-------

-------
    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
-------
-------