                             xD
                             00
                                                                                                    oo
        O
                Q
        CQ      H
        «  O  ^
                Pi
x
w
H  r.   I
Z  H  H
O  2  H
H  1=)  p{
co   i  II
WH^
a  J
3  5
 •H
 n)
 r—I
 U
 (L)

 P5
                                             bo
                                             ni
                                             h
                                             O
                                             -M
                                             CO
                                             n)
                                             o
CD
C
O
^_>
CO
cu

                                                     0)
                                                     a
                                                     o
                                                     -M
                                                     CO
                                                     0)
                                                     CO
                                                     bo
                                                     C
                                                             O
                                                             CO
 co
 O
 ft
 CO
•H
Q

T3
 fi
 O
                                       QJ
                                      -4->
                                       o!
                                                                                   o
                                                                                  •H
                                             C
                                             O
                                             rt)
                                             O
                                             CD

                                            H
                                            in
                                                  a
                                                  o
                                                  CO
                                                  C
                                                  O
                                                 O

                                                  bO

                                                 •H
                                                                                               CO
                                                                                               cu
                                                                                               In
                                                          757

-------
 o
 u
                rt
                w
                ft
                      ffi  Q  w
                      H  W  '
                      g  H  O

                      H  rt  £
                      co  O  Q
                      O  rt  Q
                      U  ft  <
                              O
                              U
 CO
         o     o
         m     I-H
          •       •
         •*     in
         CM
                                                     PO
                                                     O
                                                                          00
                                                                                 00
                                                                                       00
B

EXHIBIT


*&»
A o ft
>BING ON
IT RETR'
E-YEAR
S 2 w
P D oJ
H
H
U
>—i
H
tf
 bo
 fi
•H
^
 0)
i-H
CO


 
CO


 bo
                                     O
                                     0)
                                    CH
                                     (U

                                     fi
 fi
 V
 o
 a
 o
O

 CQ
T)
•H
I-H
 O
CO



 0
 0)
                                            CO

                                             bO
       "§
        ^
        U
       CO
                n)
                co
                O
                ft
                CO
               •H
               Q
                                                    o
                                                    ft






«— t
T3
CO
4-1
n)
^
O
fn
ft
0)
X!
O
4->
CO
rH











fl
o
•H
_j._\
aJ
r— \
n)
o
CO
W
fl
o
• H
-4->
O
53
h
-»j
CO
ti
O
O
ao
a
•I*H
f-i
2
T3
4->
CO
0)
^
a>
1
                                                                                in
                                                    CO
                                                     758

-------
              o
              a
              o
              H
              f-t
              v
              OH

                                                                        i   S     G  N
                                                                       °   
w 1
T °
1-3 0
CO
H i
W ' .--
!> 0 r\
G COSTS FOR "V
PLANTS IN 20
- WITH PONI
§ g
H^
HH
< PQ
rf pq
W &
n rr*
I-M PM
O U
CO
4->
a
cu
Jj
CQ
CO
4j
rt
cu
Requirem
T3
CU
s
J3
CO
CO
<
V
>
fi
percent i



•^
^H

4->
C
0)
&
J *
CO
cu
>
rt
percent i



ro







CO
0 manhour
o
o
o
i— i
i
m





CO
0 manhour
o
o
o*
CO
i
in
i — i





n
MM/year
*J3
•
^
t
i— i
^
CM
•be-
CO
H
co
O
U

w
J
ffl
 CO
 N
XI  fi  *H
 U  cd  O
    C  -*-1

al  2
±J  c  o
 a --j  JD
 ni J5  ni
U S  J
                                 c
                                 o
superv
 fn
 O
^3
                                      Cti
                                      4->

                                      O
 be
.s   i
                                 cu
                                 ex
                                                   c
                                                   o
                                                   H
                             M  g
                             cu  °
                             cxU
                                                   •
                                                   CO  O
                                                   o
                                                   U
                                      X  m
                                      ga
                                                   co .T5
                                                   co  U
                   •-i   o
                   -M
                   C   d
                   cd   o
                                                           o  in  o
                                                           v£)  O  ^O
                                                                     ro
                                                                                               m
                                                                                               in
                                                            S
                                                            o
                                                           H
                                                               cd  xl
                                      cu  cu  cu  v
                                      ex  ex ex  ex,

                                            vO
                                        moo
                                        .-H  O Tt
                                     CO   •   •   •
                                     -eq- -te- -te- -to-
o in     o
ro co     o
 •   •   •   •
       o
       o
                                                                        nJ
                                                                        O
                                                                       O
                                                                                     rt
                                                                                     o
                                                                                     H

co
H
CO
O
U
CO
W
J
CQ
3
rt
<


HI
^
F)
O
4J
CO
cu
a
i3






ater
ower
•£ 0,
cd
0)
(H
Xi
cu
^i
0
o
o
r— (
:eam
co



^i
n)
4->
O
J->
JD
cH

                                                                                     O
                                                                                     H

                                                                                     Q
                                                                                     tf
                                                                                     O
                                                                                                             0)   0)
                                                                                                             cx
                                                                                      m
                                                                                      in
                                                                                                             n)
                                                                                                             CO
                                                                                                             O
                                                                                                             CX
                                                                                                             CO
                                                                                                CO
                                                                                                ni
                                                                                                                 O
                                                                                                                 a
                                              759

-------
           a

           H "3

           ^O
           
_ CO CO
percent i
manhour
manhour

0 0
o o
o o
co r-* o*
' CM
ro i
O
i— i







II
f-t
ni
0)
•^
5

i
2
o
2




CO
cu
60
^
n)
41
O
,— 1
ni
S
AH
ni
U









iance
0)
4->
a
••-<
n)
s









t-
O
4->
n)
fH
o
42
n)
J

rvision
0)
£
CO
4!
t-t
O
43
ni
i— i
bo
•4-)
nJ
^
V
A,
O

                                  tn
                                            a
                                            o
                                            H
                                            ts's
                                            o
                                            u
                                            T3  CQ
                                              c
                                                     d +s  fe  «j
                                                    .B  n)  o  a>
                                                                    n)
                                                                   •4->
                                                                    O
                                                                   -4->
                                                                    I


                                                                   •§
                                                                   CO
                                                                            H
                                                                            O
                                                                            EH
                                      0
                                     £
                                                             w
                                                                            O
                                                760

-------
             EXHIBIT  V

     TYPICAL MAINTENANCE
         FOR COKE PLANT
 4—Investment-oge, million dollors x yeors
80
70
60
50
40
30
20,
10
 Cl  I   I   I   I
                  I  1  I   I
                                Fig. 2
                               I   l   I
              100            200        280
         Annual maintenance, thousand dollors
   M    =    0. 004  I x t  -  83, 100

  where   M   =    annual maintenance
            I    =    investment
            t    =    age,  years
                                   Source:  Chem.  Eng.
                    761

-------
                        EXHIBIT  VI
EFFECT OF CHEMICAL REACTION ON THE TRANSFER AND^TORAGE OF FILTER CAKE
      80-
CQ

z
<
fe
o
to
 O
 to
      60-
      40
                               (dense, fluid material)
                                                   (bulky, friable material)
                   20         40         60          80


            % CONVERSION OF LIMESTONE TO REACTION PRODUCTS
                                                                100
                              762

-------
                     EXHIBIT   VII
                     ESCALATION
                                                 % annual gain
                                                          15
                    Skilled trades
                 ENR-20-cities avg. rates for bricklayers.
                 carpenters, and structural ironworkers
1966
71     72      73
                                                 Source:  ENR
                                763

-------

-------
      THE RC/BAHCO SYSTEM FOR
     REMOVAL OF SULFUR OXIDES
   AND FLY ASH FROM FLUE GASES
              By

          J. D. McKenna
          R. S. Atkins
 COTTRELL ENVIRONMENTAL SYSTEMS, INC.

   A Division of Research-Cottrell

       Bound Brook, New Jersey



           Prepared For:


 Presentation at Second International
Lime/Limestone-Wet Scrubbing Symposium

        New Orleans, Louisiana

          November 8-12, 1971
                 765

-------
INTRODUCTION






The first widespread application of a flue gas desulfuri-



zation system appears to be underway.  There are a number



of variations on the theme but basically it appears that



a throw-away calcium based scrubbing system offers the



most feasible alternative at this point in time.  Research-



Cottrell in the belief that no one system can meet the



needs of the total market has endeavored to provide a



number of systems.  One of these systems which utilizes



a packed tower has been previously described at this



symposium.   A second and completely distinct system



now being offered by Research-Cottrell under license



from Banco will be reported on in this paper.







In 1964, AB Bahco of Sweden initiated investigation of



sulfur dioxide control and after preliminary screening



studies, decided to develop a calcium based scrubbing



system.








In 1966, AB Bahco installed a 1400 scfm pilot unit at



their central heating plant.  This facility was operated



for three heating seasons, and studies were conducted on
                             766

-------
hydrated  lime, burned  lime  and  limestone.   Results


obtained  during this operation  were previously  reported


by Gustavsson.




The first commercial installation of the system went into


operation in November, 1969, at So'dersjukhuset, a large


hospital  in Stockholm.  With the success of the first


unit, two additional units  were installed at that site.


Each unit accommodates an oil fired steam boiler with


a capacity of 33 metric tons of steam per hour.  The

                                        H
first slide shows the  installation at Sodersjukhuset.




Gustavsson's  paper presented at the 1st International


Lime/Limestone Symposium indicated that this system


was significantly beyond any other in its stage of


commercial development.  That symposium fostered the


Research-Cottrell - Bahco relationship which culminated


in their recent license agreement.  In order for Research-


Cottrell to arrive at a decision to proceed with the


licensing negotiation, a rigorous techno-economic and


market evaluation of the Bahco  SC>2 removal system first


had to be executed.  Portions of the techno-economic


evaluation will be reported on  here.
                           767

-------
PROCESS DESCRIPTION

Figure 1 illustrates the major functions of the RC/BAHCO
two-stage scrubber.  Flue gas is supplied by an I.D. fan
with the typical draft control mechanism.  A secondary
damper, which admits make-up air and automatically main-
tains optimum gas flow in the scrubber during varying
boiler operations, is connected to the vacuum  side of
the I.D. fan.

Flue gas is forced into the scrubber inlet  (1) where it
is distributed and directed against the surface of a
hydrated lime slurry.  There the flue gas reverses
direction and enters the first venturi stage  (2).  The
gas liquid contacting creates a vigorous cascade of
droplets.

Due to the large flue gas velocity in the column inlet,
the droplets are transported through the venturi scrubber
at a high concentration and turbulence.  The droplets are
separated from the flue gas in a centrif'igal force drop
collector (3).  Liquid is returned to the first stage
contact zone and the gas passes to the second stage
venturi  (4) where the gas-liquid contact process is
repeated.
                              768

-------
In the second stage drop  collector  (5),  the  gas  becomes



free of droplets and exits  through  the chimney  (6).   The



scrubbing  liquid  (13)  is  recirculated to the first  stage



contact zone  (14).  The scrubbing liquid is  milk of  lime



prepared by dissolving slaked  lime  in water.








The slaked lime is fed from a  bin  (7) by a screw conveyor  (8)



through a mixer  (9) into  a  dissolver  (10) .   From the



dissolver, the milk of lime is pumped to the second  stage



impingement zone (11)  where the  liquid height is controlled



by a level tank  (12).  The  liquid overflow is recirculated



to the mixer.







The lower scrubber stage  is provided with partially  spent



slurry from the two drop  collectors.  The overflow  returns



to the dissolver, which is  also  the level tank for the



first stage.  A regulator keeps  the level constant  in



the dissolver by supplying  fresh water.







A fraction of the return  flow  from the first drop collector  (16)



is continuously separated using  a concentration  regulator  (17)



into a thickener (18)  which automatically feeds  viscous



waste sludge into a large basin  (19).  Sludge then can be



pumped to a storage truck for  transport  to a refuse  dump.
                            769

-------
Alternately, the sludge can be pumped to a settling pond



or disposed of directly; another option would be filtra-



tion of the clarifier bottoms to achieve a higher solids



content before trucking.








PERFORMANCE





In December, 1970, a team of Research-Cottrell's technical



personnel visited Sedersjukhuset in order to observe and



test the Bahco SC>2 removal system.  At that time, only



Unit #1 was installed and operating.  The main objectives



of this visit were verification of the Bahco S02 removal



claims and also observation of the system's operational



reliability and simplicity.  Particulate removal testing



was not given priority at that time since Research-Cottrell



believed that based on its extensive venturi scrubber



experience it could reliably predict the particulate



removal efficiency.







On-site observations of the unit were made for a three



week period.  During this time, the unit operated in a



trouble free and reliable manner requiring very little



attention.  The scrubber operation was attended to by



the same operator who serviced the boiler.  His main



task with respect to the scrubber was that of measuring
                            770

-------
the pH of the outlet slurry to the sedimentation tank.



This he did hourly and this task required 10 minutes



at the most.








Lime delivery was observed twice.  The lime was pneumati-



cally delivered from the truck through a rubber hose to



the lime hopper.  This task required about thirty minutes



and was wholly performed by the truck driver.








Waste removal was observed once.  The sludge was pumped



to the removal truck and this task was performed again



solely by the truck driver.








An interview of the hospital's chief engineer indicated



that the greatest problem encountered in the first



season's operation of the Bahco scrubber was that the



dampers and flow measuring devices became inoperable



at the very low ambient temperatures prevalent  (-25°C).



The sludge handling portion of the unit, therefore, had



to be insulated.  At the end of the first season's



operation (i.e. seven months), the unit was still operable



even though there was about one inch of build up



on the wall.  The end of the season cleaning was executed



manually by three men in thirty-two hours.
                              771

-------
Assessment of Performance



While at Sfldersjukhuset, CES personnel measured overall


SC>2 removal efficiencies of the Banco system.  These


tests confirmed the efficiency claims of Bahco.





Due to the unseasonably mild weather enjoyed by Stockholm


residents during the first three weeks of December, the


boiler was only operating at 30% of full capacity.  The


average steam production was 10 metric tons per hour at

                        ^
a pressure of 24.0 Kg/cm-  (341 psi) and temperature of


360° Ci  Heavy #4 oil with a sulfur content of approximately


1".5% was consumed at an average rate of 800 liters/hour.


The scrubber was operated at an average pressure drop of


400 mm (15.7" water) and a pH of 5.5-6.0.  Lime consumption


was about 70 Ibs/hour. Make-up air, introduced via a


calibrated mechanical damper, was mixed with the flue


gas before entering the scrubber to insure a constant


liquid to gas ratio.




The gas stream was sampled simultaneously at the inlet


(before make-up air is introduced) and the outlet of the


scrubber.  The gas was absorbed in 150 ml of 3% hydrogen


peroxide for eleven minutes at a rate of 0.4 cfm.  A


suitable sample was extracted and titrated with 0.1 N


sodium hydroxide to a methyl-orange endpoint.  Three
                             772

-------
different tests verified the Bahco claims of high SC>2


removal efficiency.  These tests are shown in Table 1.


A comparison of these results with Bahco data are shown


in Figure 2.  It must be noted that the values obtained


for SOj concentration at the inlet are actual measurements


of SC>2 concentration coming from the boiler and need to


be adjusted for the dilution of fresh air.




ECONOMICS



A plot of RC/Bahco system capital costs versus gas


volume is shown in Figure 3.  Represented here are the


estimated bugetary selling price for an installed


module; thus, all equipment shown in Figure 1 are


included.  This, however, is not a turnkey cost; and,


as such, does not include any unique installation costs


such as interconnecting duct work, utility connections,

                                     *
remote instrumentation, etc.  Turnkey costs are often


significantly higher than installed costs for historical


air pollution control applications.




Operating costs for a single case have been provided in


Table 3.  Here it is shown that for a 40 megawatt unit,


the annual operating cost would be about $316,000.  The
* Total installed system cost.
                              773

-------
largest single item is the lime cost at about 57% of
the total.  The second largest is the power at about
13% of the total. The corresponding material balance is
shown in Table 2.

Figure 4 illustrates the annualized cost of operating
a RC/Bahco SC>2 scrubbing system for various flue gas
SCU concentration levels and unit sizes.  The scrubbing
annualized operating costs are also compared with an
assumed cost of a fuel switching option for meeting
pending SC>2 legislation requirements.


SUMMARY

In 1970, Bahco licensed their system for the Japanese
market.  A number of installations are already in
operation in Japan.  A list of the systems sold to
date are provided in Table 4.  The last group of
slides show the installation at Yoshinaga.  This
installation consists of three units each with a gas
handling capacity of 44,000 scfm.  The units are
installed on oil fired boilers and employ sodium
hydroxide scrubbing.  The reaction product is used in
the pulping process.


At this time, Research-Cottrell intends to apply the
                            774

-------
system to fossil fuel boilers with capacities in the



range of 60,000 to 500,000 pounds of steam.  It is



anticipated that once Research-Cottrell is fully geared



up to respond to the market,the delivery time will be



nine months.  Discussions are presently underway with



a number of companies for application of the system to



both utility and industrial boilers.
                           775

-------
REFERENCES
1.  Gleason, R. J., "Limestone Scrubbing Efficiency of
    Sulfur Dioxide In A Wetted Film Packed Tower In Series
    With A Venturi Scrubber", Paper presented at the
    Second International Lime/Limestone-Wet Scrubbing
    Symposium, November 10, 1971, New Orleans, Louisiana.

2.  Gustavsson, C. A., "Bahco SC>2-Scrubber CTB-CTK 1.5
    Pilot Plant Connected to Bahco's Central Heating Plant",
    Paper presented at 1st International Lime/Limestone
    Symposium, March, 1970, Pensacola, Florida.

3.  Gustavsson, C. A., "Bahco SO2-Scrubber, Commercial
    Installation at Sfldersjukhuset, a Swedish Hospital
    in Stockholm", Paper presented at 1st International
    Lime/Limestone Symposium, March, 1970, Pensacola,
    Florida.
                           776

-------
(M IM
O * IM
CO H
M
0) V4 r+t
£3ol

g
to a
H
4J
4) •
H a
8"
S** •
W Ml
D C 5o
OS H H
CO

K
a
o
M
a a
M W
w tj

H O -S
3 O ij
io
OIO
I 4
SH X-H
10 M
C -P
0 4)
HS
:Ti
S"o»
ft S4
(1)
•p
Q

at

i»«
4J
M
VO
0
o
o
*
M
in

o
M
VO


00
f)
O
H
H


00
•
in

o
H

O
M
91
O
3
o
0
H


O
C4


O
X
on
X
fM
H

M
O\
O
0
in
*
N
in
in
0
Ot
VO


00

o
o
tH


^
•
VO

O
o

o
(N
9l
O
fl
O
0
H


O
(N


O
X
a\
X
r>t
H

<•>
VO
e\
o
o
VO
H
ft
N
•
in
in
VO


<*

o
o
H


O
•
VO

o
H

O
OM
*""
O
en

-------
                              o
                              o  o
                              o  ir
                               «•  N"\
                              •sr  t—
                              o
                         oo
                         oo

                         LO
           O

       UJ  T
       <->    I
CSI
       
-------
   Utilities

 Power

 Make-up  Water
                               TABLE 3

                     Estimated Annual Operating Costs

                                 For

                        40MW RC/Bahco S02 System
Quantity
Required
600 KW
40 gpm
Cost
$/Unit
0.008 $/KW hr
0.3 $/l,000 gal
Annual Cost
$/yr
42,400
6,300
                                                                      (1)
   Chemicals

 Lime  (91%  CaO)
0.935 tons/hr    22.0 $/ton
                            (2)
180,000
  Other Operating  Expenses

Operators             0.5  man/shift     4.0  $/hr

Supervision  (25% labor)

Maintenance  (3% of capital  cost)

Direct Overhead  (75%  of Labor  & Maintenance)

Taxes and  Insurance (2% of  capital  cost)
                                      17,500

                                       4,400

                                      20,200

                                      31,600

                                      13,500
                          Estimated  Operating  Costs  	$315,900
 (1)  Assumes  8,760  hours/year; does not include waste disposal
      cost or gas reheat cost.

(2)   Includes  delivery; i.e.,  lime trucked approximately 50 miles.
                                 779

-------
                 CD
                 2TH;


                 P3U-
                 oa.1-
                 oo:
                 CO
           rn
           o
                                     in  zc
                                     o  CD
              CSJ
             ^^>
             rc
             o
                 OCD
                  «
                       O
                        <
CO
i
<_>UL
  C_-
I— c/:
                        CD  CD   O  CD  CD  CD  CD=>  CD
                        CD  CD   CD  CD  CD  CD  oCD  CD
                        I—  rv.   CNI  CD  CD  hn
                         ^    S    X    X    S   K

                        iv,  cr   cr  CD  CD  oo
ce:


c§°
CO

o
^
«a:
CQ

uu
O



CD
                    CJ
                    CD
       O
CD
  CD   oo
       CO  00  >-   O
         oz  0  CO       h-
       OO   •—i  >—•       •—.
                        CO
                        or
                        LU

                       ,§
                        CO
                                               oo
            CO  00
            CD
            rr:
CD
01
00
<_)  1-H
    rn
-
                                                   
-------
781

-------
 FIGURE 2
-scrubber
   782

-------
                                    FIGURE 3
0&
i5
                                    PLANT SIZE,  SCFM
                                          783

-------
                                     CM
                                    I ^v.
                                    I CD




                                    \i
                                     a:

                                    I ^
                                    I O


                                    I <

                                    i *^
                                  00
784

-------
THE BISCHOFF PROCESS—INITIAL RESULTS FROM
     A FULL-SIZE EXPERIMENTAL PLANT
      Dr.-Ing. Gerhard Hausberg
         Gottfried Bischoff
          Essen, Germany
          Presented by:
         Ulrich Kleeberg
            Prepared for
 Second International Lime/Limestone
      Wet Scrubbing Symposium
       New Orleans, Louisiana
        November 8-12, 1971
                785

-------
The Bischoff Process  -  Initial results
  from a full-size experimental plant

         by Dr.-Ing. Gerhard Hausberg
For reasons well known it is advisable
to use experimental plants of the greatest
possible size to gain knowledge on the
reactions and transitional processes taking
place in heterogeneous mass flows.

The flow rates through such experimental
units should be approximately comparable
to the partial streams passing through
each of the several units that are connec-
ted in parallel in high-capacity commercial
plants. Taking, for example, six scrubbers
connected in parallel for an 000-MW boiler,
than each group will have to handle a flue
gas rate of approximately 500 000 m5NTP/h.

Scrubbers of the same design for flow rates
of this magnitude and even of up to 10  m /
NTP/h have been in use for several years in
the iron and steel industries.

The tests we carried out in 1968 and 1969
in a pilot plant designed for a maximum
throughput rate of 5 000 m NTP/h were from
the outset regarded as preliminary tests
only. There was agreement that whatever
results would be obtained from this pilot
plant they certainly could not be used
without further experience as basis for
engineering full-size commercial plants.
                786

-------
An initial report on the resuls obtained
with the pilot plant was given in October
1969.

Although the informative value and use-
fulness of the data derived from small
scale plants are often higher than what
one would expect in view of the equipolent
influence of several laws of similarity,
the only thing it was hoped could be
gained from the pilot plant was an ade-
quate justification for planning and
building a much larger experimental unit.
Since the results obtained from the small
version were indeed encouraging, the
decision was made, early in 1970, to go
ahead with a full-size experimental plant
for a flue gas throughput of 140 000 m^NTP/h.
This plant, which was installed in the
Steag power station of Kellermann at Lunen,
West Germany, went into operation at the
beginning of February 1971.

The large scrubber unit is, basically, of
the same design as that of the small version,
which had been dismantled in the meantime
and which is now being used in a steel
plant. Owing to the fact that the full-size
unit is located immediately after the new
350-MW boiler of the Kellermann power
generating unit, minor changes in the
dimensional proportions had to be made.
The throughput ratio between the two
units is approximately 30:1.

Pig. 1 gives a size comparison between
the two experimental units.

               787

-------
The scrubber of the large unit is again
a two-stage design with centrally dispo-
sed spiral jet nozzles, some of which are
of the double-sided type. The raw gas
pipe has initially been connected to the
boiler at a point upstream of the electro-
static precipitator and enters the primary
scrubber stage at the top. The flue gases
flow at first only through the primary
scrubbing and cooling zone which has no
internals other than the spiral jet
nozzles.

The second stage is accommodated in the
same circular cylindrical housing. It
contains an adjustable annular gap
washer. The two stages are separated
from one another by a sloping plate.
The scrubbing fluid for the second
stage is injected through a spiral jet
nozzle also arranged in the centre of
the housing.

Downstream of the annular gap washer,
the gas flows through a water separator
of the axial-flow rotor type whereby
the residual water present in the form
of droplets is removed by centrifugal
effect.

The induced draft fan arranged at the
downstream end of the system is provided
with variable-pitch guide vanes. In
combination with the adjustable annular
gap washer it permits the system to be
operated at a constant pressure differen-
tial while the flue gas rate varies. The
clean gas leaving the scrubber passes
through the clean gas flues to the power
station stack. V/ith the system operating
                 788

-------
at full capacity, approximately 10 per
cent of the flue gas volume discharged
from the 350-MW "boiler passes through
the scrubber.

With the present system layout, all of
the water required for the annular gap
stage is taken from the overflow of the
settling tank. This amount is withdrawn
from the collecting tray of the secon-
dary washer and passed to the two bottom
nozzles of the primary scrubber.

Of the remaining amount of water needed
for the primary stage, only that quantity
required for the annular gap stage is
pumped to the settling tank. The larger
portion is taken from the collecting
tray of the first stage and directly
injected again into the primary scrubber.
In the tests still to be run, the layout
of the' water cycles will be varied in
several ways, although at the present
stage no details can be given. An
elevated circular cylindrical vessel
with central inlet and conical bottom
section is used for clarifying and
settling.

The next illustration (2) shows the
general layout of the system with its
water circuits.

The sludge accumulating in the conical
discharge section of the settling tank
is pumped to the dump area by a slurry
pump via a plastic pipe over a distance
                789

-------
of abt. 700 m.

Before the tests were started, three
wells were drilled in the area where
the sludge was to be dumped and the
ground water was analyzed in its ori-
ginal condition. After the tests were
begun and sludge dumped in this area,
the ground water was checked at regular
intervals.

Like in the small-scale plant, dry
pulverized white lime is still being
used as alkaline additive in the full-
size unit. The material is introduced
through a vertical pipe  arranged cen-
trally in the raw gas inlet. A propor-
tional belt weigher located at the lime
bin serves to adjust the amount of lime
fed into the system.

After the full-size plant was put into
operation, the lime-to-SOp ratio was
at first adjusted as far as possible
to the values which experience gained
with the pilot plant had shown to result
in an SOp removal efficiency of around
80 per cent. The flue gas rate was
varied in the range from 90 000 to
133 00 m%TP/h and the loss of head
accross the annular gap washer held
between 200 and 250 mm WG. So far,
the system could not be operated at
the maximum flue gas flow of 140 000 m^
NTP/h as this would have led to a loss
of head exceeding the maximum allowed
value of 250 mm WG. Pull-load tests will
be run, however, after making a minor

                 790

-------
modification on the annular gap washer.

If we enter the values obtained from
the full size plant in the SOp removal
diagram plotted for the pilot plant,
then, as we see from the next slide (3)»
the values already known are largely
confirmed by the full-size plant.

In this connection, it should be noted
however, that the raw gas from the
350-MW boiler contained between 3,0
and 5,0 g/m^NTP of S0p, that is, about
       •2             ^
1,5 g/m NTP or 0,0525 per cent by volume
more than the flue gas of the small
boiler with liquid ash discharge.

It is also apparent from Pig. 3 that the
flue gas volume, and hence, the period
of detention do not exercise any distinct
influence within the range of the through-
put rates used. This may be due to the
fact that the effects of several process
parameters partly cancel each other. Por
example, it is not possible to determine
what influence the S02 content of the
raw gas has on the removal efficiency,
as the concentration of SOp in the raw gas
from the 350-MW boiler was constantly
fluctuating within the above-mentioned
range.

Measurements made to check the dust
removal efficiency of the full-scale
plant showed that dust concentration
in the clean gas was 50 mg/nrNTI when
the system throughput was 133 000 m^NTP/h
and the dust load of the raw gas was
15 g/nrNTP.  The removal efficiency,
                791

-------
 therefore, on average was 99,7 per cent.

 In the small-scale unit, dust concentration
 in the clean gas ranged between 10 and
        •Z
 15 mg/m NTP at a mean entry concentration
 of abt. 10 g/m5NTP. The light coloured,
 yelloish grey fly ash of the 350-MW boiler
 appears to be much finer than the much
 darker ash from the small boiler with
 liquid ash discharge. So far, no dust
 analyses have been made.

 The final dust concentration of 50 mg/nrNTP
 obtained v/ith the full-size plant is far
 below the maximum allowable value of
 150 mg/m5NTP.

 With the knowledge now available on S0?
 and dust removal and its dependence on
 volume flow, it would be possible to
 build units for throughput rates of,
       . C -2
 say, 10 m NTP/h and over, based on the
 tested design and layout.

 While the tests so far have proved in
 the main that the results obtained with
 the pilot plant v/ith respect to separa-
 tion efficiency also apply to the full-
 size version, the purpose of further
 tests v/ill be to try out a series of
 designs and process variants some of
 which are aimed at an optimization of
 the dimensions of the primary scrubbing
 stage. Operational reliability will be
 the most important aspect.

 Reproducibility of guaranteed values in
continuous operation at a minimum of
                792

-------
maintenance costs is the standard by
which operational reliability must be
gauged. Therefore, one thing for which
the full-size plant was constantly
being observed and checked was the
formation of deposits and incrustations.

It was particularly in the conical section
and in the upper third of the cylindrical
shell of the scrubber that substantial
incrustations and asymmetrical deposits,
due to the given inflow conditions,
occurred.

After the design was changed to give
better flow conditions and additional
spray nozzles for intensive shell
washing, as shown in Fig. (4), were
installed, a considerable improvement
was obtained.

By varying the rate of wash water flow
at constant gas throughput, it was
proved that the rate of growth of the
deposits on the shell surface was
essentially dependent on the thickness
and velocity of the film running down
the shell surface.

A temporary disturbance in the operation
of the full-size unit occurred as a result
of insufficient clarification of the
return water. In designing the settling
tank, the dimensions had intentionally
been held rather small because plans were
for this part of the system to be extended
only at a later stage. It thus could happen
                793

-------
that the water recycled to the scrubber
was for a while entering the annular gap
washer as a slurry containing 200 g/1 of
solids. While this appeared to have no
influence on SOp and dust removal, and
the nozzles were not clogged, heavy
erosion must be expected to occur in
the long run on such items as pumps,
flow restriotors and nozzles.

After the inflow conditions and the
pump-out conditions at the discharge
cone of the elevated vessel were im-
proved, the solids concentration in
the overflow dropped to between 10
and 20 g/1. It is intended to test
still other settling tanks of modified
design and with larger dimensions. In
this connection, it is noted that the
mean settling velocity at concentrations
up to 50 g/1 was about 4 m/h and that
it slowed to 1 m/h at a solids concen-
tration of 100 g/1.

The slurry forming in the conical thicke-
ning section of the settling tank con-
tains an average concentration of solids
of abt. 300 g/1. Tests carried out by
Westfalia Separator of Oelde, \7est
Germany, have shown that this sludge
water can be largely clarified with
the aid of decanters. Residual water
in the solids discharge is as low as
abt. 30 per cent by weight. In spite
of this low water content, the mixture
still flows and can be pumped, and it
may greatly reduce handling and dumping
costs.
                 794

-------
But since it was not our intention
to content ourselves with dumping the
sludge, we requested three institutions
and specialist firms to cooperate in
analyzing the sludge and studying the
possibilities of processing it for
commercial use.

Initial results of these studies
indicate that it will be possible, for
example, to use the sludge consisting
of fly ash and reaction products as an
additive in the production of lime sand
bricks. Test bodies made by Rheinisch-
Y/estfalische Kalkwerke, Dornap, using
a 20-per cent addition of sludge were
found to have a maximum compressive
                      p
strength of 300 kgf/cm  as compared
               P
with 240 kgf/cm  for the standard lime
sand brick.

In this connection it should be noted
that gypsum in the form of dihydrate
was found at the time in the sludge of
the pilot plant in addition to the
hemihydrate. In contrast to this ,
calcium sulphate in the form of
anhydrite was in some instances found
to prevail in the full-size plant.

One reason for this may be temperature
zone differences between the two systems.

For instance, one point where dihydrate
is  converted into anhydrite is around
40 deg C. Moreover, as is known, the
conversion points may shift owing to
the presence of attendant materials,
                795

-------
such as certain constituents of the
fly ash. Part of the heat generated
during the slaking of CaO will surely
be dissipated more quickly in the small
unit than in the larger plant.

Another use of the sludge seems possible
if the gypsum produced in the process
is again obtained in the form of dihy-
drate. If a unit for dry separation of
fly ash is arranged to precede the
scrubber, then it will be possible to
produce high-quality gypsum board by
the well known Giulini process.

After completion of the first set of
test  runs, the full-size plant will,
therefore, be connected to follow the
electrostatic precipitator so that
measurements can be made for separate
removal of dust and SOp.

The fact should be emphasized here that
it is not possible at the present stage
to make a final statement on the sludge
produced, because only a few samples
were tested so far.

After completion of the tests in which
pulverized white lime is injected into
the raw gas inlet, the plant will be
operated on milk of lime. It can be
expected that the use of this additive
in the form of slaked lime or calcium
hydroxide will result in a further
improvement of S0? removal. This was
seen to be the case with the small
pilot plant in which pulverized calcium
hydroxide and milk of lime were used
                 796

-------
at times. In those tests a removal effi-
ciency of 88,8 per cent was obtained
at a concentration in the raw gas of
only 2,25 g/m NTP or approximately
0,08 per cent by volume.

The use of milk of lime will offer
still another advantage. So far, all
of the additive was introduced, as
already mentioned, upstream of the
primary scrubbing stage. The annular
gap washer, operating in a zone of
lesser concentration, received no
additional lime. A check on the pH
values of the wash, fluid passing
through the annular gap indicated
11,0 for the inflow and 5,5 for the
outflow of that stage.

This means that the annular gap
washer was working partly in the
acid range.

When we change to milk of lime, we
shall feed the additive to both
stages in varying proportions.

The first and most important result
of the tests run so far with the
full-size plant is the fact that the
information obtained with respect to
SOp and dust removal is also applicable
to the design and layout of large
commercial units.

The tests were again prepared and
carried out in close cooperation
with Steinkohlen-Elektrizitatsgesell-
schaft-AG, Essen, which has proved
so successful in the past.       /      „
                 797           ..

-------
Isvirvvoi ef$0i east Cj'i frc.Ti flue 'c^'-.s
ELpeiuiH-nKi! ur,tti tar Hi COia'i/tryii art £8X)i,i',:fn/li
                                             .1
                                                                               I  I    I   j  I    I     1  -11,-" I '"I*
                                                          f
Schematic Diagram  of Pfant for ffts
Removal of Dust and S02 from Flat Qasas
70/IB
                                               .  i



TO
SO
id
30
30-

0
t





si
X *«
1.0




y
./
:Lpfc.
iff with
»r wilfi




i
2L
tfr/ otfi
quid ash

*
m *


'iseharg
titictiarg
Of


•

t


o

—

~ uiL
tor
l,fj
* so.c

-T~'~

	

crtrfcn

-



e< .>;o-«
Cm'HrP/f, 't



-—

Wbctlw
"&
if 2 M^
G£ ; £3 1 SO/- removal from flue go$e$
\LJ? I



—

with


71/11
                                                                   SOi  and dust removal from flue gases
                                                                full-size ttpefimentol plant-raw gas Intel
                                              I   •
                                           .3
                                                   798

-------
AVAILABILITY OF LIMESTONES AND DOLOMITES
            J.J. 0'Donne11

            A.G. Sliger



 Research and Engineering Development

       The M.W. Kellogg Company

        Piscataway, New Jersey
  Presented at the Second International
  Lime/Limestone Wet Scrubbing Symposium,
  New Orleans, Louisiana, November 8-12, 1971
                    799

-------
                     INTRODUCTION
     Power in the United States is produced largely from
plants which burn coal or oil as the primary fuel.  The
use of limestone or dolomite to control sulfur emissions
from these plants is contingent upon several factors.
Among these, proximity of adequate carbonate rock deposits
to fossil fuel-fired power plants, and the relationship of
carbonate rock production to possible demand are important
considerations in establishing the relative merits of any
carbonate rock-based process.  This study, performed under
contract with the Office of Air Programs, had as its objec-
tive the determination of the availability and costs of
limestone and similar materials throughout the contiguous
United States, thus providing a basis for determining the
feasibility and economics of limestone-based SOK removal
processes for any particular power plant site.
                           800

-------
      POTENTIAL LIMESTONE DEMAND BY POWER PLANTS
     Figure 1 shows the location of the major (^200 MW)
power plants in the United States which burn either coal
or oil as the primary fuel.  Included are a few plants
which, although not yet constructed, are being designed
exclusively for either coal or oil and are scheduled to
be on stream by 1975.  Of the 275 plants shown,  84% are
coal-fired.  The oil-fired plants are all located along
the eastern coast, with most of these in the northeast.
About 90% of all power plants shown are located east of
the Mississippi River, with locally high concentrations
in the northeastern quarter of the country, particularly
in many of the major metropolitan areas.

     Based on 1969 fuel consumption statistics,  an estim-
ated 20 million tons of sulfur oxides were emitted by
power plants in the United States.  More than 40 million
tons of limestone would have been required to remove these
oxides from the stack gases.  This potential limestone
demand has been broken down by region, with the regions
                          801

-------
being defined as follows
     Region
States Included
     New England
     Middle Atlantic
     East North Central
     West North Central
     South Atlantic
     East South Central
     West South Central
     Mountain
     Pacific
Connecticut, Maine, Massa-
chusetts, New Hampshire,
Rhode Island, Vermont

New Jersey, New York,
Pennsylvania

Illinois, Indiana, Michigan,
Ohio, Wisconsin

Iowa, Kansas, Minnesota,
Missouri, Nebraska,
North Dakota, South Dakota

Delaware, Florida, Georgia,
Maryland  (incl. Washington,
D.C.)f North Carolina,
South Carolina, Virginia,
West Virginia

Alabama, Kentucky, Mississippi,
Tennessee

Arkansas, Louisiana, Oklahoma,
Texas

Arizona, Colorado, Idaho,
Montana, Nevada, New Mexico,
Utah, Wyoming

California,  Oregon, Washington
     Table 1 itemizes potential demand by region.  The East

North Central region far outranks any other region in poten-

tial limestone requirements with over one-third of the total.

With the exception of New England, all regions in the eastern

half of the country had potentially large demands for lime-

stone.
                           802

-------
                CARBONATE ROCK RESOURCES

     Deposits of carbonate rocks, including limestone,
dolomite, shell, marble, and marl, occur in some form in
every state.  Total reserves have never been estimated,
but are known to be enormous.

     Figure 2 shows the distribution of surface carbonate
rocks in the United States and includes limestone, dolo-
mite, and marble.  The map shows that surface deposits of
carbonate rocks occur throughout the nation, but are par-
ticularly in evidence in the eastern half of the country.
A band of deposits beginning in Vermont extends southward
along the Appalachian Mountains into central Alabama.
Extensive deposits are found in the states surrounding the
Great Lakes, reaching southward into northern Alabama.
Large areas of Minnesota, Iowa, and Missouri are covered
with carbonate rocks and broad outcrops occur in Kansas,
Oklahoma, Arkansas, and Texas.

     Particularly in the central lowlands, carbonate rock
deposits frequently occur as thick, horizontal formations
covering large areas.  In general, the deposits found in
western states are different.  They commonly occur as
steeply dipping or vertical beds of small areal extent.
However, notable exceptions to this are found, particularly
in Colorado, Arizona, and New Mexico where large outcrops
occur.
                         803

-------
     Limestone occurrences, including chalk but excluding
dolomite, are shown in Figure 3.  The map is similar to
Figure 2 and shows that although limestone is found through-
out the country, the more numerous and extensive deposits
occur in the eastern half of the nation.  In the western
states, the deposits tend to be discontinuous and relatively
small in areal extent.

     The formations shown in Figure 3 include limestones of
different degrees of purity.  Estimates have been made that
only about 2% of the known reserves of commercially usable
limestone is chemical grade (>95% carbonate content), and
that the bulk of these reserves will be exhausted in 40-50
years.  Much of this high purity limestone occurs in the
area extending from the Great Lakes southward to Alabama.

     Chalk deposits are shown in several of the central
states and in a curving belt through Alabama and Mississippi.
In general they are not high purity limestones, but locally
they may contain over 95% calcium carbonate.

     Figure 4 shows the location of high grade  (>25% magnes-
ium carbonate) dolomite quarries.  Although originally drawn
in 1941, it is a useful guide to the important occurrences
of dolomite.  As with limestone, the largest deposits are
located in the eastern half of the country.  Two major areas
are noted.  First, a belt of dolomite extends from Vermont
to central Alabama, along the Appalachian Mountains.  This
coincides quite well with a similar band of limestone pre-
viously noted.  Second, large formations of dolomite occur
in the states encircling the Great Lakes.  These deposits
coincide with or adjoin large limestone deposits in the
region.
                           804

-------
     The most significant deposits of marble are found
along virtually the entire length of the Appalachian Moun-
tains in the east, and as scattered occurrences in the Rocky
Mountains in the west.  Although eastern marbles are pre-
dominantly calcitic (high calcium) , dolornitic types also
occur.  Both types are found in the west.

     Shell limestone occurs primarily in Gulf Coastal
waters, but it also is found in bay waters along both the
east and west coasts.  It is usually a very pure type of
calcium carbonate.

     Marl deposits exist in several areas, notably around
the Great Lakes, and along the southeastern coastal plain.
This soft, relatively impure form of calcium carbonate
varies considerably in character, that of the Great Lakes
area being a precipitated calcium carbonate, while that of
the coastal plain is an impure shell deposit.  Limited
occurrences in other regions are generally impure chalks or
soft limestones.
                       PRODUCTION

     Crushed carbonate rock production in the United States
in 1969 was distributed, by type, as indicated in Table 2.
As shown, the total national production amounted to 652
million tons, of which over 85% was limestone.  Dolomite
ranked second with more than 10%.  Marl and marble production
were very low, amounting to less than 5 million tons combined,
or 0.7% of the total.
                          805

-------
     No limestone production was reported in four states,
viz., Louisiana, Delaware, New Hampshire, and North Dakota.
In fact, the latter three states did not produce any type
of carbonate rock.  Dolomite was produced in twenty-four
states, chiefly in the northeastern quarter of the country.

     Production of limestone and dolomite, by region, is
shown in Table 3.  Within most regions, production rates
of individual states varied from near zero to tens of
millions of tons.  The New England and Mountain regions,
however, had fairly uniform (but low) outputs.  The East
North Central region was also exceptional, in that all
states reported large quantities of limestone and dolomite,
ranging from 16 million to 55 million tons.  Nationwide,
Pennsylvania and Illinois were the leading producers of
limestone and dolomite, respectively.

     Shell was dredged from bay waters along all three
coasts.  However, 83% came from Texas and Louisiana, with
the latter being the leading producer.  Small quantities
of marl were produced in Indiana, Michigan, Minnesota,
Mississippi, Nevada, South Carolina, Texas, and Virginia.
Eighteen states, principally in eastern and western moun-
tainous regions quarried marble, with Alabama recording the
highest production at 632,000 tons.

     Over 4,700 quarries, producing  861 million tons of
crushed stone of all types, were in operation in the United
States in 1969.  Since production of crushed carbonate
rocks totalled 652 million tons, it can be assumed that,
                            806

-------
roughly, over 3,500 of these quarries produced limestone,
dolomite, and related stones.  More than one-third of all
quarries had annual production rates of less than 25,000
tons.  The large operations  (over 900,000 tons/year) pro-
duced one-third of the total crushed stone, although they
represented less than 4% of the total number of quarries.
                          END USE

     Carbonate rocks are unique among the different types
used in this country.  Not only do they find use in appli-
cations where their physical properties are important, but
also in markets which utilize them for their chemical prop-
erties and composition.  Almost two-thirds of the total
production of carbonate rocks were used for various con-
struction purposes in 1969.  Most of this stone was used
as an aggregate material in road construction.  The second
largest use was in cement manufacture, which consumed more
than 105 million tons.  About 70 million tons of high grade
limestone and dolomite were used in lime manufacture and
other applications requiring a high purity material.
                       SELLING PRICE

     The average unit value, or net selling price at the
quarry, for all crushed carbonate stones produced in the
United States in 1969 was $1.49/ton and varied by type of
stone as shown in Table 4.  The average values ranged from
a low of $1.0I/ton for marl to a high of $9.69/ton for
                           807

-------
marble.  The high unit value of marble reflects its prim-
ary use as a decorative material.

     Unit value also varies with end use.  Stone used for
construction purposes averaged $1.44/ton while stone used
in applications requiring a high purity material had an
average value of $1.69/ton.  Average prices ranged from
$0.69/ton for fill to $6.00/ton for exposed aggregate
(decorative stone).   Stone for most applications, however,
averaged under $2.00/ton.  The variation in price depends
not only on supply and demand, but also on the chemical
and/or physical properties required for the particular
application.

     Average unit values of limestone and dolomite, by
state, are shown in Table 5.  Within most states, average
prices were $1.00-2.00/ton, although spot prices ranged
from- $0.12-25.00/ton.  New Jersey and several states in
both the New England and Mountain regions reported average
values greater than $2.00/ton.  Rhode Island showed the
highest average value: $7.57/ton for limestone.  Prices
shown for the Pacific region are unusual in that those
for limestone are comparatively low while dolomite prices
are quite high.  In all cases where average unit values
were high, production of the particular stone was limited.

     With the exception of Virginia, which reported $3.92/
ton, average prices of shell were $1.00-2.00/ton.  Unit
values for marl were all below $1.15/ton.  Marble varied
widely in price.
                          808

-------
                     TRANSPORTATION

     Trucks dominated in the transportation of carbonate
rocks from quarry to consumer, accounting for almost three-
fourths of all stone.  Rail and waterway hauls, amounting
to one-fifth of the stone shipments, were about equally
divided.

     Trucks generally are used for shorter hauls of under
50-100 miles, while rail is employed for longer distances.
Where conditions permit, shipment of stone by barge or boat
is preferred, since this is usually the cheapest method of
transportation.

     So many factors influence transportation rates and
costs that it becomes very difficult to establish average
rates, even within a single area.  Most freight rates for
crushed stone in the United States, however, fall within
the ranges shown in Figures 5-7.  The lower curves corres-
pond to rates which could be obtained under favorable con-
ditions; i.e., high volume movement, periodic shipments in
lot-sized quantities, intrastate shipment or commodity
rates, etc.  The upper curves correspond to unfavorable
conditions.
              DELIVERED PRICE OF LIMESTONE
     The delivered price of limestone to a power plant is
the sum of the price of the stone at the quarry plus the
transportation charges.  To see what this actually would
                           809

-------
be, the delivered price of limestone to 37 selected power
plants was determined based on the assumption that a high
calcium limestone would be required.  The results of this
investigation are shown in Table 6.

     Most of the plants are located in the eastern half of
the United States where the major coal- and oil-fired
power capacity is found.  Prices range from $1.95-13.20/ton.
Half of the plants could be supplied at under $4.00/ton,
while all but 3 could obtain limestone at under $6.00/ton.
The latter 3 plants are located in the west in areas where
base prices are higher or limestone deposits are remote.
For several eastern seaboard plants, particularly in New
England, the availability of low-cost high calcium lime-
stone is contingent upon the acceptability of an imported
stone.  Domestic sources are either inadequate or too dis-
tant to provide a low-cost material.
                      PROJECTED COSTS
     Carbonate rocks historically have been stable, low-
priced commodities.  Based on average unit values for the
years 1960-1969, projected average base prices for 1975 are
as follows:

                                  1969         1975
     Limestone and Dolomite    $1.45/ton   $1.67-1.82/ton
     Marl                      $1.01/ton   $1.28-1.48/ton

The average value of shell has dropped considerably since
1960.  It is unlikely that it will continue to decrease
                            810

-------
through 1975.  More probably, it will parallel limestone
and dolomite but not exceed  them in value.  Average unit
values for  crushed marble  are highly variable, reflecting
the sensitivity of price to  market conditions.

     Transportation rates  during the next  few years should
rise about  6%/year, on  the average.  Estimates by type of
transportation are as follows:
               Truck           4-  6%/year
               Rail            6-  8%/year
               Water           5-10%/year
In general, rate  increases  should  follow wage  increases
granted  to  labor.  The  estimates,  offered by stone produ-
cers, are predicated  on the present  state of the economy
and a continuance  of  the present inflation  rate.  Any
changes  will,  of  course, affect these  estimates.
       SUPPLY/DEMAND RELATIONSHIP OF CARBONATE ROCKS
       	FOR POLLUTION CONTROL	

Proximity of Carbonate Rock Deposits to Power Plants

     Comparison of Figures 1-4 indicates that the major
deposits of carbonate rocks largely coincide in location
with the power plants.  This is particularly true in the
East North Central region where huge reserves of stone
occur.  Both high calcium limestone and high grade dolo-
mite abound, and many deposits are found near the major
power generation  centers throughout the region.
                         811

-------
     The New England region is not as fortunate.  While
the power plants are located primarily in coastal areas,
the rock deposits occur in the mountainous western sec-
tions.  Most of the deposits are highly crystalline stone
or marble, and many are dolomitic.  The suitability of
these materials would have to be determined before in-
cluding them as a possible source.

     Availability of stone should not be a problem for
power plants in the Middle Atlantic region.  All types
of stone occur and nearby deposits can be found.  Plants
in western and eastern Pennsylvania, particularly, are
fortunate in that large reserves of high grade stones
are present.

     In the South Atlantic region, most plants are located
near an adequate source of stone, particularly if the cry-
stalline limestones and dolomites of the mountainous areas
   <«
prove suitable.  Several coastal plants could use shell
or coral limestone, or marl.  However, for some inland
power plants in, for example, North Carolina, no nearby
deposits exist.

     Large quantites of high calcium limestone occur
throughout the East South Central region and power plants
should experience no difficulty in obtaining adequate
supplies.  Many plants located in Alabama and Tennessee
also could obtain dolomite quite easily.

      The less abundant and more widely scattered carbon-
ate resources of the western states are of minor impor-
tance, since there are few coal-fired power plants in
                         812

-------
this area.  With the exception of the two power plants
located in North Dakota, which are far removed from any
commercially important carbonate deposits, the few plants
that do exist are fairly near reserves of high calcium
stone.

Potential Demand Relative to Production
     As previously mentioned, the potential demand for
limestone by coal- and oil-fired power plants in 1969 ex-
ceeded 40 million tons.  This represents only 7.3% of the
national limestone production of 559 million tons.  Dolo-
mite production, on the other hand, was roughly 50%
greater than the potential demand.  Shell production was
only one-half of the demand.  The quantities of calcar-
eous marl and crushed marble quarried were relatively in-
significant.  Obviously, limestone is the only stone pro-
duced in sufficient quantities to warrant nationwide con-
sideration as an agent for SOX removal.  Dolomite and
shell, however, are quarried in large enough amounts to
make them important materials in some regions.  Marl and
crushed marble may be useful in certain localities where
the other rocks do not exist, but their limited occurr-
ence and production do not permit wide-scale use of these
materials.

     Table 7 shows the relationship of limestone and dolo-
mite production to potential demand, by region and state.
The third column of the table is the ratio of total lime-
stone and dolomite production to potential demand, for 1969
                         813

-------
     With the exception of New England, the eastern re-
gions of the country all have large relative supplies of
limestone and dolomite.  There are, though, some excep-
tions to this among the individual states.  New Jersey
and Delaware have a potential need which exceeds their
production.  Adjacent states, however, are large producers
of stone and could provide the necessary tonnages.
Mississippi and both of the Carolinas have low relative
supplies of stone.  If they could not otherwise be supplied,
marl deposits, which occur extensively in all three
states, could be used.  Georgia and West Virginia, with
comparatively low relative supplies, could easily obtain
needed stone from surrounding states.  It is interesting
to note that the region with the highest limestone demand,
i.e., the East North Central region, also has the highest
limestone and dolomite production.

     The New England region faces a shortage of limestone
and dolomite, with two states producing less stone than
potentially required.  The marble resources of the region
could improve the situation somewhat, but power plants
would have to rely on shipments of stone from nearby states
such as New York, or, perhaps, on imports.

     Most states in other regions of the country have ample
production.  North Dakota, with no production, is an out-
standing exception to this, however.  A few other states
have low relative supplies of stone, but either the demand
is quite small, or the production could easily be expanded
to meet the need.
                         814

-------
                     CONCLUSIONS

   €> Enormous deposits of carbonate rocks occur in the
United States, and reserves are more than adequate for the
foreseeable future.  Availability of high purity stone may
become a problem several decades hence, but, with the prob-
ability that the required quality will depend on other pro-
cess and economic factors, no shortage of suitable stone
is foreseen.

   o The major deposits of carbonate rocks occur in the
eastern half of the United States, where the vast majority
of fossil fuel-fired power plants are located.  Large re-
serves in these eastern areas provide a nearby source of
stone for most power plants.

   © Relative to the potential demand for carbonate rocks
by power plants, production of these materials is quite
large in most states.  However, current production is in-
adequate to supply the potential needs of power plants in
several Atlantic coastal regions, notably New England.

   • Limestone is the only type of carbonate rock which is
produced in large enough quantities to merit consideration
for widespread application in the removal of SOX from
stack gases.  In many areas, however, ample amounts of
other carbonate rocks are produced, particularly dolomite.

   0 Most of the power plants in the eastern half of the
United States could be supplied with high calcium lime-
stone at less than $6.00/ton.  Many could obtain stone at
                         815

-------
less than $4.00/ton.  Costs for power plants located in
western states generally would be higher, owing to the
lack of suitable, nearby deposits.

   •  Based on projections of material cost and transporta-
tion charges to 1975, the delivered price of limestone to
most power plants should not increase by more than $1.00-
2.00/ton.
                        816

-------
                      TABLE  1
POTENTIAL LIMESTONE DKMAND BY PO'./EK PLANTS IN THK UNITED STATES
Region
New England
Kiddle Atlantic
East North Central
West North Central
South Atlantic
East South Central
rfest South Central
x-
Kountair
Pacific
Total
Limestone (H Tons)
1,369
6,583
14,307
2,738
8,471
5.6A5
5
1,351
	 266
40,735
% of Total
3.4
16.2
35.1
6.7
20.8
13.9
-
3.3
0.6
100.0
                          817

-------
                        TABLE 3
PRODUCTION OF CRUSHED AND BROKEN LIMESTONE AND DOLOMITE
        IN THE UNITED STATES IN 1969, BY REGION
        Region	    Production  (MM Tons)
        New England                   2.4
        Middle Atlantic              90.9
        East North Central          185.6
        West North Central           92.5
        South Atlantic               87.5
        East South Central           81.3
        West South Central           58.3
        Mountain                     10.7
        Pacific                      18.8
           Total                    628.0
                         818

-------
                  TABLE  2
PRODUCTION OF  CKJ-HJi!) AND BROKEN C.-JIBOKATE STONES
      IN THE U::iTKi) 3TATr;M 1H 3V69, BY TYPE
                               Prpductipn (M Tons)
     Liiusstone                   558»793

     Do lorn ic                     63,330

     Shell                        19,731

     Calcareous  1'arl              2,490
                    819
           Total,  All Types      651,665

-------
                     TABLE  4

UNIT VALUE OF CRUSHED AND BROKEN  CARBONATE  STONES
      IN THE UNITED STATES  IN  1969,  BY TYPE
    Ty_£e	      Unit  Value ($/Ton)
    Limestone                      1.45
    Dolomite                       1.55
    Shell                          1.42
    Calcareous Marl                1.01
    Marble                         9.69

    Average, All Types            1.49
                   820

-------
                         TABLE  5

UNIT VALUE OF CRUSHED  AND  BROKEN LIMESTONE AND DOLOMITE
   IN THE UNITED STATES  IN 1969, BY REGION AND STATE
Average Unit Value ($/Ton)
Region and State
New England
Connecticut
Maine
Massachusetts
New Hampshire
Rhode Island
Vermont
Middle Atlantic
New Jersey
New York
Pennsylvania
East North Central
Illinois
Indiana
Michigan
Ohio
Wisconsin
West North Central
Iowa
Kansas
Minnesota
Missouri
Nebraska
North Dakota
South Dakota
South Atlantic
Delaware
Florida
Georgia
Maryland
North Carolina
South Carolina
Virginia
West Virginia
Limestone
NR(I)
1.32
4.14
7.57
1.46
2.49
1.56
1.46
1.44
1.32
1.02
1.54
1.17
1.49
1.40
1.32
1.39
1.87
-
1.22

1.31
1.50
1.57
1.62
1.51
1.52
1.62
Dolomite
4.20
- (2)
5.24
"
1.53
1.97
1.73
1.45
1.28
1.45
1.48
1.21
1.72
-
1.38
1.13
-
-
-

-
-
-
1.37
1.62
  (1)"NR" indicates that value  was  not reported
  (2)  "-"  indicates no production
                          821

-------
                       TABLE__5  (Cont'd)

UNIT VALUE OF CRUSHED AND  BROKEN LIMESTONE AND DOLOMITE
   IN THE UNITED STATES  IN  1969,  BY REGION AND STATE
  Region and Stat_e_

    Easj: Sgu,th Contra. I
      Alabama
      Kentucky
      Mississippi
      Tennessee

    West
             ___
      Arkansas
      Louisiana
      Oklahoma
      Texas
Average Unit Value  ($/Toii)
 Limestone      Dolomite
   1.17
     46
     00
     33
   1.36

   1.30
   1.35
1.60
 NR
1.17
      Arizona
      Colorado
      Idaho
      Montana
      Nevada
      New Mexico
      Utah
      Wyoming

    Pacifi c_
      California
      Oregon
      Washington
     64
     04
     10
     24
     64
   1.51
   2.25
   2.11
   1.07
   1.00
   1.28
2.83
1.77
 NR
2.66
2.91

6.00
  Total Un.i ted  States
   1.45
1.55
   (1) "NR"  indicahes  that value v/as not reported
   (2) "-•"   indicates  no production
                        822

-------
                          TABLE 6
   DELIVKKED PRICE OF HIGH-CALCIUM LTMESTC"i; TO SKLKCT-MJ PO'-.TiR PLaiTS
Power Plant

Benning
Gorgas
Cherokee
J. McDonough
Devon
Fisk
R.S. Wallace
V.'ill County
D.H. Mitchell
Wabash
Fiivcrside
Lav/rence
Cane Run
Elmer Smith
Riverside
Edgar
L Street
Delray
High Bridge
Hawthorn
Sioux v
Essex
Port Jefferson
Waterside
Allen
Lcland Olds
Tidcl
Mi and. Fort
Acme
Horseshoe Lake
Elrania
Schuylkill
Wateree
Bull Rur.
Cabin Creek
Karr^ner
LakeHide
Location
Delive re d P ri ce
Washington, D.C.
Gorgas, Alabama
Denver, Colorado
Cobb County,' Georgia
Killford, Connecticut
Chicago, Illinois
East Peorio, Illinois
Lockport, Illinois
Gary, Indiana
Terre Haute, Indiana
Iov;ana, Iowa
Lav/rence, Kansas
Louisville, Kentucky
Owensboro> Kentucky
Baltimore, Maryland
N. Weyr.outh, Massachusetts
Boston, Massachusetts
Detroit, Michigan
St. Paul, Minnesota
Kansas City, Missouri
'vest Alton, Missouri
Newark, New Jersey
Port Jefferson, New York
New York, K^w York
Belmont, North Carolina
Stanton, Korth Dakota
Brilliant, Ohio
North Bend, Ohio
Toledo, Ohio
Horseshoe lake, Oklahoma
hlrama, Pennsylvania
Philadelphia> Pennsylvania
Rockland City, South Carolina
Oak Ridge, Tennessee
Cabin Creek, t.'est Virginia
Capti.na, ,.'e.-,t Virrinia
St. Francis, Uiscoiicin
      3.23
      6.36
      4.50
      A.50*
      2.AO
      3.30
      3.30
      2.65
      2.25 (75-9A2
      1.95
      3.66
      3.00
      3.72
      3.B5
      A. 50*
      A. 50*
      2.AO
      3. CO
      4.60
      3.10
      4.50*
      4.50*
      A. 50*
      5.39
     13.20
      3.80
      2.45
      2.45
      8.00
      5.55 (92; CaC03)
      4.50*
      3.90 (se,; cacoo)
      4.24
      6.00
      4. CO (H0;; CaCOv)
      2,60
* -- Source of stone  is  outside of U.S.  (i3ahar:~jr,)
                             823

-------
                          TABLE 7
AVAILABILITY OF LIMESTONE AND DOLOMITE FOR POLLUTION CONTROL (1969)
Region and State

  New England
    Connecticut
    Maine
    Massachusetts
    New Hampshire
    Rhode Island
    Vermont
      Totals

  Middle Atlantic
    New Jersey
    New York
    Pennsylvania
      Totals

  South Atlantic
    Delaware
    Florida
    Georgia
    Maryland
    North Carolina
    South Carolina
    Virginia
    West Virginia
      Totals

  East South Central
    Alabama
    Kentucky
    Mississippi
    Tennessee
      Totals

  East North Central
    Illinois
    Indiana
    Michigan
    Ohio
    Wisconsin
      Totals

  West North Central
    Iowa
    Kansas
    Minnesota
    Missouri
    Nebraska
    North Dakota
    South Dakota
      Totals
Production
 (M Tons)
     275
     800
     750
     800
  33,457
  56,667
  90,900
  40,729
   4,334
   9,804
   4,500
   1,900
  17,829
   8,405
  87,500
  17,752
  30,158
     300
  33,109
  81,300
  54,844
  25,157
  39,066
  50,595
  15,937
 185,600
  26,200
  15,334
   4,127
  41,200
   4,663

     989
  92,500
Potential Demand
    (M Tons)
Ratio
                    9.19
                    16.9
                    4.23
                    17.8
                    14.4
                    57.8
                     365
                    6.37
                    37.7
                    42.4
                       0
                    28.3
                    33.8
                           824

-------
          V 3JSAT
(6961) J05ITMOD
HP'S  3TIMOJOQ QUA.  3MOT83MIJ 10 YTIJIHAJIAVA
oi^BH (anoT M)
61M fr
0 I>
SpttfiJ I >
ep^BJ I >
SplBJ V >
8.3* 02
I3.fr 85£
00
e.ei frv
8. SI 8V
VV.S 5fr£
8.5£ V3
IV. I eV£
Se.V I5£,I
ir • C O O O £
OV< 83£>
(anoT M)
3V3,5
00£,85
053a
05S
000,1
see
OOV,OI
OOfr,VI
02£
020,1
008,81

iBiztnsD rl^uo8 ^asW
BnBiaiuoJ
3BX9T
obBioloO
OflBbI
£ nfi -it noM
SDfiVSW
ooixsM W9M
oiiioB'I
Bin'xolilBD
nop9ttO
noipnxriaBW
            \  5S8

-------
     UJ
     Ql

     UJ
     LUl

     I
UJ
     It
     UJ

     I
     Q.
     tr.
     UJ

     X
     o

     <
cc
o
          §
          or

          m

-------

-------
                                CO
                                o
                                CL
                                UJ
                                Q

                                UJ
                                CO
                                UJ
                                      o:
                                           CO
                                           UJ

                                           fe

                                           CO

                                           O
                                           UJ
                                           H-
                                           UJ
                                           X
                                           co
                                           H
                                           (O
                                           o
                                           a.
                                           UJ
                                           o
                                          _i
                                          a
                                          X
                                          O
                                          o
                                          h-

                                          oa


                                          CO
                                          o
 to
 ro

 O
                                                 UJ
ID
m

CO
UJ
UJ
££

m

CO
o
cc
u_
Q
UJ
                                                 CC.
                                                 Q.
                                                 UJ
                                                 or
828

-------
















+
UI
oc.
CD
U_




















CO
UJ
1-
2
to
o
UJ
h;
Z
13
UI
X
h-
z

to
UI
or
or
o
UI

o
0
o

UJ
o
or
CD
i
X
CD

X
U.
O
Z
0
I
^
o
o


^^
CVJ
O)
r~-
d
•z.
cc


IS
o
cc
Lu
Q
UJ
l-

E
Q_
UJ
cc


829

-------
    CO
    h-
    2
    UJ
    X
    CO

    cc
    o
    LL_

    CO
    UJ
l/>  \—

LJ  cc

§  *
o  o


c  ?
    cc
    o
    0.
    CO
    z
    <
    (T
    O

    o:
CO
UJ

h-

h-
co

Q
UJ
H
UJ
UJ
z
o

CO

o
UJ
X
CO


01
o
                                                               o
                                                               o
                                                               o
                                                               o
                                                               00
                                                                  tn
                                                                  U
                                                               O
                                                               in
                                                                  (O

                                                                  o
                                                               o
                                                               ro
                                                               O
                                                               00
                                                       -    o
                         31IW-N01/SJ.N33 '31VH
                       830

-------
LJ
X
en
Co  w
    h-
LU  <
cc  K

1  §
    Q:
    o
    Q.
    CO
a:
l-
cc
        UJ
        h-

        h-
            Q
            t;J
    UJ
    X
U-j h-
o!
    LU

    b

    en

    Q
    UJ
    x
    en

    Q:
    o
                              m
                                                             CM
                                      31HM-NOJL/S1N3D  '31VM

                                        831
                                                                                    o
                                                                                    m
                                                                                    CM
                                                                                    in
                                                                                    CM
                                                                                    CM
                                                                                    O
                                                                                    O
                                                                                    CvJ
                                                                                    O
                                                                                    m
                                                                                    CM
                                                                                       CO
                                                                                       LU
                                                                                       LL)
                                                                                       CO

                                                                                       Q
                                                                                    O

                                                                                    O
                                                                                    in
                                                                                    O
                                                                                    in
                                                                                    m
                                                                                   - O
                                                                                  O

-------










N
UJ
0









H-
LJ
CL
1C
CO
tlL
O
u.

CO
LU
or
2:
o
<
1-
c:
o
CL.
CO
z
o:
i
i—
>_
i
ce
LJ
l-


co
LJ
1-
H
CO
Q
LJ
H-
•z.
LJ
!_,
^
LJ
O
1-
co
Q
LJ
X
CO
CC
0



                UJ
                in

                6
                           o
                           m
                           o
                           in
                           O

                           O
                           o
                           m
                           ro
                           O
                           o
                           ro
o

CJ
                              CO
                              UJ
                              _1

                              5


                              Uj"
                              h-
                              co

                              Q
                           o
                           o
                           CM
                           O
                           m
                           O
                           o
                           o
                          ' m
3HIW-N01/S1N30 '31VH

832

-------
         THE PILOT SCALE R&D AND PROTOTYPE PLANT




               OF MHI LIME-GYPSUM PROCESS
       Tsukumo Uno, Masurai Atsukawa, Kenzo Muramatsu




             Mitsubishi Heavy Industries, Ltd.




                       Tokyo,  Japan
                       Presented at




Second International Lime/Limestone Wet Scrubbing Symposium




                       Sponsored by




          Environmental Protection Agency,  U.S.A.




     Office of Air Program,  Division of Control System




      Sheraton-Charles Hotel,  New Orleans,  Louisiana







                  8-12 November, 1971
                           833

-------
                 THE PILOT SCALE R&D AND PROTOTYPE PLANT

                       OF MHI LIME-GYPSUM PROCESS




                                ABSTRACT
Tsukumo Uno, Dr. *
Masumi Atsukawa  **
Kenzo Muramatsu  **
     The lime-gypsum process is a method in which flue gas is scrubbed with

lime slurry to remove SO2 and gypsum is recovered as byproduct through air

oxidation of calcium sulfite formed in the absorber.  This process has been

attracting particular attention as an advantageous desulfurization process in

Japan because of its remarkable features - abundant and cheap absorbent,

large demand for by-product gypsum, no waste to  be  thrown away, simple de-

sulfurization system, etc.



     The technical feasibility  is  already proved by the six years operation

record of 62,500 Nm /hr plant which was constructed for a sulfuric acid plant

by our company under the license of Japan Engineering and Consulting Co.

We have conducted recently a pilot scale research and development on this

process to make the process fit to the treatment of flue gas from an oil fir-

ing boiler.  On the basis of the test results, a prototype plant, treating

100,000 Nnr/hr flue gas is under construction at a power station.  Here de-

scription is made shortly on the results of the pilot test and outline of the

plant.
  *   Manager, Hiroshima Technical Institute
  **  Manager, Environment Equipment Research Laboratory,
               Hiroshima Technical Institute
  **  Process Engineer, No.2 Project Engineering Department,
               Chemical Engineering Center

                                   834

-------
                               INTRODUCTION









     ¥e have been conducting research and development of a lime/limestone




scrubbing process since about ten years ago.  In 1962 - 1963, we carried




out a pilot test at Amagasaki No.3 P/S, Kansai Electric Power Co. which treats




5,400 Nm3/hr flue gas from an oil firing boiler.  Through this pilot test, we




have obtained technical data on gas cooling and carbon dust removal by means




of a spray  tower and on the SO2 removal by means of various types of scrubbers.




In 1964 was constructed the first commercial lime-gypsum plant treating waste




gas from sulfuric acid plant at Koyasu works, Nippon Kokan Co.  This plant has




been working smoothly for about six years producing high quality byproduct




gypsum for  sales on the market.









     Though the technical feasibility of the lime-gypsum process has proved that




it  can  treat waste gas from sulfuric acid plant as aforementioned, applica-




tion of the process to the power plant includes some problems to be solved.




That is,  we have to overcome difficulties due to the differences of the




characteristics of these plants as shown in Table 1.  This paper describes




shortly the results of pilot tests which have been carried out to solve these




problems as mentioned hereunder and an outline of the prototype plant being




under construction at Amahigashi P/S,  Kansai Electric Co.




(l)  How to design an economical and efficient scrubber suitable for treat-




     ing a large volume of flue gas with lime slurry.




(2)  How to prevent the plugging due to deposition of sludge and occurrence




     of hard gypsum scale.




(3)  How to obtain high quality gypsum without the influence of impurities




     from flue gas and in raw materials.




(4)  How to satisfy both high S02 removal and high Ca reactivity in case thin




     S02 is to be treated.





                                    835

-------
                  OUTLINE OF MHI LIME-GYPSUM PROCESS









     Though the process differs somewhat depending on conditions of waste




gases to be treated hereby,  it consists of the following sections basically




as shown in Pig.  1.




(l)  Gas cooling and dust removal section




     This is a section to clean, moistened and cool flue gas with water be-




     fore the gas is sent to absorbers.




(2)  Absorption section




     This is a section in which oxides of sulfur are absorbed by lime slurry.




        Ca(OH)2 + S02 + Aq = CaS03'|H20 + Aq                        (l)




        Ca(OH)2 + S03 + Aq = GaS04-2H20 + Aq                        (2)




        CaC02   + S02 + Aq = CaS03-|H20 + C02 + Aq                  (?)




        CaC03   + SOj + Aq = CaS04-2H20 + C02 + Aq                  (4)




        Ca(OH)2 + C02 + Aq = CaC03 + Aq                             (5)




        CaS03-|H20 + y02 + Aq = CaS04'2H20 + Aq                     (6)




     The main reaction of S02 absorption produces calcium sulfite as see;n




     in equations (l) and (3) and some amount of calcium sulfate is also pro-




     duced depending upon 803 and 02 contents of the flue gas and operating




     conditions of absorber as seen in equations (2),  (4) and (6).




(3)  pH value adjusting section




     The pH value of the spent liquor from the absorption section is adjusted




     here to 4 - 4-5 by adding a small amount of sulfuric acid.




(4)  Oxidating section




     This is the section of converting calcium sulfite to gypsum by air oxida-




     tion at 4 - 5 kg/cm^G.




(5)  Gypsum filtering section




     Gypsum is filtered out  of the slurry delivered from the oxidation tower.




     The filtrate is fed back to the cooling tower.
                                  836

-------
(6)  Impurities separating section


     The impurities brought with flue gas or raw materials (lime) are sepa-


     rated out of the system by neutralizing the waste water of the cooling


     tower with slaked lime and filtering this liquor.


     The filtrate is fed back to the process.
                  PILOT SCALE RESEARCH AND DEVELOPMENT




     The photographs of the pilot plant at which the tests on the MHI lime-


gypsum process have been conducted are shown in Fig. 2 and 3.  This plant was

designed to be  able to produce 0.7 T/D of gypsum by treating 2,000 - 3,000

Nm3/hrof flue gas and installed at Hiroshima Technical Institute in October

1969.  The test results are as follows.




1.  Selection of the scrubber type - economical,  efficient and free from

    scaling




         As they well know, the prevention of scaling trouble is the most

    important problem at the designing of lime/limestone scrubbers.

    From this standpoint,  we adopted a spray tower  as a scrubber of simple

    structure for the treatment of waste gas from sulfuric acid.  In this

    case,  the application of this process is rather easy because the amount

    of gas is rather small and also it has higher S02 content,  negligible


    amount of C02 and no particulates.




         To apply this process to  other sorts of plant as shown in Table 1,


    we have carried out a series of pilot test and found that the plastic

    grid packed tower is suitable to the process  «t± it can be operated without
                                                 *=>


                                         837

-------
    scaling trouble at a higher superficial gas velocity and lower draft




    pressure loss.   For the flue gas containing 0.1 vol % of SC>2,  90 to 98 %




    of SO 2 removal  is attainable at a superficial gas velocity of about 1.2




    x 104 mVm2-hr  and a total draft pressure loss about 90 mmH20.   The flow




    rate of scrubbing liquor is 70 m^/m^ hr,  in this case.









2.  Reaction rate in the scrubber









         The absorption reaction occurred in the scrubber is mainly control-




    led by the dissolving rate of lime as shown in Fig.  5 and the dimension




    of the scrubber can be determined by the equation (l) and (2).
                                                                    (i)
                      a
            Kd  = A-C0a-(L/G)                                        (2)





         From these equations  and Fig.  4 and 6,  it  is clear that a higher pH




    value is requested together with a  higher liquid-gas ratio  for attaining




    a higher absorption efficiency.   However,  a  lower pH value  is necessary




    for attaining a high conversion  ratio of lime to  calcium sulfite or  sul-




    fate (this is designated here as Ca reactivity).   To accommodate these




    conflicting factors, we have adopted two scrubbers combined system in




    which one is operated at a lower pH and the  other at a high pH.




    The two scrubbers are installed  in  series in relation to gas stream  - the




    former is countercurrent and the latter is cocurrent.   The  scrubbing




    liquor is at first fed to  the latter and the effluent is then sent to the




    former.  The volume of each scrubber necessary for 90 % 302 removal" and




    98 % Ca reactivity is related with  pH value  of  the latter scrubber,  as




    shown in Fig. 7.  This figure shows that the most economical point exists




    at the pH value of 9, but  practically we have to  use a lower pH at which




    the formation of calcium carbonate  is not predominant in the scrubber





                                  838

-------
    when carbon dioxide content is high in the gas to be treated.









3.  Prevention of plugging in the scrubber









         Troubles often encountered in the lime scrubbing process  are the ac-




    cumulation of soft deposit and the formation of hard scale in  the scrubber.




    The former is usually caused by the deposition of calcium carbonate or




    sulfite and it is preventable by operating the scrubber at a suitable pH




    and avoiding the stagnant portion in the scrubber.   The latter is caused




    by the growth of gypsum crystals and then the desupersaturation of scrub-




    bing liquor is the most effective countermeasure for the scaling.  For




    this purpose, addition of the seed crystals and installation of the delay




    pipe are well known.   And also to avoid stagnant flow or portions dried inter-




    mittently in the scrubber is very important because  the crystalline scale




    of gypsum often begins to grow by concentration of  the liquid  at these




    portions.  We have realized all of these countermeasures at the pilot




    test and found no sign of scaling after a 500 hours continuous operation.









4.  Oxidation of calcium  sulfite








         The oxidation of calcium sulfite is conducted  in a tower  equipped




    with the rotating cylinder type of air atomizer at  4 - 5 kg/cm2 and 50 -




    80 C.   The principle  and features of this specially devised atomizer were




    explained at the first symposium.  However,  we like to again emphasize




    that this may be the  most suitable one for the lime-gypsum process  as it is




    operated at a very high oxidation efficiency without fear of plugging.









         The rate of oxidation depends on pH value of the slurry as shown in




    Fig. g and we adopt pH 4 - 4.5 at practical  operation.    This  figure






                                         839

-------
    suggests us that HS03 ion in the slurry may take an important role in




    the oxidation reaction.  The seed crystal added in the scrubbing liquor




    serves again for the growth of gypsum crystals.
                     OUTLINE OP A PROTOTYPE PLANT









     A prototype plant of MHI lime-gypsum process treating 100,000 Nm3/hrof




flue gas from an oil firing boiler is under construction at Amahigashi P'/S,




Kansai Electric Power Company.  This plant is designed to remove 90 % of S02




from flue gas containing O.llvol % S02 and produce daily 19 tons of gypsum




for sales in which quick lime is used as absorbent.  It is scheduled to be




completed in March 1972 and operated for about one year after the guarantee




run to confirm such problems as concerned with materials of construction,




quality of byproduct, performance characteristics and economics through the




long run.  The production/demand of gypsum and distribution of limestone in




Japan are shown in Pig. 9 and 10 respectively.
                              SUMMARY








     The lime-gypsum has been attracting particular attention in Japan as an




advantageous desulfurization process because it produces gypsum which has a




great possible share in the market by the simple process using lime as ab-




sorbent which is cheap and abundant in Japan.  Here described are the test




results on the process at the pilot plant treating 2,000 NmVhr of flue gas




from an oil firing boiler, especially on the rate of absorption reaction and




the techniques for prevention of scaling.  The technical feasibility and eco-




nomics of this process will be made clear through the operation of the proto-







                                  840

-------
type plant which will be started to operate in March 1972.
                              NOTATION



Yj_n, Iout     S02 concentration at the inlet and outlet        ppm
              of the scrubber, respectively

PO            Pressure of gas in the scrubber                  atm

C0            Effective concentration of Ca in slurry          kg-mol/m3

#            Ca reactivity

Z             Effective height of scrubber                     m

G"            Molar superficial gas velocity in the scrubber   kg-mol/m2.hr

Kd.            Mass transfer coefficient based on dissolving    1/hr.atm
              of Ca
                                      841

-------
h-
U.
O
O
CO
U
s
o  ^

^5
CO  f-
ll-
u.
o
CO

o
8
1
CE D
o°.§
cE z ^
UJ ^ -J
CO 2 {/)
i
e?
z
tr .
Cfe
O Q-
PETROLEUM
REFINERY PLANT
FURNACE
o
^
o
a:
2Z
^<
W Q.



O

^
8
in

§
n
L
*•
8
in

o
o
(VJ
^
8

o
o
'

JANTITY
l03xNrr^H)
"""^ »
Q5 v"1^'^
s


8
n
^
o
r~

in
^*

i
o
n

8
n
*


%


— ^'"""*
is
(T
TEMPE


8
^
in
CJ


o

<
3

3



o

11
i
Q


N
c> ^ !2 —
^ ^ ) j
(VI /-N _«.
^5 ^^ ^^ O^
o ~

n
CJ «-' ^f ~~
. "* ™~
 l * *
o n ~ o
0 ~

£? ^ oo n
• • • ^ •
O °O 
-------
&

oo
oo
1 1 1
U
0
C£
Q_
-
(D
\
LU
•<-
<
	 1

n:
, 	 ,
I
en
U_
00








o
1—
Q_
CXL
0
/ s\
00 
^^
o
X
O

t c
tx. [
 O
< U
(D
t
00
<
ID
x-X
LU
Z)
	 1
u_
\^
U
1
^
^
<
— p>c
— ^(.
»
(


^— -
>^
3
/)
i_
V 	
>—
f-\
J)


o
JN

C
(\l
i
k
(3
00
oj
U
/
\
00

-------
                                 Fig. 2  (left)

                                 Photograph of  the  scrubbers
                                 at which pilot  tests have
                                 conducted on the lime-
                                 gypsum  process.
                                  Fig.  3  (under)

                                  Photograph  of  the  pilot
                                  plant besides  the  scrubbers
                                  at  which the tests have
                                  conducted on the  lime-
                                  gypsum  process.

                                  This  picture shows inside
                                  of  the  house which is
                                  seen  in the Fig.  2.
 -
i§
                    844

-------
SP
ON
   100
    so
5   60
o
51
 ol
O
CO
    20
   -j
   ^/
   TS
LOW •*•
                          PH
HJG-H
               REMOVAL/OPERATING COMDITION
                                     500-WOO
                                       2-7
Remarks

O
•
Grrld pitch.
2-5 Turn.
3 0 mm
4-0 w*
      0-1     0.2    04  0-£ 0.8 1.0    Z.

          EFECTIV/E Ca COWC,-Co (

      F..F- RATE OF ABSORPTION  REACTION
                         845

-------
X




X
o
-J
                            0  o
  0     20     4-0     60    80     100


                REACTIVITY  (%)
 Fig. 6  EFFECT OF  PH ON Cai REACTIVITY
               846

-------
UJ
QO
ca
a
a:
o
to
u.
o

LU
r
                        8.Qa
                        Ca Reactivity
TOTAL
     No-2
     SCRUBBER
        4    5    6     7    8     9     10


            PH  IN Nlo-2  SCRUBBER


  Fig. 7   RELATION- SCRUBBER yOLUME/pH IN No.2
                           847

-------
   NO
 v 0.8
**:
   0.6
   04
   0-2
                                          0.10
                                              s
                                              o
                                              oc.
                                              Of.
                                         0.06
                                         0.02
                                              CO
                                              I
            456
           pH IN OXIPATION TOWER

           EFFECT op PH ON OXIDATION  R/ATE
o  3
         —  Demand
         ||  Production
                                         for
                                         OTHERS
                                         for
                                         CEMENT
                                         for
                                         GrYPSUM
           8   '69   '70   '71   '72   '73  F. YEAR

           PRODUCTION/DEMAND OF
           IN  T/\PAN
                       848

-------
UME STONE, PRODUCTION CHART(1959)
   CF.9.IO)
   TOTAL  91.3QaOOQTon
                                WIT 1QQQTon
                    849

-------

-------
  WET  SCRUBBER INSTALLATIONS
                    AT
ARIZONA   PUBLIC  SERVICE  COMPANY
            POWER  PLANTS
               PRESENTED  BY:

             LYMAN  K.  MUNDTH

       VICE  PRESIDENT,  POWER PRODUCTION
       ARIZONA  PUBLIC SERVICE  COMPANY
               PHOENIX, ARIZONA

               PRESENTED  AT:

                INTERNATIONAL
    LIME/LIMESTONE WET  SCRUBBING SYMPOSIUM
   SPONSORED  BY DIVISION  OF  CONTROL SYSTEMS
     U.S.  ENVIRONMENTAL PROTECTION AGENCY
           NEW  ORLEANS, LOUISIANA
             NOVEMBER 8  - 12, 1971
                    851

-------
                   WET SCRUBBER INSTALLATIONS
               AT ARIZONA PUBLIC SERVICE COMPANY
                          POWER PLANTS
INTRODUCTION

Arizona Public Service Company is pioneering the installation of wet

flue gas scrubbing equipment on coal fired power plants with work

currently progressing at two of its operating stations, one in New

Mexico and the other in Arizona.


At the Four Corners Plant, near Farmington, New Mexico,  scrubbers

are being installed on three units with a total net generating capacity

of 575  MW (nominal) .


The other installation is at the Cholla Power Plant at Joseph City,

Arizona, a  single unit, with a rating of 115 MW  (nominal).


The Four Corners Plant is fueled with sub-bituminus coal from the

Navajo Mine which is adjacent to the plant.  This coal has a high

ash content; ranging from 6 to 25%, and averages about 22%.  The

coal has a  low  sulfur content; with a range  of 0.4 to 1.9% and aver-

ages about 0.7%.


The Cholla Plant coal is from the McKinley  Mine at  Gallup, New

Mexico. The ash content varies from 5  to 15%,  with an average of

about 8%.  The sulfur content is from 0.4 to 1.0% with an average

of about 0.5%.
                                 852

-------
I.  EQUIPMENT CONCEPT





At the time this report was prepared both the  state of New Mexico and



Arizona had particulate emission regulations, but neither state had



adopted SO   emission limits for power plants. However, both states



are currently in the process of adopting SOo emission regulations.





The wet scrubber concept of cleaning flue gases was selected because



of concern for the effectiveness of electrostatic precipitators with low



sulfur coal and for the SOo removal capabilities of the scrubbers.





       A.   Four  Corners Plant    The scrubber installation at the Four



       Corners Plant is being performed by the Chemical Construction



       Company  (CHEMICO) and will use Chemico's variable  throat,



       high energy, wet approach venturi for particulate removal. A



       pilot plant test has been conducted to obtain design data and



       operating parameters for the scrubber system.





       B.   Cholla Plant   The  Cholla installation has been awarded to



       Research-Cottrell,  Inc. for design and construction and will  use



       a high energy, flooded-disc venturi for particulate removal and



       an absorber tower following the venturi for SO  removal.  A
                                                  h


       5000 CFM pilot plant is presently being installed to optimize recycle



       rates,  verify materials of construction, and to conduct various



       chemical control tests.
                                  853

-------
II.     ADDITIVE ABSORBER







Neither plant has a completely engineered SG>2 removal system at the




present time, although the basic concepts of SC>2 removal have been




studied and provisions are being incorporated.







       A.   Four Corners  Pilot plant tests conducted at the Four




        Corners  Plant by Chemico demonstrated that approximately




        10 to 20% SO7 removal occurs simultaneously with particulate
                    £1



        removal  in the venturi. The capability exists for further SC>2




        removal  by injecting supplementary sprays, after the venturi,




        in the lower  section of the scrubber vessel and appropriate




        connections  are being provided at this time for these sprays.







        B.   Cholla Plant   In the Cholla scrubber, SO2 removal will




        take place in a second stage, utilizing a wetted film,  fixed




        packing  absorber tower with an alkaline  solution as  the absorp-




        tion medium.







III.     PROCESS CONFIGURATION







Both the Four Corners and the Cholla  scrubber systems winl  employ a




continuous loop scrubbing cycle with part of the cycle liquor bled-off




to continuously  agitated thickeners.  The clarified water will  be returned




to the scrubbing cycle for make-up.
                                   854

-------
Both installations provide for reheat of the flue gas before entering




the stack to minimize negative bouyancy effects  of the plume and to




prevent condensation of the flue gas within the stack.







Both scrubber installations are retrofits to existing plants with the




major part of the construction taking place with the units in service.







       A.  Four Corners Plant  At Four Corners, space is limited.




       The venturi  scrubber vessels (two venturi scrubber vessels




       on each unit) have been mounted above the existing mechan-




       ical dust collectors and induced draft fans.  The particular




       arrangement was dictated because of limited construction




       space and priority of  continuous plant operation during the




       scrubber installation. With two scrubbers on each unit, and




       appropriate ducting and baffles,  the unit  will be able to




       operate at 50% load if one  of the scrubbers is out of  ser-




       vice for repairs  or maintenance .






       New induced draft fans will be placed downstream from the




       scrubbers, and designed to handle the wet gas.  The fans will




       have stainless steel rotors and rubber lined casings.  The




       fans will discharge into mist eliminators  which  will help




       reduce the flue gas reheat  requirements by drying the flue




       gas before it enters the reheater.
                                  855

-------
        B.   Cholla Plant   At the Cholla Plant, the existing mechanical




        dust collectors and induced draft fans will be retained.  The




        new scrubber system will be installed in series with , and




        downstream from the  existing mechanical collectors and I. D.




       fans.   New booster fans are required to operate  in series




        with each of the existing  induced draft fans to satisfy the addi-




        tional  pressure drop requirements of the  scrubber system.  More




        construction space is available at this site, and the scrubbers




        will be installed at ground level.  Again, two scrubbers will




        be installed  to provide the capability of half load operation




        when one scrubber is out  of service.




IV. WASTE DISPOSAL







At both plants it is planned that the  concentrated solids (fly ash and




sulfates and sulfites of calcium)  in the thickener under-flow will be




transported to the  existing plant ash disposal area  in a separate trans-




port system.   Presently, wet sluice  systems  transport the fly ash to




retention ponds.







V.  PRESENT STATUS OF CONSTRUCTION







Construction and installation of the Four Corners scrubbers is nearly




complete.  The scrubbers are scheduled to be in operation by December,




31 of this year  (1971).  These scrubbers will be "tied-in" during
                                  856

-------
scheduled  maintenance outages continuing through the latter part of




this month (November) and into December.







The Cholla scrubber is scheduled to be in operation by December 31,




1972.  The design work and materials procurement is currently in




progress.  The pilot plant is being  installed at the plant at the present




time,  and scheduled to be operating after the first of the year.







VI. DESIGN CONSIDERATIONS
The nature of an electric utility requires a very high degree of reliability




in its generating facilities.  The  installation of largely untried and com-




mercially unproven pollution control equipment will most certainly de-




grade the reliability  of the generating units  on which it is  installed.




Consequently, an  extreme amount of care must be taken in the design




and selection of wet scrubbers, or any other SO2 removal equipment




to maintain this high degree of reliability.   Specific items to be con-




sidered are:







        1.  Redundancy of Critical Equipment  Because emission regu-




        lations will,  in most cases, prohibit the operation of generating




        units without the pollution control equipment in service, the same




        redundancy is required  in  the design of this equipment as is used
                                  857

-------
throughout the rest of the plant.  This means that for any




critical piece of equipment, a spare should be installed




with provisions for on-line isolation and repair.







2.  Controls   The Controls, and the control system philos-




ophy of the removal system must be compatible with, and




integrated  into,  the control system for the rest of the plant.







3.  Plant Personnel Plant personnel must be trained in  the




operation and maintenance of the  removal equipment.







4.  Materials  Careful specification of  materials of con-




struction is required throughout and each component of the




removal system must be individually studied to determine




the severity of duty to which it will be subjected.







5.  Equipment Layout Thus  far, the experience of industry with




wet type removal systems has demonstrated many problems and




poor reliability.  Consequently, careful attention must be




given to the layout and arrangement of the system.  Ready




access  must  be provided for maintenance, repair, and re-




placement  of equipment.
                           858

-------
6.  Guarantees  On the subject of guarantees, a manufacturer




may be able to design equipment  to meet given proformance




specifications and meet a certain particular guarantee.  However,




I do not believe  you will find a manufacturer who will  guarantee




the performance  of his equipment over the period of its useful




life.  This is, in effect, what the regulatory agencies require




of the utility operator.
                             859

-------
                                         ^N

                                         S*
                                         •^ c\
                                  O
                                  t—t
                                  H
                                O p re
                                O re O
                                H W £5
                                O § Q
                                1—I f-^.
                                > re ^
                                w O O
                                CO CO i-J
                                ,.  I PH
                                U H ^
                                O ^

                                Is
                                  re
                                  O
860

-------
                                                              *
                                                o
                                                I—I
                                                H


                                                §


                                             < EH

                                             ^ >5
                                             w PQ  n
                                             O CQ  O
                                             CQ
                                             CL.     I—1
                                                K Cu
                                             <  w 5;
                                                O
                                                ffi
                                                o
861

-------

-------
DETROIT EDISON FULL-SCALE DEVELOPMENT PROGRAM
        FOR ALKALI SCRUBBING SYSTEMS
              J.H. McCarthy
              J.J. Roosen

        Detroit Edison Company
          Detroit, Michigan
              Prepared for
  Second International Lime/Limestone
         Wet Scrubbing Symposium
          New Orleans, Louisiana
           November 8-12, 1971
                     863

-------
     The text material for this paper was included in Paper




No. 4c, "Detroit Edison Pilot Plant and Full-Scale Development




Program for Alkali Scrubbing Systems—A Progress Report." There-




fore the paper is not presented here.  J.H. McCarthy and J.J.




Roosen of Detroit Edison Company were the co-authors of both




presentations.
                              864

-------
                             A FULL SCALE LIMESTONE
                          WET SCRUBBING SYSTEM FOR THE
                  UTILITY BOARD OF THE CITY OF KEY WEST, FLORIDA
                                 ROBERT R.  PADRON
                                 Superintendent of Engineering
                                 Utility Board of the City of Key West, Florida
                                 KENNETH C. O'BRIEN
                                 Supervising Engineer
                                 R. W. Beck and Associates
                                 Denver, Colorado
          In early 1972, the Utility Board of the City of Key West will put
into operation a full scale limestone wet scrubbing system on the initial  37 MW
unit of its new power plant, now under construction.
          This presentation outlines the decisions which led to the installation
of the system and includes a description of the system, economics, proposed
operation, and possible operating problems.
                                        865

-------
INTRODUCTION

          Key West, an island city, is approximately five square  miles  in  area
and located 160 miles southwest of Miami, Florida.   Key West,  as  well  as the
entire State of Florida, is widely known for its tourism, fishing and  water
recreation, and is boastful of its clean and clear  water.
          The Utility Board of the City of Key West provides  electrical  energy
for Key West, the U. S.  Navy, and adjacent communities  in the  lower keys.
The generating facilities for the system are currently  comprised  of approxi-
mately 60 MW of steam power plant generating equipment  and 17  MW  of package
diesel generators.
          In early 1968, design was started on a 37 MW  power  plant to  be located
on Stock Island, adjacent to the City of Key West.   The plant  was required to
maintain firm capability to meet the growing system load.
          In recognition of the increasing emphasis on  the protection  of the
environment and pollution control by both private citizens and governmental
agencies at all levels,  the Utility Board established a policy of considering
all practical methods of controlling pollution at the new power plant.
          At the time the plant design was being completed, the air pollution
control requirements for the State of Florida were  under revision.  Consul-
tation with the Florida State Director of Air Pollution and the Southeastern
Regional Director of Air Pollution for the Federal  Government revealed that
the revised criteria would be relatively strict with regard to S02 and             '
particulate flue gas emissions.
          The fuel for the new plant was residual oil,  selected on the basis
of availability and economics.  The residual oil available at the time contained
approximately 2.75% sulfur.  The fuel oil suppliers advised that  residual  oil
would be available at some time in the future with  a sulfur content of about
0.8%.   However, even with the lower sulfur content, preliminary calculations
indicated that the future air pollution control requirements  could not be
met without some form of pollution control equipment.
          The Board's policy to control pollution,  coupled with the certain
adoption of new air pollution restrictions and public opinion, resulted in the
inclusion of an SOo and particulate scrubbing system as a part of the  steam
generator specifications.  The scrubbing system was included as an alternate
bid item to allow the scrubbing systems to be evaluated separately.
          A number of scrubbing processes were at various stages  of develop-
ment at the time, but there was some doubt as to their commercial success.  However
there appeared to be sufficient prospective economies in using a  scrubbing
system to warrant careful investigation.


DESIGN PARAMETERS

          Once it was established that some form of flue gas scrubbing should
be considered, a survey of candidate processes was  made.
          Some of the factors considered were:
          1.  Estimated capital costs.
          2.  Estimated operating costs.
          3.  State of development.
          4.  Simplicity of Process.
          5.  Availability of raw materials.
          6.  Operating flexibility.
          Based on the evaluation of these factors, the  limestone wet
scrubbing system was selected for primary consideration.

                                          866

-------
           The  scrubbing  systems  normally employ scrubbing devices with
 proprietary  designs.   For  this  reason  the specifications were developed
 on a functional  or  performance  basis,  with the primary purpose to obtain
 proposals  on equipment which would  be  suitable for continuous operation with
 a base loaded  power plant.
           Some of the  functional  requirements included were:
           1.   A guarantee  of SC>2  and particulate removal as follows:
               a.  85%  of S02 produced  when burning fuel oil with
                  a 2.75%  sulfur  content.
               b.  95%  of all particulate matter generated during
                  the  combustion  process.
           2.   The scrubbing liquid  was  to be a slurry consisting of
               sea water  and native  coral marl.  The marl (limestone)
               is  the primary constituent of all land areas in the Keys,
               and is consequently readily available.  Sea water rather
               than  fresh water  was  specified due to the extremely high
               cost  of  fresh water in the area (in excess of $2.00/1000
               gallons.)
           3.   A complete system was required, including tanks, conveyors,
               pulverizers, storage  bins, pumps, piping and controls.
               The raw  marl delivered to the system was specified to be
               a maximum  size of 2 inches in diameter.
           4.   Materials  of construction were required that would withstand
               the corrosive effects of a sea water and marl slurry  in con-
               tack  with  flue gas.   It  was further required that the scrubber
               be  designed  to safely withstand a gas temperature of  650° F
               in  the event of an  air preheater failure.
           5.   It  was required that  the spent slurry be of such quality that
               the sea  water could be discharged into the bay and meet all
               applicable pollution  restrictions once the solids had been
               precipitated out  in a settling pond.
           6.   The manufacturers were required to submit operating costs
               for the  system to be  used in evaluating the proposals.
           The  specifics  of design necessary to meet these requirements were
 established  as  the  responsibility of the manufacturer submitting the proposal.


 EVALUATION OF  PROPOSALS

           Four proposals were received for steam generating equipment, three
 of which included proposals for limestone wet scrubbing equipment.
           Values  listed  in the  proposals for the scrubbing equipment ranged
 from $345,000  to  $738,000.  Although it appeared that there was some economic
 justification  in  installing a wet scrubbing system, a bid analysis was
 completed which  indicated  that  the  lowest bid for the steam generator and wet
 scrubbing  system  was substantially  in  excess of the funds available.  For this
.reason, the  steam generator contract was awarded without the wet scrubbing
 system.
          At this time it  was learned  that the National Air Pollution Control
 Administration  was  in  a  position  to make certain funds available for a full
 scale  limestone wet scrubbing system.
                                          867

-------
          A request was made by the Utility Board of NAPCA that Key  West
be considered for the installation of a limestone wet scrubbing system, to
be jointly funded by the Utility Board and NAPCA.  A series of negotiations
took place which resulted in a jointly funded full scale limestone wet
scrubbing system for the power plant.  The system selected for the project
was designed by Zurn Air Systems.   The Zurn system was selected on the basis
of the lowest total evaluated bid  which complied in principle  with the
specifications.  The scrubbing device is described by the manufacturer as
a modified impingement type scrubber.
          Concurrently with the negotiations between the Board and NAPCA,
a series of pilot plant tests were conducted at the existing Key West Power
Plant using the Zurn Air System's  scrubbing device, which demonstrated the
feasibility of the system.
          A flow diagram for the complete scrubbing system and a schematic
diagram of the scrubbing device are shown on Exhibit I.


DESIGN AND CONSTRUCTION

          A change order to the steam generator contract was executed in
October, 1970, to include the wet  scrubbing system.  Design and construction
progressed at a rapid rate and at  the present time the installation  is
approximately 70% complete.
          A considerable amount of review during the design phase was
exercised by the Utility Board and their Consulting Engineers  as well as  by
the Office of Air Programs for the Environmental Protection Agency (previously
designated as NAPCA) to provide as much constructive input into the  design
as possible.
          Design areas of particular interest were as follows:
          1.  Material selection -
              The ability of the materials used to withstand the corrosive
              conditions was considered critical to the successful opera-
              tion of the system.   Although it was known that it might be
              necessary to replace certain parts of the system with  more
              "exotic" materials during the initial operating period,
              materials were selected which appeared to be the most
              practical from economic as well as performance aspects.
              In general, stainless steel combined with non-metallic
              coatings was used whenever corrosive conditions were
              anticipated.  Some material selections were based on
              experience gained during the pilot plant tests.
          2.  Safety -
              The performance of the wet scrubbing device is directly
              related to the slurry level in the scrubbing device.
              Close control of the slurry level is required.  Interlocks,
              automatic overflow devices, and an automatic flue gas
              bypass were included in the design to avoid the possibility
              of the slurry level  rising above the tubes and consequently
              blocking off the flue gas flow.  Additional interlocks and
              safety devices were  included to protect materials not
              designed to withstand the high inlet gas temperatures  in
              the event of sea water supply failure.
                                         868

-------
          3.  Operation -
              The control system was designed to allow normal  continuous
              operation of the scrubbing system under a relatively  constant
              load.  Automatic operation and simplicity were  considered
              to be key design criteria for the controls.   Provisions  for
              instrumentation required during the Office of Air Programs'
              test were included.
          Additional design criteria were established by the  manufacturer
as a result of the pilot plant tests being conducted in Key West.
          The general arrangement of equipment for the system is shown on
Exhibits 2 and 3.
          The entire power plant, including the scrubbing system,  is
scheduled for start-up in the early part of 1972.


PROPOSED OPERATION

          No operating experience other than that gained during the operation
of the pilot plants is available since the wet scrubbing system for Key  West
is still under construction.   Therefore, the procedures outlined in this
section are tentative, and will  be modified as required when  the system  is
put into operation.
          The various components of the wet scrubbing system  and the  material
flows are shown on the flow diagram, Exhibit I.
          In general, the wet scrubbing system is divided into two  sections:
The first section is the marl preparation plant.  Here coral  marl,  2  inches
in diameter or less, is received in a pick-up hopper, and after initial
crushing to 3/4 inch size is  placed in the storage bin.  The  bin has
sufficient capacity to hold a 2-day supply of marl under full  load  conditions.
The material handling equipment up to this point is operated  from local  manual
control stations, on an intermittent basis.
          The marl is then fed through a roller mill which reduces  the particle
size to 90% - 325 mesh, and then on to the surge bin, ready for mixing with
sea water.  The roller mill feed rate is not easily adjustable and  for this
reason must be set for the average expected load condition.  It must be  started
nanually and will run continuously.
          The second section  of the system is the actual scrubbing  section.
Here marl is fed in correct proportion to sea water from the  surge  bin into
the mixing tank to form the desired slurry concentration.
          The slurry is pumped to the two scrubbers, each handling  half  the
:lue gas.  The partially spent slurry is pumped out of the scrubber,  some
jeing recirculated to the scrubber and the remainder being pumped to the
settling ponds.   The slurry flow to the scrubbers and the slurry level in
;he scrubbers are automatically controlled, based on manual set points
established in the main power plant control room.  The set points are
jetermined by the unit load and readings from the S02 monitor located
 n the stack.
          The flue gas passes from the induced draft fan through a  spray
;ection where the gas is cooled from 350° F to 160° F.  It then passes
;hrough the scrubbers where the S02 and particulates are removed, and
;hen to the stack.  The flue  gas pressure drop across the scrubber  is
 bout 12 inches  F^O and is a  function of the slurry level in  the scrubber.


                                         869

-------
          The scrubbing section can be run on a relatively automatic basis
from the main control room once the equipment has been started.   A number
of automatic controls are included to allow safe operation of the unit.   A
bypass around the scrubbers is provided for maintenance purposes.
          The Office of Air Programs of the Environmental  Protection Agency
will be performing tests on the system during the first yeac. of  operation.
It is anticipated that these tests will aid in establishing the  optimum  flow
rates and equipment operation.
          Two settling ponds will  be provided.  One will be emptied while
the other is in use.  It is anticipated that a front-end loader  will be
required to handle this material.


POSSIBLE OPERATING PROBLEMS

          Anticipated operating problems, particularly in the areas of corrosion,
scaling and disposal of spent slurry, are major factors that must be taken  into
consideration in formulating operating procedures.
          The problems of corrosion and scaling were observed in the initial
operation of the pilot plants and resulted in material and design changes.   It
is anticipated that it will be necessary for the contractor to make a number
of additional changes once the system is put into operation.
          To combat corrosion, the stainless steel  scrubbers were epoxy  lined
and the tube material was changed to fiberglas.  It is possible  that the change
of tube material will also solve the scaling problem, but if not, it was dis-
covered that by installing sprays and maintaining the tubes in a constant wet
condition the scaling problem would be eliminated.
          Disposal of 50 tons per day of spent slurry is an obvious operating
problem and has not been completely resolved to date.  The initial thinking
was that of utilizing the product to fill approximately twenty acres of  City
owned baybottom land adjacent to the steam plant site.  Although this matter
is still being pursued, the State of Florida's recent stand of opposing
the filling of any submerged land productive to marine life has  made it  a
less plausible solution.
          Initially the slurry will be placed on land above sea  level to avoid
this problem.  Other methods of disposal are currently being investigated.
          Wind tunnel tests were conducted during the design of  the system
tg determine if a reheat system would be required for the flue gas.  The
results were inconclusive.  It will therefore be determined during the initial
operation of the plant if this will be a problem.


SYSTEM ECONOMICS
          The cost information shown on Exhibit IV summarizes the estimated
capital investment requirements and operating costs for the scrubbing
system.  A cost evaluation, also shown on this plate, relates the system
costs to the plant capability and output, and also compares them to the
cost of burning low sulfur fuel.
                                          870

-------
CONCLUSIONS

          The mandate has been issued to limit S02 and  participate  emissions
from power plants.
          The limestone wet scrubbing process  does not  appear to  be the
universal answer in reducing these emissions,  but is  considered to  be  the
best economical  selection for the Stock Island Plant.   It is  likely thac
the selection of limestone wet scrubbing equipment will  become economically
more attractive  as the demand for low sulfur fuel  increases.
          A certain number of operating problems  must be expected during
the initial phases of operation, but it is  anticipated  that the solutions  to
these problems are available within the limits of the current technology.
                                          871

-------
H 'A J. OZ HOJ.VA313
i3«na
HdO.01 5
>-.'
0


n


Ul

z.
t-



y
\
\
872

-------
0.1
I

z
LU
5
LU
O
C£
<£
LU
LU|
O
   CC
   o
   (0
   Ul
                              /--•  -'
                            xf    x
                                 x, ;-•
^x
                                          v.
                                                  tur
                               873

-------
874

-------
         o olo
         o olo
         o olo
o
o;
NNUAL OPE]


U



COSTS

"

ne stone)

^.
t,
1


hi
V
°S
££

.5 D
'x ""*
3 .
*^ ei

t
00
C
•O
C
"rt
bi
0}
rt
*
j-
fM
g
2
U
00
C
1
u

*V h.
JD


in
'rt
CU
£
(4
O
C

V
.S
2


1
"
t/) P
J 'o

0 |
 «$•


t)
sf conduit wiring
nstructlon including wet
control center
Mechanical co

t/i



3
S
a
00
C
o.
'S.
1
S.
o
hi
Ling equipment for marl and
scrubbing con!
control tubing
Material hand:
spent slurry






•+Q



TOTAL
Engineering

r-










(4
0)
O
o
o

0



n)
2
S
o
r-
00
X
in
00

X
^
O
o
0

r-




X
J<
(D
^3
1
r-
vO
o"
l(






«
V

w "1
x: S
> >-.
v c
U («
*• 5
* ^
gl
§ 5
u ^
c °
0 <=
11
f4 ^3
•° ^
* u
sS;
^ Z
rt
S

a.
'5
4)
00

1
3
hi
u
T3
r4
1
c
'o
^*
in
*>
i ma. ted
t.
c


n
S
.2

.3
0.
(4
U
"o
hi
rd
O
>*
hi
S.
o

*»
(4
-o
5
S K _
   •3   S
   'o
       §s
       *> pq
                                                                                .S

                                                                                'a
W
to
O
U
O
g
H

U 2
°* t
° 3
Q 2
U] O
2 ^
yj Z
< **•




-
w
O
u
rt
•d
0
*j
C
V
rt
h
C*
C
S
u
3
C
00
1)
T)
hi
m
V
d



|
§
'x
g
"c
conte
)H
3






4-1
M
0
u
I
0
u
p.

OO
1
hi
3
*3
U)
£
0
O
in
O *•
U w
o
rt u
^ h

1_l
VOTUZ1
V
•S
M
C
j5
*• "§
V) hi
0 0
U (0
K ^
U °
? >.
0 h
P*. ti
>.s
   .
      , S £ .
             -
             >< 3  «
             2 i!  v

^
^
rt
a
U
"c
fS
S
^_,
o
U

d
rt
U
O
o
o
o"

* «
•C 0
o U
i 1
^ * I
o ** rt
Tp O ^
r- [-^ ^

0
h
«.

OB
0
pe rating c
o
*rt
3
G
G
<

<4

M (Nl


O
O
in
r-

^
OB
O
nt annual <
u
rt
^
3
CT
Ul

*•



«
>*
o
o
00
„**
^2-
^
1^
« H
IB
4
5.
n)
" in
o W
H
O
H
O
O
U,
^ S
jj hi
1 1
a +?
f 3
c S
9 'M
« U
f-a- 6
? V 0
^ O -H
-C U *
*j ui O
*. fl 0.
S^ 2
•2 « a
c ^ c
03 o
•O j T3
« TJ «
fl C ^
W (4 CQ

— —,
C —
* S
"^
11
« 8
1 ••
h ^
31
6^
o «
j5
c «
M O
« **
C T3
0 0
T3 G
V 3

-------

-------
    AIR  POLLUTION CONTROL AT THE
  NORTHERN STATES POWER COMPANY
SHERBURNE  COUNTY  GENERATING  PLANT
                     By

                  J. A. NOER
               Mechanical Engineer
     Plant Engineering and Construction Department
          Northern States Power Company
              Minneapolis, Minnesota
                A. E. SWANSON
            Director - Nuclear Activities
        Black & Veatch, Consulting Engineers
               Kansas City, Missouri
                Presented Before
             The Second International
   LIME/LIMESTONE WET SCRUBBING SYMPOSIUM
              New Orleans, Louisiana
                    1971
                     877

-------
s—
z

-------
                   ACKNOWLEDGEMENTS
The  authors  express their  appreciation  for  the suggestions by  the
many people from  Northern States  Power  Company and Black &
Veatch who reviewed the paper.
                        879

-------
                               ABSTRACT
The  paper describes  the Air Pollution  Control System  and related  consider-
ations  of the two-unit  Sherburne County Generating Plant. The plant will be
located near the  Mississippi River  about 40  miles northwest  of Minneapolis
and  will have  a net  capacity  of 1360  MW.  Fuel will be low  sulfur western
coal  delivered   by unit  train.  Commercial  operation  of  the two  units  is
scheduled for 1976 and 1977 respectively.

Northern  States Power Company  (NSP) encouraged the  formation of the
Citizens  Advisory Task Force comprising concerned citizens  and  represen-
tatives of public agencies and conservation organizations. This group's contri-
bution in  the  plant  site selection  and  in the development  of environmental
protection  criteria is discussed.

The  authors  present the background and considerations which led NSP to the
decision  to use  a limestone  wet-scrubber for  paniculate collection  and SC>2
removal. The scrubber  selected is the Combustion Engineering Tail-End  Lime-
stone  Additive  System  using  a  10  inch  layer  of glass  marbles  as  the con-
tactor.  The  paper describes  the  components and  physical  arrangement  of
equipment as  well as  simplified  flow  diagrams  and operating features. The
status  of the  scrubber system development   including  effluent  control, cost
estimates, and  schedule requirements is  given.
                                880

-------
                AIR POLLUTION CONTROL  AT  THE
               NORTHERN  STATES POWER  COMPANY
            SHERBURNE COUNTY GENERATING PLANT
INTRODUCTION
Northern States  Power Company (NSP) will  construct a two-unit, 1360 MW
electric generating plant near Becker, Minnesota for commercial operation in
1976  and 1977. The fossil  fuel plant, named Sherburne County Generating
Plant, will burn low sulfur western  coal.

The  two previous NSP base-load generating plants  are nuclear  plants, and it
was NSP's desire to  build a nuclear plant for commercial  operation in 1976
and 1977. Due to  the regulatory uncertainties at both the State and Federal
level,  the growing public  resistance  to siting of nuclear plants, and finally the
amount  of nuclear  generating capacity in the Company's system, a decision
was made to proceed with a fossil fuel plant. Through  concerted efforts to
involve the public in  environmental planning, the use of low sulfur  coal,  and
the use  of  advanced  technology  for pollution  control, it  appears  that  a
coal-fired  plant  will  provide  acceptable  environmental protection and avoid
delays associate^, with nuclear plants.
                                  881

-------
PLANT DESCRIPTION
The  Sherburne  County  Generating  Plant  is  located in  Becker Township,
Sherburne  County, Minnesota,  approximately 40  miles north  of Minneapolis.
(See FIGURES 1 and 2.) The site area is about  1,300 acres.
The  site is now  cultivated land with  an average  elevation of  965 feet  above
mean sea  level. The land  is nearly  level, with a  maximum elevation differ-
ential of about  10 feet. The main plant facilities are set back  more than half
a mile  from the Mississippi River, so that an undisturbed zone  is maintained
along the  waterfront. (See FIGURE 3.)  The  normal  river  water elevation  is
920  feet mean sea level.
The  plant  will comprise two  electric  generating  units  (Units No. 1 and No. 2)
each rated  at 680,000  kilowatts net  electric generating  capacity. The  steam
generators,  turbine generators, wet scrubbers, and  associated  auxiliary equip-
ment will  be  enclosed in the main building structure.  Unit  No. 1 is scheduled
for  commercial operation  on  May  1, 1976, with  Unit No. 2 scheduled for
operation one year later.
Black & Veatch  (B&V),  Kansas  City, Missouri,  is  the Architect-Engineer and
has  overall design responsibility for  the  plant. Management  of design, con-
struction, and quality assurance  is the responsibility of NSP's  Plant Engineer-
ing and Construction Department.
Capital  expenditures  are estimated to be  over $360,000,000  for  the  plant.
This includes  more than $25,500,000 for the air pollution control system.
The  steam  generators will  be  supplied by  Combustion Engineering,  Inc., and
will  be of the balanced-draft type designed for burning pulverized coal.  Each
steam  generator  will be rated  at  4,985,000  pounds  of  steam  per hour and
will  require a gross heat input of 6,757 million  Btu per hour.
Each turbine  generator  will have  a gross  capacity  of 720,000  kilowatts and
will  be supplied  by General  Electric  Company.  The  turbine inlet  steam
pressure will  be  a  nominal 2,400  pounds per square inch (psig). The inlet
                                   882

-------
steam  temperature  will  be  1,000  degrees Fahrenheit  (F)  and  the  reheat
temperature  will be  1,000 F. Each  generator  will  have a rating  of 800,000
kVA at 0.90 power factor.
The  furnace will be  equipped  with  tilting  tangential  burners  and will be
designed for low combustion temperatures which  are  expected to reduce to
the practical minimum  the  formation of nitrogen oxides.
Each  steam generator will  be equipped  with  a wet flue-gas scrubber, capable
of removing 99  per cent  of the  paniculate  matter and 50 per cent  of the
sulfur  dioxide in the combustion gases. The scrubbers are described  in greater
detail later in the paper.
Combustion  gases leaving the scrubbers  of the two  units will  be emitted to
the atmosphere  from a single chimney at least 550 feet tall.
Each  electric generator will be  connected through a power transformer  to  a
transmission  substation. The substation  will  feed  four 345 kV  transmission
lines supplying power to  the existing NSP electrical transmission system.
The  turbine  condensers  will be  cooled  with water from  cooling  towers
operated as  part  of a  closed-cycle circulating  water system. The  towers will
be  of  the  wet, mechanical-draft,  cross-flow type  and  will be  oversized to
minimize visible  plumes.  The cooling  tower for each unit will consist  of 20
cells,  each  equipped with  a 28-foot  diameter  fan capable of moving about
1,500,000  cubic  feet  of  air   per  minute.  The  air-vapor mixture will be
discharged  from individual fan  stacks about 60 feet above grade. The circu-
lating  water flow  rate for  each unit will be  approximately 240,000 gallons
per minute  (gpm).  The  cooling tower for Unit No. 1  will be located 3,000
feet  from a similar tower for Unit  No. 2  to reduce  the  combination of plumes
from  the two towers.
Fuel  for the plant will  be  low sulfur  (0.8 per cent  or less)  subbituminous
coal  from  the Colstrip area of Montana.  Coal shipment  is to  be  made by
unit  trains. Coal  handling facilities at the plant  site  will include automatic
sampling facilities,  rotary  car  dumper,  stacker-reclaimer,  crushing  facilities,
                                    883

-------
and  a  system of conveyors  for transfer of coal  between  elements  of the
system. A generally circular railroad  track, which will handle the unit trains,
will  enclose the  coal  storage area at the plant site. Storage at the plant site
will  provide  a  90  day  supply  of coal.  This  will require approximately
1,600,000 tons of coal,  stored to  a  depth of about 40  feet.  The  maximum
design burning rate is 814 tons per hour  for the two units.
Ash  from  coal combustion will be collected in  the bottom  ash hoppers and
scrubbers associated with each unit.  Ash from these collection  points  will be
sluiced  to a  water-filled ash  storage basin  formed  by an earthen  dike.  The
bottom of the storage basin will be  sealed  to  minimize  seepage into  ground
water.
Water  sources for operation of  the  plant will  be  the Mississippi  River and
wells. After maximum reuse, all  process  return water from  the various  plant
systems will  be directed to a common water holding basin  with a minimum
retention  time of 24 hours. Water for the ash handling systems will be reused
process water.  All  process return water  which  cannot  be  recycled  will be
treated  to  meet  applicable water  quality standards prior to  release  to the
river at a single point of discharge.
BACKGROUND OF  ENVIRONMENTAL CONTROL
In  Minnesota,  as in  most other states, there has been no effective method of
resolving  conflicting  viewpoints  concerning  the  environmental  aspects  of
power plant siting and  development.  In  an effort to find a new  approach to
resolution  of  conflict,   NSP  discovered that  many environmental  conflicts
involved in its  past  construction programs  arose  from  a  lack  of early public
participation in  its environmental  planning activities. Northern  States Power
Company  also   realized  that  the   public  hearing  process is  an inadequate
method  for  communicating with the  public or involving the public  in  plan-
ning and decision  making  processes. As  a  result of evaluating  past environ-
mental planning programs, NSP  created the Citizens Advisory Task  Force to
                                    884

-------
provide a  forum  for  in-depth  discussions  on  environmental problem  solving
and power plant  siting.
Participating in the new open-planning  group were representatives  from the
Minnesota  Conservation  Federation, the  Scientific Study Areas Committee of
the  Minnesota  Academy of  Science,   Minnesota-Wisconsin  Boundary  Area
Commission, Minnesota  Chapter of the  Sierra Club,  Minnesota Committee for
Environmental   Information,   Minnesota   Environmental   Control   Citizens
Association, Minnesota Chapter of the  Isaac Walton League, Minnesota Envi-
ronmental  Defense Council, Save  Lake Superior Association,  Zero Population
Growth,  St. Croix River Association, Clear Air — Clear  Water Unlimited, and
the League  of Women Voters.
Northern  States  Power  Company  believes  that it is  important to  seek  out
those individuals who can effectively represent the  concerned public to assist
in major decisions relating  to  power plant siting and pollution  control. There
are two  reasons  for  this  approach: (1)  the installation  will  more  closely
represent  the  public  interest;  that  is,   it  will be  consistent  with what the
public appears to  be willing  to  pay for,  and (2) delays  in  securing permits
and  licenses may be avoided by  using  the  advice of those who represent the
public.
After about six weekly  meetings,  the Citizens Advisory Task Force generally
lost  its hostility  toward  NSP and developed a deep concern  in accomplishing
the objective  of plant site  selection.  The group  took under advisement  four
alternate sites to determine which site should be  selected for a 680 MW fossil
fired  power  plant  to be in service in   1976.  The Task Force recommended
the Sherburne  County site  and NSP concurred, even though another location
was the Company's first choice.
Northern  States Power Company  conducted a thorough  investigation of alter-
nate methods of air  pollution  control for Sherburne County, in line with the
following Task Force guideline contained in the group's  report:
     Emission  of all air-pollutants should be reduced to the minimum
     technologically  feasible.   The  cleanup should  include  not  only
                                     885

-------
     paniculate  matter  but  also  gaseous  pollutants such  as sulfur
     oxides. The Task Force noted the Company's stated intention to
     use low  sulfur western  coal  in  its next plant,  but cautions the
     Company  against  regarding  this  use  as removing  the need for
     sulfur oxide  cleanup  as  it becomes  technologically feasible for
     other coals. Adequate space should be provided for future addi-
     tions of air pollution control equipment to accommodate possible
     future   technological  developments   and  revised  air  emission
     standards.

The  collection  of fly  ash from low  sulfur coal is  difficult  with a  "cold"
electrostatic precipitator  (with a flue  gas temperature at the precipitator inlet
of about 300 F)  due to  the  high resistivity of  the ash  at normal  exit gas
temperatures.  There are also problems of  energizing  a  "cold" precipitator at
low  loads and  startup.  For  these reasons, a  "cold"  precipitator  was  not
considered practical. The  alternatives considered were  a "hot"  precipitator
(with a  flue gas  temperature  at the  precipitator  inlet  of about  700 F) or a
wet  flue gas scrubber.  Even though economics  were somewhat in favor of a
"hot"  precipitator,  NSP  selected the  wet  scrubber  because there is  a high
degree  of confidence in  collecting particulates and there would be some
reduction in  the sulfur dioxide emissions.
SCRUBBER DESCRIPTION

The air  pollution  control system includes  a tail-end limestone  wet scrubber
designed to remove both  particulate and  809 from the boiler flue gases. The
manufacturer  guarantees  that the  system  will  remove  99 per  cent  of the
particulates and  50 per cent of the SO 2 entering the scrubber  if either high
calcium limestone or dolomitic limestone  is  used as an additive.

The flue  gas  enters  each module below a  bed  of glass marbles at a design
temperature of 290 F. (See FIGURE 4.)  As the gas turns upward,  large particles
are separated  immediately. Sprays provide a constant  supply of water to the
underside of the bed of marbles.  Water entrained  in the gas stream  floods the
spaces between  the marbles.  Violent  mixing and physical contact occurs which
                                     886

-------
results  in  the  capture  of  substantially all  of  the  particles  larger  than 5
microns and a high proportion of the  particles smaller  than 5 microns. The
cleaned and cooled  moisture laden  gas  goes up thraugh  demisters which
remove entrained  water.  The  gases  are then reheated from 120 F to about
163 F  to  (1)  provide  dry gas  for protection  of the ID fans, (2) reduce visual
pluming  at the stack  exit  and  (3)  enhance buoyancy which increases  the
effective stack  height. The water  containing  the  particulates overflows  the
turbulent bed and is drained to a clarifier for particulate concentration.
The  SC>2  is  removed from  the flue gas  by  chemical  reaction  with  the
limestone  additive  in  the  marble bed contactor.  The  overall reaction can be
shown as:
            CaCO? + H0O + SO9 	*• CaSO, + CO9 + H9O
                 J     £        £             j      £•    £
                            and continuing
           2CaSO3 + O2-*-2CaSO4

The  calcium sulfate and calcium sulfite are only  slightly soluble in water and
are therefore mostly  precipitated. The precipitated solids are carried with  the
scrubber bed drain water  along  with the  fly ash  to  the clarifier  for  concen-
tration  prior  to disposal in  the  ash  storage area.   The  recirculating tank
is  the  source  of  the  scrubber spray water  and  is  made up of the fol-
lowing:

     * Clarifier  overflow
     • Additive slurry
     * Makeup water to replace  evaporation  and blowdown
A  soot blower is located  in the gas  inlet  duct to each module to clean  the
scrubber at the dry  gas — moist  gas interface.  Soot  blowers  also clean  the
heat transfer surface  of the  reheater.  Water washers are provided to clean  the
demister section. Experience gained  from operation of the  scrubbers at Unit 4
of the Lawrence Station  on The Kansas Power and Light  Company  system
indicates that  this cleaning equipment will be adequate.
                                     887

-------
Velocity  of  the  gases through  the  bed of  marbles  must be  maintained at
above  75  per cent  of  rated  flow  to  create  the required  turbulence  for
paniculate  removal. This is  accomplished by automatically removing scrubber
modules from service  in groups of three as the load is reduced.
The  modules  are arranged in four groups  of three as shown on  FIGURE 5,
Plan View. Flue gas crossover ducts at the air heater outlet and ID fan inlet
permit flexibility  of  ID  fan and module operation. The arrangement  will
enhance the reliability of operation of the  flue gas  scrubber system.
The  additive will  be  wet ground and sluiced  to the recirculation tank.  The
feed quantity will be controlled  as a  function of  the amount  of SO? in the
inlet to the scrubber. It is  expected  that  100 per cent  of the stoichiometric
ratio will  be required for 50 per cent SO^ removal.  Dolomitic limestone (40
per cent MgCO-j — 50 per  cent  CaCO-j)  is  being  considered as  an  additive
because it  is available near  the  plant. High calcium limestone would have to
be  shipped from areas  remote  from  the  plant.  The MgCOj  will react  with
SC>2 to  form  MgSO? and  MgSO^. The solubility of MgSO^  is very high,
which  may complicate the  ability of the scrubber  to form a solid precipitate
of the 804  ion. This could result  in a  less  desirable  blowdown from the
scrubber  for eventual  discharge to the  environment.
RESEARCH AND DEVELOPMENT
The  scrubber  components  are well along  in  the development stage.  The
scrubber system development, howe/er, is still in progress, with emphasis on
the following items:
       • Reducing quantity of the scrubber  system effluent
       • Determining  quality of the effluent and the  effect on river
         water quality
       • Selecting type and determining amount of the additive
       • Confirming system and component  reliability
       • Confirming scrubber  performance  with  respect to the guar-
         antee

-------
A task force  comprising  personnel  from NSP, B&V, and Radian Corporation
is working closely with Combustion Engineering to  solve these  problems.
The  highest priority is placed on minimizing the impact of the effluent with
respect to both quantity  and  quality.  Tests are  being  conducted  on  the
Combustion  Engineering  pilot system  in  Windsor,  Connecticut,  to minimize
the blowdown from  the scrubber system by using a  seeded crystallizer tank
to precipitate calcium  sulfate and  calcium  sulfite. The tests are being  run
initially  with  a  high  calcium limestone  additive.  Further  tests are planned
with  dolomitic  limestone  and western coal  fly ash.  Radian  Corporation is
supplying input with  respect  to reviewing bench  studies and  analyzing pilot
system  scrubber  chemistry.  The  design  of the   scrubber  components   and
scrubber system  is under  careful review  to maximize reliability and perfor-
mance.
TABLE I  and TABLE II  summarize the important river water  quality para-
meters  that  can be  affected  by the  scrubber effluent. The  total dissolved
solids would be slightly increased immediately downstream  of the plant,  but
this  effect  would be  hardly  noticeable  at  the  Minneapolis water  intake
approximately 35 miles downstream.
                                 TABLE  I
                        RIVER  WATER ANALYSIS
             AT SHERBURNE COUNTY GENERATING PLANT
                                  Prior to Plant      After Plant Goes
                                    Operation        Into Operation
       River Flow, cfs                               1,223*      3,322**
       Dissolved Solids, mg/1
          Ca                           40.1         43.6       41.4
          Mg                           15.6         18.1         16.5
          Na                           4.8         4.8         4.8
          M(HC03)                     181.0        178.4       180.1
          S04                          10.0        30.2        17.4
          Cl                            3.6         3.6         3.6
          NO^                          1.9         1.9         1.9
          SiO^                          5.8         5.8         5.8
       Total                           262.8        286.4       271.5
       Carbonate Hardness               148          146         148
       Noncarbonate Hardness             16         37         24
        * Flow exceeded 90 per cent of the time
       * * Flow exceeded SO per cent of the time
                                    889

-------
                                 TABLE II
                        RIVER WATER ANALYSIS
         AT RIVER INTAKES FOR MINNEAPOLIS AND ST. PAUL
                                  Prior to Plant      After Plant Goes
                                   Operation         Into Operation
       River Flow cfs                              1,729*       4,781**
       Dissolved Solids, mg/1
          Ca                           46.9        49.4         47.8
          Mg                           15.4        17.1         16.0
          Na                           2.1         2.1          2.1
          M(HC03)                     207.4       205.6        206.7
          S04                          10.8        25.1         16.0
          Cl                            4.0         4.0          4.0
          N03                          0.7         0.7          0.7
          Si02                          8.6         8.6          8.6
       Total                           295.9       312.6        301.9
       Carbonate Hardness               170         168         169
       Noncarbonate Hardness             10          25          16
        * Flow exceeded 90 per cent of the time
       ** Flow exceeded SO per cent of the time

Northern  States Power Company  is actively involved in air  pollution control
planning.  There  are  a number  of  areas  that  NSP believes require  more
research  and development  in the  near future. Much effort has been made to
develop systems that will  provide SC>2  control and particulate removal. These
areas will require more attention.  NOX emission is becoming  a major concern,
and NO   reduction systems will need  to be developed.
The water effluent from scrubbing  systems must be minimized or eliminated.
if there  is an  effluent,  it  must be of acceptable quality. It is not prudent to
exchange  one problem for another. Much  effort has been made with respect
to particulate  removal  and  SO2  removal, but  very  little concern has been
given  to  the blowdown effluent from  scrubbing systems.
The design  and development of scrubbing systems must consider reliability of
operation since the scrubbing  system  is an integral part of  the power plant.
Unless the scrubbing system is reliable,  the plant itself cannot be reliable.
                                     €90

-------
                             REFERENCES
California Institute of Technology, Environmental Quality Laboratory, People,
Power and Pollution: Environmental and Public Interest Aspects of Electric
Power Plant Siting, September 1,  1971. Report No. 1.

Black &  Veatch, Northern  States  Power  Company Environmental  Report,
Sherburne County Generating Plant, May  24, 1971.
                                 891

-------
          LIST OF  FIGURES
FIGURE 1    PLANT SITE LOCATION MAP
FIGURE 2    PLANT SITE VICINITY MAP
FIGURE 3    PLANT SITE ARRANGEMENT
FIGURE 4    FUNCTIONAL SCHEMATIC
           FLUE GAS SCRUBBER MODULE
FIGURE 5    PLAN OF
           FLUE GAS SCRUBBER
           MODULES
                 892

-------
                                    SHE\BUR|NE NATIONAL
                   T. -CtlOtm- -   —V   -(WILDLIFE  REFUGf
 PLANT  SITE
-St+ERBttRNE COUNTY.,
 GENERATING PLANT
                                                     PLANT  SITE LOCATION  MAP
                                                          5    0   5   10   15
                                                               SCALE MILES

                                                               INTERSTATE HIGHWAY
                                                               U.S. HIGHWAY
                                         893
FIGURE I

-------
   ;-- - -|30  ^J
         B  ~-
   - SHERBURNE  COUNTY ^
     GENERATING PLANT „,
EXISTING NUCLEAR
PLANT
                              PLANT  SITE VICINITY MAP
                                        SCALE - MILES
                                 CONTOUR INTERVAL - 20 FEET
                                   DATUM IS MEAN SEA LEVEL
             894
FIGURE  2

-------
EFFtJpT
TREftTMT
FACILITY
                          PLANT SITE ARRANGEMENT
            895
                                                FIGURE 3

-------
                                                O  OD
                                                CO  CD
                                                •<  O
                                                Z  CO
                                                O
                                                -  CO
                                                I—  •<
                                                O  CD
896

-------
                                                              ID FAN
AIR HEATER GAS OUTLETS
INLET DAMPER
   (TYP)
                                        STEAM GENERATOR
                                                  PLAN OF
                                             FLUE GAS SCRUBBER
                                                  MODULES
                          897
    FIGURE 5

-------

-------
ONTARIO HYDRO'S PROTOTYPE LIMESTONE  SCRUBBER FOR
        S02 REMOVAL FROM CLEAN FLUE  GAS
                   J.W. James
                 Ontario Hydro
                Toronto, Canada
                 Prepared for
     Second International Lime/Limestone
            Wet Scrubbing Symposium
             New Orleans, Louisiana
              November 8-12, 1971
                      899

-------
      ONTARIO HYDRO'S PROTOTYPE LIMESTONE SCRUBBER FOR
      S02 REMOVAL FROM CLEAN FLUE GAS - by J.W. JAMES
ABSTRACT

          Ontario Hydro is currently designing a 30 MW demonstration

limestone scrubber and expects to install it at one of its

stations near Toronto.

          The system, which is planned for initial operation in

1973, will use limestone slurry in a spray tower contactor to

scrub SO- from clean flue gas.  The gas will be cleaned by an

existing 99% efficient electrostatic precipitator.  This prototype

will be sufficiently large to allow confident upscaling to a full

size 150 MW module, and should be particularly applicable to

existing units fitted with fly ash collectors.

          The design of the prototype is based on a system devised

by Ontario Hydro's Research Division and tested in a 4000 cfm

pilot plant with flue gas from a coal-fired boiler.

          The reasoning involved in selection of this system is

presented, along with the considerations to be investigated

during the tests.
Presented at the Second International Lime/Limestone Wet Scrubbing

Symposium, New Orleans, November 11, 1971

                                  900

-------
                  ONTARIO HYDRO'S DEMONSTRATION  LIMESTONE
                               /AL OF S02 Ff
                                J.  U. JAMES
SCRUBBER FOR REMOVAL OF S02 FROM CLEAN FLUE GAS
1.       Most of us are now aware that the route to the development of
    a successful S0? removal  process leads through a beautiful  garden
    having many inviting and  expensive pathways.   Choice of the route
    to follow is influenced by faith, hope, and much scientific
    endeavour; and is fraught with concern of being led down the wrong
    garden path.  - Having had some exposure to this process,  I would
    like to speak briefly today on the recent choice at Ontario Hydro
    in favour of a 30 MW scrubber to demonstrate the removal of S02
    from one of its coal-fired boilers.

2.       During the past few years, our top management have been
    anxious to develop an S02 removal process,  through work on  a large
    demonstration unit installed in one of our own plants.  Many systems
    have been analyzed with this objective in mind.  For a variety of
    reasons, all were rejected.  Some were already being developed in
    demonstration units, while others were either not suited to our
    needs or appeared to hold little promise.  During this period of
    search, we entered into two multi-sponsor agreements to support
    well known major Research and Development projects.  However, it was
    not until the completion of our Research Division's most encouraging
    work on a 4000 cfm pilot plant, using limestone slurry in  a spray
    tower scrubber, that we had a viable candidate for a large  demon-
    stration system for development on one of our generating units.
3.       The pilot plant and its performance is described in a
    paper entitled "Sulphur Dioxide Removal by Limestone Slurry in
                                     901

-------
    a Spray Tower"  by A\  Saleem,  D.  Harrison  and  N.  Sekhar.  This
    paper was presented during Tuesday's  session.   The  4000  cfm  pilot
    plant achieved  about 75% removal  of S0?  from  clean  flue  gas  which was
    drawn off the ductwork following the  electrostatic  precipitator of
    a 300 MW coal-fired unit.   The spray  tower operated with a high
    degree of reliability; and inspection,  following a  1000  hour
    continuous run, indicated  no  hard scale  formation,  although  some
    surfaces did have a soft deposit.
 4.          In general, the large scrubber will be  similar to the
    pilot, and its  design will be based on  the optimized process
    variables from  the pilot work.  Provision will  be made to operate
    at off-design conditions as indicated by the  demonstration tests.
    - Aerodynamic model tests  may be necessary to establish  the  large
    scrubber vessel shape.  These would optimize  vessel  configuration
    to accommodate  structural  and space considerations,  while ensuring
    a satisfactory gas velocity profile in  the scrubber and  demister
    over a 25 to 100% gas flow range.
5.           The choice of limestone slurry  in a spray tower  contactor
    to scrub clean  flue gas from 30 MW of generation was based on  the
    following reasoning:
    a)      the limestone slurry process  appeared to have a  better
            chance of early development than other  SCL  removal processes.
    b)      the choice of a spray tower scrubber  was based on the  pilot
            plant experience.  Some of its advantages are:
            - it holds promise of fairly high reliability.   This
              results from the simplicity of the  tower, which has
                                      902

-------
             a minimum surface on which deposit can form or
             collect.  Also, deposit formations on the demister
             and scrubber surfaces are expected to be sufficiently
             soft to allow removal by slurry or water sprays.
           - it is much less sensitive to gas flow turndown than
             either fixed or moveable bed scrubbers.  This is
             particularly important on our generating system
             because of the cycling demand on fossil units.
           - it has a very low flue gas pressure-drop, and this may
             allow it to be retrofitted to existing units with
             minimum modification to the ID fans.
           - its SCL removal efficiency is expected to be at least 70%
             and limestone consumption is about 1.3 stochiometric.
             This is considered to be a satisfactory performance if
             high reliability is achieved.
c)      It was decided to demonstrate SCL removal while scrubbing clean
        flue gas since it reduced the chemical problems introduced by
        fly ash.  All of Ontario Hydro's existing fossil plants are
        fitted with high efficiency electrostatic precipitators, thus
        a clean gas system can be directly applied to its existing
        plants.  For new power plants, the environmental pressures for
        dry disposal  of fly ash may result in continued specification
        of precipitators.  Alternatively, substitution of a high efficiency
        venturi scrubber for particulate removal  is not expected to
        adversely affect the spray tower operation.
d)      The 30 MW size was selected to provide the fastest and most
        economical means of developing a full-scale scrubber.   It was

                                  903

-------
          assumed that full-scale scrubber modules are unlikely
          to exceed 150 MW in size during the next few years.
          Further, it was thought that a successfully operating
          30 MW scrubber could be scaled-up to full  size with
          confidence, and applied to a       number of operating
          boilers.  Alterations to the 30 MW unit during develop-
          ment will require considerably less time and cost than
          for a larger scrubber.
6.        Areas to be investigated in the demonstration unit are those
  common to most limestone slurry systems.   Some of these are  noted
  below:
  a)  Scrubber
      - The optimized gas velocities and L/G ratios derived in the
      pilot module are to be verified.
      - a major consideration is  gas velocity distribution in  the
      scrubber and demister.  Attempts will be made to measure and
      optimize this at critical locations.
  b)  Demister
      Test data are needed in at  least three areas:
      - The effects of velocity maldistribution and demister cleaning on
      the amount of solid and liquid particles escaping the demister.
      - The optimization of moisture removal versus pressure drop.
      - The methods of on-loaid cleaning.
  c)  Reheater
      Initially the unit will have a direct fired reheater. Upon
      successful demonstration of the scrubber/demister, other methods
      of reheating will be tried.
                                  904

-------
       d)  Waste Slurry
           This will be discharged to a divided settling pond where
           the slurry is expected to decant to 80% solids.  The liquor
           will be recycled to the process and the solids disposed
           off-site.
           Items requiring investigation include:
           - The control of dissolved constituents in the recycled liquor.
           - The most economical means of dewataring the slurry on site.
           Alternatives to the settling pond include a clarifyer, or a
           vacuum filter.
           - The environmental problems of the solids disposal.
       e)  Materials
           Corrosion rates will be investigated for various materials in
           contact with the liquid phase.
       f)  Assessment of process availability and reliability.
7.      Regarding design and construction status, preliminary design of
   the 30 MW unit is in progress and the commitment to final design and
   construction is scheduled for next March.  The unit is expected to be
   commissioned and ready for testing by September 1973.
8.      Ladies and Gentlemen, I am well aware that many plans for the removal
   of S02 have come and gone.  However, the plan presented here today is
   based on considerable recent R&D, and is one in which we have great
   confidence.  I trust that at a future Symposium, we can present
   demonstrated results of the successful operation of this 30 MW spray
   tower scrubber.
                                       905

-------

-------
            LA CYGNE STATION AIR QUALITY SYSTEM
                            by




                       D. T. McPHEE




             KANSAS CITY POWER & LIGHT COMPANY




                  KANSAS CITY, MISSOURI
                       presented at




SECOND INTERNATIONAL LIME/LIMESTONE WET SCRUBBER SYMPOSIUM




                  SHERATON-CHARLES HOTEL




                  NEW ORLEANS, LOUISIANA







                   November 8-12, 1971
                           907

-------
                  LA CYGNE STATION AIR QUALITY SYSTEM









     La Cygne steam electric generating station is a 820 raw,  3500 psi




cyclone-fired fossil-fuel unit utilizing a local low grade coal and is




scheduled for commercial service in 1973.  This is a joint venture be-




tween Kansas City Power & Light Company of Kansas City,  Missouri, and




Kansas Gas and Electric Company of Wichita, Kansas.  It  is estimated to




cost between $180 and $190 million and each of the companies  will have




a 50% interest in the plant.  Kansas City Power & Light  Company is the




operating agent and will operate the plant.  The fuel will consist of




some 2 million tons per year of a low grade bituminous coal obtained




from strip mines within a few miles of the plant.  Some  of the charac-




teristics of this coal are 22% ash, 5-1/4% sulfur, 10,000 Btu per pound.




In 1968 when the two companies decided to construct this plant, the




decision was made that an adequate air quality system would be installed




to fulfill our environmental responsibilities.






     This decision has been a bit difficult to carry out in that we have




been faced with a moving target in trying to identify our environmental




responsibility and as originally anticipated, have had to carry out a




crash program to develop a technology for removal of S02 as well as




particulate matter.  I do not want to belabor the moving target part




of the problem, but bear in mind that it certainly has been and still




is substantial.  With a vast amount of emotionalism, and a fair degree




of bureaucratic and political nonsense that invariably creeps into




important new broad base concerns of man, it has been a  bit difficult




to accurately and realistically identify the environmental problem.
                                    908

-------
However, regardless of this and regardless of the fact that a new tech-




nology must be hurriedly ushered onto the electrical generating scene,




Kansas City Power & Light Company and Kansas Gas and Electric Company




keenly feel that the new dimensions of our environmental responsibility




requires that we forge ahead with an adequate air quality system for




La Cygne.






     A substantial development program was carried out preparatory to




the decision on the type of equipment.  Some $300,000, mostly with




Ebasco Services, Inc. and Chemical Construction Company, went into




this effort over an 18-month period.  We also took a look at the sulfur




and sulfuric acid market for the metropolitan area of Kansas City.   The




results of this investigation were not favorable for a recovery type of




system.  Our review of the status of technology also indicated that some




type of lime or limestone non-recovery type of system should be used.




Discussions were carried out with Chemical Construction Company, Babcock




& Wilcox and Combustion Engineering for providing a system that would




remove particulate matter and sulfur dioxide to insure ambient air




quality levels required by the Environmental Protection Agency's primary




and secondary standards.






     Our final decision was to use Babcock & Wilcox equipment consisting




of a venturi scrubber for removal of particulate matter and an absorber




for sulfur dioxide removal.  The schematic arrangement of this system is




as per attached Exhibits C and E.  The first stage of the system is a




variable throat venturi scrubber for particulate removal.  The second




stage is a packed type of absorber using hollow plastic spheres for the




packing material.   Particulate matter is removed in the first stage by






                                    909

-------
means of a water spray.  Sulfur dioxide is removed in the second stage




of the absorber section by precipitating calcium sulfate and calcium




sulfite.  The gas then passes through de-misters and steam reheat coils




where it is reheated 25 degrees.  The gas stream from the seven identical




parallel modules is combined in a plenum from which it is discharged by




six induced draft fans, 7000 hp each, through a 700 ft chimney.






     Some 500,000 tons of limestone is expected to be used per year.




This limestone, of approximately 927, calcium content, will come from




local quarries and will be delivered by the supplier to a limestone




hopper adjacent to the coal receiving hopper.  It will be handled by




the plant coal handling system for the initial part of its route to




the limestone facility.  Two full capacity 110-ton per hour wet ball




mills will be available to grind the limestone to a fineness permitting




907. passage through a 325 mesh screen.






     Since the S02 removal system will produce low pH water and since




the limestone and the fly ash that accumulate in the slurry may have




abrasion characteristics, the material throughout the system has to




have special consideration.  The venturi section will be made of #316




stainless steel and lined with 2" refractory material.  The sump tanks




under the venturi and absorbers will be made of carbon steel with a #316




stainless sheath bonded to the inner surface.  These tanks will also have




some refractory material lining in areas where abrasion may be a problem.




The absorbers will be made of 316 stainless including the wire mesh




baskets that contain the hollow plastic balls.  The de-mister in the top




of the absorber will be made of fiberglass.  The steam reheat coils will




be made of 5/8" diameter stainless steel tubing.  The breeching from the






                                     910

-------
reheater  to the fans and to the stack will be carbon steel.  It is expected




that no corrosion will take place in this area since the reheat of the




gas should prevent the condensation of moisture on the walls of the




breeching.  The pumps and the piping in the recirculation system will




be rubber lined for both corrosion and abrasion protection.






     The total cost of the air quality system is estimated at $32.5 million.




It is expected that 99% efficiency will be achieved for particulate matter




removal and 807» for removal of sulfur dioxide.  The design of the system




is basically complete.  Construction is underway and testing is scheduled




to begin in August and September of 1972.  We are certain that there are




challenging times ahead for making this an effective system.  We are




convinced that this will be accomplished; however, recognize that the




operating and maintenance costs will be substantial which, of course,




will have to be added to the electric rate payers' bill.
                                     911

-------
                                              GAS
                                             OUTLET
                                                                          FROM LIMESTONE
                                                                          SLURRY STORAGE
                  VENTURI
                RECIRCULATION
                   PUMP
  ABSORBER
RECIRCULATION
   PUMP
TO SETTLING POND
                                                                            RECYCLE AND
                                                                           MAKE-UP WATER
                             VENTURI-ABSORBER MODULE
                                               912
                                                                                         Exhibit C

-------
                                                        X
                                                        UJ
                                           s
                                           UJ
                                           fc
                                            a
                                            CO
                                            GO
                                            a
                                            UJ
                                            I
913

-------

-------
SULFUR DIOXIDE SCRUBBER  SERVICE  RECORD
UNION ELECTRIC COMPANY—ME RAMEC  UNIT  2
         J.P. McLaughlin, Jr.
       Union Electric Company
         St. Louis, Missouri
              Prepared for
  Second International Lime/Limestone
         Wet Scrubbing Symposium
          New Orleans, Louisiana
           November 8-12, 1971
                   915

-------
            SULFUR DIOXIDE SCRUBBER SERVICE RECORD

            UNION ELECTRIC COMPANY—MERAMEC UNIT 2



     The following summary shows total days during which

the precipitator was blanked off to direct the boiler flue

gas through the sulfur dioxide scrubber.  The type operation

during each test period is indicated by showing the number

of days the unit was either out of service, firing gas, or

firing coal.

                    No   Gas Firing      Coal Firing      Total
 From       To     Load  50-125 MW   50-55MW 100-110 MW  Days

9-9-68   10-5-68

11-11-68 12-5-68

2-15-69  3-2-69

6-16-69  6-21-69

10-3-69  10-10-69  —

11-24-69 12-22-69

2-16-70  3-25-70*  10 1/2

8-31-70  9-6-70

5-2-71   6-4-71

Total
*Unit was shut down temporarily from March 6 to March 17 but
 not converted to precipitator operation.

9/27/71
                                          J.F. McLaughlin, Jr,
                              916
4 17
5 1/2
8
1/2
— —
2
10 1/2
— —
15 1/4
30 32 3/4
—
11
6
1
3
6
6
2
3
41


1/2
1/2
1/2

1/4
1/2
3/4

4
9
1
3
3
20
20
4
14
79


1/2

1/2

1/4
1/4

1/2
25
25
16
5
7
28
37
6
33
183

1/2





3/4

1/4

-------
                    WILL COUNTY UNIT 1
                  LIMESTONE WET SCRUBBER
                            by
                       D. C. QIFPORD
                COMMONWEALTH EDISON COMPANY
                     CHICAGO, ILLINOIS
                       presented at
SECOND INTERNATIONAL LIME/LIMESTONE WET SCRUBBER SYMPOSIUM
                  SHERATON-CHARLES HOTEL
                  NEW ORLEANS, LOUISIANA
                    NOVEMBER 8-12, 1971
                            917

-------
          Commonwealth Edison in order to gain technical under-
standing and to determine the economics and feasibility of sulfur
dioxide removal, embarked on the installation of two separate and
different facilities for the removal of sulfur dioxide from boiler
flue gas.

          One system is a pilot plant at our State Line Station
that will produce elemental sulfur.  This is a joint research
project with Universal Oil Products of their sulfoxel process.

          The second system is a full size limestone wet scrubber
that will remove particulate and sulfur dioxide at our Will County
Station, Unit 1.  This is the system I will discuss.  I will cover
a progress status report as well as the technical aspects of this
system.

          In January of 1970, the existing electrostatic precipita-
tor was found inadequate to meet the existing particulate emission
standards.

          Accordingly, in the spring of 1970, we contracted with
Bechtel Corporation to investigate the sulfur removal systems
available and to recommend a system that had the greatest chance
of success.

          Bechtel recommended a wet scrubber system using limestone
or lime.  A specification was then prepared by Bechtel and released
for bid.  Of the nine bidders that were solicited, only seven
proposals were received.  After detail study and bid evaluation
with consideration of the project schedule, Babcock and Wilcox
was given authorization to begin the detail engineering in Septem-
ber, 1970.  A formal purchase order was issued in November, 1970,
with a project completion deadline of December 31, 1971-  This
completion date was established by the Illinois Commerce Commission
as part of a recent rate case.

          The Babcock and Wilcox designed process is guaranteed to
remove 98$ of the fly ash and 7-6$ of the sulfur dioxide, but is
anticipated to remove 99$ and 83$ respectively.  These efficiencies
are based on a dus't inlet loading of 1.355 grains per standard
cubic foot at 70 degrees P.  and burning lj.$ Illinois sulfur coal.
In considering scrubbers, the pressure drop across a scrubber
with the same dust removal capability differs greatly between a
cyclone boiler with its smaller dust sizing and a p ilverized fuel
boiler with its larger dust sizing.

          The Will County wet scrubber is being backfitted on a
163 net megawatt Babcock and Wilcox radiant cyclone boiler that
was put in service in 1955*

          The wet scrubber is indicated on this property plat
of Will County Station.
                                 918

-------
          The wet scrubber, like gaul, is divided into three
parts; a limestone milling system, the wet scrubber, and the
sludge disposal area.

          The milling system as shown on slide 2 consists of a
limestone conveyor, two 260 ton capacity limestone storage silos,
two full sized Allis Chalmers wet ball mills, and a slurry storage
tank.  Each silo when full can supply the wet scrubber for 2l|.
hours of operation.  The limestone required should be high in
calcium carbonate, above 97^«  It should be noted that the reactivity
of the limestone is not necessarily related to the chemical analysis
of the limestone.

          Each wet, ball type mill Isdesigned to pulverize 12
tons of limestone per hour so that 95>^ will pass through a 325
mesh screen.  The mill output product is in the form of a water
slurry with 20fo solids.  The slurry is piped to the 1± hour capacity,
62,^00 gallon slurry storage tank where it is pumped to the wet
scrubber system.

          The wet scrubber system Is made up of two identical sys-
tems each taking half the boiler flue gas.  Each system consists
of two recirculation tanks, slurry recirculation pumps, a Venturi
fly ash scrubber, a sump, a sulfur dioxide absorber, flue gas
reheater, and ID booster fan.

          For clarity I have broken the wet scrubber system into
two subsystems, a gas system and a slurry system.

          Slide 3 shows the flue gas path.  Flue gas passes from
the  boiler after the precipitator and goes to the Venturi.  Here
the gas is forced through a pressure spray of water coming from
nozzles on each side of the venturi.

         The gas pressure drop through the venturi xa 9 inches
of water.  The removal of fly ash is effected by the collision of
the particles with small water droplets (the ability to collect fly
ash is a function of water droplet size).

          From the venturi the gas turns through the s unp and
then upwards into the absorber.  Here the sulfur dioxide is removed
as the gas at greatly reduced velocity is forced through two
separate stages of plastic  spheres.  These spheres, coated with
limestone slurry provide a wetting surface for the chemical reaction.
They also act as cleaners to prevent buildup of solids.  The
abosrber outlet has a chevron type demister.  The gas pressure
drop through the absorber is 6 inches of water.  Space for a third
stage of plastic spheres is available if found necessary.

          From the abosrber the flue gas is reheated from 128
degrees F to 200 degrees F to give the gas buoyancy and to limit
condensation in the fans, ducts and existing steel brick lined
stack.  The bare tube reheater is divided into three sections, the
                                 919

-------
first Is made of 30lf stainless steel, the other two sections are
corten steel.  Each reheater has four sootblowers to maintain
tube cleanliness.

          To compensate for the draft loss across the wet scrubber,
all ID booster fan Is used that discharges to the existing boiler
ID fan.  It is intended that the new booster fan maintain a zero
differential across the bypass damper and that the existing ID
fan will continue to control the furnace pressure.

          Slide If shows the slurry recirculation system.  There
are three venturi recirculation pumps and four absorber recircula-
tion pumps connected to their own header.  Normal operation will
be with each venturi and absorber system isolated from each other,
but the flexibility is there to enable online pump maintenance.
The sump is designed so that there is little or no mixing of the
venturi and absorber recirculation flows.  The fresh limestone
slurry at 20$ solids is added to the absorber recirculation tank
where water dilutes the slurry to 8$ solids.  The spent or waste
slurry is taken off the venturi pump discharge line.

          Tank level differences are compensated by an inter-
connection between the two tanks.  Each tank holds lj.2,000 gallons
and provides a reaction hold up time or four minutes in the
absorber recirculation tanks and six minutes in the venturi
recirculation tanks.  Space Is available to add two more tanks
to increase the hold up time to six minutes for each absorber
system if the system performance characteristics dictate.

          The flow of slurry to each venturi is 5800 gallons per
minute and to the absorber is P-750 gallons per minute.  This
gives a liquid to gas ratio of l8.1j. to 1 and 28 to 1 respectively.

          The variable throat venturi and the three sections in
the absorber allows a load range from 30 to 100$ boiler load.

          The waste slurry Is pumped to a settling pond and all
the water runoff is recycled to the we.t scrubber and milling
system.  With our pond arrangement there could be a limited
blowdown but Illinois law allows a dissolved solids limit on
water discharges to the canal of only 750 parts per million.  With.
total recycle the dissolved solids concentration Is expected to be
2500 parts per million.  To compound the difficulty of any blow-
down, the canal water used for makeup has a concentration of 625
parts per million.

          Slide 5 is a sketch of the entire system showing the
milling system, Venturis, absorbers, recirculation tanks, pumps,
re heaters, fans, and duct arrangement.

          The materials used for the construction of the system-are:

            Flues from the boiler to the venturi, carbon steel
            The venturi, cart-on steel with plasite 7122 and two
            inch kaocrete
                                 920

-------
            The sump, corten steel with flake line 103 'and two
            Inch kaocrete

            The sump bottom, lined with firebrick

            The absorber, rubber lined corten steel

            Flues from absorber to re heater, corten steel with
            flakeline 103

            Flues from reheater to the ID booster fan, corten steel

            ID booster fan, corten steel housing, carbon steel wheel

            The pumps, recirculatlon tanks, valves, and all piping
            in contact with slurry above six Inches in diameter is
            rubber lined carbon steel

            Any piping less than six inches in diameter is 316 L
            stainless steel

          The power requirement for the entire wet scrubber system
is nine megawatts or 5.1$ of the unit gross capacity of 177
negawatts.  This is nearly equivalent to the auxiliary power
consumed by the rest of the unit.  The eleven largest power con-
sumers are:

               Two ID booster fans, 22^0 HP each

               Two limestone mills, lj.00 HP each

               Four absorber recirculation pumps, 200 HP each

               Three venturi recirculation pumps, 350 HP each

          The controls for the  wet scrubber system are all located
on a ten foot long central control board that has approximately
as many Instruments as the present boiler board.  This new control
board will allow complete start up, operation, and shut down of
the mill and wet scrubber system remotely.

          Slide 6 shows that the estimated cost for the system is
in excess of $8 million, with equipment about $1|,750,000, erection
about $2,600,000, and professional engineering about $7^0,000.
This amounts to %9 per net kilowatt hour.

          The limestone cost delivered is about $5.00 per ton.
Full load operation requires 1$ tons per hour or 130,000 tons
per year.
                                   921

-------
          With sludge production anticipated at 19 tons per hour
at full load, it will be necessary to dispose of 166,000 tons per
year.  The cost to get rid of the sludge will approach $5»00 per
ton.  This cost per ton would take the sludge from the pond and
convert it from toothpaste consistancy to a solid, stable, non-
reverting material.

          The above costs related in cents per million BTU's are
as follows:

            Carrying charges on $8.1 million for 15 years, 11.5

            Limestone, 5«0

            Sludge disposal, 6.5

            Manpower (one shift position), 1.0

            Auxiliary power, 2.0

for a total of 26 cents per million BTUs.  This total does not
include maintenance or property tax.

          So far I have just talked about the equipment and design.
Construction presents a great many probelms both physically and
schedule wise.  Slide 7 shows how it was necessary to sandwich
the scrubber between the boiler house and service building with a
substantial canteliver.  Also shown on the slide is the complexity
of the duct arrangement.

          Due to foundation problems the equipment erection did not
start until mid May, 1971.  Judicious use of overtime will allow the
erection of equipment to be completed by the end of this December.

          To conclude my presentation, here are some slides that
I took at various stages of construction.
                                  922

-------
  c c
  s s
      ooo
        ,1
        0)3
        •i
923

-------
                                                                      Figure  2
                                .MILLING SYSTEM
eclaim
bpper
                                                                      Slurry
                                                                     Storage
                                                                       Tank
                                                                              I
                                         Recycle
                                         Tank &
                                         Pumps
To Wet
Scrubbe
                                        924

-------
                                                        Figure '•
                      FLUE GAS PATH
Boiler
                   Stack
                 Existing
                 ID Fan
           Byp iss
           Dam >er
 (TF-J
\
     \

Electro).
Precip

                                          ID
                                      Booster
                                                   \
                                 Rdheater
                                   ^
                          Deznister

                                              Absorber
                                             QQOQQOQOQQ
                                         Sump
                        925

-------
                                                                        Pi gun
                            SLURRY RECIBCTJLATIOK SYSTEM
To Sludge
Waste Pond
                                            Absorber
                                        Sump
  Venturi
ecirculatlon
   Tank
                                                         I
   Absorber
Rfecirculation
    Tank
        Venturi Pumps
                                                From ^
                                                 Systt
                                  Absorber Pumps
                                       926

-------
                                             E
                                             o  .t:
                                             c
                                             o   '
                                             M  C
                                            =5  o  T
                                            UJ  ~  U
                                                oo.

                                            ££  <
                                             o •£
                                            O ^
927

-------
                   Estimated Costs
                  Wet Scrubber System
Investment
     Equipment, buildings and foundations
     Erection
     Professional Engineering
 Figure 6
 $4,750,000
  2,600,000
    750,000
 $8,100,000
Operating
     Carrying charges on $8,100,000 for 15 yrs.
     Limestone at $5.00/ton (130,000 tons)
     Sludge disposal at $5.00/ton (166,000 tons)
     One shift position
     Auxiliary power
 11.50/MBTU
  5.0^/MBTU
  6.50/MBTU
  1.00/MBTU
  2.00/MBTU
*26.0
*Note:  This does not include maintenance or property tax.
                                928

-------
929

-------

-------
Chemical Construction Corporation

    Pollution Control Division
A SUMMARY REPORT - CHEMICO'S COMMERCIAL SYSTEMS INSTALLATIONS

               AT ELECTRIC POWER  GENERATING STATIONS
              H. P. WILLETT - Vice President
              I. S. SHAH      - Chief, Process Engineering and Development
                         Pollution Control Division
                     Chemical Construction Corporation
                   320 Park Ave. ,  New York,  N. Y. 10022
                                Presented At
                      2nd International Lime-Limestone
                         Wet Scrubbing Symposium
                           New Orleans,  Louisiana
                           November 8 - 12,  1971
                                     931

-------
unemicai construction Corporation

    Pollution Control Division

       A Summary Report of Chemico's Commercial Sysfrejtru3*4nstallations

                     At  Electric Power Generating Stations

          H. P.  Willett,  Vice President Pollution Control Division
          I. S.  Shah,  Chief,  Process Engineering And Development

Chemico's extensive research and development work.   Bench scale, pilot plant

 cale and prototype scale,  is aimed at developing capabilities to offer suitable

 olutions - both technically and economically feasible - for pollution problems of the

 tility industry.  The major pollution problems of the utility industry  are emissions of

 .y ash,  sulfur dioxide (SC^)  and nitrogen oxides (NOX).  Chemico has  developed and

 3 continuing to develop technology for:

           (a)  Fly Ash Removal

           (b)  Simultaneous removal of Fly ash and SC>2 using  lime -

               •limestone throw away processes

           (c)  SC>2  recovery using Magnesium base SO2  recovery process, to

               produce saleable products.

 i this presentation,  we would  like to summarize the various projects, one in oper-

 •J.on andthe others unaer construction, in the utility industry, and briefly describe the

 nportant  features for each installation.  The various projects are summarized in

 able I.



 ly Ash Removal - Holtwood Station of Pennsylvania Power and  Light  Company

 oiler No. 17 is  a pulverized coal fired balanced draft  boiler having generating capacity

 ' 72 MW.   The coal used is Anthracite, dredged from the river basin, having an ash co

  17  - 20% and a sulfur content of 0. 5-0. 6%.  The boiler  was originally equipped with


                                          932

-------
Chemical Construction Corporation



    Pollution Control Division





Fly Ash Removal - Holtwood Station of Pennsylvania Power and Light Company (Cont'd




mechanical collectors and electrostatic precipitator for dust collection.  The scrub-




ber system is designed to handle 358,000 ACFM of flue gas leaving the air heater at




360°F and -8" WG.  This  represents 80% of the total flue gas.  The balance, 20%




of the gas, flows through the  precipitator and partially blanked mechanical collector




section.  The hot gases leaving the precipitator, mixes with saturated gas leaving the




scrubber, thus providing a reheat of approximately 35-40  F.  The mixed reheated




gas is exhausted to the atmosphere through existing I.D. Fans and stack.
The liquor system consists of a thickener, recycle pump tank, and neutralization




tank (to neutralize thickener underflow with lime).  The underflow after neutraliza-




tion is then sent to an ash pond through existing fly ash disposal pumps.









The scrubber system is designed to reduce the outlet dust loading to 0. 04 grains/SCF




dry when the inlet dust loading is 4.5 grains/SCF day or less.  In case the inlet




dust loading is higher than 4.5 grains/SCF day, the scrubber system will provide




99% efficiency.










"'he scrubber  system  is  successfully  meeting  the guaranteed




performance, even though the ash content of coal has increased from the design value o





17% (4. 5 gr/SCFD) to 38% (10 gr/SCFD).










Fly Ash Removal-Four Corners Station -  Arizona Public Service Company




Jnits No. 1, 2 and 3 are pulverized coal fired balanced draft boilers having generating




                                      933

-------
Chemical Construction Corporation



    Pollution Control Division




 Fly Ash Removal - Four Corner Station - Arizona Public Service Company(Conttd)




 capacity of 175 MW, 175 MW and 225 MW respectively.  The  coal used contains 28. «°,




 ash and 0. 6% sulfur.  The boilers were originally equipped with mechanical collecto




 Each of the three boilers will be equipped with two scrubbers, I. D. Fans,  (wet)




 mist eliminators,  and reheaters.  Each scrubber handles a  flue gas volume of




 407, 000 ACFM at 340°F and -10" W. G. , in  the case of units 1 and 2,  and a flue gas




 volume of 515, 000 ACFM at 340°F and -10" W. G.,  in the case of unit 3.  The flue ga




 from units 1 and 2 are discharged to the atmosphere through one common existing st




 whereas unit 3 has its own existing  stack.   Adjustable throat mechanism are provide




 to maintain constant dust removal efficiency at varying loads.









 A common liquor system, for all three boilers, consists of  thickeners,  pump tanks




 and existing ash pond.  The thickener overflow, and liquor from ash pond are re-




 turned to scrubber system.









 The scrubber  system is designed to provide an outlet dust loading of 0. 4 grains/SCF




 with an inlet dust loading of 12 grains/SCFdry.  'The flue  gas is  reheated by




 20°Ftoavoid    condensation of water vapor in t h e high  velocity  stack.




 The system will be in operation before  the end of this year.









 Fly Ash Removal - Dave Johnson Station, Unit 4,  Pacific  Power and Light Company




 Unit  no  . 4  is a new pulverized coal fired balanced draft boiler having generating




 capacity of 360 MW.  The coal fired has an ash   content  of 16% and  sulfur cor
                                         934

-------
lical Construction Corporation
Pollution Control Division
r Ash Removal - Dave Johnson Station, Unit 4, Pacific, Power and Light Company(Ccn)
%.  The total flue gas of 1, 487,100 ACFM at 270°F and -12" W. G,  and having
t loading of 12 grains /SCF dry, leaving the air heater is handled by three
•ubbers, three wet I. D.  Fans,  and one common low velocity wet stack.  The
urated flue gas is not reheated.        The bleed liquor from the scrubber system
Dumped to an ash plant,  and cleared liquor from the plant is returned to the scrubbers.
'ustable throat mechanisms are provided in each scrubber,  to maintain constant
:ssure drop  to  achieve constant dust  removal efficiency at varying boiler loads.


3  scrubber system is designed to provide outlet dust loadings of 0. 04 grains/SCFdry
•vided the inlet dust  loading is 12. 0 grains/SCFdry or less.   The plant is under con-
uct ion  and scheduled for start-up early in 1972.


2  Recovery - Mystic Station,  No. 6 Unit, Boston Edison Company,  and Essex
smical Company's Acid Facility at Rum ford, Rhode Island
'.t No. 6, rated at 155 MW generating  capacity is equipped with air heaters, elec-
static precipitator (de-energized when burning oil fuel),  two induced  draft fans
 a stack.  At 155 MW rating,  the flue gas volume leaving the I. D.  Fans is
, 000 ACFM at SOOOp and +1" W. G.  The SC-2 loading is 1410 ppm (by volume,
 gas basis).  The flue gas leaving the two I. D. Fans enter  the two new F. D.  Fans,
 one Venturi Type SO2 absorber.  SC>2 is absorbed by MgO  slurry in the absorber,
ning a slurry of MgSOs,  MgSO4 and unreacted MgO.  The bleed from absorber is
t  to a centrifuge, to produce a cake containing approximately 5% surface moisture.
                                      935

-------
hemical Construction Corporation
   Pollution Control Division

SC>2 Recovery - Mystic Station, No. 6 Unit, Boston Edison Company,  and Essex
Chemical Company's Acid Facility at Rumford, Rhode Island  (Cont'd)

The centrifuged cake is then dried in a drier by removing both crystalline and
surface water.  The dry product is  stored in an existing fly ash silo and then trucked

away to Essex Chemicals sulfuric acid  facility at  Rumford,  Rhode Island.  The dry

product containing MgSC^, MgSO4 and  MgO is calcined in a calciner, to produce

SC>2 rich (12 - 16%) flue gas,  and regenerate MgO.  The flue gas after proper cleaning.

enters the sulfuric acid plant to produce 98% H2SO4 acid.


The regenerated MgO is returned by truck to Mystic Station of Boston Edison, for    '

reuse in the absorber,    Make up MgO will be added to the  system at Boston

Edison.


The SO2 recovery process plant will reduce the inlet SO2 concentration of 1410 ppm

'n the boiler flue gas to less than 150 ppm which is equivalent to burning less than

). 3 percent sulfur content fuel oil.


Approximately 50 tons/day of  crystal    MgSO3,  MgSO4 and  MgO will be produced

it the power plant.   The recovered sulfur dioxide from the power plant stack flue gas
vill be equivalent to  the entire feed requirement of the 50 tons/day sulfuric acid  plant

it Rumford, Rhode Island.  The crystals will be shipped by truck only five days a

veek and only during the  8 hour day shift. At the  acid  plant, approximately 20 tons/day

)f MgO will be regenerated.
                                          936

-------
cal Construction Corporation
ollution Control Division

 Recovery  - Mystic Station, No.  6 Unit, Boston Edison Company, and Essex
mical Company's Acid Facility at Rumford, Rhode Island (Cont'd)

 plants at Mystic Station and Essex Chemical are under construction and

eduled for start-up before the end of 1971.
 Ash and SC"2 Removal - Phillips Station of Duquesne .Light Company

.he Phillips  station,  there are 6 boilers having a total generating capacity of

 MW.  Each of the boiler^ is a pulverized coal fired balanced draft unit, and

resently equipped with mechanical collector-precipitator and separate stack.

 coal burnt has 21% ash and 2. 3%  sulfur.


• flue gas leaving the air heater from each boiler enters a manifold. The total

 gas in the  manifold is 2,190, 000 ACFM at 362OF and -22" W. G. ,  and the  dust

ding and SO2 loading are 5.  9 Grains/SCFdry and 1370 ppm.   This total volume

lue gas is handled by four scrubbing trains, each consisting of first stage scrub-

,  wet I. D.  Fan, and a mist eliminator.


ommon reheater and common  stack are provided for  all the scrubbing trains. One

ubbing  train also includes a second stage absorber in place of the mist eliminator.

lis two stage scrubber-Absorber, simultaneous fly ash  and SO. will

amoved using lime  as absorbing agent.  The other trains will remove only fly

using water as  scrubbing liquor.   The bleed from each  scrubber is sent to an

pond,  and if found necessary, additional lime will be used to neutralize the
                                      937

-------
Chemical Construction Corporation



    Pollution Control Division




 Fly Ash and SO2 Removal - Phillips Station of Duquesne Power and- Light Company(O




 total bleed.  The  clear liquor from the ash pond is returned to scrubber,









 To maintain pressure drop across the  scrubber throat, to attain constant




 dust removal efficiency at varying load conditions,  the venturi scrubbers are provide




 with adjustable throats.  As the load decreases, the throat area is reduced by clos-




 ing the throat,  and as the load increases the throat area is increased
 The system is designed to provide an outlet dust loading of 0. 04 grains/SCFdry, whet




 the inlet dust loading does not exceed 5. 9 grains/SCFdry.  The outlet SC>2 concen-




 tration from one scrubber train will be 274 ppm or less.  Design and engineering




 work is in progress,  and the plant is scheduled for start up during February 1973.









 Fly Ash Removal - Elrama Station of Duquesne Light Company	




 At the Elrama station, there are 4 boilers, having a total generating capacity of




 494 MW.   All the boilers are pulverized coal fired balanced draft  units,  and are




 presently equipped with mehcanical collector-precipitator.  The coal is  similar




 to that  used at Phillips Station.









 The flue gas leaving the air-heater from each boiler enters a manifold.  Tne total




 flue gas in the manifold is 2, 211, 000 ACFM at 304°F, and -22" W. G.  The inlet




 dust loading and SO2 concentration are  7. 32 grains/SCFdry and 1570 ppm.  Five
                                           938

-------
   wuiiau ui/uuii
'Dilution Control Division




 Ash Removal - Elrama Station of Duquesne Power and Light Company (Cont'd)




ubbing trains will handle the total gas flow,  and each train consists of a




ubber, wet I. D. Fan, and mist elimi nator.  A common reheater and common




• stack are provided for  all the scrubbing trains.  All scrubbers are provided




i adjustable throats to maintain efficiency at varying loads.   At this plant,




ially only fly ash will be removed using water as scrubbing  liquor.  The bleed




TL scrubbing  system is sent to the ash pond,  and clear liquor from pond is re-




led to scrubber.  The system is designed to provide an outlet duct loading of




74 grains /SCF dry, provided the inlet dust loading is 7.  32 grains /SCFdry or




3.  The plant is  scheduled for start up during February 1973.









kerson Station - Potomac Electric Company




s unit no. 3 of the Dickerson Station is a pulverized coal fired balanced draft




er rated at a generating capacity of 195 MW.  The coal fired has an ash content




3% and a sulfur content of 3. 0%.  The  boiler is presently equipped with mechani-




collector-precipitator system for dust removal.









ty  percent  of the total gas flow will be treated  in the prototype scrubber-absorber




tern, where  the fly ash is removed in the first  stage using water as scrubbing liquor




1 cleaned flue gas then enters the absorber where SO2 is removed using MgO




~ry as absorbing liquor.   The slurry of MgSOs, MgSO4  and MgO is centrifuged,




cake is dried in the drier.    The  dry crystals are then trucked to  an acid faci-




of Essex chemicals, where  upon calcination, MgO is regenerated,  and flue  gas
                                      939

-------
   Pollution Control Division



 Dickerson Station  - Potomac Electric Company (Cont'd)



 rich in SC>2 is produced to make 98% H2SO4 acid.  Approximately 50 tons/day



 of acid will be produced.







 The ductwork is so arranged that the flue gas either before the precipitator or



 after the precipitator can be withdrawn.  The flue gas volume will be 295, 000



 ACFM at 259°F and -11. 0"W. G. , and the dust loading and SO2 concentration are



 5. 95 Gr/SCF dry and 1850 ppm respectively.  The  system will provide an outlet



 dust loading of 0. 03 Gr/SCF dry, and an outlet SO   concentration of 185 ppm or les
                                                 LJ






 A n  adjustable throat mechanism will be provided to maintain constant efficiency



 at varying loads.  Approximately 15°F reheat will be provided using  off gases



from the dryer. At present,  engineering and design is  in progress and the plant is



 scheduled for start up in 1973.







 Fly Ash and SO2 Removal At a Power Generating Station In Japan



 A 155 MW generating  station burning pulverized coal having 20% ash and 3% sulfur,



 in a balanced draft boiler is presently equipped with mechanical collectors-pre-



 cipitator for dust removal.  A flue  gas volume of 451, 000 ACFM at 277OF and 0"



 W. G. , .having a  dust loading of 0. 25 Gr/SCF dry, and SO2 concentration of 2350.



 ppm will be handled in one two stage scrubber train.  The gas volume represents



 75% of the total flue gas.  A 5QOF reheat will be  provided  using fuel oil burners.
                                         940

-------
cal Construction Corporation



'Dilution Control Division





 Ash and SO2 Removal At a Power Generating Station In Japan (Cont'd)




"bide sludge will be used for simultaneous SC>2 and fly ash removal in a two




?e venturi  scrubber system, including a delay tank.  The bleed from the




ubber  system will be pumped to the ash pond, and clear liquor from the ash




d returned to the  scrubber system.










j scrubber system which handles flue gas leaving the existing pracipitator will




vide 90%-fSO2 removal and an outlet dust loading of 0. 025 grains/SCFD. The




it is under construction and is scheduled for start up during March 1972.  This




it is designed for initial disposal consumed  alkali,  but for future conversion to





lufacture gypsum.









imary




>mico designed systems to date include four plants for fly ash removal using




er as scrubbing liquor,  one plant each for simultaneous  removal  of fly ash




 SC>2 using lime slurry and carbide sludge,  and two SO  recovery  plants using MgO




~ry process, one each for coal fired and oil fired  generating stations.  Chemico




igned systems will handle a total flue  gas volume of 10, 090, 100 ACFM resulting




n a total generation capacity of 2293 MW.  Chemico systems designed   to date




. remove 4320 Tons/day of fly ash and 400 Tons/  day of SO2, from flue  gases,




thus help reduce  air pollution and clean the air.
                                       941

-------
O O O   ^
 - >J I—I
CO ft i   r-
steam Adjustab'
in efficiency at
et I D Fans
locity wet stack
djustable throats
00 to 3 01 <
C 4-1 >
*4 C.
3 n w o w
6 T3 — i C
4J (0 10
(0 O O - U,

0) in 00 01 O
1-. 4-1 C .C
TO -. (U h-l
(n O >> 1-
CO 1- M *J
0 SI to 0 4i
 Z. 3





o o
00 >*-i u"i C
C >w ^- O =0
BJ h O
11 X 00 ^
C ti E u
1* T3 (D b-
00 O
00 4J >
TJ C O ^
01 rt -^ tn Q
H 3 u x

r- (0 > U TO
- QJ e u


O CN U -C
am c X
>, — CO 01

O - in ^~. a>


£ §S,5.S
0
C
CO
*J

OJ
oil burners

c



(TJ

-------
PROBLEMS RELATED TO SCALING  IN LIME/LIMESTONE  WET SCRUBBING
                     A.V. Slack,  Chairman
                        Participants:

               A.V. Slack and J.D. Hatfield
               Bela M. Pabuss
               Joan B. Berkowitz
               J.R. Martin
               Philip S» Lowell
                               943

-------
                              SUMMARY
      PROBLEMS REIATED TO SCALING IN LIME/LIMESTONE WET SCRUBBING

      Second International Lime/Limestone Wet Scrubbing Symposium
                         New Orleans, Louisiana
                           November 8-12, 1971
      Participants:  A. V. Slack, Chairman
                     B. M. Fabuss
                     J. Berkowitz
                     A. L. Plumley
                     J. R. Martin
                     P. S. Lowell


                                 SUMMARY


          One of the major problems in removing S02 from waste gases by
lime/limestone slurry scrubbing is scaling in the scrubber and other slurry
handling equipment.  Although study of the problem dates back to work in
England in the 1930's, much remains to be learned about the problem--par-
ticularly in regard to the effect of differing conditions among the units
producing the S02-laden waste gas.

          The mechanisms involved in scrubber scaling are complex, much
more so than for scaling of boilers and desalination equipment.  Among the
design and operating factors to be considered are (l) amount of S02 absorbed
per unit of slurry passed through the scrubber, (2) content of calcium sulfite
and calcium sulfate crystals in the recirculated slurry, (3) degree of de-
supersaturation accomplished outside the scrubber before return of the slurry,
(k) pH levels at various points in the circuit, (5) scrubber design as related
to tendency of solids to settle out of the slurry onto surfaces, (6) scrubber
design as related to the scouring effect of the slurry, (7) degree of oxi-
dation in the scrubber, and (8) point of lime introduction into the circuit.

          Scaling can result from deposition of calcium sulfate, calcium
sulfite, or calcium carbonate--although carbonate scaling is rare.  Calcium
sulfate is the usual scaling species but calcium sulfite is often encountered,
particularly when lime is the absorbent.

          The consensus is that lime gives more scaling than limestone, but
no conclusive reasons for the difference have been advanced.  The generally
lower pH in limestone systems seems to be a factor, at least in regard to
sulfite scaling.  At the present level of development, lime systems must
be operated with blowdown (dilution with water) or at low pH (less than the
stoichiometric amount of lime) to reduce scaling to an acceptable level.
                                    944

-------
Closed-loop operation (no blowdown to watercourses) is feasible with
limestone but careful attention to operating conditions is essential to
avoid scaling.  For adequate S02 removal, the slurry circulation rate
and slurry solids content must be much higher for limestone scrubbing
than for lime; since this should also reduce scaling, it may be that the
particular level of operating variables required for good S02 removal
with limestone is the reason for limestone superiority in regard to
scaling.

          The driving force for scaling is supersaturation.  The formula
developed in the early English work for avoiding scaling was limiting the
degree of supersaturation developed in the scrubber to a level low enough
to avoid crystallization on scrubber surfaces; homogeneous crystallization
on sulfite and sulfate crystals was promoted, however, by carrying a large
surface area of such crystals in the slurry.  The supersaturation developed
in the scrubber was then released in delay tanks before return of the
slurry to the scrubber.

          The limiting upper value in the range of supersaturation that can
be tolerated is that at which bulk nucleation takes place in the solution
even in the presence of seed crystals.  In recent work it has been deter-
mined that this level (for calcium sulfate) is about l.J for the nonstoichio-
metric Ca++/S04~ ratio in the scrubber solution.  (Supersaturation is defined
here as aCa  aS04~ / KspCaS04> where a is activity and Ksp is the solubility
product constant.)  The critical level for calcium sulfite has not yet been
determined.

          Data from recent successful pilot plant operation (nonscaling)
indicates a supersaturation level of 1.19 at the scrubber outlet, well below
the l.J critical level.  The value decreased to 1.02, near the saturation
level, in the delay tanks before return to the scrubber.   Sulfite super-
saturation was much higher--8.15 at the scrubber outlet and 6.^9 returning
to the scrubber--indicating  that the critical level is much higher than
for sulfate.

          The effects of scouring by the slurry or mechanical accumulation
of carbonate or sulfite crystals on surfaces followed by knitting together
or converting to sulfate have not yet been adequately evaluated.  The
composition of the scrubber surface does not seem significant from tests
so far, and little is known regarding the effect of additives or of ionic
strength.

          Further study is needed on all phases of the scaling problem.
In the meantime, the best course appears to be (l) use of limestone rather
than lime, (2) high recirculation rate, (j) high solids content in slurry,
(k) adequate delay time,  and (5) use of scrubbers of the spray or mobile-
bed type.   Since both of these scrubber types have major drawbacks--low
mass transfer rate and excessive packing wear, respectively—it may be better
to use a very open type of fixed packing.  Recent work with a set of wire
screens as packing has given excellent absorption, relatively low wear,  and
no apparent scaling.

                                     945

-------

-------
        REMOVAL OF SULFUR DIOXIDE FROM STACK GASES
           BY SCRUBBING WITH LIMESTONE SLURRY:
        OPERATIONAL ASPECTS OF THE SCALING PROBLEM
                            By
              A. V. Slack and J. D. Hatfield
             Division of Chemical Development
                Tennessee Valley Authority
                  Muscle Shoals, Alabama
               Prepared for Presentation at
Second International Lime/Limestone Wet Scrubbing Symposium
     Sponsored by the Environmental Protection Agency
                  New Orleans, Louisiana
                    November 8-12,  1971
                            947

-------
               REMOVAL OF SULFUR DIOXIDE FROM STACK GASES
                  BY SCRUBBING WITH LIMESTONE SLURRY:
               OPERATIONAL ASPECTS OF THE SCALING PROBLEM
                                   By
                     A. V.  Slack and J.  D.  Hatfield
                    Division of Chemical Development
                       Tennessee Valley Authority
                         Muscle Shoals,  Alabama
                                ABSTRACT


          Scaling in lime-limestone scrubbing for S02 removal is  a very
complicated process; much more is involved than the simple crystallization
of calcium sulfate from solution in scaling of boilers and desalination
equipment.

          The possible effects of the following process variables on scaling
are discussed.

    *Degree of desupersaturation in the surge tank.

    *Amount of S02 absorbed per unit volume of liquor recirculated.

    *Use of limestone rather than lime as the absorbent.

    "Various factors related to scouring, such as slurry rate,  slurry
     velocity, solids content of slurry, particle size of solids,
     scrubber type (e.g., mobile-bed vs fixed packing), and impingement
     angle between slurry and surface.

    *Factors affecting tendency of solids to silt onto surfaces,  in-
     cluding surface roughness and presence of transverse surfaces
     that catch solids.

    *Point of lime introduction into the scrubber circuit.

    'Presence of fly ash in slurry.

    *Degree of oxidation in scrubber.

    *Nature of scrubber surfaces (e.g., plastic vs steel) in regard to
     strength of bond developed between crystal nucleus and surface.
                                    948

-------
     Use of additives that weaken the bond between crystal and
     surface.

          At the present state of the art, th° most effective measures for
avoiding scaling appear to be (l) adequate delay time for desupersaturation
in surge tank,  (2) use of limestone rather than lime, (3) high recirculation
rate (needed anyway with limestone for good S02 removal), (U) elimination
of surfaces that collect solid particles from the slurry, (5) high solids
content in slurry, and (6) use of scrubbers designed to maximize the scouring
action.
                                    949

-------
               REMOVAL OF SULFUR DIOXIDE FROM STACK GASES
                  BY SCRUBBING WITH LIMESTONE SLURRY:
               OPERATIONAL ASPECTS OF THE SCALING PROBLEM
                                   By
                     A.  V.  Slack and J.  D.  Hatfield
                    Division of Chemical Development
                       Tennessee Valley Authority
                         Muscle Shoals,  Alabama
          Because of marketing problems,  the power and smelter industries
are generally turning to throwaway processes—production of a waste solid—
as a means of coping with the S02 emission problem.   Lime or limestone is
the preferred absorbent since only a very low cost material can be con-
sidered when there is no return from sale of product.

          Since the use of limestone as sorbent in a dry system has been
generally unpromising, most of the current effort is centered on absorption
of the S02 by a slurry of lime or limestone.  Although this has shown con-
siderable promise, and is the method that has been selected by most of the
power and smelter companies that are planning to install full-scale S02
removal facilities, there are some major technical problems that remain
unsolved.  The main one of these is scaling, that is,  growth of an adherent
crystalline deposit on scrubber surfaces that eventually causes shutdown
because of interference with gas or liquid flow.

          In this paper, the effect of operating factors, both chemical
and physical, on the scaling problem will be reviewed.  The material
presented is based mainly on small-scale and pilot plant studies carried
out at TVA.
The Basic Problem

          The reaction of S02 with CaO or CaC03 in a scrubbing operation
produces mainly crystalline CaS03-0.5H20.  There is some dissolved sulfite,
however; the equilibrium concentrations of the various species produced
vary with pH, which depends on whether CaO or CaC03 is the absorbent and
on whether a countercurrent or backmixed scrubber is used.  For CaO and
a backmixed scrubber, which gives the highest pH, the principal dissolved
sulfite species are HS03~, S03", and CaS03(aq) with CaS03(aq) preponderant.
With CaC03 and a countercurrent scrubber, which gives the lowest pH, the
HS03~ concentration is higher and the S03~ lower.

          If oxygen is present in the gas, as it is in most situations, there
will be some oxidation of dissolved sulfite during passage of the solution
through the scrubber.  Part of the resulting sulfate remains in solution, the
amount depending on several factors.  The remainder crystallizes as CaS04-2H20,
which in most cases is the species that causes scaling—by crystallizing on
equipment surfaces.

                                      950

-------
           Solubility data for sulfite  and  sulfate  are  given in  Table  I.
 Of more importance  in regard to scaling, however,  is the  tendency  of  both
 calcium sulfate and sulfite  to supersaturate.   Lessing2,  in development
 of the ICI-Howden process,  found that  CaS04-2H20 would supersaturate,  in
 a simulated  solution,  by  about four  times  the  saturation  concentration.
 In recent  TVA pilot plant tests,  the indicated degree  of  supersaturation
 at the scrubber outlet has  averaged  about  1.8  to 1.9 times  saturation
 concentration.   The degree  of calcium  sulfite  supersaturation has  appeared
 to be  even higher;  these  results will  be checked in further tests.
                                  TABLE  I

                 Effect  of  pH on Solubility3  in  the  System

                     CaO-SOg-SOa-HgO  at 50°C  (l22°F)


                           	Parts per million	
                 pH          Ca           S02P
7-0
6.0
5.0
4-5
4.0
3-5
3.0
2-5
6?5
680
731
841
1,120
1,763
3,135
5,873
23
51
302
785
1,873
4,198
9,375
21,999
1,320
1,314
1,260
1,179
1,072
980
918
873
                a Solution  saturated with CaS03'0-5H20
                  and CaS04-2H20.
                  Sulfite.
                c Sulfate.
          The basic operational factors involved in  scaling are  illustrated
by Figure 1.  The flowsheet shown is for a countercurrent  scrubber and  for
CaO or CaC03 introduction into the recirculation tank.  Variations from this
include (l) backmixed or cocurrent scrubbing and (2)  injection of limestone
into the boiler, in which case the resulting CaO enters the scrubber with
the gas.
1 Slack, A. V., Falkenberry, H. L., and Harrington, R. E.  "Sulfur Oxide
  Removal from Waste Gases:  Lime-Limestone Scrubbing Technology."  Paper
  presented at 70th National Meeting, American Institute of Chemical Engineers,
2 Atlantic City, New Jersey, August 29-September 1, 1971.
  Lessing, R.   J.  Soc. Chem. Ind. 5J_, 373-88 (Nov.  1938).

                                       951

-------
                I
                 TO  STACK
GAS FROM
 BOILER
           SCRUBBER
           MAKE UP  H20
      CoO OR
       Cocoa
B
              RECIRCULATION
                 TANK
                                         THICKENER
                                          OR POND
                                                                           OVERFL'
                                                                           TO WATf
                                                                             COURS
                                                      -*. WET SOLIDS
                                   FIGURE  1
                       Flow System Involved  in  Scaling
               Operation of the system will  be  discussed  in terms of the slurry
     composition at points  A,  B,  and  C in the recirculation loop.  At A the slurry
     contains the solid species CaS03-1/2H20, CaS04-2H£0, and CaC03 (or CaO).
     The liquid phase hopefully is  unsaturated  with  sulfite-sulfate species because
     of the water addition  just before A and insufficient time for solid sulfite
     and sulfate to resaturate the  solution.

               In the scrubber,  S02 is absorbed and  forms various dissolved sulfite
     species.  At the inlet pH involved in limestone scrubbing (about 6.0), the
     main species present,  HS03~,  is  in equilibrium  with  a much  smaller amount
     of S03=.  The pH decreases as  the solution flows down through the counter-
     current scrubber,  thereby bringing Ca++ into  solution.  This causes the
     solubility product of  Ca++ and S03- to  be  exceeded (under normal scrubbing
                                        952

-------
 conditions)  with the result  that  a  driving  force  for CaS03-0.5H20 cry-
 stallization is  developed.   Crystallization does  not necessarily occur,
 however,  because the CaSOo/0. 5HpO tends  to  supersaturate.  Even a mild
 tendency  to  supersaturation  may have  a major effect because  the rapid
 passage of the solution through the scrubber leaves little time for
 nucleation and crystal  growth.

           Even if crystallization takes  place  it  can occur on  the surface
 of calcium sulfite crystals  already present in the slurry (homogeneous
 crystallization)  or on  other solids (CaS04-2H20,  CaC03,  fly  ash; hetero-
 geneous crystallization).  If the driving force for crystallization becomes
 high  enough,  however, the  capacity  of these mechanisms  to hold sulfite in
 a  harmless form  will be exceeded  and  crystallization on scrubber surfaces
 will  occur.   This is usually expressed as "critical degree of  supersaturation,"
 that  which must  not be  exceeded if  scaling  is  to  be avoided.

           The decrease  in  pH as the solution flows through a countercurrent
 scrubber  increases the  amount of  total sulfite species  present at saturation.
 Hence the solution can  hold  more  sulfite in the lower part of  the scrubber
 without exceeding the critical degree of supersaturation.

           Calcium sulfate  also tends  to  supersaturate in the scrubber and
 also  has  a critical degree of supersaturation.  However, since its solubility
 decreases with pH (when the  solution  is  saturated with  sulfite) there is no
 advantage from the pH drop in the scrubber.

           At  point B the solution is  supersaturated with both  calcium sulfite
 and sulfate.   In  the recirculation  tank  the pH rise from CaO or CaC03 addition
 reduces sulfite  solubility and promotes  desupersaturation.   The retention
 time  in the  tank  and in the  thickener (or pond) also aids in desupersaturation
 by allowing  time  for crystallization  to  take place.  The objective is to
 have  the  solution at or near saturation  at  point  C.  Although unsaturation
 would obviously be desirable,  such  a  goal seems impracticable except by water
 addition  close to the scrubber inlet.   Unsaturation without water addition
 could be  achieved only  by removing  all the  sulfite and  sulfate crystals and
 then  crystallizing further by usual crystallization techniques; there are
 several process and economic considerations that make this course undesirable.

           Lessing proposed the following equation for rate of  scaling.

                          K  =  (G! - C)/(C -  C2)
                                      pt

                where   C  =  concentration of CaS04'2H20  at time t

                G!  and  C2 =  initial and final  CaS04-2H20 concentrations

                         p =  amount of sulfate  crystals

 The value of K, about 1.5 in the  ICI work,   depends somewhat on the type of
 sulfate crystals;  small, thin  ones have more surface area and therefore are
more  effective in  dissipating  supersaturation  than are blocky ones.

                                    953

-------
 Use  of Diluting Water

           Of the  various process  steps available for reducing scaling,
 addition of  water to the slurry entering  the scrubber is one of the more
 effective.   For example, if  the slurry is diluted by 20%, i.e., addition
 of 20  gal water per 100 gal  of slurry liquid phase, the capacity of the
 resulting liquid  to take up  S02 in  the scrubber is increased by about
 on the basis of solubility alone  (assuming that without the water addition
 the  saturated liquid phase can absorb only that amount resulting from a
 drop in pH from 6-0 to 5,5 in the scrubber).  In addition, the benefit of
 supersaturation is increased in proportion to the volume of water added.

           The main disadvantage of  diluting water is that liquid must be
 removed from the  system (blowdown)  to maintain the liquid volume in the
 system at a  constant level.  Eventually this blowdown must be drained to a
 watercourse,  carrying with it dissolved sulfite and sulfate plus soluble
 constituents  introduced by the limestone  and the boiler gas (Mg, Cl, Na, K) .
 The  water pollution aspects  of this are discussed in another of the papers
 in this symposium ("Potential Water Quality Problems Associated with Lime/
 Limestone Wet Scrubbing for  S02 Removal from Stack Gas" by James S. Morris).

           Since there are some unavoidable losses of water from the system,
 a certain amount  of water can be  added without need to blow down part of the
 scrubbing solution.  A water balance for  typical scrubber conditions is shown
 in Figure 2.   The  toal allowable makeup, 0.7^  ton water per ton of coal
 burned,  is quite  small in comparison with the amount of liquid recirculated.
 The  makeup is equivalent to  a blowdown of only about 1.0%.


 Reduction in pH

           There is an increasing body of  evidence to the effect that reduction
 of pH  in the  scrubber decreases scaling.   No conclusive explanation for the
 effect  appears to have been  advanced.

          When CaO is used,  the reduction in pH can be accomplished by cutting
 back on the  CaO:S02 ratio to less than stoichiometric.   This reduces S02
 absorption,  of course; the question then  is whether the reduction in S02
 absorption required to avoid scaling will make it difficult to meet S02
 emission regulations.   No data appear to  have been reported on the point.
 Dilution with water can be combined with  pH reduction,  of course, to improve
 absorption without incurring scaling.

           The beneficial effect of  low pH may be associated  with the deposition
 of solid  CaC03 that can occur at high pH.  In small-scale continuous tests at
 TVA,  use  of  Ca(OH)2 rather than CaC03 as  the feed material (in countercurrent
 scrubbing) resulted in rapid deposition of CaC03 in the upper part of the
 scrubber  near the slurry inlet.   The deposit also contained CaS04'2H20; it
may  be  that  CaC03 was converted to CaS04-2H20 in place.  In the work reported
by Lessing,  the accumulated  scale contained as much as lQ% CaC03, which also
 indicates  the possibility of CaC05-CaS04 conversion on the scrubber surfaces.

                                    954

-------
STACK  GAS, 11.7
 (0.57 H20)
                          STACK  GAS, 12.1
                          (O.99 H20j 0.42 H20 PICKED
                          UP IN SCRUBBER)
                SCRUBBER
                                    r
    MAKE UP H20, 0.76
 ASSUMING 50% H20 IN
 FINAL SETTLED OR
 FILTERED SOLIDS AND
 RECIRCULATION  RATE
 OF 50 GAL/MCF
CIRCULATION
    TANK
                                 1.86 H20
                                 0.33 SOLIDS
   FILTER
  OR POND
                                                     H20, 69.6
                                                     SOLIDS, 12.3
                                                           .53  H20
                                                0.33 WASTE SOLIDS;
                                                0.34 H20  (0.33 AS H20,
                                                0.0061  AS  CoS03'0.5H20,
                                                0.0043 AS CaSO4'2H20)
                              FIGURE 2
 Water Balance for S0g Removal from

              Quantities are tons per ton of coal burned
           The use of CaC03 instead of CaO,  which also gives a lower pH level
  in the system, has generally reduced scaling.  In the TVA  tests mentioned
  above, there was no deposition when CaC03 slurry was used  instead of Ca(OH)2-
                                 955

-------
          The pronounced effect of pH on sulfite solubility may also be a
factor.  A given reduction in pH in a low pH range (say, from pH 6 to k)
gives more increase in sulfite solubility than a similar reduction at a
higher level (say, from pH 9 to 7)-


Solids Content of Slurry

          Much of the available data on scaling comes from the early work
in England on the ICI-Howden process1?2>3>4.  One of the more important
variables in this work was the content of calcium sulfite and sulfate
crystals in the recirculating slurry; the concentration specified for the
full-scale unit constructed at the Fulham station was 3 to 5$ each of
calcium sulfite and calcium sulfate.

          Although a large number of sulfite-sulfate crystals circulating
in the scrubber loop provides surface on which dissolved sulfite and sul-      '
fate tend  to crystallize preferentially, this alone does not seem to be
sufficient for preventing scaling.   In the ICI work it was necessary to
adjust other factors also to get nonscaling operation.  In TVA pilot plant
tests, scaling occurred even though CaC03 slurry was used and the solids
content of the slurry was about 15$.


Desupersaturation in Surge Tank

          Another variable found important in the ICI work was retention
time of the slurry before return to the scrubber.  It was considered necessary
to provide enough time to dissipate the supersaturation developed in the
scrubber, since it is apparent that any degree of supersaturation at point C
in Figure 1 reduces the capacity of the solution for absorbing S02 in the
scrubber without incurring scaling.

          ICI specified a retention time of about 2.5 minutes in the recir-
culation tank.   The effects of delay time and crystal concentration on
supersaturation, as reported by Lessing, are shown in Figure 3-  Conclusive

* Lessing, R.  J. Soc. Chem. Ind. 57, 373-88 (Nov. 1938).
  Rees, R. L.  J. Inst. Fuel XXV (lt8), 350-57 (March 1953)-
3 Hewson, G.  W., Pearce, S. L., Pollitt, A., and Rees, R. L.  Soc. Chem.
  Ind. (London), Chem. Eng. Group,  Proc. 15, 67-99 (l933)«
4 Pearson, J. L., Nonhebel, G., and Ulander, P. H. N.  J. Inst. Fuel VIII
  (39), 119-156 (February 1935)-
5 However, the species distribution was 5.0/0 CaS03'0.5H20 and 1.5/0 CaS04-2H20
  (remainder ash), and therefore the amount of CaS04-2H20 may not have been
  sufficient.  Moreover, the excess of calcium sulfite may have been harmful;
  Lessing pointed out that the presence of sulfite crystals can cut in half
  the beneficial effect of sulfate crystals on sulfate desupersaturation.
                                  956

-------

9000
6000
2
a 7000
•
z
2 6000
5
8 5°°°
0 4000

-------
however, that the solution returning to the scrubber is still supersaturated
in regard to both sulfate and sulfite; further data will be obtained in an
effort to resolve the question.

          It should be noted that the rise in pH in the hold tank due to the
dissolution of CaO or CaC03 results in a much lower solubility of calcium
sulfite (Table l) and consequently a very high degree of supersaturation
can be developed even if the solution entering the hold tank is only saturated
with calcium sulfite.  For example, if the pH rises from 5.0 to 6.0 in the
hold tank and the entering solution at pH 5.0 is saturated with calcium sul-
fite, the 302 ppm of S02 in the incoming stream is about six times the
saturation amount in the returning stream (at pH 6.0) to the top of the
scrubber.   Unless sufficient delay time is permitted to precipitate the
excess sulfite,  it is not surprising that the returning solution will like-
wise be supersaturated.   The extent to which the solution entering the hold
tank is supersaturated will serve to increase further the supersaturation
in the solution returning to the top of the scrubber.  The opposite effect
is obtained with calcium sulfate; a pH rise from 5 to 6 in the hold tank
increases the sulfate solubility (Table l) by about 5$.  However, since the
resulting decrease in supersaturation is small, delay time is necessary to
dissipate the supersaturation  before return of the solution to the scrubber.


Slurry Circulation Rate

          A further process variable emphasized in the ICI work was amount
of slurry circulated per unit of S02 absorbed.  After optimization of surge
tank retention time and solids content of the slurry, the critical degree
of supersaturation in the scrubber could be determined.  From this the amount
of slurry circulation necessary to avoid exceeding the critical value was
calculated.  For the inlet S02 concentration involved in the ICI tests (about
1000 ppm S02), the circulation rate required was about 130 gal per Mscf of gas.

          This amount of circulation is almost intolerable for the situation
in the eastern part of the United States, where the high sulfur content of
the coal produces an inlet S02 concentration two to three times that in the
British work.  For an inlet S02 content of 3000 ppm, which is not uncommon,
the required slurry rate would be about 400 gal/Mcf to give a liquor:S02
ratio similar to that used by ICI.  Capital and operating costs for such a
pumping load would be extremely high.

          This objection would not hold for the western part of the United
States, where the sulfur content of the coal is quite low, for example, 0.6$.
For O.S'/D sulfur, the slurry rate equivalent to the ICI practice would be
only about 50 gal/Mcf.

          In use of CaO in this country, the general practice has been to
operate with a relatively low L/G (gal/Mcf) because good S02 absorption can
be obtained with CaO at low liquor rate (on the order of 10-30 gal/Mcf).
                                     958

-------
Hence scaling has been promoted by the low liquor:S02 ratio.   For CaC03,
which is much less reactive than CaO, it has been necessary to use higher
L/G (on the order of 40-6o) to get good absorption,  which is favorable to
reduction of scaling.  This is another factor,  in addition to lower
system pH, that makes CaC03 a better absorbent in regard to scaling.

          One possibility would be to use two or more scrubbing stages
with separate recirculation circuits.  Enough delay time could be de-
signed into each circuit to desupersaturate the solution so that the
sulfite-sulfate make in each circuit would not be large enough to cause
precipitation and scaling.   Such a system would be expensive but might
solve the problem.
Erosive Effect of Slurry

          There is some evidence that the erosive or scouring effect of
the slurry may be a very important factor in scaling.   In the TVA pilot
plant work, severe scaling was encountered when stack gas was scrubbed with
CaC03 slurry in a crossflow scrubber.  There was no significant scaling,
however, in spray and mobile-bed scrubbers.   The slurry circulation rate
and slurry solids content were somewhat higher in these tests,  but the main
difference was the intense scouring effect of the high velocity sprays in
the spray scrubber and the bouncing spheres  in the mobile bed--as compared
with the relatively slow flow of slurry through the crossflow.   The thin
spines of CaS04-2H20 formed on the crossflow packing are shown in Figure k.
It would be expected that the erosive action in the other two scrubbers
would break off such spines as fast as they formed.

          Even without solid particles in the liquor,  it would be expected
that high velocity flow of liquid past surfaces would discourage adherence
of nuclei in the formative stage.  This point was emphasized in the ICI work.

          There are numerous factors that may affect the magnitude of the
scouring effect.

     1.   Slurry pumping rate obviously is important, but has generally
         been fixed more by S02 removal requirement than by other
         considerations.

     2.   Slurry velocity.   In a spray scrubber a relatively high liquor
         velocity from the spray nozzles is  necessary for good spray
         distribution.   Velocity probably is lowest in the fixed packing
         type.   Data on effect of liquor velocity are not available.

     3.   Solids content of the slurry should be as high as practicable.
         However,  the 12 to l^% used in the  TVA and ICI work for other
         reasons  (mainly to provide crystal  surface) may be as high as
         should be attempted.   The slurry not only scours crystals away
         from surfaces but also erodes the surfaces themselves.  TVA

                                     959

-------
      FIGURED




_; a tg  Scale^ on_Scrubber Packing






        960

-------
         has under way a test program aimed specifically at the
         erosion problem.  This work may indicate that a lower
         solids content should be used to avoid excessive erosion.

         Particle size of solids will also be varied in the TVA
         tests, since small particles should not be as erosive as
         large ones.  They may also be less effective in removing
         scale.
Silting
          One of the more puzzling aspects of scaling is the effect of
silting, that is, the accumulation of solids on scrubber surfaces by
mechanical means — either by settling onto transverse surfaces or into
crevices or by fine particles being caught bodily on rough surfaces.   If
calcium sulfate crystals are accumulated in this way, dissolved sulfate
should crystallize on them as it does on crystals in the bulk slurry.  The
difference is that nucleation on surfaces is bypassed when crystals accumu-
late on surfaces by mechanical means; crystallization on the accumulated
crystals can cement them together and onto the surfaces, thus providing
an additional mechanism for scaling.

          Calcium carbonate and calcium sulfite crystals can accumulate
on scrubber surfaces in the same way.  Calcium carbonate is not stable
in the system except at high pH.  In reacting with the solution, however,
it may become converted in place to calcium sulfite or sulfate and thus
cause scaling.  Calcium sulfite can also form a relatively stable form of
scale although the scale found in TVA tests has been mainly sulfate.

          No data appear to be available on the effect of silting.  In the
TVA crossflow scrubber work, relatively loose deposits of CaC03 formed in
the packing, presumably by silting.  Calcium sulfate also formed but it is
not clear whether or not the silting contributed to the sulfate scale
formation.   However, the fact that the spray and mobile-bed scrubbers, in
which silting would not be likely, did not scale is a possible indication
that silting is a factor.  Further work on the point appears desirable.


Degree of Oxidation in Scrubber

          It seems logical that the "make" of sulfite and sulfate in the
scrubber, per volume of solution, should have a major effect on the degree
of scaling; as noted earlier, this was the basic consideration in the ICI
work.  The limited data on the point from U.S. work are confusing.  In the
recent TVA tests, only 10 to 20$> oxidation of sulfite occurred in the scrubber
(as indicated by the solid phase composition) yet the scale was mainly sulfate.
                                    961

-------
Work by others has produced sulfite scaling,  however,  even at a degree
of overall oxidation on the order of 50$.   It is not clear what makes the
difference and whether it is better to promote or inhibit oxidation as
far as scaling is concerned.

          It is obvious that much more study is needed in this area,  since
there are ways to change oxidation rate if such a change would be helpful.
Data are needed on the relationship between sulfite and sulfate in regard
to (l) supersaturation driving force needed under various conditions  to
initiate nucleation, (2) activation energy of nuclei formation, (3) strength
of bond in adherence to surfaces, (k) rate of crystal growth after nucleation,
and the effect of increasing amounts of homogeneous surface for growth, and
(5) effect of other dissolved constituents and of ionic strength.
Nature of Surfaces
                                                                              <
          The type of construction material and condition of the surface
may be significant.  ICI adopted wood packing and pointed out that corrosion
roughening of steel surfaces promoted scaling.   Today wood is not favored
as a packing material and steel surfaces likely will be covered with rubber
or plastic to reduce corrosion and erosion.  In the TVA pilot plant tests,
scale grew well on polypropylene packing and in small-scale work scale
formation occurred on glass.

          It does not seem likely that type of construction material will
be a major factor in preventing scaling.  However, it may be important enough
to be a trade-off alternative in determining the most economical solution to
the problem.


Use of Additives

          In the desalination field various additives have been proposed
for reducing or avoiding scaling.  Several mechanisms can be considered,
including nucleation inhibition, weakening the bond between crystal and
surface, formation of films that alter surface properties, and alteration
of crystal habit to change crystal growth pattern.  A major drawback is that
such additives are likely to be expensive and that they will be lost from
the system with the liquor in the discarded wet solids.  However, the cost
might be justified if a major benefit resulted.  Experimental work is indi-
cated since experience in the desalination field will not likely be applicable
because of the wide differences between the two systems.
                                     962

-------
Summary

          Although a considerable amount of experience has accumulated on
scaling in lime-limestone systems, much remains to be learned regarding the
mechanisms involved and the best way to arrive at a reliable and economical
method for eliminating the problem.  The iCI-Howden formula of adequate
retention time in the recirculation tank, high crystal content in the slurry,
and high recirculation rate per unit of S02 absorbed does not appear appli-
cable to high-sulfur coals because of the extremely high pumping load
required; however, it may be usable for low-sulfur coal.   Since the ICI
method was developed from basic chemical considerations,  there is no obvious
way to improve on it from the standpoint of process chemistry except for
using limestone instead of lime, which, for some reason as yet not clearly
identified, reduces scaling considerably.

          There are, hox-jever, some mechanical factors that may be helpful.
With slurry composition and scrubber design aimed at producing an intense
scouring effect, it appears that scaling can be controlled at slurry re-
circulation rates no higher than those required for good S02 absorption.

          The next step should be optimization of the system to give the
most economical combination of conditions.   This will require a considerable
amount of research and development on the quantitative effect of the various
factors and on the basic mechanisms involved.
                                   963

-------

-------
    CALCIUM SULFATE SCALING
           Bela M. Fabuss
  Lowell Technological Institute
         450 Aiken Street
    Lowell, Massachusetts 01854
            Prepared for
Second International Lime/Limestone
       Wet Scrubbing Symposium
        New Orleans, Louisiana
         November 8-12, 1971
                  965

-------
                     CALCIUM SULFATE SCALING

                               by

                         Bela M. Fabuss

       LOWELL TECHNOLOGICAL INSTITUTE RESEARCH FOUNDATION
                        450 Aiken Street
                    Lowell, Massachusetts  01854
          Distillation processes are usually regarded as the most

effective means of producing potable water from saline or brackish

water.  The evaporative desalination is often hampered by the

formation of calcium sulfate scale on the heat transfer surfaces.

This scale may be deposited at low temperatures in the form of

gypsum, at high temperatures as hemihydrate and may undergo trans-

formation to anhydrite.

          Figure 1 shows a summary of the solubility curves for

these three modifications.  It can be seen that both the anhydrite

and the hemihydrate show a strong inverse solubility.  Figure 2

shows pilot and demonstration plant data of the Office of Saline

Water projects plotted on this diagram.  It clearly indicates

that scale-free operation was frequently achieved well above the

arhydrite solubility curve and even in some instances above the

hemihydrate solubility curve.

          Figure 3 shows several sea water heating runs at a

series of heating rates and on two different surfaces, stainless

steel and an epoxy resin.  Figure 4 shows a series of experiments

when the supersaturation was achieved by evaporation at a constant

temperature.   Finally, Figure 5 summarizes these data giving two

sets of precipitation curves superimposed on the calcium sulfate
                                966

-------
solubility diagram.  In summary, this work clearly shows that the



precipitation limits and scale formation depend on the operating



conditions of the unit.  Concentration of sea water by non-



boiling heat transfer permits operation at significantly higher



supersaturations than by boiling heat transfer.  The effect of



other variables such as the heating surface materials, additives,



and heating and evaporation rates was slight.  Scaling was con-



trolled primarily by kinetic factors, determined by the residence



time of the solution in the unit.



          In applying these considerations to lime scrubbing of



stack gases, let us take a look at the equilibrium concentrations



of the ions and molecules in the system.  We are dealing here only



with the dissolved ions.  Figure 6 shows the calculated concentra-



tion of the ions in the solution at equilibrium.



          If we consider that scaling occurs by calcium sulfate



precipitation and not by calcium sulfite or hydroxide deposition,



this can occur only at high sulfate ion concentrations since the



calcium ion concentration is controlled by the solubility equi-



librium of calcium sulfite.  Thus, the scaling should depend on



the rate of oxidation and on the pH of the solution.  The rate



of oxidation strongly depends on the pH of the solution.  Above



pli. 7 and at high suspension concentrations, the suspensions were



practically stable and little or no oxidation occurred.  Even at



pH 3, the oxidation of CaSO., suspensions proceeded slowly after



a significant induction time.



          Based on the presented evidence, we would like to draw



the following tentative conclusions:
                               967

-------
     (1)  Lliniinating CaSO^ ,  Ca(OH),., and CaCO^ as potential



scale formers,  scaling should occur only at high conversions at



low pH values.



     (2)  The scaling is most probably the result of a sequence



of processes:  dissolution of calcium sulfite , oxidation in the



dissolved state, precipitation of calcium sulfate with potential



further conversion to anhydrite scale.



     (3)  The kinetics of each of these processes must be studied



to identify the rate controlling step.



     (4)  There is sufficient evidence from desalination practice



that even if the oxidation to sulfate cannot be prevented, scale-



free operation can be achieved by proper selection of operating



variables affecting the equilibria and kinetics of the process.
                                 968

-------
   o
   o
   c\j
UJ
o:
LU
OL
S
UJ
   O
   CD
   O
   OJ
                                       O
                                       o
                                       ro
                                       O
                                       CD
                                       OJ
CD
          in
                            UJ
                            CC
                            Q
                                       O
                                       <\J
                                       OJ
                                          UJ
                                          o:
                                     (T
                                     UJ
                                     CL
                                     S
                                     LU
                                                CD
                                                M
                                       O
                                       O
                     ro
—   o
                                       O
                                       CO
                 NOIlVdlN30NOO
                      969

-------
IO
   ro        CJ
NO I lVdiN3DNOD
O
                   970

-------
X  X
o  o
0-  CL
LJ  LJ
O  LJ   «
E
0
o
ro
CD
O



c c
E "E
o o
0 0
CO 00
CD CD
D •
1
O
lO
c
"E
o
o
O
O
*



.E "^ h
"E ^v*>.
o ^A/ly" "~ """•--
lO (
*
*
lilt
J LJ
h-
0
00
)

O
o
O O O O o^

-------
     to o> o
     10 ro si-
     — \ _) ID — >
     d .E .£
     _,.  CJ
     51 CM ro
          I
                             CO

                             O
                             O
o
ro
O
OJ
O
                                       J-


                                       w



                                       CD
% 'QBlVlldlOBdd
               972

-------
   O
   O
   C\J
   o
   CO
   o
o
o
LU
a: o
=> ^

<
a:
LU
a. o
2 o
Ld ""
   O
   00
   O
   CO
   O
   O
   OJ
     10
                  Q
                        I
          I
I
ro       <\j       —

  NOIlVdiN30NOO
                                                     O
                                                     CM
                              o
                              O
                              O
                              ro

                                U_
                                o

                              O   -
                              CD LU
                              N EC
               o
               c\J
                              O
                              CO
                              O
                              si-
   H

   o:
   LJ
                                LU
                                                     O
O
(D
                                       to

                                       W
                                       «
                                       D
                                       CD
                             973

-------
K)
°  i
 CM
       CJ
             to
O


 CVJ
                    ro
                                 in

                           CDNODJOOT
                                      974

-------
REVIEW OF SCALING PROBLEMS  IN LIMESTONE BASED WET
              SCRUBBING PROCESSES
             By:  Joan B. Berkowitz
          Arthur D.  Little,  Inc.
        Cambridge, Massachusetts
               November  11, 1971
                                                 Arthur D Little, Inc

-------
           REVIEW OF SCALING PROBLEMS IN LIMESTONE BASED WET
                          SCRUBBING PROCESSES

I.  General Considerations
     Scaling involves the precipitation of insoluble salts from aqueous
solutions onto process equipment surfaces.  Any salt may precipitate if its
solubility limit is exceeded at some point in the processing stream.  From
the point of view of thermodynamics or equilibrium, solubility is a
function of specific concentrations of the precipitating ions, total
ion concentration, local temperature, and pH.  Kinetically precipitation
will not necessarily occur, even if the theoretical solubility limit is
exceeded for a given salt, since some degree of supersaturation is
typical of crystallization phenomena generally, and in many practical
cases a very high degree of supersaturation can be sustained.  Precipita-
tion per se is not scaling.  A precipitate becomes a scale when it
attaches itself to a solid surface, either by nucleation and growth
directly on the surface or by migration of particulates from the bulk
of the solution to the walls.  A precipitate which forms within the body
of a solution can be carried in suspension within the processing stream
and will not result in scale formation unless it is carried to the walls
and tends to adhere there.
     The salts most comonly found as components of scale are:
(anhydrite), CaS0^.2H20 (gypsum)? CaSO^.1/2 H20 (hemihydrate) ;
CaS03.2H20; and Mg(OH)~.  In fact, any wet process in which calcium or
magnesium sulfites or sulfates must be handled is prone to scaling
problems.  For example, saline water distillation processes, wet phosporic
acid manufacturing processes, as well as limestone/dolomite wet scrubbing
processes for removal of sulfur dioxide have all been very much troubled
by the deposition of scale on heat transfer and other surfaces.  Accumula-
tion of scale in pipelines, orifices, and other flow passages results
in plugging of the equipment, often to the point where it becomes
inoperable.  In the wet phosphoric acid process, calcium sulfate formed
by reaction between calcium phosphate ore and sulfuric acid, typically
crystallizes in lines to the extent that periodic shutdown is necessary.
                                 976                            Arthur D Little, Inc

-------
     The so-called scale forming compounds listed above have two signifi-
cant characteristics in common.  First, the salts exhibit inverse
solubility behavior; i.e., they become less soluble as solution tempera-
ture increases.  Second, the salts tend to form relatively stable, super-
saturated solutions.  The inverse solubility as a function of temperature
is probably the major factor responsible for ordinary boiler scale, and
for the scaling of heat transfer surfaces in saline water evaporation
plants.  Scaling under relatively isothermal conditions may often be
ascribed to uncontrolled precipitation from highly supersaturated
solutions unto receptive surfaces of process equipment.
II.  Wet Limestone Scrubbing Processes
     A generalized wet limestone scrubbing process is schematically
depicted in Figure 1 and represents several alternative methods of
operating an SCL scrubbing process using calcium based reactants.  As in
the wet limestone/dolomite injection process developed by Combustion
Engineering and Union Electric, the limestone can be calcined in the boiler
and hydrolized as it is removed from the gas stream in the scrubber.
Alternatively, as in the Howden-ICI process, lime or limestone can be
added outside the scrubber loop, thereby allov;ing greater flexibility of
scrubber pH and scaling control.
     A.  Scale Control in the Howden-ICI Process
     The introduction of alkali outside of the scrubber loop is not in
itself sufficient to prevent scaling.  It does, however, permit the
application of a number of simple scale control techniques.  In one of
the early wet scrubbers, which was set up in Fulham around 1935, a scale
2-3 inches thick was found on scrubber surfaces within 72 hours after
start-up.  The key to eliminating the problem lay in the understanding
and control of supersaturation behavior in calcium sulfate and sulfite
solutions, the primary products or  "make"  of the wet scrubbing process.
It was recognized that calcium sulfite and calcium sulfate solutions
exhibit an apparent stability under conditions of fairly high super-
saturation.  If the scrubbing liquor is never permitted to become
anything more than slightly supersaturated, then sulfite and sulfate
precipitation in the scrubber loop will be very slow.  The slight super-
saturation can then be destroyed by precipitation at preselected sites
                                                                Arthur D Little, Inc

-------
                          o
                           CM
                          I
                                   1
                                  fc
                                  I
                                                                                           m
                                                                                           CO
                                                                                           D
                                                                                           cc
                                                                                           cc
                                                                                           cc
                                                                                           D
                                                                                       LU


                                                                                       O

                                                                                       V)
                                                                                       LU

                                                                                       5

                                                                                       _l

                                                                                       CC

                                                                                       O

                                                                                       LU
to
.O
.0



I
"8
JQ
                       	 I


                      /g I
                       3 «2 3
                                              II
                                              ht
                                              c/5 UJ
                                                                     O)
                                                                    .£ .*

                                                                    2 g
                                                                                           LU
                                                                                           CC
                                                                                           D

                                                                                           C3
                        8
                    u. (D CD
                                             978
                                                                                    Arthur D Little,!

-------
so as not to interfere with normal scrubber operation.
     The ICI group applied their understanding of supersaturation to
effectively overcome scaling in their wet scrubbing operations.  A three-
pronged approach was used.  First, the  "make" of calcium sulfite and
calcium sulfate per pass through the scrubber loop was controlled, by
empirical adjustment of absolute flow rates and L/G ratios, so that the
solutions formed were only slightly supersaturated.  Second, a delay
tank was introduced where supersaturation of the scrubbing liquor could
be dissipated prior to recirculation.  Third, suspended crystallites of
calcium sulfite and sulfate, 3-5% of each, were carried in the circulating
liquor to provide sites for homogeneous nucleation.
     The above three measures taken to control supersaturation in the
ICI wet scrubbing process went a long way towards alleviation of scaling
problems.  Other design factors, however, had to be taken into account
before the problem could be completely eliminated.  The solubility of
calcium sulfite is decreased dramatically as pH is increased, even in
the range 6 to 7.  The pH must therefore be controlled so that the
solubility change does not occur in the scrubbing tower where calcium
sulfite might precipitate onto the packing.  In the ICI process, the lime
or limestone slurry is added just before the delay tank where the
increase in pH assists in dissipating supersaturation.  The rate of
addition of alkali is adjusted so that the pH at the bottom of the
scrubber is maintained at about 6.2.  The total pH change through the
scrubber is therefore from about 6.8 at the top to 6.2 at the bottom.
     It has been implicit in the discussions so far that solutions and
slurries are homogeneous in composition.  It is naturally of prime
importance that such uniformity in composition be maintained, at least
to the extent that supersaturations are not exceeded in localized areas
within the scrubbing tower.  In the ICI work deep plates inserted in the
lowest section of the scrubber tower provided for even gas distribution.
The plates were also in a region of high sulfur loadings and the main-
tenance of high liquor velocities as well contributed to elimination of
scaling in the system.  It is interesting to note that in spite of the
substantial progress made by ICI in the 1930's towards prevention of
scaling by proper adjustment of design parameters, ICI was still troubled
by occasional scaling problems when wet scrubbing operations were resumed
                                    979                          Arthur D Little, Inc

-------
in the 1950' s.  The problems were usually the result of incomplete
irrigation of the scrubbing tower grids due to accidental blockage of
flow elsewhere in the system.
     Finally some materials and surface finishes are more resistant to
nucleation, growth, and adherence of scale than others.  In the ICI work,
it was found that scaling could be minimized by constructing scrubber
grids of smooth, planed red deal wood.  The critical factors are probably
corrosion resistance and surface smoothness.
     B.  Scaling in Limestone Injection Wet Scrubbing Processes
     Although some of the ICI work involved the use of lime slurries,
the bulk of the effort by far was concentrated on limestone additions.
It is generally believed that lime is more efficient than limestone for
removal of SCL.  However, the use of externally calcined lime adds
substantially to the cost of scrubber operations.  By introduction of
limestone directly into the boiler, calcination of limestone may be
accomplished very inexpensively.  Unfortunately simultaneous introduction
of lime and flue gas into the scrubber circuit has introduced scaling
problems which are yet to be brought under control .
     The key mechanism responsible for S09 absorption in limestone
injection wet scrubbing process is believed to be the reaction of SCL(g)
in the flue gas with a circulating saturated slurry of calcium sulfite,
resulting in the formation of calcium bisulfite in solution:
           CaS03 (sat. soln.) + S02(g) + HZ) (]_) -> Ca(HSC>3)2 (soln-)
A sudden increase in pH in local areas, where hot lime particles from the
boiler first contact the scrubber liquor, can force calcium sulfite out
of solution with resultant plugging problems.  The lack of pH control at
the bottom of the scrubber may be a major factor contributing to scaling
in limestone injection systems.  Solution to the problem is not easy and
might require major design changes.
     In pilot plant experience with limestone injection wet scrubbing
processes, scaling has been more the rule than the exception.  In one
installation which has been described in the literature, calcium sulfate
deposited on overflow drain screens in the scrubber and drastically
restricted water flow.  Scaling and plugging were also encountered in
the marble bed scrubber and in the reheater.  The most serious scale

                                     980
                                                                Arthur D Little, Inc

-------
problem was encountered at the scrubber inlet where temperature is at
a maximum.  It might be anticipated that this would be a critical position,
due to the inverse temperature dependence of calcium sulfate solubility.
Scaling at the scrubber inlet may be controllable through the use of
saturated sprays to pre-cool the flue gas and to avoid the formation of
a sharply defined wet/dry interface.  This approach has been suggested
by Bechtel, and is expected to be tested in the pilot plant later this
year.
     While the specific scale prevention methods devised by ICI for
limestone slurry scrubbing are not all directly applicable to boiler
calcined limestone injection scrubbing, the insights into the factors
responsible for scale formation can provide guidance to the development
of appropriate control techniques.  The factors of primary importance
are pH, both local and global; gas and liquor distributions; liquor
velocities at scrubber surfaces; "make" of calcium sulfite and sulfate;
scrubbing liquor composition; and materials of construction.  Under EPA
sponsorship, we are currently building a laboratory bench scale scrubber
to explore the effect of these factors on scaling behavior.
III.  Scale Composition
     The principal components of the scale formed in the ICI limestone
slurry process, before the scale control methods were optimized, were,
gypsum  (CaS04.2H20), 60-90%; CaS03.l/2H20, 1-40%; and calcite (CaC03),
1-5%.  The ratio of sulfate to sulfite in the scale seems to depend on
the degree of oxidation of sulfite in the scrubber circuit.  This in
turn seems to be highly sensitive to catalysis by trace quantities of
transition metals.
     Very little information seems to have been published about the
composition of scales formed in the boiler calcined limestone injection
processes.  The absorption process is generally described in terms of
sulfite-bisulfite reaction, but in the presence of fly-ash, oxidation
of sulfite to sulfate is expected to be quite rapid.  Since control of
scale depends to some extent on composition and supersaturation behavior,
the extent of sulfate formation could be an important and possible crucial
process parameter.  When a limestone slurry is used as a reactant, it is
hardly surprizing that CaCO., might be a component of the scale.  When
limestone is calcined in the boiler prior to introduction to the scrubber,

                                   981                           Arthur D Little, Inc

-------
the calcining process is complete and virtually no CaCO.,(s) is  carried
in the flue gas mixture.  Any CaCO_ found in the scale would  thus have
to be due to the reaction of CO- with scrubber liquor components.  The
role of C0~ in the injection scrubbing process is very much in  need  of
clarification.  The critical step in the ICI process is  supposed to  be
the reaction of CaCO- with CO- in the flue gas to form the bicarbonate
which subsequently reacts rapidly with SO- to form sulfite.   In the
injection process, the absorption mechanism seems to involve  primarily
sulfite/bisulfite rather than carbonate/bicarbonate.  If there  is a  real
difference in mechanism, a different approach to control may  be required.
                                     982                         Arthur D Little, Inc

-------
DEPOSITION PROBLEMS AND SOLUTIONS  IN  THE COMBUSTION
 ENGINEERING LIME/LIMESTONE WET  SCRUBBING SYSTEMS
                 J.R. Martin
        Combustion Engineering,  Inc.
               Prepared for
   Second International Lime/Limestone
          Wet Scrubbing Symposium
           New Orleans, Louisiana
            November 8-12, 1971
                    983

-------
              Lime/Limestone Wet Scrubbing Symposium
                   Thursday, November 11, 1971
                   Session on Scaling Problems

                           J. R, Martin
                   Combustion Engineering, Inc.
     It would appear that Combustion Engineering has been asked to

participate in this session on scaling problems in lime/limestone

vet scrubbing because of our vast experience in producing scale.

The deposition problems which we have experienced in the C-E - APCS

fall into two general categories.  The first area includes all those

deposits which are mechanical in nature (i.e., deposition of solids

due to drop out at low gas velocities).  The second area is limited

to scale formation as a result of chemical reaction.  My initial

remarks will be related to the first type of deposition; mechanical.



Mechanical Deposition

     The system schematic shown in Figure I of the C-E - APCS

(limestone-furnace injection) at Kansas Power and Light Corapary,

Unit #k will serve as the reference for this discussion.  The C-E

APCS at Union Electric, Mersmec Station is similar except for

employing a clarifier rather than a pond; therefore, the statements

herein are considered to be applicable to both systems.



     The first deposition probiera that was encountered was the

mechanical plugging of the scrubber inlet.  The inlet became ^0-

50 percent plugged within 2-H hours of  operation of the  system.

Figures II and III show the scrubber inlet plugged and  clean.

The clean inlet is as a resuK of insta] "1 ing  a sootblower which

prevents excessive build-up of d<-porits.  This de-position is
                                  984

-------
caused bv "the wet-dry interface at the scrubber inlet.   The inlet •




deposit is typically composed of flyash, calcium sulfate, and cal-




cium oxide bound together in a cemcntitious mixture.   The calcium




sulfate which has been found in the inlet deposit does  not exhibit




the crystalline type properties of calcium sulfate scale.  It is




our conclusion that the calcium sulfate found in the inlet deposit




is a result of the removal of sulfur trioxide in the boiler by the




injected limestone; this is verified by the fact that the composi-




tion of the inlet deposits are quite similar to that of the dust




entering the APCS.









     Another area of mechanical deposition is the area under the




marble bed.  This area includes the underbed spray system, the




marble bed structural supports, the gas straightening vanes, and




the scrubber walls.  Most of this deposition is also a result of




wet-dry interfaces.  The flue gas entering the scrubber is 300-




350°F and is cooled down to its saturation temperature  (llO-128°F)




in the area under the marble bed.  During this cooling, some of the




hot dry flue gas impinges on partially wetted surfaces  and deposi-




tion of the dust being carried by the flue gas can result.









     The reheater and demister (shown in Figure l) are  two components




of the APCS with which we have experienced deposition and scaling




problems.  The deposition of mud (c?iemically uncombined solids) on




the demister occurs in normal operation of the system to a minor




extent.  This build-up of solids IL cleaned by the utilization cf a




demister wash system, but vheu the marble bed is not operating





                                985

-------
correctly, excessive "build-up of solids can occur which the  vash




system cannot cope vith.  Additionally, calcium sulfate scale is




formed in the demister when the scrubber becomes supersaturated




vith calcium sulfate.  The liquid vhich the demister is separating




has the highest concentration of calcium sulfate in the scrubber




and therefore, the greatest tendency to scale is at this point.




The problem of supersaturation of calcium sulfate and the resulting




scaling will be discussed in subsequent remarks.  Also the APCS




reheater builds up calcium sulfate scale when the demister is not




operating properly.  The excess liquid impinging on the reheater




is evaporated leaving anhydrous calcium sulfate.  This deposit is




very hard even withstanding sand blasting.









Chemical Scale






     Basically, we have formed three types of chemical scale in the




C-E - APCS limestone scrubbing process:  calcium carbonate,  calcium




sulfate, and calcium sulfite.  These three types of scale have been




found in many locations throughout the APCS.  Ky remarks on  this




subject will be limited to where we have formed these different




scales, what we think the mechanism or reaction is that causes




them to form, and how we have eliminated or minimized their  formation.









     The marble bed is where we have experienced our most severe




scaling problems.  Scaling of both calcium sulfate and calcium




sulfite has occurred on the overflow pots, recycle piping, and an




the marble bed.
                                986

-------
Calcium Sulfate




     Figure IV is the marble "bed at Meramec Station,  Union Electric


Company APCS, after about 2\ hours of operation in the  fall of 1968.


The deposition on the overflov pots is a mixture of calcium sulfate


scale and flyash.  The flyash appears to get trapped in the deposit


as the calcium sulfate is scaling.  The deposit has a definite


crystalline shape and reflects light similar to chips of glass.


The calcium sulfate scale also was found on the scrubber walls.





     This problem of calcium sulfate scale was not encountered in


our earlier pilot plant work and was therefore unexpected.  Ini-


tially, we thought the scaling might be due to the retrograde


solubility that calcium sulfate exhibits (i.e., liquid  tempera-


ture in the scrubber is higher than the liquid temperature in  the


clarifier).  This theory was weakened when it was determined that


the form of calcium sulfate we were scaling was gypsum and not
                                  *

anhydrous.  There is no  change in gypsum's solubility in the


temperature range that the C-E - APCS operates.





     Next, we switched to the use of a dolomitic limestone and the


higher magnesium seemed to depress the overall calcium in solution


to a point where it did not scale calcium sulfate.   In  the late


fall of 1969} the calcium sulfate scaling problem occurred again.


This occurrence was probably due to longer sustained periods of


operation of the APCS.   It was at this tine that we added some


additional dilution water to the system in order to maintain the




                                 987

-------
calcium sulfate concentration below the scaling level.   The rate




of  dilution water we determined that was required to run Meramec




was equivalent to about two gallons per thousand acfm.   We are




not suggesting that this is the best way to prevent sulfate




scaling, but at that time it alleviated the problem and enabled




further operation of the system.









     To date, we have never seen sulfate scaling in the scrubber




at Kansas Power and Light, Lawrence No. k.   Tests this year in




our laboratory suggest that the. pond in Kansas (we go directly




from the scrubber to the pond) acts as a desaturating vessel,  in




other words, the scrubber effluent which is supersaturated in




calcium sulfate solution going to the pond will "be reduced to




saturation given sufficient time.  In fact, the calcium sulfate




concentration in the spray water from the pond has never gone




higher than 1200 ppm.  Whether this is due to the action of the




pond or that we have not operated *for a long enough period of




time to completely saturate the pond is not known.









Calcium Sulfite







     The sulfite scaling problem we encountered occurred when  we




tried to recycle the solids from the bottom of t?.e scrubber to




above the bed.  Figure V shows a recycle nozzle at Kansas Power




and Light; the spray pattern of this no7.£le is a hollow cone;  it




looks like an inverted umbrella.  All the deposit on the bottom




of the nozzle is pure calcium Fulfite.  We have explained the
                                  988

-------
formation of this type of scale as follows:  there is a localized




area highly concentrated in calcium hydroxide where the recycle




slurry enters the bed.  This highly alkaline slurry raises the pH




of the bed liquid causing a shift in the bisulfite-sulfite equi-




librium.  Scaling of calcium sulfite results because of its rela-




tively low solubility.  Later, when we installed a system where




part of the pot effluent water was pumped to the recycle tank and




controlled the recycle slurry pH, ve found that we could operate




the recycle system without sulfite scale.









     Figure VI shows another view of the same marble bed.  The




diagonal white line down the middle of the figure is where the




division plate is located under the marble bed; it is a loca-




lized area of low gas velocity which readily plugs during normal




operation.  The other light areas are the type of deposit which




results when the recycle slurry pK is not controlled.









     Figure VII shows the marble bed at K. P. & L., Lawrence Ho. h




after we revised our recycle system and rearranged the overflow




pots.  ¥e have been able to run for a two-and-one-half week period




last June with this revised above-bed recycle system without any




sulfite scale formation in the scrubber system.









Calcium Carbonate





     The other type of scaling problem we hr-ivc exr/ffiencc-d is




calcium carbonate scaling.  Tn January of 1971 ">•'•:> incurred
                                  989

-------
severe scaling of calcium carbonate at Kansas Power and Light




Company.  Vie were trying to run some tests on the system to




determine the effect of recycle at the time.   One particular




series of tests required running the system without recycle.




During this test very high limestone feedrates to the furnace




were required to obtain reasonable sulfur dioxide removal.




Since there was no recycle of the scrubber reject slurry, the




scrubber effluent slurry to the pond rose to a pH of 10.




Gradually, the pond pH started rising and when it rose to 10,




we developed a serious problem of calcium carbonate scaling




in the spray nozzles of both scrubbers.









     The scrubber effluent comes in at the left of the pond




(Figure VIII is a picture of the pond) and then the liquid




travels around the pond to the spray water pumps.  The pond




is simultaneously used for the plant's cooling tower blowdown,




their bottom ash blowdown, and their pyrites blowdown.  At the




time that, the CaCOo scaling problem occurred, the cooling tower




blowdown was coming into the pond at the A^CS spray water pump




suction.  The cooling tower blowdown had been pumped into this




area of the pond since the system was started up in 1968 with-




out any problems.









     The cooling tower blcwdown water has about ^00 Dprn of




bicarbonate.  Further, the auentit'r of cooling tower blowdown




water entering the por;d i p, roughly equivalent to 1500 gpm whereas:




the spray water to the APC3 is 3oOO gpi.i.  It has been theorized







                                  990

-------
that the pH rise which occurred in the pond last  January  caused a




shift in the bicarbonate-carbonate equilibrium and since  calcium




carbonate is a very insoluble compound it  became  saturated  and




subsequently scaled in the spray piping system as it  left the




pond.  Naturally, the first place where the problem would become




serious would be in the orifice of the spray nozzle.









     After analyzing the problem, we moved  the location where the




cooling tower blowdown enters the pond to  the same place  where




the scrubber effluent slurry enters the pond.   In this way  the pH




rise which may or may not occur depending  on the  APCS mode  of




operation will take place on the inlet side of the pond,  thereby




allowing the bicarbonate-carbonate shift and the  resulting  pre-




cipitation of the calcium carbonate to have sufficient time to




take place in the pond rather than the scrubber.   Subsequent




operation of the system has not produced any calcium  carbonate




scale.
                                   991

-------
&a-9

-------
H



 
-------
\"

-------
                                                      H

                                                      0)
                                                      ^1
                                                      3
                                                      d
                                                      •rH
995

-------
                                                 >

                                                 Q)
                                                 tr
                                                 •H
996

-------
                                                        H

                                                        >

                                                        Q)
                                                        tn
                                                        •r-l
997

-------
998

-------
                                                       H
                                                       H
                                                       H
                                                       tJ>
                                                       •H
999

-------

-------
    adian  Corporation
  8500 SHOAL CREEK BLVD. • P. O. BOX 9948 • AUSTIN, TEXAS 78757 • TELEPHONE 512/454-9535
USE OF CHEMICAL ANALYSIS AND SOLUTION
  EQUILIBRIA IN PREDICTING CALCIUM
  SULFATE/SULFITE  SCALING POTENTIAL

                by:
         Philip S. Lowell
           Presented at:

 SECOND INTERNATIONAL LIME/LIMESTONE
      WET SCRUBBING SYMPOSIUM
        8-12 November 1971
      Sheraton-Charles Hotel
      New Orleans,  Louisiana
          Sponsored by:

  Environmental Protection Agency
      Office of Air Programs
    Division of Control  Systems
              1001

-------
        \sOrpOr3ilOn  ssoo SHOAL CREEK BLVD •  p. o 60x9943 • AUSTIN. TEXAS 73757 • TELEPHONES^ -154 9535
1.0       INTRODUCTION

          This paper presents experimental  data  from a TVA
pilot plant.  These data are consistent with  a proposed
quantitative measure of scaling potential.  Both calcium
sulfate and calcium sulfite precipitation could  be  antici-
pated in the system studied.  The possibility of scale
formation in the scrubber or on vessel walls  existed for
both sulfate and sulfite, while sulfite scaling  could also
take place on limestone feed crystals, a phenomenon known
as "blinding".

          The principle object of understanding  scaling phenomena
is to be able to design and operate a process that  does not
scale.  This requires that three types of information be known:

             A description of the individual  major
             pieces of equipment used in the  process
             including kinetic data, equilibrium con-
             ditions, and mass and heat balances.

             A description of the entire system.

             The ability to predict in quantitative,
             measurable terms the scaling potential
             in all the different parts of  the system.

          The equipment descriptions for particular types of
scrubbers have been discussed in Papers la  and Ib and the first
two requirements have been treated in general by Dr. Ot;tmers in
Paper Ic.
                              1002

-------
 Radian Corporation
SHOAL CRFLK BLVD • P O BOX 7918 • AUSTIN, TEXAS 78757 • TELEPHON C 512 454 "'533
          The third requirement for design of a  scale-free
process, a quantitative measure of scaling potential will be
discussed in more detail in Section 2.0.  The proposed method,
which was first presented by Mr. J. L. Phillips  in Paper Id
involves some function of the relative supersaturation, i.e.,
the quotient of activity product and solubility  product
constant.

          The experimental results presented here were obtained
at the Tennessee Valley Authority's Colbert Steam Plant.  A
pilot plant is being operated there in support of a full-scale
plant to be designed for the Widow's Creek facility.  The
objectives of the Colbert pilot operation are to obtain design
and operating data and to find out whether the scrubber and
ancillary vessels can be operated without scale  formation.
The results during an extended period of operation at relatively
constant conditions showed that no scaling occurred.  Radian
collected data in the middle of this period.

2.0       A QUANTITATIVE MEASURE OF SCALING POTENTIAL

          In this section, the basis for the proposed scaling
potential function will be discussed and some basic laboratory
results will be referred to.  Although laboratory studies have
been conducted only for calcium sulfate dihydrate precipitation,
it is expected that the same kind of behavior will be shown for
calcium sulfite in future work.  While anhydrous calcium sulfate
is the thermodynamically stable form above 40°C, the formation
kinetics are much too slow for it to be of significance in any
system.
                               1003

-------
 Pad \r. PS
                                             Tf
-------
                    ^oo SHOAL CPLCK BLVD •  P o Box??-i8 • AUSTIN TEXAS 73757 • TELEPHONIST 1549535
                              FIGURE  2-1
RELATIONSHIP BETWEEN ACTIVITY COEFFICIENT AND IONIC STRENGTH
                                  1005

-------
 Radian Corporation
•ibOO SHOAL CREEK 81.VD • P O. BOX 9948 •  AUSTIN, TEXAS 78757 • TELEPHON: 512 454-9535
Subsaturation:         a  ++acn=   <  K                    (2-4)
                        v_>a   oU^       Spp cr\
(dissolution ten-                        <-asu4
  dency)

Supersaturation:       ^a^SOr   >  Kspc SQ              ^^
(precipitation ten-                          4
  dency)

          Since the magnitude  of  K   varies widely for different
compounds, it is convenient  to "normalize" the means of express-
ing the degree of saturation by using the quotient a1aa/K
                                                          sp
Then the relationships in 2-6,  2-7 and 2-8 are  true for any com-
pound formed from cation 1 and anion 2.

Equilibrium:           ^a-  = 1                         (2-6)
                         sp

Subsaturation:         a, a,,     ,                          /o -7\
                       ~v—  < 1                         (2-7)
(dissolution ten-        sp
  dency)

Supersaturation:       'a, a,     ,                          ,9 Q\
                        T,    > 1                         ^Z -O )
                         sp

The quotient a, a5/K    has been termed the relative Supersaturation,
                   sp
r»  It is proposed that some function of the relative supersatura-
tion is a valid and useful quantitative  description of the tendency
towards scale formation.

          The phenomena of scaling and nucleation are somewhat
related.  Precipitation on an  existing crystal requires only
that the relative Supersaturation be greater than one.  Very
small crystals are more soluble than large crystals.  There is,
therefore, some value  of r greater  than  one that  must be attained
 in the solution before a nuclei  can be  formed.

                                1006

-------
 Radian Corporation
CC^k, SL'.D •  ~ O 3CX '743 • AUSTIN TIXAS 73757 • TELEPHONE 512 4549535
          Scaling is somewhat similar to nucleation in that a  new
species is being formed.  If the electrostatic forces and  lattice
spacing of the hoat surface is very close to that of the scaling
species a value 01 r close to one might be sufficient to cause
scale nuclei.  If the host surface is very dissimilar the  value
of r will probably be near that required for bulk nucleation.
While this is perhaps a rather folksey description of a complex
phenomenon, it does appear to have merit in correlating the
results.  The important point being made is that value of  r
required for nucleation should have some relationship to scaling.

          Results obtained by J. ~L. Phillips in the Radian
laboratory (Paper Id) show clearly that for calcium sulfate
there is indeed a critical value of relative supersaturation
beyond which nucleation and a greatly increased rate of preci-
pitation takes place.  The results are illustrated in Figure
2-2 which shows that beyond a value of ar -j_facn=/K   of 1.3 to
                                        U3   ^^4  Sp
1.4 the rate of calcium sulfate dihydrate precipitation markedly
increases.  Photographs of the crystals formed at values of the
ratio greater than this critical value clearly indicate the
formation of small new crystals where nucleation has occurred.
Again, the data for calcium sulfite precipitation have not yet
been obtained.  However, the same behavior is anticipated  for
the sulfite.

3.0       EXPERIMENTAL CONFIRMATION

          The equipment and flow arrangements for the pilot
plant scaling studies are given in Figure 3-1.  Again it should
be pointed out that scale formation was not a problem in the
system shown.  This can be explained by examining the values of
relative supersaturation for calcium sulfate and calcium sulfite
throughout the system.   For similicity, let rl be the relative
                               1007

-------
I <•

I-'

( a
 1 6
  u
i Or—

.9
\
.8
.7
.6
5
.4
.3
.2
.1
1
Si
i
, — -f
45
Effi
-h-f-
ir£
r-
j -
_i
13
rrrr
.:;i
>

1 r i '
t-
i-
•t
Ij
1
t
t'
~G
*
4
I
S
4
'}
<
4
•r
1-
C
r"
i
i
•H
-f-r
1
t,
-1
U
J
^
4
1)
g
•H
i-l
•H
S
1
QJ
4J
1
4-J
4-1
O,
0
0)
PH
lit
t • :
%
0 ~
;!j!
J i ,1...
'RE
Jon
4'™
rtiT
"tTp"
-rH-r
-T — L-
-H 	
7-1 1:
rtt;

M -i
::: 1 ."
:.: 1
TT1
I
CII
stc
C
IF-
~T | .
— L 	
: — -~

::'
'IG
'IT
a
)ic
).5
TTTT
....
	
,,r
.mz-
— ~
> ,'~
••
URI
AT]
t 4
hie
% E
z^rv
:~
" r i
. \"-.



• t
: 2
:ON
1-5°
>me
>ee
T — :
-- t
. 	 J
	 ]
-'•Er
—
. 	
:.:v)
•'==1
=~
:_r
-~
_..
N



-2
01
C
tri
d S
	
x
• 1 1
. ,
' C
.c
>lu
-I--— -
-* ,-L-l
•-rr.|
...
:.^.|
-
--U-i
aSC
So]
rr>
TUT
_ 	
.._
_
--—
r^ 	
. ...
-•-
—
I;-.-'
• _:
I.:!.
:;::|f,|--:.
:r,
)4-
-Ut
r
TTT
.
— -T»
. 	 ,-
	 ;
"-I ^
:•_:!
2Ha
ior
Z^t"r-^-
,-, — .
p 	
— -'
— -
-
-O
r:-
^/
/-'
r *
0 g
i
i- •-
- 1- -
* — -- f-t-— —
*"- r "~
	 r 	
" " "*" r r ' " ' '
i
-. ~.rl _ HT-
	 1 	
71^71:-
-V--
^ . , i _, 	
:->_^
. : 	 J — ,/
ZT r ' ' tl
• "\~S
--- - /-
1 /- —
/ - -

. ~~'.
;.::
_ -— «-
—
i 	
r~ 1
^^ ;
--t-
f i :-:
I -.
' —
•:''
Jr|
i
"~~"'
— *.
I -r, :
	
•::r
--
—
.....
:=.
r -

0
..„
	 p
z^n
.-::r|
~-
	
—

» i - -
—
1 	 :
f— - -
zrrr
—
t:- -
--— -
	

- ' t -
-itn.
!~* *
, —
r—
-
rzr~~r
_ _^.
Supersaturation - araagr) /Ks
, . . | . ,.. j . . . , f. , | . . . j, , , , j. . . . | . .,.(
...._„..., ,.,.., .—.,..-„ „..» 	 ..., ^^-...^
J_«X X • ^- J. • ^J
1

t-, 	
.. —
— —
-- —
— rr*
•— -
-—
^—-
,=
. 	 —
t---
k— -
-•-
•— -
.....
' r
Hi

| • ' • • • -•] i ' • r |

—
tTr-
t..i _
—
vn
:r-)
— ^
'
— t- -
1
"
—
~~
.-•1
!i:
. . 1 1
rxi.
_^_
~
t —
*
1 	
1 	
-
^
:
.— ,

, i . -ii •'.

::
:.--
~— Z
—
-r~
	
=
1 	 1
- —
~ -
	 _|
p
__.
--
	 ~ .
	 • —
:" |
. ;. 	 i
-r-
	
- — _.
' i •



ft!!

~~~
-:
	
	
t 	 .— —




• •
.
::,
•~
-
- • —
- — .
— .
-— |
m;:


•

,...
—
h__.
~:_:
r_ -
._.
.::-
.. .
-
:=-
-
—
:.-.
:;•


'

_-;-:.
•— ;
- — -
	 _


i1'
i 	
." —
I--
t-"-
~.~7
1 	
.:::
-


."4 ^"~ "1.5 ""
                                                  1008

-------
to
5.

g
                                                                            i
                                                                           CO

                                                                           w

                                                                           p
                                                                           O
                                        1009

-------
 Rrdian Corporation
3MOAL CPi LK BLVC • P il BOX 1715 • AUSTIN T'/AS 73757 • TE LE°HC NT 5' 2 4SJ 9535
supersaturation with respect to calcium sulfate and ra be the
relative supersaturation with respect to calcium sulfite, i.e.,
                                         'SpCaSO
                                                    3
respectively.  Using analytical chemistry methods described by
K. Schwitzgebel (Paper No. 9a) the molalities at appropriate
points in the system were measured.  Using a programmed equili-
brium model, activities of the ionic species were calculated and
it was thus possible to describe r: and ra in any piece of
equipment in the system.  These values of rx and rs are given
in Figure 3-1.

          From the figure it can be seen that the slurry from
the scrubber is caught in the hold tank.  In the hold tank rl
and Ty decrease (de-supersaturation occurs) .  The hold tank
overflows into the delay tank.  Here further de-supersaturation
occurs and rx becomes almost unity.  The highest value for r1
occurs in the clarifier where there is a relatively long hold
time in contact with the air and the solids settle out.  The
reason for this appears to be that the clarifier is a poor
liquid-solid mass transfer device.  The total sulfur in the
clarifier (sulfite plus sulfate) stayed essentially the same
which indicates that little or no precipitation occurs.  However,
since the solution is in contact with a large air volume for a
relatively long time period, oxidation from sulfite to sulfate
occurs.  Therefore, rx increases and r2 decreases.  The same
events apparently take place in the limestone feed tank where
the total sulfur remained essentially constant but r1 increased
and r2 decreased.
                              1010

-------
 p^^:~.
                                    B''X?948 •  AUSTIN. TEXAS 73757 • IFLEPHONr 5l2 <5-f 9535
          From Figure 3-1 it can be seen that rx for calcium
sulfate stays below the critical value of 1.3 to 1.4 in every
part of the system.  On the basis of these values of the  relative
supersaturation we would predict that calcium sulfate  scale
formation would not take place.  A possible exception  would be
the clarifier outlet.  The pilot plant observations are in
agreement xvith our predictions in that no sulfate scale forma-
tion occurred.

          Since the laboratory investigations for the  calcium
sulfite system have not yet been conducted there is no critical
vnluc of r_ with v.Mch to compare the pilot plant values.  The
pilot plant data showed however that a value of r2 up  to  at least
8 can be tolerated without the formation of calcium sulfite scale,
In addition there is probably some value of Ty above which sul-
fite blinding occurs in the system.

4,0       SUMMARY

          Laboratory investigations and pilot plant observations
indicated that sulfate scale-free operation will occur at values
of relative supersaturation of less than 1.3.  Pilot plant
observations also indicated that sulfite scale-free and blinding-
free operation occurs at values of relative supersaturation with
respect to calcium sulfite less than 8.

          These types of information are required for  use in
the design procedure.  After heat and material balances have
been made throughout a system, the scaling potential indicators
can be used to predict operability with regard to scald formation.
                              1011

-------
 Radian Corporaticn
~;i • AUSTIN Ti/AS 78757 • TELEPHONIST 454-953$
                           ACKNOWLEDGEMENTS

           This work was  funded by  the Tennessee  Valley Authority,
Division  of  Power Research,
                                  1012

-------
SAMPLING AND ANALYTICAL METHODS
     J.A. Dorsey, Chairman
         Participants:
      Klaus Schwitzgebel
      E.A. Burns and A. Grunt
      Terry Smith and Ronald Draftz
      Terry Smith and Hsing-Chi Chang
      R.M. Statnick and J.A. Dorsey
      Gene W.  Smith
                10.13

-------
                                 SUMMARY
                       SAMPLING AND ANALYTICAL METHODS
                         J.A.  Dorsey,  Chairman
     The concluding session of the symposium dealt with  the measurements
programs developed primarily for the  EPA prototype scrubber tests at the T
Shawnee Steam Generating Station in Paducah, Kentucky.   The methods are
also applicable  to other lime/limestone process development studies.
The EPA program is designed to acquire  extensive data on the composition
of the process streams under varying operating conditions.  The data will
be utilized to perform a complete evaluation of the process chemistry,
define operational problems, and model the process.  This program results
in a requirement for extensive sampling and analysis with a high degree of
accuracy.  While the basic chemistry of the process is rather straight-
forward, in actual operating practice  the system presents a three-phase
system that does not achieve equilibrium in the scrubber pass.  This failure
to achieve  equilibrium  (coupled with reactions  producing undesirable
soluble species, side reactions producing  shifts in the  sulfite-
sulfate oxidation rates,  and potential  scale-producing  species)
presents  significant  sampling and analytical difficulties.

     Sampling problems were discussed  in the papers by K. Schwitzebel and
by G. Burns as they relate to separation of  the  liquid and solid
phases in the unstable slurry from the  scrubber downcomer.  Several
techniques were devised, one based on  centrifugal separation followed by
filtration and one employing only filtration.  Both of these techniques
provide  separations  in less than  15 seconds.  Sampling  techniques for
gaseous and particulate species in the  gas phase were discussed in
papers by R.  Draftz and by R. Statnick.

     Characterization of the slurry components requires  analysis of nine
ionic species in both the liquid and  solid phases.  This produces a labo-
ratory load of over 450 analyses per day.  K. Schwitzgebel discussed accu-
rate referee methods and manual field  methods for the required analysis.
G. Burns presented evaluations of instrumental methods and development of
on-line monitors for the slurry solids  and liquor samples.

     The gas-phase analysis of sulfur  and nitrogen oxides was discussed

                                1014

-------
by R.  Statnick.  An evaluation of instruments suitable for continuous
monitoring was presented.  The development  of a manual-size selective
particulate sampling system  for suspended particulate in the gas phase
was described by R. Draftz.

     Finally, a discussion of the proposed  EPA methods for new source
performance  standards was presented  by G. Smith. These methods or an
equivalent will be required  to define compliance with emission standards.
Hence, anyone developing  control systems should incorporate similar tests
into  his program,in addition to the tests being used for engineering
analysis.  A copy of the  December 23, 1971, Federal Register, containing
the final version  of  the methods, is included with  the symposium
papers.
                                 1015

-------

-------
                                                  • • ^^ M^
                                                  B ^^ • •

                    8500 SHOAL CREEK BLVD. • P. O. BOX 9948 • AUSTIN. TEXAS 78757 • TELEPHONE 512/454-9535
                   DEVELOPMENT AND FIELD  VERIFICATION OF
                     SAMPLING AND ANALYTICAL METHODS
                                 FOR SHAWNEE
                                    By:
                            Klaus Schwitzgebel
                               Presented at:

                   SECOND INTERNATIONAL LIME/LIMESTONE
                          WET SCRUBBING SYMPOSIUM
                            8-12 November  1971
                          Sheraton-Charles Hotel
                          New Orleans, Louisiana
                               Sponsored by:
                     Environmental Protection  Agency
                          Office of Air Programs
                      Division of Control  Systems
                                 1UI7
;HEMICAL RESEARCH •  SYSTEMS ANALYSIS •  COMPUTER SCIENCE • CHEMICAL ENGINEERING

-------
 Radian Corporation
8500 SHOAL CREEK BLVD • P. O. BOX 9918 • AUSTIN, TEXAS 78757 • TELEPHONE 512 - 454.9535
1.0       INTRODUCTION

          The work presented here describes  the  analytical and
sampling techniques for the forthcoming  test facility at  Shawnee.
One of the main objectives is the collection of  engineering
design information for lime/limestone based  SOE  removal processes
Therefore, the demand for accuracy of the  analytical  chemistry
methods is more stringent than the demand  for accuracy in control
processes.

          The problem areas in analyzing unstable  slurry  streams
are sampling, sample handling and analysis.   Two kinds of
analytical methods were selected, referee  methods  (used also as
back-up) and rapid field methods.

          The ultimate use of the analytical results  is for
chemical engineering purposes.  The chemical engineer uses
the data to describe:

              vapor-liquid mass transfer characteristics
              in the scrubber

              solid-liquid mass transfer rates throughout
              the system

              scaling potential

          A mathematical description of  the  reaction  kinetics
of the mass transfer steps as a function of  the  liquor composi-
tion is a prerequisite for the process engineering of limestone
based sulfur dioxide removal processes.  The driving  force term
in these rate equations is a function of the difference between
actual and equilibrium conditions.  The  driving  force term for
                              1018

-------
  Radian Corporation
8500 SHOAL CREEK BLVD. • P O. BOX 9948 • AUSTIN, TEXAS 78757 • TELEPHONE 512 - 454-9535
liquid-solid mass transfer  is a  function  of  the  difference
between actual and equilibrium activities.   The  driving  force
term for gas-liquid mass  transfer  is  a  function  of  the difference
between actual and equilibrium vapor  pressures.

          The solutions are non-ideal.  Concentrations are  not
suitable quantities for describing the  driving forces.   Thermo-
dynamic concentration, or activity, must  be  used.   Only  the
total amount of a species in solution is  measured by  chemical
analysis.  The activities of the species  of  interest  differ
markedly from chemical analysis values  due to complexation  and
the deviation of the solution from ideality.  An example is
given showing how the activities of the important species can
be extracted from the results of the  chemical analyses.

2,0       PROBLEM DEFINITION

          The basic equipment arrangement for limestone
injection wet scrubbing (LIWS) processes  is  shown in  Figure 1.
The three streams entering  the system are flue gas, particulates
and make-up water.  Three streams  leaving the unit  are cleaned
stack gas, solid waste products, and  scrubbing liquor.   The
composition of the incoming streams provides a means  of  predict-
ing the liquor composition  on a qualitative  basis.  The  important
species in the LIWS process are:

Group I                      Group  II                  Group  III
calcium                      sodium                trace  elements
sulfite                      potassium            iron
sulfate                      magnesium            cobalt
                             chloride              nickel
                             nitrate               copper
                             nitrite               manganese
                             carbonate

                               1019

-------
GAS SPECIES
   FG, SG
                 STACK GAS
                    SG
                           WATER
                           MAKEUP
                             WM
1. S02 '
2. C02
3. NOX
4. H20

D. U2
6 CO
' ,. SCRUBBER
7. N2 s
1 1
FLUE GAS
FG * A
1
SCRUBBE
BOTTOMS
SB
/
                              SCRUBBER FEED
                                  SF
                                                 y  v L
                                                 Limr-.'-mara t-rmwr
                              PROCESS
                               V/ATER
                             HOLD TANK
                                 P
                                       SLURRY RECYCLE  SR
LIMESTONE
FLY  ASH
 SOLIDS
   LA
 r. CoO
2. MgO
3. CaS04
4. MgS04
5. CoS05
6. MgSOs
7. CoC03
8. MgC03
9. FLY ASH
10. SOLUBLE No
II. SOLUBLE Cl
 SCRUBBER
 EFFLUENT
HOLD TANK
    E
                                                CLARIFIER
                                                 LIQUID
                                                  CL
                                       CLARIFIER
                                         FEED  .
                                          CF
CLARIFIER
    C
                                  CLARIFIER
                                  BOTTOMS
                                    CB
                               FILTER
                                 F
                                                                FILTER
                                                                LIQUID
                                                                 FL
                                   FILTER
                                   BOTTOMS
                                    FB
                                  PROCESS SOLID SPECIES
                                    (CF.SR.CB,  FB, SF)
                             I. CoO
                             2. Co(OH)2
                             3. CoC03
                             4. CcS03 •  xH20
                             5. CoS04 •  xH20
                            6. MgO
                            7. Mg(OH)2
                            8. MgC03-xH20
                            9. MgS03  • xH20
                            10. FLY ASH
   PROCE^
   LIQUID
   SPECIE
SB.CF.SR,
FB.CL.FL,

  I.H*
 2. OH~
 3. HSOj
 4. SOf
 5. SOf
 6. HCOj
 7. COf
 8. HS04
 9. H2S03
 10. H2C03
 II. Co++
 12. CoOH"
 13. CoS03
 14. CaC03
 15. ColiCC
 16. CoS04
 17. CoNOj
 18. N03
 19. Mg + +
20. MgOH'
21. MgSO/
22. MgHCC
23. MgSO:
24. WgC03
25. No+
26. NoOH
27. NoC03
28. NoHCO
29. NoS04
30. NoN03
31. cr
    FIGURE  1    WET  SCRUBBING  SCHEME
                               1020

-------
 Radian Corporation
8500 SHOAL CREEK BLVD • P. O. BOX 9948 • AUSTIN, TEXAS 78757 • TELEPHONE 512 - 454-9535
          The components  listed  in  Group  I  are  the  most
important.  They dominate the  process  by  participating in the
gas-liquid and liquid-solid mass  transfer steps.  The species
listed under Group  II contribute  to the process performance
in three ways.  First,  they influence  solubilities  which  are
dependent on the ionic  strength  of  the solution.  Second,  they
form ion pairs with Group I compounds.  Finally,  they influence
the driving force for the mass transfer rates.   The components
in this group form very soluble  compounds with  the  exception of
magnesium hydroxide and calcium  carbonate.   In  a closed loop
operation there is a buildup of  the soluble compounds, since the
only stream in which they can  leave the scrubbing unit is the
liquor adherent to the  solids.  This fact must  be kept in mind
when selecting analytical methods.   The procedures  must give
accurate results in those cases where  the soluble species build
up to a high level.  The  implications for  the  selection of methods
for sulfate and sulfite will be  discussed later.

          The third group is comprised of species leached
from the fly ash and impurities in  the limestone.   The concentra-
tion of these elements  is never very high, since it  is limited
by the solubility of the  hydroxides in the  alkaline parts of
the scrubbing unit.  Their importance  is  based  on the fact that
they are excellent catalysts for  sulfite  oxidation,  even  if
present in the parts per  billion  range.

          The process simulations,  discussed  earlier by
D. M.  Ottmers, Paper #lc,  gave a  valuable basis  for estimating
anticipated concentration ranges.   Estimation was necessary
since no data on a closed loop system  operated  over an extended
period of time were available at  the time of  analytical method
development.
                              1021

-------
 Radian Corporation
8500 SHOAL CREEK BLVD. • P. O. BOX 9948 • AUSTIN, TEXAS 78757 • TELEPHONE 512 - -64-9535
          As a general rule,  the higher  the  accuracy demand
of an analysis, the higher  are  its  costs.  This  fact raises the
question as to the ultimate use of  the analytical  results.   The
accuracy requirements for routine,  day-to-day  operation are
less stringent than the requirements  for process analysis.   One
key objective of the forthcoming tests at  Shawnee  is the collec-
tion of engineering design  information.  From  an engineering
point of view the following areas are of ultimate  interest:

              gas-liquid mass transfer rates in
              the scrubber

              dissolution and precipitation  rates
              as function of  liquor composition

              scaling potential.

          The driving force term in the  mass transfer equations
describing these rates is a function  of  the  difference of the
actual process condition and  the equilibrium condition of the
system„  In other words, the  rates  are a function  of the
difference of two activity  expressions.  The closer  the system
operates to equilibrium the more severely  analytical errors will
influence extracted rate correlations.   For  LIWS processes  the
analyses of the species listed  in Group  I  are  therefore the most
important.  Error propagation calculations showed  that the  error
in these analyses should not  be greater  than about 2%.  The con-
centration of the species influencing the  ionic  strength (Group
II) must be known within about  4%.  The  accuracy requirements for
the trace elements effective  as catalysts  are  still  less stringer
Twenty to fifty percent is  considered to be  sufficient.
                              1022

-------
 Radian Corporation
8500 SHOAL CREEK BLVD. • P. O BOX 9948 • AUSTIN, TEXAS 78757 • TELEPHONE 512 --154 9535
          The analytical results are  influenced  by three steps:

              solid-liquid  separation
              sample handling
              actual analysis

These problem areas will be discussed next.

3.0       SAMPLING

          The scrubbing system can be divided  into an  acidic
and a basic part.  The environment is acidic in  the scrubber
itself and in the pipe between the scrubber and  the effluent
hold tank.  The solutions circulated  in  the rest of the  system
are alkaline.  For sampling purposes  it  should be noted  that
the scrubbing slurry, especially in the  acidic part of the  system,
is not in thermodynamic equilibrium.   The  sorbent tends  to  dissolve
and sulfite and sulfate tend to precipitate.   The technique often
used to sample this stream  is collection of a  slurry sample in a
beaker and filtration through a Buchner  funnel.   This  technique
results in only semi-quantitative results  for  the follow-on
chemical analysis for three reasons:

              loss of acidic gases (SO,, , C03)
              especially if a vacuum  is  used

              solid-liquid mass transfer during
              the sampling procedure

              sulfite oxidation by air oxygen.

Because of these sources of error most of  the  pilot plant data
presently available must be considered to  be qualitative in nature
and not suitable for the extraction of engineering design informa-
tion.  In-line, positive pressure filtration was  the sampling

                               1023

-------
 Radian Corporation
8500 SHOAL CREEK BLVD • P O BOX 97-18 • AUSTIN TEXAS 78757 • TELEPHONE b!2 4549535
method selected after field tests  at  several pilot units  (see
Figure 2).  The sampling apparatus consists of a positive  pres-
sure pump, a membrane filter holder and lines and valves to
control sampling and purge rates.  Flow rates used in the  tests
were about 1300 ml/min.  The residence time of the slurry  is
about 2.3 seconds in the filter and approximately seven seconds
in the entire  sampling  equipment.

          The degree of mass transfer in the filter cake,  which
is by nature a good contacting device, was checked by taking
consecutive samples and plotting the chemical analysis results
as a function of the filtered volume.  Extrapolation to zero
volume of filtrate represents the  true aqueous phase composition.
With the exception of carbonate, the amount of solids dissolved
or precipitated in the  filter cake was within the experimental
error of the chemical analyses.

          Loss of acidic gases is  avoided by the positive  pressure
filtration, and air oxidation of sulfite is prevented by fixing
the sample immediately.

4.0       FIXING OF THE FILTERED LIQUID

          After filtration care must be taken that the liquid
samples do not undergo  further change.  This is especially true
for the sulfite analysis.  Sulfite losses can occur by  :

              evaporation from acidic samples
              oxidation by air oxygen
              interaction with nitrites

Nitrites can be formed  by absorption of NO and N0a from the  flue
gas.  All three sulfite losses can be avoided by quenching the
sample in a solution of pH = 6 with knoxvn iodine content.
                                1024

-------
      J-l
      cu
      O
     a
      0)
      4J
        -00-
               B
CO

•H 0)

•H 3
tO CO
O to
P-i CU
   J^
   P-I
             a
           CU -rl
                                        B
                                        H
                                        En
 PI
•H
 ctf
 ^
H

 bO
 C
•H
!-l
 Cu
 S
 C(J
00
                                              CO
                                               ciJ
                                               O
                                              •H
       CO
1025

-------
 Radian Corporation
8500 SHOAL CRLTK BLVD • P. O BOX 9948 •  AUSTIN TEXAS 78757 •  TCLEPHONC 51? •!5'i-9535
          Carbonate losses  from  acidic  liquid can be avoided
by quenching the sample  in  a  solution  of pH = 10.  EDTA must
be added to the buffer in order  to  avoid calcium carbonate
precipitation at this pH.

          Sulfate in the presence of  sulfite is determined
as the difference between the total sulfur and the sulfite
sulfur.  In order to avoid  sulfite  losses and sulfate precipi-
tation the sample for total sulfur  analysis is quenched in a
HP Op-water solution.  Hydrogen peroxide oxidizes the sulfite.
Dilution with distilled  water prevents  sulfate precipitation in
the sample bottle.

5 • °       LTQUID PI-IASE ANALYSIS  METHODS

          The literature was  surveyed  through 1970 for analytical
methods which might be applicable to  solutions of interest.  The
sources consulted were:

              Kolthoff and  Elving,  "Treatise on Analytical
              Chemistry"

              Biannual Reviews on Analytical Chemistry

              1969 Book  of  ASTM  Standards

              FWPCA Methods for  Chemical Analysis of
              Water and  Wastes

              Chemical Abstracts

              Pertinent  Original Articles
                                1026

-------
 Radian Corporation
8SOO SHOAL CRfcEK BLVD. • P. O. BOX 9748 • AUSTIN, TEXAS 78757 • TELEPHONE 512 - 154-9535
          Promising methods  for application  to  the  analysis of
wet scrubbing liquors were checked  in  the  laboratory  and  through
visits to manufacturers.  Two  types  of  analysis methods were
sought:  referee  (and back-up) methods  and rapid  routine  procedures
5.1       Back-Up Methods

          The literature review  revealed  that  the  choice  of  method
for sulfate, calcium and sulfite in  the key  or Group  I  species  is
rather limited.  The methods  published for sulfate analysis  can be
divided into five groups.

          1.  Gravimetric Procedure
          2.  Direct Titrimetric Procedures
          3.  Indirect Titrimetric Methods
          4.  Colorimetric Techniques
          5.  Turbidimetric Procedures

          All these methods,  with few exceptions,  are based  on
the formation of insoluble barium or lead sulfate.  They  all,
therefore,  show the same potential interferences,  namely,  copre-
cipitation  errors, occlusion  of  foreign salts,  and errors  due to
supersaturation ;   The techniques  most widely used  are the  gravi-
metric method (ASTM referee method), direct  titration in  a water
ethanol mixture using 133(010,,.);,  or BaClP as  titrant and thorin
as end point indicator, the barium chloranilate method  (FWPCA
method), and the turbidimetric technique.

          The gravimetric procedure was rejected for  two  reasons.
First, it is extremely time consuming, and second,  there  are
interferences expected in scrubbing solutions  with high salt back-
ground.  Alkali metals cause  errors due to occlusion.  Calcium
causes serious errors due to  coprecipitation.   Nitrate is  reported
                                  1027

-------
        Corporation
flOO SHOAL CRICK BLVD •  P O BOX 99-53 • AUSTIN. TEXAS 7P757 • TELEPHONE 512 - 454 9515
to cause  errors by occlusion.  The  occlusion and coprecipitation
problems  due  Co cations can be avoided by use of ion exchange
resins.

           The direct titration using thorin as end point  detector
was rejected  due to severe anion  interferences.  Figure 3  shows
the errors  caused by several common anions.
                                             FIGURE 3
                                        Reference:  Fritz, J.  S.,
                                        and S.  S. Yamamura, Anal.
                                        Chcm.,  27_, 1461-1464~(T9"55)
          il'B of tilfntio'i of ?,n!f,'i!o in presence of
          coincorUralH-tis at cotiVJion among
Compensation  for these anion interferences can be made by
standardizing the titrant solution with a Hr SO.,  standard con-
taining  the  foreign ions at a concentration corresponding  to
that of  the  unknown solution.  This technique may be useful in
routine  analytical work, but is  not acceptable for sulfate
determination in varying environments.
                                  1028

-------
 Radial! COTOrStiOn  SMO SHOAL CRUKBLVD • PO.BOXWB • AUSTIN, TLXAS 7875? • TELEPHOW bi? - in-9535
          The  turbidimetric  sulfate determination is recommended
by ASTM mainly as a control  procedure where concentration and
type of impurities present in  the  xvater are relatively constant.
Kelly and Baldwin [Chora.  & Ind_. ,  1283-1285 (1969)] automated
this technique using an autoanalyzer.   They report results more
consistent than with manual  operation.   Compared with gravimetric
techniques, however, they found  deviations of ± 10%, which is
unacceptable.

          Extensive laboratory effort was devoted to the barium
chloranilate method recommended  by FWPCA.  The laboratory results
revealed ni trate and chloride  interference if these anions are
present at higher concentrations as well as a critical dependence
on pll.  In addition, inconsistencies were found if different
batches of reagents were  used.   The time required for complete
reaction and precipitation to  take place as well as the time
required to separate the  very  fine barium sulfate precipitate
from the acid chloranilate solution gave very little hope for com-
plete, fast automation of the  method for very accurate determinations
These precipitation difficulties were one of the main reasons that
the barium chloranilnte method was also abandoned by Technicon.

          The method ultimately  adopted is an ion exchange
alkalimctric procedure.   Sulfate in aqueous solutions is deter-
mined as sulfuric acid after passage of the sample through a
hydrogen form cation exchange  resin.   The aqueous acid mixture
obtained after the cation exchange is  evaporated to a few milli-
liters on a steam bath or a  hot  plate.   After this step all the
acids a.nd the water are driven off by  evaporation at 75° C.   Only
1L SO  and other nonvolatile  acids  such as H5rO.: remain.  If
]J:,P(\ is absent, the sulfuric  acid can be titrated directly.
If II,PO.  is present, !L S0: is  driven off at 275°C and the H3PO.t
.is determined  by alb :liircl;ri c  titration.  The method is free
(: r om c a I: i o n i n t e r f c r < • n c e .
                                1029

-------
 Radian Corporation
8500 SHOAL CRCEK BLVD • P O BOX 9943 • AUSTIN, TEXAS 78757 • TELEPHONE 512 454 9535
          Phosphoric acid presents the most  severe  anion
interference.  This acid must be determined  separately  if  it
is present in large amounts.  Table 1 compares  the  results
obtained with the volumetric method to the results  obtained
with the gravimetric procedure.  The method  was  checked in
Radian's laboratory using high salt backgrounds.  Laboratory
results are presented in Table 2.  Results in analyzing field
samples agreed within 270 with the X-ray fluorescence  technique
discussed later.

          The methods for sulfite determination  in  the  presence
of nitrite are also rather limited.  The  normal  iodine  thio-
sulfate procedure gives erroneous results due to nitrite-sulfite
interaction at the low pH values used in  the procedure.  The best
method found was to quench the filtered liquid  in a buffer of pH
= 6 containing a known amount of iodine.  The back  titration must
be done at this pH value with arsenite instead  of thiosulfate,
since thiosulfate is partly oxidized to sulfate  in  the  presence
of nitrite at pH =6.  A dead-stop technique for the  end point
detection was chosen.  At low sulfite concentrations  this  leads
to better results than the starch indicator  normally  used.

          The most convenient method for  the determination of
the third species in Group I, calcium, was found to be  atomic
absorption.  A 5% HC1, 1% LaCl3 solution  used to dilute the
samples into the optimum range for A.A. measurement x\?as found
to suppress all the interferences.

          Table 3 summarizes the referee  methods found  to  be
most suitable for scrubbing liquor anlaysis.
                               1030

-------
 Radian Corporation
P500 SHOAl CRf [K BLVD
               O. BOX 99.48
                      AUSTIN, TEXAS 737S7
                                  TLLLPMGNC 512
                               TABLE 1
        GRAVIMETRIC AND VOLUMETRIC DETERMINATIONS OF SULFATE

o
Sample
1
2
3
4
5
6
7
8
9
10
11
12
Sea Water
(one sample)


Sexv/age0
(one sample)


IN VARIOUS SAMPLES
Sulfate Found, n
Gravimetric
427
421
391
188
167
161
104
96.5
87.5
60.2
39.4
27.0

2,650
2,640
2,640

125
124+
124

Volumetric
426
419
390
188
165
162
101
95.7
87.0
59.7
40.1
26.4

2,630
2,630
2',630

124
124
124
a. Natural  waters (surface  streams, wells, reservoirs, etc.)
b. Diluted  1  to 50 prior  to ion exchange.
c. Filtered and broininated  prior to ion exchange.
                                 1031

-------
                              TABLE 2

                DETERMINATION OF SULFATE IN LIMESTONE

        INJECTION SIMULATION SOLUTIONS USING DIFFERENT RESINS
Exper.
t
1
4> U> N>
Amberlite
5
6 x
7 1
Q
8
9
10
X
Uu
f*j
13
SAMPLE
10 ml Simulation Soln.*
n n
" plus H202
11 plus Na2Si03
10 ml Simulation Soln.*
it n
" plus H202
11 plus Na2Si03
10 ml Simulation Soln.*
n ii
11 plus HS02
11 plus Na2S103
10 ml Pure K2S04 Soln.
ml 0.0502N
NaOH used
3.92
3.94
3.96
3.95
3.95
3.95
3.96
3.96
3.95
3.96
4.04
3.96
3.97
m mole
Theory
0.0990
n
it
tt
it
11
n
it
it
it
11
11
2 S04
Exper.
0.0984
0.0989
0.0994
0.0991
0.0991
0.0991
0.0994
0.0994
0.0991
0.0994
0.1014
0.0994
0.0996
Percen
Error
-0.7
-0.1
+0.4
+0.1
+0.1
+0.1
+0.4
+0.4
+0.1
+0.4
+2.3
+0.4
+0.6
Resin column dimensions:  1.2 cm I0D.  x 18 cm high.
Resins used:  Amberlite CG-120, Dowex SOW, and Rexyn 101.
              All were 100-200 mesh size.
Simulation solution contained:
0.009904 M K2S04
0.150 M Ca(N03)2
0.100 M NaCl
0.05 M HC1
Samples 3, 7, and 11 contained 2.2 m moles H202 which was added as  5  dro
of 30% solution.
Samples 4, 8, and 12 contained O.lm rnole Na2SiOa  which was added as  1 m
of 1 M solution.
                                  1032

-------
•(
•1
D
y
3










E
P
E
•H
4-J
CX
O




































h
o
cd
5-J
p
o
o
<
d
O E
•H a
4-J CX;
CO
5-1 CJ
•!-) K
£ C
QJ ftf
O 5-1
d
0
o
G
o
•H
4J
cd
4-J
a!
o;
Pi
P
r4
•I-J
C:
f ;
H




o

4J
CD
^
CD
CD
5-1
CD

CD
&






B-? B^ ^S
cn m CN








in
•
rH VO P-.
VI VI VI


1 1
0 0
4J 4J
O <3
rd co
CM ^ H
O W O
5-1 0- U
4~> n P
O C\; rH
CD t-i •.-(
CX P,
CO *
r-l CJ
• Q) T<

J-i
O
1
•H
rH
cd
*^
rH

4-J
0!
P
r/j

d
0
•H
CO















o
•H
£_l
4-J
CD
E
O
4-J
co cd
C M
E 4J
P -H
rH -!J
• O
O O
•H
0) 4-J
^ ') Cj
c. E
cj o
.a 4.1
0 p
;>•' (vj
Wv_x




















0
O
rH
r4
QJ
4-J
d
M

O
•H
£
4J
Q)
E
P
5-i
4-J
co
d
O
O
1
0
o
cn

4-1
(I)
5-J
P
pq
CXM
rj
!-i
bO
0
d
c j
CJ
o
o





T3
O
r]
4J
Q)
)2^

d
o
•H
4J
P
rH
O
J>
w
rH
Cd
C
O
•H
4-1
Cj
d




















d
o
•H
CO
•r-l
Q
0
5-1
PM

|
nj

4-J
•r-)
H

O
•H
5-1
4-1
CD
E
•H
P
O
5^
CD
f£
0
4-J
cd
5-i
4J
•H
4J

O
•H
4-.J
1 j
E
O
4-J
rj
cj
*-^
^_^
O
•H
5-t
-P
Q)
r"
P!
O
•H
4-J
d
CJ
4-J
O
CX
^
Ci
0
•H
4-J
in
CN
i
m

CO
P
4-J
cd
5_(
CO
CM
•<]

co
CO
«5
rH
O

T3
d
CO

m
ffi
21

O
4J

d
o
•H
4-J
O
r^
T3
CJ
&4
Cu
E
CO
d
•H
bO
^
























d
o
•H
4J
rd
rH
rH
•H
4-J
CO
•H
Q
O
O
o
o
o
rH
QJ
4-J
QJ
E
O
4-J
O
cx
o
5-1
4J
CJ
QJ
CX
CO

0
•H
5-i
4J
QJ
E
•H
5-1
O
rH
O
CJ

M
0
•
^>
•
£3




















0
•H
a
o
5-1
4-J
o
s
d
!_)
r\
O
rd
4-J
•H
5
                                                                           B-S
                                                                           CO
                                                                                                                            O
                                                                                                                            o
                                                                                                                            o
                                                                                                                            o
                                                                                                                            o
                                                                                                                            rH
                                                              0)
                                                              4-J
                                                              QJ
                                                              E
                                                              O
                                                                            a
                                                                            o
                                                                            5-1
                                                                            4-1
                                                                            O
                                                                            QJ
                                                                            CX
                                                                            CO
                                                       O O
                                                       CN r-l
                                                                           ,0
                                                                        ^n  pt
                                                                        CX CM
                                                                        CX
                                                                           O
                                                                        OO
                                                                        CN CN
                                                                         I   I
                                                                        m o
                                                                        co
                                                                        cx cx

                                                                        §§
                                                                        r-lrJ
                                                                                      o o

                                                                                      •U 4J
                                                                                      CM CM
                                                                                      r4 5-1
                                                                                      O O
                                                                                      CO CO
                                                                                      O O
                                                                                     •H-H
                                                                                      S E
                                                                                      O O
                                                                                      -U 4-1
                p
                O
               CO

                CO
                OJ
               •H
                CJ
                C)
                fl
               CO
                            +     rcj    bO
                             ^   U   >^
  (V
O
CO
  cn
O
CO
  re
O
 nj
4-J
 O
H
l  n
O
O
                                                                                   co  d

                                                                                   ctf
                                                                                  4J   «
                                                                                   CD  QJ
 QJ   «•>
 O  P-H

 5-J
H
                                                                         1033

-------
                    ssoo SHOAL CKCCKBLVD • p. o. BOX 99<8 • AUSTIN, TEXAS 73757 • TELEPHONE 512- 454-9535
5 .2       Routine Field Methods

          The selection of routine field methods was based  on
the type and number of analyses per day required at peak  load
operation „  This breakdown is shown in Table 4.  Table  5  gives
the time requirements per day in the event that the back-up
methods are used.  Duplicate analyses were assumed if not in-
dicated otherwise.  The last column in Table 5 shows the  total
man-rninute/day necessary for each type of analysis.  The  costs
show the following pattern:  Total S > Ca > Mg > S02 >  COS  > Cl
= K, with total sulfur being the most expensive determination.

          X-ray fluorescence appeared to be a technique which
cut the expenses drastically.  However, no data were found  in
the literature describing the use of this technique in  analyzing
liquors of the composition encountered in lime/limestone  based
scrubbing processes.  Figure 4 shows the principle of this  tech-
nique.  The specimen is radiated by a primary X-ray source.  The
elements present in the sample emit characteristic secondary
emission lines whose wavelengths and intensities are measured
using an analyzer crystal and a counter.

          Quantitative X-ray fluorescence analysis is subject to
interferences as are most of the other analytical procedures.
The intensity of the emission line of an element can be reduced
or increased by the other elements present in the sample.   An
increase is observed if secondary excitation occurs.  A reduction
of intensity is caused by absorption effects.  Correction factors
must therefore be determined and the measured intensities correctei

          Tables 6 and 7 present preliminary results in analyzing
simulated scrubber solutions for sulfur and calcium.  The RMS
errors for the two most important key species are quite acceptable
Preliminary results indicate that chlorine, potassium,  and magne-
sium also can be determined by this technique.

                               1034

-------
Radian Corporation
8E.OO SHOAL CRtTK BLVD •  P O BOX 99<3 • AUSTIN, T[XAS 78757 • TELEPHONE 512 - 451 9535
                                      TABLE  4
Species
Ca-H-
Mg4^
K+
Na+
Total S
SOS
Cl"
CO,
Total N
N0~
NO;
NUMBER OF ANALYSES PER DAY
Number of Analyses
Sought at Steady State
53
48
9
9
53
36
9
53
9
9
9
AT PEAK LOAD OPERATION
Analyses During
Line Out
27
27
27

27
27
27




Total
80
75
36
9
80
63
36
53
9
9
9
459
                                        1035

-------
m

w
rJ
PQ
    co
    Q
    O
    ffi
    w
    w
    Pi
    w

    w
    Pi

    p

    (H
    C/J
CO
H

w

s
    o-
    w
    Pi

    w

    M
    H

    Q
    W
    H
    (H


    fc
    W
K*1
CO
Q
^~^
«
d
•H
^Ej
i
C
H


tH
4J
O
H








o o o o o
r^ oo o LTI o^
CM O •
cO
Q
CO
OJ
CO
>-
CO
c
^
c
•H
CO
•H
CO
fx.
tH
cd
d
QJ
E
•H
H





13
O

4J
QJ
S





4-J
bC
O
CO

CO
0)
•H
O
cu
0
CO







00 CN O O H 0)
CO T3 CO 13
V— / \~-S
O MD CM O> CTs
VD o r*^
rH rH





o m m o o
CO rH rH CO CM






13 /-N
0 0
& -H 2 d OJ
4-J O £-1 O3
0 O 0) -H 4J O -H cr1
bC-f-< S M OJ 4J 4J -H
d J-i 4J E co d
CTJ4J d QJO drH X
^ 
O i — 1 I> CJ CL< QJ •
H,
~-^ CO
Q
d ^^

S M
D
0 0
rH ffi
00 1
d
m cO
rH S
m
•
CO
CM

j_l
O














































                                                       1036

-------
"nary X-ray Beam
)n!inuum + Anodo Characteristic)
Secondary Emission Lines
(Sample Characteristic)
ay Tube
velength Measured, X
                                                               Analyzing Crystal
                                                               Changer
                                                                Flow
                                                                Proportional
                                                                Counter
                                                                          >29
                                                               -Scintillation
                                                                Counter
v A  =  2d-sin0
                                 FIGURE  4

                                    1037

-------
  Radian Corporation
Sample
Number
  14
  15
  13A
  16
  17
  18
  19A
  20
  2-1
  22
  23
  24
  25
  26
  27
  28
  29
  30
8409 RESEARCH BLVD. • P O. BOX W8 • AUSTIN, TEXAS 78758 • TELEPHONE 512 - 454-7S35
                                TABLE  6
Results of Sulfur Analyses
Nitrogen and
Corrected
Magnesium Interference
for
Only

Corrected Values
Sulfur Content
(mmoles/jO
25
25
25
25
25
25
50
50
50
50
50
50
50
50
50
50
50
50
Instru-
ment I
25.3
25.2
24.7
24.6
24.9
25.2
48.9
49.2
49.0
49.2
49.7
50.2
49.6
49.9
49.7
50.3
50.4
51.4
RMS =
% Error
1.2
0.8
-1.2
-1.6
-0.4
0.8
-2.2
-1.6
-2.0
-1.6
-0.6
0.4
-0.8
-0.2
-0.6
0.6
0.8
2.8
1.8
Instru-
ment II %
25.5
25.1
	
24.8
25.6
25.5
	
52.2
50.6
50.0
51.8
49.9
49 „ 4
50.8
50.0
50.9
49.9
53.0
RMS =
Error
2.0
0.4
	
-0.8
2.4
2.0
	
4.4
1.2
0.0
3.6
-0.2
-1.2
1.6
0.0
1.8
-0.2
6.0
2.4
                                       1038

-------
       Corporation   wot RESEARCH BLVD. . P.O. BOX r>48 . AUSTIN, TEXAS 73753 • TELEPHONE $12 • 454-9535
                               TABLE  7
Results of Calcium Analyses Corrected
Sample
Number
86
77
78
79
80
81
82
90
91
92
Nitrogen
Chlorine,
Calcium Content
(mmoles/ A)
10
25
25
25
25
25
25
50
50
50
, Magnesium
and Sulfur

Instru-
ment I
10.0
25.0
25.0
25.0
25.0
25.0
25.0
50.3*
50.2*
50.0*
, Potassium,
Interferences
Corrected
% Error
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.6
0.4
0.0
for
Values
Instru-
ment II
9.7
25.0
25.0
24.9
25.0
24.8
24.6
50.2
49.8
49.2

70 Error
3.0
0.0
0.0
-0.4
0.0
-0.8
-1.6
0.4
-0.4
-1.6
                                       RMS  =  0.2             RMS =1.3
•*•
 50  mmoles/4  Calcium values  exceed  linear  range
                                   1039

-------
 Radian Corporation
            SHOAL CRT! K BLVD
                               AUSTIN, TLXA5 78757
                                          TEILPHONL '
          Table 8 compares the results of the proposed  referee
method and of X-ray fluorescence in analyzing total  sulfur  in
samples taken from different streams at TVA's Colbert  Station
pilot plant.  The agreement can be judged as being good.

          A big advantage of X-ray fluorescence not  yet mentioned
is the speed of analysis, especially if a minicomputer  is used  to
perform the matrix interference corrections.  The estimated saving,
in time for the Shawnee tests are reflected in Table 9.  The man-
minutes/day are cut by a factor of four as compared  to  the  referee
methods.
6.0
SOLID ANALYSIS
          The solid analysis comprises three  steps:

              phase identification
              solids dissolution
              analysis of the liquid phase.

          The phase identification of crystalline compounds
collected on the filter previously described  is most conveniently
done using X-ray diffraction.  Important solid species  (other
than fly ash) potentially present are:

                                     	Magnesium Compounds
Calcium Compounds
1.
2.
3.
4.
5.
6.
7.
8.
9.
CaO
Ca(OH)
CaC03
CaC03
CaS03 •
CaS04 •
CaS04 •
g-CaSC
v-CaSC

2
(aragonite)
(calcite)
%H3 °
2H20
%HaO

'4
                                     1.
                                     2.
                                     3.
                                     4.
                                     5.
                               MgO
                               Mg(OH)s
                               MgCCv
                               Mgso3- :
                               MgS03-
                                1040

-------
   col
   w
    Q
    i-3
    W
    M  a\
    PM  M
    O  H
       CO
    col
DO  W[

y
.-)
pq
<£
H
    Kl
     t
    H  H
W
«
 -i
i—i
O
o

H
    53  HI
O
c
0)
M
(!)
m
m
•M
Q
E-£
cu
o
c
0!
CJ
W
0) ^
i , «. 	
M ^^-
O W
P CD
i—4 t— 1
Pn O
xi
!•— < f3
"3
f^
><
w
•r4
W
>i
•^ di
^*-^ r*
^^~ \-4
W <|
cu
r-4 r-4
O Ctf
6 o
C3 . 1
vS- P
f-i
CO  00 00 O> CM
vO CO  v£> vO ^- ^~ 4J
•H ~^ \ 
rQ  H J^ CO
C
Pi O O P4 cd Pi!
> H H 0 H 0
/*~N /^~S X~S /^^\ X*^S X*^>
r-l CM CO 
-------































CTi

w
h-1
ft-
%
1 — 1
rt
G
rd

0)
•M
Cd
O
rH
3"
T3

j^
c;
>
0)
£



*
CO

0)
s
rd
H


• •
co
•H
CO
rd
s
















Ctj
Q
^
G
•r-
1
d
rt
S
r-l
K(
t>
O
H



>*,
rl
O
~--~.
co
CD
CO
,">
r-l
$
d
<£.


co
•r-l
CO-
pi
r--!
rj
CJ

•^
•
G
«r-|
^£j




1T3
0
rC.

CD
r"* -i

T3
r-l
0
•r-
f^H


4-|
'oi
b
CO

CO
CD
•H
O
D
CO






O O O
00 CO VO
rH CM ~,
G I >,^i CO
O B r-l rt
3 rj w T3
T) kl G co G
CD <3 ^ (U 6
r: G Ki cj 'o o
O O PH CD f
rH G rH \-
W) H W) 0
CO) C 4-
••H 4J -H Q
CO 0) CO T
^-^TTJ "•— '
vO O> CTv O
O
rH

f












O O O
rH CO CM

f"
cu a
'U p ^
c G cr c
Cd O '>H 'f~
•rH C C
C C 4J ,G ,E
o o rt cj c
•H -H rH a; a
4-1 ID 4J rH H t-
2 0 CJ-rH
rH rd D 4-1 •
O 4-' "O CO > >
> JU CD -H • __







2

rH
rt
o) 4J 1 rel
O 0 O C
CJ H ^^
0} £>-i
Q cri
\ Q
• "^^^
G co
3 D
O
O ffi
CO I
CM G
st" 1
|J-V
,
O

N
* J— i
: o
^

4
j
j
j
3

N




















)
J
r


1
4
)
)
\
•












•0


1042

-------
 Radian Corporation
8500 SHOAL CREEK BLVD • P O BCX 97-13 • AUSTIN, TEXAS 78757 • TELEPHONE 512 454-9535
The characteristic diffraction patterns  for these  compounds
are listed in the "Inorganic Index to the Powder Diffraction
File".

          The diffraction patterns of the collected  solids can
be obtained using a film or a goniometer technique.  The  princip]e
of these methods is outlined in Figures  5 and  6.

          The next stop in solids analysis is  the  chemical
analysis of the individual compounds.  Carbonate is  determined
on a weighed sample by the evolution technique.  Sulfite  is
determined by dissolving a x\7eighed sample in an iodine  solution
of known iodine content.  Calcium, magnesium and total  sulfur
are determined in a sample dissolved in  a weak mineral  acid  con-
taining HP0S for the oxidation of sulfite.  The total calcium,
magnesium, and sulfate in solution can then be determined by X-ray
fluorescence or the referee methods described  earlier.  Typical
analysis results are presented in Section 7.

7•°       FINDINGS ON PILOT PLANT STUDIES

          The sampling, sample handling  and referee  methods
were developed and tested by analyzing data from several  pilot
units.  Samples were collected during

              GAP in-house studies

              pilot plant runs at the Tidd Plant in
              Brilliant, Ohio

              pilot plant runs at Key West

              pilot plant studies at TVA's Colbert
              Steam Plant

              pilot plant studies at Shawnee

                              1043

-------
K3di3P COrpOr3*IOn   *w> RESEARCH BLVD •
                           P.O. BOX ws • AUSTIN, TEXAS 78/ss • TELEPHONE 512 - 454-9535
             X-ray Beam
                                Debye Cones
                                  —Cylindrical Film
FIGURE  5  - Principle of the Debye-Scherrer Method Using
           a  Cylindrical Camera
   X-Ray Tube
FIGURE 6  -  Use of the Goniometer Technique to Obtain
            X-Ray Spectra from Powders.   The intensity
            of  the reflected radiation is monitored by  a
            counter and plotted with  a strip chart recorder
            as  function of the angle  29.   The specimen  is
            rotated with half the speed  of the counter.
                           1044

-------
 Radian Corporation
foOO St'OAL CPTCK BLVD • P O BOX 9^,3 •  AUS1I N. TEXAS 78757 • TFLCPHONI 512 - 454-9535
Results will be presented  for  the  pilot  studies  at  TVA's  Colbert
steam plant.  The system arrangement  at  Colbert  is  shown  in
Figure 7.  Samples were taken  and  analyzed  at  the  scrubber
effluent (sample point 2),  scrubber spray  (sample  point  1),
affluent hold tank F-12 overflow  (sample point 3),  and the
process liquor tank F-13 (sample point 4).  The  results  of the
liquid and solid phase analyses are shown  in Tables 10 and 11.
The buildup of inerts is very  small in this arrangement  since
nost of the fly ash was removed by the raw water spray.

          The accuracy of  the  methods is reflected  in  the
imbalance .
                .     .           . • z.
             /.  i    i/pos   i.\  i   i/neg
          The pH measurements and analytical  results  shown  in
Table 10 were used to calculate this imbalance.  The  imbalance
should ideally be zero for zero errors  in  the analytical  deter-
ninations.  Another source of ionic imbalance is the  presence
of species for which no analysis was made.

          The results of the solid phase analyses  are presented
in Table 11.  The concentrations of the  solid  species  add  up to
nearly 10070 with exception of the solids of the scrubber  spray
which contains most of the fly ash.  Compounds leached  from
the fly ash for which no analysis was made may be  responsible
for the low values found .
                                1045

-------
                         cfl
                         60

                         O
                         O
                         o
                         CO
 (U
•H

•H

 cfl
r-l
O
                                  4-1
                                  O
                                  cy
                                  ,0
                                  u
                                  C/3
                                  CJ
                                  a
                                  o
                                  4J
                                  en
                                  cu
                                  1-4
                                  <:
                                  £

                                   i
                                  t^-
                                  O
                                  M
1046

-------
        g
        H
          -4
        to

        c
    rt
    J3
    E
.0
r-f

O
       B
      ^	^
      w
        o
        H
        I  Ci

        :) o
                    eg

                    o
                                 o^
                                 CXI
                    CM

                    O


                    CS


                    O
               3

               o
              t/]
                                                     c
                                                     o
                                                    CJ
                                   1047

-------
w
w
    CO
    M
    CO
    w
CO

O
M

O
CO
    o
    CO
    EH
     v:

     K
        CO
        CM
        OO
Pi
         4-J
         C

         r-t
     •U


     CJ
     ^3
     r-t
     O
                    r-:

                                                          :s«
                             IS
                                                O OO
                                                S33
                                                         <-4 OO *4 [

                                                         ox".s" S*
                                                                                   1048

-------
 Radian Corporation
8 . 0       USE OF THE RAW  DATA

          It was mentioned  earlier that the results of the
chemical analyses have no value per  se.  They gain their value
in the chemical engineering framework within which they are used.
Dominant points of interest are:

             mass transfer  characteristics in the
             scrubber

             solid- liquid mass transfer rates

             scaling potential

          There Core,  the da La presented in Tables 10 and 11 must be
processsecl further,   As r-n  example, suppose one wishes to predict
the scaling tendency of the scrubber effluent given in Table 10.
This task is solved  by considering the ionic equilibria in the
aqueous phase.   The  results of the chemical analysis listed in Table
10,  the pH value and  the temperature were used as inputs for compu-
ter calculations.   The resulting activities of the individual ionic
specJes for the scrubber effluent are listed in Table 12.
          The  activities  of  Ca,  SO^" , and  SO^ are given as
 7.25xlO"3,  1. 22x10-*, and  5.84xl(T3 respectively.  The ratios
 of activity product  to  solubility  product constant at 38.34°C
 for CaS03-  ^H., 0 and  CaS04 • 2Ha 0 are 10.6 and  1.79 respectively.
 This  shoxtfs  that the  solution is highly supersaturated with
 respect  to  CaSO:; ' %H=0  and moderately supersaturated with respect
 to CaS04 • 2EP0.  These  numbers will be of value in conjunction
 with  scaling studies to define scaling tendency.
                              1049

-------
LJ






co
w
M
O
M
Pi
co

h
[_•-!
i--- 1
iv ,
f— -1
C

CO
K
I-!
C-,'
(•" '
^fS
\- '
H
O
"^-v

ki
.' •>,
^. *
U '
n p
r
C."1
t —
r ' •
M
Pi
I, I
co
M
Q

l

CNl
r— 1

K
i~3
r/"
r'-
<




















10
UJ
O
0
a-
1*1
•
CC
Ki

UJ
if
ZJ r-j
f- C)
•3 1
rr L^
i '
c c
-C j
^-s u- ( -
4-J l~ •
C "'
-• O O
c$ — s «"7 fv
H a <-> x
2-
-
CU
CO
•• — ' h'; f-,
f i ( j
1 1
r D
c~ a
c rj
t o r_>
r- f- .
• •
— < cr
1 1 ii
 r • L , Cj c;
L f" c; t.
t < r^ u c
O r- a tn
O C' f~> '^
h- •-< f-'i 1 T r--'
H II II II
2
D C-
~3 (M t v' r i f .•
c o c- --


K— •
r^
'z
U'
t-4
LJ
f-<
U
U.
Ii . l-H f— 1 t— * r-l *-•< «-H *— ^ t-H ^-( f-4 ^ } 'C\ r— 4 r-* ,"_) ^ J r— 1 f ; ^- 4 ,— « ^^ t ^ •— ' r"' ', J .— 1 O •"
c r. ( c c r r C i r, t c c c. c- c f. f- r c r c r: r t r '_ c r
LJ(llllllll|*+ll + *l + l||-»|4+|+(
F- r~ r-: r" C1 i . " 13 ^. f>- n- C^ i." c- fi-. r .-r c: CO n rt; 0^ cr 7^ r ^ ;•-. re
y- .i iT c_; i i c — < ^ , j- r • L" •-« *-, f L: •-> <-• i. r-i C-' r t". •— « i — • •-< i •-< (.
i_ iv o r- M cr rj v: c; o r- t" O >..• f- o t" r- c • r- i<< »— C r-- f_-. «.. " i" r
--,.-<-'•- r~- r~ • -. - r- • •> r~ r- ^-c t— < • r- — < ^-- i— ^-< r-- ••", r- —* r- ^-< ,~< r- ,j r-
t—i
i
u
<3
- u u. i , J f. ^ u. i i i.. i r- r- C". r r- ir. r\ i.; .^, o' 3 1.1 c. r- J f- t".
D 1- 'I Cf O C.J IT C) (D -J) O Ci 'J3 O d '_3 C-1 O LJ O CD O O O rl> C3 C.^ "« —
' "' * ' 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 i I 1 1 l
ij. > r-i: r^ i- C" Lf- — < !•'• >. i •••-- .-< c "~ ;r ; v, t i LI t-~ i > L- ;:• — i c ^ C' •"• u-
>-i > • f , r- ( >- 1 ( - r- LJ t r^i '- -j i. ;--«;••' r j .-i T--, r j i>.. — 1 ^ r . t- t • • a u
n i— — ' f ' — < ' •-' f- i j a r- •—> ' •• ( .. " < ^ <~ f. : (_.~ 3 i • n" i i " j- —4 f • i , •-* r-
^ H «? f f~^ t ' r~* I I ' »-4 f ^- ( r-^ p ' fv_ r . f i^. 4 r^_ ^ _. j , • r, j f , | —^ t t ^ '• * ^^
(—
ID
_J
C
f'
1. -
— .
<_
UJ
n
c: >- u r-: •- n <>j j- 'r r-i i. t r-- r-. <--. r-- r- - , i ^ f. , <-• ^ L' ^ f. ^r r>, c")
<; l_ C-. C^ C . ^ - C~ i . 1 [_, C- C^ t "> C^ C^' C_, L " O C t ' ^J C. - t J C- * ^ Cv C ? '—» •— '
'- 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 | 1 1 1 1 1 1 1 1 1 1
.j c i " ^- . ' - i " J L_ t L. r-- c i "• v. .r r. i c. a j ^ •• j .-.—.( r- • t
-J C_' c, OT C
: o i o f --!(-• '-'- ' c i. t. + c (.- c' 3 1.1 r •* c i , T n c> + e o
r, ( ! •* D" i , c" c i - c~ c i f ' t ., «--; «•* « «j «.-t .= «j cr c i' i' i ' LI- *= <= *-
u T. 2r c: ;r L-- i.'i 3 <_ : n: j J t e..' c> i • e> c i^ r ;-.:.:> i > r: ;: i'









Cj

-------
      4   I

      C  —>
                             c;
                              I
                             (Nl
                             01
                             r\)
                              i
                             LL)

                             tt
     r-  r~>
     o  o
      I    I
                       1.1
                       -J
                       •X
                       Cj

                       !
                  I
                 r-
                 r-

                 
                             o

                             vr
r-!
CO
M
Q
     o
c.  o

C'  r-
o   .—i
r-i  r-
X
d
c_
u.
W
E-i
.-i  Ci
ii  n

CM  (si
c.  o
l/l  O
a   a
LJ
a
I/-.
                                                                               1051

-------
       IsUrpUTdllOn  SBOO SHOAL CREEK BLVD  • PO BOX 9949 • AUSTIN, TEXAS 73757 • TELEPHONE 512- 454.9535
          Another  number of importance for engineering calculation
is the partial pressure of S0a.   From Table  12  it  is  seen that
PCQ  in the  scrubber  effluent is 1.46xlO~s atm  or  about 1.5 ppm.
This is a necessary input for SOS vapor-liquid  mass  transfer
calculations.  In  similar fashion other activities will be re-
quired for solid-liquid mass transfer rates.
                          ACKNOWLEDGEMENTS

          The work  presented here was sponsored  by the Office
of Air Programs,  Environmental Protection Agency under Contract
CPA 70-143, Mr. Julian Jones, Project Officer.
                                1052

-------
    ON-STREAM CHARACTERIZATION OF THE LIMESTONE/DOLOMITE

                    WET SCRUBBER PROCESS
                  E. A. Burns and A. Grunt
                        TRW SYSTEMS
       Chemistry and Chemical Engineering Laboratory
                 Redondo Beach, California
                        1053

Second International Lime/Limestone Wet Scrubbing Symposium
                    November 8-12, 1971
                   New Orleans, Louisiana

-------
      ON-STREAM CHARACTERIZATION OF THE LIMESTONE/DOLOMITE

                      WET SCRUBBER PROCESS
                               by

                    E. A. Burns and A. Grunt
                          TRW SYSTEMS
         Chemistry and Chemical Engineering Laboratory
                   Redondo Beach, California
                            ABSTRACT

     The development of control  methodology for sulfur oxide
and particulates from power plant emissions by wet scrubbing
requires accurate and reliable measurements of process vari-
ables.  Planned OAP process demonstration studies will result
in a requirement for a large number of chemical analyses re-
quiring 1) automatic instrumental methods and 2) associated
data acquisition and processing capabilities which exceed
current instrumental capabilities.  This paper describes acti-
vities undertaken at TRW Systems under Contract 68-02-0007
toward the development of methods suitable for optimization
and control of the wet limestone and dolomite scrubbing pro-
cesses by continuous onstream analytical methods.  Emphasis
was placed on development of continuous on-line methods for
slurry sampling and separation that do not disturb the chemical
steady state condition.  Establishment of sampling requirements
and an effective means for total phase separation in a period
less than thirty seconds were accomplished.

     Analytical instrumental methods having capability of con-
tinuous or slug flow analysis within two minutes were identi-
fied for characterization of the separated solid matter and
liquor.  Analytical methods were identified which permit con-
tinuous X-ray analyses of solid constituents for sulfur, calcium,
magnesium and iron contents.  Liquid phase analyses methods were
established for instrumental analysis of acidity, sulfite, sul-
fate, calcium, magnesium and carbonate contents.  A new method
for rapid analysis of sulfite content based on furfural bleaching
is being carried to a state of prototype analytical instrument
development.  In addition, approaches for total complete on-line
analysis of other wet limestone scrubber constituents have been
identified.
                             1054

-------
                               INTRODUCTION
     The development of control methodology for sulfur oxide and particu-
lates from power plant emissions by limestone/dolomite wet scrubbing re-
quires accurate and reliable measurements of process variables.  Efficient,
proven methods for many of these measurements have not yet been developed.
The monitoring of the complex chemistry involved in this scrubbing process
and associated sampling of representative samples in quiescent and dynamic
mixtures of liquors, slurry and solids are in themselves challenging analyt-
ical problems.  In addition, planned OAP process demonstration studies will
result in a requirement for a large number of chemical analyses requiring
1) automatic instrumental methods and 2) associated data acquisition and
processing capabilities which exceed current instrumental capabilities.
     The chemistry of the process is not sufficiently understood at the
present time because of the lack of definitive mass balance information in-
volving the chemical species existing in the scrubbing solution.  The de-
velopment of suitable on-stream analysis methods will provide a means to
fill this gap through detailed characterization of the process.  High ana-
lytical accuracy is not a requisite of the needed methods but rather they
must be adaptable to instrumental techniques that will be reliable, repro-
ducible, cost effective and employ hardware requiring little maintenance.
This paper describes activities undertaken at TRW Systems under Contract
68-02-0007 for development of methods suitable for optimization and control
of the wet limestone/dolomite scrubbing processing by continuous on-stream
analytical methods.
                    SAMPLING AND ANALYSIS REQUIREMENTS
     As described in earlier papers presented at this symposium, the po-
tential chemical species in limestone/dolomite wet scrubbing processes are
numerous.   Table I lists the major and minor species that could be present
in the limestone/dolomite wet scrubber slurry.  The number of chemical vari-
ables  studied were  limited to  only  the  key parameters affecting the oper-
ation  of  the limestone/dolomite  scrubber in order to  scope the present
program at a manageable size.  As a result, activities were focused on es-
tablishing instrumental criteria and identification of instruments suitable
for on-line monitoring of the following chemical species or characteristics:
                                    1055

-------
                                   TABLE I
               POSSIBLE LIMESTONE/DOLOMITE SLURRY  COMPONENT DISTRIBUTION3
Major Components
Liquid Phase Solid Phase
Ca2+ CaO
Mg2+ "9°
HS03" Ca(OH)2
S042' Mg(OH)2
HC03" CaS03
S032" CaS04
C032" CaC03
MgC03




Minor Components
Liquid Phase Solid Phase
K+ MP04
PO/ SiO,
4 <-
N03" CaF2
Na+ PbS
Fe3+ A1203
Fe2+ S8
Mn2+ ZnS
Cl' Na20
N02" MS04
N03" FeS2
Ti02
C02°3
         Distribution of components is dependent on pH and temperature.
     •    Calcium ion  concentration
     •    Magnesium  ion  concentration
     t    Sulfite ion  concentration
     •    Sulfate ion  concentration
     •    Carbonate  ion  concentration
     •    pH
     •    Ionic  strength
The analysis methods were  selected for evaluation based on their  applica-
bility for characterizing  slurries, filtered solutions and dry and/or  separ-
ated solids on a continuous  basis.
     On review of the  scrubbing process variables several sampling  require-
ments were identified  relating to  characterization of the scrubbers mixture.
These sampling requirements  are shown in  Table II.  During a literature  re-
view phase, sampling,  separation and quenching of reactants were  identified
as major problem areas that  had to be resolved prior to application of any
analytical techniques  to on-stream analysis.  A system capable of handling
                                     1056

-------
                                 TABLE II
                  LIMESTONE SLURRY SAMPLING REQUIREMENTS
             •     Slurry Solids Content - 0 to 15% w/w
             •     Slurry Sample Quantity - <_!% of stream flow
             •     Instream Sampling Rate - <2X stream velocity
             •     Phase Separation
                   •    100% removal of >0.5y particles in liquid
                   •    Lag Time - <30 seconds
             •     Sampling Rate - 30 samples/hr, min
             •     Analysis Time - 2 min, max.
             •     Easily Maintained
a sampling rate of 30 samples per hour necessitated the use of a rapid
separation of slurry and isolation of solid and liquid phases and was a key
milestone prior to developing analytical  techniques.  The sampling rate was
established assuming specific combinations of scrubber designs and analysis
location and sample frequency.  An example which fulfills this requirement
is three different scrubber design processes sampled every 30 minutes at
five different locations.  Variations of sampling locations up to eight and
sampling frequency of up to 15 minutes cover a wide range of samples to be
analyzed.  For the purpose of establishing the ability of an instrument to
meet the continuous on-loop analysis requirements a total of 30 samples per
hour was taken as a nominal value.
     The sampling requirements, process characteristics and the need for
rapid analysis established the instrument requirements identified in
Table III.  During the early phases of the program, analysis error require-
ments for the key chemical constituents present in the limestone/dolomite
scrubber were based on estimates provided by the OAP Project Monitor of con-
centration range and relative error of the methods required for 20% sulfur
mass balance closure as determined by the Bechtel Corporation.  These data
are presented in Table IV and were used to guide the direction of the pro-
gram pending updating of these requirements in concurrent programs by the
Radian Corporation and Bechtel Corporation.  It is interesting to compare
the cost associated with analyzing this number of samples by alternative
                                     1057

-------
                               TABLE III
                        INSTRUMENT REQUIREMENTS
           •    Selective Ca +,  Mg++,  H+,  S03=,  SO^
           t    Continuous or plug flow analysis
           •    Minimum analytical  lag time
           •    Facile calibration
           t    Routine operation
           t    Rugged construction
           •    Low maintenance
           t    Acceptable accuracy
                                 TABLE IV
                       LIQUID ANALYSIS REQUIREMENTS
Concentration
Range mM
Mg++
Ca++
so3=
so4=
co3=
Na+
K+
cr
i
i
i
i
i
i
i
i
- 1000
250
150
500
20
500
500
500
Maximum Allowable
Relative Error*
3
3
3
15
15
15
15
15
*For 20% sulfur mass balance closure

laboratory procedures as opposed to on-line instrumentation.   Table  V lists
the estimated labor for laboratory analyses of sulfite,  calcium,  magnesium,
sulfate and pH and solids analysis for calcium, total  sulfur  and  magnesium.
The labor hours per sample are estimated to be between 2.1 and 5.2 hours.
The extrapolation of these values to an hour, daily, and monthly basis
show that a large amount of labor is required for minimum characteriza-
tion of a limestone/dolomite scrubber.   When the cost of this labor is
                                   1058

-------
                                 TABLE V
                 ESTIMATED LABOR FOR LABORATORY ANALYSES
Liquid Analyses
Sampling, Hours
S03=, (Titr), hrs
Ca , (AA) Hours
Mg++, (AA) Hours
S04~, (Grav) Hours
pH, Hours
Solids Analyses
Ca
S (X-ray) Hours
Mg
Total Labor Manhours/Sample
For 30 samples/hour, manhours
Per 8-hour day, manhours
Per 30 days, manhours
Minimum
0.1
0.2
0.2
0.2
0.7
0.1

0.6

2.1
63
504
15,120
Maximum
0.1
0.5
0.3
0.3
3.0
0.1

0.9

5.2
156
1,248
37,440
projected together with its necessary supervision, the use of automated on-
line analyses is readily justified on a cost saving basis alone without con-
sideration of advantages of reproducible sampling, calibration and sample
representation.
                      SLURRY SAMPLING AND SEPARATION
     During the course of this program several  vendors were contacted to
determine whether they had equipment available  which could separate a lime-
stone/dolomite slurry meeting the following operating parameters:
     •    Flow rate - to 300 Ib/minute (a portion of this flow could
          be diverted prior to the separator)
     •    Solids, % - 0.5 - 15
     0    Particle size, micron - 5 - 300
     •    Density of solids (unpulverized)  g/ml  - 2.7 -  2.9
     0    Density of liquid, g/ml - 1.005 - 1.080
                                    1059

-------
     •    Temperature, °F - to 150
     •    System to exclude air during and after separation -
          both phases
     •    Time to effective separation - 15 seconds
     Ten companies replied positively that they had equipment which might
fit these operating parameters.  The separation principles identified in-
cluded continuous discharge centrifuges, in-line filter cartridges, belt
filters, and a continuous cyclone cone centrifuge.   Laboratory evaluation
of these principles was undertaken using spent slurry obtained from the Key
West  Electric Company  and equipment  sold by deLaval, Sharpies and  Demco.  A
summary of the findings are shown in Table VI.  It  was  found that neither
                                 TABLE VI
          SUMMARY OF LABORATORY EVALUATION OF SEPARATION METHODS
     Continuous centrifugation - deLaval Laboratory Gyro-tester
     Performance - 30 sec. operation at 0.5 gpm feed - 3% Zurn slurry
     Results - very nearly clogged
     Cone centrifuge (cyclone) - Demco 18mm cone
     Performance - continuous - pretreatment device
     Results - very promising
     Solid bowl centrifuge/cone-Sharples solid bowl/Demco
     Performance - minimum one hour continuous operation
     Results - slight turbidity
     Polishing filter - Accu-flow in-line convoluted cartridge
     Performance - high capacity - quick interchange
     Results - optically clear output
the cone centrifuge nor a combination of the solid bowl  centrifuge-centri-
fugal  cone provided clear-cut separation as indicated by slight cloudiness
in the discharge fluids.  An optically clear  fluid would demonstrate excel
lent solids rejection and is needed for any subsequent colorimetric
                                     1060

-------
  INLET
                             OVERFLOW
characterization of the liquid phase.  However, inclusion of a polishing
filter, such as an Accu-Flow in-line convoluted cartridge filter downstream
resulted in a high capacity unit providing continuous transparent liquid
for periods as long as several hours depending on the initial solid loading.
     The cyclone cone separator was fabricated by Demco to meet TRW's design
requirements and is shown schematically in Figure 1.  The device consists
                                     of an 18-mm cone fabricated from 316
                                     stainless steel and possesses an ad-
                                     justable orifice control.  The unit
                                     operates with a 35 psi minimum pressure
                                     differential with an inlet feed velocity
                                     of 46 ft/second and a volume demand of
                                     1  gpm.  Throttling the underflow to
                                     cause an overflow to underflow ratio
                                     of 45, resulted in an overflow to under-
                                     flow solids content ratio of 0.0204.
                                     Consequently, operation of the Demco in
                                     this mode permitted rejection of approx-
                                     imately 98% of the original solids con-
                                     tent.  The solution containing 2% of
                                     the original solids is readily handled
                                     through   banked parallel polishing
                                     filters to provide optically clear
                                     liquid.  The life time of the filters
                                     are at least one hour and use of a
                                     parallel bank system permits back
                                     flushing to reactivate a spent filter
                                     when it is isolated from the flow loop.
                                          A continuous stage separation con-
                                     cept has been devised which is capable
                                     of achieving "instantaneous quenching
                                     of reaction" within an arbitrary allotted
                                     time of 15 seconds in such a manner as
                                     to present "dry" stream of slurry solids
                                   SEAL
                            •UNDERFLOW

Figure 2.  DEMCO Centrifugal Separator
                                      1061

-------
for continuous analysis.  In the conceptual  design shown in Figure 2,  the
first stage utilizes a liquid/liquid/solid centrifugal  separator,  such as
one of the deLaval PX solids ejecting centrifuges.  The separator  would be
fed by the slurry stream, from the point in the scrubber process under
scrutiny.  A second heavy liquid phase such as a Freon, trichlorethylene or
other heavy inert solvent would be added to the slurry  as it entered the
separator.  As shown in the schematic drawing, the light, clear aqueous
phase is separated from an annular zone near the center, the denser non-
aqueous phase is ejected from an intermediate zone while the solids, essen-
tially free from aqueous liquid contamination are continuously discharged
from the outermost zone and transferred to the quartz filter carrier belt.
In the filter/drying housing, which is the second stage, residual  inert sol-
vent is volatilized in a heated high pressure dry nitrogen stream  before the
solids pass into the solids analyzer.
                             ANALYSIS METHODS
     During the course of this program many alternative instrumental
methods were considered as candidates for adaptation to continuous on-line
analysis.  The details of these evaluations will be documented in  the final
report to Contract 68-02-0007.  A summary of our recommendations  is provided
in Table VII which identifies methods of analysis for liquid phase and solid
phase components.  The applicability of the recommended methods have been
confirmed through analysis of known simulated slurry mixtures and  actual
slurry mixtures and constituents obtained from operating wet limestone scrub-
bing units at the Key West Electric Company, Kansas Power and Light Company
and Shawnee Power Plant.  As can be seen from Table VII, considerable use is
made of X-ray and atomic absorption methods which handle samples with mini-
mum pretreatment.  Colorimetric methods have been recommended for  sulfur (IV)
content in solution and tentatively for nitrite and nitrate.  The  turbidi-
metric method for dissolved sulfate is a standard method which requires the
addition of barium salts after acidification and heating to remove carbon-
ate and bisulfite interferences.  Brief discussions are provided  below on
the continuous X-ray analysis methodology and the colorimetric dissolved
sulfur (IV) analysis method.  This information is provided because of the
special consideration and testing that were required to ensure acceptable
results for analysis in a limestone slurry environment.  The other methods
                                     1062

-------
x
*
                                               o
                                               l_>
=>
h-

**
                                               1
                                               0
                                               s
                                               o

                                               z
                                               13
                                               1/1
                                                                  O


                                                                  to
                                                                  oo
                                                                 -o
                                                                  C
                                                                  13
                                                                 g
                                                                 g.
                                                                 g
                                                                 cn
                                                                (T3

                                                                £
                                                                
                                                                S-
                                                                CD
      1063

-------
                                TABLE VII
            RECOMMENDED METHODS FOR ON-LINE ANALYSIS OF LIQUID
                 AND SOLID PHASE WET SCRUBBER COMPONENTS
 Liquid Phase
     •    S (IV) (HS03~ + SO,
     f
     t
Ca
Mg
Fe
K+
            ++
+ Fe
     •    NaT
     •    Total Solids
     t    pH
     •    C03=


     t    Total Sulfur
     t    N02"
     •    N03"
 Solid Phase
     t
     t
Total Sulfur
Ca
Mg
Fe
Si
Al
so2, co2
Colorimetric (Furfural Bleaching)
Atomic Absorption
Atomic Absorption
Atomic Absorption
Atomic Absorption
Atomic Absorption
Conductivity
Electrometry
Acidification, Heat + IR Determination
of Evolved Gas
Acidification, Heat  and Turbidimetric
X-ray (limit 9.4 mM)
Colorimetric* (Brucine)
Colorimetric* (Brucine)

X-ray
X-ray
X-ray
X-ray
X-ray
X-ray
Pyrolysis + IR Determination*
tentative projected method
are relatively straightforward and are considered routine by those experi-
enced in process control analysis and monitoring.
                         X-RAY ANALYSIS EQUIPMENT
     Commercially available analysis equipment was evaluated with the identi-
fication of the Applied Research Laboratories (ARL) process control  X-ray
quantometer (PCXQ) as being suited for continuous on-line analysis of
                                      1064

-------
selected species for both liquid, solid or slurry phases.   This  instrument
was evaluated for applicability of analysis of these mixtures by determina-
tion of synthesized simulated mixtures.  The equipment can be obtained with
a slurry presenter and can handle up to 15 slurry streams  sequentially in an
automative mode for elements from magnesium upwards in the periodic table.
Nine spectrometric channels of information are available and nine elements
in a slurry stream can be simultaneously detected and analyzed.   A bulk
density monitor is incorporated into the system along with a fixed external
standard.
     In the ARL unit the limit of detection for sulfur is  0.03%.  The sensi-
tivity of sulfur for on-line  slurry units utilizing helium X-ray path and
a Kapton cell window was determined to be significantly less than the 0.25%
absolute value considered to be the lowest reasonable value likely to be
encountered in the slurry mixture.  This detection limit corresponds to a
sulfur content of 9.4 mM in the liquid phase.  Consequently, X-ray cannot
be used for determining dissolved sulfate and sulfite concentrations which
total less than 9.4 mM.  The determined repeatability of the unit was 0.4%
relative, far less than the 3% which has been viewed as a  requirement.
     As expected, the key for obtaining good X-ray information is to estab-
lish elemental calibration curves using comparable matrix  materials which
will be present in the analysis sample.  This requires incorporation of both
limestone and flyash to ensure comparable matrices.  The effect  of particle
size on analytical accuracy is most striking when the size if greater than
40 microns.  However, the case of spent limestone/flyash solids, approxi-
mately 90%, have a particle size below 30 microns.  For the solids analyzed
from Key West Electric Company, Shawnee Power Plant and Kansas Power and
Light Company, more than 95% of the particles were less than 40  microns.
Consequently, the variability that can be introduced by particle size will
not play a significant role in the analysis of these mixtures.
     Comparison of the on-line analysis capability of the  ARL unit with re-
presentative laboratory analysis X-ray equipment shows that considerably more
analyses can be accomplished using the ARL PCXQ as is seen in Table VIII.
Extrapolation of these data to obtain operating costs show the ARL unit has
a 4.5-fold advantage over the best competitor oer analysis ($1.33 vs $6.00)
and a 3.4-fold cost per element analyzed advantage (18<£ vs. 62<£).
                                     1065

-------
                                TABLE VIII
         ELEMENTAL ANALYSIS CAPABILITIES OF CANDIDATE X-RAY UNITS


X-Ray
A.R.L.
A.R.L.
G.E. (
Kevex


Unit
(on-line)
(lab)
lab)
(lab)
Estimated Maximum
Number of Analyses
Per 8-hour Shift
240a
160a
160a
45b
Number of
Elements
Per Analysis
8
7
1
a-50 (16)
Total
Elemental
Analyses
^1800
'vlOOO
-v 160
* 640
aTwo-minute residence period in spectrometer
 Ten-minute residence period in spectrometer
                DETERMINATION OF DISSOLVED SULFUR DIOXIDE
     During the review of candidate analytical  methods for the determination
of dissolved sulfur dioxide (HS03~ and SO-~) it was determined that no sat-
isfactory methods existed for determining concentrations in the range to be
found in the limestone slurry mixture (see Table IV).   Consequently, a new
method based on bisulfite bleaching of the furfural UV absorption was de-
veloped to facilitate this analysis.  This method is based on the chemical
equation b in Equations 1 - 4 and depends on the bleaching of the 276 nM
absorption of furfural by reaction with bisulfite.
          C4H3OCHO  +  HS03"  t  C4H3OCHOHS03"                           (1)

                                        "  +  H+                        (2)

                                                                        (3)

                                                                        (4)
The absorbance, A, at 276 nM is directly related only to the amount of fur-
fural in solution when the pH of the media is maintained around 4.0 in ac-
cordance with the Lambert-Beer-Bouguer Law.
          A = abcp                                                      (5)
          where a  =  molar absorptivity of furfural
                b  =  optical path length
                Cp =  concentration of uncombined furfural

                                       1066
4. "^
_/


HS03- 1
/I 1 Ov/ o 1 ' ~
4.
^ H +
+
^ H+ +
4

HS03

S03~

-------
The equation of the bleaching reaction (Equation 1) is governed by the
formation constant, K
              cF[HS03~]
          where c. = concentration of furfural -sulfite adduct
                     (CA = co - CF}
                     "] = concentration of uncombined bisulfite
Combining Equations 5 and 6 results in a relationship of absorbance and bi-
sulfite ion as shown in Equation 7.
  =   _   [HSO "]  +  -—
A   abcQ       3        abcQ
                                                                        (7)
It is interesting to note that this method was first developed for the de-
termination of furfural and prior to this study has not been used for the
determination of bisulfite.  The reason for this is because in most situ-
ations colorimetric procedures are used for determining low concentrations
of chemical species but in the limestone scrubber case the concentration of
bisulfite (1 - 150 mM) is too large for trace analysis methods (without mass-
ive dilution) and not readily adaptable to common macro titrimetric proce-
dures (without using large volumes and dilute titrants).
     Detailed studies of the effect of pH, diverse ions,  temperature, and
time to constant color development has resulted in the selection of a single
reagent addition consisting of furfural, phosphate buffer and sulfamic acid
(to remove trace concentrations of nitrite interference).  The reproducibility
of the method has been determined to be better than 2% relative or 0.2mM
absolute whichever is higher.  This method is currently being adapted to a
plug flow analyzer system.
                     LABORATORY BENCH SCALE SCRUBBER
     A basic modular designed bench scale test loop wet scrubber (see
Figure 3) was fabricated to permit evaluation of the recommended methods
under simulated use conditions.  A loop system was selected because of the
necessity of: 1) closely approximating the full scale operating unit, 2) ac-
curate control, and 3) producing stable (equilibrium) and unstable
                                     1067

-------
              GASEOUS NITHOGEN
              MAKEUP CONNECTION
V]GAS RELIEF VALVE Y
GAS BLOCK VALVE 1
t"4
|| f"| ||ROTOMETERS
1 [ 1 ] |_J 0 - 5 CFM
V, V, Vi WASHOUT
£? Q* q? CONNECTION
(\ (l l\ ^V
I IsOj I koJ Io2 jr
^ 	 >
rr^—\2J*-m
10 CFM BLOWER,
W/ DRIVER
6P - 15" H,O MAX
1/3 HP
WASHOUT
CONNECTION
^
X ^ <0
) PISTON TYPE
— - -or:04 v " dh
MINOR GAS ADDITION FACILITY ~*~ - - c«rlP?-t - • , • -r

V-103
\
,T£O—
1/20 HP ELECTRIC STIRRER<^\
M-IOI vf
V
TANK HEATER WITH 	 ^3
THERMOSTAT (1KW) J
E-IOI 	 -J
(INSULATED)
_) t
^s V^
.X* SAMPLE /WASHOUT AL
P*^ CONNECTION
~~ O?«S
15 GALLON - 304 SS TANK
SOLIDS H
10 LB CAP
v-ic
	 TV" 1
V/ioi
Y TANK BLOCK VALVE
,... SAMPLE /WASHOUT J[
'•*•' CONNECTION f^>
VA,
_. STAR VALVE
r \ AR yj' DRIVER
\^_J 1 01

CARTRIDGE FILTER |
'ijo-
!02Vy
TANK HEATER WITH 13
THERMOSTAT (1 KW) ^J
E - 102 	
j
u
O,ri
LINES 1 "." SS TUBING
0.028" WALL
OPPER
ACITY
2
/ SLURRY FLOWMETER
CALIBRATION SY-PASS
til
PISTON TYPE
FLOW METER Q
EMERGENCY
BY -PASS VALVE
b"b

' / ELECTHIC STIRKER
/ M-102
+1
15 GALLON,
304 SSTANK
TANK BLOCK VALVE
»X. SAMPLE WASHOUT
~V*~ CONNECTION
OAR ,03
1 \MC
PROCESS FEED TANK y 	 K
T - 102
(INSULATED)
3-
-OlS,
H RUPTURE
DISC
3YNO VARIABLE SPEED PUMP
P-101
1 - .3 GPM
             Figure 3.  Bench Scale Scrubber Analysis Loop
(non-equilibrium) conditions for evaluating candidate instruments  under
known, controllable conditions with realistic  compositions.
     The system consists of a bench scale  Venturi  scrubber with  a  second
stage packed bed, fitted with a recirculating  gas  stream.  The  pressure
drop associated with the packed bed is  about 0.5-inch of water,  the  pressure
drop due to the Venturi is about 1-inch  of water  and  the pressure  drop
                                       1068

-------
associated with the ducting is 0.4-inch of water.   The  packed  bed  is  9-inch
deep and has a diameter of 4-inch.   The ducting is  2-inch  I.D.  throughout.
The Venturi has a throat size of 1-inch.
     The recirculating gas stream is moved via blower K-101.   The  composi-
tion of the recirculating gas stream is controlled  by the  Minor Gas Addition
facility.  This facility allows the addition of small  amounts  of gases  via
rotometers and bottled gas.  Gases  such as SO^, COp and 0~ are controlled  in
this manner.  Nitrogen is occasionally bled into the system to make up  that
amount which has been absorbed by the circulating  slurry.   The level  of
slurry in the liquid separator V-103 is controlled  in this manner.  The
composition of the gas stream is monitored by gas  analyzer AR-104, which
gives compositions of S0?, CCL and  0? in  the circulating stream.   Flow  of
the gas stream is given by a differential  pressure  cell, FR-101.
     The  liquid  slurry exits  the Venturi scrubber via the liquid separator
V-101.  The  temperature  in the downcomer is measured and recorded by
TR-105.   Analyses  of  the  slurry  is  also provided in the downcomer by
analyzer  AR-105  (type of  instrument  to be determined during Contract
68-020-0007.   The  liquid  stream  from the liquid separator dumps into a
15-gallon delay  tank, T-101,  where  it  is agitated with  a 1/20 horsepower
electric  laboratory stirrer  (M-101)  and the temperature is adjusted and
controlled  by  a  tank  heater  E-101.   The temperature is measured and re-
corded by TR-101.  The residence time  in this  delay tank is about one
hour with a  design slurry flow of 0.2  gpm.
     The  liquid  slurry travels to the  process  feed tank along one of two
routes.   It  can  travel along the straight transfer section, or it can be
diverted  through a filter.   The  purpose of the filter  is to take out
solids from  the  circulating  slurry.  The composition of the slurry exit-
ing the delay  tank is monitored  and  recorded by AR-101.

     The  solids  content of the slurry in Process Feed Tank (T-102),  a
15-gallon, 304 stainless  steel tank  equipped with a 1/2 horsepower
laboratory stirrer, is adjusted  by  adding limestone from the solids
hopper via a star  valve.  The composition of the tank is monitored and
                                    1069

-------
recorded by the process analyzer AR-102.  The temperature in this tank
is maintained by tank heater E-102 and is measured and recorded by
TR-102.
     The adjusted slurry from T-102 is transported along the transfer
line with a positive displacement pump, P-101.  This pump has the
capacity of 0.1 to 0.3 gpm.  This range is required so that the liquid to
gas ratio present in the packed bed Venturi scrubber is capable of being
changed.  The characteristics of this stream are given in Table VII.  Ac-
curate flow of the pump output is adjusted via the recycle stream to T-102.
The flow is monitored and recorded on FI-103 which receives a signal from
a positive displacement piston type flow meter.  The temperature in this
section of line is recorded on TR-103.  The  flow then  splits, part  going
through a counter-current flow section packed with 1/4-inch  Raschig
rings.  The other part of the flow goes to the throat of the Venturi.
The flow which goes to the packed section is measured and recorded on
a flow indicator FI-104, which receives its signal from a positive
displacement flow meter.
     All liquid lines present in the bench loop simulator are of 1/4-inch
polypropylene, with an  .028-inch wall.  Utilizing this type of tubing,
the flow velocity will be about 1.8-feet per second.
     This unit has been used to test the applicability of the recommended
methods under controlled conditions.  These current studies have confirmed
the fact that 1) the constituents of flyash catalyze the oxidation of sul-
fite to sulfate and 2) the presence of dissolved oxygen in the slurry (al-
though only 0.5 mM at 125°F) contributes significantly to sulfite oxidation.
                                 SUMMARY
     Methods have been  identified which are suitable for rapid sampling and
on-line chemical analysis of the principle constituents limestone/dolomite
wet scrubber solutions.  Utilization of the identified methods in process
demonstrations will permit rapid chemical changes in the scrubber slurry
as a function of process variables and provide needed basic information for
subsequent process optimization.
                                      1070

-------
                              ACKNOWLEDGMENT
     The authors wish to acknowledge the assistance of J.  Craig  (Zurn
Engineering), Lee Bruton (Kansas Power and Light Company), Jim Martin
(Combustion Engineering), and Joe Barkley (Tennessee Valley Authority)  for
their cooperation and assistance in acquisition of limestone scrubber slurry
and solid samples as well as Bob Statnick, OAP Project Monitor for his  guid-
ance and encouragement.  In addition, the authors wish to  acknowledge mem-
bers of the TRW Systems Chemistry and Chemical Engineering Laboratory for
their efforts on this program.
                                    1071

-------

-------
PARTICULATE EMISSIONS PROM TWO LIMESTONE WET  SCRUBBERS
                    Terry Smith
                    Ronald Draftz
         Walter C. McCrone Associates,  Inc.
                  493 East 31st Street
                Chicago, Illinois 60616
                   Prepared for
       Second International Lime/Limestone
              Wet Scrubbing Symposium
               New Orleans, Louisiana
                November 8-12, 1971
                          1073

-------
     PARTICULATE EMISSIONS FROM TWO LIMESTONE WET SCRUBBERS
                      Terry Smith and Ronald Draftz*
                   Walter C. McCrone Associates,  Inc.
                            493 East 31st Street
                         Chicago, Hlinois 60616
Abstract
       During field tests on a full-scale flooded bed scrubber and a pilot plant
scrubber, data were collected on the variation in mass emission levels and the
size distribution and chemical composition of the particle emissions.
       A filter sample and three runs with an Andersen stack sampler were
taken at the outlet of the flooded bed scrubber.  We found thai half of the  mass
emissions from this scrubber are smaller than 1. 6 /.'tn in diameter.  Using elec-
tron diffraction, we determined that 70% of these emissions are hydrated crystals
of calcium sulfate.
       Filter samples taken from the pilot plant scrubber showed that  the parti-
cles emitted from this scrubber are even smaller than those from the flooded
bed scrubber:  The mass average diameter is 0.8 /zm.  Here, again, a large
majority of the particle is calcium sulfate.
       The size,  shape and quantity of small calcium sulfate crystals indicate
that much of the emissions from limestone wet scrubbers are being produced by
the  evaporation of fine droplets  containing dissolved solids.
* Presented by Ronald G. Draftz
                                    1074

-------
Introduction
       During our program (1) to evaluate particulate sampling methods for wet
scrubbers, field tests were conducted on a full-scale flooded bed scrubber at
Kansas Power and Light (KPL) Lawrence, Kansas, and the Zurn pilot plant
Dustraxtor® scrubber at TVA's Shawnee plant.  Both scrubbers were operating
on coal-fired power plants.
       Although our primary task was to evaluate the effectiveness of various
methods of determining particle mass concentration and size distribution,  in the
outlet and inlet of wet scrubbers,  we were also interested in obtaining some data
on the  composition and size distribution of the particulate in the scrubber outlet.
We did not attempt to determine the exact mass loading from either unit.
Sampling Methods
       With the exception of the particulate collection device and the diameter of
the sampling probe, all samples were collected using a standard EPA sampling
train (2).  A summary of the  sampling conditions of the various particulate col-
lectors is shown in Table 1.
                                TABLE 1
                     Summary of Sampling Conditions
    Conditions
      KPL Flooded-bed
       scrubber outlet
                         Zurn pilot
                       scrubber outlet
particulate collector
sampling time (rain)
flow rate (SCFM)
collector temperature (° R)
probe diameter (in.)
filter
30
 3.68

 3/8
Andersen stack sampler
   1     4       10
   1.0   1.25    1.25
   760   760     760
   1/4   1/4     1/4
cyclone + filter
     45
      1.07
    695
      1/4
                                   1075

-------
       Tests at the outlet of the flooded-bed scrubber included collecting parti-
cles with glass-fiber filters, an improved Andersen Stack Sampler, and a 7-stage
cascade  impactor.  The Andersen Stack Sampler tests were designed to evaluate
particle  re-entrainment using three sampling times,  1, 4, and 10 minutes.  An
experimental cascaded cyclone and high-temperature membrane filter were used
for sampling the outlet of the Zurn scrubber.
       None of the samples were collected under isokinetic conditions because
the particulate matter encountered in these  tests is so small that particle size bias
was assumed to be negligible.  Since the samples were collected from only one
point they are not completely representative of the stack emissions.  However,
the samples from KPL were taken 24 inches inside one stack, and the velocity
profile was very flat, leading us to believe that the particulate was evenly dis-
tributed.  The outlet of the Zurn scrubber is only 8 in.  in diameter so that fixed
point sampling should be representative.

Analysis Methods
       Particulate collected on the glass-fiber filter at the flooded bed scrubber
was sized using an optical microscope in  conjunction with a Millipore IIMC auto-
matic image analyzer, set to measure Feret's diameter   Transmitted light  il-
lumination was used with a 100X oil-immersion objective on the microscope. Five
samples  of particulate were removed from different locations on the filter and
mounted  in glycerol on glass slides.  Glycerol was chosen because its refractive
index is  significantly different from those of the major components found in the
samples, calcium sulfate and glass spheres.  A total of 2,919 particles were
counted in 10 size intervals from 0. 2 fj.m to 4.5 nm..  Number fractions in each
interval were converted to mass fractions using the cube of the geometric mean
of the interval.
                                    1076

-------
        The size distributions of the particles collected by the Andersen Stack
 Sampler were obtained by first determining the fraction of the total mass col-
 lected on each stage.  The particulate collected on each plate was weighed to the
 nearest 0.01 milligrams using a semi-micro analytical balance.  The characteris-
 tic cut-off point,  d  ,  for each stage was calculated using the data supplied by
                                                                            3
 2000,  Inc., for operating conditions of 760°R and a particle density of 2. 5 g/cm
 (3). The cumulative distribution was produced by plotting the d   for each stage,
                                                            ou
 against the sum of the fractions collected below that stage.  More precise methods
 for determining size distributions have been reported which take into account the
 variation of d  with the size distribution (4, 5).  However, for our purpose, such
             O \}
 precision was unnecessary.
        The particles collected by the experimental cascaded cyclone  and filter
 were sized with a different method because of the extremely small size of the
 particle sampled. Particles from several samples removed from the cyclone and
 filter were photographed at magnifications of 10,OOOX and 5,OOOX repsectively,
 with a Cambridge Stereoscan IIA scanning electron microscope  (SEM).  The par-
 ticles were then sized  using an epidiascope attachment to the IIMC automatic
 image analyzer.  Again Feret's diameter was used and weight fractions were
 obtained in the manner described above.
        The fractional efficiency curve was obtained from the fractional mass
 distributions of the cyclone and filter catch.  The mass collection efficiency of the
 cyclone is defined as:

                       _  Mi - Mo _  (Me + Mo) - Mo
                           Mi            Me + Mo

where Mi, Me, and Mo are the inlet mass, mass of the cyclone  catch, and out-
let mass respectively.
                                    1077

-------
It is easy to show that the fractional collection efficiency of size X is given by
                                         C
                           E  =
                            x            F  (1 - K)
                                   C  +
                                     x       K

where C  and F  are the mass fractions at size X in the cyclone catch and out-
        xx
let or filter catch.
       Extensive electron diffraction analysis of individual particles from the
filter sample from the flooded-bed scrubber was performed using an RCA EMU 4
transmission electron microscope and later confirmed for the total sample using
x-ray diffraction analysis.   Elemental analysis of the cyclone samples were
performed using the SEM and an energy dispersive  analyzer for x-ray fluorescence
analysis.

Results
       The four particle-size distribution obtained at the flooded-bed scrubber
are shown in Figure 1.  The microscopically determined size distribution shows
that 50% of the mass emissions are smaller than 1.6 /urn in diameter.  Variations
in mass loading  in the stack resulted in the difference between the three distri-
butions obtained with the Andersen Stack Sampler.
       Figure 2 is a transmission electron micrograph of particles from the
filter sample.  The large cubic particle measures 0. 38 jim on a side, and the
numerous small particles of y-calcium sulfate and gypsum are less than 0.1 jum
in diameter.   The complete results of  the selected area electron diffraction analy-
sis are given in  Table 2.
                                    1078

-------
                                TABLE 2
           Approximate Composition of Particles Collected from a
                       Flooded-Bed Scrubber Outlet
          Chemical species                        Concentration

          CaSO, •  2H O                              ~ 70%
               4    2
          CaSO,                                    ~ 20%
               4
          CaCO                                     < 5%
               O
          Fe O and Fe  O                           < 5%
            o  4      2* o
          SiO  (glass spheres)                       < 5%
          CaO                                       < 5%
       The analysis of the cyclone sample collected from the Zurn scrubber
showed that 83% of the mass of particulate emitted from that scrubber are
smaller than 0. 74 p,m in diameter.  X-ray fluorescence analysis on these samples
indicated that the major elements are calcium and sulfur with only minor amounts
of silicon and iron,  which is similar to the particle composition of the flooded-
bed scrubber.

Conclusions
       Because short sampling times are necessary to avoid particle re-
entrainment, the Andersen Stack Sampler is not useful for determining particle
size-distribution from wet  scrubbers.
       The high concentration of small hydrated calcium sulfate crystals indicates
that much of the particulate emissions from limestone wet scrubbers are being
produced by evaporation of droplets containing dissolved calcium and sulfate ions.
However,  scrubber efficiency for solids seems very good  since very little flyash
was found in our samples.
                                   1079

-------
References
  This work is supported by Environmental Protection Agency contract
  EHS-D-71-25.
2
   Federal Register, Standards of Performance for New Stationary Sources,
   EPA, Volume 36,  Number 159, Part 11, 17 August 1971.
3
   2000, Inc.,  Instructions for Andersen Stack Sampler,  Salt Lake City, Utah.
4
  Kubie, G., A note on a treatment of impactor data for some aerosols,
  Aerosol Sci.  2,  23-30 (1971).
5
  Soole, B. W.,  Concerning the calibration constants of cascade impactors,
  with special reference to the Casella MK. 2, Aerosol Sci.  2, 1-14 (1971).
                                    1080

-------
            1081
(uir/)
9c-7

-------
FIGURE 2 Transmission electron micrograph of particles emitted
          from the wet scrubber, (50, OOOX).
                                 1082

-------
DESIGN CRITERIA FOR A SIZE-SELECTIVE  SAMPLER
        FOR LIMESTONE WET SCRUBBERS
                Terry Smith
                Hsing-Chi Chang
     Walter C. McCrone Associates, Inc.
             493 East 31st Street
            Chicago, Illinois 60616
               Prepared for
   Second International Lime/Limestone
          Wet Scrubbing Symposium
           New Orleans, Louisiana
            November 8-12, 1971
                    1083

-------
        DESIGN CRITERIA FOR A SIZE-SELECTIVE SAMPLER

                 FOR LIMESTONE WET SCRUBBERS


                                by
                  Terry Smith and Hsing-Chi Chang
                 Walter C. McCrone Associates,  Inc.
                         493 East 31st Street
                       Chicago, Illinois 60616
SUMMARY

       The reasons for and difficulties in developing a gravimetric size-selective

sampler for use with a limestone wet scrubber are outlined.  The expected gas

stream conditions at the sampling site are described.

       A parallel cyclone sampler which meets the design requirements is

discussed.  Methods of obtaining representative samples of particulates from

the gas stream, of accurately sizing the samples, and of determining the max-

imum number of size cuts which can be obtained have been developed for this

sampler and are described.  Also included is a discussion of how data from

the sampler can be used to determine the particle-size collection efficiency

curves for a scrubber.


INTRODUCTION

       Particle-size information is  useful in the design of all particulate col-

lection devices.  The effects of size distribution of limestone moeties on gas-

solid reaction kinetics  makes particle-size data vital to the development of

the limestone wet scrubber.  As reported in  a previous paper,  the only  com

mercially available size-selective sampler for use at stack conditions,  the


                                  1084

-------
Andersen stack sampler, does not perform satisfactorily.   As a result, the


development of a parallel cyclone sampler for use at the  EPA Alkali Scrubbing

                                   2
Test Facility at Shawnee was begun.
Conceptual Design of a Size-Selective Sampler


       A sampler should,  of course, provide accurate results and avoid the


particle  reentrainment problem encountered with the Anderson stack sampler.


It should also measure those segments of the particle-size distribution which


will provide information about the fractional removal efficiency of the wet


scrubber and the specific surface of the particulate.  The sensitivity and res-


olution of the sampler should be adequate to detect significant changes in the


particle-size distribution.


       A small cyclone followed by a filter meets these  requirements, and,


by using several of these in parallel, a gravimetric size-selective sampler is


obtained. The filter provides a stable low terr weight collection media upon


which the particulate  matter which passes the cyclone can be accurately weighed.


The cyclone thereby acts as a particle-size selector for the filter.


       Since a single sampling probe must transport an  unbiased sample of


particulate from the stack to the cyclone for fractionation, the first task in


designing the sampler is determination of the conditions  that produce a mini-


mum of sample bias by particulate deposition in the transpost tube.  Laboratory


experiments later confirmed during a field test, proved that dust  deposition



can be reduced to 2% by  mass with proper selection of transport tube diameter


and a transport nozzle having a radius of  curvature of 4  diameters.  Transpost



tube diameter should be  selected to obtain a Reynold's number of approximately



                                  1085

-------
15,000 for the tube.



       The next task in the design of the cyclone sampler is the determination



of the number of stages that can be used and the selection of the particle-size



cut-off point (d  ) for each stage.  Each stage of the sampler views one portion
               o U


of the particle-size spectrum: the number of stages that can be used,  then, is



limited by the line width, or resolution of each stage, the line width being the



uncertainty in knowing the collection efficiency of the device.  At best, the



cut points for the stages can be one line width apart over the entire size range.



A more reasonable spacing would be three line widths.



       Two factors affect the resolution of the collector:  the error in controlling



the collection parameters and the error in the calibration method.  It is



possible to estimate these errors from theoretical considerations and thereby



determine the approximate resolution of a cyclone.  By using the method of


                   3

sensitivity analysis  to analyze the mathematical prediction equation of the



cyclone collection efficiency, we found that reasonable errors in controlling



the collection parameters (flow rate, viscosity, temperature, etc.) lead to



variations in the cut point of 2-10%,  If a scanning electron microscope is



used for calibration, the  calibration error can easily be kept below 0.1 /urn.



The sum of the calibration error and the variance in the collection parameters



reduce the size resolution of a cyclone having a cut point of 1 p.m to a resolution



of  0.1-0.14/^m.  For larger cut points, the size  resolution is  limited by  the



variance of the collection parameters; for a cut point of  4.5 Mm  the size
resolution is about 0.14-0.49 fiin.  By applying these limitations to the size
distribution obtained at the outlet of the Kansas Power and Light Scrubber, some
                                   1086

-------
logical choice of the number of stages can be made.  Table 1 shows that six




cut points can be placed, three resolution elements apart, along the mass dis-




tribution.
                                TABLE 1




             Cut Points and Size Resolution of Cyclone Stages
Stage
1
2
3
4
5
6
Cut point
4.48
3.16
2.10
1.35
0.80
0.52
3 Resolution
elements (/nm)
1.47
0.96
0.70
0.5
0.39
0.3
% of particulate
mass below d50
90
75
60
39
15
3
% of particulat
area below dg0
98
90
86
65
35
10
        The outlet of the precollector functions as a gas manifold to divide the




gas  into seven branches.  Since the gas flow in the precollector outlet is a




vortex, the best aerodynamic method of dividing the flow is the use of 7




tangential outlets.  An inverted cone in the middle of the outlet manifold is




used to maintain the vortex motion all the way to the top of the  manifold.  By




incorporating the filter holders into the cover plates of the small cyclones,




particle losses due to deposition in the connecting tubing are  minimized.
                                   1087

-------
       When the sampler is used at the scrubber inlet where large particles




are present, a precollector to scalp large particles is used in front of the parallel




stages.  It is necessary to prevent large particles from entering the small




cyclone where gas velocities are high and particle bounce can lead to the escape




of large particles.  The cut point for the precollector then should be small




enough to remove most of the large particles but not so small that there is sub-




stantial overlap between the collection efficiency curve of the first stage and




that of the precollector.  We found that a cut point for the precollector of 6.75 Mm




meets the requirements.




       By addition of a filter as one of the parallel stages, the mass concentration




of the particulate can be determined and the total flow rate of the sampler can




be easily varied to maintain isokinetic conditions.







Methods of Designing Cyclones




       So that the parallel cyclones can be systematically designed,  an adequate




understanding of cyclone performance  is necessary.   A design  technique which




optimizes and adjusts all cyclone parameters has been developed based on the




work  of several German researchers.   ' '   This  design  method not only




predicts the performance of the cyclone but also adjusts the cyclone  geometry




so that a minimum amount of energy in the form of a pressure  loss is used to




collect particles of a given size.  The  result of  the optimization process is




reduced turbulence in the cyclone,  which should lead to sharp collection ef-




ficiency characteristics.
                                 1088

-------
       With a computer to aid in the calculations, three types of cyclones have




been designed for the sampler.  One type of cyclone serves for stages 1,  2, and




3 while another is used for stages 4, 5, and 6.  The third type is used as the




precollector.  We estimated that the flow rate through stage 6 would have to




be greater than  0.75 cfm  to obtain an accurately weighable sample of par-




ticulate in 30 minutes at the lowest mass concentration expected.  Table 2




shows the operating conditions for each stage and the precollector.









                               TABLE 2




                 Operating Conditions for Each Cyclone

Stage
1
2
3
4
5
6
precollector
Flow rate
(cfm)
0.475
0.675
1.00
0.400
0.650
0.875
5.025
Pressure drop
(in. of HO)
Lt
0.114
0.248
0.605
2.63
7.79
15.31
0.165
       Maintaining a constant cut point for each cyclone is a complex task




since the cut point depends on the gas velocity, density,  viscosity,  and wall




friction in the cyclone.  In practice it has been found that as particulate de-




posits on the walls of the cyclone, the pressure loss in the cyclone drops,




                                  1089

-------
indicating a reduction in gas velocity.  This leads to the idea that perhaps the




best way of maintaining a constant cut point would be maintaining a constant




pressure loss in the cyclone.  Therefore the pressure drop across each stage is




monitored by magnihelic differential pressure gages.




       A conceptual design for the entire  sampling train shows in Figures 1 and 2.




The sampling train consists of three units: the sample box which contains the




cyclone, a control unit which contains pumps, and gas metering equipment and




a cooling supply system for the water vapor traps.






Results of Preliminary Testing




       A cyclone having near optimum geometry was designed and constructed.




A field test on a pilot-plant wet scrubber, discussed in a previous paper, was




carried out to determine the performance of  the cyclone.  We had calculated




that for a flow rate of  0.8 cfm the small cyclone would have  a pressure drop




of 27 in water; however, we found its actual drop to be 17 in. water, indicating




that the frictional losses in the cyclone walls had been underestimated.  The




design equations also predicted that 50% collection efficiency would occur at




0.44-Mm particle diameter for a 1-cfm  flow rate. As is seen in Figure 3,




the actual 50% collection efficiency occurred for  0.74 Mm.  This is, again,




due to the increased wall losses causing reduced velocities and collection ef-




ficiency in the cyclone.




       A measure of the steepness of the  collection efficiency curve for a




device is given by the geometric standard derivation of the collection efficeincy




S where
                                 1090

-------
                     Particle diameter at 50% efficiency

                   Particle diameter at 84.13% efficiency




We are gratified to note that the optimization procedure for the cyclone design



produced a very steep efficiency curve having a geometric standard deviation


                                                                    789
of 0.94. This surpasses the performance reported for several cyclones  '  '


                               10
as well as for inertial impactora
Conclus ions
       Undoubtedly a parallel cyclone sampler can be built using the design



techniques developed thus far.  The  inability to accurately predict performance



merely means that emperical calibrations methods will have to be used.  Before



better predictions can be made, however, a better understanding of friction



losses in cyclones will have to be obtained.
                                   1091

-------
Footnotes


 Smith, T.M., and R.G. Draftz,   Particulate Emissions from two limestone wet
 scrubbers,  delivered at 2nd International Symposium on Limestone Wet Scrubbers
 New Orleans,  La.  (8-12 November 1971).

2
 This work is supported by EPA contract EHS-D-71-25.

g
 Schenck, H., Theories of Engineering Experimentation, pp. 50-51,  McGraw-Hill,
 New York,  1968.

4
 Barth, W.,  Calculation and design of cyclone separators on the basis of recent
 investigations,  Brennstoff-Warme-Kraft 8, 1-9 (1956).

g
 Muschelknautz, E., Design of cyclone separators in the engineering practice,
 Staub-Reinholdt Luft 30,  1-12 (1970).

/»
 Muschelknautz, E., and W. Krambrock, The aerodynamic coefficients of the
 cyclone separator as based on recent, improved measurement, Chem,-Ing.-Tech.
 42, 247-255 .(1970).

17
 Statrmand,  C. J.,  The design and performance of cyclone separators, Trans.
 hist.  Che IT. Engrs. 29, 356-383 (1951).

g
 Lipman, M., and A. Kydonieus, A multistage aerosol sampler for extended
 sampling intervals, Am.  hid. Hyg. Assoc. J., 730-7 (1970).

g
 Freudenthal, P.,  High collection efficiency of the Aerotec-3 cyclone for
 submicron particles, Atmos.  Environ.5, 151-4 (1971).


 Cocchman, J. C., and H. M.  Moseley, Simplified method for determining
 cascade impactor stage efficiencies, Am. Lid. Hyg. Assoc. J., 62-67 (1967).
                                 1092

-------
1093

-------

-------
                      FIGURE 3  Size distribution of TVA samples collected
                                  with small cyclone
»0   .1

-------

-------
INSTRUMENTAL METHODS FOR FLUE GAS ANALYSIS
       R.M. Statnick and J.A. Dorsey

       Process Measurements Section
         Control Systems Division
      Environmental Protection Agency
         For presentation at
 Second International Lime/Limestone
        Wet Scrubbing Symposium
         New Orleans, Louisiana
          November 8-12,  1971
                    1097

-------
The application of continuous monitoring instrumentation to
pilot plant and fuel scale evaluations of control technologies
require careful consideration of the effects of the source and
sampling system on overall accuracy of the measurement.  In
general, to monitor the mass flow rate of pollutants in flue
gas, the following three problem areas must be considered.

          1.  Sample Acquisition
          2.  Sample Handling
          3.  Instrument Selection

1.  Sample Acquisition;

    The major problem in the precision determination of pollutant
    mass flow rate (Ibs/hr) is the variation of the species
    concentration which may exist spacially in a large duct as
    a result of air infiltration and poor mixing (stratification).
    In the course of OAP's extensive studies of coal fired power
    plant effluents, a large number of carbon dioxide concentration
    profiles within large ducts were obtained at various sampling
    locations.

    Typical sampling locations are illustrated in Figure 1 for a
    power plant.  The two most common sampling locations are at
    the inlet and at the outlet of control equipment.  Infiltration
    of air generally occurs within and post the air pre-heater.
    In Figure 2 and Table I, examples of homogeneous and stratified
    carbon dioxide profiles are shown.  As can be seen in Figure 3,
    the probability of determining the concentration of a species
    within 15% with a single probe is fifty percent.  With nine
    probes, one achieves a 99+% probability of determining the
    pollutant concentration within 15%.

    Stratified flow in pilot plant operations can be avoided by
    installation of gas mixing equipment such as venturi, perforated
    plate, vanes, etc. at a cost of about $500-$1500 @ 3000 acfm.
    This expedient will assure the most accurate measurement
    possible of the concentration of the pollutant with a single
    point sample probe.  To determine the total gas flow, a gas
    metering venturi is acceptable.

    At a full scale power plant, it is unrealistic to modify the
    plant; therefore, the mass flow rate can best be determined by
    careful selection of the sampling site and verification of
                               1098

-------
    sampling conditions by complete characterization of the
    velocity profile and determination of whether stratification
    exists.

    The best sites for pollutant concentration and determination
    of the velocity profile are not necessarily identical; one
    approach which can be used is:

    a.  Determination of SC>2 or NOx concentration along with the
        CC>2 concentration post the economizer but prior to the
        air preheater.

    b.  Determination of the velocity profile at a location move
        amenable to a velocity traverse.  This will yield the
        total gas volumetric flow at the location.

    c.  The volumetric flow of gas at the economizer is given by
        volumetric flow at economizer = volumetric flow (from
        velocity data) x CO? (at traverse)
                         CC>2 (at economizer)

2.  Sample Handling

    Having selected and evaluated a site, the next consideration
    is extraction of samples.

    Probes;

    The sampling probe design is dictated by temperature.   It
    can be as simple as a 1/2-inch O.D. stainless steel tube for
    temperature 350°P and above,  or as complex as a shielded in
    the stack filter-probe combination for temperatures below
    350°F to minimize reactive losses of SC>2

    Filters;

    All instruments which utilize optical principles to determine
    the pollutant concentration require the removal of particulate
    matter.  Particulate matter is removed by passing the  particulate
    laden-gas stream through a positive filter.   A silicon carbide
    filter which has a 90% collection efficiency at 5/1  have been
    found to be practical.   These filters can be mounted internal
    or external of the ducting; since the average working  life of
    the filters is 3-4 weeks (at 5-7 grains/scf  and 2 cfh  flow),
    the external stack filter is recommended for ease of replacement
                              1099

-------
of the filter and minimum down time.  The filter assembly
should be maintained at 300-350°F to eliminate reactive S02
losses with the filtering media.

A potential problem particular to wet scrubbers might be
observed post the device.  If there is substantial liquid
re-entrainment and poor mist eliminator efficiency, the
saturated scrubber liquor droplets will enter the probe and
be collected on the filtering media.  At 300-350°F, evaporation
of the water will leave a residual deposit of CaSC>4, CaSC^J^O,
and CaCOs platlets which will plug the filter.  Filter pluggage
was observed during manual particulate sampling at Kansas
Power and Light.  It will also occur using the filters describe
above; to reduce the probability of this type of pluggage, high
turbulence in the probe to promote droplet evaporation is
recommended.

SamplingLines and Water Vapor Condensors

The sampling lines can be constructed of heat traced teflon
or stainless steel.  The sampling lines should be maintained
above the dew point of the flue gas, 300-350°F.  Stress
corrosion has been observed in the sampling lines  (304 stainles
steel) used at the TVA dry limestone injection tests.  The need
for frequent replacement of the stainless steel tubing makes th
heat traced teflon, although initially more expensive, the most
desirable material of construction.  The use of rubber on PVC
tubing is not recommended since absorption of SC>2 on tube walls
occurs in these materials.

For those who select non-dispersive infrared (NDIR) as the
pollutant monitor, a condensing system is required to remove
water vapor which is a positive interference.  A sample NDIR
has a rejection ratio of 100; that is, 100 ppm of water vapor
yields a signal equivalent to 1 ppm of SC>2 •  A condenser held
at 0° .+ 1°C will contribute a _+ 4 ppm signal of SC>2 equivalent
water vapor.  At 1000 ppm SO2 or greater, typical of the
scrubber inlet this error is insignificant, but with 90% contro
of a 1000 ppm inlet S02 concentration, it will yield a ± 4%
error in the SO2 level at the outlet.

Gas Pumps

Either bellows type or other leakless air moving pump is
acceptable.  The pump is located prior to the detection system.
                           1100

-------
    Response Time

    All of the components of a sampling system have been covered,
    these elements should be so constructed such that the desired
    system response time is achieved.  The system response time
    is defined as:

         The time interval from a step change in pollutant
         concentration at the probe inlet to a recording of
         90% of the ultimate recorded output.

    By suitably adjusting the volume of the sample handling
    system and/or the volumetric flow through the sampling system
    a wide range of system response times are achievable.  For a
    typical NDIR sampling analysis system, the sample handling
    system has a total volume of 0.5 cu ft (assuming 100' 1/2-inch
    I.D. tubing, filter, and 0.3 cu ft cooler volume).  The
    response time of the system, assuming plug flow and 2 cfh
    pumping capacity, normally supplied with instrument, is 15
    minutes.

    In Figure 4, a high volume pump about 20 cfh is used to
    extract the sample and draw it through the filter and cooler.
    The water and particulate free stream is then sampled by the
    2 cfh pump.  This will reduce the system response time to 1.5
    minutes with a sample system of 0.5 cu ft.

3.   Instrumentation

    The Office of Air Programs had funded field evaluation of
    commercially available sulfur dioxide and nitrogen oxide
    monitors.  These studies were conducted at coal fired power
    plants; the sulfur oxide study at a steam station burning
    0.5% sulfur coal; the nitrogen oxides study at a coal fired
    station with approximately 200 ppm NOx emissions.  Preliminary
    results of these studies are shown in Table II.

    The table also shows that for sulfur dioxide NDIR's or DNUV1s are
    of comparable accuracy and reliability.  For NO, NDIR; for
    NO2 NDUV; presently only NDUV for total NOx; however, this is
    batch operation and the 10 minute reactor time must be added
    to system response time.  In addition to the instrumentation
    described above, flue gas analysis could be performed by gas
    chromatography, mass spectroscopy, dispersive infrared,  etc.
    None of these are presently available as commercially tested
    units.
                              1101

-------
    Instrument types, found to be effective for data collection
    for engineering analysis for control systems development of
    SC>2 and NOx flue gas control equipment, include:

         S02—NDIR and NDUV,
         NO —NDIR and NDUV, and
         NOx—NDUV.

    The NDUV is sensitive to NO2 and the commercially available
    instrument provides integral catalytic oxidation of NO to NO2.

Overall Conclusions

The overall conclusions which can be reached are:

    1.  Extreme care must be exercised in the choice of sampling
        location and each location should be completely characteriz

    2.  The probe, filter, and sampling lines must be held above
        the dew point of the gas stream.

    3.  Using vendor supplied sample conditioning equipment and
        gas pumps, system response times are greater than 10 minute

    4.  For sulfur dioxide monitoring, NDIR's or NDUV's are
        effective.

    5.  For nitrogen oxides, NDUV is effective if 10 minute cycle
        times are acceptable; for continuous analysis NDIR's are
        effective  (note: total nitrogen oxides will be NO; NDIR's
        are insensitive to NO2).
                                1102

-------
                                                               Low level
                                                              economizer
                                  Sampling points  B
                                                   Sampling points A
                                                                    Burners
Figure 1  Boiler  Outline for Corner-Fired  Unit Showing Sampling  Posit

                                   1103
ions

-------
                        HOMOGENEOUS
1.00
1.03
1.00
1.00
1.01
1.01
0.98
0.98
0.98
1.00
1.00
1.01
1.01
1.01
1.01
t
3 '3"
if
A ??' kl
                      Avg C02 = 11.7%


                          CV  =  1.9%
                        STRATIFIED
0.
0.
0.
0.
74
82
75
82
0
1
1
1
.92
.01
.03
.00
1.
1.
0.
0.
00
02
99
93
1.
1.
1.
0.
00
01
02
93
0
1
1
1
.97
.00
.01
.01
0.
0.
0.
0.
95
90
90
85
                                                    t
                                                    4'8"
                                                    I
                      Avg CO2 = 12.6%


                          CV =   9.3%





Random selection from six plants.
Figure 2.  Normalized CO2 Traverse Data at Dust Collector  of
           Coal-Fired Power Plants.
                            1104

-------
TABLE  I.   OBSERVED COEFFICIENT OF VARIATION FOR  CO2  TRAVERSE
                FOR VARIOUS COAL-FIRED PLANTS
Plant
No.
1
2
3
4

5

Type of
Boiler Firing
Horizontally opposed
Cyclone
Spreader stoker
Corner

Vertical

Dust
Collection
Equipment
C
E
C
C, E

C, E

Sampling
Location
I
O
I
0
I
0
I
0
I
0
No. of
Traverse
Points
24
12
24
24
18
9
18
12
24
12
co2
%
9.3
2.3,
4.6
3.2
1.5
1.02
8.8
0.97
7.1
3.2
(CV)
1.4






]=Cyclone
]=Electrostatic precipitator
:=Dust  collector inlet
)=Dust  collector outlet
                              1105

-------
      10
                        PROBABILITY FOR
                         3 PROBES ACROSS
                         CENTER OF DUCT*
                  PROBABILITY FOR A
                  •SINGLE PROBE IN THE
                  CENTER OF THE DUCT
        01    2345678
            NUMBER OF PROBES UTILIZED
Figure 3.   Probability of Obtaining an Accuracy Within
           15% of 9-Point Analysis for 02 in a Large Duct
                           1106

-------
PROBE

mt^m
FII

/TER

(


CONDENSER
1

2(
_JO_
2:
20 CFH
                                2 CFH
                               MONITOR
     Figure 4.  Fast-Response Sampling System.
                         1107

-------
          TABLE II.   SULFUR DIOXIDE MONITORS
Detection Principle

NDIR
NDUV
Conductrometric
Couldmetric
Ele c tr ochemlca-1
Instrument ^ '
Response Time
Good
Good
Poor
Good
Good
Reliability(2)
MTF
351 hrs.
322 hrs.
67.1 hrs.
569 hrs.
Poor
Accuracy

Good
Good
Good
Good
Good
                      Nitrogen Oxide Monitor
Detection Principle

NDIR
NDUV
Electrochemical
Response
Time
Good
Good
Good

Reliability, MTF- (2)
Very Good(3)
Very Good^3^
PrtrtT


Accurac'
Good
Good
Good

(1)   Time interval  from a step change in the pollutant in concentration
     at the instrument inlet to a recording of 90% of the ultimate recorded
     output (>3 sec.).

C2)   Mean time between failure

(3)   No failures during test
                               1108

-------
EPA RECOMMENDED SOURCE TEST METHODS FOR NEW  SOURCE
          PERFORMANCE STANDARDS TESTING
                  Gene W. Smith
           Applied Technology Division
         Environmental Protection Agency
                 Prepared for
     Second International Lime/Lime stone
            Wet Scrubbing Symposium
             New Orleans, Louisiana
              November 8-12, 1971
                          1109

-------
     EPA RECOMMENDED SOURCE TEST METHODS FOR NEW SOURCE
              PERFORMANCE STANDARDS TESTING
     The text of this paper consisted of an explanation of

the Standards of Performance for New Stationary Sources

proposed by the Environmental Protection Agency and established

by the Clean Air Act as Arranended.  Gene Smith used the Federal

Register, Vol. 36, No. 247—Thursday, December 23, Part II

as his reference material, which he handed out during the

symposium.
                           1110

-------
                  THURSDAY, DECEMBER 23, 1971
                  WASHINGTON, D.C.

                  Volume 36 • Number 247


                  PART II
                  ENVIRONMENTAL
                     PROTECTION
                       AGENCY
                  Standards of Performance for
                    New Stationary Sources
No. 247—Pt. H	1
                  mi

-------
24876
     RULES AND REGULATIONS
   Title  40—PROTECTION OF

           ENVIRONMENT

Chapter  I—Environmental  Protection
               Agency
      SUBCHAPTER C—AIR PROGRAMS
PART 60—STANDARDS OF PERFORM-
   ANCE   FOR  NEW   STATIONARY
   SOURCES
  On August  17, 1971 (36 F.R. 15704)
pursuant to section 111 of the Clean Air
Act  as  amended, the   Administrator
proposed standards of performance for
steam  generators,  Portland  cement1
plants, Incinerators, nitric acid plants,
and  sulfuric acid  plants.  The proposed
standards, applicable to sources the con-
struction or modification  of which was
initiated after August 17,  1971, included
emission limits for one or more of four
pollutants   (particulate matter,  sulfur
dioxide,  nitrogen  oxides, and sulfuric
acid mist) for each source category. The
proposal included  requirements for per-
formance testing,  stack gas monitoring,
record keeping and reporting, and pro-
cedures by which EPA will provide pre-
construction review and  determine the
applicability of the standards to specific
sources.
   Interested parties were afforded an
opportunity to participate  in the rule
making by submitting comments. A total
of more than 200  interested parties, in-
cluding Federal, State, and local  agen-
cies, citizens groups, and commercial and
Industrial organizations submitted com-
ments. Following  a review  of the pro-
posed regulations  and consideration of
the  comments, the regulations, includ-
ing the appendix,  have been revised and
are being promulgated today. The prin-
cipal revisions are described below:
   1. Particulate   matter   performance
testing procedures have been revised to
eliminate the requirement for impingers
in the sampling train. Compliance will be
based only  on material collected in the
dry  filter and the probe  preceding the
filter. Emission limits have been adjusted
as appropriate to reflect  the change in
test methods. The adjusted standards re-
quire the same degree of particulate con-
trol as the originally proposed standards.
   2. Provisions have been  added whereby
alternative  test methods can be used to
determine compliance. Any person who
proposes  the  use  of an  alternative
method will be obliged to  provide evi-
 dence that  the  alternative  method is
equivalent to the reference  method.
   3. -The definition of modification, as it
pertains to increases in production rate
and changes of fuels, has been clarified.
Increases in production rates up to design
capacity will not be considered a modifi-
cation nor will fuel switches if the equip-
ment was originally designed to accom-
modate such fuels. These provisions will
eliminate inequities where equipment had
been put into partial operation prior to
the proposal of the standards.
   4. The definition of a new source was
 clarified to include construction  which
is completed within an organization as
well as  the more common situations
•where the facility is designed and con-
structed by  a contractor.
  5. The provisions regarding requests
for EPA plan review and determination
of construction or modification have been
modified to emphasize that the submittal
of such requests and attendant informa-
tion is  purely  voluntary. Submittal of
such a request will not bind the operator
to supply further information; however,
lack of sufficient information may pre-
vent the Administrator from rendering
an opinion. Further provisions have been
added to the effect that information sub-
mitted voluntarily for such plan review
or determination of applicability will be
considered confidential, if the owner or
operator requests such confidentiality.
  6. Requirements for notifying the Ad-
ministrator prior to commencing con-
struction have been deleted. As proposed,
the provision would have required notifi-
cation prior to the signing of a contract
for construction of a new source. Owners
and operators  still  will be  required  to
notify the Administrator 30 days prior to
initial  operation  and to  confirm the
action within 15 days after startup.
  7. Revisions  were incoporated  to per-
mit compliance testing to be deferred up
to 60 days after achieving the maximum
production  rate but no longer than 180
days after initial  startup. The proposed
regulation  could  have required  testing
within  60 days after startup but  defined
startup  as  the  beginning  of  routine
operation. Owners or operators  will  be
required to notify the Administrator at
least 10 days prior to compliance testing
so that an EPA observer can be on hand.
Procedures  have been modified  so that
the equipment will have to be operated
at  maximum expected production rate,
rather than rated capacity, during com-
pliance tests.
   8. The criteria for evaluating perform-
ance testing results have been simplified
to  eliminate  the requirement that all
values be within 35 percent of the aver-
age. Compliance  will be based  on  the
average of three repetitions conducted in
the specified manner.
   9. Provisions were  added to  require
owners or operators of affected facilities
 to  maintain records of compliance tests,
 monitoring equipment,  pertinent anal-
 yses, feed rates, production rates, etc. for
 2 years and to make such information
 available on request to the Administra-
 tor. Owners or operators will be required
 to summarize the recorded data daily
 and to convert recorded data into the
 applicable units  of the standard.
   10. Modifications were  made to  the
 visible  emission  standards for steam
 generators, cement plants, nitric acid
 plants,  and sulfuric  acid  plants. The
 Ringelmann standards have  been  de-
 leted; all limits will be based on opacity.
 In every case, the equivalent opacity will
 be at  least as  stringent as the proposed
 Ringelmann number.  In  addition, re-
 quirements have  been altered for three
 of the source categories so that allowable
 emissions will be less than 10  percent
 opacity rather than  5 percent or  less
 opacity.  There  were  many comments
that  observers  could  not  accurately
evaluate emissions of 5 percent opacity.
In addition, drafting errors in the pro-
posed visible emission limits for cement
kilns  and steam generators were cor-
rected. Steam generators will be limited
to visible emissions not greater than 20
percent opacity and cement kilns to not
greater than 10 percent opacity.
  11.  Specifications  for monitoring de-
vices  were clarified,  and directives for
calibration were  included. The instru-
ments are to be calibrated at least once
a day, or more often if specified by the
manufacturer.  Additional  guidance  on
the selection and use of such instruments
will be provided at a later date.
  12.  The requirement for sulfur dioxide
monitoring  at steam  generators  was
deleted for  those  sources which  will
achieve the standard by burning low-sul-
fur fuel, provided that fuel analysis is
conducted and recorded daily. American
Society  for  Testing  and  Materials
sampling  techniques are  specified for
coal and fuel oil.
  13.  Provisions were added to the steam
generator standards to cover those In-
stances where  mixed fuels are burned.
Allowable emissions will be determined
by prorating the heat input of each fuel,
however, in the case of sulfur dioxide, the
provisions allow operators the option of
burning  low-sulfur  fuels   (probably
natural gas) as a means of compliance.
  14.  Steam generators fired with lignite
have  been exempted from the nitrogen
oxides limit. The revision was made in
view of the lack of information on some
types of lignite burning. When more in-
formation is developed, nitrogen oxides
standards may be  extended  to lignite
fired  steam  generators.
  15.  A provision was added to make it
explicit that the  sulfuric acid plant
standards will not  apply to scavenger
acid plants. As stated in the background
document, APTD 0711, which was issued
at the time the proposed standards were
published, the standards were not meant
to apply to such operations, e.g., where
sulfuric acid plants are used primarily
to control sulfur dioxide or other sulfur
compounds  which  would  otherwise be
vented into the atmosphere.
   16. The  regulation  has been  revised
to provide that all materials submitted
pursuant to these regulations will be di-
rected to EPA's  Office of General En-
forcement.
   17. Several other  technical  changes
have  also been made. States and inter-
ested parties are urged to make a careful
reading of these regulations.
  As  required by section 111 of the Act,
the standards  of performance  promul-
gated herein "reflect the degree of emis-
sion  reduction which (taking into ac-
count the cost of achieving such reduc-
tion)  the Administrator determines has
been   adequately demonstrated".  The
standards of performance are  based on
stationary source testing conducted by
the Environmental  Protection Agency
and/or contractors and on data derived
from various other sources, including the
available technical literature. In the com-
ments on the proposed standards, many
questions were  raised  as  to  costs and
                              FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
                                                           1112

-------
                                               RULES AND  REGULATIONS
                                                                                                                     24877
demonstrated  capability of control sys-
tems to meet the standards. These com-
ments have  been evaluated and investi-
gated,  and  it  is  the  Administrator's
judgment that emission control systems
capable  of meeting  the standards have
been adequately demonstrated.and that
the  standards  promulgated herein  are
achievable at reasonable costs.
  The regulations establishing standards
of performance for steam generators, in-
cinerators, cement  plants,  nitric  acid
plants, and sulfuric acid plants are here-
by promulgated effective on publication
and apply to sources, the construction or
modification of which  was commenced
after August 17, 1971.

  Dated: December  16, 1971.

       WILLIAM  D.  RTJCKELSHATJS,
                     Administrator,
    Environmental Protection Agency.

  A  new  Fart  60 is added to  Chapter I,
Title 40, Code  of Federal Regulations, as
follows:

        Subpart A—General Provisions
Sec.
601   Applicability.
60.2   Definitions.
60.3   Abbreviations.
60.4   Address.
605   Determination  of  construction  or
        modification.
60.6   Review of plans.
60.7   Notification and recordkeeping.
60.8   Performance tests.
60.9   Availability of information.
60 10 State authority.

   Subpart D—Standards of Performance for
       Fossil Fuel-Fired Steam Generators
60.40 Applicability and designation of af-
        fected facility.
60.41  Definitions.
60.42 Standard for particulate matter.
6O.43  Standard for sulfur dioxide.
60.44 Standard for nitrogen oxides.
60.45  Emission and fuel monitoring.
60.46  Test methods and procedures.

   Subpart E—Standards of Performance for
                Incinerators
60.50 Applicability and designation of af-
        fected facility.
60.51  Definitions.
60.62  Standard for particulate matter.
60.53  Monitoring of operations.
60.54 Test methods and procedures.

   Subpart F—Standards of Performance for
           Portland Cement Plants
60 60  Applicability and   designation   of
        affected facility.
60.61  Definitions.
60.62  Standard for particulate matter.
60.63  Monitoring of operations
60.64  Teat methods and procedures.

Subpart G—Standards of Performance for Nitric
                Acid Plants
60.70 Applicability and designation of af-
        fected facility
60 71  Definitions.
60 72  Standard  for nitrogen  oxides.
60.73  Emission monitoring.
60.74  Test methods and procedures.

Subpart H—Standards of Performance for Sutfuric
                Acid Plants
60.80  Applicability and designation of af-
        fected facility.
60.81  Definitions.
Sec.
60.82
60.83
60,84
60.85
Standard for sulfur dioxide.
Standard for acid mist.
Emission monitoring.
Test methods and procedures.

  APPENDIX—TEST METHODS
Method 1—Sample and velocity traverses for
      stationary sources.
Method 2—Determination of stack gas veloc-
      ity and volumetric flow rate  (Type S
      pitot tube).
Method 3—Gas analysis  for carbon dioxide,
      excess  air, and dry molecular weight.
Method 4—Determination  of moisture in
      stack gases.
Method 5—Determination  of   parttculate
      emissions from stationary sources.
Method 6—Determination of sulfur dioxide
      emissions from stationary sources.
Method 7—Determination of  nitrogen oxide
      emissions from stationary sources.
Method 8—Determination  of  sulfuric  acid
      mist and  sulfur  dioxide  emissions
      from stationary sources.
Method 9—Visual determination of the opac-
      ity  of emissions  from stationary
      sources.
  AUTHORITY: The provisions  of this Part 60
issued under sections 111, 114, Clean Air Act;
Public Law 91-604, 84 Stat. 1713.

   Subpart  A—General Provisions

§ 60.1   Applicability.
  The provisions of this part apply to
the owner or operator of any stationary
source, which contains an affected facil-
ity the construction or modification of
which is commenced after the date of
publication in this part of  any proposed
standard applicable to such facility.
§ 60.2  Definitions.
  As  used in this part, all terms not
defined herein shall have  the meaning
given them  in the Act:
  (a)  "Act"  means  the Clean  Air Act
(42 U.S.C. 1857  et seq., as amended by
Public  Law  91-604, 84 Stat. 1676).
  (b)  "Administrator" means the  Ad-
ministrator  of the Environmental Pro-
tection Agency or his  authorized repre-
sentative.
  (c) "Standard"  means a standard of
performance proposed or promulgated
under  this part.
  (d)  "Stationary  source" means  any
building, structure, facility, or installa-
tion which emits or may  emit any air
pollutant.
   (e)  "Affected  facility"  means, with
reference to  a stationary source, any ap-
paratus to which a standard is applicable.
  (f) "Owner or operator"  means any
person who  owns,  leases, operates, con-
trols, or  supervises an affected  facility
or a stationary source  of which an af-
fected facility is a part.
  (g) "Construction" means fabrication,
erection,  or  installation of an  affected
facility.
  (h) "Modification" means any physical
change in, or  change in the method of
operation of. an affected facility  which
increases the  amount of  any  air pol-
lutant  (to which  a standard  applies)
emitted by such facility or  which results
in the emission of any air  pollutant (to
which a standard applies) not previously
emitted, except that:
   (1)  Routine maintenance, repair, and
replacement  shall  not  be  considered
physical changes,  and
   (2)  The following shall not be consid-
ered   a  change  in  the  method  of
operation:
    (i)  An  increase in  the  production
rate, if such increase does not exceed the
operating design capacity of the affected
facility;
   (ii)  An increase in hours of operation;
   (iii) Use  of an alternative fuel or raw
material if, prior to the date  any stand-
ard under this part becomes applicable
to  such facility, as provided by § 60.1,
the affected facility is designed to  ac-
commodate  such alternative  use.
   (i)  "Commenced" means that an own-
er  or  operator has undertaken a  con-
tinuous program   of   construction  or
modification or that an owner or opera-
tor has entered into  a binding agree-
ment or contractual obligation to under-
take and complete, within  a  reasonable
time, a continuous program of construc-
tion or modification.
   (j)   "Opacity" means  the  degree  to
which emissions reduce the  transmission
of light and obscure the view of an object
in the background.
   (k)  "Nitrogen oxides"  means all  ox-
ides of nitrogen except nitrous oxide, as
measured by test  methods  set forth In
this part.
   (1)   "Standard of normal conditions"
means 70°  Fahrenheit  (21.1"  centi-
grade) and  29.92 in. Hg  (760 mm. Hg).
   (m)  "Proportional sampling" means
sampling at a rate that produces a con-
stant ratio of sampling rate to stack gas
flow rate.
   (n)  "Isokinetic    sampling"   means
sampling in which the linear velocity of
the gas entering the sampling nozzle is
equal  to  that  of  the  undisturbed  gas
stream at the sample point.
   (o)  "Startup" means  the  setting  hi
operation of an affected facility for any
purpose.

§ 60.3  Abbreviations.
   The  abbreviations used in  this part
have   the  following meanings  in  both
capital and lower case:
B.t.u.—British thermal unit.
cal.—calorie (s).
c.f.m.—cubic feet per minute.
COa—carbon dioxide.
g—gram(s).
gr.—grain(s).
mg —milligram (s).
mm.—millimeter(s).
1.—liter (s).
nm—nanometer(s), —10-' meter.
Pg.—microgram(s), 10-« gram.
Hg.—mercury.
in—inch(es).
K—l ,000.
lb.—pound (s).
ml —milliliter(s).
No.—number.
%—percent.
NO—nitric oxide
NOj—nitrogen dioxide.
NOX—nitrogen oxides.
NM.1—normal cubic meter.
s.c.f.—standard cubic feet.
SO,—sulfur dioxide.
H2SO,—sulfuric acid.
SO,—sulfur trioxide.
                              FEDERAL REGISTER, VOL  36, NO  247—THURSDAY, DECEMBER  ?3, 1971
                                                         1113

-------
24878
     RULES  AND  REGULATIONS
ft.3—cubic feet.
ftJ—square feet.
mm.—minute(s).
hr.—hour(s).

§ 60.4  Address.
  All applications, requests, submissions,
and reports under this part shall be sub-
mitted in triplicate and addressed to the
Environmental Protection Agency, Office
of General Enforcement, Waterside Mall
SW., Washington, DC 20460.
§ 60.5  Determination of construction or
    modification.
  When requested to do so by an  owner
or operator, the Administrator will make
a determination of whether actions taken
or Intended to be taken by such owner or
operator constitute construction or modi-
fication  or the  commencement thereof
within the meaning of this part.

§ 60.6  Review of plans.
  (a) When  requested to  do so by  an
owner or operator, the Administrator will
review plans  for construction or modifi-
cation  for  the  purpose  of providing
technical advice to the owner or operator.
  (b)  (1)  A separate request shall  be
submitted for each affected facility.
   (2) Each request shall (i)  identify the
location of such affected facility, and (ii)
be accompanied by technical information
describing the  proposed  nature, size,
design,' and method of operation of such
facility,  including  information on any
equipment to be used for measurement or
control of emissions.
   (c)  Neither a request for plans review
nor advice furnished by the Administra-
tor in response to such request shall (1)
relieve  an owner or  operator  of legal
responsibility for compliance with any
provision of this part or of any applicable
State or local requirement, or (2) prevent
the Administrator from implementing or
enforcing any provision of this part or
taking any other action authorized by the
Act.
§ 60.7  Notification and record  keeping.
   (a)  Any owner or operator subject to
the provisions of this part shall furnish
the Administrator written notification as
follows:
   (1) A notification  of  the anticipated
date  of initial  startup  of  an  affected
facility not more than  60  days or less
than 30 days prior to such date.
   (2) A notification  of  the  actual date
of initial startup of an  affected facility
within 15 days after such date.
   (b)  Any owner or  operator subject to
the provisions of  this part shall maintain
for a period of 2 years  a  record  of the
occurrence and duration of any startup,
shutdown, or malfunction in operation of
any affected  facility.
§ 60.8  Performance tests.
   (a)  Within 60*days after achieving the
maximum production rate at which the
affected facility will be operated, but not
later than 180 days after initial startup
of such facility and at such other times
as may be required by the Administrator
under section 114 of the Act, the owner
or operator of such facility shall conduct
performance test(s) and furnish the Ad-
ministrator a written report of the results
of such performance test(s).
   (b) Performance tests  shall  be  con-
ducted and  results reported in accord-
ance with the test method set forth in
this part or equivalent methods approved
by the Administrator;  or where the Ad-
ministrator  determines  that emissions
from the affected facility are  not sus-
ceptible  of  being measured  by  such
methods, the  Administrator shall  pre-
scribe  alternative  test procedures for
determining  compliance  with   the re-
quirements of this part.
   (c) The owner or operator shall permit
the Administrator to conduct perform-
ance tests at any reasonable time,  shall
cause the affected facility to be operated
for purposes of such  tests under  such
conditions as the Administrator  shall
specify based on representative perform-
ance of  the affected  facility, and  shall
make  available  to  the  Administrator
such records  as  may be  necessary  to
determine such performance.
    Safe sampling platform (s).
   (3)  Safe  access to  sampling   plat-
form (s).
   (4)  Utilities for sampling and testing
equipment.
   (f)  Each  performance test shall con-
sist of  three repetitions of the applicable
test method. For  the purpose of deter-
mining compliance with an applicable
standard of performance, the average of
results of all repetitions shall apply.
§60.9   Availability of information.
   (a)  Emission  data  provided to,  or
otherwise obtained by, the Administra-
tor in accordance with the provisions of
this part shall be available to the public.
   (b) Except as  provided in paragraph
(a) of this section, any records, reports,
or information provided to, or otherwise
obtained by, the Administrator in accord-
ance with the provisions of this part
shall be available to the public, except
that (1) upon a showing satisfactory to
the Administrator by any person that
such records,  reports, or information, or
particular  part   thereof  (other  than
emission data),  if made  public,  would
divulge methods  or processes entitled to
protection as  trade secrets of such per-
son, the Administrator  shall  consider
such records,  reports, or information, or
particular part thereof, confidential in
accordance with  the purposes of section
1905 of title 18  of the  United States
Code, except that such records, reports,
or information, or particular part there-
of, may be disclosed to other officers, em-
ployees,  or authorized representatives of
the United States concerned with cairy-
ing out the provisions of the Act or when
relevant  in any  proceeding under ths
Act; and (2) information received by the
Administrator solely for the purposes 01
§160.5 and 60.6  shall  not be  disclosed
if it is identified by the owner or opera-
tor  as being  a  trade secret  or com-
mercial or financial information which
such  owner   or  operator  considers
confidential.
§ 60.10  State authority.
  The provisions of this part shall not
be construed in any manner to preclude
any State or political subdivision thereof
from
  (a) Adopting and enforcing any emis-
sion standard or limitation applicable tj
an  affected facility, provided  that such
emission standard  or  limitation is not
less stringent  than the standard appli-
cable to  such facility.
  (b)  Requiring  the owner or operator
of an affected facility to obtain  permits.
licenses,  or approvals prior to initiating
construction, modification, or operation
of such facility.

Subpart D—Standards of Performance
for Fossil-Fuel Fired Sfeam Generators
§ 60.40  Applicability and designation of
     all'ected facility.
  The provisions  of this subpart are ap-
plicable  to each  fossil  fuel-fired steam
generating unit of more than 250 million
B.t.u. per hour heat input, which is the
affected  facility.
§ 60.41  Definitions.
  As used in this subpart,  all terms not
defined herein shall have  the meaning
given them in the  Act, and in  Subpart
A  of this part.
  (a)  "Fossil  fuel-fired steam generat-
ing unit" means a furnace or boiler used
in  the process of burning fossil fuel
for  the  primary purpose  of  producing
steam by heat transfer.
  (b)  "Fossil  fuel"  means natural gas,
petroleum, coal and any  form of solid,
liquid,  or gaseous  fuel  derived  from
such materials.
  (c)  "Particulate  matter" means any
finely  divided  liquid or solid  material,
other than uncombined water, as meas-
ured by  Method  5.
§ 60.42  Standard for particulate matter.
  On and after the date on which the
performance test required to  be  con-
ducted by § 60.8 is initiated  no owner
or  operator subject to  the provisions of
this part shall discharge  or  cause the
discharge into the atmosphere of par-
ticulate  matter which is:
   (a)  In excess  of  0.10 Ib. per million
B.t.u. heat input  (0.18 g. per million calj
maximum 2-hour average.
   (b)  Greater than 20 percent opacity,
except that 40 percent opacity  shall be
permissible for not more than 2 minutes
in  any hour.
   (c)  Where the  presence  of  uncom-
bined water is the  only reason  for fail-
ure to meet the requirements of para-
graph (b) of  this section  such failure
shall not be a violation of this section.
                              FEDERAL REGISTER,  VOL. 36, NO. 247—THURSDAY,  DECEMBER 23, 1971


                                                                  1114

-------
                                             RULES AND  REGULATIONS
                                                                                                               24879
§ 60.43  Standard for sulfur dioxide.
  On and after the date on which the
performance test required to be  con-
ducted by § 60.8 is initiated  no owner
or  operator  subject  to the provisions
of this part shall discharge or cause the
discharge into the  atmosphere of sulfur
dioxide in excess of:
   (a) 0.80 Ib. per million B.t.u. heat in-
put (1.4 g. per million cal.), maximum 2->
hour  average, when liquid fossil fuel is
burned.
   (b) 1.2 Ibs. per million B.t.u. heat input
(2.2 g. per million cal.),  maximum 2-
hour  average, when  solid fossil  fuel is
burned.
   (c)  Where different fossil  fuels  are
burned simultaneously in any combina-
tion,  the  applicable  standard shall  be
determined by  proration. Compliance
shall  be determined using the following
formula:
             y(0.80)-l-z(1.2)
                x+y+z

where:
  x Is the percent of total heat input derived
    from gaseous fossil fuel and,
  y is the percent of total heat input derived
    from liquid fossil fuel and,
  z is the percent of total heat input derived
    from solid fossil fuel.

§ 60.44  Standard for nitrogen oxides.

  On and  after the date on which the
performance  test  required to be  con-
ducted by I 60.8 is initiated no owner or
operator subject to the provisions of this
part shall  discharge or cause the dis-
charge into the atmosphere of nitrogen
oxides in excess of:
  (a) 0.20  Ib. per million B.t.u. heat in-
put (0.36 g. per million cal.), maximum
2-hour average, expressed as NO2, when
gaseous fossil fuel  is burned.
  (b) 0.30  Ib. per million B.t.u. heat in-
put (0.54 g. per million cal.), maximum
2-hour average, expressed as NO2, when
liquid fossil fuel is  burned.
  (c) 0.70  Ib. per million B.t.u. heat in-
put (1.26 g. per million cal.), maximum
2-hour average, expressed  as  NOa when
solid fossil fuel (except lignite) is burned.
  (d) When  different  fossil  fuels  are
burned simultaneously in any combina-
tion the applicable standard shall be de-
termined by proration. Compliance shall
be  determined  by using the following
formula:
        x(0.20) +y(0.30) +z(0.70)
                x+y+z
where:
  x is the percent of total heat input derived
    from gaseous fossil fuel and,
  y is the percent of total heat input derived
    from liquid fossil fuel and,
  z is the percent of total heat input derived
    from solid fossil fuel.

§ 60.45  Emission  and fuel monitoring.

  (a) There  shall  be  installed,  cali-
brated, maintained, and operated, in any
fossil fuel-fired steam generating unit
subject to  the  provisions  of this  part,
emission   monitoring  instruments  as
follows:
  (1) A photoelectric or  other  type
smoke  detector  and  recorder,  except
where  gaseous  fuel  is  the  only  fuel
burned.
  (2) An instrument  for  continuously
monitoring and recording sulfur dioxide
emissions, except where  gaseous  fuel is
the only fuel burned, or where compli-
ance is achieved through low sulfur fuels
and representative  sulfur  analysis  of
fuels are conducted daily in accordance
with paragraph (c) or (d) of this section.
  (3) An instrument  for  continuously
monitoring and recording  emissions  of
nitrogen oxides.
  (b) Instruments and sampling systems
installed and used pursuant to this sec-
tion shall be capable of monitoring emis-
sion levels within ±20  percent  with a
confidence level  of 95 percent and shall
be  calibrated in accordance  with  the
method(s)  prescribed by the  manufac-
turer^)  of such  instruments;  instru-
ments shall be subjected  to manufactur-
ers recommended zero adjustment and
calibration procedures at least once per
24-hour operating period unless the man-
ufacturer^)  specifies or   recommends
calibration at shorter intervals, in which
case such specifications or recommenda-
tions shall be followed.  The  applicable
method specified in the appendix of this
part shall be the reference  method.
  (c)  The sulfur content of solid fuels,
as burned, shall be determined  in accord-
ance with the following  methods of the
American  Society   for   Testing  and
Materials.
   (1) Mechanical sampling by Method
D 2234065.
   (2) Sample preparation  by  Method D
2013-65.
   (3) Sample  analysis  by Method  D
271-68.
   (d) The sulfur content of liquid fuels,
as burned, shall be determined in accord-
ance with the American Society for Test-
ing and Materials Methods D 1551-68, or
D 129-64, or D 1552-64.
   (e) The rate of fuel burned for each
fuel shall be measured daily or at  shorter
intervals  and  recorded.  The heating
value and ash content of fuels shall be
ascertained  at least once per  week and
recorded. Where the steam generating
unit is  used  to generate electricity, the
average electrical output and the mini-
mum and maximum hourly generation
rate shall  be measured and recorded
daily.
   (f) The  owner or  operator  of  any
fossil fuel-fired steam  generating  unit
subject to the  provisions  of  this  part
shall maintain a file of all measurements
required by this part. Appropriate meas-
urements shall be reduced to the units
of  the  applicable standard daily,  and
summarized monthly. The record of any
such   measurement(s)   and  summary
shall be retained for at least 2 years fol-
lowing  the date of such measurements
and summaries.
§ 60.46  Test methods and procedures.
   (a) The provisions of  this section are
applicable to performance tests  for de-
termining emissions of particulate mat-
ter, sulfur dioxide,  and  nitrogen oxides
from fossil  fuel-fired steam generating
units.
  (b) All performance tests shall be con-
ducted while the affected facility is oper-
ating at or above the maximum steam
production rate at which such facility
will be operated and while fuels or com-
binations  of   fuels  representative  of
normal operation are being burned and
under such other relevant conditions as
the Administrator  shall specify  based
on  representative performance of the
affected facility.
  (c) Test  methods  set  forth in the
appenlfet  to  this  part  or equivalent
methods approved by the Administrator
shall be used as follows:
   (1) For  each repetition, the average
concentration of particulate matter shall
be  determined by  using Method  5.
Traversing during sampling by Method 5
shall  be according  to Method 1. The
minimum sampling time shall be 2 hours,
and minimum sampling volume shall be
60 ft.3 corrected to standard conditions
on a dry basis.
   (2) For  each repetition, the SO* con-
centration  shall be determined by using
Method 6.  The sampling site shall be the
same as for determining volumetric flow
rate.  The  sampling point in  the duct
shall  be at the  centroid of  the  cross
section if the cross sectional area is less
than 50 ft.2 or at a point no closer to the
walls than 3 feet if the  cross sectional
area is 50 ft.' or more. The sample shall
be extracted at a rate proportional to the
gas velocity at the sampling point. The
minimum sampling time shall be 20 min.
and minimum sampling volume shall be
0.75 ft.3 corrected to standard conditions.
Two samples shall constitute one repeti-
tion  and  shall  be  taken  at  1-hour
intervals.
   (3) For  each repetition the NO, con-
centration  shall be determined by using
Method 7.  The sampling site and  point
shall be the same as for SO». The  sam-
pling time  shall be 2  hours,  and four
samples  shall  be taken  at 30-minute
intervals.
   (4) The volumetric flow rate of the
total effluent shall be determined by using
Method  2  and traversing according to
Method  1.  Gas analysis  shall be per-
formed by  Method  3, and moisture con-
tent shall  be  determined- by  the con-
denser technique of Method 5.
   (d) Heat input, expressed in B.t.u. per
hour, shall be determined during each 2-
hour testing period by suitable fuel flow
meters and shall be confirmed  by a ma-
terial balance over the steam generation
system.
   (e) POT each repetition, emissions, ex-
pressed in  Ib./lO* B.t.u. shall be deter-
mined by dividing  the emission rate in
Ib./hr. by  the heat input. The  emission
rate shall be determined by the equation,
lb./hr.=Q.xc  where,  Q,=volumetric
Sow rate of the total effluent in ft.'/hr. at
standard conditions, dry basis, as deter-
mined in accordance with paragraph (c)
(4) of this section.
   (1) For  particulate matter, c=partic-
ulate concentration in lb./ft.3, at deter-
mined in accordance with paragraph (c)
(1) of this  section, corrected to standard
conditions, dry basis
                             FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23,  1971


                                                        1115

-------
24880
     RULES AND REGULATIONS
  (2) For SO, c=SO* concentration in
Ib./f t.3, as determined in accordance with
paragraph  (c) (2)  of this  section, cor-
rected to standard conditions, dry basis;
  (3) For NO*, c=NO, concentration in
lb./ft.s, as determined in accordance with
paragraph  (c) (3)  of this  section, cor-
rected to standard conditions, dry basis.

Subpart E—Standards of Performance
           for Incinerators
§ 60.50   Applicability and designation of
     affected facility.
  The provisions of this subpart are ap-
plicable to each incinerator of more than
50 tons  per day charging rate, which is
the affected facility.
§ 60.51   Definitions.
  As used in this subpart, all  terms not
defined  herein  shall have the meaning
given them in the Act and in Subpart A
of this part.
  (a) "Incinerator" means any furnace
used in the process of burning solid waste
for the primary purpose of reducing the
volume  of the waste by removing com-
bustible matter.
  (b) "Solid waste" means refuse, more
than  50 percent of which  is  municipal
type waste consisting  of a mixture  of
paper, wood, yard  wastes,  food wastes,
plastics, leather, rubber, and other com-
bustibles, and noncombustible materials
such as glass and rock.
  (c) "Day" means 24 hours.
  (d) "Particulate  matter" means any
finely divided liquid or solid  material,
other than uncombined water, as meas-
ured by Method 5.
§ 60.52   Standard for paniculate matter.
  On and after the date on which  the
performance test required  to be con-
ducted by § 60.8 is initiated,  no owner
or operator subject to  the provisions of
this part shall  discharge or  cause  the
discharge into  the  atmosphere of par-
ticulate matter which is in excess of 0.08
gr./s.c.f. (0.18  g./NM")  corrected  to  12
percent CO2, maximum 2-hour average.
§ 60.53   Monitoring of operations.
  The owner or operator  of any  in-
cinerator subject to the provisions of this
part shall maintain a file of daily burn-
ing rates and hours of operation and any
particulate emission measurements. The
burning rates  and  hours  of operation
shall be summarized  monthly.  The
record(s) and summary shall be retained
for at least 2 years  following the date of
such records and summaries.
§ 60.54  Test methods and procedures.
   (a) The provisions of this section are
applicable to performance tests for de-
termining emissions of particulate matter
from incinerators.
   (b) All performance  tests  shall  be
conducted while the affected facility is
operating at  or above the  maximum
refuse charging rate at which such facil-
ity will be operated and the solid waste
burned  shall be representative of normal
operation and under such other relevant
conditions as  the  Administrator shall
specify  based  on  representative per-
formance of the affected facility.
  (c) Test methods set forth in the ap-
pendix to this part or equivalent methods
approved by the Administrator shall be
used as follows:
  (1) For  each repetition, the average
concentration of particulate matter shall
be determined by using Method 5. Tra-
versing during sampling  by Method 5
shall be according to Method 1. The mini-
mum sampling time shall be 2 hours and
the minimum sampling volume shall be
60 ft.3  corrected to- standard conditions
on a dry basis.
  (2) Gas  analysis shall  be performed
using the integrated sample technique of
Method 3, and moisture content shall be
determined by  the condenser technique
of Method 5. If a wet scrubber is used,
the gas analysis sample shall reflect flue
gas conditions after the scrubber, allow-
ing for the effect of carbon dioxide ab-
sorption.
  (d) For  each repetition  particulate
matter emissions, expressed in gr./s.c.f.,
shall be  determined  in accordance with
paragraph (c) (1) of  this section cor-
rected to 12 percent CO,, dry basis.

Subpart F—Standards of Performance
     for Portland Cement Plants

§ 60.60  Applicability and designation of
    affected facility.
  The  provisions of the subpart are ap-
plicable to the  following affected facili-
ties  in Portland cement  plants: kiln,
clinker cooler,  raw  mill  system, finish
mill system, raw mill dryer, raw material
storage,  clinker storage, finished prod-
uct  storage,  conveyor  transfer points,
bagging and bulk loading  and unloading
systems.

§ 60.61  Definitions.
  As used in this subpart, all terms not
defined herein shall have the meaning
given them in the Act  and in Subpart A
of this part.
  (a)  "Portland cement  plant"  means
any facility manufacturing Portland ce-
ment by either the wet or dry process.
  (b)  "Particulate  matter"  means any
finely  divided  liquid or solid  material,
other than uncombined water, as meas-
ured by  Method 5.
§ 60.62  Standard for particulate matter.
  (a)  On and  after the date on which
the performance test required to be con-
ducted by § 60.8 is  initiated no owner
or operator subject to the provisions of
this part shall discharge  or cause  the
discharge into the atmosphere of par-
ticulate matter from the  kiln which is:
  (1) In excess of 0.30 Ib.  per ton of feed
to the kiln (0.15 Kg.  per metric ton),
maximum  2-hour average.
  (2)  Greater  than 10 percent opacity,
except that where the presence of uncom-
bined water is the only reason for failure
to meet the requirements for  this sub-
paragraph, such failure shall  not be  a
violation of this section.
  (b)  On  and  after the date on which
the performance test required to be con-
ducted by 5 60.8 is Initiated no owner
or operator subject to the provisions of
this part shall discharge or cause the dis-
charge into the atmosphere of particulate
matter from the clinker cooler which is:
  (1) In excess of 0.10 Ib. per ton of feed
to the kiln  (0.050 Kg. per metric ton)
maximum 2-hour average.
  (2) 10 percent opacity or greater.
  (c) On and after the date on which the
performance  test required  to  be con-
ducted by § 60.8 is  initiated no  owner
or operator subject to the provisions of
this  part  shall  discharge or cause  the
discharge into the atmosphere of partic-
ulate matter from any affected facility
other than the  kiln and clinker cooler
which is 10 percent opacity or greater.
§ 60.63   Monitoring of operations.
  The owner or operator of any portland
cement  plant subject  to  the provisions
of this part shall maintain a file of daily
production rates and kiln feed rates and
any  particulate  emission measurements.
The  production  and,feed rates shall be
summarized monthly. The record (s) and
summary  shall be retained  for at least
2 years following the date of such records
and  summaries.

§ 60.64  Test methods and procedures.
   (a) The provisions of this section are
applicable to performance tests for de-
termining emissions-of particulate mat-
ter  from  Portland cement  plant kilns
and  clinker  coolers.
   (b) All performance  tests  shall  be
conducted while the affected-facility is
operating at or above the maximum
production rate  at which such facility
will  be  operated and  under such other
relevant conditions as the Administrator
shall specify based on representative per-
formance  of  the affected facility.
   (c) Test methods  set forth in the ap-
pendix to this part or equivalent meth-
ods approved by the Administrator shall
be used as follows:
   (1) For each repetition,  the average
concentration of particulate matter shall
be determined by using Method 5. Tra-
versing during  sampling by Method 5
shall be according to Method 1. The mini-
mum sampling time shall be 2 hours and
the  minimum sampling volume shall be
60 ft.* corrected to standard conditions
on a dry basis.
   (2) The volumetric flow  rate  of the
total effluent shall be determined by us-
ing Method 2 and traversing according to
Method 1. Gas  analysis shall be per-
formed using the integrated sample tech-
nique of Method 3, and moisture content
shall be determined by  the condenser
technique of Method 5.
   (d) Total kiln feed  (except fuels), ex-
pressed in tons  per hour on a dry basis,
shall be determined during  each 2-hour
testing  period  by suitable  flow  meters
and shall  be confirmed by a material
balance over the production system.
   (e) For  each repetition, particulate
matter emissions, expressed in Ib./ton of
kiln feed shall be determined by dividing
ttie  emission rate in Ib./hr. by the kiln
feed. The emission rate shall be deter-
mined by the  equation, lb./hr.=Q«xc,
                             FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
                                                             1116

-------
                                            RULES AND REGULATIONS
                                                                      24881
where Q.=volumetric  flow  rate of the
total effluent in f t.'/hr. at standard condi-
tions, dry  basis,  as determined in ac-
cordance with paragraph (c) (2) of this
section, and,  c=particulate  concentra-
tion in lb./ft.*, as determined in accord-
ance  with  paragraph  (c) (1)  of  this
section, corrected to standard conditions,
dry basis.

Subpart G—Standards of Performance
        for Nitric Acid  Plants
§ 60.70  Applicability and designation of
    affected facility.
  The provisions  of this subpart are
applicable to each nitric acid production
unit, which is the affected facility.
§ 60.71  Definitions.
  As used in this subpart, all terms not
defined herein shall have the  meaning
given them in the Act and in Subpart A
of this part.
  (a)  "Nitric   add  production  unit"
means any facility producing weak nitric
acid by either the pressure or atmos-
pheric pressure process.
  (b)  "Weafc  nitric add" means add
which is 30 to 70 percent in strength.
§ 60.72   Standard for nitrogen oxides.
  On  and  after the date on which the
performance  test  required  to  be  con-
ducted by  § 60.8  is initiated no owner
or operator subject to  the provisions of
this part shall discharge or cause the
discharge Into the atmosphere of nitro-
gen oxides which are:
  (a)  In excess of 3 Ibs. per ton of acid
produced   (1.5 kg.  per  metric  ton),
maximum 2-hour  average, expressed as
N02.
  (b)  10 percent opadty or  greater.
§ 60.73   Emission monitoring.
  (a)  There  shall be  installed,  cali-
brated, maintained, and operated, in any
nitric add production  unit subject to
the provisions  of this subpart, an instru-
ment  for continuously  monitoring  and
recording emissions of nitrogen oxides.
  (b)  The instrument  and  sampling
system1 installed and used pursuant to
this section shall be capable of monitor-
ing emission levels within ±20 percent
with a confidence level of 95 percent and
shall  be  calibrated in  accordance with
the method(s) prescribed by the manu-
facturer (6)  of such  instrument,  the
instrument  shall   be   subjected  to
manufacturers' recommended zero  ad-
justment and  calibration procedures at
least once per 24-hour operating period
unless the  manufacturer (s)  specifies or
recommends calibration at shorter In-
tervals, in which case such specifications
or recommendations shall be followed.
The applicable method specified in tbe
appendix of this part shall be the ref-
erence method.
  (c) Production rate and hours of op-
eration shall be recorded daily.
  (d) The owner  or operator of  any
nitric acid production unit subject to the
provisions of  this  part shall maintain
a file of all measurements required by
this subpart. Appropriate measurements
shall be reduced  to the units of the
standard daily and summarized monthly.
The record  of any  such measurement
and summary shall  be  retained for at
least 2 years following the date of such
measurements and summaries.
§ 60.74   Test methods and procedures.
  (a) The provisions of this section are
applicable to performance  tests for de-
termining emissions  of  nitrogen oxides
from nitric acid production units.
  (b) All performance  tests shall  be
conducted while the affected facility is
operating at or above the maximum acid
production rate at which  such facility
will be operated and under such other
relevant  conditions as the Administra-
tor shall specify based on representa-
tive performance of the affected facility.
  (c) Test methods set forth in the ap-
pendix to this part or equivalent methods
as approved by the Administrator shall
be used as follows:
  (1) For each repetition the NO, con-
centration shall be determined by using
Method 7. The sampling site  shall be
selected according  to Method 1 and the
sampling point shall be the centroid of
the stack or duct. The  sampling time
shall be 2 hours and four samples shall
be taken at  30-minute intervals.
  (2) The volumetric flow rate of the
total  effluent  shall  be  determined  by
using Method 2 and traversing accord-
ing to Method 1. Gas analysis shall be
performed   by  using  the  integrated
sample  technique  of  Method 3,  and
moisture content shall be determined by
Method 4.
  (d) Add produced, expressed in tons
per hour of 100 percent nitric acid, shall
be determined during each 2-hour test-
ing period by suitable flow meters and
shall be  confirmed by a  material bal-
ance over the production system.
  (e) For  each  repetition,   nitrogen
oxides emissions,  expressed in Ib./ton
of 100 percent nitric acid, shall be de-
termined by dividing the emission rate
in  Ib./hr. by the  add produced. The
emission rate shall  be determined by
the   equation,  lbyhr.=QsXc,   where
Qa=volumetrlc flow  rate of the effluent
In ft.'/hr. at  standard  conditions, dry
basis, as determined  in accordance with
paragraph (c) (2)  of this  section,  and
c=NO, concentration in lb./ft.', as de-
termined in accordance  with paragraph
(c) (1) of tills section, corrected to stand-
ard conditions, dry basis.

Subpart H — Standards of Performance
       for  Sulfuric Acid Plants

§ 60.80   Applicability and designation of
     affected facility.
  The provisions of this subpart are ap-
plicable to  each sulfuric acid production
unit, which is the affected facility.

§ 60.81   Definitions.
  As used in this subpart, all terms not
defined  herein  shall have the  meaning
given them in the Act and in Subpart A
of this part.
  (a) "Sulfuric acid production  unit"
means  any faculty  producing sulfuric
acid by the contact process by burning
elemental sulfur, alkylation acid, hydro-
gen  sulfide, organic sulfides and  mer-
captans, or acid sludge, but does not in-
clude facilities  where conversion to sul-
furic acid is utilized primarily as a means
of preventing emissions  to  the atmos-
phere of sulfur dioxide or other sulfur
compounds.
  (b) "Acid  mist" means sulfuric acid
mist, as measured by test methods set
forth in this part.
§ 60.82   Standard for sulfur dioxide.
  On and  after the  date on which the
performance test  required  to  be con-
ducted by § 60.8 is initiated no owner or
operator subject to the provisions of this
part shall  discharge  or  cause  the dis-
charge  into  the  atmosphere of sulfur
dioxide  in excess of 4 Ibs. per ton of acid
produced (2 kg. per metric ton), maxi-
mum 2 -hour average.
§ 60.83   Standard for acid mist.
  On and  after the  date on which  the
performance  test  required  to  be con-
ducted by § 60.8 is initiated no owner or
operator subject to the provisions of this
part shall  discharge  or  cause  the dis-
charge into the atmosphere of acid mist
which is:
  (a) In excess of 0.15 Ib. per ton of acid
produced   (0.075  kg.  per metric  ton),
maximum  2-hour average, expressed as
  (b) 10 percent opacity or greater.
§ 60.84  Emission monitoring.
  (a) There shall  be  installed,  cali-
brated, maintained, and operated, in any
sulfuric acid production unit subject to
the provisions  of this  subpart, an  in-
strument  for continuously monitoring
and recording emissions of sulfur dioxide.
  (b) The instrument and sampling sys-
tem installed and used  pursuant to this
section shall be capable of monitoring
emission levels within ±20 percent with
a confidence level of 95 percent and shall
be  calibrated in accordance  with the
                             RDERAl MGISlCt. VOL 36,  NO. 147—THURSDAY, DECEMBER 23, 1971
                                                          1117

-------
24882
                           RULES AND REGULATIONS
 ^irPili^fU
 »«ag|l§l8||sa|s5
 *j!b3J|aM|«*3
 lfSg|j)if|JSgI:i3g

 *  I'llilWf
 2i!2?g||lr!r^I
 «B3«MO
                                                                    o
                                                                    ui
                                                                    o



                                                                    2
                                                                     O
                                                                     i
3

II
5S
                                    1118

-------
                                RULES  AND  REGULATIONS
                                                                                                            24883
      QJ CO
     J3

      E
tu        s-
W JJ J-   Q)
s- cr tu 03 -M
o> -I— .a   at
> o E c: E
fd ex n o TO
             i—  CM  VO
                      en ,u")

                      r*"» o
                   *3-  O  CM  CO



                   i—  CM  CM  CO
LD CO O)


*£ 10 CO
             »-*  M  VQ CO »—
                   co i~  in



                   CM CM  CO
                                                                                63
             ro  cn  r--. r-» en
                         CM  <^0  O
                                   U3 CM  CO
             i —  «3"  CO
                      LD  cn


                      CM  UD
                                                   LT) 1T>  I—
                                                co co  en
                   en UD i —
                                                f — CO CM
                          o  in  un o co  CM  ro
                   t— f-. in  in
                   ^- i— CM  CO
                             in  N- in f—  r-
«— r- CO
             co co en
             r*v  o o co

             vo  in in ro
          10  10  s, co cn  o  t-
                                                                                                               IU
                                                                                                               U
                                                                                                               tu
                                                                                                               a
                                CM cd
                                i- CO
                                '" 12
                                *a o


                                is

                                ^- m
                                0 >

                                •*- £
                                "^; c
                                0 0
                                o


                                "I
                                 . '5
                                CM O
                                 '  ,C
                                •r- CO
                                
o 2
CO £
CD
5 cn
2, to
.? CD
U. (Q
                                                                                                               «

                                                                                                               6
                                                                                                               O
                                            1119

-------
24884
      RULES AND  REGULATIONS
  2.2.2  For  rectangular stacks divide the
cross section into as many equal rectangular
areas as traverse points, sucll that the ratio
of the length to  the width at the elemental
areas Is between one and two. Locate the
traverse points at the centroid of each equal
area according to Figure 1-3.
  3. References.
  Determining Dust Concentration in a Gas
Stream, ASME Performance Test Code  #27,
New York, N.Y.,  1957.
  Devorkin,  Howard,  et al., Air  Pollution
Source Testing Manual, Air Pollution Control
District, Los Angeles,  Calif. November 1963.
  Methods for  Determination  of Velocity,
Volume, Dust and Mist Content of  Gases,
Western Precipitation Division of Joy Manu-
facturing  Co., Los Angeles, Calif.  Bulletin
WP-50, 1968.
  Standard Method for  Sampling Stacks for
Particulate Matter, In:  1971 Book of ASTM
Standards, Part  23, Philadelphia,  Pa. 1971,
ASTM Designation  D-2928-71.

METHOD  2	DETERMINATION  OP  STACK  GAS
  VELOCITY AND VOLUMETRIC FLOW RATE (TYPE
  S PTTOT TUBE)

  1. Principle and applicability.
  1.1  Principle.  Stack gas velocity is deter-
mined from tne gas density and from meas-
urement of the velocity head using a Type S
(Stausohelbe or reverse type) pitot tube.
  1.2  Applicability. This method should  be
applied only when  specified by the test pro-
cedures for determining compliance with the
New Source Performance Standards.
  2. Apparatus.
  2.1  Pitot tube—Type S  (Figure  2-1), or
equivalent,  with a coefficient within  ±5%
over the working range.
  2.2  Differential pressure gauge—Inclined
manometer, or equivalent, to measure velo-
city head to within  10%  of the minimum
value.
  2.3  Temperature gauge—Thermocouple or
equivalent attached  to the  pitot  tube to
measure stack temperature to within 1.5 % of
the  minimum  absolute stack temperature.
  2.4  Pressure gauge—Mercury-filled TJ-tube
manometer, or equivalent, to measure stack
pressure to within 0.1 in. Hg.
  2.5  Barometer—To  measure atmospheric
pressure to within 0.1 in. Hg.
  2.6  Gas analyzer—To analyze gas composi-
tion for determining molecular weight.
  2.7  Pitot  tube—Standard  type,  to cali-
brate Type S pitot tube.
  3. Procedure.
  3.1  Set up the apparatus as shown in Fig-
ure 2-1. Make sure all connections are tight
and leak free. Measure the velocity head and
temperature at  the traverse points specified
by Method 1.
  3.2  Measure  the static  pressure  in the
stack.
  3.3  Determine  the stack  gas molecular
weight by gas analysis and appropriate cal-
culations as indicated in Method 3.
                                      PIPE COUPLING
                     TUBING ADAPTER
  4.  Calibration.

  4.1  To calibrate the pitot tube,  measure
the velocity head at some point in a flowing
gas stream with both a Type S pitot tube and
a standard type pitot tube with known  co-
efficient. Calibration  should be done in  the
laboratory and the velocity of the flowing gas
stream should  be  varied over the normal
working range.  It is  recommended  that  the
calibration be repeated after use at each field
site.
  4.2  Calculate the  pitot tube  coefficient
using equation 2-1.
                                  *•   o  ,
                       Apt,,,  equation 2-1
where :
  Cp,esl=Pitot  tube coefficient  of Type  S
            pitot tube.
   Cn,td=Pitot  tube coefficient of standard
            type pitot tube (if unknown, use
            0.99) .
   Ap,tt = Velocity head measured  by stand-
            ard  type pitot tube.
  Apte«t=: Velocity head measured by Type S
            pitot tube.
  4.3  Compare the coefficients of the Type S
pitot tube determined first with one leg and
then the other pointed downstream.  Use the
pitot tube only if the two coefficients differ by
no more than 0.01.
  5.  Calculations.
  Use equation 2-2 to calculate the stack gas
velocity.
                                             where-
                                                (Va)«
                                                                                                                      Equation 2-2

                                                                                                   = Stack gas velocity, feet per second (f.p.s ).
                                                                                                 C0=Pltot tube coefficient, dimenslonless.
                                                                                             (T8)avB.=Average absolute stack gas tempeiature,
                                                                                                     °
                                                                                                   = Average velocity head of stack gas, inches
                                                                                                     H,0 (see Fig. 2-2).
                                                                                                 P,=Absolute stack gas pressure, inches Hg.
                                                                                                 Ma=Molecular weight of stack gas (wet basis),
                                                                                                     Ib /Ib.-mole.
                                                                                                       Md(l— B,o)+18B,0
                                                                                                 Md=Dry molecular weight of stack gas (from
                                                                                                     Methods).
                                                                                                Bwo= Proportion by volume of water vapor in
                                                                                                     the gas stream (from Method 4).

                                                                                            Figure 2-2 shows a sample recording sheet
                                                                                         for velocity traverse data. Use the averages
                                                                                         in the last two columns of Figure 2-2 to de-
                                                                                         termine the average stack gas velocity from
                                                                                         Equation 2-2.
                                                                                            Use Equation 2-3 to calculate the stack
                                                                                         gas volumetric flow rate.
                                                                                           Q.=3600
    Figure 2-1.  Pitot tube-manometer assembly.
                                                                                                                       Equation 2-3
                                                                                          where:
                                                                                            Q.=Volumetric flow rate, dry basis, standard condi-
                                                                                                 tions, ft.'/hr.
                                                                                             A = Cross-sectional area of stack, ft.'
                                                                                           T,td*=Absolute temperature at standard conditions,
                                                                                                 630° R.
                                                                                           Patd^AbsoIute pressure at standard conditions, 29.93
                                                                                                 inches Hg.
                                 FEDERAL REGISTER,  VOL. 36, NO. 247—THURSDAY, DECEMBER 23,  1971


                                                                       1120

-------
                          RULES  AND  REGULATIONS
                                                                    24885
  6. References.

  Mark, L. S., Mechanical Engineers' Hand-
book, McGraw-Hill Book Co., Inc., New York,
N.Y., 1951.
  Perry, J.  H.,  Chemical  Engineers' Hand-
book, McGraw-Hill Book Co., Inc., New York,
N.Y., I960.
  Shigehara, K. T., W.  F. Todd,  and W. S.
Smith, Significance of Errors in Stack Sam-
              pling Measurements. Paper presented at the
              Annual Meeting of the Air Pollution Control
              Association, St. Louis, Mo., June 14-19, 1970.
               Standard Method for Sampling Stacks for
              Particulate Matter, In: 1971 Book of ASTM
              Standards, Part 23, Philadelphia,  Pa., 1971,
              ASTM Designation D-2928-71.
               Vennard, J. K., Elementary Fluid Mechan-
              ics, John Wiley & Sons, Inc., New York, N.Y.,
              1947.
  PLANT,

  DATE
  RUN NO.
  STACK DIAMETER, in.
  BAROMETRIC PRESSURE, in.
  STATIC PRESSURE IN STACK (Pg), in. Hg._

  OPE R ATORS	
                              SCHEMATIC OF STACK
                                 CROSS SECTION
         Traverse point
             number
Velocity head,
   in. H20
                                                              Stack Temperature
                                AVERAGE:
                       Figure 2-2. Velocity traverse data.
         FEDERAL REGISTER,  VOL.  36, NO. 247—THURSDAY, DECEMBER 23, 1971
                                      1121

-------
24886
                                                 RULES AND  REGULATIONS
METHOD 3	GAS ANALYSIS FOE CARBON DIOXIDE,
  EXCESS AIR, AND  DRY MOLECtTLAR WEIGHT

  1. Principle and applicability.
  1.1  Principle. An integrated  or  grab  gas
sample  is extracted from a sampling point
and analyzed for  its components using  an
Orsat analyzer.
  1.2  Applicability. This  method should be
applied  only when specified by the  test pro-
cedures for determining compliance with the
New Source Performance Standards. The test
procedure will indicate whether a grab sam-
ple or an integrated sample is to be used.
  2. Apparatus.
  2.1  Grab sample (Figure 3-1).
  2.1.1  Probe—Stainless  steel   or  Pyrex1
glass, equipped with a filter to remove partic-
ulate matter.
  2.1.2  Pump—One-way  squeeze  bulb,  or
equivalent,   to  transport gas   sample  to
analyzer.
  1 Trade name.
                                             2.2  Integrated sample (Figure 3-2).
                                             2.2.1  Probe—Stainless  steel  or Pyrex1
                                           glass, equipped with a filter to remove par-
                                           ticulate matter.
                                             2.2.2  Air-cooled condenser or equivalent—
                                           To remove any excess moisture.
                                             2.2.3  Needle valve—To  adjust  flow  rate.
                                             2.2.4  Pump—Leak-free,  diaphragm  type,
                                           or equivalent, to pull gas.
                                             2.2.5  Bate meter—To  measure a  flow
                                           range from 0 to  0.035 cfm.
                                             2.2.6  Flexible bag—Tedlar,1 or equivalent,
                                           with a capacity of 2 to 3 cu. ft. Leak test the
                                           bag  in the laboratory before using.
                                             2.2.7  Pitot tube—Type  S,  or equivalent,
                                           attached to the probe so that the sampling
                                           flow rate can be  regulated proportional to
                                           the stack gas velocity when velocity is vary-
                                           ing  with  time  or a  sample traverse is
                                           conducted.
                                             2 3  Analysis.
                                             2.3.1  Orsat analyzer, or  equivalent.
                  PROBE
                                           'FLEXIBLE TUBING
                                                                       TO ANALYZER
  LTER (G
FILTER (GLASS WOOL)
                                         SQUEEZE BULB




                         Figure 3-1.  Grab-sampling train.

                                             RATE METE?


                                   VALVE

          AIR-COOLED. CONDENSER       /        PUMP

     PROBE
 FILTERlGLASSYIIOOL}
                                                                  QUICK DISCONNECT
                                   RIGID CONTAINER'
                 Figure 3-2. Integrated gas • sampling train.
  3. Procedure.
  3 1  Grab sampling.
  3.1.1  Set up the equipment as shown in
Figure 3-1, making sure all connections are
leak-free. Place the probe in the stack at a
sampling point and purge the sampling line.
  3.1.2  Draw sample into the analyzer.
  3.2  Integrated sampling.
  3.2.1  Evacuate the flexible bag. Set up the
equipment as shown  in Figure 3-2 with the
bag  disconnected. Place  the probe  In  the
stack and purge the sampling line. Connect
the bag, making sure that all connections are
tight  and that there are no leaks.
  3.2.2"- Sample at a rate proportional to the
stack velocity.
  3.3  Analysis.
  3.3.1  Determine the CO2, O2, and CO con-
centrations as soon as possible. Make as many
passes as are necessary to give constant read-
ings. If more than ten passes are necessary,
replace the absorbing solution.
  3.3.2  For grab sampling, repeat the sam-
pling and  analysis until three consecutive
samples vary no more than 0.5  percent by
volume for each component being analyzed.
  3.3.3  For integrated sampling, repeat the
analysis of toe sample until three consecu-
tive analyses vary no more  than 0.3 percent
by  volume  for  each  component   being
analyzed.
  4.  Calculations.
  4.1  Cartoon dioxide. Average the three con-
secutive runs and report the result to the
nearest 0.1% CO-
  4.2 Excess air". Use Equation 3-1 to calcu-
late excess air, and average  the runs. Report
the result to the nearest 0.1%  excess air.

%EA =

        (%02)-0.5(%CO)
0.264(% N,)-(% 02)+0.5(% CO)X1UU

                             equation 3-1
where:
  %EA=Percent excess air.
   %O3=Percent oxygen by volume, dry basis.
   %N3=Percent nitrogen  by volume, dry
           basis.
  % CO=Percent carbon  monoxide  by  vol-
           ume, dry basis.
  0.264=Ratio of oxygen to nitrogen in air
           by volume.
  4.3  Dry molecular weight. Use Equation
3-2 to  calculate dry molecular weight  an-mole.
  % COn=Percent carbon dioxide  by volume,
           dry basis.
    %Oa=Percent  oxygen  by  volume,  dry
           basis.
    %Ni=Percent  nitrogen  by volume, dry
           basis.
    0.44=Molecular weight of carbon dioxide
           divided by 100.
    0.32=Molecular weight of oxygen divided
           by 100.
    0.28=Molecular  weight of  nitrogen and
           CO divided by 100.
                                 FEDERAL REGISTER, VOL.  36, NO. J47—THURSDAt, DECEMBER 23.  1971
                                                                      1122

-------
       RULES AND REGULATIONS
                                                 24887






in
§
I
Q
C 0
*5

at
tz
"a.
rt
u>
a>
w
'o
S
„.:
••a-
a)
3
O)
LL

2















O
S te
3 P
















u.
§


























a
i
S3
K
fl_
O
OC
^ £
S 8
UJ £C
fe <
O co




UJ
cc
5
OC
UJ
^
S u-
uj a
H-
ec
UJ
£
S

(3
z
(-•
S
In.e
DC *E
S <
*~ CO
U; -
<
g*"


I
o
o
o ^
M
S ~-
                       aa-SB'^j.
                        SS jajSa




                       Ipfl1
   cs
-»§"S«3§«aS^§i5 .
^iriiilig!i°ir
                      Ss"  0«
                      S§S¥S °
                      •o$33aa
                      3 1 » 2 £2
                      &= a fe &-g ,
s « t>>
SOP
                                                   o
                                                   o
                                                   _


                                                   I
                                         > «
               1123

-------
24888
                                                  RULES AND  REGULATIONS
4.2  Oas volume.
       17 71 -
         '
            in. Hg V  Tm  /  equation 4-2
where:
  Vmc =Dry gas volume through meter  at
          standard conditions, cu. ft.
  Vm =Dry gas volume measured by meter,
          cu. ft.
  Pm = Barometric pressure at  the dry gas
          meter, Inches Hg.
  P.td=Pressure»t standard conditions, 29.92
          Inches Hg.
  T»td=Absolute temperature at  standard
          conditions, 530° B.
  Tm = Absolute temperature at meter (° F+
          460),  °B.
4.3   Moisture content.

T>        *W9    I T3         • WO
                               -+(0.025)

                             equation 4-3
where:
  Bwo=Proportion by volume of water vapor
          in the gas stream, dimensionless.
  Vwc =Volume  of  water  vapor  collected
          (standard conditions), cu. ft.
  Vmc =Dry  gas  volume  through  meter
          (standard conditions), cu. ft.
  BWM=Approximate volumetric  proportion
          of water vapor in the gas stream
          leaving the implngers, 0.025.
  5. References.
  Air Pollution Engineering Manual, Daniel-
son, J. A. (ed.), U.S. DHEW, PHS, National
Center for Air Pollution Control, Cincinnati,
Ohio, PHS Publication  No.  999-AP-40, 1967.
  Devorkln,  Howard, et al., Air Pollution
Source Testing Manual, Air Pollution Con-
trol District, Los  Angeles,  Calif.,  November
1963.
  Methods for Determination of  Velocity,
Volume, Dust and Mist Content of  Gases,
Western Precipitation Division of Joy Manu-
facturing Co., Los Angeles, Calif., Bulletin
WP-60, 1968.

METHOD  5—DETERMINATION  OF  PARTICULATE
   EMISSIONS FROM  STATIONARY SOURCES

  1. Principle and applicability.
  1.1  Principle. Particulate matter is with-
drawn Isokinetioally from the source and its
weight Is determined gravimetrically after re-
moval of uncomiblned water.
  1.2  Applicability. This method is applica-
ble for the determination of particulate emis-
sions from  stationary   sources only when
specified by the test procedures for determin-
ing  compliance  with New  Source  Perform-
ance Standards.
  2. Apparatus.
  2.1  Sampling train.  The design specifica-
tions.of the  particulate sampling train used
by  EPA (Figure 5-1)  are described in APTD-
0581. Commercial models  of this train  are
available.
  2.1.1   Nozzle—Stainless  steel (316)  with
sharp, tapered leading  edge.
  2.1.2   Probe—Pyrex1  glass with a heating
system capable of maintaining a minimum
gas  temperature  of  250° F.  at the exit end
during  sampling  to prevent condensation
from occurring.   When length  limitations
(greater than about 8 ft.) are encountered at
temperatures less  than  600° F., Incoloy 825 *,
or equivalent, may be usedv Probes for sam-
pling gas streams at temperatures in excess
of 600° F. must have been approved by  the
Administrator.
  2.1.3   Pitot tube—Type S, or  equivalent,
attached to probe   to monitor  stack  gas
velocity.
  2.1.4  Filter  Holder—Pyrex»  glass  with
heating system capable of maintaining mini-
mum temperature of 225° F.
  2.1.5  Implngers / Condenser—Four impin-
gers connected in series with glass ball Joint
fittings. The first, third, and fourth  impin-
gers are  of the  Greenburg-Smitn  design,
modified by replacing the tip with a 
-------
                                                   RULES  AND REGULATIONS
                                                                                  24889
       PLANT	.	

       LOCATION	

       OPERATOR __,

       DATE	

       BUN NO.	

       SAMPLE BOX N0j_

       METER BOX NO,

       METER AH.,	

       C FACTOR	
                   AMBIENT TEMPERATUflf _

                   BAROMETRIC PRESSURE_

                   ASSUMED MOISTURE. *__

                   HEATtR BOX SETTING	

                   PROBE LENGTH, »	

                   NOZZLE DIAMETER, ln._

                   PflOBt HEATER SETTING,
                                   SCHEMATIC Of STACK CROSS SECTION
TRAVERSE POINT
NUMBER












TOTAL,
SAMPLING
TIME
(•). min.













AVERAGE
STATIC
PRESSURE
IPS). fc H9














STACK
TEMPERATURE
ITS)."F














VELOCITY
HEAD
I*PS>.














PRESSURE
DIFFERENTIAL
ACROSS
ORIFICE
METER
(AH),
In, H2O














GASSAMPU
VOLUME
IVm) It3














GAS SAMPLE TEMPERATURE
AT DRY GAS METER
INLET
tT(H}n>1,*F












A«g.
OUTLET
IT-ou.l.-f












Avg.
Avg.
SAMPLE BOX
TEMPERATURE.














TEMPERATURE.
OF GAS
LEAVIHG
CONDENSER OR
LAST IMPINCER














     Tm = Average dry gas meter temperature,
            °R.
    Pbir = Barometric pressure at the orifice
            meter, inches Hg.
     AH = Average pressure drop across the
            orifice meter, inches  H.O.
    13.6 = Specific gravity of mercury.
    Plld= Absolute pressure at standard con-
            ditions, 29.92 inches  Hg.

  6.3  Volume of water vapor
  4.2  Sample recovery. Exercise care in mov-
ing the collection train from the test site to
the  sample  recovery area to minimize the
loss of  collected  sample or  the  gain of
extraneous  particulate matter.  Set aside  a
portion of the  acetone  used in the sample
recovery as a blank for analysis. Measure the
volume of water from  the first three  im-
pingers,  then discard. Place the  samples in
containers as follows:
  Container  No. 1. Remove the  filter from
its holder, place in this container,  and seal.
  Container  No.  2. Place  loose  particulate
matter   and  acetone  washings  from  all
sample-exposed surfaces prior to the filter
in this container and seal. Use a razor  Wade,
brush, or rubber policeman to lose adhering
particles.
  Container  No.  3. Transfer  the  silica gel
from the fourth impinger to the original con-
tainer and seal. Use  a rubber policeman as
an  aid  in   removing silica  gel  from  the
Impinger.
  4.3  Analysis. Record the data required on
the  example sheet shown in  Figure  5-3.
Handle each sample container as follows:
  Container  No. 1. Transfer the  filter  and
any loose particulate matter from the sample
container to a tared glass  weighing dish,
desiccate, and dry to a constant weight. Re-
port results to the nearest 0,5 mg.
  Container  No.   2.  Transfer  the  acetone
washings to  a tared beaker and evaporate to
dryness at ambient temperature  and pres-
sure. Desiccate and dry to a constant weight.
Report results to the nearest 0.5 mg.
  Container No. 3.  Weigh the spent silica gel
and report to the  nearest gram.
  5. Calibration.
  Use  methods  and equipment which have
been  approved  by  the  Administrator  to
calibrate  the orifice  meter,  pitot tube,  dry
gas  meter, and probe  heater. Recalibrate
after each test series.
  6. Calculations.
  6.1  Average dry gas meter temperature
and average orifice pressure  drop.  See data
sheet (Figure 5-2).
  6.2  Dry  gas 'volume.  Correct the sample
volume measured  by  the dry  gas  meter to
standard conditions (70° F., 29.92 inches Hg)
by  using Equation 5-1.


V    -V  /T....\(Fb"+ibV
 '"d    "VT,J\   P.ld    /
                        (0.0474 51^) V,.

                              equation 5-2

where :
  VwlU= Volume of water vapor in the gas
           sample   (standard   conditions) ,
           cu. ft.
    Vi0 = Total volume of liquid collected in
           impingers and silica gel (see Fig-
           ure 5-3 ) , ml.
    P«jO= Density of water, 1 g./rnl.
   MH,O= Molecular  weight of water, 18 lb./
           Ib.-mole.
      B = Ideal  gas   constant,  21.83  inches
           Hg — cu.  ft./lb.-mole-°R.
    T,ta= Absolute temperature  at standard
           conditions, 530° R.
    P,,4= Absolute pressure at standard con-
           ditions, 29.92 inches Hg.

  6.4  Moisture content.

                      V
                      '"id
                                    ,
                                     13-6
                              equation 5-1
where :
  Vm,td= Volume of gas sample through the
           dry  gas  meter  (standard  condi-
           tions) , cu. ft.
     Vra= Volume of gas sample through the
           dry  gas  meter  (meter  condi-
           tions) , cu. ft.
   T.,d= Absolute temperature at  standard
           conditions, 530* R.
                              equation 3-3
wheie'
  Bwo
      — Pioportkm by volume of watei vapor in the ^as
         stieam, dimensionless.
 ^"btd =Volume of water in the gas sample (stand-aid
         conditions) , cu. ft.
 ^"Vd = Volume of gas sample through the dry gas motcr
         (standard conditions) , cu. ft.

  6.6  Total particulate  weight. Determine
the total particulate catch from the sum of
the  weights  on  the analysis  data  sheet
(Figure 5-3) .
  6.6  Concentration.
  6.6. 1  Concentration in gr./s c.f .
                                                     c'.=  0.0154
                              equation 5-4
where:
    c'.= Concentration of particulate matter in stack
         gas, gr./s.c.f., dry  basis.
   M.=Total amount of particulate matter collected,
         mg.
 ^matd=Volume of gas sample through dry gas meter
         (standaid conditions), cu. ft,
                                 FEDERAL  REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER  23.  1971

-------
 24890
                                                    RULES  AND  REGULATIONS
                              PLANT.

                              DATE_
                              RUN NO.
CONTAINER
NUMBER
1
2
TOTAL
WEIGHT OF PARTICULATE COLLECTED,
mg
FINAL WEIGHT

:xi
TARE WEIGHT

X
WEIGHT GAIN




FINAL
INITIAL
LIQUID COLLECTED
TOTAL VOLUME COLLECTED
VOLUME Of LIQUID
WATER COLLECTED
IMPINGER
VOLUME.
ml




SILICA GEL
WEIGHT,
9



g* ml
  CONVERT WEIGHT OF WATER TO VOLUME BY DIVIDING TOTAL WEIGHT
  INCREASE  BY DENSITY OF WATER.  (1 g. ml):
                                                       =  VOLUME WATER, ml
                       Figure5-3.  Analytical data.

  6.6.2  Concentration In Ib./cu. ft.

                       c ^(453^00 rZj:)M"^OOA,win_A  M,,
where1
     cfl=Concentiatioii of pattieulate matter in stack
          gas, Ib./s.c.f., diy ba.sis.
  453,600=Mg/lb.
               ^"ttd          equation 5-5

    Mn = Tot,U amount of paiticulate matter collected,

  Vm,,j= Volume of gas sample through dry gas meter
          (standard conditions), cu. ft.
6 7  Isokinetic variation.
T 	
            iB2O
                                       -xioo
when1'
     I = Perccnt of isokinctic sampling.
    Vjc=Total volume of liquid collected lii Impingets
         and silica gel (Sec Fig. 5-3), nil.
   pH2o=Density of water, 1 g./ml.
     K=Ideal gas constant, 21.83 inches Hg-cu. ft./lb.
         moio-°R.
  MH,O =Molecular weight of water, 18 Ib /Ib.-mole.
    Vm = Volume of gas sample through the di y gas meter
         (metei conditions), eu. ft.
    Tm= Absolute average dry gas meter temperature
         (see Figiue6-2),°K.
   Fbar=Baiomctue pressuie at sampling site, Indies
         UK.
    AH=Aveiagc pressuie drop across  the orifice {see
         Fig. 5-2), inches H2O.
    T,=Absolute aveiagc stack gas temperature (see
         Fig. 5-2),°It.
     0=Total sampling time, min.
    V,=8tack gas velocity calculated by Method  2,
         Equation 2-2, ft /sec.
    l\=Absolute stack gas piossure, inches Hg.
    An = Cross-sectional area of nozzle, sq. ft.

   6.8  Acceptable   results.  The  following
range sets the limit on acceptable isokinetic
sampling results:

If 90%
-------
RULES AND REGULATIONS
                                                  24891
                     n 8« *~a  •* '• 3 s
                     o  ~S&* -S«~
                                                    o
        1127

-------
24892
                                         RULES AND REGULATIONS
nitrous oxide,  are  measure  eolorimetrically
using  the  phenoldlsulfonic  acid   (PDS)
procedure.
  1.2   Applicability. This method is applica-
ble for the measurement of  nitrogen oxides
from stationary sources  only when specified
by the test procedures for determining com-
pliance  with   New  Source  Performance
Standards.
  2. Apparatus.
  2.1   Sampling. See Figure 7-1.
  2.1.1  Probe—Pyrex1 glass,  heated, with
filter to remove particulate matter. Heating
is unnecessary if the probe remains dry dur-
ing the purging period.
  2.1.2  Collection  flask—Two-liter,  Pyrex,'
round  bottom  with  short neck and 24/40
standard  taper opening, protected  against
Implosion or breaKage.
  1 Trade name.
                                     2.1.3  Flask  valve—T-bore stopcock  con-
                                   nected  to a 24/40 standard taper  Joint.
                                     2.1.4  Temperature gauge—Dial-type ther-
                                   mometer, or equivalent, capable of measur-
                                   ing 2° F. intervals from 25' to 125' P.
                                     2.1.5  Vacuum  line—Tubing  capable  of
                                   withstanding a vacuum of 3 inches Hg abso-
                                   lute pressure, with "T" connection and T-bore
                                   stopcock, or equivalent.
                                     2.1 6  Pressure gauge—U-tube manometer,
                                   36  inches,  with   0.1-inch  divisions,  or
                                   equivalent.
                                     2 1.7  Pump—Capable of producing a vac-
                                   uum of 3 inches Hg absolute pressure.
                                     2.1.8   Squeeze bulb—Oneway.
                                     2.2  Sample recovery.
                                     2.2 1  Pipette or dropper.
                                     2.2.2  Glass storage  containers—Cushioned
                                   for shipping.
        PROBE
        A
      fILTER

  GROUND-GLASS SOCKET,
      g NO. M/S
                              f LASK SHIELD-, ,\
GROUNO-GLAS:
 STANDARD TAPER,

J SLEEVE NO, 24/40
                           GROUND-GLASS
                           SOCKET, § NO. 12,5
                           PYREX
                                                                   FOAM ENCASEMENT
                                                             BOILING FLASK -
                                                             2 LITER. ROUND-BOTTOM, SHOUT 1CCK.
                                                             WITH J SLEEVE NO. 24/40
                         Figure 7-1, Sampling uain, l)ask valve, and flask.
  2.2.3  Glass wash bottle.
  2.3  Analysis.
  2.3.1  Steam Datn.
  2.3.2  BeaKers or casseroles—250  ml., one
for each sample and standard  (blank).
  2.3.3  Volumetric pipettes—1, 2, and 10 ml.
  2.3.4  Transfer pipette—10 ml. with 0.1 ml.
divisions.
  2.3.5  Volumetric flask—100 ml., one for
each sample, and 1,000 ml. for the standard
(blank).
  2.3.6  Spectrophotometer—To measure ab-
Eorbance at 420 urn.
  2.3.7  Graduated cylinder—100  ml.  with
1.0ml. divisions.
  2.3.8  Analytical  balance—To measure to
0.1 mg.
  3. Reagents.
  3.1  Sampling.
  3.1.1  Absorbing solution—Add  2.8 ml. of
concentrated  H,SO, to  1 liter of  distilled
water. Mix well and add 6 ml.  of 3 percent
hydrogen peroxide. Prepare a fresh solution
weekly and do not expose to  extreme heat or
direct sunlight.
  3.2  Sample recovery.
  3.2.1  Sodium  hydroxide  (IN)—Dissolve
40 g. NaOH in distilled water and  dilute to  1
liter.
  3.2.2  Red litmus paper.
                                      3.2.3  Water-—Deionized, distilled.
                                      3.3  Analysis.
                                      3.3.1  Fuming sulfurlc acid—15 to 18% by
                                    weight free sulfur trioxide.
                                      3.3.2  Phenol—White  solid reagent grade.
                                      3.3.3  Sulfuric acid—Concentrated reagent
                                    grade.
                                      3.3.4  Standard solution—Dissolve 0.5495 g.
                                    potassium nitrate  (KNOS) in distilled water
                                    and dilute to 1 liter. For the working stand-
                                    ard solution, dilute  10  ml.  of the resulting
                                    solution to 100 ml. with distilled water. One
                                    ml. of  the working standard  solution  is
                                    equivalent to 25 /ig. nitrogen dioxide.
                                      3.3.5  Water—Deionized, distilled.
                                      3.3.6  Phenoldlsulfonlc   acid   solution—
                                    Dissolve 25  g. of pure white phenol in 150 ml.
                                    concentrated sulfurlc acid on a  steam bath.
                                    Cool, add 75 ml.  fuming sulfuric acid, and
                                    heat at  100° C. for 2 hours. Store in a dark,
                                    stoppered bottle.
                                      4. Procedure.
                                      4.1 Sampling.
                                      4.1.1  Pipette 25 ml. of absorbing solution
                                    Into a sample flask.  Insert the flask valve
                                    stopper  into the flask with the  valve in the
                                    "purge"  position. Assemble the  sampling
                                    train as shown. In Figure 7-1 and place the
                                    probe at the sampling point. Turn the flask
                                    valve and the pump valve to their "evacuate"
positions. Evacuate the flask  to at least 3
inches Hg absolute pressure. Turn the pump
valve to its "vent" position and turn off the
pump. Check the manometer for any fluctu-
ation in tne mercury level. If there is a visi-
ble change over the  span of one  minute,
check for leaks. Record the initial volume,
temperature, and barometric pressure. Turn
the flask valve to its  "purge"  position,  and
then  do the  same with  the  pump valve.
Purge the probe and the vacuum tube using
the squeeze bulb. If condensation occurs in
the probe and flask valve area, heat the probe
and purge-.until the condensation disappears.
Then turn the pump valve to Its "vent" posi-
tion.  Turn  the flask  valve to Its "sample"
position and allow sample to enter the flask
for about  15 seconds. After collecting the
sample, turn the  flask valve to  its "purge"
position  and disconnect the flask from the
sampling  train.   Shake  the  flask for  5
minutes.
  4 2  Sample recovery.
  4.2.1  Let the flask  set for a minimum of
16 hours and then shake the contents for 2
minutes. Connect  the flask to  a  mercury
filled U-tube  manometer,  open the valve
from the flask to the manometer, and record
the flask pressure  and temperature along
with  the barometric pressure. Transfer the
flask  contents  to  a container  for shipment
or to a 250 ml.  beaker for analysis. Rinse the
flask  with  two portions of distilled water
(approximately 10 ml.) and  add  rtnse water
to tne sample. For a blank use 25 ml. of ab-
sorbing solution and the same volume of dis-
tilled water as used in rinsing the flask. Prior
to shipping or  analysis, add sodium hydrox-
ide (IN) dropwlse into both the  sample and
the blank  until  alkaline to  litmus paper
(about 25 to 35 drops in each).
  4.3  Analysis.
  4.3 1  If  the sample has been shipped in
a container, transfer the contents to a 250
ml. beaker using a small amount of distilled
water. Evaporate the solution to dryness on a
steam bath  and then cool. Add 2  ml. phenol-
disulfonlc acid solution to the dried residue
and triturate thoroughly with a glass  rod.
Make sure the solution contacts  all the resi-
due. Add 1 ml. distilled water and four drops
of concentrated sulfuric acid. Heat the solu-
tion on a steam bath for 3 minutes  with oc-
casional  stirring. Cool, add  20 ml.  distilled
water, mix well by stirring, and add concen-
trated ammonium  hydroxide dropwise with
constant stirring  until  alkaline to litmus
paper. Transfer the  solution  to a 100 ml.
volumetric flask and wash the beaker three
times with  4 to  5 ml. portions  of  distilled
water. Dilute to the  mark  and mix thor-
oughly. If the sample  contains solids, trans-
fer a  portion of the solution to a clean, dry
centrifuge  tube, and  centrifuge, or  niter a
portion of the solution. Measure  the absorb-
auce of eacn sample  at 420 nm. using the
blank solution as a zero. Dilute  the sample
and the blank with a suitable  amount of
distilled water if absorbance falls outside the
range of calibration.
  5. Calibration.
  5.1  Flask volume. Assemble the flask  and
flask valve  and fill with water to the stop-
cock.  Measure the  volume of water to ±10
ml. Number and record the volume on the
flask.
  5.2  Spectrophotometer. Add 0.0 to 16.0 ml.
of standard solution to a series of beakers. To
each beaker add 25 ml. of absorbing solution
and add sodium  hydroxide (IN)  dropwlse
until  alkaline to litmus paper (about 25 to
35 drops).  Follow the analysis procedure of
section 4.3 to collect enough data to draw a
calibration curve of concentration In /«g. NO»
per sample versus absorbance.
  6. Calculations.
  6.1  Sample volume.
                                FEDERAL REGISTER, VOL. 36, NO. 247—THURSDAY, DECEMBER 23, 1971
                                                                      1128

-------
                                                  RULES AND  REGULATIONS
                                                                                                                            24893
v..=-
          P.ui

where:
   Vsc= Sample volume at standard condi-
          tions (dry basis), ml.
  T
-------
24894
RULES AND REGULATIONS

  51ili
  -r* -w ^  '

1 3!
< g
f i
3 °





*
?
WED MOIS
5
•"





O
J

I « B
1 I !
S o ^
<£ £ S



1

£
2
i
S
*

















I i





I







O ™
ill



g
m
S

SS SECTION
5
•t
S
i
K




' 6
! E

sj
C

^

i*

I
2 £
IS
Ii
i
i-
u.
t*.
3 5
^^
^~
-'i
III

Et*o =
"
Cj I ^
Ifi-

s
ill
I*i
r















































































































































































































































































I

i i











M
                                                                                        o
                                                                                        Ul
                                                                                        a
                                               1130

-------
                            RULES AND  REGULATIONS
                                                         24895
  Bom, Jerome J., Maintenance, Calibration,
and Operation  of  Isokinetic  Source Sam-
pling Equipment, Environmental Protection
Agency, Air Pollution Control Office Publi-
cation No. APTD-0576.
  Shell Development Co. Analytical Depart-
ment,  Determination of Sulfur Dioxide and
Sulfur Trioxide in Stack Gases, Emeryville
Method Series, 4516/59a.
METHOD  9	VISUAL  DETERMINATION  OF THE
  OPACITY  OF EMISSIONS  FROM  STATIONARY
  SOURCES
  1. Principle and  applicability.
  11  Principle.  The relative opacity of  an
emission from  a stationary  source is  de-
termined  visually by a qualified  observer.
  1.2  Applicability. This method  is  appli-
cable for  the determination of the relative
opacity of visible emissions from stationary
sources only when specified by test proce-
dures  for determining  compliance  with the
New Source Performance Standards.
  2. Procedure.
  2.1  The qualified observer stands at ap-
proximately two stack heights, but  not more
than a quarter  of a mile from the base of
the stack with, the sun to his back. From a
vantage point perpendicular  to the plume,
the observer  studies the point of greatest
opacity in the plume. The data required in
Figure 9-1 is recorded every 15 to 30 seconds
to the nearest 5 % opacity. A minimum of 25
readings is taken.
  3. Qualifications.
  3.1   To certify as an observer, a candidate
must  complete a smokereading course  con-
ducted  by  EPA, or equivalent;  in order to
certify  the  candidate must assign opacity
readings in 5% increments to  25 different
black plumes and 25 different white plumes,
with  an error  not to exceed 15  percent on
any one reading and an average error not to
exceed  7.5  percent  in  each category.  The
smoke  generator used  to  qualify  the  ob-
servers must be equipped with  a calibrated
smoke indicator or light transmission meter
located in  the source  stack  if  the smoke
generator is to determine the actual opacity
of the emissions. All qualified observers must
pass  this test  every  6  months in order to
remain certified.
  4. Calculations.
  4.1   Determine the average opacity.
  5, References.
  Air Pollution Control District Rules  and
Regulations, Los Angeles County Air Pollu-
tion Control District, Chapter 2,  Schedule 6,
Regulation 4, Prohibition, Rule 50,17 p.
  Kudluk, Rudolf, Ringelmann Smoke Chart,
TJ.S. Department of Interior, Bureau of Mines,
Information Circular No. 8333, May 1967.
                                              fil-rl, tnml.nn
                                              Optc'W ;
                                                    Sum of iws. racarcfod
                                                    Total no. reading*
                              Figure 9-1.  Field data,

                        [FR Doc.71  18624 Filed 12-22-71:8:45 am]
    U. S. GOVERNMENT PRINTING OFFICE: 1972	746461/4IO2
           FEDERAL REGISTER, VOL. 36, NO.  247—THURSDAY,  DECEMBER 23,  1971
                                           1131

-------

-------